NT ae NAH CHAR eafie dh waeek®. Magia HOR chr th ge she pie PES ree ere ene Dees rae mar free ete Soheyeetes ses nF tae ope re kena area Veen at nat her aw ye FE Er teeeh HF HMDA BNE ery nenttert Bien hea Weta Beds Prpetubeg stan Vonagecto apd Fr Ye Pagaes fave red eet ea een cet Ba bo Rnd in cite weet ls ik tnt Nea tte Om Pg mies + ean topes Ee ties? Paha fue anaes bye tte en me pee ie eset ty ene roenpieh ane F 2 en Pej ee eagle he tre SRE 9 ONE Peveie se eer ran pet dete Ee Pape sobs pereps tt oa athe ean ota fet soya ee po SP ANTS we: Cree shy ye A Seren _ pe FE rSevhay % Pitre Nevertire: SPAM STEN TON: ee ee a Aetna Ata ae ree See reeves (fee nT D Grern ee yeide. east Deshp a Ap Swett nn h! EIS tae Peed de PS FL aha EN ot Su eer Per aans Shak. Jee nwa ies ute te TEM se ow oe Beg chet en eeatind EMER Oo TEEN EH tian aught Aig PORTA gas we sete. Semin Nn Gey ed ALAN ASIA ee er eee ert nananee ay nes FRSA AMAT TAR, Nan oer Seka Tatas ered St Moe S tare tte MAN ANN YBa erm er AAD Aryan eVeN EA, A See ent erae i FH MAR) nt 6 CURA GAN PARA RE as er eee a sO ed re BAAD DS \AWNDEDT AU SEN es UO yD ad a eM VET Vie Na A RAE REMS DS Pie got FE foie re te ee ete neta Sn Dae ees "A, aliowie teen voc fs Pe RMN Da asf TE aA analy Sire) eee area eat Cera WN ORV a a bate Ny RAS sett ete dN RE ee Pea eM ate Te Die Mea a Cac ynas SNAG A Re SeN we) 26D, Ratner tA hows a Beste: TCT Dea atone Bee rsdes onthe GAN CRE aea , ern Penne N= Wed oo TREN Maa Te Me mare oR GATENGS HER SM GR eT Re TL Age ROA wm hath HOY OM A pepeineone enna ir Sot mre. Ponies nets S = CAA SA me aan ORT ake VN Hae Pa caeeei Mt d SPA Thattaik LYRA Neonates Ne aber tantinhe€ ett Se VA eT PERC BLT NE ote hamtcatane EN em BONA AD Re DAM ee Seems es stamina Fa ANOER Dag Neate AN weaheeeshy & “ sak amet tite PUY apa caeew Fea 1 ON SOT RETR Oe eeaN Ws Vite ere Sey aun shat vanes os Fe Siero tet erate ad coy Day ee eyes ae Soe ate Ve scat peemerey Fas ere! sree Sane wether er a ee Sens = “hy poe ; ar Sao, ~. a joe — i J ay GF.¢ § — = = SN Na oO oO ie! = ries hay eae. 2 S Ss i 2 INOSHLINS S3IUVUE WON = = $ ~ LIBRARIES SMITHSONIAN INSTITUTION NOLLNIILSNI NV as 5 a 2 e NI” NVINOSHLIWS satuvagi7 LI z c S) oo NS 2 NIAN INSTITUTION NOLLALILSNI = rc S = < SNS OE HSO — z = NS fo) oO = a = > WW = _ Oo o \N SS ris 0 >. > (= % re 2 : & he z = as Fi = eae” a > ' = = = a = n m = un > | = = = yaks y = ON es SMITHSONIAN INSTITUT rm 2 a = ® NOILALILSNI NVINOSHLINS $31 S o ZX. D = ” SMITHSONIAN — oe z 2 = = z : oo a ames ¥ = = F 3 GY 6 d 2 | = > = 5 x a ik Bd Be =. ix = tt ff: = r= Ore w 5 oO Tp = z = : ZY = é 2 = 3 Kal a ae fue So Gs 2 > z | NVINOSH Pa “i, = — a . 2 SMITHSONIAN _ INSTITUTION ,NOLLALILSNI_ SI 0 2 es ILSNI_NVINOSHLINS S3IYVYSIT_ on = ws o — HSONIAN _INSTITUTION NOILNL = oe a - Ww oe = = = _ uw ‘ NN = =f e.Vvegey a <= (c=: « pe : 2gX | : SGiy = : : : . 7's, ma = SK. “4 —_ (oo) : = 4 = < Cc SS o = oO dg fo) -_ > = aay OA : é : BK 0 5 - @ LIWS=S31¥VUSIT LIBRARIES. SMITHSONIAN” INSTITUT f = = + VINOSH oe o g a ES SMITHSONIAN” INSTITUTION | NOLLNLUSNITN ul z & S} Yy, @ SE eee ee ee ene cS a 2 5 GLY, ® = a © ly OO = 2 =) > - Uf hx ) o Vy, ® =, > WE ip fer 2 = So, os AE > ~ x Nx WS i a Ls ‘ef A m Jo ae a =_ MQ (97) SS 2 = = 0 Ki OR a m WY g @ = IWS S3LUVUS : = » - = Ni NVINOSHL QS ANE OE = nn? m z n TITUTION NOIMLALILS w XS 2 a = WS, S31¥VEGIT_ LIBRARIES SMITHSONIAN INS go ae g “SP ov ~ NVINOSHLI ztlt = = LLMLILSNI_ NVIN | | 2 < = > = HSONIAN INSTITUTION NO z g < = 2 =I . 8 z Sills = 2 = 3 ob 2 ; 8 = = | L Cues Gr é a wn ey ~ SA = oa s > YO 3 co) 3 ~ 2 = 2 Hace S - = Ss * Ww = a au: 2 IBRARIES. SMITHSONIAN INSTITUT sus So a 2 © NOLINLILSNI- NVINOSHIIWS S31YVY 8 ITAL P Z ee, “LIBRARIES SMITHSONIAN INSTITUTION | NOIL NY “ 2 us = OSHLINS = ” z uw & s = = We i a a - Yhy, : = SY: = : E WA = : OG: : = W's : a i . - Lf pad eS oo WT sc S 2 * |ES SMITHSONIAN INSTITUTION NOILAL ae = ae = NI NVINOSHLINS S31uvugI7. LIBRAR = S) w WS as HSONIAN INSTITUTION NOILNLILSNI | . - g e 2 GQ = : é eS : ES a a= = o Qs = 6 fiw ~s = YA y OO XY, Ww - _— Paes, — me 2 WY E MQ 2 s 2° = : > = > S > = 5 = “” * 2 n ‘ s a) Fs MITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IMVYIT_LIBRARIES SMITHSONIAN z o > we = ” : <” aan s CN wW uw ; ; “S we ZW = = =. oof Lp 5 < = \ = ‘ P dip ~ = a S iN ce S = UyZs s Zz = | ra * ag _ “es = VINOSHLINS _S31¥VYEI7 LIBRARIES” SMITHSONIAN” INSTITUTION — NOILNLILSNI_NVINOSHLINS » oO ae, (@) — oO a Ss c b = o = t, ow = o = \ = ee) 2 Vifly, >) = 0 \ XE SW 7 > - LL. = > we = AN 5 2 = Of? - 2 OWN & Wr 2 - gr % m YW g _. mw i (ap) < w oO 2 MITHSONIAN INSTITUTION NOILNLILSNI _NVINOSHLIWS, $3 luvugi7_ LIBRARIES SMITHSONIAN z ss = 4; < = Ws z . = 2 = * a it fy, = 4 Yip = = 2 8 NN 7723 3 Ys NX? : rT MIN G6 ts ff, TC fe) Yt LZ xX SSS Oo x E QO 2. 7RE 27S P* = NS 2 : Fe z 2 ze 3 Lt pr 2iYVUsiT_LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI_ NVINOSHLIWS uw F ts Zz us Z u = e pre 77) = ay = < 5 e = - = 3 . hg a ‘s Ws = <= \ = << = a a = S 5 SGU ERR 2 2 2 2 é 2 “Hy = WW 2 = Z = 2 Be ue _ 2 _ 2 - : MITHSONIAN INSTITUTION NOILOLULSNI_NVINOSHLIWS Sa luvugiy SMITHSONIAN wf fg 3 . 4 XE — Z = Lif fg 3 Lip at 5 2 YX 5 - > = Yi > > ES = - a Ze fa Ee YW E F a aS fer 2 = 2 NS i MITHSONIAN INSTITUTION NOLLALILSNI_ NVINOSHLUIWS, $3 ryyUgiy as BRARI ES SMITHSONIAN _ < ‘ = , < = yy =< * = & 3 1 §%,% 2 YG? z Z = 2,7 fF" 27S f° OR Zz | 2 z : : a oe /INOSHLINS _LIBRARI ES SMITHSONIAN _INSTITUTION pNOULALILSNI_ . os 7 uw = us Fe ve] fad es ~~ a ie = te _< ar os aoe oe a a co c ow Cc o c x be 7 wy 3 : 7 : 1 - : uty "f A 7 ; ‘ ne - - ve 1 > " (aye nT ; ; : es Yeon - ae ; q : on rT i is rn iT Ae : ; ms : * - = rm a 7 ' * 7 _ ' f ve F ‘ : ‘ 1 mal : : : : * 4 - - ‘x oo ern ' ah > Y : a 7 a oe Le ‘ a : vie) : 3 : an) : » e ve a oo” as — : vey wer? : ‘ Aw) one aft ‘ ‘ : 7 Mu ea en a i kul ZOOLOGICAL SCIENCE An International Journal VOLUME 5 1988 published by The Zoological Society of Japan os Ae CONTENTS VOLUME 5 REVIEWS de Pomerai, D.I.: The transdifferentiation of neural retina into lens in vitro............ 1 Tsuneki, K.: The neurohypophysis of cyclo- stomes as a primitive hypothalamic center of Vertebrates. aioe saucreshonainn.cene ate gaaen 21 Meusy, J.-J. and G. G. Payen: Female repro- duction in malacostracan Crustacea ...... 217 Nishioka, R. S., K. M. Kelley and H. A. Bern: Control of prolactin and growth hormone secretion in teleost fishes................. 267 Gause, G. G.: Taxon-specific crystallins.... 727 Kuroda, H., S. Obata, K. Takemoto, M. Ishi- guro and H. Sato: The mechanism and physiological function of electrical changes during fertilization of sea urchin gametes 733 De Santis, R. and M. R. Pinto: The pathway of sperm-egg interaction in ascidians: biology and! CHEMISULY & 6.02 ea Sete owes ees 919 Urano. A.: Neuroendocrine control of anur- an anterior preoptic neurons and initiation of mating behavior.................0.0.0e eee 925 Gremigni, V.: Planarian regeneration: An overview of some cellular mechanisms ... 1153 Mohri, H. and N. Hosoya: Two decades since the naming of tubulin — The multi-facets of CHUUNIN eee cece ae eden Son sees 1165 Special Issue on Advances in Cell Division Research Sakai, H.: General introduction to the special issue on Advances in Cell Division Research SVN ets ne cee Oe ches ne earn oee as 505 Dan, K.: Mechanism of equal cleavage of sea urchin egg: transposition from astral mechan- ism to constricting mechanism ........... 507 Mazia,D.: Mitotic poles in artificial par- thenogenesis: a letter to Katsuma Dan... 519 Inoué, S.: The living spindle .............. 529 Nakano, Y. and Y. Hiramoto: Measurement of spindle birefringence by the optical in- tegration method.....................005. 539 Hamaguchi, Y.: Jn vivo cytochemistry in cell GIVISION eee eae nen sacs ee ce aay ea 545 Yoneda, M.: Computed profiles of compress- ed sea-urchin eggs with elastic membranes 553 Yamao, W. and T. Miki-Noumura: Effect of hexyleneglycol on meiotic division of starfish OOCVIES et ccnst sees ara nus ae i pea es 563 Longo, F., W. H. Clark, Jr. and G. W. Hinsch: Gamete interactions and sperm incorporation in the nemertean, Cerebratulus lacteus .... 573 Schattern, H., C. Howard, G. Coffe, C. Simer- ly and G.Schattern: Centrosomes, cen- trioles and post-translationally modified mic- rotubules during fertilization ............. 585 Palazzo, R.E., J.B. Brawley and L.I. Rebhun: Spontaneous aster formation in cytoplasmic extracts from eggs of the surfclam DeForest auc ttece Oe easiness aera eee 603 Ohta, K., M. Toriyama, S. Endo and H. Sakai: Mitotic apparatus-associated 51-kD protein INIMILOSIS aera ae end eect sree eres 613 Sato, H. and J. Bryan: The thermodymanics of molecular association in the mitotic spindle with or without heavy water (DO) ...... 623 Harris, P. J.: Metaphase to anaphase transi- tion of sea urchin eggs examined in caffeinein- duced monasters ..............0.02 cece eee 639 Kojima, M. K.: Marked elongation of the anaphase spindle by treatments with local anesthetics in sea urchin eggs ............ 645 Sluder, G.: Control mechanisms of mitosis: The role of spindle microtubules in the timing of mitotic events ..............0. 02. e eee ee 653 Sawada, T.: The mechanism of ooplasmic segregation in the ascidian egg........... 667 Kawamura, K.: The contraction wave in the cortex of dividing neuroblasts of the grass- DOPPCln dese sas eee ees aes oe ere 677 Sawai, T.: Participation of the subcortical and interior cytoplasm in cleavage division of MEWUWERLS frase ces sae ne detent oe me 685 Ohnuma, M. and I. Mabuchi: Partial purifica- il tion and characterization of a factor which dissociates 45K protein-actin complex from Sea Urchin: 80.4.2 cineca: eee aden 691 Bonder, E. M., D. J. Fishkind, J. H. Henson, N.M.Cotran and D.A. Begg: Actin in cytokinesis: Formation of the contractile APPATACUS. .daiscoddians doa caste scutes 699 Schroeder, T. E. and J.J. Otto: Im- munofluorescent analysis of actin and myosin in isolated contractile rings of sea urchin eggs Re en iptale wr snie aL catenin eet: 713 ORIGINAL PAPERS Physiology Azuma, K.: Hypersensitivity after offset of adapting light in vertebrate photoreceptors Takei, Y., J. Okubo and K. Yamaguchi: Effects of cellular dehydration on drinking and plasma angiotensin II level in the eel, Anguilla Japonica ...... 6... c cece 43 Ozaki, M.: A possible sugar receptor protein found in the labellum of the blowfly, Phormia FE QUIVG charac ais vice whe sie cto Paehors cath 9 hey Sa aoe eho 281 Okano, Y., E. David, K. Honda and S. Inoué: Auditory evoked potentials dynamically re- lated to sleep-waking states in unrestrained LatS ome iacts ais cevac ete sevemtigarar etter treater 291 Tazaki, K.: The anatomy and physiology of the stomatogastric nervous system of Squilla. II. The cardiac system ................... 299 Obika, M.: Ultrastructure and physiological response of leucophores of the medaka Oryzias latipes: 2 octane we antcnngonen 311 Lindstrom, M., H. Nilson and V.B. Meyer- Rochow: Recovery from light-induced sen- sitivity loss in the eye of the crustacean Mysis relicta in relation to tempertature: a study of ERG-determined V/log I relationships and morphology at 4°C and 14°C.............. 743 Naitoh, T., K. Takeuchi and I. Takabatake: Mode of melanosome migration in teleostean melanophores ...........-..0. sees eee eee es 759 Yasuyama, K., T. Kimura and T. Yamaguchi: Musculature and innervation of the internal reproductive organs in the male cricket, with special reference to the projection of un- paired median neurons of the terminal abdo- minal ganglion: . 3. 20c.i0o3 concen ee 767 Khin Maung Saing: Functional innervation of the intrinsic thumb muscles of the fruit bat Pteropus medius.........0. 0.0. c ccc ees 781 Inoda, T., H. Ohtake and M. Morisawa: Activation of respiration and initiation of motility in rainbow trout spermatozoa ... 939 Hidaka, T. and S. Yukiyama: Excitatory and inhibitory junction potentials recorded from the red muscle of marine teleost, puffer fish deaths a dvctiaNaad ld eae Bee ee a 947 Negishi, S.:_ The involvement of microtubules in the light response of medaka melanophores suikGisiates GAUMIAAE oA ASST SUR es 951 Grundstrom, N., H. Sundgren, J.-O. G. Karl- sson, and H. Elwing: A simple and efficient method for photometric estimation of the state of pigment aggregation in fish mela- NOphores, 2. asec bate eee a ee 959 Endo, Y.: Non-synaptic release of transmit- ter-containing vesicles from the enteric neurons of the rat small intestine ........ 965 Nishi, T., M. Kobayashi, M. Isomura, H. Ishi- da and Y. Shigenaka: Direct evidence for axopodial fusion preceding cell-to-cell con- tact in a heliozoan Echinosphaerium (COM- MUNICATION) (ce. ee 179 Srivastav, A. K. and L. Rani: Phosphocalicic response to vitamin D; treatment in freshwa- ter snake, Natrix piscator (COMMUNICA- ‘HION) Ganesan: tote iae eee 893 Cell Biology Zama, N. and H. Katow: A method of quan- titative analysis of cell migration using a computerized time-lapse videomicroscopy 53 Iwasaki, S. K.Miyata and K. Kobayashi: Fine structure of the filiform papillar epithe- lium from the tongue of the frog, Rananigro- VAGCULAID 5 3.5 anak AS adn Dae ee LER 61 Suganuma, Y. and H. Yamamoto: Conjuga- tion in Tetrahymena: Its relation to concana- valin A receptor distribution on the cell surface Iwasaki, S. and K. Kobayashi: Fine structure of the dorsal tongue surface in the Japanese toad, Bufo japonicus (Anura, Bufonidae) 331 Okamoto, M.: Fine structure of the iris mus- cle in the Japanese common newt, Cynops pyrrhogaster, with special reference to in- NETVALON Sasa es cen ck ones cata oes 337 Fujishima, M. and K.Hoshide: Light and electron microscopic observations of Holos- pora obtusa: A macronucleus-specific bacter- ium of the ciliate Paramecium caudatum. 791 Ishida, H., Y. Shigenaka and M. Imada: _ Fib- rillar system and possible control mechanism for the cycle of contraction and elongation of Spirostomum ambiguum .............065. 973 Ebitani, N. and T. Kubo: An established marine fish cell line with high plating efficien- cy (COMMUNICATION) ............... 183 Genetics Saitoh, M. and Y. Obara: Meiotic studies of interracial hybrids from the wild population of the large Japanese field mouse, Apodemus SPECIOSUS SPECCIOSUS 1.2.66. eee 815 Nakamura, T.: Female heterogametic sex- determination in Xenopus laevis as recon- firmed by repeated diploid gynogenesis (COMMUNICATION) ...............-5. 187 Naruse, K., A.Shimada and A. Shima: Gene-centromere mapping for 5 visible mutant loci in multiple recessive tester stock of the medaka (Oryzias latipes) (COM- MUNICATION) @ ie ek scat eke es 489 Shimada, A., A. ShimaandN. Egami: Estab- lishment of multiple recessive tester stock in the fish Oryzias latipes (COMMUNICA- TON) Bee secre as hah ene Somes gee 897 Ota, H., T. Hikida, M. Matsui and M. Hasega- wa: Karyotype of a scincid lizard, Carlia fusca, from Guam, the Mariana Islands (COMMUNICATION) osu o5 seu es ees 901 Immunology Nagata, S.: T cell-specific antigen in Xenopus identified with a mouse monoclonal antibody: Biochemical characterization and species dis- tHIDUION os cee 5 castes uaeebis ed neg sanss i Nagata, S.: T cell-specific XTLA-1 antigens from Xenopus laevis tadpole and froglet are not identical (CAMMUNICATION)..... 493 Biochemistry lil Suzuki, T., R. Muramatsu, T. Kisamori and T. Furukohri: Myoglobin of the shark Galeus nipponensis: Identification of the exceptional amino acid replacement at the distal (E7) position and autoxidation of its oxyform. 69 Tsuneoka, M., K. Maruyama and K. Ohashi: In vitro dimerization of I-protein, an A-I junctional component of skeletal muscle MYONDIUIS weteis cs st oon eeeeeecen een. 347 Kawamura,S. and M. Murakami: Lightin- duced Michaelis constant increase is rapid and inherent in cGMP phosphodiesterase in frog rod outer segments.................. 801 Hung, F. and Y. Shaoyi: Isolation and iden- tification of crucian (Carassius auratus L.) hemoglobin and its subunits.............. 809 Developmental Biology Fujisawa, H. and S. Amemiya: Temper- aturedependence in reaggregation of cells dissociated from sea urchin embryos with different seasonal growth ................ 85 Mitsunaga, K., Y.Fujino and I. Yasumasu: Probable participation of mitochondrial Ca** transport in calcification of spicules and morphogenesis in sea urchin embryos.... 93 Numakunai, T., Z. Hoshino and S. Kajiwara: Spawning of three intraspecific groups of the ascidian, Halocynthia roretzi (Drasche), in the wild, and fertilization among them... 103 Tahara, U.: the lamprey, Lampetra reissneri (Dybowski) CE Tar Ree ea gree arirey ae Sera eee 109 Tsunemoto, M., O. Numata, T. Sugai and Y. Watanabe: Analysis of oral replacement by scanning electron microscopy and im- Normal stages of development in munofluorescence microscopy in Tetrahyme- na thermophila during conjugation ....... 119 Iwamatsu, T., T. Ohta, E. Oshima and N. Sakai: Oogenesis in the medaka Oryzias latipes — stages of oocyte development .. 353 Tsuchiyama-Omura, S., B. Sakaguchi, K. Koga and D. F. Poulson: Morphological fe- atures of embryogenesis in Drosophila mela- nogaster infected with a male-killing spiro- plasmas ees osc dcr te cteee nthe ee ee eee sien 375 Suematsu, N., H. Takeda and T. Mizuno: Glandular epithelium induced from urinary iv bladder epithelium of the adult rat does not show full prostatic cytodifferentiation .... 385 Uchiyama, H. and T. Mizuno: Sexual dimorphism in the genital tubercle of the duck: Studies on the normal development and IStOSEMESIS ions i ..cy sade eee s peewee 823 Yamamoto, M., M. Ishine and M. Yoshida: Gonadal maturation independent of photic conditions in laboratory-reared sea urchins, Pseudocentrorus depressus and Hemicentro- tus pulcherrimuS .......0. 000 cece eee 979 Yamamoto, M.: Normal embryonic stages of the pygmy cuttlefish, /diosepius pygmaeus paradoxus Ortmann...................00- 989 Mizuno, T., H. Takeda, N. Suematsu, N. Hiro- naka and I. Lasnitzki: Absence of androgen receptors in the prostatic glandular epithe- lium derived from testicular feminization mutant (7fm) mice.................. eee 999 Yasumasu,S., I.Iuchi and K. Yamagami: Medaka hatching enzyme consists of two kinds of proteases which act cooperatively (COMMUNICATION) .................. 191 Sivasubramanian, P.: Interspecific trans- plantation of developing tissues and their subsequent differentiation in flies (COM- MUNICATION )............. eee eee eres 497 Reproductive Biology Awaji, M. and J. Hanyu: Effects of water temperature and photoperiod on the begin- ning of spawning season in the orange-red typesmedak as acacucttere senile aeie eiotyetctorste 1059 Kosaka, T., M. Obata, T. R. Saito and K. W. Takahashi: Effects of adult male cohabita- tion on precocious puberty in early weaning female guinea pigs (COMMUNICATION) Endocrinology Engstr6m, W., E. Dafgard and S. Falkmer: Comparative effects in vitro of Myxine, Squalus, avian and mammalian insuiins on DNA-synthesis in 3T3 mouse fibroblasts . 133 Ueda, H., T. Kosaka and K. W. Takahashi: Effects of long-term progesterone treatment on synchronized ovulation in guinea pigs 139 Endo, K., T. Masaki and K. Kumagai: Neuroendocrine regulation of the develop- ment of seasonal morphs in the Asian comma butterfly, Polygonia c-aureum L.: Difference in activity of summer-morph-producing hor- mone from brain-extracts of the long-day and short-day pupae... io262..cas5 ses snlqeaeias 145 Hyodo, S., M. Fujiwara, S. Kozono, M. Sato and A. Urano: Development of an in situ hybridization method for neurohypophysial hormone mRNAs using synthetic oligonuc- leotide: probes'.).......4. s24..i eee eee 397 Seki, T., S. Kikuyama and M. Suzuki: Effect of hypothalamic extract on the prolactin release from the bullfrog pituitary gland with special reference to thyrotropin-releasing hormone (TRH) & ..c.3005030. 26742 ee 407 Hirohama, T., H. Uemura, S. Nakamura and T. Aoto: Atrial natriuretic peptide (ANP)- immunoreactivity and ultrastructures of car- diocytes in fish «. «2.0.0. «.sieqseee eeeee 833 Tanaka, S., H. Iwasawa and K. Wakabayashi: Plasma levels of androgens in growing frogs of Rana nigromaculata ....... 06.60 c cece 1007 Oota; Y. and I. Koshimizu: Vascular supply of hypophysis in the turtle, Geoclemys PECVES Ib... ova ciinls o Heals dens SAGER eee 1013 Yamashita, T., K. Kawamoto, and S. Kawashi- ma: Fetal and postnatal development of arginine vasopressin-immunoreactive neurons in the mouse ..................5. 1019 Hyodo, S., M. Fujiwara, M. Sato and A. Ura- no: Molecular and immuno-histochemical study on expressions of vasopressin and oxytocin genes following sodium loading 1033 Nishida, M., J. Kawada, H. Ishizuka and S. Katsura: Goitrogenic action of manganese on female mouse thyroid through three PENETAtIONS c26 how cee des wena eee 1043 Masaki, T., K. Endo and K. Kumagai: Neuroendocrine regulation of the develop- ment of seasonal morphs in the Asian comma butterfly, Polygonia c-aureum L.: Is the factor producing summer morphs (SMPH) identical to the small prothoracicotropic hor- mone (4K-PTTH)?................ 00 cee 1051 Srivastav, A. K. and S. P. Srivastav: Corpus- cles of Stannius of Clarias batrachus in response to 1,25 dihydroxyvitamin D3 admi- nistration (COMMUNICATION )........ 197 Fujimori, M., Y.Sasayama and C. Oguro: Translocation of “Ca from the endolympha- tic sacs to the bone in Rana nigromaculata (COMMUNICATION) ...............-5. 201 Endo, Y. and T. Endo: S-100 protein-like im- munoreactive cells in the brain-midgut endoc- rine system of the insect Periplaneta america- na (COMMUNICATION)............... 905 Kobayashi, Y. and S. Kawashima: Changes of acetylcholinesterase activity in rat supraop- tic nucleus cell bodies during water depriva- tion (COMMUNICATION) ............. 911 Morphology Fujikara, K., S. Kurabuchi, M. Tabuchi and S. Inoue: Morphology and distribution of the skin glands in Xenopus laevis and their response to experimental stimulations.... 415 Tsuneki, K. and M. Ouji: Absence of blood vessels in the brain of six species of primitive salamanders ......cosjed0. 108 boos he deeda es 847 Chiba, A. and Y. Honma: agranular cells in the gummy shark (Mustelus manazo) Adenohypophysis............... 1065 Amasaki, H., M. Daigo and N. Meguro: Morphological observations of the large in- testine in the common vole, Microtus arvalis Fine structure of Pallas (COMMUNICATION) ........... 205 Behavior Biology Pandey, S.C. and S.D. Pandey: Sexual maturation in female wild mice: Combined effect of adults’ urinary chemosignals and minimum time of exposure to stimulus subst- ances for bringing the effects............. 153 Verrell, P. A.: Sexual interference in the alpine newt, Triturus alpestris (Amphibia, Urodela, Salamandridae) ................ 159 Ooka-Souda, S., H. Kabasawa and S. Kinoshi- ta: Circadian rhythms in locomotor activity of the hagfish, Eptatretus burgeri. II. The effect of brain ablation................... 431 Ooka-Souda, S. and H. Kabasawa: Circadian rhythms in locomotor activity of the hagfish, Eptatretus burgeri. V1. Hypothalamus: a locus of the circadian pacemaker?........ 437 Weldon, P.J.: Feeding responses of Pacific Vv snappers (genus Lutjanus) to the yellowbel- lied sea snake (Pelamis platurus)......... 443 Daumae, M. and T. Kimura: Factors regulat- ing urination patterns in male and female mice (Mus musculus)..........0.00000 0005 855 Ebino, K. Y., K. Yoshinaga, T. R. Saito and K. W. Takahashi: Coprophagy as an innate behavior in the mouse ................... 863 Kohda, Y. and M. Watanabe: Preference of striped backgrounds by striped fishes (COM- MUNICATION )..............0.000 0 eee 501 Saito, T. R., K. kamata, M. Nakamura and M. Inaba: Maternal behavior in virgin female rats following removal of the vomeronasal organ (COMMUNICATION )............ 1141 Ecology Hanzawa, N., N. Taniguchi and K. Numachi: Geographical differentiation in populations of Japanese dace Tribolodon hakonensis deduced from allozymic variation ........ 449 Konishi, K. and R. Quintana: The larval stages of three pagurid crabs (Crustacea: Anomura: Paguridae) from Hokkaido, Japan Ohgushi, R., S. Yamane and S. F. Sakagami: Ecological distribution and _havitat-linked density of colonies of stenogastrine wasps in tropical S. E. Asia ............ 0... e eee eee 869 Takahashi, H. and H. Iwasawa: Interpopula- tion variations in clutch size and egg size in the Japanese salamander, Hynobius nigrescens 1073 Meserve,L. A. and M.A.R. Gonzalez: Thyroid status and ambient temperature as influences on weaning in young mice..... 1083 Matsumoto, T.: Colony composition of the wood-feeding cockroach, Panesthia australis Brunner (Blattaria, Blaberidae, Panes- thiinae) in Australia (COMMUNICATION) Blas Soyeuse aera aee nan tiakroregoi at dpe hae 1145 Egami, N., O. Terao and Y. Iwao: The life span of wild populations of the fish Oryzias latipes under natural conditions (COM- MUNICATION) sesccéin. canons dedeeee ce 1149 Asada, N.: Invation of Drosophila albomi- cans to the mainland of Japan (COM- MUNICATION)...........2..: eee eee eee 915 vil Taxonomy Nakasone, Y.: Land hermit crabs from the Ryukyus, Japan, with a description of a new species from the Philippines (Crustacea, Decapoda, Coenobitidae) ................ 165 Sawada,I. and A.L.Molan: Two new hymenolepidid cestodes, Vampirolepis mola- ni sp. n. and V. iraqensis sp. n., from Iraqi Dats eter. coe co puarenoacueis titesoers eters 483 Kristensen, R. M. and Y.Shirayama: Plici- loricus hadalis (Pliciloricidae), a new lori- ciferan species collected from the Izu- Ogasawara Trench, Western Pacific...... 875 Uchikawa, K., K. Nakata and F. S. Lukoschus: Mites of the genus Myobia (Trombidiformes, Myobiidae) parasitic on Apodemus mice in Korea and Japan, with reference to their immature stages................. eee eee 883 Hirayama, A.: A ghost shrimp with four- articulate fifth pereopods (Crustacea: Caprel- lidea: Phtisicidae) from northwest Australia By arora Siiyscleals-ddienenece edt ns heattpsioeias, eee a Eee 1089 Xia, Z. W. and M. J. Toda: The Drosophila immigrans species-group of the subgenus Drosophila (Diptera: Drosophilidae) in Yun- man, ‘Chima so ac dheydestssyoec ccsstuelaste een eee 1095 Nakasone, Y.: Larval stages of Coenobita purpureus Stimpson and C. cavipes Stimpson reared in the laboratory and survival rates and growth factors of three land hermit crab larvae (Crustacea: Anomura) ............ 1105 Hayashi, T. and M. Matsui: Biochemical dif- ferentiation in Japanese newts, genus Cynops (Salamandridae):...............00. eee 1121 Others Proceedings of the 59th Annual Meeting of the Zoological Society of Japan.............. 1189 Book TevieWS¥rs..7 i: ls see eon eee 1340 ANNOUNCEMENTS .. «+. «oe 6. efe 8h ety tafe get ae 1342 Author index’, . 3 s.«.svernsc< cess eyergncues scene 1345 Instructions to Authors................-.05- 209 Erratum aes: cose ce sod specs specs. exe texte ae eae ee eee 212 aL * 7OOLOGICAL ~ SCIENCE me eee ‘ a 5c. et * An ‘International Journal ZOOLOGICAL SCIENCE The Official Journal of the Zoological Society of Japan Editor-in-Chief: The Zoological Society of Japan: Hideshi Kobayashi (Tokyo) Toshin-building, Hongo 2-27-2, Bunkyo-ku, Managing Editor: Tokyo 113, Japan. Tel. (03) 814-5675 Seiichiro Kawashima _ (Hiroshima) Officers: Assistant Editors: ; President: Nobuo Egami (Tsukuba) Takeo Machida : (Hiroshima) Secretary: Hideo Namiki (Tokyo) Sumio Takahashi (Hiroshima) Treasurer: Tadakazu Ohoka (Tokyo) Kazuyoshi Tsutsui (Hiroshima) Librarian: Shun-Ichi Uéno (Tokyo) Editorial Board: Howard A. Bern (Berkeley) Walter Bock (New York) Aubrey Gorbman (Seattle) Horst Grunz (Essen) Robert B. Hill (Kingston) Yukio Hiramoto (Chiba) Susumu Ishii (Tokyo) Yukiaki Kuroda (Mishima) Koscak Maruyama (Chiba) Roger Milkman (Iowa City) Hiromichi Morita (Fukuoka) Kazuo Moriwaki (Mishima) Tokindo S. Okada (Okazaki) | Andreas Oksche (Giessen) Hidemi Sato (Nagoya) Hiroshi Watanabe (Shimoda) | Mayumi Yamada (Sapporo) Ryuzo Yanagimachi (Honolulu) ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and communications in the broad field of Zoology. The journal was founded in 1984 as a result of unification of Zoological Magazine (1888-1983) and Annotationes Zoologicae Japonenses (1897-1983), the former official journals of the Zoological Society of Japan. ZOOLOGICAL SCIENCE appears bimonthly. An annual volume consists of six numbers more than 1000 pages including an issue containing abstracts of papers presented at the annual meeting of the Zoological Society of Japan. MANUSCRIPTS OFFERED FOR CONSIDERATION AND CORRESPONDENCE CONCERN- ING EDITORIAL MATTERS should be sent to: Dr. Seiichiro KAWASHIMA, Managing Editor, Zoological Science, Zoological Institute, Faculty of Science, Hiroshima University, 1-1-89 Higashisenda-machi, Naka-ku, Hiroshima 730, Japan, in accordance with the instructions to authors which appear in the first issue of each volume. Copies of instructions to authors will be sent upon request. SUBSCRIPTIONS. ZOOLOGICAL SCIENCE is distributed free of charge to the members, both domestic and foreign, of the Zoological Society of Japan. To non-member subscribers within Japan, it is distributed by Business Center for Academic Societies Japan, 6-16—3 Hongo, Bunkyo-ku, Tokyo 113. Subscriptions outside Japan should be ordered from the sole agent, VSP, Utrechtseweg 62, 3704 HE Zeist (postal address: P.O. Box 346, 3700 AH Zeist), The Netherlands. Subscription rates will be provided on request to these agents. New subscriptions and renewals begin with the first issue of the current volume. All rights reserved. No part of this publication may be reproduced or stored in a retrieval system in any form or by any means, without permission in writing from the copyright holder. © Copyright 1988, The Zoological Society of Japan Publication of Zoological Science has been supported in part by a Grant-in-Aid for Scientific Publication from the Ministry of Education, Science and Culture, Japan. ZOOLOGICAL SCIENCE 5: 1-19 (1988) © 1988 Zoological Society of Japan REVIEW The Transdifferentiation of Neural Retina into Lens in vitro oS Davip I. DE POMERAI \ Department of Zoology, University of Nottingham, Nottinghara,, j NG7 2RD, United Kingdom “Stig I. INTRODUCTION; PRECEDENTS During long-term monolayer culture, chick embryo neuroretinal (NR) cells lose most of their differentiated characteristics and convert exten- sively into both melanised pigment cells [1] and crystallin-containing lens-fibre-like cells (lentoids; [2]). These changes are best described by the term ‘transdifferentiation’ [3,4], because the NR cells giving rise to these foreign phenotypes remain incompletely characterised. But clearly no retinal cell would ever give rise to lens or pigmented progeny during the course of normal development in situ. The confusingly similar terms ‘metaplasia’ and ‘cell-type conversion’ are best reserved for those systems where the initial cell type is both fully differentiated and postmitotic. The best example fulfilling these criteria is Wolffian lens regeneration from dorsal iris follow- ing lentectomy in adult newts (reviewed in [5]). Lens removal, coupled with a stimulating influence emanating from the neural retina (possibly a growth factor; [6]), causes the marginal cells of the iris to re-enter mitosis and become depigmented. Those dorsal iris cells which divide fastest (cell cycle time<48hr) later convert into crystallin- expressing lens cells, whereas those which divide more slowly (cell cycle time >72 hr) later withdraw from mitosis, become repigmented and so revert to an iris phenotype. This decision is not irrevocable, however, since serial lentectomy (i.e. removing an earlier lens regenerate) will cause some dorsal iris Received October 5, 1987 cells to convert into lens even though they had previously decided not to do so [7]. Yamada and McDevitt [8] suggest that those cells which convert into lens must pass through a critical number (> six) of cell divisions within the 20-day prolifera- tive period following lentectomy, whereas cells passing through fewer than four mitoses (including ventral iris cells) would instead redifferentiate as pigment cells. Increased synthesis of extracellular matrix components is associated with the depig- mentation/proliferation and repigmentation phases of this process in vivo, but not with lens formation [9]. Cultures prepared from dispersed iris cells can also give rise to lens tissue in vitro, but this process neither requires NR influence, nor is it restricted only to dorsal iris (cells from ventral iris are equally lentoidogenic in culture; [10]). This relaxation may arise because iris cells of either type can pass through an unrestricted number of mitoses in vitro, although they do so on average more slowly than during lens regeneration in vivo. Similarities between chick NR transdifferentia- tion and Wolffian lens regeneration from newt iris include both the endpoint attained (lens-fibre-like cells) and the neural (optic cup) origin of both starting tissues. Differences include the embryonic nature and heterogeneity of cell types in the NR system, in contrast to adult iris tissue composed of postmitotic and fully differentiated melanocytes. An intermediate case is provided by the transdif- ferentiation of chick embryo tapetal cells into lens in vitro, where the initial tapetum consists solely of melanin-containing pigmented epithelial (PE) cells. As shown by Eguchi and Okada [11] using 2 D. I. DE PoMERAI clonal culture techniques, a single pigmented cell can give rise to crystallin-containing lentoids among its progeny, demonstrating a definitive switch of cell type in vitro. Lower vertebrates show extensive powers of regeneration, and transdifferentiations between adult cell types have been demonstrated e.g. in Hydra [12] and in various marine medusae [13,14]. However, the differentiated states of most verte- brate cell types seem much more stable, and few instances of transdifferentiation have been de- scribed outside the eye system [15]. Neural crest cells give rise to a very diverse range of cell types in vivo, but this may reflect cellular heterogeneity in the neural crest population with respect to dif- ferentiation potential [16]. Rat pancreatic cells can apparently convert into hepatocytes in vivo as a consequence of drug treatment or copper deple- tion/repletion [17], and different types of amphib- ian chromophore can interconvert to some extent in vitro [18]; however, these changes occur be- tween cell types sharing a common embryonic ancestry. The interconvertibility of embryonic NR and tapetal phenotypes is another example in this category, since both tissues derive from the optic cup. But this will not suffice to explain the ability of optic cup tissues (neural) to convert into lens (normally induced from head ectoderm) in vivo or in vitro. Such an ability is also shared by cultured cells from chick embryo pineal [19], a neural structure homologous to the median eye of lower vertebrates. At least in certain reptiles (e.g. Anolis; [20]) this median eye includes a cellular lens of neural derivation, containing authentic lens crystallins. II. TISSUE-SPECIFIC FUNCTIONS a. Retinal markers The chick retina comprises a single glial cell- type, the radially arranged Miiller fibres, plus several different neuronal cell types arranged in layers (viz., photoreceptor, horizontal, bipolar, amacrine and ganglion cells). These latter express a wide variety of differentiated functions, includ- ing several neurotransmitter-synthesising enzymes (e.g. choline acetyltransferase, CAT; glutamic acid decarboxylase, GAD) and corresponding receptor/uptake systems [21-24]. Most retinal neurones (apart from photoreceptors) express tetanus toxin receptors on their surfaces, and like other electrically excitable cells they can be stained by Merocyanin 540. Although many of these neuronal markers are expressed transiently in monolayer cultures of NR cells, both the markers and the neuronal cells themselves are lost in a characteristic sequence during the first few weeks in vitro [24]. The differentiated characteristics of NR glial (Miller) cells include the enzymes glutamine synthetase (GS) and carbonic anhydrase II (CA). Retinal GS activity is induced by glucocorticoid hormones (e.g. cortisol), first appearing at about the 16th day during embryonic development; immunologically detectable GS is confined to the Miiller cells and absent from neurones. Retinal explants or aggregate cultures of dissociated NR cells can be induced to express GS precociously by treatment with cortisol from about the 7th day onwards, but hormonal induction of GS is negligi- ble in dispersed monolayer cultures of such cells [25]. This implies a requirement for histotypic contacts between retinal neurones and glia in order for GS to be expressed in the latter, such contacts being disrupted in sparse monolayer cultures but retained in explants and restored in aggregates. In dense monolayer cultures of NR cells, cortisol treatment induces transient GS activity [26]; im- munocytochemical staining shows this GS expres- sion to be limited to those glial cells in intimate contact with neurones [27]. By contrast, CA is found in many neuronal cell types as well as in glia during early retinal development, but later dis- appears from the former and becomes glial- specific, after which CA expression increases only within the Miiller cells [28]. Recent studies with cloned GS and CA DNA probes show that the respective patterns of GS and CA activity closely parallel the expression of their corresponding mRNAs [29]. b. Lens markers Terminally differentiated lens fibres lose their nuclei and most cell organelles, their cytoplasm becoming filled with very high levels of crystallins. Lens Formation from Neural Retina in vitro 3 Lens cells also express specific membrane proteins, including the MP26 species characteristic of gap junctions in the lens [30]. The crystallins are characteristic structural polypeptides which together comprise the vast bulk of soluble lens protein; they are present only at trace levels in non-lens tissues [31]. Crystallins fall into four main classes, designated a, 8, yand 6, of which the first two are found throughout the vertebrates. 6 Crystallin is confined to birds and reptiles, where it largely or wholly replaces the y class prominent in fish, amphibia and mammals [32, 33]. In chick embryo lens there are two 6 _ crystallin polypeptides of Mr 48 and 50 kd [34], at least six 2 crystallins ranging from 21 to 35 kd, and two main a species of 19 and 20 kd [35]. DNA sequences corresponding to most of these chick crystallins have now been cloned [36-41], and DNA sequ- ences published for the aA and two 6 crystallin genes [42-45]. The chick @A crystallin gene contains two introns and encodes an unexpectedly long (~ 1600 nt) mRNA with an extensive 3’-terminal untrans- lated region. The aA promoter includes both TATATATA and CAAT boxes, located respect- ively at —-26 to -31 and —-67 to —70 bp relative to the cap site [42]. DNA sequences responsible for lens-specific expression of this gene have been identified by fusing aA 5’-promoter fragments onto the structural gene encoding chick 6 crystallin, and then microinjecting the chimaeric genes into mouse lens cells in culture. Since mice have no endogenous 6 gene, expression of the constructs can be monitored by immunocytochemistry. This analysis reveals an important sequence element located between —242 and —189, which can func- tion when reversed in orientation or even (to some extent) when moved 1.7 kbp downstream from the cap site; this element may thus fulfil an enhancer- like function [46]. Expression from the @A pro- moter seems to be wholly restricted to lens, with no detectable expression in non-lens cells [42]. The chicken genome contains two closely re- lated 6 crystallin genes designated 61 and 62, which are linked 4.2 kbp apart in the same orientation, 5’- 6 1-spacer- 6 2-3’. Both genes are highly complex and are split at homologous posi- tions by 16 introns each. Three exons (7, 12 and 15) are identical between 61 and 62, while most of the rest (apart from exons 1 and 2) show >70% sequence identity; even the introns (apart from B, E and I) contain extensive sequence homologies [44, 45]. The 61 mRNA sequence includes two possible translation initiation codons (AUG) lo- cated 19 amino acids apart in the predicted protein sequence; these would yield two § polypeptides of Mr 50,750 and 48,900, respectively [44]. Selective use of these alternative initiation codons may explain how changes in ionic balance can modify the relative synthetic rates of the 48 and 50 kd 6 polypeptides [47]. This derivation of both polypeptides from 61 now appears much more likely than one species arising from 61 and the other from 62. Hybridisation experiments using intron-specific probes suggest that in lens the 6 1 gene is at least 100-fold more active transcrip- tionally than §2 [42, 45]. Notably, the S’- promoter region of the 5 1 gene contains a CAAT box (-67 to -70) and two consensus core enhancer sequences (at -308 and +350), whereas the §2 gene lacks these features [48]. TAAAA boxes are found 5’ to the 61 (-24 to -28) and 6 2 (-26 to —30) genes, and both promoters support similar levels of transcription by an in vitro cell-free system [42]. However, when the 61 and 62 promoter regions (approx. 350 bp) are fused to bacterial chloramphenicol _acetyltransferase (bChAT) genes and introduced into chick lens cells, the § 1 promoter supports significantly (~ 5-fold) higher bChAT expression than does the 6 2 promoter [48]. Microinjection of a complete 6 1 gene sequence (plus 5’-flanking regions) into mouse cells of various types demonstrates that § expression is mainly lens-specific, i.e. much higher in lens than in non-lens cells such as fibroblasts [49]. Promoter deletion experiments indicate that high-level ex- pression of the §1 gene in mouse lens cells requires Only the promoter region from -80 to +35, whereas the 20-fold lower level of expression in fibroblasts requires an extra 12 bp of 5’-flanking sequence (from position —92; [50, 51]). Transient expression assays on ¢ 1 genes introduced into a wide variety of mouse cell types, show that high-level 6 expression is possible only in lens and embryonic epidermal cells, whereas other cell 4 D. I. DE POMERAI types permit only low-level expression (even in brain or retinal glia, or in tapetal cells; [51]). The 61 promoter region includes a GC-rich inverted repeat bracketing the CAAT box between —78 and -59; if various GC-rich DNA fragments are coinjected along with the ¢ 1 gene into mouse lens cells, then 6 expression is greatly reduced, sug- gesting that these GC-rich sequences may compete for the ubiquitous transcription factor Spl [52]. Consensus sequences for the TGGCA-binding factor are also found in the 61 promoter (-—60 to -48). Finally, brief mention should be made of recent studies of 6 expression in transgenic mice [53] and fish [54] carrying introduced $1 genes (neither mice nor fish have 6 genes of their own). In the transgenic fish (medaka), low levels of 6 expression are found in many tissues and there is no restriction to lens, perhaps because the DNA signals for lens-specific expression of the chick 6 1 gene have diverged too far from those used by the fish’s own crystallin genes [54]. By contrast, in transgenic mice the introduced 6 1 gene is expres- sed chiefly in lens tissue, albeit at lower levels than in transient 6 1 expression assays using mouse lens cells. However, 0 expression is also unexpectedly detectable in certain pyramidal neurones of the anterior piriform cortex [53]. II. FOUNDER CELLS AND EXTRALENTICU- LAR CRYSTALLINS Although a single melanised tapetal cell can give rise to lens cells among its clonal progeny [11], a similar analysis of NR transdifferentiation gives less clear-cut results. Clonal cultures of 8 day NR produce three types of colony, respectively con- taining lentoids, or pigment cells, or neither [55]; but no single cell generates both lens and pig- mented progeny in the same colony. In clonal cultures of 3.5 day NR cells, however, some colonies eventually develop lentoids, some pig- ment cells, some neither and some both [56]. Moreover, at earlier stages most colonies contain both neuron-like and epitheloid cells, presumably derived from a single neuroepithelial precursor. Thus many of these early NR cells appear to be multipotent, since up to four distinct phenotypes can emerge amongst their progeny, two of which are foreign to NR. In mass cultures of later (6 to 10 day) NR, two major categories of cell emerge after a few days in vitro [2]. These are: (i) flattened epitheloid (E) cells adhering to the substrate (presumably Miller glia or their precursors); and (ii) rounded neuron- like (N) cells, which often form clusters intercon- nected by neurite processes, and which attach preferentially to the underlying E cells (unless using highly adhesive substrates such as polylysine; [23, 57]). The lentoids which appear after 4-5 weeks in these cultures are likely to be derived from E cell progenitors, since N cells are mostly lost before this stage [2]. Several lines of evidence confirm this; thus lentoid appearance is promoted in E-cell cultures prepared by the selective eli- mination of N cells, using either mild dissociation [57] or the neurotoxin chinoform-ferric chelate [58]. In a reaggregate culture system (see section IV below) which speeds up transdifferentiation, CA-expressing retinoglial cells from 13 or 16 day NR convert directly into crystallin- and MP26- expressing lentoid cells [30, 59]. Moreover, pre- treatment of the NR tissue with the gliotoxin a-aminoadipic acid greatly reduces lentoid forma- tion [30]. In a similar reaggregate culture system, retinal N cells pre-labelled using tetanus toxin do not subsequently express ¢ crystallin, as shown by double immunofluorescent staining; thus few if any lens cells are derived directly from N cells [60]. With NR material from much earlier (3.5 day) embryos, however, there is some evidence for N cells converting into lentoids. This has come both from time-lapse photography and from chimaeric cultures combining quail N cells with chick E cells or vice versa; in either combination, both quail and chick 6 crystallins are expressed in late cultures, suggesting that some cells in the N fractions (all of which can be stained with Merocyanin 540) do go on to form lens [61]. Related questions are raised by the presence of crystallin transcripts (and proteins; [31]) in a variety of embryonic non-lens tissues. Broadly these fall into three categories: (i) Tissues such as early chick embryonic heart and liver which show no potential for transdif- ferentiation, nor for enhanced crystallin expres- sion in vitro. Nevertheless, a small proportion of Lens Formation from Neural Retina in vitro 5 cell clusters in such tissues contain nucleus- confined 6 transcripts detectable by in situ hybrid- isation, whereas neighbouring cells contain none [62, 63]. (ii) Tissues where a low level of 6 expression in vivo becomes enhanced during in vitro culture, but lentoids never develop and other crystallin types (e.g. a) are not expressed. This category includes 6 day chick embryo brain [64] and 3.5 day embryo limb bud cells [65]. Some 30% of cells in 3.5 day adenohypophysis contain § protein, but in this case § expression decreases during in vitro cul- ture, and also during later development in vivo [66, 67]. (iii) Tissues which are able to transdifferentiate when cultured in vitro, forming lentoids which express high levels of several crystallin types. This group includes not only chick embryonic NR and tapetum (see above) but also 3.5 day embryo brain [68] and 8 day quail embryo pineal [19]. The three neural tissues (NR, brain, pineal) can also trans- differentiate into pigment cells, while cultured explants of early embryonic tapetum can convert into neuroretinal derivatives [69]. Several studies have momitored crystallin tran- scripts in non-lens tissues, initially by solution hybridisation with cDNAs prepared from abun- dant lens mRNAs [70]. The levels of such transcripts in chick NR tissue drop 10-fold between 3.5 and 8 days of embryonic development, and become almost undetectable by hatching [71]. This trend apparently correlates with the declining ability of NR from older embryos to transdiffer- entiate into lens [72]. More recent studies using a cloned § probe confirm the decline of 6 tran- script levels in later embryonic NR tissue [73], and also detect § transcripts in other non-lens tissues such as 3.5 day embryo brain and limb bud. Most of these extralenticular § transcripts are larger than the 2kb § messenger [63] and presumably represent nuclear precursors, although 6 mRNA- sized transcripts predominate in the case of 3.5 day NR [73]. Jn situ hybridisation shows that most extralenticular § transcripts are indeed confined to the nuclei of certain cell groups [62], though cytoplasmic hybridisation is also detectable in 3.5 day NR [74]. Although it is tempting to link extralenticular (ectopic) 6 transcription with transdifferentiation potential, this relationship does not hold good in all cases ((i) and (ii) above). Could ectopic 6 expression merely reflect a general leakiness of transcriptional control for this gene (cf. low expression of 61 genes introduced into mouse fibroblasts, reviewed earlier)? But this would predict a uniformly low level of 6 expression in all nuclei, rather than the clustering of strongly positive cells surrounded by a majority of nonex- pressing cells, as revealed by in situ hybridisation [62]. Nevertheless, the proportion of 6- expressing cells is much greater in those tissues capable of transdifferentiation (e.g. ~15% in 3.5 day NR) than in those unable to do so (e.g. 0.1% in 3.5 day heart; [63]). Transcripts of aA crystallin are barely detectable in both NR and tapetum from 8 day embryos, whereas 6 RNAs are far more abundant in NR, and transcripts encoding the 25 kd f crystallin are undetectable in either tissue [41, 63, 75]. An antiserum directed against the a-crystallin fraction from chick lens stains only Miller glia and their precursors in sections of chick NR tissue [76]; however, it remains unclear whether this anti- serum recognises (i) a@ crystallin itself, or (ii) a related polypeptide, or (ili) an unrelated protein copurifying with lens a crystallin and also present in glia. It is worth noting that lens epithelial cells contain glial fibrillary acidic protein, previously thought to be confined to neural-derived cells [77]. A recent report [78] shows that 8-9 day chick embryo NR contains a subclass of glial-like cells immunostaining positively for 6 crystallin. These 6 -positive cells form a loose meshwork at the retinal/optic nerve boundary (Fig. 1); they are of glial morphology, and some at least express glial markers such as CA and GS (although GS remains undetectable in most NR glial cells until much later). It remains to determine (i) the level and function of 6 crystallin in these cells, and (ii) whether they also express other lens markers (e.g. aA crystallin or MP26). Plausibly, these ¢- positive glia might represent the elusive founder cells which act as lens precursors in transdif- ferentiating NR cultures. If so, the transformation of cell phenotype involved in this system may be less fundamental than was previously supposed. 6 D. I. DE PoMERAI Fic. 1. Boundary cells immunostaining positively for 6 crystallin in sections of 9 day embryonic neural retina. Panels a and b show adjacent sections stained with anti- 6 -crystallin (followed by an FITC-linked second antibody), and with haematoxylin/eosin respectively. Panels c and d show sections from a different preparation stained with anti- 6 -crystallin before (c) and after (d) the adsorption of the antiserum with a lens lysate. Experimental details are given in reference 78. Photographs kindly supplied by Dr. P. Linser. Magnification 375. duced under optimal conditions in vitro (sections But could a small population of such founder cells IV and V), even assuming a significant advantage directly give rise to all the lentoidal tissue pro- Lens Formation from Neural Retina in vitro 7 over other E cells in terms of survival and/or proliferation? One intriguing possibility, based on the model of Pritchard [79], is that 6 -positive glia might act as “leader cells” in vitro, encouraging their neighbours (initially §-negative) to join them in the process of conversion into lens. Overall, there seem to be significant overlaps between the gene sets active in lens and those actually or potentially expressed in NR glial cells [76, 78]. IV. CELL-SURFACE CHANGES ASSOCI- ATED WITH TRANSDIFFERENTIATION As noted in section IIa, histotypic contacts between retinal neurones and Miller cells are prerequisite for the latter to respond to cortisol by expressing GS mRNA and protein [25, 28, 29]. Disruption of such contacts, for instance by cultur- ing NR cells as dispersed monolayers, results in a loss of GS inducibility associated with a rapid depletion of intracellular cortisol receptors [80]. Both features are retained in aggregate cultures prepared from freshly dissociated NR cells, and notably some aspects of retinal tissue organisation are restored under such conditions. Such aggre- gate cultures of NR cells also fail to transdifferenti- ate into lens even after 28 days in vitro [81]. Nor is any transdifferentiation observed when dissociated NR cells are maintained for 4 days in monolayer culture, then redissociated and cultured as reag- gregates for a further 24 days [81]. However, if the monolayer phase of culture is extended to 8-10 days prior to reaggregation, then lentoids form within the reaggregates much more rapidly and extensively than in parallel cultures maintained as monolayers throughout [59, 81]. This is true not only for 8-9 day NR, but also for 13 and even 16 day NR, where glial cells are postmitotic [59]. As discussed in the previous section, the lentoidal tissue formed in this reaggregate system derives from NR glial cells (sensitive to a-aminoadipic acid; [30]) but not from tetanus-toxin-binding neurones [60]. These studies suggest a radical change in the state of cell determination among NR glial cells between the 4th and 10th days of monolayer culture; furthermore, this ‘transdetermination’ can occur even if mitosis is inhibited by using serum- free medium during this period [81]. The levels of 6 crystallin mRNA in 30-day reaggregate NR cultures (i.e. 10 days as monolayers, then a further 20 days as reaggregates) reach 27% of those in newly hatched chick lens; this compares with 0.7% in 30-day monolayer cultures (where most lentoid formation occurs after 30 days) and less than 0.02% in 30-day aggregate cultures of the same NR cells [82]. 6 Crystallin mRNA is just detect- able (at <0.01% of the level in lens) after 10 days of monolayer culture, suggesting that transdeter- mination may coincide with the first onset of 6 transcription in some NR cells. However, 6 crystallin may not be the most informative marker as regards this transdetermina- tion event. Studies by Moscona and co-workers [30] show that the lens membrane protein MP26 is absent from fresh retina or newly dissociated NR cells; however, it becomes immunologically detect- able in a few scattered cells after only 3 days of monolayer culture, and is expressed by most of the glial-like cells after 5 to 7 days. In lens, the MP26 protein is associated with gap junctions, although these are not found in the lentoids formed in reaggregate NR cultures [83]. @A crystallin tran- scripts are also detectable after only 7 days in monolayer NR cultures [41]. One important feature of this system is a dramatic change in the affinities of different retinal cell types for each other. Aggregates of freshly dissociated NR cells display an interspersion of glial (Miller) cells among retinal neurones of various types, reflecting a preferential affinity between glial and neuronal cells which is mediated in part by the cell-surface protein R_ cognin (expressed by all retinal cell types; [84]). In monolayer culture, however, the NR glial cells (but not neurones) rapidly lose both R cognin and their affinity for retinal neurones [83]. As a result, when such monolayer cultures are dissociated and reaggregated after 8 to 10 days in vitro, the modified glial cells adhere preferentially to each other, forming CA-positive cores which rapidly develop into crystallin-positive lentoids. Retinal neurones are excluded from cores, instead forming peripheral shells which do not participate in lentoidogenesis [30, 59, 83]. The loss of both R 8 D. I. DE PoMERAI cognin and affinity for neurones can be delayed for several days in monolayer cultures of NR cells by treatment with retinoic acid [85]. Lentoidogenesis, even in so-called ‘monolayer’ NR cultures, is frequently associated with multilayering of the E cells, and can be promoted by artificially folding the cell sheet [86]; this effect may also operate via increased contacts between modified glial cells. Other important cell-surface phenomena in- clude those mediated by the extracellular matrix (ECM), whose role is particularly clear in the tapetal transdifferentiation system. Growing tapetal cells on a collagen rather than plastic substrate blocks the appearance of lens cells but reinforces melanisation [87]. Moreover, treatment of tapetal cells with hyaluronidase promotes their transdifferentiation into lens (see next section; [88,89]), an effect which is presumably due to the degradation of ECM components. This situation recalls the enhanced synthesis of ECM associated with iris cell proliferation and repigmentation, but not with lens conversion, following lentectomy in adult newts [9]. Recent studies in our laboratory suggest that the ECM is also profoundly modified during lentoidogenesis in monolayer NR cultures, and is affected by medium hexose levels (Flor- Henry and de Pomerai, in preparation). V. MEDIUM INFLUENCES ON TRANSDIF- FERENTIATION Several early studies noted the differential effects on NR transdifferentiation of various medium formulations [86, 90], and of specific supplements such as bicarbonate (which promotes pigment cell formation; [91]) or insulin (which promotes both growth and lentoidogenesis; [92]). A combination of horse serum and high glucose, used to promote neuronal survival and differentia- tion in many neural culture systems, exerts a similar influence on chick NR cells in vitro, but also strongly inhibits transdifferentiation into lens [57]. This was used to assay the state of cell determination in NR cultures, by transferring them from standard into nonpermissive medium or vice versa at various times, then later assaying the amount of 6 crystallin produced [93]. Subse- quently, this approach was refined by combining minimum essential medium (MEM) with both horse and foetal calf sera; this allows extensive transdifferentiation at the normal glucose concen- tration (6mM; FH), but blocks lentoid formation when supplemented with glucose to 18 mM final (FHG; [94]). NR cultures changed from nonper- missive FHG into permissive FH medium can only transdifferentiate into lens if transferred on or before 21 days in vitro, but fail to do so if transferred later than 24 days. Conversely, cul- tures changed from FH into FHG medium are blocked (no § production) if transferred prior to the 15th day in vitro, but are able to transdiffer- entiate extensively if transferred on the 18th day or later. These experiments argue against the possibility of cell selection, whereby a subpopulation of lens precursor cells would overgrow the other cell types present under FH conditions, but would be elimi- nated or selectively disadvantaged under FHG conditions. Rather, determinative events appear to act between the 15th and 21st days of culture, such that NR cells become committed either to follow a lens differentiation pathway, or else not to do so [94]. The only external variable here is the concentration of glucose, although this could have multiple effects on the heterogeneous NR cell population in culture. In fact, all parameters of glucose metabolism studied (including glucose uptake, utilisation of the pentose shunt, lactate production and glycogen accumulation) are strong- ly stimulated in FHG as compared to FH cultures [95]. However, when inhibitors or antagonists of these metabolic processes are added as continuous supplements to FHG cultures, the effects vary markedly. At one extreme, iodoacetate inhibits lactate production but scarcely stimulates transdif- ferentiation [96]; at the other, ouabain both reduces glucose uptake and promotes lentoid formation almost to FH levels. Transdifferentia- tion is also enhanced by forskolin (an activator of adenyl cyclase) or dibutyryl cAMP, both of which reduce glycogen accumulation in FHG cultures via cAMP-stimulated glycogenolysis. The intermedi- ate levels of glycogen and 6 crystallin in such cultures, relative to those in FH and FHG con- trols, suggest an inverse relationship between the two markers [95]. Lens Formation from Neural Retina in vitro 9 Because glycogen is a differentiated feature of retinal Miiller cells, its accumulation in vitro might preclude E cells from later converting into lens. Enhanced glial differentiation leading to reduced transdifferentiation may also underlie the marked inhibition of lentoidogenesis observed in cortisol- supplemented dense NR cultures, which transient- ly express GS activity [26]. However, the known involvement of retinal neurones in this latter process (section IIa) poses the question of whether neuronal influences might similarly affect glycogen production and so mediate in the glucose block on transdifferentiation. Jn situ histochemistry shows that both glycogen and its main synthetic enzyme are localised in NR glial cells underlying clusters of neurones [95], reminiscent of the situation with cortisol induction of GS in dense monolayer cultures [27]. Moreover, neurones survive better under FHG conditions and show prolonged ex- pression of CAT as compared with FH controls [94]. An inhibitory effect of retinal neurones on transdifferentiation can also be demonstrated by recombining N and E cell fractions together; low N:E ratios allow extensive lentoid development, whereas high N:E ratios block this process [97]. Recent studies of glucose effects on transdif- ferentiation in neurone-stripped E-cell cultures suggest that the glucose block may be mediated partly via enhanced neuronal survival/differentia- tion, and partly by direct effects of high glucose on the E cells (Tobal et al., in preparation). Reducing the level of glucose in MEM below 6 mM results in poor survival of NR cells. But if glucose-free MEM is supplemented with 6mM fructose, then 6 crystallin production is strongly stimulated (by ~4-fold) and lentoids appear ear- lier than in 6 mM glucose control cultures (Flor- Henry and de Pomerai, in preparation). We are currently studying how these changes in hexose regime affect the extent and type of ECM produc- tion. A second effective means for inhibiting NR transdifferentiation into lens involves the use of medium 199 in place of MEM [98]. Medium transfer experiments show that cultures main- tained for 20 or even 30 days in 199 can still transdifferentiate extensively about 10 days after transfer into MEM, i.e. such cultures never be- come committed not to form lens (in contrast to the high glucose block). The reasons for this contrast will be discussed in the next section. We have tentatively identified the inhibitory agent in 199 as acetate, since supplementation of MEM with acetate significantly reduces NR conversion into lentoids (Flor-Henry, unpublished). Another issue to be addressed briefly in this section concerns the influence of serum factors on NR transdifferentiation into lens. de Pomerai and Gali [99] found that different types of serum exert a much greater influence on this process than do different batches of the same serum type. Specifi- cally, adult serum (from chicken, horse or even newborn calf) supports little if any crystallin accumulation, whereas foetal calf serum (FCS) or embryo extract (bovine or chick) allows extensive lentoid formation. The active agents in embryonic/ foetal sera appear to be of low molecular weight, able to diffuse out through a dialysis membrane. Thus, the dialysis medium (MEM plus low-MW serum components) supports transdifferentiation, whereas the dialysed serum (macromolecular frac- tion, plus MEM) does not. The active low-MW fraction is partially sensitive to both heat and immobilised trypsin treatments [100], suggesting that it could include small polypeptides such as growth factors. By contrast, adult sera appear to contain macromolecular components which inhibit transdifferentiation [99]. In the parallel tapetal system, modifications of the medium conditions can be used to obtain three types of culture, viz., (i) dedifferentiated tapetal cells, none of which express melanin or crystallin proteins; (ii) repigmented tapetal cells, all of which are fully melanised; and (iii) transdifferentiated lentoidal cells expressing abundant crystallins [88, 89]. The key variables include phenylthiourea (which inhibits melanin production) and hyalu- ronidase (presumably acting on ECM compo- nents), whose presence together in the medium favours (i) and (iii), but whose absence favours (ii). High culture densities also promote (iii), ascorbic acid is helpful both for (ii) and (iii), while dialysed FCS is used throughout. The availability of pure cell populations in each of these three states should greatly facilitate future molecular studies [75]. 10 D. I. DE POMERAI CRYSTALLIN REGULATION DURING TRANSDIFFERENTIATION VI. During the later stages of NR or PE transdif- ferentiation, the levels of putative crystallin mRNAs (hybridising to abundant lens cDNAs) increase by at least two orders of magnitude above the traces detectable earlier [101-103]; in vitro translation confirms that the most abundant trans- latable mRNAs in late NR cultures are those encoding crystallins. A longstanding question in developmental biology asks whether a given phe- notype can only be attained by way of an invariant and coordinated programme of gene expression. If such were the case for lens, then crystallin genes should be activated as a battery and expressed in the same order during transdifferentiation as during normal lens development. The data on extralenticular crystallin expression cited earlier (section III) cannot easily be reconciled with this simplistic view. For instance, increasing levels of 6 expression in 6 day embryonic brain cultures do not entrain any detectable expression of other crystallins such as a [64]. Evidence from earlier immunological studies also suggests that the var- ious crystallin classes must be regulated independ- TABLE 1. from NR or PE in vitro ently, e.g. during lentoid formation in NR as compared to lens epithelium cultures in vitro [104]. Recent investigations by Clayton and co-workers [63] fully confirm this view, using three specific DNA probes to monitor the levels of 6, aA and £25 crystallin mRNAs during normal lens develop- ment and during transdifferentiation of both NR and tapetal (PE) cultures. These findings are summarised in Table 1. Whereas 6 is the first crystallin detectable in the 2 day embryonic lens rudiment, in both NR and PE transdifferentiation systems aA crystallin mRNA is expressed many days before the appear- ance of 6 mRNA. 6 Transcripts are easily detectable in fresh NR tissue (day 0), but apparently disappear during the early stages of culture. Detectable 6 mRNA does not reappear until about 20 days in NR cultures and thereafter increases steeply; in PE cultures this sharp rise in 6 mRNA occurs even later. Fresh NR and PE tissues contain only faint traces of aA transcripts. However, in NR cultures significant levels of aA mRNA (plus a prominent precursor) are already expressed after 7 days in vitro, rising by 21 days to a moderate level which does not further increase up to 42 days. A similar but slightly later trend is Crystallin mRNA levels during lens development in vivo and during lentoidal transdifferentiation Normal lens NR cultures PE cultures Stage development in vitro in vitro Tinie Gnveulture aA! B25? o aA! B25 x aA! B25 ot 2 day embryo —- _ + + (+) — + (+) — (+) 0 (fresh tissue) 8 day embryo + (+) +++ + 7 days 15 day embryo + + +++4++] (4+) ae = 14 days Hatching (21 days) ++ ++ +44 + ata (+) ++ + — 21 days 1 month posthatch +++ +++ ++ ++ + ++ ++ + (+) 28 days 1 year posthatch +++ +++ = ++ ae Sbaatctat eat. ote aR + 35 days ++ + +44 44+ «+ + 35 days ++ + FHHH+ 44+ +44 444 42 days ND ND ND ++ +++ ND 49 days Relative levels of these three mRNAs are expressed very approximately as follows: ND, not done; —, undetectable; (+), trace; +, low level; ++, moderate level; +++, high level; ++-++, very high level. Unless otherwise indicated, these estimates were derived (subjectively) from Northern blots in reference 63. Additional sources include: 1 reference 41; *, reference 127; *, reference 105; and *, reference 75. Note that absolute mRNA levels cannot be compared accurately between the three probes used, nor between tissues. The relative mRNA levels in lens are at least an order or magnitude higher than those attained in NR or PE cultures, hence the numbers of +’s are not comparable across the 3 sets of data. Lens Formation from Neural Retina in vitro 11 apparent for aA mRNA in PE cultures. The contrast in 825 mRNA levels is much more strik- ing; this mRNA is expressed only at low levels in later NR cultures, but shows a curious two-phase accumulation profile in PE cultures, being present at fairly low levels between 14 and 35 days, then rising to high levels at 42 and 49 days. The data just cited imply independent transcriptional reg- ulation for these three crystallin genes [63]; they are not coordinately expressed. Only the steep rise in 6 mRNA levels in late NR and PE cultures correlates temporally with the appearance of lentoids, whereas other lens ‘markers’ (e.g. aA transcripts or MP26; see section IV) may appear long before the overt emergence of lens-fibre-like cells. In situ hybridisation studies confirm that 6 transcripts are undetectable during the first to third weeks of NR culture [74, 105]. After about 20 days, a few clusters of cell nuclei are found to be expressing § transcripts, but no hybridisation is apparent over the cytoplasm. Nuclear RNAs extracted from NR cultures at this stage contain only a4kb 6 precursor but no fully processed (2 kb) & mRNA [105]. This nuclear restriction phase persists until about 25 days, but by 30 days 6 hybridisation is abundant over both nuclei and cytoplasm in many areas of the culture, though groups of strongly hybridising nuclei can still be observed in non-lentoidal regions [105]. These and other indications suggest that NR cells become able to synthesise 6 transcripts at high levels a few days before they acquire the capacity to process these transcripts (to the 2kb § mRNA) or to export § messengers from the nuclei [74, 105]. Non-permissive high glucose (FHG) cultures show very little hybridisation with a 6 probe at late (35-40 day) stages, suggesting that 6 tran- scripts are either not synthesised at all (a transcrip- tional block) or else are rapidly degraded within the nuclei [105]. By contrast, in late 199 cultures 6 transcripts can be detected in Northern blots of total cellular but not cytoplasmic RNAs. Jn situ hybridisation confirms that many nuclei in such cultures contain § transcripts, but these are either not exported to, or else are rapidly degraded in, the cytoplasm. Both 6 precursors and the pre- dominant 2kb 6 mRNA are present in late 199 nuclei; thus medium 199 must act on 6 mRNA export or stabilisation, but not on 6 transcript synthesis or processing [105]. This affords a simple explanation for the observation that transferring late 199 cultures into MEM medium always results in transdifferentation about 10 days later [98]. In the tapetal transdifferentiation system, 0 precursors (but not mRNA) are readily detectable in fully dedifferentiated cultures, but disappear rapidly from repigmenting cultures, and are only present at trace levels in fresh tapetum. By contrast, dense lentoidal cultures of these cells express the 2kb 6 mRNA and corresponding protein at very high levels [75]. If crystallin genes are regulated differently during lentoidal transdifferentation in vitro as compared to lens development in vivo, one can ask whether there are any correlative differences in the DNA or chromatin of the crystallin gene regions. DNA methylation patterns in the 6 locus (as revealed by digestion with methylation-sensitive or -insensitive isoschizomer restriction enzymes) do not correlate with the levels of 6 expression in a range of chick tissue types [106]. Nor does the pattern of 6 gene methylation change detectably during NR transdifferentiation, despite a 100-1000 fold increase in § expression [107]. Nevertheless, there is evidence that the § locus does become hypomethylated in postmitotic lens fibres in vivo, an event which occurs after the onset of high 6 expression [108, 109]. Although gene expression bears no direct relationship to DNA hypomethyla- tion [110], there is a clearer link with the pattern of DNase I-hypersensitive sites (HSS) in the S’ regions near active genes. In the chick vitellogenin gene system, one specific HSS appears only when expression is induced hormonally [111]. Alterna- tive sets of HSS are associated with (i) low-level constitutive expression of the chick lysozyme gene in macrophages, and (ii) high-level hormone- induced expression of the same gene in oviduct [112]. We have initiated studies along these lines, using a form of direct end-labelling to look for DNase I-generated sub-bands among the major restriction fragments of the 6 locus [113]. As shown in Figure 2, our data indicate several HSS in the 5’ region of the 6 1 gene, only one of which (at 1000) shows the same location in both lens and 12 D. I. DE PoMERAI transdifferentiated NR cultures. Another HSS lying close to the transcription initiation site differs only slightly in location between the two tissues. However, the two sites at -2000 and —3750 in lens are apparently replaced by a single site at —1500 in transdifferentiated NR cultures [113]. Another relevant finding from this study suggests a radical change in the chromatin conformation of the 6 locus during NR culture, from largely DNase-I- resistant (nucleosomal chromatin) in fresh tissue and early 10 day cultures, to largely DNase-I- sensitive (smooth-fibre chromatin) in later 25 and 35 day cultures [113]. This conformational change in the 6 locus seems to occur within most cells, and may prepare the way for the widespread appearance of 6 transcripts in many nuclei be- tween 20 and 30 days. The gene transfer approach has recently been used in attempts to assay the potential for crystal- lin expression in cultured cells undergoing transdif- ferentiation. One interesting observation is that mouse retinal glial cells (which do not themselves transdifferentiate into lens in vitro) express trans- fected § genes only at low levels in early cultures, but at much higher levels in late cultures, suggest- ing that the latter have moved at least some way towards transdifferentiation [114]. A more de- tailed experimental study used chick a, chick 6 or viral promotors respectively linked to bChAT reporter genes [115], the fusion constructs being introduced into chick NR cells cultured under lentoidogenic (low N:E ratio) or non-lentoid- ogenic (high N:E ratio) conditions. The levels of bChAT expression from the viral promoter are high in early cultures of both types, but decline to low levels by 10 days. By contrast, bChAT expression from the @ or 6 promoters is low initially, but rises steeply between 10 and 20 days only in the lentoidogenic cultures (expression remains low in the blocked cultures). These results suggest that onset of expression for the introduced fusion genes somewhat precedes the onset of high-level transcription from the endogenous crys- tallin genes (particularly in the case of 6). In these experiments the a@ construct is expressed much more efficiently than the 6 construct after 20 days in vitro (cf. 21-day values in Table 1); it would be interesting to determine whether the reverse would be true after 35 or 40 days of culture. Possible relationships between the onset of expression for such fusion genes and the determinative phenomena discussed earlier (sec- tions IV and V) deserve careful investigation. Vil. LINKS WITH OTHER SYSTEMS Although there are few well-documented in- stances where differentiated vertebrate cells give rise to new phenotypes (see section I), there are at least two processes which can alter the pattern of gene expression in such cells either transiently or permanently. These are (i) the induction of a stress response by heat shock or other environ- mental insults [116], and (ii) oncogenesis associ- ated with the activation of endogenous proto- oncogenes [117]. In this context, it is noteworthy that a crystallins share limited sequence homolo- gies with small heat shock proteins [118], and some A/y crystallins with certain oncogenes [119]. The only such instance reported for 6 crystallin, however, is an intriguing antigenic relationship Lens 5’ C ere hae eS Ll NR ee ee ee een 5 10 kbp Fic. 2. Approximate locations of DNase-I-hypersensi- tive sites in the 5’-flanking region of the 61 gene in lens (upper panel) and in late transdifferentiated cultures of NR cells (lower panel). Nuclei from such material were treated with extremely low levels of DNase I and then digested completely with Bam HI, Hind III or Eco RI restriction en- donucleases. The presence of DNase-I-generated sub-bands was noted upon Southern blotting the DNA fragments and probing with a full-length 61 cDNA clone, péCRI17 [39]. The DNase-I- hypersensitive sites indicated by arrowheads show the simplest derivation of the observed sub-bands from restriction fragments of known size [113]. The site marked C is shared both by lens and late NR cultures, and that which lies just within the 6 1 gene may also be shared. a v — 4 r Lens Formation from Neural Retina in vitro 13 Fic. 3. Co-localistation of pp 60°° and 6 crystallin in NR-derived lentoids. NR cultures were stained with rabbit anti-pp60"*”* followed by PAP, and then with rat anti- 6 -crystallin followed by rhodamine-linked anti-rat IgG. Panels a and b show negative controls in which preimmune rabbit serum replaced the anti-pp60"*”* (panel a), and where the rat anti- 6 treatment was omitted (panel b). Panels c and e show PAP staining for pp60°°” in lentoids, while panels d and f show rhodamine immunofluorescence for 6 crystallin in the same fields. Magnification <200. Full details are given in reference 137. Note that (unlike 6 ) pp60°°”* is not uniformly present throughout lentoids. 14 D. I. DE PoMERaI with feather keratin [120]. a. Stress response In most cell types from bacteria to man, expo- sure to a temperature increase of ~5°C above ambient, or to various toxic agents (arsenite, Cd**, aminoacid analogues), results in a suspen- sion or diminution of the normal protein synthetic pattern and in the rapid induction of a small number of characteristic heat shock proteins (HSPs). The major HSPs are of approximate Mr 70 kd, 85-90 kd, and one or more in the 22-28 kd range. The induction of these HSPs is regulated transcriptionally, but they themselves exert gener- alised post-transcriptional effects on the expres- sion of normal cellular proteins, and rescue a heat-shock-induced block on pre-mRNA splicing ({121]; note that most HSP genes are intron-free). Even if cells are maintained continuously at the higher temperature, the heat shock response even- tually fades out, with HSP expression diminishing and the normal pattern of protein synthesis being resumed. This results from complex autoregula- tory interactions among the HSPs [122]. Carr and de Pomerai [123] investigated whether a stress response (e.g. to adverse conditions experienced during culture) might play any role in NR transdifferentiation. Under standard (FH) permissive conditions, transient exposure of NR cultures to heat shock does not result in precocious crystallin expression, although both 70 and 85 kd HSPs are readily induced. A 24 kd HSP, induced by sodium arsenite (but not by heat) in NR cells and by heat in lens, comigrates with one of the 8 crystallins on SDS gels, but proves to be immuno- logically related [124, 125]. However, NR cells cultured in medium 199 display an intriguing response to heat shock. In the first place, 199 cultures maintained continuously at 43°C (heat shock temperature) readily transdifferentiate into lentoids and express high levels of 6 crystallin, in sharp contrast to the block on both apparent at 37°C [126]. Secondly, 6 crystallin synthesis can be induced by transferring late 199 cultures from 37°C to 43°C, although the kinetics of this induction are much slower (several days) than those for the HSPs (a few hours). Further experiments suggest (i) that actinomycin does not block increased 6 synthesis if added after the commencement of the heat shock response, and (ii) that agents such as sodium arsenite can elicit a similar response in late 199 cultures kept throughout at 37°C. Thus HSP induction somehow relieves a block on the export or stabilisation of § mRNAs which normally operates in late 199 cultures at 37°C [105]. An effect via pre-mRNA splicing seems much less likely, since the bulk of 6 transcripts confined to late 199 nuclei are already in the 2kb § mRNA size-class. It is curious that § synthesis should be stimu- lated at all following heat shock (a response expected only from authentic HSPs). In fact, this is also true in chick embryo lenses, where 6 synthesis increases by some 20% during HSP induction [124]. During post-hatching stages of chick lens development, when 6 protein synthesis is low although 6 mRNA remains present [127], heat shock treatment of lens epithelial explants induces a very marked (2-5 fold) stimulation of 6 synthesis [126]. It will be interesting to determine whether these heat shock responses also involve the export and/or stabilisation of 6 mRNAs that were previously confined to the nuclei. There is also the intriguing possibility that stress might transiently exert similar effects when culturing cells which show ectopic 6 transcript expression in vivo. b. Oncogene expression Cellular proto-oncogene products function at various levels in the transduction of signals across the cell membrane and in the control of cell proliferation [128]. In the latter context, transcrip- tion of the c-myc proto-oncogene has been demon- strated both in dedifferentiated tapetal cultures and in prelentoidal NR cultures, but not in crystallin-expressing lentoidal cultures from either source [125, 129]. The correlation between multi- layering of the NR cell sheet (loss of contact inhibition?) and lentoidogenesis may suggest another tenuous link with cancer cells. The proto-oncogene c-src encodes a membrane- associated tyrosine kinase (pp60°°"") homologous to the transforming protein of Rous Sarcoma Virus (pp60"*’). Whereas the latter induces rapid cell proliferation and transformation, the former does Lens Formation from Neural Retina in vitro 15 not, even when grossly overexpressed by NR cells infected with an RSV variant carrying the c-src gene in place of v-src [130]. The normal function of the c-src gene may be linked with differentiation rather than proliferation, since high levels of pp60°°” are expressed during the differentiation of postmitotic neurones in chick NR [131] and cere- bellum [132], as well as during smooth muscle development [133] and early neural tube induction [134] in the chick embryo. Maximal expression of pp60°°”* protein and its kinase activity occurs in chick NR tissue at around the 13th day of embryonic development [131, 135]; this appears to be regulated transcriptionally, since c-src mRNA levels also peak at this time [136, 137]. During the early stages of NR culture, pp60°°”* is expressed at moderate levels in neuro- nal cells, as shown by in situ immunocytochemis- try. However pp60°° kinase activity increases sharply in late lentoidogenic (FH) cultures but falls in blocked FHG or 199 cultures [137]. Immunocy- tochemical staining confirms that most of the pp60°*”* protein in late FH cultures is confined to ientoidal structures (Fig. 3). In lens, pp60°*’° kinase activity increases during embryonic de- velopment, and is maximal in the lens epithelial fraction of newly hatched chick lens [137]. These data suggest an elevated level of pp60°°’* expres- sion during the onset of lens fibre differentiation, but a lower level in established lens fibres (both in vivo and in transdifferentiated NR cultures). No- tably, lens fibre cells are highly elongated (as are smooth muscle and neuronal cells), and it may be that pp60°°”’ mediates in the cytoskeletal and/or cell surface reorganisation required during cell elongation. We are currently using a c-src-carrying RSV variant (kindly supplied by Dr. R. Jove) to determine whether high levels of pp60°°” expres- sion might enhance NR transdifferentiation into lens. If so, this would draw another sharp contrast between the c-src and v-src gene products. RSV infection of chick lens cells [138] or quail NR cells [139] blocks lens fibre differentiation and crystallin expression. This block depends on the v-src function, since it can be relieved by shifting to a non-permissive temperature in the case of RSV mutants encoding a temperature-sensitive v-src product. With ts-RSV transformed quail NR cells, such a shift results in appearance of 6 plus a crystallin mRNAs and a protein within 48 hours, whereas 6 protein and lentoids only appear after a further 5 days. By contrast, quail NR cells transformed with a different retrovirus (Mill Hill 2) permanently express both a and 6 crystallins and also contain stem cells for lentoid production [139]. VIII. CONCLUDING REMARKS The preceding pages of this review have tried to summarise current knowledge of NR _transdif- ferentiation into lens (and closely related systems). Sections III to VII have dealt with the main events in approximately chronological order. In this final section, I wish to pose some of the many questions which remain open. In the first place, it is apparent that the end result of NR transdifferentiation, namely a close approximation to a lens-fibre phenotype, is arrived at as the culmination of a complex sequence of events which initially bear little relationship to lens formation in vivo. There can be no question of simply switching on some master programme for lens development. How such diverse routes con- verge on a common endpoint remains a fascinating question. Further investigations are needed to elucidate the nature of the NR founder cells which give rise to lentoids in vitro. Are they restricted to progeny of the 6 -positive ‘boundary’ cells identified in vivo [78], or could such cells influence their neighbours in culture to join them in lentoidogenesis [79]? This may become amenable to analysis if the 6 - positive cells can be identified and tagged in early NR cultures. The significance of ectopic 6 expression is another problem requiring attention. What is the pattern of 6 gene regulation in the expressing cells, and is there a distinction of kind or only of degree between the levels of 6 transcription observed in those tissues able to transdifferentiate into lens and in those unable to do so? How easy is it for retinal glial cells to switch on a subset of lens markers (e.g. 6 and aA crystallins, MP26), even under conditions which do not lead on to lentoid formation? Clarification is needed as to whether the deter- 16 D. I. DE PoMERAI minative events revealed by reaggregation experi- ments (effective between 4 and 10 days) are different in kind from those implied by medium transfers (occurring between 15 and 21 days). The former may coincide temporally with MP26 and aA expression, while the latter correlate more closely with the onset of 6 expression, or at least with the widespread conformational change of 6 - locus chromatin from DNase I-resistance to sensi- tivity. However, it may be less confusing to think of determination in this system as progressive rather than multistep, since the timing of any commitment to form lens may be largely a function of the stabilising or destabilising effects of later culture conditions. Thus increased glial cell con- tacts in reaggregates are evidently conductive to lentoid formation, whereas high glucose levels favour more normal glial differentiation through a stimulation of glycogen production. Either way, a single discrete determination decision appears unlikely. The control of crystallin gene expression during NR transdifferentiation raises several fur- ther questions. It is generally assumed (though not yet proven) that the rate of 6 gene transcription must increase greatly when © transcripts begin to accumulate within the nuclei. RNA processing, export and stabilisation controls would subse- quently amplify this primary transcriptional event [74, 105]. If this model is approximately correct, then what factors are responsible for activating 6 transcription? Positively acting transcription fac- tors would be expected to interact with the 5’-flanking regions of the 6 1 gene, a feature which would prove amenable to foot-printing analysis. The locations of protein-binding and DNase-I- hypersensitive sites in the 6 gene region will need to be compared in detail between lens tissue and transdifferentiated cultures. Preliminary work along these lines suggests partially overlapping sets of DNase-I-hypersensitive sites (Fig. 2). Conceiv- ably, the common site(s) might correlate with high-level 6 expression (a feature of both lens and NR lentoids), while the differences might reflect the fact that (some) NR cells originally expressed only low levels of 6 [78]. Could transdifferentia- tion involve some kind of ‘overshoot’ in the expression of such lens markers, since low levels seem to be tolerated in several non-lens tissues, but high levels are appropriate only to lens? The link between heat shock and _post- transcriptional events (6 mRNA export and/or stabilisation) in late 199 cultures poses the follow- ing question; do HSPs bind directly to intranuclear 6 mRNAs and move with them out into the cytoplasm? (It is known that HSPs do move from nucleus to cytoplasm during recovery from heat shock; [140]). The generality of such post- transcriptional controls over crystallin expression also remains to be clarified, both in lens and non-lens tissues. Finally, the role of c-src expression in develop- ing lens fibres requires more detailed exploration. Is pp60°°" also expressed in lentoids derived in vitro from tapetum? Can c-src-substituted RSV stimulate lentoid development under appropriate NR culture conditions? Why indeed does c-src expression affect NR cells so differently from v-src expression [130, 137]? Some elucidation may be forthcoming from studies of quail NR cells trans- formed with ts-RSV or Mill Hill 2 viruses, which should provide homogeneous cell populations after a few subcultures. These will afford a unique opportunity to follow the time course of molecular events during transdifferentiation in a largely synchronous system [139]. ACKNOWLEDGEMENTS The author is grateful to his associates (Mr. M. McLaughlin, Mr. M. Flor-Henry, Dr. K. Tobal and Dr. K.C. Perry) for permission to quote unpublished findings, and for many stimulating discussions in this field. The Cancer Research Campaign and SERC are thanked for financial support. Particular thanks are due to Dr. P. Linser for the photographs in Fig. 1, to Mrs. R. M. Clayton for permission to use her data in Table 1, and to Mrs. E. O. Wigginton and Ms. J. Barton for typing the manuscript. REFERENCES 1 Itoh, Y., Okada, T.S., Ide,H. and Eguchi, G. (1975) Dev. Growth Differ., 17: 39-50. 2 Okada, T. S., Itoh, T., Watanabe, K. and Eguchi, G. (1975) Dev. Biol., 45: 318-329. 3. Okada, T. S. (1980) Curr. Topics Dev. Biol., 16: 349-390. 4 Okada, T. . (1983) Cell Differ., 13: 177-183. 24 25 26 27 28 29 Lens Formation from Neural Retina in vitro 17 Yamada, T. (1977) Control Mechanisms in Cell- type Conversions in Newt Lens Regeneration. (Monogr. Dev. Biol., ed. by A. Wolsky, Vol. 13), S. Karger, Basel. Cuny, R., Jeanny, J.-C. and Courtois, Y. (1986) Differentiation, 32: 221-229. Borst, D. E., Di Rienzo, S. M. and McDevitt, D. S. (1981) J. Cell Biol., 91 (supplement): 28a. Yamada, T. and McDevitt,D.S. (1984) ferentiation, 27: 1-12. Kulyk, W. M. and Zalik, S. E. (1982) Differentia- tion, 23: 29-35. Eguchi,G., Abe,S. and Watanabe, K. (1974) Proc. Natl. Acad. Sci. U.S.A., 71: 5052-5056. Eguchi, G. and Okada, T. S. (1973) Proc. Natl. Acad. Sci. U.S.A., 70: 1495-1499. Bode, H., Dunne, J., Heinfeld,S., Huang, L., Javios,L., Koizumi,O., Westerfield,J. and Yaross, M. (1986) Curr. Topics Dev. Biol., 20: 257-280. Schmid, V. and Adler, H. (1984) Cell, 38: 801- 809. Schmid, V. and Adler, H. (1986) Curr. Topics Dev. Biol., 20: 117-135. Okada, T. S. (1986) Dev. Growth Differ. , 28: 213- 221. Weston, J. A. (1986) Curr. Topics Dev. Biol., 20: 195-210. Rao, M.S., Scarpelli, D.G. and Reddy, J.K. (1986) Curr. Topics Dev. Biol., 20: 63-78. Ide, H. (1986) Curr. Topics Dev. Biol., 20: 79-87. Watanabe, K., Aoyama, H., Tamamaki, N., Yasu- jima, M., Nojyo, Y., Ueda, Y. and Okada, T. S. (1985) Cell Differ., 16: 251-257. McDevitt, D. S. (1972) Science, 175: 763-764. Crisanti-Combes, P., Pessac, B. and Calothy, G. (1978) Dev. Biol., 65: 228-232. Guerinot, F. and Pessac, B. (1979) Brain Res., 162: 179-183. de Pomerai, D.I. and Carr, A. (1982) Exp. Eye Res., 34: 553-563. de Pomerai, D.I., Kotecha, B., Flor-Henry, M., Fullick, C., Young, A. and Gali, M. A. H. (1983) J. Embryol. Exp. Morphol., 77: 201-220. Linser, P. J. and Moscona, A. A. (1979) Proc. Natl. Acad. Sci. U.S.A., 76: 6476-6480. de Pomerai, D.I., Carr, A., Soranson, J. and Gali, M. A. H. (1982) Differentiation, 22: 6-11. Linser, P.J. and Moscona, A. A. (1983) Dev. Biol., 96: 529-534. Linser, P. J. and Moscona, A. A. (1981) Natl. Acad. Sci. U.S.A., 78: 7190-7194. Vardimon, L., Fox,L.E. and Moscona, A. A. (1986) Proc. Natl. Acad. Sci. U.S.A., 83: 9060- 9064. Dif- Proc. 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 Si 52 Moscona, A. A., Brown, M., Degenstein, L., Fox, L. and Soh, B. M. (1983) Proc. Natl. Acad. Sci. U.S.A., 80: 7239-7243. Clayton, R. M., Campbell, J.C. and Truman, D. E. S. (1968) Exp. Eye Res., 7: 11-29. Clayton, R. M. (1974) In “The Eye”, vol. 5. Ed. by H.Davson and L.T. Graham, Jr., Academic Press, New York & London, pp. 399-494. Piatigorsky, J. (1984) Cell, 38: 620-621. Reszelbach, R., Shinohara, T. and Piatigorsky, J. (1977) Exp. Eye Res., 25: 583-593. Thomson, I., Wilkinson, C. E., Burns, A. T. H., Truman, D. E. S. and Clayton, R. M. (1978) Exp. Eye Res., 26: 351-362. Bhat, S. P. and Piatigorsky, J. (1979) Proc. Natl. Acad. Sci. U.S.A., 76: 3299-3303. Bower, D.J., Errington, L.H., Wainwright, N. R., Sime, C., Morris, S. and Clayton, R. M. (1981) Biochem. J., 201: 339-344. Hejtmancik, J. F. and Piatigorsky, J. (1983) J. Biol. Chem., 258: 3382-3387. Nickerson, J. M. and Piatigorsky, J. (1984) Proc. Natl. Acad. Sci. U.S.A., 81: 2611-2615. Yasuda, K., Nakajima, N., Isobe, T., Okada, T. S. and Shimura, Y. (1984) EMBO J., 3: 1397-1402. Errington, L.H., Bower,J., Cuthbert,J. and Clayton, R. M. (1985) Biol. Cell, 54: 101-108. Yasuda, K. and Okada, T. S. (1986) Oxford Surv. Eukaryotic Genes, 3: 183-209. Ohno, M., Sakamoto, H., Yasuda, K., Okada, T. S. and Shimura, Y. (1985) Nucleic Acids Res., 13: 1593-1606. Nickerson, J. M., Wawrousek, E. F., Hawkins, J. W., Wakil, A.S., Wistow, G.J., Thomas, G., Norman, B. L. and Piatigorsky, J. (1985) J. Biol. Chem., 260: 9100-9105. Nickerson, J. M., Wawrousek, E. F., Borras, T., Hawkins, J. W., Norman, B.L., Filpula,D.R., Nagle, J.W., Ally, A.H. and Piatigorsky, J. (1986) J. Biol. Chem., 261: 552-557. Okazaki, K., Yasuda, K., Kondoh, H. and Okada, T.S. (1985) EMBO J., 4: 2589-2595. Shinohara, T. and Piatigorsky, J. (1977) Nature (Lond.), 270: 406-411. Borras, T., Nickerson, J. M., Chepelinsky, A. B. and Piatigorsky, J. (1985) EMBO J., 4: 445-452. Kondoh, H., Yasuda, K. and Okada, T. S. (1983) Nature (Lond.), 301: 440-442. Hayashi, S., Kondoh, H., Yasuda, K., Soma, G., Ikawa, Y. and Okada, T. S. (1985) EMBO J., 4: 2201-2207. Kondoh, H., Hayashi, S., Takahashi, T. and Oka- da, T. S. (1986) Cell Differ., 19: 151-160. Hayashi, S. and Kondoh, H. (1986) Mol. Biol., 6: 4130-4132. Cell 18 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 D. I. DE PoMERAI Kondoh, H., Katoh, K., Takahashi, Y., Fujisawa, H., Yokoyama,M., Kimura,S., Saito, M., Nomura, T., Hiramoto, Y. and Okada, T. S. (1987) Dev. Biol., 120: 177-185. Ozato, K., Kondoh, H., Inohara, H., Iwamatsu, T., Wakamatsu, Y. and Okada, T. S. (1986) Cell Differ., 19: 237-344. Okada, T. S. (1977) Dev. Growth Differ., 19: 47- 55. Okada, T. S., Yasuda, K., Araki, M. and Eguchi, G. (1979) Dev. Biol., 68: 600-617. de Pomerai, D.I. and Gali, M. A.H. (1981) J. Embryol. Exp. Morphol., 62: 291-308. Shinde, S. L. and Eguchi, G. (1982) In “Problems of Normal and Genetically Abnormal Retinas”. Ed. by R.M. Clayton, J. Haywood, H. W. Read- ing and A. Wright, Academic Press, London & New York, pp. 37-48. Moscona, A. A. and Degenstein, L. (1980) Cell Differ., 10: 36-41. de Pomerai, D.I., Takagi,S., Kondoh,H. and Okada, T. S. (1984) Dev. Growth Differ., 26: 111- 119. Kondoh, H., Takagi, S., Nomura, K. and Okada, T. S. (1983) Roux’ Arch. Dev. Biol., 192: 256-261. Jeanny, J.-C., Bower,D.J., Errington, L. H., Morris, S. and Clayton, R. M. (1985) Dev. Biol., 112: 94-99. Clayton, R. M., Jeanny, J.-C., Bower, D. J. and Errington, L. H. (1986) Curr. Topics Dev. Biol., 20: 137-151. Takagi, S. (1986) Roux’ Arch. Dev. Biol., 195: 15- 21. Kodama, R. and Eguchi, G. (1982) Dev. Biol., 91: 221-226. Barabanov, V.M. (1977) SSSR, 224: 195-198. Ueda, Y. and Okada, T. S. (1986) Cell Differ., 19: 179-185. Nomura, K. (1982) Differentiation, 22: 179-184. Tsunematsu, Y. and Coulombre, A. J. (1981) Dev. Growth Differ., 23: 297-311. Jackson, J. F., Clayton, R.M., Williamson, R., Thomson, I., Truman, D. E.S. and de Pomerai, D. I. (1978) Dev. Biol., 65: 383-395. Clayton, R. M., Thomson, I. and de Pomerai, D. I. (1979) Nature (Lond), 282: 628-629. de Pomerai, D. I. and Clayton, R. M. (1978) J. Embryol. Exp. Morphol., 47: 179-193. Agata, K., Yasuda, K. and Okada, T.S. (1983) Dev. Biol., 100: 222-226. Bower, D.J., Errington, L.H., Pollock, B. J., Morris, S. and Clayton, R. M. (1983) EMBO J., 2: 333-338. Eguchi, G. (1986) Curr. Topics Dev. Biol., 20: 21- Dokl. Akad. Nauk. 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 37. Moscona, A. A., Fox, L., Smith, J. and Degen- stein, L. (1985) Proc. Natl. Acad. Sci. U.S.A., 82: 5570-5573. Hatfield, J. S., Skoff, R.P., Maisel, H. and Eng, L. (1984) J. Cell Biol., 98: 1895-1898. Linser, P. J. and Irvin, C. K. (1987) Dev. Biol., 121: 499-509. Pritchard, D. J. (1986) Foundations of Develop- mental Genetics. Taylor Francis, London. Saad, A.D., Soh,B.M. and Moscona, A. A. (1981) Biochem. Biophys. Res. Comm., 98: 701- 708. Okada, T.S. Nomura, K. and Yasuda, K. (1983) Cell Differ., 12: 85-92. Yasuda, K. Okuyama, K. and Okada, T. S. (1983) Cell Differ., 12: 177-183. Ophir, I., Moscona, A. A., Loya,N. and Ben- Shaul, Y. (1985) Cell Differ., 17: 149-157. Ophir, I., Moscona, A.A. and Ben-Shaul, Y. (1983) Cell Differ., 13: 133-141. Degenstein, L. and Moscona, A. A. (1986) Exp. Eye Res., 43: 93-102. Clayton, R. M., de Pomerai, D. I. and Pritchard, D. J. (1977) Dev. Growth Differ., 19: 319-328. Yasuda, K. (1979) Dev. Biol., 68: 618-623. Itoh, Y. and Eguchi, G. (1986) Cell Differ., 18: 173-182. Itoh, Y. and Eguchi, G. (1986) Dev. Biol., 115: 353-362. Agata, K., Kondoh, H., Takagi, S., Nomura, K. and Okada, T. S. (1980) Dev. Growth Differ., 22: 571-577. Pritchard, D. J., Clayton, R. M. and de Pomerai, D. I. (1978) J. Embryol. Exp. Morphol., 48: 1-21. de Pomerai, D. I. and Clayton, R. M. (1980) Dev. Growth Differ., 22: 49-60. de Pomerai, D. I. and Gali, M. A. H. (1981) Dev. Growth Differ., 23: 229-236. de Pomerai, D. I. and Gali, M. A. H. (1982) Dev. Biol., 93: 534-538. Karim, S.A. M., Flor-Henry, M. and de Pomerai, D. I. (1987) Cell Differ., 22: 29-46. Karim, S. A. M. (1987) Ph.D. Thesis, University of Nottingham. Takagi, S., Kondoh, H., Nomura, K. and Okada, T. S. (1983) J. Embryol. Exp. Morphol., 73: 97- 109. Gali, M. A. H. and de Pomerai, D. I. (1984) Dif- ferentiation, 25: 238-246. de Pomerai, D. I. and Gali, M. A. H. (1982) Dev. Growth Differ., 24: 233-243. Gali, M. A. H. (1983) Ph. D. Thesis, University of Nottingham. Thomson, I., Wilkinson, C. E., Jackson, J. E., de 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 Lens Formation from Neural Retina in vitro 19 Pomerai, D. I., Clayton, R. M., Truman, D. E. S. and Williamson, R. (1978) Dev. Biol., 65: 372- 382. Thomson, I., de Pomerai, D. I., Jackson, J. F. and Clayton, R. M. (1979) Exp. Cell Res., 122: 73-81. Thomson, I., Yasuda, K., de Pomerai, D. I., Clayton, R. M. and Okada, T. S. (1981) Exp. Cell Res., 135: 445-449. de Pomerai, D. I., Pritchard, D. J. and Clayton, R. M. (1977) Dev. Biol., 60: 416-427. Carr, A. and de Pomerai, D. I. (1985) Dev. Biol., 111: 119-128. Bower, D. J., Errington, L. H., Cooper, D. N., Morris, S. and Clayton, R. M. (1983) Nucleic Acids Res., 11: 2513-2521. Errington, L. H., Cooper, D. N. and Clayton, R. M. (1983) Differentiation, 24: 33-38. Grainger, R. M., Hazard-Leonards, R. M., Sama- ha, F., Hougan, L. M., Leek, M. R. and Thom- sen, G. H. (1983) Nature (Lond.), 306: 88-91. Sullivan, C. H. and Grainger, R. M. (1986) Proc. Natl Acad. Sci. U.S.A., 83: 329-333. Bird, A. P. (1986) Nature (Lond.), 321: 209-213. Burch, J. B. E. and Weintraub, H. (1983) Cell, 33: 65-76. Fritton, H. P., Igo-Kemenes, T., Nowock, J., Strech-Jurk, U., Theisen, M. and Sippel, A. E. (1984) Nature (Lond.), 311: 163-166. McLaughin, M. (1987) Ph. D. Thesis, University of Nottingham. Kondoh, H. and Okada, T. S. (1986) Curr. Topics Dev. Biol., 20: 153-164. Kondoh, H., Ueda, Y., Hayashi, S., Okazaki, K., Yasuda, K. and Okada, T. S. (1987) Cell Differ., 20: 203-207. Schlessinger, M. J., Ashburner, M. and Tissieres, A. (1982) Heat Shock: from Bacteria to Man. Cold Spring Harbour Press. Bishop, J. M. (1983) Ann. Rev. Biochem., 52: 301-354. Ingolia, T. D. and Craig, E. A. (1982) Proc. Natl. Acad. Sci. U.S.A., 79: 2360-2364. Crabbe, M. J. (1985) Biosci. Rep., 5: 167-174. 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 Kodama, R. and Eguchi, G. (1983) Dev. Growth Differ., 25: 261-270. Yost, H. J. and Lindquist, S. (1986) Cell, 45: 185-193. Di Domenico, B. J., Bugaisky, G. E. and Lind- quist, S. (1982) Cell, 31: 593-603. Carr, A. and de Pomerai, D. I. (1985) Dev. Growth Differ., 27: 435-445. Collier, N. C. and Schlessinger, M. J. (1986) Exp. Eye Res., 43: 103-118. Carr, A. (1985) Ph.D. Thesis, University of Not- tingham. de Pomerai, D. I. and Carr, A. (1987) Dev. Growth Differ., 29: 37-46. Hejtmancik, J. F., Beebe, D. C., Ostrer, H. and Piatigorsky, J. (1985) Dev. Biol., 109: 72-81. Heldin, C. H. and Westermark, B. (1984) Cell, 37: 9-20. Eguchi, G., Agata, K., Itoh, Y. H. and Itoh, Y. (1985) Cell Differ., 16 (supplement): 127S. Iba, H., Jove, R. and Hanafusa, H. (1985) Mol. Cell. Biol., 5: 2856-2859. Sorge, L. K., Levy, B. T. and Maness, P. F. (1984) Cell, 36: 249-257. Fults, D. W., Towle, A. C., Lauder, J. M. and Maness, P.F. (1985) Mol. Cell. Biol., 5: 27-32. Barnekow, A. and Bauer, H. (1984) Biochim. Biophys. Acta, 727: 94-102. Maness, P. F., Sorge, L. K. and Fults, D. W. (1986) Dev. Biol., 117: 83-89. Cotton, P. C. and Brugge, J. S. (1983) Mol. Cell. Biol., 3: 1157-1162. Vardimon, L., Fox, L. E. and Moscona, A. A. (1986) Mol. Cell. Biol., 6: 4109-4111. Ellis, D. K., Carr, A. and de Pomerai, D. I. (1987) Development, 101: 847-856. Jones, R. E., de Feo, D. and Piatigorsky, J. (1981) Vision Res., 21: 5-9. Simonneau, L., Crisanti, P., Lorinet, A. M., Alliot, F., Courtois, Y., Calothy, G. and Pessac, B. (1986) Mol. Cell. Biol., 6: 3704-3710. Velazquez, J. M. and Lindquist, S. (1984) Cell, 36: 655-662. - 2 # ih ame , ar Mou “es The sales : oy ay hea ‘ i, yy oe are) ae) ar © 7 jaa: Wy { ; i Hoey” a ‘s { 7 i f i ; ts ‘ a “ ~ 7 4 t —= : 7 : : si Ve + i i a ' me ie + Ba 4 let a a a y PIL ee of a - > 1 : V oi Ds 7 : 7 ce ea o ZOOLOGICAL SCIENCE 5: 21-32 (1988) © 1988 Zoological Society of Japan REVIEW The Neurohypophysis of Cyclostomes as a Primitive Hypothalamic Center of Vertebrates KAZUHIKO TSUNEKI! Department of Biology, Shimane University, Matsue, Shimane 690, Japan INTRODUCTION The function of the adenohypophysis of gnatho- stome vertebrates is regulated by the hypothala- mus usually by means of releasing or inhibiting neurosecretory hormones (regulating hormones) liberated to portal vessels which irrigate the adenohypophysis. The activities of the pars inter- media of gnathostomes and of the pars distalis of teleost fishes also are regulated commonly by direct innervation of secretory cells. The region of the hypothalamus where regulating hormones are released to portal vessels (in non-teleostean gnathostomes) or delivered directly to the pars distalis (in teleosts) is called the median eminence [1]. Therefore, the median eminence is considered as a device where external and internal environ- mental information integrated in the hypothala- mus is transferred to the central endocrine organ (adenohypophysis) in the form of neurosecretory regulating hormones. This information transfer system constitutes the most fundamental neuroen- docrine integration circuit of gnathostome verte- brates and enables them to adapt appropriately to changing environments. A related question of some interest is how such an information transfer system evolved in the ancestral vertebrates. Among living vertebrates, agnathan cyclostomes (hagfishes and lampreys) are the closest relatives to the ancestor of vertebrates Received June 1, 1987 ' Present address: Department of Biology, College of General Education, Osaka University, Toyonaka, Osaka 560, Japan. and therefore the hypothalamo-hypophysial region of these animals has been investigated by com- parative endocrinologists who anticipated that the hypothalamo-hypophysial system of cyclostomes might represent the most primitive condition of the central neuroendocrine circuit of vertebrates. The present article reviews recent advances made in the study of the hypothalamo-hypophysial system in cyclostomes, with special attention focused on the median eminence as the essential component of this system. In recent years, Gorbman [2, 3] also has reviewed the characteristics of the cyclo- stome hypothalamo-hypophysial system. Various problems related to this topic have been reviewed by Gorbman [4] (on reproduction), Nozaki [5] (on immunocytochemistry in brain and pituitary), and Chiba and Honma [6] (on the brain ventricular system). CHARACTERISTICS OF CYCLOSTOME NEUROHYPOPHYSIS Hagfish The diencephalic region of the hagfish is shown in Figure 1. The hypothalamo-hypophysial system is also illustrated diagrammatically in Figure 2. The neurohypophysis is a flattened sac connected to the hypothalamus by a short neurohypophysial stalk. The lumen of the sac reaches the hypothala- mic third ventricle. In the neurohypophysis, three regions may be distinguished topologically. The anterior dorsal wall is a region rostral to the stalk. The posterior dorsal wall is a region caudal to the 22 K. TSUNEKI np, nasopharyngeal duct; sco, subcommissural organ. Fic. 1. a. Sagittal section of the diencephalon of the hagfish, Eptatretus burgeri. Arrows indicate the neurohypophysial sac. The adenohypophysis (ah) is embedded in connective tissue. The left side is rostral. x33. b. Sagittal section of the diencephalon of the brook lamprey, Lampetra reissneri. Arrows indicate the neurohypophsis. In contrast to the hagfish, the ventricle is spacious. The left side is rostral. ah, adenohypophysis; cp, mesencephalic choroid plexus; nh, nasohypophysial duct. x43. dw Fic. 2. Diagrammatic illustration of the pituitary region of the hagfish (left) and the lamprey (right). Black area is stained densely with paraldehyde fuchsin and dotted area is stained moderately with paraldehyde fuchsin. ah, adenohypophysis; an, anterior neurohypophysis; cpd, caudal pars distalis; dw, dorsal wall of the neurohypophysis; pi, pars intermedia; pn, posterior neurohypophysis (pars nervosa); rpd, rostral pars distalis; s, stalk; t, third ventricle; vw, ventral wall of the neurohypophysis. stalk. The ventral wall constitutes the ventral region of the neurohypophysial sac. The wall of these three regions is rather thin and is composed of an ependymal layer and a subependymal nerve fiber layer. The fiber layer of the posterior dorsal wall is densely stained by classical neurosecretory stains such as Gomori’s paraldehyde fuchsin. The fiber layer of the anterior dorsal wall is moderately Neurohypophysis of Cyclostomes 23 stained with paraldehyde fuchsin and that of the ventral wall is not stained or only weakly and sporadically stained with paraldehyde fuchsin. Blood vessels are richly distributed around the dorsal surface of the posterior dorsal wall. The adenohypophysis is a flat disk composed of cell nests. It is located just under the neurohypo- physis, being separated from it by a rather thick layer of connective tissue (about 50 to 60 ~m in thickness in the Western Pacific hagfish, Eptatretus burgeri). The adenohypophysis does not differ- entiate a pars intermedia and most of its cells are chromophobic in traditional light microscopical stainings. The hagfish hypothalamo-hypophysial system does not appear to have a median eminence-like structure at first glance. Nevertheless, some authors have sought to identify a possible hagfish median eminence. Thus, Olsson [7] in his light microscopical study tentatively identified the me- dian eminence in the vascularized brain floor just rostral to the neurohypophysial stalk in the Atlan- tic hagfish, Myxine glutinosa. In the Eastern Pacific hagfish, Eptatretus stouti, Gorbman et al. [8] studied hypothalamo-hypophysial vasculariza- tion with an ink-injection method and found a prehypophysial vascular network (in the chiasma- tic region) which is drained by a portal system to the caudodorsal wall of the neurohypophysis. The significance of this peculiar portal system was unknown. There appeared to be only a few vertical blood vessels connecting the ventral wall of the neurohypophysis with the adenohypophysis [8]. Later, in the same species, Nishioka and Bern [9] studied fine structure of the vascularized chiasmatic region and considered this area as a possible median eminence, although they could not detect neurosecretory axon terminals. Under these circumstances, it is fair to say that the median eminence of M. glutinosa or of E. stouti could not be sufficiently characterized to be identified with certainty. In 1972, Kobayashi and Uemura [10] studied the ultrastructure of the neurohypophysial sac of E. burgeri and revealed many axon terminals contain- ing secretory granules in the ventral wall of the neurohypophysial sac. They also found about 20 vertical blood vessels connecting the ventral wall of the neurohypophysis with the adenohypophysis. The number of vertical blood vessels issuing from the ventral wall of the neurohypophysial sac appeared to be slightly higher in E. burgeri (a shallow-water inhabitant) than in E. stouti (a deep-water inhabitant). According to Kobayashi and Uemura [10], the axon terminals in the ventral wall of the neurohypophysis of E. burgeri could be classified into three types by the size of their granules (about 65, 80, and 110 nm in diameter). In the dorsal wall, axon terminals containing much larger granules were frequent. In gnathostomes, it was well known that the median eminence posses- ses axon terminals with granules predominantly smaller than 100 nm (probably containing regulat- ing hormones and/or monoamines) and the neural lobe (pars nervosa) possesses axon terminals with granules predominantly much larger than 100 nm (probably containing neural lobe hormones such as arginine vasotocin). Encouraged by _ those findings, Kobayashi and Uemura [10] concluded that the ventral wall of the neurohypophysis of the hagfish is a structure comparable to the median eminence of gnathostomes and the dorsal wall represents the neural lobe. In E. stouti, Hender- son [11] also revealed neurosecretory axon termi- nals in the ventral wall of the neurohypophysial sac. Following these electron microscopical works, it was histochemically demonstrated in E. burgeri that the fiber layer of the ventral wall contains acetylcholinesterase but that of the caudodorsal wall does not [12]. This histochemical study shows at least the difference in the neurotransmission mechanism between the ventral wall and the dorsal wall. Afterwards, with the aid of a computer, Tsuneki et al. [13] classified the secretory granules found in the ventral and dorsal walls of the neurohypophysis of FE. burgeri and substantiated the theory of Kobayashi and Uemura [10] (Fig. 3). The problem of whether the ventral wall of the hagfish neurohypophysis should be designated the median eminence depends on the definition of median eminence. If we adhere to the definition of the median eminence as a structure where portal vessels to the adenohypophysis originate, then the ventral wall of the hagfish neurohypophysis cannot be considered as the median eminence because of the absence of a typical portal system directed to 24 K. TSUNEKI Fic. 3. a. Outer layer of the dorsal wall of neurohypophysis of the hagfish, Eptatretus burgeri. 1, type 1 axon containing granules from 140 to 200 nm in diameter; 2, type 2 axon containing granules from 80 to 95 nm; cl, capillary lumen; ec, endothelial cell; f, fibroblast-like cell; pvs, perivascular space. 12,900. b. Outer layer of the ventral wall of the neurohypophysis of the hagfish, Eptatretus burgeri. Neurosecretory axon terminals contain granules of about 85 to 110 nm in diameter. These terminals do not directly contact the underlying connective tissue (ct), but are interposed by ependymal end feet (ef). Arrows indicate small vacuoles in the ependymal end feet. 24,000. the adenohypophysis. However, if the median _ the neural lobe terminate in the vicinity of the pars eminence is considered as the region where distalis, then the ventral wall of the hagfish neurosecretory axons other than those leaving for neurohypophysis should be termed the median Neurohypophysis of Cyclostomes Pes eminence. In any event, there is no doubt that the ventral wall of the hagfish neurohypophysis repre- sents a primitive median eminence-like structure, at least on anatomical grounds. If we do not consider the ventral wall of the neurohypophysial sac as the median eminence-equivalent region, then we cannot explain the existence of neurosecretory axon terminals in this region. These morphological findings on the hagfish hypothalamo-hypophysial system prompted func- tional studies. Since the adenohypophysis is sepa- rated from the ventral wall of the neurohypophysis by a rather thick and poorly vascularized connec- tive tissue, the problem arises of how the contents of the granules in the axon terminals in the ventral wall are delivered to the adenohypophysis. The most probable mechanism is by diffusion. Blood vessels located between the neurohypophysis and adenohypophysis may be involved in material transfer only to a limited extent because of their extremely small number compared to the median eminence of gnathostomes. The problem of diffu- sion in the hagfish was experimentally investigated by Kobayashi and his colleagues. In 1975, Nozaki et al. [14] found that intraventricularly injected peroxidase reaches the adenohypophysis through ependymal cells of the ventral wall of the neurohy- pophysis and the connective tissue layer. More recently, Tsukahara et al. [15] demonstrated that intraventricularly injected trypan blue reaches the connective tissue layer and ferric ions even pene- trate into the adenohypophysial cells within ten minutes after injection. These studies clearly show that at least some substances can reach the adenohypophysis through diffusion across the con- nective tissue layer between the neurohypophysis and the adenohypophysis. In M. glutinosa, European authors also have studied the relation between the neurohypophysis and the adenohypophysis. Fernholm [16] revealed a contact between the ventral wall of the neurohy- pophysis and the adenohypophysial tissue in 6.5% of the animals he examined. This anatomical relation is reminiscent of the close apposition between the neural lobe and the pars intermedia in lampreys and many gnathostomes. Schultz and Adam [17] and Schultz et al. [18] reported the occurrence of a few nerve terminals in the dorso- lateral part of the adenohypophysis. Therefore, direct information transfer through innervation may also operate in some individuals but this system apparently plays a minor role if any. Lametschwandtner [19] studied the vascularization in the hypothalamo-hypophysial region with a corrosion cast method and confirmed the earlier ink-injection study by Gorbman ef al. [8] that showed the absence of a typical portal system in E. stouti. Recent immunohistochemical studies on the hagfish hypothalamo-hypophysial region have re- sulted in some odd findings. Although immunohis- tochemical studies of the distribution of arginine vasotocin yielded the same results as classical Gomori’s staining [20, 21], the attempts to find immunoreactive LHRH, TRH, substance P, en- dorphin, a-MSH, and ACTH in tissue sections all failed to give positive results in the species so far studied [20-24]. With an immunological method, LHRH was detected in small amounts in extracts of the whole brain of the South African hagfish, Eptatretus hexatrema [27], but not in E. stouti [28]. Immunoreactive somatostatin does not exist in the neurohypophysis in EF. burgeri and E. stouti, although it is demonstrable in the brain [20, 21]. Synthetic TRH had no effect on thyroidal activities either in vivo [25] or in vitro [26]. Peculiarly enough, there are no reports of the distribution of monoamine fluorescence in the hagfish neurohy- pophysis, although monoamine oxidase was his- tochemically detected in the ventral as well as the dorsal wall of the neurohypophysis [12]. There- fore, ultrastructurally demonstrable secretory granules in the axon terminals in the ventral wall are totally unknown as to their contents except for a few large granules (about 160 to 200nm in diameter) probably containing arginine vasotocin. However, it is hardly conceivable that there are granules without any biologically active sub- stances. They may contain substances which do not cross-react with the antisera currently used or are simply insufficient in amount to be detected immunocytochemically. Gorbman [29, 30] reexamined the old Dean- Conel’s preparation of embryos of E. stouti and demonstrated that the hagfish adenohypophysis is not of ectodermal origin, but instead it is peculiarly 26 K. TSUNEKI of endodermal origin. However, no one could argue that the structure concerned is not the adenohypophysis (but see [31]). Although a FMRF-amide-like substance is the only substance so far successfully revealed immunocytochemically in the hagfish adenohypophysis [24], the adenohy- pophysis contains many secretory cells with elec- tron-dense granules although the number of gran- ules per cell is apparently small. In M. glutinosa, either two [32] or five [33] types of secretory cells have been demonstrated. In EF. burgeri, three types of secretory cells are found [34, 35]. Bioas- say experiments revealed ACTH-like and TSH- like activities in extracts of the pituitary of FE. stouti [36, 37]. Hypophysectomy did not cause a signif- icant change in thyroidal activity [38, 39], but it slightly interfered with the normal development of the gonad, especially in E. burgeri [40]. There- fore, the hagfish possesses a definite adenohy- pophysis, although it is apparently poorly dif- ferentiated both histologically and functionally. Lamprey The neurohypophysis of the lamprey forms the ventral floor of the third ventricle (Figs. 1 and 2). The posterior part of the neurohypophysis is densely stained with paraldehyde fuchsin except for ependymal cells. The ventral surface is covered with a capillary plexus and a pars intermedia. This portion of the neurohypophysis is frequently called the complete neurohypophysis, but actually it is a part of the neurohypophysis and corresponds to the neural lobe both topologically and in the paraldehyde fuchsin-stainability of nerve fibers which release their contents into the general circulation. The anterior part of the floor of the third ventricle also consists of an ependymal layer and a nerve fiber layer. This part of the neurohy- pophysis is usually called the infundibulum. It is separated from the underlying pars distalis by a layer of connective tissue which has a thickness of only about 2 to 8 ym in the non-parasitic brook lamprey, Lampetra (Lethenteron) reissneri. The rostral region of the anterior part of the neurohy- pophysis is moderately stained with paraldehyde fuchsin. Although the anterior part of the neurohypophysis is not vascularized [41], this region was compared to the median eminence of gnathostomes topologically [10, 42]. This assertion is rather natural, because the relation between the neurohypophysis and the adenohypophysis in lam- preys is anatomically similar to that in chondros- teans and holosteans in which the anterior part of the neurohypophysis forms an unequivocal median eminence [43]. The adenohypophysis of lampreys is composed of three regions; the rostral pars distalis, caudal (proximal) pars distalis, and pars intermedia. In contrast to the hagfish, there are many chromophilic cells, especially in the rostral pars distalis and pars intermedia. The ultrastructure of the anterior part of the neurohypophysis (anterior neurohypophysis) was first studied in Lampetra (Entosphenus) tridentata [44]. This region contains many neurosecretory axon terminals. The predominant types of granules in the axons are 65 to 100, 95 to 140, and 140 to 220 nm in diameter, respectively. The axon terminals with the smaller granules are more frequently encountered than in the posterior neurohypoph- ysis (neural lobe) where the predominant type of granules is about 160 to 200 nm in diameter. The axon terminals in the rostral region of the anterior neurohypophysis frequently abut on the basal lamina of connective tissue, while those in the caudal part of the anterior neurohypophysis mostly end on the intervening ependymal end feet. In the parasitic river lamprey, Lampetra (Lethenteron) japonica, neurosecretory axon terminals occa- sionally abut directly on the basal lamina of connective tissue even in the caudal part of the anterior neurohypophysis (unpublished). In the ventral wall of the neurohypophysis of the hagfish, ependymal end feet usually intervene between neurosecretory axon terminals and the basal lami- na of connective tissue. (See Oota et al. [45] for the significance of ependymal intervention in release of neurosecretory materials.) In Lampetra (Lampetra) fluviatilis, Belenky et al. [46] also studied the ultrastructure of the anterior neurohy- pophysis (the “proximal neurosecretory contact region” in their terminology) and obtained essen- tially the same results as in L. tridentata. These ultrastructural characteristics of the lamprey ante- rior neurohypophysis may offer support for the earlier suggestion that this region might be the lamprey median eminence (Fig. 4). The surface Neurohypophysis of Cyclostomes 2] es. Fic. 4. a. ese ¥; Outer layer of the rostral part of the anterior neurohypophysis of the river lamprey, Lampetra Japonica. Axon terminal | contains granules from 120 to 150 nm and axon terminal 2 contains granules from 110 to 130 nm in addition to numerous synaptic vesicles. These axon terminals directly abut on the basal lamina of connective tissue (ct). ep, ependymal process. x 20,300. b. Outer layer of the caudal part of the anterior neurohypophysis of the lamprey, Lampetra japonica. Axon terminal 1 contains granules from 130 to 170 nm and axon terminal 2 contains granules from 85 to 110 nm. In the area shown, only ependymal end feet (ef) abut on the basal lamina of connective tissue (ct). fine structure of ependymal cells of the anterior neurohypophysis of the lamprey also differs from that of the neural lobe in the degree of ciliation [47]. The substances contained in the granules in the axon terminals of the anterior neurohypoph- ysis may reach the adenohypophysis by diffusion through the thin connective tissue sheet, because x 20,300. virtually no blood vessels exist between the ante- rior neurohypophysis and the pars distalis. Ultra- structural studies also did not reveal nerve fibers in the pars distalis [44] or the pars intermedia in L. tridentata and L. fluviatilis [48, 49]. In L. japonica, bundles of nerve fibers are found close to the pars intermedia, under the capillary plexus, but their 28 K. TSUNEKI destination is not clear (unpublished). They are apparently rare and may play a minor role, if any, in the regulation of pars intermedia activity. In lampreys, immunohistochemical distributions of biologically active substances in the hypothala- mo-hypophysial region have been rather exten- sively studied. The neural lobe contains im- munoreactive LHRH (in L. japonica and L. tridentata), met-enkephalin (in L. tridentata), growth hormone (in Petromyzon marinus), prolac- tin (in P. marinus), as well as arginine vasotocin (in L. fluviatilis) (22, 23, 50-52]. The rostral part of the anterior neurohypophysis contains im- munoreactive arginine vasotocin (in L. fluviatilis), LHRH (in L. japonica and P. marinus), enkepha- lin (in Lampetra (Lethenteron) lamottenii), and substance P (in P. marinus) [23, 50, 53, 54]. The caudal part of the anterior neurohypophysis con- tains immunoreactive arginine vasotocin (in L. fluviatilis), LHRH (in L. japonica and P. mari- nus), and enkephalin (in L. lamottenii) [23, 50, 54]. The LHRH system is only faintly stained immunohistochemically in the ammocoetes larvae of Lampetra (Lampetra) richardsoni [55]. TRH and LHRH were also detected immunologically in the lamprey brain [28, 56]. However, the amino acid sequence of lamprey LHRH appears to be different from that of mammalian LHRH [28]. Monoamine fluorescence occurs in the anterior neurohypophysis as well as in the neural lobe [57, 58]. Although the intensity of fluorescence is stronger in the neural lobe, the histochemical activity of monoamine oxidase is stronger in the anterior neurohypophysis than in the neural lobe [59]. The presence of monoamines in the anterior neurohypophysis was further confirmed by auto- radiographical study with *H-dopamine [58]. Therefore, the median eminence-equivalent ante- rior neurohypophysis of lampreys is an active secretory organ at least in adults, as determined not only ultrastructurally but also immunohis- tochemically and by routine histochemistry. Somatostatin is not detected in the neurohy- pophysis, but in P. marinus somatostatin- immunoreactive neurons are abundant in the ventral hypothalamus and they are _ liquor- contacting neurons [52]. (See [6] and papers cited therein for liquor-contacting neurons in lampreys.) Based on this immunocytochemical observation, Wright [52] suggested the possibility that somato- statin is released into the third ventricle, is absorbed by ependymal cells, and reaches the pars distalis by diffusion through connective tissue. The differential distribution of secretory cells in the dorso-ventral direction in the rostral pars distalis in lampreys further supports the possibility of diffu- sion of hypothalamic substances [44]. Ex- perimental studies on the hypothalamic regulation of adenohypophysial activity are scarce. Although the apparently normal functioning of heterotopi- cally implanted pituitary in L. fluviatilis [60] appears discouraging, the effectiveness of exoge- nously administered LHRH in accelerating ovula- tion in P. marinus [61] may be encouraging for future research. The highly secretory nature of the lamprey adenohypophysis is beyond doubt. Ultrastructural- ly, it contains several types of secretory cells with abundant granules [61-63]. Some cell types also possess abundant secretory granules even in larvae [63, 64]. Hypophysectomy experiments also re- vealed that the adenohypophysis is involved in the induction of secondary sex characters, spawning, color change, metamorphosis, and so on (reviewed in [60, 65], see also [66]). Immunocytochemical identification of secretory cell types is currently being actively pursued in P. marinus by Wright [52, 67, 68]. This topic is beyond the scope of this review and here it may be sufficient to note that there are TSH-like cells in the rostral pars distalis and LH-like, GH-like, and prolactin-like cells in the caudal pars distalis of the upstream migrants. Nozaki [5] summarized their studies on the im- munocytochemical distribution of proopiomelano- cortin-related peptides such as_ enkephalin, ACTH, and MSH in the lamprey adenohypophysis as well as other substances. According to him, there is some difference between L. tridentata and P. marinus and the interpretation of results are complicated although there are stainable cells. Proopiomelanocortin-related peptides also occur in the adenohypophysis of L. lamottenii [54]. In Lampetra (Okkelbergia) aepyptera and L. fluviati- lis, bioassay experiments demonstrated ACTH- like activity in extracts of the lamprey adenohy- pophysis [69, 70]. Neurohypophysis of Cyclostomes 29 These observations taken together suggest that the lamprey hypothalamo-hypophysial system is highly active compared to that of the hagfish and rather resembles that of gnathostomes. However, it must be emphasized that the lamprey anterior neurohypophysis (median eminence-equivalent re- gion) is not supplied with portal vessels or any other kind of significant blood vessels. EVOLUTIONARY PERSPECTIVES As reviewed in the previous section, the median eminence-equivalent structure has been revealed ultrastructurally both in the hagfish and lamprey. In lampreys, immunohistochemical studies also demonstrated the presence of hypothalamic reg- ulating hormones in this region. The adenohy- pophysis of both hagfish and lamprey is a secretory organ at least ultrastructurally. In lampreys, the adenohypophysis appears to be active also func- tionally. Therefore, the neurohypophysis- adenohypophysis relation seen in cyclostomes is fundamentally similar to that seen in gnathos- tomes. The most significant difference is the absence of portal circulation in cyclostomes. The virtually avascular or at most only sporadically vascularized median eminence-like organ of cy- clostomes may represent the most ancient type of vertebrate median eminence and the neurosecre- tory regulating hormones might reach the ade- nohypophysis by diffusion at the stage before the development of portal vessels and direct innerva- tion. It is unlikely that the condition in cyclostomes was brought about by degeneration from the stage where the neurohypophysis and adenohypophysis were intimately linked anatomically. Lampreys are fundamentally anadromous animals and repro- duce only once at the end of their life cycle. Hagfishes are purely marine animals and they reproduce more than once in their life cycle. In spite of these basic differences in life strategy, the fundamental relation between the neurohypoph- ysis and adenohypophysis is the same, that is, the common absence of typical portal vessels, in lampreys and hagfishes. Therefore, this condition is better interpreted as primitive rather than degenerate. Immunohistochemical failure in de- tecting biologically active substances in the hagfish hypothalamo-hypophysial region is usually ex- plained by degenerative condition of these animals living in the deep sea under constant temperature and continuous darkness. However, it is zoological nonsense to claim that the hagfish is more degener- ate than protochordates (see below). The basic hagfish should be sought in shallow water species such as E. burgeri. M. glutinosa is apparently degenerate compared to E. burgeri. E. burgeriisa strictly seasonal breeder [71-73] and shows a distinct nocturnal activity rhythm [74, 75]. They are highly voracious and active animals, and do not look degenerate although they are certainly spe- cialized to some extent in their burrowing habit. The apparent poor development of the hagfish hypothalamo-hypophysial system may be ex- plained not by degeneration, but by primitiveness associated with some specialization. It is a current popular belief that the lamprey and hagfish are very different animals with totally different lineages and the difference in the hypothalamo-hypophysial system is taken as one piece of evidence supporting this interpretation (see [76] for references). In a variety of animal taxa, the primitive group frequently shows greater morphological diversity than the advanced group, but such diversity should not be necessarily taken as the evidence of di- or polyphyletic origin of the primitive group. The old concept that the cyclo- stomes constitute a monophyletic group should be reexamined [77], although the virtually avascular median eminence apparently represents symple- siomorphy. Finally, it may be asked how the hypothalamo- hypophysial relations in vertebrates originated. Two possibilities may exist; first, the hypothalamo- hypophysial system evolved as an entirety, and second, the pituitary independently developed and later it was incorporated under hypothalamic influence. As far as the embryologically intimate relation between the hypothalamus and the pitui- tary is concerned, the first possibility seems to be more likely. The embryological relation between these two organs may be more profound than usually considered, because the hypothalamo- hypophysial unit actually appears as a real unit from the neuroectodermal ventral neural ridge in the chick [78]. The development of the unsepara- 30 K. TSUNEKI ble hypothalamo-hypophysial system as a central information transfer system might be one of the factors that endowed the original vertebrates with the great possibility for future development and more elaborate refinement of this system might have opened the way for the original gnathostomes to vast prosperity. However, it is unknown how the relation between the central nervous system and the central endocrine organ developed during the transition from the invertebrate level of body organization to the vertebrate level of organiza- tion. A possible clue might be sought in extant protochordates, but the evidence is equivocal. In amphioxus, the epithelial cells of Hatschek’s pit, a possible homologue of the vertebrate ade- nohypophysis, possess secretory granules [79]. Hatschek’s pit also is claimed to contain im- munoreactive LH [80]. It originates from the preoral pit of larvae, although the real germinal origin of Hatschek’s pit itself is unclear. In adults, Hatschek’s pit is located below the notochord on the right side in the oral vestibule and is rather remote from the brain vesicle although the ventral area of the brain vesicle contains neurosecretory axon terminals [81]. A detailed study of the vascular system around the brain vesicle and Hatschek’s pit is not available, but the small size of amphioxus might make diffusion workable in transporting material as well as the involvement of blood vessels. Nerve fibers may occur in Hats- chek’s pit but the terminals have not been conclu- sively demonstrated [79, 82]. In ascidians, both the neural gland and the cerebral ganglion appear to develop from the neural tube of larvae during metamorphosis [83]. Therefore, they are probably ectodermal in origin and develop together as the hypothalamo- hypophysial unit of the chick. In adults, the neural gland directly abuts on the cerebral ganglion either dorsally or ventrally. In ascidians, the circulatory system is of the open type and diffusion across sparse connective tissue may be important in transporting materials. Recent immunocyto- chemical studies have demonstrated various bio- logically active substances, including LHRH and somatostatin, in the cerebral ganglion of ascidians [84, 85]. Pestarino reported in a series of papers [86-88] that the neural gland of Styela plicata contains immunoreactive ACTH, MSH, prolactin, and so on. Therefore, the cerebral ganglion-neural gland complex of ascidians appears to be more developed. immunocytochemically than the hypothalamo-hypophysial system of the hagfish. However, the presence of various biologically active substances in the ascidian neural gland is somewhat puzzling, because the cells of the neural gland do not contain secretory granules at the ultrastructural level at least in Ciona intestinalis [89]. These studies in protochordates are intriguing in themselves, but do not necessarily solve the problem of the origin of the hypothalamo- hypophysial system of vertebrates. Neither asci- dians nor amphioxus (Acrania) possess a head. The head supplied with several sensory organs and a complex central nervous system is a prerequisite for the emergence of active animals such as vertebrates. The development of the hypothala- mo-hypophysial system as a central information transfer system must have waited for the develop- ment of a complex head during the transition from invertebrates to vertebrates. ACKNOWLEDGMENT I would like to express my sincere thanks to Prof. H. Kobayashi for his reading of the manuscript. My works cited in this review indeed were carried out under his guidance. I would like to thank also Dr. S. R. Vigna for reading the manuscript and Prof. M. Ouji for his continuous advice and encouragement. REFERENCES 1 Kobayashi, H., Matsui, T. and Ishii, S. (1970) Int. Rev. Cytol., 29: 281-381. 2 Gorbman, A. (1980) Can. J. Fish. Aquat. Sci., 37: 1680-1686. 3 Gorbman, A. (1980) In “Hormones, Adaptation and Evolution”. Ed. by S. Ishii, T. Hirano and M. Wada, Japan Sci. Soc. Press, Tokyo, pp. 81-92. 4 Gorbman, A. (1983) In “Fish Physiology, IX-A”. Ed. by W.S.Hoar, D.J. Randall and E.M. Donaldson, Academic Press, Orlando, pp. 1-29. 5 Nozaki, M. (1985) In “Evolutionary Biology of Primitive Fishes”. Ed. by R. E. Foreman, A. Gorb- man, J.M. Dodd and R. Olsson, Plenum Press, New York, pp. 433-454. 6 Chiba, A. and Honma, Y. (1986) In “Indo-Pacific 24 25 26 2 28 29 30 31 Neurohypophysis of Cyclostomes 31 Fish Biology”. Ed. by T. Uyeno, R. Arai, T. Taniuchi and K. Matsuura, Ichthyol. Soc. of Japan, Tokyo, pp. 76-85. Olsson, R. (1959) Z. Zellforsch., 51: 97-107. Gorbman, A., Kobayashi,H. and Uemura, H. (1963) Gen. Comp. Endocrinol., 3: 505-514. Nishioka, R.S. and Bern,H.A. (1966) Comp. Endocrinol., 7: 457-462. Kobayashi, H. and Uemura, H. (1972) Gen. Comp. Endocrinol., Suppl. 3: 114-124. Henderson, N. E. (1972) Acta Zool. (Stockh.), 53: 243-266. Tsuneki, K., Urano, A. and Kobayashi, H. (1974) Gen. Comp. Endocrinol., 24: 249-256. Tsuneki, K., Adachi, T., Ishii,S. and Oota, Y. (1976) Cell Tissue Res., 166: 145-157. Nozaki,M., Fernholm,B. and _ Kobayashi, H. (1975) Acta Zool. (Stockh.), 56: 265-269. Tsukahara, T., Gorbman, A. and Kobayashi, H. (1986) Gen. Comp. Endocrinol., 61: 348-354. Fernholm, B. (1972) Gen. Comp. Endocrinol., Suppl. 3: 1-10. Schultz, H. J. and Adam, H. (1975) Norw. J. Zool., 23: 297-306. Schultz, H. J., Patzner, R. A. and Adam, H. (1979) Cell Tissue Res., 204: 67-75. Lametschwandtner, A. (1982) Mikroskopie (Wien), 39: 35-42. Nozaki, M. and Gorbman, A. (1983) Cell Tissue Res., 229: 541-550. Tsukahara, T., Shioda, S., Nakai, Y. and Kobayashi, H. (1983) Zool. Mag. (Tokyo), 92: 386- 392. Crim, J. W., Urano, A. and Gorbman, A. (1979) Gen. Comp. Endocrinol., 37: 294-305. Nozaki, M. and Kobayashi, H. (1979) Arch. Histol. Jpn., 42: 201-219. Jirikowski, G., Erhart, G., Grimmelikhuijzen, C. J. P., Triepel, J. and Patzner, R. A. (1984) Cell Tissue Res., 237: 363-366. Tsuneki, K. and Fernholm, B. (1975) Acta Zool. (Stockh.), 56: 61-65. Dickhoff, W. W., Crim, J.W. and Gorbman, A. (1978) Gen. Comp. Endocrinol., 35: 96-98. King, J. A. and Millar, R. P. (1980) Endocrinology, 106: 707-717. Sherwood,N.M. and Neuropeptides, 6: 205-214. Gorbman, A. (1983) Am. Zool., 23: 639-654. Gorbman, A. and Tamarin, A. (1985) In “Evolu- tionary Biology of Primitive Fishes”. Ed. by R. E. Foreman, A. Gorbman, J. M. Dodd and R. Olsson, Plenum Press, New York, pp. 165-185. Dodd, J. M. and Dodd, M. H. I. (1985) In “Evolu- tionary Biology of Primitive Fishes”. Ed. by R. E. Gen. Sower,S. A. — (1985) 32 33 34 35 36 37 38 39 40 Di 58 Foreman, A. Gorbman, J. M. Dodd and R. Olsson, Plenum Press, New York, pp. 295-319. Fernholm, B. (1972) Z. Zellforsch., 132: 451-472. Patzner, R. A., Erhart,G. and Adam, H. (1982) Cell Tissue Res., 223: 583-592. Oota, Y. (1974) Rep. Fac. Sci., Shizuoka Univ., 9: 67-78. Tsuneki, K. (1976) Acta Zool. (Stockh.), 57: 137- 146. Buckingham, J.C., Leach, J.H., Plisetskaya, E., Sower, S. A. and Gorbman, A. (1985) Gen. Comp. Endocrinol., 57: 434-437. Dickhoff, W. W. and Gorbman, A. (1977) Gen. Comp. Endocrinol., 31: 75-79. Henderson, N. E. (1976) Can. J. Zool., 54: 180- 184. Matty, A. J., Tsuneki, K., Dickhoff, W. W. and Gorbman, A. (1976) Gen. Comp. Endocrinol., 30: 500-516. Patzner, R. A. and Ichikawa, T. (1977) Zool. Anz. (Jena), 199: 371-380. Gorbman, A. (1965) Arch. Anat. Microsc. Mor- phol. Exp., 54: 163-194. Ruthle, H.-J. and Sterba, G. (1966) Z. Zellforsch., 70: 136-168. Hansen, G. N. (1971) Kong. Dans. Vidensk. Selsk. Biol. Skr., 18: 1-64. Tsuneki, K. and Gorbman, A. (1975) Gen. Comp. Endocrinol., 25: 487-508. Oota, Y., Kobayashi, H., Nishioka, R. S. and Bern, H. A. (1974) Neuroendocrinology, 16: 127-136. Belenky, M. A., Konstantinova,M.S. and Pole- nov, A. L. (1979) Cell Tissue Res., 204: 319-331. Shioda, S., Honma, Y., Yoshie, S. and Hosoya, Y. (1977) Arch. Histol. Jpn., 40: 41-49. Tsuneki, K. and Gorbman, A. (1975) Cell Tissue Res., 157: 165-184. Polenov, A. L., Belenky, M. A. and Konstantino- va, M.S. (1974) Cell Tissue Res., 150: 505-519. Goossens, N., Dierickx,K. and Vandesande, F. (1977) Cell Tissue Res., 177: 317-323. Nozaki, M. and Gorbman, A. (1984) Gen. Comp. Endocrinol., 53: 335-352. Wright, G. M. (1986) Cell Tissue Res., 246: 23-31. Nozaki, M. and Gorbman, A. (1986) Gen. Comp. Endocrinol., 62: 217-229. Dores, R. M., Finger, T. E. and Gold, M. R. (1984) Cell Tissue Res., 235: 107-115. Crim, J. W., Urano, A. and Gorbman, A. (1979) Gen. Comp. Endocrinol., 38: 290-299. Jackson, I. M. D. and Reichlin, S. (1974) Endocri- nology, 95: 854-862. Tsuneki, K., Kobayashi, H., Yanagisawa, M. and Bando, T. (1975) Cell Tissue Res., 161: 25-32. Belenky, M. A., Chetverukhin, V. K. and Polenov, 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 32 A. L. (1979) Cell Tissue Res., 204: 333-342. Tsuneki, K. (1974) Cell Tissue Res., 154: 17-27. Larsen, L.O. and Rothwell, B. (1972) In “The Biology of Lampreys, 2”. Ed. by M. W. Hardisty and I. C. Potter, Academic Press, London, pp. 1- 67. Sower,S.A., Dickhoff,W.W., Gorbman, A., Rivier, J. E. and Vale, W. W. (1983) Can. J. Zool., 61: 2653-2659. Bage,G. and Fernholm, B. (1975) (Stockh.), 56: 95-118. Percy, R., Leatherland, J. F. and Beamish, F. W. H. (1975) Cell Tissue Res., 157: 141-164. Wright, G. M. (1983) J. Morphol., 176: 325-339. Hardisty, M. W. and Baker, B. I. (1982) In “The Biology of Lampreys, 4B”. Ed. by M. W. Hardisty and I. C. Potter, Acacemic Press, London, pp. 1- 115. Joss, J. M. P. (1985) Gen. Comp. Endocrinol., 60: 58-62. Wright, G. M. (1983) Cell Tissue Res., 230: 225- 228. Wright, G. M. (1984) Gen. Comp. Endocrinol., 55: 269-274. Eastman, J.T. and Portanova, R. (1982) Comp. Endocrinol., 47: 346-350. Baker, B. I. and Buckingham, J.C. (1983) Gen. Comp. Endocrinol., 52: 283-290. Kobayashi, H., Ichikawa, T., Suzuki, H. and Seki- moto, M. (1972) Jpn. J. Ichthyol., 19: 191-194. Patzner, R. A. (1978) Acta Zool. (Stockh.), 59: 57— 61. Tsuneki, K., Ouji, M. and Saito, H. (1983) Jpn. J. Acta Zool. Gen. K. TSUNEKI Ichthyol., 29: 429-440. Fernholm, B. (1974) Mar. Biol., 27: 351-356. Ooka-Souda, S., Kabasawa, H. and Kinoshita, S. (1985) Zool. Sci., 2: 749-754. Hardisty, M. W. (1982) In “The Biology of Lam- preys, 4B”. Ed. by M. W. Hardisty and I. C. Potter, Academic Press, London, pp. 165-259. Yalden, D. W. (1985) Zool. J. Linn. Soc., 84: 291- 300. Takor,T.T. and Pearse, A. G. E. Embryol. Exp. Morphol., 34: 311-325. Tjoa, L. T. and Welsch, U. (1974) Cell Tissue Res., 153: 175-187. Zhiyi, Z., Yitao, Z. and Dayuan, C. (1982) Kexue Tongbao, 27: 1233-1234. Obermiller-Wilén, H. (Stockh.), 60: 187-196. Welsch, L.T. and Welsch, U. (1978) Zool. Jb. Anat., 100: 564-578. Elwyn, A. (1937) Bull. Neurol. Inst. N. Y., 6: 163- 177. Georges, D. and Dubois, M. P. (1980) C. R. Acad. Sci. Paris, D290: 29-31. Fritsch, H. A. R., Van Noorden, S. and Pearse, A. G. E. (1982) Cell Tissue Res., 223: 369-402. Pestarino, M. (1984) Gen. Comp. Endocrinol., 54: 444-449, Pestarino, M. (1985) Cell Tissue Res., 240: 497- 500. Pestarino, M. (1985) Gen. Comp. Endocrinol., 60: 293-297. Lane, N. J. (1971) Z. Zellforsch., 120: 80-93. (1975) J. (1979) Acta Zool. ZOOLOGICAL SCIENCE 5: 33-42 (1988) Hypersensitivity after Offset of Adapting Light in Vertebrate Photoreceptors Katsu AZUMA Department of Biology, Osaka Medical College, Sawaragicho 2-41, Takatsuki, Osaka 569, Japan ABSTRACT— Bullfrog and albino rat retinas were superfused with physiological solutions containing chemicals in order to suppress slow component (slow PIII) of distal PIII. Twenty uM of [Ba**] in the superfusate suppressed the slow PIII of frog, but concentration of higher than 500 “M was required to suppress that of rat. A gliotoxin, L-a aminoadipate (L-aAA), abolished the b-waves of both frog and rat and the slow PIII of frog. However, the gliotoxin did not suppress the slow PIII of rat. Measurements of the fast PIII isolated by these chemicals and of the light-induced decrease of [K*], were carried out in frog retina with Ag/AgCl electrodes and K *-selective microelectrodes, respective- ly . At 0.09 mM[Ca?*], (low Ca) or in the presence of 50 «M IBMX (phosphodiesterase inhibitor), the amplitude of fast PIII in the early dark after the offset of adaptation light was larger than that in the initial dark (called hypersensitivity). The hypersensitivity observed at low Ca was suppressed by reducing [Na ],. The low Ca or IBMX induced a remarkable overshoot of [K*], at the offset of the adaptation light. A decrease of [K*], induced by a light flash was remarkable in the dark after the offset of the adaptation light at low Ca or in the presence of IBMX. The hypersensitivity of rat photoreceptors was investigated by measuring distal PIII or fast PIII isolated by a capacity-coupled method. It was © 1988 Zoological Society of Japan confirmed that low Ca or IBMX could induce the hypersensitivity of photoreceptors of rat. INTRODUCTION It has been generally accepted that visual sensi- tivity decreases in light (light adaptation) and subsequently recovers in the dark (dark adapta- tion) in photoreceptors of both invertebrates and vertebrates. Several authors, however, have re- ported an unusual adaptive property of inverte- brate photoreceptors, 1. e., an enhancement of sensitivity after the offset of adapting light. This phenomenon has been named facilitation [1, 2]. On the other hand, no similar phenomenon has been found at normal condition in vertebrates. However, when an isolated frog retina was super- fused with low Ca solution, the photoreceptors exhibited an increase of sensitivity in the dark after the offset of adaptation light. This phenomenon has been called “hypersensitivity” by Azuma and Azuma [3]. On the basis of experiments using Accepted June 19, 1987 Received April 11, 1987 stimuli and adapting lights of different wavelengths, Azuma and Azuma [3, 4] have suggested that the hypersensitivity is mainly due to red rods. A phenomenon like hypersensitivity was briefly reported in an isolated skate retina [5] and confirmed in a frog retina [6]. Furthermore we found that IBMX could induce hypersensitivity even in normal [Ca**], [4]. As these inhibitors have been reported to change the response prop- erties of photoreceptors [7], studies on hypersensi- tivity may give useful information about visual excitation and adaptation. In the previous experiments [3, 4], we measured the transretinal potential of the frog by the following methods and assumed it as a receptor potential. The distal PIII was separated by an aspartate treatment from frog ERG [8, 9], led off by means of extracellular Ag/AgCl electrodes and capacity-coupled with 0.3-1 sec time constant in order to eliminate the slow PIII which originated in Miller cells [10, 11]. Recently it has been reported that the slow PIII is suppressed in the 34 K. AZUMA presence of Ba’t [12, 13]. The suppression of slow PIII by Ba’* is due to its blocking effect on the gx(K*-conductance) of Miller cells [13]. On the other hand, several authors have reported of the gliotoxic effects of L-a aminoadipate (L-aAA). When the reagent was applied to the isolated retina of skate [14] or, was intravitreously injected to the eyes of frog [15], chicken [15, 16] and rat [17], the loss of ERG (b-waves) with the extensive damage of Miller cell structures was observed. Therefore, in this study, we tried to eliminate the slow PIII of frogs and of albino rats by the treatments of Ba*t and L-eAA instead of capac- ity-coupled method [3, 4], and reexamined the phenomenon of hypersensitivity. In this paper, we have also examined the effects of [Na*], on the hypersensitivity, and measured [K*], in the interstitial space of frog retina as the first step to elucidate ionic mechanism of hypersen- sitivity. Furthermore, we have used albino rat retinas whose photoreceptors are mainly rods. It is interesting whether hypersensitivity is obseved or not in mammalian retina. All the present results support our hypothesis that hypersensitivity is caused by the increase of Na*-gradient across rod photoreceptor membrane [3]. MATERIALS AND METHODS Measurements of transretinal potentials of frog and rat Adult bullfrogs (Rana catesbeiana) and albino rats (100-150 g) were dark adapted overnight. The frogs were pithed and the rats were anesthe- tized with diethylether. The eye balls of these animals were enucleated under dim red light. In order to measure transretinal photoreceptor potential, retinas were isolated from the eye balls. The isolated retina was spread on a piece of nylon mesh and inserted into a superfusion chamber (a gift from Prof. W. Sickel of Cologne Univ.). Transretinal photoresponses were led off by means of Ag/AgCl electrodes embedded in the chamber and amplified by a FET instrumentation amplifier (Teledyne Philbrick 4253). Measurements of [K*], of frog retina In order to measure the interstitial [K*], of the frog isolated retina, it was fixed receptor-side up in a small chamber that had a transparent bottom. This chamber was set on the stage of an inverted microscope. The interstitial [K*], was measured by K*-selective double-barrelled microelectrodes and a differential electrometer (WPI FD -223). The microelectrode was prepared following the method described by Fujimoto and Kubota [18]. One barrel was an ion-selective electrode and contained a K*-selective liquid exchanger (WPI, IE-190) in the tip, and the remainder of this barrel was filled with 1M KCl. The other barrel was a reference electrode filled with 1M LiCl. The K*-selective microelectrode was advanced toward the receptor surface of the retina from above by using a hydraulic microdrive, until the electrode tip touched the surface. The electrode then was advanced in a few mum steps, and its tip was positioned at the depth where the amplitude of the light-evoked decrease in [K*], was maximal. Immediately after an experiment, the K * -selective microelectrode was calibrated using solutions hav- ing various [K*] and a fixed background of 110 mM [Na‘*]. Using the calibration data and the log-linear, least-squared error regression analysis [19], parameters (A and S) of the following equation were obtained: VK+=A: logio([K*],+[Na‘]./S)+V. where Vx+ is the differential potential between the two barrels (K *-selective barrel positive), A is the logarithmic slope and S is the selectivity coefficient for K* over Na‘, and V2o is a constant. The value of A was 55-58 mV/decade, and S was about 50. Vx+ referenced to an Ag/AgCl electrode was displayed on an oscilloscope (Sony/Tektronix Corp. 5103N) and recorded permanently on a pen recorder (TOA Electronics LTD., FBR-252A) through a transient memory (Kawasaki Electronica Co., Ltd., HR-1200). Solutions of superfusion (superfusates) The superfusate of frog retina (pH 7.6) con- tained 80mM NaCl, 2.5mM KCl, 25mM NaHCO;, 1.2 mM MgCl, 25 mM glucose, 3 mM HEPES and an appropriate concentration of Hypersensitivity of Photoreceptors 35 CaCl. The superfusate of rat retina was almost the same as that of the frog retina, except for the concentration of NaCl (110 mM) and KCI (5 mM). A low [Na*] superfusate was prepared by the equimolar substitution of choline-chloride instead of NaCl. In order to suppress slow PIII, L-aAA or BaCl, with Na-aspartate was used. The superfu- sate was run off at the rate of 1 ml/min. The temperature of the superfusate was maintained at 21+1°C for frogs and at 25+1°C for rats. The procedures of stimulations of retinas were previously described in detail [3]. The unattenu- ated intensity of the stimulus light (500 nm) was 6.3X10~7 W/cm’. RESULTS Suppression of slow PIII by chemicals Figure 1 shows the effects of [Ba*t], on frog distal PIII. The superposed photoresponses drawn in (a) were recorded after 10 min of superfusion 100uVv 5sec Fic. 1. The effect of [Ba’*], on frog distal PHI. The superposed responses, drawn in (a), were elicited by 0.5 sec-stimuli at four different intensities with 2 min-interval after 10 min of the superfusion with the physiological solution containing 5mM Na- aspartate. After recording these, the superfusate was changed to the one that contained 5 mM Na- aspartate plus 10 ~M Ba*+. The responses drawn by the continuous line in (b), were obtained by the same stimuli as in the case of {a) after 10 min of the superfusion. The responses drawn by dotted lines indicate those obtained after 10 min of further superfusion with the solution that contained 5 mM Na-aspartate plus 20 .M Ba**+. The numeral on each response indicates the stimulus intensity ex- pressed by the negative logarithm of the (neutral density) filter. The downward deflection means that ganglion side of the retina becomes negative with respect to the receptor side. with the physiological solution containing 5mM Na-aspartate. Each response contained a remark- ably slow component (slow PIII). After recording these responses, the superfusate of the retina was switched to the one that contained 5 mM Na- aspartate plus 10 ~M Ba’. As shown in (b), the superfusion of 10 min with the Ba*t solution suppressed slow PIII component (continuous lines) and the further superfusion of 10 min with the one that contained 5 mM Na-aspartate plus 20 uM Ba*t was more effective (dotted line). Thus [Ba’*], higher than 20 «M could suppress the slow PIII of the frog distal PIII. On the other hand, the slow PIII of rat was barely suppressed by Ba?* of concentration higher than 500 “M (data not shown). Figure 2 shows the effect of L-aAA on frog distal PIII. These superposed responses were elicited by the same stimuli as in Figure 1 after 10 min (a) and 120 min (b) of the superfusion with the physiological solution containing 10 mM L-aAA instead of 5 mM Na-aspartate. The recovery phase of each response elicited by an identical stimulus was more rapid in (b) than in (a). Thus 10 mM L- aAA abolished frog b-wave within 10 min, and completed the suppression of the slow PIII within 120 min. However, concentration higher than 10 mM L-aAA solution hardly helped to shorten the superfusion time necessary for suppression of the slow PIII. L-aAA at 2 mM abolished frog b-wave, but not the slow PIII. In the case of rat retina, L- aAA at 10 mM could also suppress b-wave, but not (a) (b) 100uVv mercer : _ io — Ssec Fic. 2. The effect of L-eAA on frog distal PII]. The superposed responses were obtained by the same stimuli as in Fig. 1 at 10 min (a) and at 120 min (b) after the superfusion with the physiological solution containing 10 mM L-aAA. The downward deflec- tion of responses and the numeral on each response mean the same as in Fig. 1. 36 K. AZUMA slow PIII. The isomer, D-aAA, suppressed neither the b-wave nor the slow PIII of both the animals. Hereafter, the residual fast component of the distal PIII of frog after treatment with the chemi- cals will be called fast PIII which is assumed to be the response of receptor origin. Hypersensitivity by measuring chemicals-isolated fast PIII in frog photoreceptors Figure 3 shows the effects of [Ca**], and light ; : 50uv t i i i yer | oe Gee oe ee ee LA —_ —L 1 pt (b) | Pe ae Ale Hae ied, 100uV ti ae 2 ry oe ee ee LA —L1_ pred ca ES Da Les La Hee Fic. 3. The effects of [Ca?*], and light adaptation on the amplitude of L-aAA-isolated fast PIII. The records in (a) were obtained after 120 min of the superfusion with the physiological solution contain- ing 10 mM L-aAA and 0.9 mM Ca’** (control solu- tion). After recording, the superfusate was changed to the one that contained 10 mM L-aAA and 0.09mM Ca?t. The records in (b) were obtained after 10 min of superfusion with the low Ca solution. The isolated frog retina was stimu- lated by the light flash (—3log units, 0.5 sec) before, during and after the light adaptation. The horizontal lines labeled LA indicate a 5 min light adaptation (white light, 8 x 10~> lux). The fast PIII responses during and after the adaptation were recorded at 30 sec, 1, 2, 3 and 4 min after the onset of the adaptation light and at 10, 30sec, 1, 2, 4, 6 and 8 min after the offset of the light, respectively. The bottom traces in (a) and (b) are flash light monitors. The time base is discontinuous. The downward deflection of responses means the same as in Fig. 1. adaptation on the fast PIII isolated by L-aAA. At 0.9 mM [Ca**],, as shown in (a), the amplitude of fast PIII response at 10 sec-dark after the offset of adaptation light was smaller than that in the initial dark, and the amplitude of the response recovered after 8 min of the offset. As shown in (b), reducing [Ca’*], from 0.9mM to 0.09mM led to an increase (about 1.3 times) in the amplitude of fast PIII elicited by the same stimulus light. The amplitude at 10 sec-dark after the offset of adaptation light was 1.4 times larger than that in the initial dark, and then the amplitude of re- sponse recovered after 8 min of the offset (called hypersensitivity). The offset of the adaptation light caused a large upward deflection of transre- tinal d.c. potential (hereafter called off-response). The amplitude of the off-response was 1.3 times larger than that of the initial downward deflection (hereafter called on-response) induced by the onset of the adaptation light. In this paper, we defined the gain of fast PIII as the ratio of the amplitude of fast PIII at 10 sec-dark after the offset of adapting light to that in the initial dark. Also we termed the ratio of off-response to on-response as the off-/on- response. Figure 4 shows the correlation between y =0.56x + 0.57 r=0.89 gain of fast PIII 0.5 1.0 1.5 2.0 of f-/on-response Fic. 4. The correlation between the gain of fast PIII and the magnitude of off-response. The isolated retina was superfused with the physiological solu- tion containing 20 4M Ba*+, 5mM Naz-aspartate and various Ca** concentrations. The superfused retina was stimulated by the flash light before and after the light adaptation. The duration of adapta- tion light was varied from 5 min to 20min. The straight line was drawn following to the first order equation (shown in the figure) obtained by the regression analysis. The intensities of stimulus and adaptation lights were the same as in Fig. 3. Hypersensitivity of Photoreceptors 37 Relative amplitude (fast Pill) 0 5 0 5 0 Time (min) Fic. 5. The effect of [Nat], on hypersensitivity. The isolated retina was superfused with the solutions containing 20 ~M Bat, 5 mM Na-aspartate, 0.09 mM Ca** and various Nat concentrations. The superfused retina was stimulated by the flash light before and after the light adaptation. In reducing [Na‘],, the equivalent to reduced [Na*] was replaced by choline. The symbol + at each [Na*], indicates an amplitude of fast PIII response evoked by the stimulus in the dark before the onset of the adaptation light. The amplitude of fast PIII at each [Na*], was normalized by that in the dark measured originally at 110 mM [Na*],. The data were obtained from the same retina. The conditions of stimulus and adaptation lights were the same as in Fig. 3. the gain of fast PIII and the off-/on-response. This result was obtained from several retinas super- fused with the solution containing various Ca?* concentrations. The correlation coefficient be- tween them was about 0.9, which was obtained by a linear, least-squared error regression analysis. It can be said that the hypersensitivity (the gain of fast PIII>1) occurs in the case of off-/on-response larger than 0.8. Figure 5 shows the effect of [Na*], on the hypersensitivity. The hypersensitivity was remark- able at low [Ca**] solution containing 110 mM [Na*] (normal concentration), where the gain of fast PIII was about 1.5. Hypersensitivity was less remarkable at 80 or 60 mM [Na™ ], and could not be observed any more at 40mM [Na‘],. The recovery of [Na‘], to normal concentration in- duced again a remarkable hypersensitivity (the gain of fast PII=1.5). The off-/on-response also decreased with reducing [Na*], (data not shown). Figure 6 illustrates the effect of [Ca?*], on the gain of fast PIII at various [Na*],. In normal [Na*],, lowering [Ca?*], from 0.9 mM to 0.6 mM induced hypersensitivity (the gain of fast PIII>1), but in 80 mM [Nat ],, lowering [Ca?*], to 0.4 mM Pill gain of fast tca2*}_ (mM) FI ) .6. [Ca?*],-dependence of hypersensitivity at four different Na* concentrations. Each value obtained from the retina superfused with the solutions con- taining the various concentrations of Na‘ and of Ca’*+ was normalized by the one obtained from the same retina superfused with the control solution (110 mM Nat and 0.9mM Ca?*). The conditions of stimulus and adaptation lights were the same as in Fig. 3. was required in order to induce hypersensitivity. In 40 mM [Na‘* ],, hypersensitivity scarcely occur- red even at 0.01 mM[Ca**], (the gain of fast 38 K. AZUMA PII[=1). Thus, the hypersensitivity induced by low Ca was markedly inhibited by reducing [Na*],. On the other hand, the change of [K*], little influenced the hypersensitivity (data not shown). Hypersensitivity of rat photoreceptors Figure 7 shows the effect of [Ca?*], on rat distal PIII responses before and after light adaptation. At 1mM [Ca**],, the distal PIII response at 10 sec-dark after the offset of the light was smaller than that in the initial dark, and then the ampli- 200uV (b) 10 30 1 2 4 Fic. 7. The effects of [Ca?*], and adapting light on the distal PIII of albino rat. The d.c. record in (a) was obtained after 10 min of the superfusion with the physiological solution containing 10mM Na- aspartate and 1 mM Ca**. After the recording, the superfusate was changed to the one that contained 10mM Na-aspartate and 0.1mM Ca**. The re- cord in (b) was obtained after 10 min of the super- fusion with the low Ca solution. The downward deflection means the same as in the case of Fig. 1. The bottom traces in (a) and (b) are flash light monitors. Horizontal bars labeled LA indicate 5 min-light adaptation (white light, 2107? lux). PIII responses elicited by the flash light (—1.5 log units, 0.5 sec) were recorded at 2 min before the onset of adaptation light and at appropriate peridos after the offset of the adaptation light (indicated beneath each flash light monitor). The time base is discontinuous. The d.c. recordings were obtained from the same retina (each response includes slow PIII component). tude of the response recovered within several minutes(a). Reducing [Ca**], from 1 mM to 0.1 mM led to the increase (about 1.2 times) in the amplitude of distal PIII response. In addition, the amplitude at 10sec-dark after the offset of the light was 1.5 times larger than that in the initial dark, and the off-response was 1.8 times larger than on-response(b). Though data were not shown, the hypersensitivity of rat photoreceptors was induced by IBMX as already reported in the frog retina [4]. Figure 8 shows the stimulus-response curves for the fast PIII of the rat in 0.1 mM, 1 mM and 2 mM [Ca?*],. In the case, the fast PIII was measured by the capacity-coupled method as described in the previous paper [3], because the slow PIII of rat retina could scarcely be suppressed by chemicals already described. The smooth curves in the figure were obtained from calculations based on the following equation: V=Vmaxl / (I'+1,’)—(1) (Naka and Rushton [20]), where V is the response amplitude related to a stimulus intensity (I), Vmax is the saturated response amplitude and I, is the half saturation 60 A = oO nN oO Amplitude (pV) =3 = -] 0 Log relative intensity Fic. 8. The effects of [Ca?*], on the stimulus-response curves of rat fast PIII. The isolated retina was superfused with the physiological solution contain- ing 10 mM Na-aspartate and various Ca?* concen- trations. Fast PIII component was elicited by 0.5 sec-stimuli of various intensities. The measure- ments were carried out consecutively with the same retina. Hypersensitivity of Photoreceptors 39 constant. The costants, I, and n, were determined by using the log-linear, least-squared error method to fit each set of data points. The values of I,(n) obtained were —2.3 (1.3), —2.3 (1.0) and —2.4 (1.0) for 2mM, 1mM and 0.1mM [Ca’*],, respectively. Reducing [Ca**], from 2 mM to 0.1 mM caused the increase (1.3 times) of Vmax. Figure 9 shows stimulus-response curves at 0.1 mM [Ca?*], in the dark (curve 1), in the light at 20 sec (curve 2) and 8 min (curve 3) after the onset of adaptation light for 10min, and at 10 sec-dark after the offset of the light (curve 4). The smooth curve in each condition was obtained from calcula- tions based on the equation 1. The values of I,(n) were —2.5 (1.1), —2.3 (1.7), —2.2 (1.5) and —2.6 (1.1) for curves 1, 2, 3 and 4, respectively. As indicated by the change from curve 1 to curve 2, the light adaptation decreased Vmax by half. During the light adaptation, I, decreased slightly, and V,yax increased about 1.5-fold (from curve 2 to 60 n (=) ips) oO Amplitude (uv) =3 = 2 =\ 0 Log relative intensity Fic.9. The effect of adapting light on the stimulus- response curve of rat fast PIII at 0.1 mM [Ca**],. Curve 1 was obtained from the measurement in the dark. Curves 2 and 3 were obtained from the measurements after 20 sec and 8 min of the onset of adaptation light, respectively. Curve 4 was obtained from the measurement at 10 sec-dark after the offset of the light. These experiments were carried out by 0.5 sec-stimuli of various intensities and constant adapting light (2 x 10~3 lux) from the same retina. 3). I, at 10 sec-dark after the offset of the light was slightly smaller than that at the initial dark condition, but Vj,ax increased markedly by about 1.5 times (from curve 1 to 4). Measurements of [K*], in the frog retina Figure 10(a) shows the light-induced changes of [K*], at normal [Ca?*],. The onset of adaptaition light induced the downward deflection of Vx+ which was corresponding to the decrease of [K*], in the interstitial space of the frog retina. The decrease reflects the reduction of K*-efflux through rod membranes by the adaptation light [19]. In the dark after offset of the adaptation light, [K*], tended to recover to the initial dark (a) 2 LA LA — I 1 1 Fic. 10. Light-induced changes of [K*], in the isolated retina at two different [Ca**],. (a) 0.9mM [Ca?*],. (b) 0.09 mM [Ca**],. The isolated retina was superfused with the solution containing 20 u.M Ba?+ with 5mM aspartate and 0.9mM Ca?+ or 0.09 mM Ca**. Horizontal bars labeled LA indi- cate 5 min-light adaptation. K+-responses elicited by 1 sec-stimuli were recorded at 2 min before the onset of the adaptation light and at appropriate periods after the offset of the light (indicated above each response). The bottom traces in (a) and (b) are flash light monitors. The intensities of stimulus and adaptation lights were the same as in Fig. 3. 40 K. AZUMA (a) a 1 Vt 1mV (b) Vit 1mV LA Fic. 11. sextetirn lh Dees A Ae Se The effect of IBMX on light-induced change of [K*], in the isolated retina. In (a) K*-response in the superfusate with 50 ~M IBMX (curve 2) is compared with that in the normal superfusate (curve 1). Record in (b) was obtained from the retina superfused with the IBMX contained solution. In part (b), horizontal bars labeled LA indicate 5 min-light adaptation. K‘-responses elicited by the flash light were recorded at 2 min before the onset on the adaptation light and at appropriate periods after the offset of the light (indicated above each response). The bottom trace is the flash light monitor. The conditions of stimulus and adaptation lights were the same as in Fig. 10. level, which indicated the recovery of K * -efflux. The figure also shows that the decrease of [K*], evoked by a light flash (hereafter termed K*- response) is influenced by light adaptation. At normal [Ca?*],, K*-response became markedly smaller during light adaptation (data not shown), and recovered gradually in the dark after the offset of the adaptation light. Figure 10(b) shows the effect of the adaptation light on K*-response at low Ca (0.09 mM). K*-response became smaller during the illumination (data not shown), but the K*-response at 30 sec-dark after the offset of the light was larger than that in the initial dark, and then K*-response recovered within 6 min. The figure also shows that the level of [K*], remark- ably overshoots the initial dark level after the offset of the adaptation light. In the previous paper [4], we reported that IBMX induced hypersensitivity of frog photore- ceptors even at normal [Ca?*],. We tested whether IBMX caused the transient increase of K*-response after the offset of adaptation light. As shown in Figure 11(a), the amplitude of K*- response was enlarged, and its time course was elongated by the addition of 50 ~M IBMX. The similar effects of the reagent on fast PIII were observed (data not shown). Figure 11(b) shows the effect of adapting light on K*-response at normal [Ca**], in the presence of 50 ~M IBMX. The amplitude of K*-response at 30 sec-dark after the offset of the adaptation light was about 1.4 times larger than that in the initial dark, and the overshoot of [K*], was also remarkable. DISCUSSION In this experiment, L-aAA abolished the b- wave of frog ERG rapidly, and the slow PIII of it slowly, but D-aAA did not. This suggests that the L-aAA is an agonist of the amino acids (aspartate and glutamate) which can suppress the b-wave. The suppression effect of the L-aAA on slow PIII which originates in Miiller cell [10, 11] is consistent with other authors’ result [15] showing that the Hypersensitivity of Photoreceptors 41 aAA can cause the severe destruction of frog Miiller cell. As discussed by Casper and Reif- Lehler [16], the difference in the effects between L- and D-aAA may be a result of the differences in conformation and receptor interactions. On the other hand, neither of the two isomers of aAA was effective on the slow PIII of albino rat (see results). This seems to contradict with other authors’ result [17] showing that intravitreal injec- tion of the DL-aAA to the rat causes morpho- logical changes indicating the destruction of the Miiller cells. The discrepancy may be due to the differences between isolated retina (this experi- ment) and living eye (other authors), or because aAA can not destroy the membrane itself of the Miiller cells. Ba’t suppressed the slow PIII of frog at concentration higher than 20 uM, which was con- sistent with Matsuura’s result [13], and barely suppressed the slow PIII of the rat at concentration higher than 500 uM. As Ba’? is considered to be a blocker of the g, of Miiller cell [12, 13], it can be said that Ba’* is more accessible to the K*- channel of frog Muller cell than to that of rat Miiller cell. In this report, the hypersensitivity was observed in measuring chemicals-isolated fast PIII of frog. The hypersensitivity was also observed in rat retinas at low Ca or in the presence of IBMX, as in frog retinas. These findings corroborate the con- clusion described previously [3] that the hypersen- sitivity is a phenomenon in rod photoreceptor itself. A similar phenomenon was reported in skate retina [5]. Therefore it can be said that the hypersensitivity is popular in the vertebrate photo- receptor. In measuring [K*], of frog retina under the conditions causing the hypersensitivity, K*- response increased markedly in the dark after the offset of adaptation light, where the large over- shoot in [K*], was observed at the offset (see results). This is consistent with Oakley’s result [21]. The overshoot indicates the rapid increase of K*-efflux, and the increase of K*-response is corresponding to the reinforcement in suppression of K*-efflux by flash light. As suggested by Oakley and Steinberg [22], the change of K*- efflux is caused by that of Na*-influx through rod membrane. Therefore it can be said that the large overshoot in [K* ], indicates the transient increase in Na*-dark current. It was confirmed that IBMX induced hypersensi- tivity of the photoreceptors of both frogs and rats. The reagent induced a similar effect on [K*], change as that of low Ca (see results). As already shown, reducing [Na*], suppressed the hypersen- sitivity, and the alternation of [K*], was less effective. From these results, we consider the cause of hypersensitivity as follows. The hypoth- esis is composed of three assumptions. 1) The g,. of rod photoreceptors in the dark increases under such condition as low Ca or the existence of IBMX [23], which intensifies Na‘-influx and K*-efflux and then leads to the increase of [Na*], and the decrease of [K*];. 2) Light adaptation causes the decrease in g., which recuces both Na *-influx and K *-efflux, and then leads to the decrease of [Na * ]; and the increase of [K*];, i.e., the increase of Nernstian potentials of the two ions across rod membranes. This is because Na *-K *-pump works even during the light adaptation. 3) In the dark following the light adaptation, the recovery of g, to the initial dark level is faster than that of the Nernstian potentials which increases during light adaptation. If these assumptions are reasonable, the following events will be expected. The large Na*-dark current is observed immediately after the offset of the adaptation light, and the increase of fast PIII amplitude (hypersensitivity) occurs in the early dark after the offset. The first and second assumptions are supported by the data of X-ray microanalysis [24]. The data have shown that reducing [Ca** ], from 1.8 mM to 0.18 mM induces the increase of [Na* ]; and the decrease of [K*]; in rod photoreceptors, and that light adaptation causes a highly significant reduction of [Na‘* ]; and an increase of [K*];. Lowering [Na*], may give the rod photorecep- tors the following effects. The lowering induces the decrease of Na‘ -dark current [23] resulting in the decrease of Na‘-influx. The difference of Na*-influx between dark and light adaptations under such condition is markedly less than that under normal [Na*], condition. Therefore, low [Na*], condition can not induce the increases of Nernstian potentials of Na * and K* during light adaptation. 42 This may be a reason why the lowering in [Na*], suppresses the hypersensitivity. As the change of [Na*], gives the effect on the activity of Na*-Ca**+ exchange pump [23], it was not excluded that the pump is not related to the hypersensitivity. ACKNOWLEDGMENT I wish to thank Professors N. Iwasaki and M. Fujimoto for continuous encouragement and technical advice. Iam indebted to Dr. M. Azuma, with whom I worked in the early phases of this project, for technical assistance and helpful discussions. 10 REFERENCES Hanani, M. and Hillman, P. (1976) Adaptation and facilitation in the barnacle photoreceptor. J. Gen. Physiol. , 67: 235-249. Ventura, D. F. and Puglia, N. M. (1977) Sensitivity facilitation in the insect eye. A parametric study of light adapting conditions. J. Comp. Physiol., 14: 35- 49. Azuma, M. and Azuma, K. (1979) The increase in sensitivity following light illumination in frog photo- receptors. Vision Res., 19: 1171-1175. Azuma, M. and Azuma K. (1982) The action of phosphodiesterase inhibitors on the hypersensitivity of frog photoreceptor. Vision Res., 22: 151-155. Dowling, J. E. and Ripps, H. (1972) Adaptation in skate photoreceptors. J. Gen. Physiol. , 60: 698-719. Hanawa,I., Ando,H. and Takahashi, K. (1981) Enhancement of visual response after illumination in the isolated frog retina. Exp. Eye Res., 32: 719- 727. Brown, J. E. and Waloga, G. (1981) Effect of cyclic nucleotide and calcium ions on Bufo rods. In “Molecular Mechanism of Photoreceptor Transduc- tion”. Ed. by W. Miller, Academic Press, New York, pp. 369-380. Furukawa, T. and Hanawa,I. (1955) Effects of some common cations on electroretinogram of the toad. Jpn. J. Physiol., 5: 289-300. Murakami, M. and Kaneko, A. (1966) Differentia- tion of PIII subcomponents in cold-blooded verte- brate retinas. Vision Res., 6: 627-636. Witkovsky, P., Dudek, F. E. and Ripps, H. (1975) Slow PIII component of the carp retina. J. Gen. Physiol., 65: 119-134. K. AZUMA 11 14 15 16 17 19 20 PAL 22 23 24 Fujimoto, M. and Tomita, T. (1979) Reconstruction of the slow PIII from the rod potential. Invest. Ophthalmol. Visual Sci., 18: 1090-1093. Bolnick, D. A., Walter, A. E. and Sillman, A. J. (1979) Barium suppress slow PIII in perfused bullfrog retina. Vision Res., 19: 117-119. Matsuura, T. (1984) Effects of barium on separately recorded fast and slow PIII responses in bullfrog retina. Experientia, 40: 817-819. Szamier,R.B., Ripps,H. and Chappell, R. L. (1981) Changes in ERG-wave and Miller cell structure induced by alpha-aminoadipic acid. Neurosci. Lett., 21: 307-312. Bonaventure, N., Roussel,G. and Wioland, N. (1981) Effects of D, L-a-amino adipic acid on Miller cells in frog and chicken retinae in vivo: relation to ERG b-wave, ganglion cell discharge and tectal evoked potentials. Neurosci. Lett., 27: 81-87. Casper, D. S. and Reif-Lehrer, L. (1983) Effects of alpha-aminoadipate isomers on the morphology of the isolated chick embryo retina. Invest. Ophthal- mol. Vis. Sci., 24: 1480-1488. Pedersen, O. and Karlsen, R. L. (1979) Destruction of Miiller cells in the adult rat by intravitreal injection of DL-a-aminoadipic acid: an electron microscopic study. Exp. Eye Res., 28: 569-575. Fujimoto, M. and Kubota, T. (1976) Physicochem- ical properties of a liquid ion exchanged micro- electrode and its application to biological fluids. Jpn. J. Physiol., 26: 631-650. Oakley, B., II (1983) Effects of maintained illu- mination upon [K*], in the subretinal space of the isolated retina of the toad. Vision Res., 23: 1325- 1337. Naka K.I. and Rushton, W.A.H. (1966) S- potentials from colour units in the retina of fish (Cyprinidae). J Physiol., 185: 536-555. Oakley, B., II (1984) Effects of low [Ca?+], upon {K*], during and after maintained illumination of the isolated retina of the toad. Vision Res., 24: 815- 819. Oakley, B., II and Steinberg, R. H. (1982) Effects of maintained illumination upon [K*], in the subretinal space of the frog retina. Vision Res., 22: 767-773. Hodgkin, A. L., McNaughton, P. A., Nunn, B. J. and Yaw, K. W. (1984) Effect of ions on retinal rods from Bufo marinus. J. Physiol., 35: 649-680. Somlyo A. P. and Walz, B. (1985) Elemental dis- tribution in Rana pipiens retinal rods: Quantitative electron probe analysis. J. Physiol., 358: 185-195. ZOOLOGICAL SCIENCE 5: 43-51 (1988) Effects of Cellular Dehydration on Drinking and Plasma Angiotensin II Level in the Eel, Anguilla japonica YosHIO TAKEI, JUNKO OxuBo and KEN’ICHI YAMAGUCHI! Department of Physiology, Kitasato University School of Medicine, Sagamihara, Kanagawa 228, and ‘Department of Physiology, Niigata University School of Medicine, Niigata, Niigata 951, Japan ABSTRACT— An intra-arterial injection of 0.5 ml of 7% NaCl, 14% NaCl, 65% sucrose, or 61% sucrose in 0.9% NaCl into the dorsal aorta of freshwater (FW) eels, which theoretically causes cellular dehydration by 2.8% (7% NaCl and sucrose solutions) or 5.7% (14% NaCl), consistently inhibited drinking for 1 hr after injection, compared with controls injected with 0.9% NaCl. Drinking was not stimulated by any of the injections for up to Shr. Plasma angiotensin II (AII) level increased consistently 15 min after any of the injections, and the increase became smaller after 4 hr. Plasma Na level increased for 4 hr after the injection of hypertonic saline, whereas a decrease was observed after the injection of hypertonic sucrose. Drinking was also inhibited after injection of 0.5 ml of 7% NaCl into eels adapted to 1/3 seawater (SW), which is isosmotic to plasma, or into those exposed to 1/3 SW or SW for 1.5 hr. Plasma AII level increased in all the experimental groups after 15 min, but the increase was significant only in 1/3 SW-adapted eels. Plasma Na level increased for 4 hr after injection of 7% NaCl in 1/3 SW-exposed and SW-exposed eels, but the increase was no more significant after 4 hr in 1/3 SW-adapted eels. Collectively, drinking was decreased and plasma AII was increased by the stimuli to cellular dehydration in the eel. These results are the converse of those obtained for mammals and birds, in which administration of a hypertonic solution of NaCl induces drinking and © 1988 Zoological Society of Japan reduces plasma AII level. INTRODUCTION It is generally accepted that intravascular ad- ministration of hypertonic solution of solutes which are impermeable to the cell membrane and, thus, causes cellular dehydration, induces drinking in mammals [1], birds [2, 3] and reptiles [4], while intravascular administration of hypertonic NaCl inhibits renin release in mammals [5] and birds [6]. Thus, administration of hypertonic NaCl is dip- sogenic in mammals and birds, even though the levels of angiotensin II (AII), another potent dipsogen [1], are suppressed by such treatment. In fishes, Hirano [7] reported that the rate of drinking is increased in freshwater (FW) eels by a slow infusion of hypertonic NaCl solution. However, our preliminary data suggest that injection of Accepted June 16, 1987 Received May 15, 1987 hypertonic NaCl failed to stimulate drinking in FW eels [8]. Thus, the effect of hypertonic solutions on drinking in fishes remains to be established. Furthermore, it is unknown how increased plasma osmolality, or cellular dehydration, influences plasma AII levels in the eel. AII has been shown to stimulate drinking in the eel [9] and other fishes [10-13]. The present study was undertaken to examine further the effect of cellular dehydration on drinking and plasma AII level in the eel. As stimuli to cellular dehydration, we made a single injection of hypertonic NaCl and sucrose instead of infusion, to avoid dilution of the osmotic load during the infusion by water influx across the gill. We used not only FW eels but those adapted or exposed to 1/3 seawater (SW) or full-strength SW as experimental animals, since the influx of water across the gill after the osmotic load might be smaller in these eels than in FW eels. However, 44 Y. TAKEI, J. OKUBO AND K. YAMAGUCHI SW-adapted eels were removed from the experi- ment because we preliminarily found that Na ions loaded were excreted immediately after injection of hypertonic NaCl into these eels. MATERIALS AND METHODS Animals Cultured Japanese eels, Anguilla japonica, were purchased from a local dealer. They were kept in groups of 20 in 1-ton, FW tanks for more than 1 week before use. Some eels were transferred to a 0.5-ton, 1/3-SW tank, and acclimated for more than 2 weeks before use (1/3 SW-adapted eels). Water in the tank was continuously filtered, aerated and thermoregulated at 18+0.5°C. Eels were not fed after purchase, and weighed 197+1 g (mean +SEM, n=102) at the time of experiments. Surgical procedures After the eels were anesthetized with 0.1% tricaine methanesulfonate (Sigma), a vinyl tube (o.d.:2.0mm) was inserted into the esophagus, and 2 polyethylene tubes (o0.d.:0.8mm) were inserted into the dorsal and the ventral aorta, respectively. The blood stream through these aortae was not occluded by the cannulation as described previously [14]. The eels that bled more than approximately 0.05 ml were excluded from the experiment. After the surgery, the eels were transferred to a plastic trough through which aerated and thermoregulated (18°C) water was circulated. In this condition, the cannulated eels usually survived more than 2 weeks. The circulat- ing water through the trough could be changed from FW to 1/3 SW or SW by turning a 3-way stop cock. The catheter placed in the esophagus was connected to a drop counter for continuous measurement of drinking rate [7]. The drunk water that dropped from the esophageal catheter was not reintroduced into the stomach. The catheters in the aortae were connected to syringes filled with saline containing Ca heparin (10 U/ml). Eels were allowed to recover for more than 18 hr post-operatively. Experimental protocol Four different groups of eels were used in this experiment, eels adapted to FW, those adapted to 1/3 SW, those exposed to 1/3 SW for 1.5 hr and those exposed to SW for 1.5 hr. The FW eels were injected with 0.5 ml of one of the solutions of 0.9% NaCl, 7% NaCl, 14% NaCl, 65% sucrose, and 61% sucrose in 0.9% NaCl into the dorsal aorta in 1 min, whereas 0.5 ml of blood was withdrawn simultaneously from the ventral aorta at the same rate as the injection. The blood was collected into a chilled syringe which contained 12.5 «l each of 125mM_~ disodium EDTA, 25mM __ o- phenanthroline and 0.2% neomycin sulfate (Wako Chemicals, Tokyo), and used for radioimmuoassay of AII. The sucrose solutions were approximately isosmotic to 7% NaCl (ca. 2.50Osm). The 1/3 SW-adapted eels, 1/3 SW-exposed eels and SW- exposed eels were injected with 0.5 ml of either 0.9% or 7% NaCl, and 0.5 ml of blood was also withdrawn simultaneously as described for FW eels. The injection-withdrawal procedure was repeated 15 min and 4hr after the initial proce- dure, but the injection at these times consisted of 2% dextran (molecular weight : 60,000—90,000, Wako Chemicals, Tokyo) in 0.9% NaCl. Dextran was added to maintain colloidal osmotic pressure. Before each procedure, 5041 of blood were collected into capillary tubes for measurements of the hematocrit and concentrations of Na and K ions in plasma. Water intake was measured every 5 min after injection by means of the number of drops that emerged from the esophageal catheter (0.03 ml/ drop). The measurement was continued for up to 5 hr because in another poikilothermal animal, the iguana, drinking response to cellular dehydration occurred much more slowly than in_ the homeothermal mammals and birds [4]. For radioimmunoassay of AII, 0.5 ml of blood with- drawn at time 0, 15 min and 4 hr was centrifuged at 2,500 g at 2°C for 15 min. Immunoreactive AII was extracted from plasma with acetone and petroleum ether, and the concentration was deter- mined as reported previously [15]. The antibody used in this assay exhibited 100% cross-reactivity with natural eel Asn!-Val° AII [16, 17], and the Drinking and Angiotensin Responses in Eels dilution curve of the assayable AII, extracted from pooled eel plasma, was superimposable on the standard curve of authentic Asp'-Ile? AII. The concentrations of Na and K ions in plasma were determined with an atomic absorption spec- trophotometer (Hitachi 180-80) after 1/2,000 and 1/100 dilutions, respectively. Double distilled water collected into the capillary tube and diluted as above was used as a blank for spectropho- tometry. All determinations were made in dupli- cate. Analysis of data Variation of data is a common problem when we attempt to quantify some parameter accompanying behavior, because the behavior is so vulnerable to environmental influences. In fact, the water intake of the eel measured in the present experiments was also variable, and it seems that the variation masks the actual change in water intake after injection. Thus, the change in water intake in each eel during the periods from 1 to Shr after injection of the hypertonic solution was classified into an increase, no change, or a decrease compared with the intake during the same time period before injection, and statistically compared with that of controls injected with 0.9% NaCl by the nonparametric Fisher’s 1/3. SW- exposed 45 exact probability test. Actual water intakes before and after injections were also given in the text. The changes in plasma AII, Na and K levels and the hematocrit after injection of hypertonic solu- tions were expressed as ratios to the values before injection (zero-time value) to make the actual changes clearer. The zero-time values were also given in the text. By doing this, it is also possible to compare the degree of the changes among these parameters after injection. The changes in plasma All, Na and K levels and hematocrit after injec- tion of the hypertonic solution were compared with those of controls by the nonparametric van der Waerden test [18]. Data that fell outside the range of other data of the same group was excluded by means of the Smirnov-Grubbs test. Significance was determined at P<0.05. All results are expressed as means +SEM. RESULTS Drinking rate Thirty one of 79 eels did not drink at all in FW, and the mean water intake of FW eels was 0.04+0.01 ml/5 min/eel (n=79). However, FW eels started drinking copiously upon exposure to aa FW 1/3 SW- adapted 20} g a c& 10 = vn per \ { ite) ~~ 0 i= ro| b b oe ~&—— 4 co 10 — £ of C Cc @ wy oOo = a BC ste 0 1 2 3 4 5 0 1 Time id 3 4 5 after Fic. 1. /c Haan, a (0) 1 2 3 4 0 1 2 3 4 5 injection (hr) Changes in water intake after injection of 0.5 ml of (a) 0.9% NaCl, and (b) 7% NaCl in FW eels (a, n=12; b, n=9), 1/3 SW-adapted eels (a, n=14; b, n=9), 1/3 SW-exposed eels (a, n=9; b, n=8), and SW-exposed eels (a, n=4; b, n=5). The change in water intake after injection of 7% NaCl was also expressed in terms of the difference from the mean intake of controls injected with 0.9% NaCl (b-a) to make the effect of 7% NaCl clearer (c). Each 5-min intake after injection was subtracted by the 5-min intake calculated from the 30-min intake before injection. Due to the great variation of water intake among individuals, standard error of the mean was not given. The actual water intakes are given in Table 2. Smaller arrows indicate the time of exposure to 1/3 SW or SW, and larger arrows indicate the time of injection. *5-min water intakes were 2.84 ml (a), 4.07 ml (b), and 1.23 ml (c). 46 Y. Takel, J. OkuBo AND K. YAMAGUCHI 1/3 SW (1.58+0.25 ml/initial 5 min/eel; latency= 46+1 sec, n=17) or SW (4.27+0.55 ml/initial 5 min/eel; latency=S50+ 15 sec, n=9) (Fig. 1). This initial drinking was followed by a decreased rate of drinking for 1-2 hr. Thus, the eels exposed to 1/3 SW or SW for 1.5 hr were drinking at the rate of 0.85+0.11 ml/5 min/eel (n=17) and 0.84+0.09 ml/5 min/eel (n=9), respectively. Following this temporary decrease in drinking rate, the rate increased again to higher and constant levels of approximately 3 ml/5 min/eel and 1 ml/5 min/eel, respectively, in SW-exposed and 1/3 SW-exposed eels (Fig. 1). Thus, 1/3 SW-adapted eels were drinking at a constant rate of 1.35+0.08 ml/5 min/eel (n=23) at the time of injection. Change in drinking rate after osmotic stimuli Although the drinking rate of FW eels before injection of hypertonic solutions was variable (Table 1), it was apparent that the drinking rate was slightly inhibited for 1-2 hr after injection of hypertonic solutions, as compared with a slight increase in drinking rate in controls injected with 0.9% NaCl. The decrease in drinking rate was statistically significant for 1 hr after injection of any of the hypertonic solutions compared with the change in controls (Table 1). The decrease con- tinued up to 4 hr after injection of 14% NaCl, and up to 5 hr after injection of 61% sucrose in saline. TABLE 1. 61% sucrose in 0.9% NaCl in FW eels The eels were slightly hyperactive for a few minutes after injection of hypertonic solutions, but it was apparent that the behavior thereafter was quite normal. Injection of 0.9% NaCl had little effect on drinking in FW, 1/3 SW-exposed and SW-exposed eels, but it inhibited drinking for 1 hr after injection in 1/3 SW-adapted eels (Fig. 1a). The drinking rate appears to have increased in SW- exposed eels after injection of 0.9% NaCl, but this is a natural increase which should have occurred without injection as mentioned above. On the other hand, injection of 7% NaCl clearly inhibited drinking in 1/3 SW-adapted and 1/3 SW-exposed eels, and inhibited the natural increase in SW- exposed eels (Fig. 1b). Therefore, when the change in drinking rate after injection of 7% NaCl was corrected by subtraction with the change after injection of 0.9% NaCl, the inhibition of drinking became more evident in all groups of eels (Fig. 1c, Table 2). The degree and duration of the inhibi- tion were greater in 1/3 SW-adapted, 1/3 SW- exposed, and SW-exposed eels than in FW eels, probably due to the greater rate of drinking before injection. Statistical analyses revealed that the inhibition was significant for 1 hr in FW eels, for 2— Shr in 1/3 SW-adapted eels, for 1-Shr in 1/3 SW-exposed eels, and for 1-3 hr in SW-exposed eels (Table 2). Changes in water intake after injection of 0.9% NaCl, 7% NaCl, 14% NaCl, 65% sucrose, and Number Water intake during each time period after injection (ml) Injection of eels —1-0 hr 0-1 hr 0-2 hr 0-3 hr 0-4 hr 0-5 hr 0.9% NaCl 11 0.63+0.53 0.9140.59 1.43+40.69 1.9740.87 2.904+1.07 4.44+1.67 (10 0 1) (9 0 2) (9 0 2) (9 0 2) (9 0 2) 7% NaCl 9 0.22+0.15 0.09+0.06 0.82+0.49 1.34+0.69 1.74+0.76 2.38+0.86 (4 1 4)* (6 1 2) (6 1 2) (6 1 2) (6 1 2) 14% NaCl 12 0.26+0.16 0.16+0.07 0.29+0.17 0.50+0.24 0.8040.33 1.99+0.83 (2 8 2)* G72) (4 6 2)* (4 6 2)* (5 6 1) 65% _ sucrose 10 0.09+0.08 0.08+0.05 0.74+0.38 1.59+0.67 2.3140.96 3.664+1.52 (3 4 3)* (6 3 1) (7 3 0) (7 3 0) (7 3 0) 61% sucrose 11 0.81+0.31 0.1140.05 0.24+0.11 0.9340.45 1.7340.86 3.0941.58 in saline (0 4 7)* (0 4 7)* (1 3 7)* (3"3"5)* (2 3 6)* Since the drinking rate of each animal was so variable, the change in water intake of each animal during the time periods of 1-5 hr after injection was classified into an increase (+), no change (0), or a decrease (—) compared with the intake during the corresponding time period before injection, and analyzed by the nonparametric statistics. compared with the change in controls injected with 0.9% NaCl. In parentheses are numbers of eels that showed different drinking responses (+, 0, —). *P<0.05 Values are means+SEM. Drinking and Angiotensin Responses in Eels 47 TaBLE 2. Changes in water intake after injection of 7% NaCl in FW eels, 1/3 SW-adapted eels, 1/3 SW-exposed eels and SW-exposed eels Group Number Corrected water intake of each time period after injection of 7% NaCl (ml) of oO eels eels 0-1 hr 0-2 hr 0-3 hr 0-4 hr 0-5 hr FW 9 —0.81+0.06 —0.59+0.49 —0.61+0.69 —1.14+0.76 —2.06+ 0.86 (0 0 9)* (2 0 7) (3 0 6) (4 0 5) (3 0 6) 1/3 SW 9 —5.0442.82 —13.76+4.39 —19.95+4.99 —26.59+6.56 —28.31+8.47 adapted (4 0 5) (0 0 9)* (0 0 9)* (1 0 8)* (0 0 9)* 1/3 SW 8 —6.0341.49 —10.2641.79 —9.88+2.31 —12.5243.23 —17.58+4.41 exposed (0 0 8)* (0 0 8)* (0 0 8)* (0 0 8)* (0 0 8)* SW 5 —12.904+2.18 —31.10+5.26 —41.72+10.00 —56.08+16.67 —64.18+20.41 exposed (0 0 5)* (0 0 5)* (0 0 5)* (1 0 4) (1 0 4) The water intake after injection of 7% NaCl was corrected by subtraction with the mean intake of controls injected with 0.9% NaCl. For reference, see Fig. lc. Due to the variation of individual water intake, changes in water intake during the time periods of 1-5 hr after injection were classified into an increase (+), no change (0), or a decrease (—) compared with the intake during the corresponding time period before injection, and analyzed by the nonparametric statistics. In parentheses are numbers of eels that showed different drinking responses (+, 0, —). *P<0.05 compared with the intake before injection. Values are means+SEM. TABLE 3. Levels of angiotensin II, Na and K ions in plasma, and hematocrit in FW eels, 1/3 SW-adapted eels, 1/3 SW-exposed eels and SW-exposed eels before injection of hypertonic solutions Hels Angiotensin II Na K Hematocrit (pg/ml plasma) (mM) (mM) (%) FW 109.7+12.9 140.9+2.8 2.69+0.13 28.2+1.0 (47) (32) (32) (52) 1/3 SW-adapted 101.8+24.1 155.7+3.4* 2.50+0.19 20.3+1.5* (20) (23) (22) (22) 1/3 SW-exposed 135.9+32.7 132.8+2.8* 2.60+0.12 24.0+2.1* (15) (19) (19) (18) SW-exposed 182.4+31.8* 137.0+6.6 2.12+0.15* 25.9+2.6 (72) ( 9) ( 9) ( 9) These values correspond to the zero-time values in Figs. 2 and 3. parentheses. means +SEM. Plasma All, Na, and K levels and hematocrit The levels of AII, Na and K ions in plasma, and the hematocrit of FW eels, 1/3 SW-adapted eels, 1/3 SW-exposed eels and SW-exposed eels before injection of hypertonic solutions are shown in Table 3. These values are the means of zero-time values in Figures 2 and 3. It was found that plasma AIlI level tended to be higher in 1/3 SW and SW-exposed eels than in FW eels, but the differ- ence was significant only in SW-exposed eels. *P<0.05 compared with the corresponding value for FW eels. Numbers of animals are in Values are Plasma Na level was significantly higher in 1/3 SW-adapted eels than in FW eels, but it was lower in 1/3 SW-exposed eels. Changes in plasma AII, Na and K levels and hematocrit after osmotic stimuli Plasma AII level invariably increased 15 min after injection of any of the hypertonic solutions in FW eels compared with controls injected with 0.9% NaCl (Fig. 2). The increase became smaller after 4 hr, but it was still significant in eels injected 48 Y. TAKEI, J. OKUBO AND K. YAMAGUCHI with 14% NaCl or 65% sucrose. Plasma Na level increased for 4hr after injection of hypertonic solutions of NaCl, while the level decreased for 4 hr after injection of hypertonic solutions of su- crose. The changes in plasma K level were not consistent after injection of hypertonic solutions (Fig. 2). Hematocrit invariably decreased after injection of hypertonic solutions, and the degree of the decrease at 15 min was greater after injection of sucrose solutions than after injection of NaCl solutions. In controls injected with 0.9% NaCl, plasma AII, Na and K levels did not change but the hematocrit decreased after injection. Plasma AII level increased in 1/3 SW-adapted eels 15 min after injection of 7% NaCl as observed = i=) [) 15min RY Ahr after injection a oO vos Ye ooo oO oOo Change in each parameter (ratio to time 0) 14%NaCl 65%sucrose 51%sucrose in saline 0 0.9%NaCl 7% NaCl Fic. 2. Changes in plasma angiotensin II (AIT), Na and K levels and hematocrit (Ht) 15 min (plain columns) and 4 hr (shaded columns) after injection of 0.5 ml of 0.9% NaCl (Al, n=12; Na, n=6; K, n=6; Ht, n=12), 7% NaCl (AII, n=7; Na, n=8; K, n=8; Ht, n=8), 14% NaCl (AII, n=9; Na, n=12; K, n=12; Ht, n=10), 65% sucrose (AII, n=10; Na, n=3; K, n=3; Ht, n=11) and 61% sucrose in 0.9% NaCl (AII, n=9; Na, n=3; K, n=3; Ht, n=11) in FW eels. Each change is expressed in terms of a ratio to the value before injection (zero-time value). The zero-time values of AII, Na, K and Ht in FW eels are shown in Table 3. *P<0.05 compared with any changes in the corresponding controls injected with 0.9% NaCl. Values are means+SEM. in FW eels, while the increase was not significant for 4 hr after injection in 1/3 SW and SW-exposed eels (Fig. 3). Plasma Na level increased in all groups of eels for 4 hr after injection of 7% NaCl, except in 1/3 SW-adapted eels after 4 hr. Plasma K level and hematocrit invariably decreased 15 min after injection of 7% NaCl, but the levels were restored to normal after 4 hr except in SW-exposed eels (Fig. 3). In controls injected with 0.9% NaCl, plasma AII, Na and K levels did not change in 1/3 SW-adapted eels, 1/3 SW-exposed eels and SW- exposed eels except for the AII level 4 hr after injection in SW-exposed eels (ratio to zero-time value, 1.54+0.24, n=3). The hematocrit consist- [—) 15min abc bc WY Ahr after injection Ed Change in each parameter (ratio to time 0) 1/3 SW FW adapted 1/3 SW SW exposed — exposed Fic. 3. Changes in plasma angiotensin II (AII), Na and K levels and hematocrit (Ht) 15 min (plain columns) and 4 hr (shaded columns) after injection of 0.5 ml of 7% NaCl in FW eels (AII, n=7; Na, n=8; K, n=8; Ht, n=8), 1/3 SW-adapted eels (AI, n=12; Na, n=9; K, n=9; Ht, n=9), 1/3 SW-exposed eels (AII, n=8; Na, n=9; K, n=9; Ht, n=9), and SW-exposed eels (AII, n=4; Na, n=5; K, n=5; Ht, n=5). Each change is expres- sed in terms of a ratio to the value before injection (zero-time value), and corrected by subtraction with the mean change of controls injected with 0.9% NaCl. The zero-time values of AII, Na, K and Ht in each group of eels are described in Table 3. *P<0.05 compared with the zero-time value. Values are means+SEM. Drinking and Angiotensin Responses in Eels 49 ently decreased after injection of 0.9% NaCl as observed in FW eels: 0.86+0.04 and 0.80+0.02 (15 min and 4hr after injection, n=13) in 1/3 SW-adapted eels, 0.88+0.03 and 0.81+0.02 (n=9) in 1/3 SW-exposed eels, and 0.90 + 0.04 and 0.88+0.05 (n=4) in SW-exposed eels. DISCUSSION Intravascular administration of hypertonic solu- tions of NaCl or sucrose, which theoretically causes cellular dehydration by 1.60% in rats, 2.15% in dogs and 1.23% in humans, induces copious drinking [1]. The administration of hyper- tonic NaCl is also dipsogenic in birds (pigeons, [2]; quail, [3]) and a reptile (iguana, [4]). Assuming that total body water is 65% of body weight in birds [19] and 74% in reptiles [20], the minimal cellular dehydration for elicitation of drinking appears to be 1.03% in pigeons, 0.73% in quail, and less than 2% in iguanas, when the value is calculated by the equation given by Fitzsimons [1]. Thus, the administration of hypertonic NaCl, which causes cellular dehydration by as much as 2%, is potently dipsogenic in all classes of truly terrestrial vertebrates. In the present experiment, however, an intra-arterial injection of 0.5 ml of 7% NaCl, 65% sucrose or 61% sucrose in saline into FW eels, which theoretically causes cellular de- hydration by 2.81%, failed to induce drinking for 5 hr after injection, and actually inhibited drinking for lhr. The injection of 14% NaCl was not dipsogenic, either, in FW eels. Thus, the drinking response to stimuli to cellular dehydration appears to differ in the eel from the responses in terrestrial animals. When the changes in drinking rate, plasma Na and AII levels in FW eels after hypertonic saline are compared closely with those of quail, the weak dipsogenicity of cellular dehydration in the eel becomes more evident. In the quail, injection of 0.5 ml of 7% NaCl injected in the same protocol as employed in this study induced immediate copious drinking although plasma AII level decreased to less than half [21]. Plasma Na level returned to the control level within 15 min after injection in the quail, but, as illustrated in Figure 2, plasma Na level was still higher than the control level even after 4 hr in FW eels. This is probably due to the poor ability of FW fishes to excrete excess salts [20]. Thus, stimuli to cellular dehydration con- tinued for more than 4hr after injection of the hypertonic solution in FW eels. Further, plasma AII level increased after the injection in the eel. Even with these favorable conditions for elicitation of drinking, eels did not increase drinking rate for 4 hr after injection of hypertonic solutions. It is possible that the injection of hypertonic solutions into FW eels increased the osmotic influx of water across the gill, and the resultant expan- sion of blood volume inhibited drinking. Howev- er, the expansion of blood volume after the same degree of osmotic load was shown to be smaller in the eel than in the quail [22], while drinking was induced in the quail [21] and inhibited in the eel as in the present study. Furthermore, in eels adapted or exposed to 1/3 SW or SW, the expansion of blood volume might be smaller after injection of 7% NaCl than in FW eels, but drinking was also inhibited after injection of 7% NaCl into these eels. Thus, it is possible that the eels are more sensitive to the inhibition of drinking by expansion of blood volume, and/or that the osmotic mecha- nism for stimulation of drinking is less developed in the eel than in terrestrial vertebrates. It is rather unexpected that plasma Na concen- tration was smaller in esophagus-cannulated eels after exposure to 1/3 SW and SW for 1.5 hr (Table 3). This result indicates that these eels did not suffer from cellular dehydration when injections were made 1.5 hr after the exposure. We also obtained preliminary data that when FW eels with esophageal fistula were exposed to SW, plasma Na levels decreased for 1-2 hr, then increased grad- ually, whereas if drunk SW was reintroduced into the esophagus by a pulse injector synchronized with a drop counter, plasma Na concentration increased within 15 min after exposure to SW (unpublished observations). However, the in- crease continued thereafter linearly until death in the former eels, while the latter eels could adjust the plasma Na to the physiological level and could survive thereafter. It is without doubt that esopha- gus-cannulated eels in FW or isosmotic 1/3 SW did not suffer from cellular dehydration. Hirano [7] observed that a slow infusion of 50 Y. TAKEI, J. OKUBO AND K. YAMAGUCHI hypertonic NaCl into FW eels induced drinking in some FW eels. This result appears to be inconsis- tent with the present result. One possible inter- pretation of this discrepancy could be the differ- ence in the route of administration. We did not adopt infusion because eels are aquatic species and, thus, the hypertonic solution given might be diluted by an influx of environmental water across the gill during the slow infusion. Another possibil- ity arises from the experimental procedures. We withdrew 0.5 ml of blood simultaneously with the injection for radioimmunoassay of AII. We chose these procedures in order to follow the changes in drinking rate, plasma AII, Na and K levels at the same time in a single animal. We found that these procedures themselves did not cause significant changes in plasma AII, Na and K levels as exemplified by the changes in controls injected with 0.9% NaCl, but the drinking rate appears to be more vulnerable to the influence of these procedures (Fig. 1). However, it should be emphasized that the inhibition of drinking was invariably observed in FW, 1/3 SW-adapted, 1/3 SW-exposed and SW-exposed eels after injection of hypertonic solutions compared with controls injected with isotonic saline according to the same protocol. It is generally accepted that, in mammals, the macula densa responds to increased levels of Na or Cl ions in the tubular urine by inhibiting renin release from the juxtaglomerular cells. Thus, the administration of hypertonic solution of NaCl results in a decrease of plasma AII level [5]. The inhibitory role of the macula densa in renin release appears to be active in birds also, because an increased load of NaCl decreased plasma renin activity in anesthetized chickens [23], and de- creased plasma AII level in the conscious quail [21]. In the present study, however, injection of 7% and 14% NaCl elevated plasma AII levels in FW eels and 1/3 SW-adapted eels. Since the primary factor that determines plasma AII level appears to be plasma renin activity [5], it is assumable that renin release was not inhibited, but rather stimulated by injection of the hypertonic NaCl solutions in the eel. Consistent with this assumption is a report that the macula densa does not exist in fishes [6]. It appears that an increase in plasma osmolality itself is stimulatory for renin release, because hypertonic sucrose consistently increased plasma AII levels in the eel as shown in the present study (Fig. 2), and increased plasma osmolality by dextran or human serum albumin, but not by NaCl, increased renin release in anesthetized dogs [24]. Thus, it appears that the injection of hypertonic NaCl solution increased renin release in the eel due to the absence of the macula densa. The present results seem to support the role of the macula densa in the control of renin release from the phylogenetic point of view. A direct measurement of renin release is being undertaken in the eel to substantiate this assump- tion. On the other hand, injection of 7% NaCl failed to cause significant increases in plasma AII level in 1/3 SW and SW-exposed eels. This result could be due to a depletion of stored renin in the juxtaglo- merular cells after exposure to 1/3 SW or SW, since it has been shown that plasma AII level increases after exposure to SW [8], as confirmed by the result of SW-exposed eels in the present study (Table 3). Responses of renin to external osmolal- ity have also been reported in the eel; plasma renin activity was increased by the transfer of eels from FW to SW [25], and decreased by the transfer from SW to dilute media [26, 27]. In summary, the present study showed that the drinking rate was rather inhibited, and plasma AII level was increased after an intra-arterial injection of hypertonic solutions of NaCl and sucrose in the eel. These results are in contrast to those obtained in terrestrial animals that drinking is induced by intravascular administration of hypertonic solu- tions in mammals, birds and reptiles, and plasma AII level is decreased by intravascular administra- tion of hypertonic NaCl solutions in mammals and birds. ACKNOWLEDGMENTS The authors express their appreciation to Dr. Tetsuya Hirano, Ocean Research Institute, University of Tokyo, for his valuable advice and critical reading of this manuscript. This investigation was supported in part by a Grant-in-Aid (574307) from the Ministry of Education, Japan. 10 12 13 Drinking and Angiotensin Responses in Eels 51 REFERENCES Fitzsimons, J. T. (1979) The Physiology of Thirst and Sodium Appetite, Cambridge Univ. Press, Cambridge, pp.32-94, p. 149 and pp. 158-165. Kaufman, S. and Peters,G. (1980) Regulatory drinking in the pigeon Columba livia. Am. J. Physiol., 239: R219-R225. Kobayashi, H. and Takei, Y. (1982) Mechanisms for induction of drinking with special reference to angiotensin II. Comp. Biochem. Physiol., 71A: 485- 494. Fitzsimons, J. T. and Kaufman, S. (1977) Cellular and extracellular dehydration, and angiotensin as stimuli to drinking in the common iguana Jguana iguana. J. Physiol. (London), 265: 443-463. Keeton, T. K. and Campbell, W. B. (1981) The pharmacologic alteration of renin release. Pharma- col. Rev., 32: 81-227. Sokabe, H. and Ogawa, M. (1974) Comparative studies of the juxtaglomerular apparatus. Int. Rev. Cytol. , 37: 271-327. Hirano, T. (1974) Some factors regulating water intake in the eel, Anguilla japonica. J. Exp. Biol., 61: 737-747. Takei, Y., Uemura, H. and Kobayashi, H. (1985) Angiotensin and hydromineral balance: With special reference to induction of drinking behavior. In “Current Trends in Comparative Endocrinology”. Ed. by B. Lofts and W. N. Holmes. Hong Kong Univ. Press, Hong Kong, pp. 933-936. Takei, Y., Kobayashi, H. and Hirano, T. (1979) Angiotensin and water intake in the Japanese eel, Anguilla japonica. Gen. Comp. Endocroinol., 38: 466-475. Beasley, D., Shier, D. N., Malvin, R. L. and Smith, G. (1986) Angiotensin-stimulated drinking in marine fish. Am. J. Physiol., 250: R1034-R1038. Carrick, S. and Balment, R. J. (1982) The renin- angiotensin system and drinking in the euryhaline flounder, Platichtys flesus. Gen. Comp. Endocrinol., 51: 423-433. Kobayashi, H. Uemura, H., Takei, Y., Itazu,N., Ozawa, M. and Ichinohe, K. (1985) Drinking in- duced by angiotensin II in fishes. Gen. Comp. Endocrinol., 49: 295-306. Malvin,R.L., Schiff,D. and Eiger,S. (1980) Angiotensin and drinking rates in the euryhaline killifish. Am. J. Physiol., 239: R31-R34. Hirano, T. and Hasegawa, S. (1983) Effects of angiotensins and other vasoactive substances on drinking in the eel, Anguilla japonica. Zool. Sci., 1: 106-113. 15 16 18 19 22 24 25 26 27 Yamaguchi, K. (1981) Effects of water deprivation on immunoreactive angiotensin II levels in plasma, cerebroventricular perfusate and hypothalamus of the rat. Acta Endocrinol., 97: 137-144. Hasegawa, Y., Nakajima, T. and Sokabe, H. (1983) Chemical structure of angiotensin formed with kidney renin in the Japanese eel, Anguilla japonica. Biomed. Res., 4: 417-420. Khosla, M. C., Nishimura, H., Hasegawa, Y. and Bumpus, F. M. (1985) Identification and synthesis of [1-Asparagine, 5-Valine, 9-Glycine] angiotensin I produced from plasma of American eel Anguilla rostrata. Gen. Comp. Endocrinol., 57: 223-233. Conover, W. J. (1980) Practical Nonparametric Sta- tistics, John Wiley & Sons, New York, 2nd ed. Skadhauge, E. (1981) Osmoregulation in Birds, Springer Verlag, Berlin, Heidelberg and New York, p. 4. Bentley, P. J. (1971) Endocrine and Osmoregula- tion, Springer Verlag, Berlin, Heidelberg and New York, p. 6 and pp. 218-226. Takei, Y., Okawara, Y. and Kobayashi, H. (1988) Control of drinking in birds. In “Progress in Avian Osmoregulation”. Ed. by M. R. Hughes and A. C. Chadwick. Leeds Philosophical and Literary Society Ltd., Leeds (in press). Takei, Y. and Hatakeyama, I. (1987) Changes in blood volume after hemorrhage and injection of hypertonic saline in the conscious quail, Coturnix coturnix japonica. Zool. Sci., 4: 803-811. Nishimura, H. and Bailey, J. R. (1982) Intrarenal renin-angiotensin system in primitive vertebrates. Kidney Int., 22: $185-S192. Hall, J. E. and Guyton, A. C. (1976) Changes in renal hemodynamics and renin release caused by increased plasma oncotic pressure. Am. J. Physiol., 231: 1550-1556 Sokabe, H., Oide, H., Ogawa, M. and Utida, S. (1973) Plasma renin activity in Japanese eels (Anguilla japonica) adapted to seawater or in dehydration. Gen. Comp. Endocrinol., 21: 160- 167. Henderson, I. W., Jotinsankasa, V., Mosely, W. and Oguri, M. (1976) Endocrine and environmental influences upon plasma cortisol concentrations and plasma renin activity of the eel, Anguilla anguilla L. J. Endocrinol., 70: 81-95. Nishimura, H., Sawyer, W.H. and Nigrelli, R. F. (1976) Renin, cortisol and plasma volume in marine teleost fishes adapted to dilute media. J. Endocri- nol., 70: 47-59. od eit ae Ltt fu ZOOLOGICAL SCIENCE 5: 53-60 (1988) A Method of Quantitative Analysis of Cell Migration Using a Computerized Time-lapse Videomicroscopy Nosuo ZAMA and Hipeki Katow! Laboratory of Computer Science and ‘Laboratory of Biology, Rikkyo University, Nishi-Ikebukuro, Tokyo 171, Japan ABSTRACT—A quantitative analysis of the cell migration in vitro has been realized by examining a distance of the migration performed by the cells, and the results are represented by a format called the “cell migration pattern”. We have composed a system which is capable to calculate the cell migration patterns, and to carry on such statistical tests as the parametrics as well as the non-parametrics to compare and evaluate the significant differences among the cell migration patterns. The system consists of an inverted phase contrast microscope, a video camera which is connected with the microscope, a monitor TV, a time-lapse video cassette recorder, a position analyzer, and a micro- computer. This system can trace 7 cells at a time and calculate above migration criterion. The system was applied to analyze the migratory behavior of the primary mesenchyme cells of the sea urchin blastulae in different culture conditions as a model case. Its desired functions were fully demonstrated © 1988 Zoological Society of Japan in the present study. INTRODUCTION The cell migration is one of the elemental processes in the morphogenesis of the multicellular animals, and the importance of the extracellular matrix as a substratum for the migration has long been recognized (e, g. [1-3]). However, despite the recent increasing demand that the mor- phogenetic movements, particularly the cell migra- tion, should be analyzed quantitatively [4, 5], it has not been fulfilled mainly because of technical difficulties. In order to obtain proper data for the quantitative analysis one has to incubate the cells in various culture conditions, to record and com- pare their behavior, and finally to examine whether there are any significant differences among them using appropriate statistical tests. One of the difficulties associated with the analysis, moreover, was that there have been few methods to organize raw data. In order to deal with these difficulties, it has been proposed that the cell migration is quantified by 1) examining a distance Accepted July 6, 1987 Received May 15, 1987 " To whom requests of reprints should be addressed. of the migration performed by the cells, 2) organizing these data into a proper format, such as the “cell migration pattern” [4, 5], and by 3) comparing the patterns statistically. However, those whole procedures required truly time- consuming laborious works to complete. Develop- ment of a new method, therefore, has been awaited to adequately and promptly handle those processes. Recently, the technology of computer graphics has been remarkably developed with the improve- ments of a computer performance along with the progress of the computer application techniques. Since this technology enables us to send the images to a computer as input data and process them in desirable manner for the researchers, we have composed a new system which processes the images of moving cells, and automatically calcu- lates the results in a short period. The system is capable to handle up to 7 cells simultaneously. The present system was applied to analyze the migration of the primary mesenchyme cells of the sea urchin, Pseudocentrotus depressus, in two different culture conditions as a model case. These culture conditions resulted in visually similar cell migration patterns [6] so that proper statistical 54 N. ZAMA AND H. Katow treatments were required. The present system has succeeded to evaluate the present model case, and to demonstrate the desired functions. MATERIALS The eggs of the sea urchin, Pseudocentrotus depressus, were collected by intracoelomic injec- tion of 0.5M KCl, and raised in an artificial sea water, Jamarin U (Jamarin Laboratory, Osaka) until the mesenchyme blastula stage at 20°C. The primary mesenchyme cells were separated from the mesenchyme blastulae with the method pre- sented previously [4, 5]. The cells were incubated in the artificial sea water containing 40 g/ml of horse plasma fibronectin, which was kindly pro- vided by Dr. M. Hayashi, Ochanomizu Women’s University, Tokyo. This culture condition was designated as an experiment group 1. Aliquots of cells were incubated in a culture medium which was composed of the same amount of fibronectin and 40 g/ml of a synthetic peptide, Arg-Gly-Asp- Ser, which is known to inhibit the fibronectin dependent cell migration with proper concentra- tion [7] (Experiment group 6). The experiment numbers were retained to be the same as that of the previous study [6] in this paper in order to keep the integrity of the data with that study. The synthetic peptide has been kindly provided by Dr. K. M. Yamada, National Institutes of Health, U. S.A. The effect of the synthetic peptide on fibronectin dependent cell migration has been studied previously in the sea urchin primary mesenchyme cells [6]. Those experiment groups were chosen as model case in this study since it has been known from the previous studies that they possess visually similar cell migration patterns (Fig. 6), so that the statistical examinations were strongly required to evaluate any significant differ- ences between them. The outcome of the ex- aminations, therefore, was to provide whether there was any effect of the synthetic peptide on the fibronection dependent primary mesenchyme cell migration in vitro. In order to apply the technology of the computer graphics, an inverted phase contrast microscope with Nomarski optics (Nikon, Tokyo) was con- nected with a video camera (WV-1500, National, Tokyo), a monitor TV (WV-5400, National, Tokyo), a time-lapse video cassette recorder (VHS type, AG-6010, National, Tokyo), a position analyzer (VPA-1100, Nippon Jimu Koki Co. Ltd., Tokyo), and with a microcomputer (CZ-802C, Sharp Co., Osaka) (Figs. 1 and 2). Although the system was capable to handle as many cells as the researchers desire at a time, regarding human errors that were predicted to happen more fre- quently correlated with the increasing number of cells to be handled at a time during a process to pick up the initial data, the number of cells to be Fic. 1. system. Images from a microscope are recorded with a time-lapse video cassette recorder (2) through a video TV camera (1) and displayed in a monitor TV (3). The images are also sent to a microcomputer (6) through a position analyzer (4). The images are superimposed in the computer CRT with the coordinates of the cells collected through a light pen (5). The composition of the cell migration analysis CRT display micro-computer system CZ 8022 TV camera position analyzer time-lapse video recorder Fic. 2. A chart indicating flows of informations with arrows in the system. microscope Quantitative Analysis of Cell Migration Bp) tracks of migrating cells cell migration pattern Fic. 3. handled at a time was set to be at most 7. After tracing the tracts of the cells in the computer display, cathode ray tube (CRT), the system prints out the results of the moving distances and tracts of individual cells, and stores these data in the floppy disks. Then the computer calculates the cell migration patterns from the stored data, and carries on statistic tests. The tests include a X 7- test and an F-test as the parametric tests, and two non-paramatric tests, a Kolmogorov-Smirnov test and a Mann-Whitney test (Fig. 3). METHODS AND RESULTS General explanation of the system The migration of the cells was recorded using the monochrome video camera by the time-lapse video cassette recorder through the inverted phase statistical tests a A summary of the information flow. contrast microscope which was installed with Nomarski optics. The recorded images of the migrating cells were played back chronologically, and the static images were obtained by manually ceasing the running of the video tapes at every 5 min of recorded time, which was indicated in the video display by a time-generator installed in the time-lapse video cassette recorder (Fig. 4a). Each location of the cells was detected on the static image displayed in the microcomputer CRT through the position analyzer by pointing the cells with a light pen one by one (Fig. 2). The light pen was installed in the position analyzer which had change-over switches and push buttons. The position analyzer also was able to carry on two operating modes, an axis-setting mode and a measurement-mode, which can be changed from a mode to another with the change-over switch. Prior to operating the analyzer, one has to initiate 56 N. ZAMA AND Pee = a, a=) ——_ (5. 959-c4_ = sa IES A —_ 2 = = SS = === ss Se = = SS ae EO : = 1-4 aot i =o 2 = = = Js Zz _ — 2 — SS =e = lee Fic. 4. (a) The superimposed images of the cells and H. Katow Pama ~s 1 oa ee, P a s BC x t ba ae % \ +... ; el ae a s ferent eel ; fee | C _y S : 2 | _—- KT / 3 Pe +5 is \ aN ‘ ., gh r eg \ ‘hg ar a ee: \ t nee Bs “on b their migration tracts shown with bars. The initial locations of the cells were indicated with crosses by the numbers. By reading the time displayed at the upper right corner of the CRT, the video tape was chronologically ceased to run manually to collect the new locations of the cells. (b) The cell migration tracts were extracted from the superimposed images in the CRT, and were printed out for further analysis. the procedure by setting the switch to the axis- setting mode and then point a place in monitored image with the light pen in order to set up the axis of the coordinate which was programmed to display with a cross in the computer CRT as well as in the monitor TV. After changing the mode to the measurement, a process of data input was started to repeat the pointing of original locations of each cell displayed in the computer CRT with the light pen (A data input process) (Fig. 4). The values of the coordi- nates of each location of the cells were sent to the microcomputer. The processing CZ-802C micro- computer superimposes any images coming from the external device, such as the video cassette recorder in this system, to the images already displayed in the computer CRT (Fig. 4a). The images of tracts of the migrating cells were extracted from the superimposed images in the computer CRT to the computer and were printed out after the completion of the data input process (Fig. 4b). The input raw data were stored in floppy disks and then used for calculation by the computer according to the program that will be mentioned elsewhere in this paper (Fig. 3). The calculated data including the statistically treated ones were finally printed out on the paper. Five-inch floppy disks were used to store the values of coordinates of the cells as well as the results of the process, and 3.5-inch floppy disks were used to store the program written with Hu-BASIC [8]. They were necessary for processing the data and the system to control the microcomputer [9]. Operating process The processing program for the microcomputer was designated in accordance with the following operating process. In this system every experiment was assigned numerically with an experiment group number (Group number) first, and as an experimental condition was identical with the others the group number of the experiment was defined to be the same as those. The numbers after the experiment group number in printouts showed that of trials in one experiment group (Experiment number) (Fig. 5). The computation was initiated according to the following steps. 1) To put the computer system on action, and load necessary computer program to the com- puter. 2) To switch on the video cassette recorder, and make the computer superimpose the picture of the recorded images of the cells in the computer CRT. 3) To locate original positions of the cells. As one makes a picking-up program run, the comput- er asks one the origin of the coordinate on the head of the first picture frame of the video display. Quantitative Analysis of Cell Migration Si, ABC D E F G 1 1 % 45,899701 2,8662313 1,.6747438 10.77¢33 1 1 2 14,242641 ,70266504 .56458997 35 {1 1 3 435.29052 2,.83@6575 2.5057077 15,264337 1 {1 4 1@.828427 6767767 ,56041984 4,2426407 1 1 3 30,083405 1.8802128 1.3275264 35.8309519 1 1 6 5.242641 .99266504 .67934708 2.236068 1 1 #7 67,28308 4.2051925 3,5019643 12.041595 1 £ 8 22.594553 1.4121596 ,80654566 3,6255513 1 1 9 30,028532 1.8767832 1.0592243 4.1231956 1 1 10 22,906114 1,4316321 1.1199926 2,236068 1 1 11 8,8284271 .5517767 .67232029 3,1622777 1 { 12 32.719362 2.0449601 1.3546211 7,2801099 $ 1 13 32,397817 2,0248636 1,0737949 3 { £ 14 20.307136 1.269196 .483506808 3.1622777 {1 1 15 9 3625 .49607837 2.236068 1 { 16 39,492@56 2,4682935 1.844927: 10 1 1 17 25,217478 1.5760923 .71689478 6 1 1 18 49.379369 3.0862106 2.1698789 6.7082039 1 1 19 13,828427 .8642767 .38592642 3 1 1 20 8 .3 .3 4.2426427 { 1 21 42.004057 2.6252535 1.3282911 12 1 1 22 98.621634 6,.1638521 5.97159942 5,3851648 1 $ 23 37,575963 2.3484977 1.8157804 3 1 1 24 32,14433 2.0090206 1.6028051 11,313708 1 4 25 22,4181) 1.4011319 1,1139828 4,4721359 1 2 1 24.021912 1,2643112 .65524001 3.6055513 { 2 2 74,485114 3,9202692 2,0900554 9.2195444 1 2 3 24,307136 1.2793229 74204446 7,9710678 Fic. 5. Print out of the initially treated data sent to “floppy disk 2” from “floppy disk 1”. A; group numbers. B; experiment numbers. C; point num- bers which are put onto every cells examined. In this case 25 cells were examined. D; actual total migration distances, which are the sum of total length of bars drawn between every points traced according to a cell’s migration tract. E; average migration distances performed by a cell during 100 min of examination period in this case. F; S. D. of the average migration distance. G; short cut dis- tance between the original locations and the last one. The coordinate was set to be x, y=30, 150 in this system (Fig. 4). The number of cells one can pick up from the frame at a time was anywhere up to 7 as has been stated before in this paper. The program was made so that one can input the coordinate values of the cells to computer as one taps a key of the input-keyboard after pointing the cells with the light pen, and the input was confirmed with the display of a cross on each point (the cell) in the computer CRT (Fig. 4). The crosses retained at the original locations of the cells through the process until completion of the data transfer to the secondary floppy disks for restoration for further data processing (Fig. 3). In a secondary picture frame of the video display, 5 min of computer CRT display-time after the first picture frame, one points the cell’s new locations following the numbers displayed near the crosses represented the original locations of each cell, and tapping the key at each time after pointing the new locations of the cells with the light pen, which was confirmed by a bar drawn between the original point and the new one (Fig. 4). This procedure was repeated through the last picture frame which was set initially by researchers as an observation period. For instance, if one is to collect the locations of the cells every 5 min for 100 min of the observation period, the number of the picture frames will be 21 including the first one, a time zero. The process is needed to be repeated according to the number of cells to be examined, such as 5 times when one handles 35 cells in total and 7 cells at a time. During the procedure the computer displayed the tract of each cell with the bars of different colors according to the cells (Fig. 4a). The total number of cells examined and that of input procedure were confirmed by the computer momentarily on the CRT display at the end of each process. The data picking-up program was termi- nated by tapping a key of the keyboard. The coordinate values of the cells obtained by the above procedure were stored in 5-inch floppy disks (Floppy disk 1). 4) Calculation. The data in the “Floppy disk 1” were the starting point of the following proce- dures. The first step of the procedures was “calculation of moving distance” as was labeled accordingly in a block of Figure 3 chart. This step consisted of the following two substeps. a) To calculate the migration distances of the cells from the coordinates of them at each mo- ment. b) To put all the data of different disks together into one disk (Floppy disk 2) in order to organize the data as the same experiment group (Fig. 5). The second step signed “drawing of cell migra- tion pattern” in a block of Figure 3 chart was started from the floppy disk 2. This step consisted of the following two substeps. a) To draw a histogram of the moving dis- tances of the cells in an experiment. b) To convert the histogram to the cell migra- tion pattern. The floppy disk 3 stored all the histograms and the cell migration patterns carried on in a series of experiments (Fig. 6). 58 N. ZAMA AND H. Katow EXPGRa B ea D §.2113284 9.3@1S515 e@,2iis3e4 9,9431522 3.4895372 17,9854701 20,6122¢8 3,3550594 39,4679498 {5.199718 {1.003636 53,666566 A H t ! 1 19,348291 6,4542622 73,914957 { 6.5619858 1,.3036692 79,576923 1 8.1152845 6.300304 937,592307 { 3.4807692 4,.9808435 91,173076 t 48076923 .83271673 91,6939846 { 1.23 2,155@635 92,993846 ! »48076923 183271673 93.384615 1 {.4807692 {.6528395 94,865384 ! 1,4423077 2,49815@2 96,397692 1 @ Q 96,307692 { 695153845 1,66534334 97,25923 { @ @ 97,28923 { Q @ 97,26923 t 2.25 2.2776084 939,51923 1 @ Q 39.51923 1 @ ] 99,51923 { 148076923 ,83271673 99,999999 (pm) %MIGRATION? 82, 142299 Fic. 6. The cell migration patterns of experiment group I(a) and 6(b). A B 5 0 5 WSLZS7ELS: wAlSS72tS.nneiestate 5 3.29036825 2.6528895 4,723357 S $.9983507 .9@163935 13.82172! ) {6.086066 3,58€0655 29.997787 c) 16.7592049 1.997951 46,659935 5 {2.182377 .232377 £9,342213 5 {4,446722 1,9467215 73.298934 . 3,2940164 .6659936 79.972935! 5 2.6246721 1,552278 31,557623 5 3.7233607 {.3256393 91.290983 § 1.3735 1,375 $2. 165983 6 2,9893443 ,39934425 96.955328 5 2.069672! .43922785 38,.i25 § .625 625 38.75 6 Q Q 98,75 6 1,23 1.25 100 6 @ a 100 ) @ Q 122 i) a a 102 6 @ Q 120 6 Q Q 122 0 50 (pm) b “MIGRATION= 33, 276629 In both (a) and (b) the top numeral columns indicate and experiment group number (A), a proportion of the cells clarified in a group indicated with percent (B), S. D. of the B(C), and an integral from the first cell group indicated with percent (D). The bottom graphs indicate the cell migration patterns with S. D. s, which are shown with vertical bars. The very bottom of the graphs indicates the percent migration values. EXPERIMENT GROUP NUMBER 1 TOTAL NUMBER OF CELLS 114 MEAN 23.718702 UNBIASED VARIANCE 356.19723 EXPERIMENT GROUP NUMBER 6 TOTAL NUMBER OF CELLS 142 MEAN 29, 745062 UNBIASED VARIANCE 225,69918 VARIANCES ARE DIFFERENT FF= 1,39 FQ= 1,5777353 MEANS ARE DIFFERENT Fic. 7. The print out of the parametric test results. FF and FO indicate the results of F-test. The conclu- sion of the F-test is indicated with terse statement. 5) Statistical treatments. For the parametric tests, the following procedures were pursued (Fig. 7). a) To test the equality of the variances by an F-test. b) If the two variances are equal, after estima- tion of the population variance a t-test was applied to examine the equality of the means. c) If the two variances are unequal, a Welch test was applied to examine the equality of the means. In case of the primary mesenchyme cells, however, it was regarded to be inadequate to evaluate above data solely based on the parametric statistics which premise a standardized cell migra- tion behavior [10, 11] due to rather poorly under- stood migratory behavior of the cells. Under the Quantitative Analysis of Cell Migration 59 consideration, in addition to above two parametric tests further two non-parametric tests, a Kolmo- gorov-Smirnov test and a Mann-Whitney test were adopted as well (Fig.8). Therefore, one can examine the equality of variances of two results and that of the means [10, 11]. The statistical tests were applied to both the data in the “Floppy disk 2” and “Floppy disk 3”, and were particularly useful to clarify any potential and significant differences among the cell migration MANN-WHITNEY TEST Ri= 9329, 264 R2= 11770. 736 V@= 3279,2639 SIG= 496,@124 202 4,.2381368 MANN-WHITNEY TEST (5% SINGLE-SIDE) DIFFERENCE 18 SIGNIFICANT KOLMOGOROV SMIRNOV TEST 2(1) 2 +23 7,.3918663E-02 61313134 + 24646227 »23758879 + 26355122 + 20734711 » 14403373 + 12300126 . 10086223 1,6128625E-02 2.1863177E-23 -1.1899433E-02 -1,8173077E-02 -2,4423077E-02 -1,4807692E-02 ~2.7307692E-02 -2,7307692E-92 -4,807692E-03 @ MAX= ,26355122 KAI= 13,891849 KOLMOGOROV-SMIRNOV TEST(S%-SINGLE) DIFFERENCE IS SIGNIFICANT Fic. 8. The print out of two non-parametric tests, a Mann-Whitney test and a Kolmogorov-Smirnov test. In Mann-Whitney test R1, R2 indicate orders in experiment group 1 and 6, respectively. Conclu- sion of the calculation with 5% error was printed out after listing intermediate calculations (UO, SIG, ZO) with terse statement. In Kolmogorov- Smirnov test intermediate calculations were dis- played after Z(1), and MAX is smaller than KAI, which means difference of experiment group 1 and 6 is approved as is indicated with the last terse statement as a conclusion. patterns. One of these statistical treatments was available by tapping a key of the keyboard. The results with 5% error were displayed instan- taneously in computer CRT and printed out upon researcher’s request (Figs. 7 and 8). In the present model case the experiment groups 1 and 6 were significantly different with each other despite the visually similar cell migration patterns and rather close percent migration values (Fig. 6). Therefore, it was concluded that the synthetic peptide | Arg-Gly-Asp-Ser interfered _ the fibronectin dependent migration of the primary mesenchyme cells in this concentration in vitro. DISCUSSION Two methods have been reported to input the image informations into the computer in order to process the informations for particular analysis. A method is to convert whole picture frame of images to the digital informations, and then input them to the computer (the digitization method) [8, 9, 12]. The other is to pick up only necessary informations from each image displayed with approriate extracting devices (the direct extracting method) [12]. Applying the former method, one carries on a great variety of processes, since the whole in- formations on the images can be referred. The amount of informations to be processed by this method, however, becomes enormous and there- fore costly hardwares are required to be deployed. Moreover, the input data obtained often contain unnecessary informations as well. The latter method which was adopted in the present research was shown to be rather simple. The performance of the system was indicated to be sufficient for the present research aims. For the statistical tests, the non-parametric tests were used to examine the equality of the medians of two different kind of distributions, and the parametric tests were used to examine the variances and the means of the data. The medians, however, were not always adequate to adopt as a sole statistical criterion, because, if small portion of an experiment group which involves extremely large migration distances in a parameter, the median does not represent proper feature of the 60 N. ZAMA AND H. parameter. In this case seeking the mean of the data which is calculated based on the parametric tests is to be more accurate [9]. In the model case the statistically proved inter- ruptive effect of the synthetic peptide onto fibronectin dependent primary mesenchyme cell migration has been supported by further observa- tions performed in higher concentrations of the synthetic peptide [6]. Therefore, an adequacy of the present system’s function was fully proved. The present computer program will be available from our laboratories upon researchers requests. ACKNOWLEDGMENT This research has been supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture, Japan to H. K. (Nos. 61540535 and 61304009). REFERENCES 1 Hynes, R. O. (1981) Fibronectin and its relation on cellular structure and behavior. In “Cell Biology of Extracellular Matrix”. Ed. by E. D. Hay, Plenum Press, New York, pp. 295-334. 2 Katow, H. (1987) Sea urchin primary mesenchyme cells. In “Developmental Systems and Cell Be- havior”. Ed. by H. Katow and H. Mizoguchi, Baifu- kan, Tokyo, pp. 3-33. (In Japanese) 3 11 12 Katow Solursh, M. (1985) Migration of sea urchin mesen- chyme cells. In “Developmental Biology: A Com- prehensive Synthesis”. Vol. 2. Ed. by L. W. Brow- der, Plenum Press, New York, pp. 391-432. Katow,H. and Hayashi,M. (1985) Role of fibronectin in primary mesenchyme cell migration in the sea urchin. J. Cell Biol., 101: 1487-1491. Katow, H. (1986) Behavior of sea urchin primary mesenchyme cells in artificial matrices. Exp. Cell Res., 162: 401-410. Katow, H. (1987) Inhibition of cell surface binding of fibronectin and fibronectin-promoted cell migra- tion by synthetic peptides in sea urchin primary mesenchyme cells in vitro. Dev. Growth Differ., 29: 573-589. Yamada, K.M., Akiyama,S.K., Hasegawa, E., Humphries, M.J., Kennedy,D.W., Nagata, K., Urushihara, H., Olden, K. and Chen, W.-T. (1985) Recent advances in research on fibronectin and other cell attachment proteins. J. Cell. Biochem., 28: 79-97. Manual of Hu-BASIC (1984) Sharp Co. Ltd., Osaka. (In Japanese) Manual of microcomputer CZ-802C (1984) Sharp Co. Ltd., Osaka. (In Japanese) Kendall, M.G. and Stuart, A. (1973) The Ad- vanced Theory of Statistics. Vol. 2, 4th ed., Charles Griffin, London. Gibbons, J.D. (1985) Nonparametric Statistical Inference. 2nd ed., M. Dekker, New York. Foley, J. and Van Dam, A. (1982) Fundamentals of Interactive Computer Graphics. Addision-Wesley Reading, Massachusetts. ZOOLOGICAL SCIENCE 5: 61-68 (1988) © 1988 Zoological Society of Japan Fine Structure of Filiform Papillar Epithelium from the Tongue of the Frog, Rana nigromaculata SHIN-ICHI IWASAKI, KEN MIyaTA and KAN KOBAYASHI Department of Anatomy, School of Dentistry at Niigata, The Nippon Dental University, Niigata 951, Japan ABSTRACT— Light and transmission electron microscopies were used to investigate the ultrastructure of filiform papillar epithelium from the tongue of the frog, Rana nigromaculata. A large part of the epithelium was composed of cells which contained many small, electron-dense granules. Ciliary cells and mucous cells were situated among these cells. Plasma cells and cells which contained many fat dropletes were rarely seen within the epithelium. The possible functional roles of these epithelial cells were discussed. INTRODUCTION Physiologists frequently use frog tongues in their studies [1-3] because of the many taste organs or sensory discs on the fungiform papillae which are distributed among the filiform papillae on the dorsal surface. Electron microscopic studies of the structure of the sensory papillae include those of Graziadei [4], Graziadei and DeHan [5], and Diring and Andres [6], all of whom have reported that almost the entire surface of each sensory disc is covered with microvilli. However, more recent studies [7-9] have revealed that most of the surface of the sensory discs is covered with a honeycomb structure which can be seen after removal of the layer of superficial mucus. Howev- er, reports describing the features of the lingual epithelium apart from the sensory papillae are quite few in number [7-9]. In particular, the histological structures in the lingual epithelium of the frog have been practically ignored. In the present study, light and transmission electron microscopies were used to investigate the ultra- structure of the tongue of the frog, Rana nigroma- culata. Scanning electron microscopic observations have revealed significant differences in the dorsal surface of the tongue between frogs [7-9] and Accepted August 11, 1987 Received May 15, 1987 mammals [10-12]. These differences appear to be related to differences in the function of the tongue. The present study was also designed to clarify the cellular features of the mucosal epithelium, and a brief discussion of its possible functional roles is included. MATERIALS AND METHODS Tongues from five male and five female adult frogs, Rana nigromaculata, were used in the present study. These frogs were collected in the area around the city of Niigata in June and July of 1984. The animals were perfused from the heart with Karnovsky fixative which contained glutaral- dehyde and paraformaldehyde [13], under MS- 222 anesthesia. The tongue was then removed and refixed in the same fixative. After rinsing in 0.1M cacodylate buffer, the samples were post- fixed in phosphate-buffered 1% osmium tetroxide solution [14] at 4°C for 1.5 hr. Postfixation was followed by dehydration, embedding in Epon- Araldite, ultrathin sectioning and double staining with uranyl acetate and lead citrate. The speci- mens were then observed under a transmission electron microscope (Hitachi H-500). Thick sec- tions from the blocks embedded in Epon-Araldite were stained with 0.2% toluidine blue in 2.5% Na,CO3. Micrographs of the sections, taken with an Olympus BH-2 light microscope, were com- pared with the transmission electron micrographs. 62 S. IwasakI, K. MitayA AND K. KoBAyYASHI RESULTS When sections of lingual tissue embedded in Epon-Araldite were examined by light microscopy (Fig. 1), the dorsal mucosa of the tongue was seen to be composed of filiform and fungiform papillae. The filiform papillae were somewhat smaller than the fungiform papillae; however, their number was larger than that of fungiform papillae. The epithelium of the upper part of the filiform papillae was thicker than that of the basal part. Translu- cent cells which were filled with mucus occupied about 40% (by volume) of the epithelium. A high proportion of mucous cells was seen particularly in the upper area of the filiform papillae. The connective tissue of the lamina propria and the smooth muscle penetrated deeply into the center of each papilla. Relatively large numbers of the glandular structures were distributed within the lamina propria. By transmission electron microscopy (Figs. 2- 8), there were predominantly three types of epithelial cells observed in the mucosal epithelium of the tongue dorsum. One type of epithelial cell contained electron-dense, oval or round granules; another type of cell was the mucous cell; and the third type was the ciliated cell. Cells containing electron-dense granules occupied over 50% (by € , ‘ b } A 4 @: Foal ae ee Light micrograph of the lingual dorsal epithelium in Epon-Araldite-embedded tissue Fic. 1. volume) of the epithelium, and were situated over the entire area of the filiform papillar epithelium. Mucous cells occupied about 40% (by volume) of the filiform papillar epithelium. However, the volume of each mucous cells was several times larger than that of cells which contained electron- dense granules. Therefore, the actual number of mucous cells was much lower. These cells were located mainly on the upper part of the filiform papillae. Ciliated cells occupied about 5% (by volume) of the filiform papillar epithelium. These cells were scattered over the entire epithelium. The basal lamina was intercalated between the epithelium and the lamina propria throughout the mucosa of the filiform papillae. Some of the cells with electron-dense granules, located in the upper area of the filiform papillar epithelium, contained these granules only in the region just beneath the free surface of the cells (Fig. 2). In other cells, these granules were distributed throughout the cytoplasmic area (Fig. 3). In both types of cells, the nucleus was located in the basal or central area of the cells (Figs. 2 and 3). Some cells contained not only many electron- dense, oval or round granules but also a few electron-lucent granules (Fig. 2). Rough-surfaced endoplasmic reticulum was well-developed in the cytoplasm of some granular cells (Figs. 2 and 3). from a frog, Rana nigromaculata. Fi: filiform papilla, Fu: fungiform papilla, CT: connective tissue, SM: smooth muscle, Gl: glandular structure, arrows: mucous cells. Ultrastructure of Tongue Epithelium of Frog 63 pre PX, x “A Fic. 2. Transmission electron micrograph of the epithelial cells in the upper part of a lingual filiform papilla from a frog, Rana nigromaculata. N: nucleus, rER: rough-surfaced endo- plasmic reticulum, dG: electron-dense granules, IG: electron-lucent granules, arrow: micro- ridges. Cells in which the cytoplasm was filled with mucous granules were frequently observed among the epithelial cells which contained electron-dense granules (Fig. 4). In the basal part of the filiform papillae, the thickness of the epithelium was somewhat re- duced. Cells which contained many electron- dense, small granules also contained a few elec- tron-lucent granules. These cells mainly occupied the basal region of the epithelium. In this area, cells in which the cytoplasm was filled with mucus were rarely observed (Fig. 5). In both the upper and basal regions of the filiform papillae, microridges, which have been formerly demonstrated by scanning electron mi- croscope [9], were widely found on the free surface of the granular cells (Figs. 2-4, arrows). The cell surface which faced adjacent cells had abundant cellular processes (Figs. 3-5). Fic. 3. Transmission electron micrograph of cells which contain a large number of electron-dense granules in the epithelium on the upper part of a lingual filiform papilla. N: nucleus, rER: rough- surfaced endoplasmic reticulum, dG: electron- dense granules, CP: cellular processes, arrow: mi- croridges. 64 S. IwasakI, K. MitayA AND K. KoBAYASHI Fic. 4. Electron-dense granular cells and a mucous cell in the upper part of a filiform papilla. N: nucleus, trER: rough-surfaced endoplasmic reticulum, dG: electron-dense granules, mG: mucous granuies, arrow: microridges. Ciliated cells were scattered among these granu- lar cells in both the upper and basal regions of the filiform papillar epithelium. The ciliated cells had cilia and microvilli, also previously revealed by scanning electron microscopy [9], on their free surfaces. Just beneath the free surface of these cells, basal bodies (Figs. 6 and 7, BB) and elec- tron-dense vesicles (Fig. 7, arrow) can be recog- nized. Many lysosomes, mitochondria, rough- surfaced endoplasmic reticulum and free ribo- somes were distributed throughout cytoplasm (Figs. 6 and 7). In the basal region of the epithelium, the basal lamina was located between the epithelium and the lamina propria (Fig. 8, arrow). In the basal region of the cytoplasm of granular cells, there were a few electron-dense granules and the rough-surfaced endoplasmic reticulum was well-developed (Fig. 8). In the basal portion of the cytoplasm of ciliated cells, mitochondria and glycogen granules were abundant (Fig. 8). Plasma cells (Fig. 9) and cells which contained many fat droplets (Fig. 10) could also be found in the epithelium, although they were very few in number. 2 a Fic. 5. The epithelium of the basal part of a filiform papilla. N: nucleus, dG: electron-dense granules, IG: electron-lucent granules, Ci: ciliated cell, CP: cellular processes. Ultrastructure of Tongue Epithelium of Frog Fic. 6. A ciliated cell in the filiform papillar epithelium. 65 rER: rough-surfaced endoplasmic reticulum, BB: basal bodies, Ly: lysosomes, CP: cellular processes, thick arrow: cilia, thin arrow: microvilli. Fic. 7. Higher magnification of the free-surface side of a ciliated cell. BB: basal bodies, Ro: ciliary rootlet, M: mitochondrion, G: glycogen granules, arrows: electron-dense vesicles. DISCUSSION Anuran tongues play an important functional role in the uptake of food, in the sense of taste [1- 3], and in the secretion of mucus [9, 10]. In mammals, tongues function mainly as the organ used for uptake of food and for the sense of taste [4], and are not so important for secretion of mucus. Salivary glands are the main mammalian organs which serve this function. Furthermore, in a few species of reptiles [15, 16], mucous glands are located in a restricted area of the tongue, and the taste organs are located on various areas of the oral mucosa [16, 17]. Thus, it seems that there may be variations in the function and structure of the tongue based on the evolution of animals. The present study indicates that mucous cells occupy about 40% (by volume) of the filiform papillar 66 S. Iwasakl, K. MitayA AND K. KoBAyYASHI Fic. 8. Basal region of a granular cell of a filiform papilla. rER: rough-surfaced endoplasmic reticulum, dG: electron-dense granules, M: mitochondria, G: glycogen granules, arrow: basal lamina. Fic. 9. A plasma cell located within the epithelium of a filiform papilla. N: nucleus, rER: rough-surfaced endoplasmic reticulum, M: mitochondria. epithelium of the frog. Furthermore, the glandular structures, which are analogous to the lingual glands of mammals, were observed under light microscope in the lamina propria beneath the filiform papillae. These results demonstrate that the lingual dorsal mucosa of the frog is the main organ for the secretion of mucus. The present study also indicates that a large part of the epithelium of the filiform papillae consists of cells which contain many electron-dense, oval or round granules. The functional role of these granules is not obvious, but there are three possibilities. One is that these granules may be immature forms of mucous granules; another is that they may be serous granules. Yet still other possibility is that they may be organelles which are analogous to the Paneth cell granules of the gastrointestinal tract [18]. The results of the Ultrastructure of Tongue Epithelium of Frog 67 Fic. 10. A cell with many fat droplets within the epithelium of a filiform papilla. N: nucleus, F: fat droplets. 1 c present study show that some of the lingual epithelial cells contain both electron-dense and electron-lucent granules. However, there was no evidence that there were any transitional stages between electron-dense granules and mucous granules. Thus, we are inclined to suggest that the electron-dense granules are a different type of granule from the mucous granule. The fine structure and electron-density of these granules seem to be similar to those of serous granules observed in the salivary glands of mammals [19- 21]. However, the exact function of these granules remains to be elucidated. The small electron- lucent granules found between electron-dense, oval or round granules may be secretory versions of electron-dense granules. Mature mucous cells, in which almost all the cytoplasm is filled with mucous granules, are similar to the goblet cells found in the gastrointes- tinal tract and trachea of mammals [22, 23]. Ciliated cells were observed to contain many lysosomes. This indicates that these ciliated cells may be involved in phagocytosis, as are similar cells in the human trachea [24, 25]. Thus, the lingual epithelium may have some functional role in addition to its role in the sense of taste. In mammals, a greater part of the lingual dorsal epithelium is stratified squamous epithelium, and 2 a various degrees of keratinization of the epithelium have been observed in different areas of the tongue [26-28]. In frogs, on the other hand, keratinization was not recognized at all in the lingual epithelium, and the epithelium was com- posed of various kinds of cells, i.e., electron-dense granular cells, ciliated cells, and cells of connective tissue origin. These differences may be partially based on the undifferentiated state of the frog’s tongue as opposed to that of the mammalian tongue. It is possible that the lingual dorsal epithelium of the frog contains cells which are similar to those located in the mucosal epithelium of the gastrointestinal tract and/or the trachea of mammals. This hypothesis will be examined in more species of anurans in the future. In our previous studies using scanning electron microscopy [8, 9], we clearly showed that micro- ridges are widely distributed on the cell surface of the filiform papillar epithelium. The present study indicates that the granular cells have both micro- ridges on the free surfaces of cells, and cellular processes on the other surfaces which face the adjacent cells. Thus, the microridges may be the result of an altered pattern of arrangement of these cellular processes. The cellular processes seem to function as connecting structure between adjacent cells [29]. As suggested by Sperry and Wassersug 68 S. Iwasaki, K. MitayA AND K. KoBAyYASHI [30], microridges or cellular processes on the free surface probably facilitate the spreading and hold- ing of mucus. 15 REFERENCES Taglietti, V., Maffini, S. and Casella, C. (1971) The recovery cycle of gustatory fibres during chemical stimulation of the tongue. Arch. Sci. Biol., 55: 155- 164. Sato, T and Beidler, L. M. (1975) Membrane resist- ance change of the frog taste cells in response to water and NaCl. J. Gen. Physiol., 66: 735-763. Akaike, N., Noma, A. and Sato, M. (1976) Elec- trical responses of frog taste cells to chemical stimuli. J. Physiol., 254: 87-107. Graziadei, P.P.C. (1969) The ultrastructure of vertebrate taste buds. In “Olfaction and Taste”. Ed. by C. Pfaffmann, Rockefeller Univ. Press, New York, pp. 315-330. Graziadei, P. P.C. and DeHan,R.S. (1971) The ultrastructure of frogs’ taste organs. Acta Anat., 80: 563-603. Diiring, M. v. and Andres, K. H. (1976) The ultra- structure of taste and touch receptors of the frog’s taste organ. Cell Tissue Res., 165: 185-198. Jeager, C. B. and Hillman, D. E. (1976) Morpholo- gy of gustatory organs. In “Frog Neurobiology”. Ed. by R. Linal and W. Precht, Springer-Verlag, Berlin, pp. 588-606. Iwasaki, S. and Sakata, K. (1985) Fine structure of the lingual dorsal surface of the bullfrog. Okajimas Folia Anat. Jpn., 61: 437-450. Iwasaki, S., Miyata, K. and Kobayashi, K. (1986) Studies on the fine structure of the lingual dorsal surface in the frog, Rana nigromaculata. Zool. Sct., 3: 265-272. Svejda, J. and Skach, M. (1971) Die Zunge der Ratte im Raster-Elektronenmikroskop (Stereo- scan). Z. mikrosk.-anat. Forsch., 84: 101-116. Steflik, D.E., Singh, B.B., Mckinney, R. V. Jr. and Boshell, J. L. (1983) Correlated TEM, SEM, and histological observations of filiform papillae of the cow tongue. Acta Anat., 117: 21-30. Iwasaki, S., Miyata, K. and Kobayashi, K. (1987) Comparative studies of the dorsal surface of the tongue in three mammalian species by scanning electron microscopy. Acta Anat., 128: 140-146. Karnovsky, M.J. (1965) A __ formaldehyde- glutaraldehyde fixative of high osmolality for use in electron microscopy. J. Cell. Biol., 27: 137A-138A. Millonig, G. (1961) Advantages of a phosphate buffer for OsO, solutions in fixation. J. Appl. Physics, 32: 1637. Iwasaki, S. and Miyata, K. (1985) Scanning electron microscopy of the lingual dorsal surface of the 18 19 20 21 22 23 24 25 26 27 28 29 30 Japanese lizard, Takydromus tachydromoides. Oka- jimas Folia Anat. Jpn., 62: 15-26. Schwenk, K. (1986) Morphology of the tongue in the tuatara, Sphenodon punctatus (Reptilia: Lepido- sauria), with comments on function and phylogeny. J. Morphol., 188: 129-156. Iwasaki, S., Miyata, K. and Kobayashi, K. (1985) Fine structure of the oral epithelial cell surface in the Japanese lizard, Takydromus tachydromoides. Jpn. J. Oral Biol., 27: 956-964. Cheng, H. (1974) Origin, differentiation and renew- al of the four main epithelial cell types in the mouse small intestine. IV. Paneth cells. Am. J. Anat., 141: 521-536. Hand, A. R. (1971) Morphology and cytochemistry of the Golgi apparatus of rat salivary gland acinar cells. Am. J. Anat., 130: 141-158. Riva, A. and Riva-Testa, F. (1973) Fine structure of acinar cells of human parotid gland. Anat. Rec., 176: 149-166. Ichikawa, M. and Ichikawa, A. (1977) Light and electron microscopic histochemistry of the serous secretory granules in the salivary glandular cells of the Mongolian gerbil (Mongolian meridianus) and rhesus monkey (Macaca irus). Anat. Rec., 189: 125-140. Cheng, H. (1974) Origin, differentiation and renew- al of the four main epithelial cell types in the mouse small intestine. II. Mucous cells. Am. J. Anat., 141: 481-502. Dalen, H. (1983) An_ ultrastructural study of tracheal epithelium of the guinea-pig with special reference to the ciliary structure. J. Anat., 136: 47- 67. Krsti¢, R. V. (1984) Illustrated Encyclopedia of Human Histology. Springer-Verlag, Berlin, pp. 424-425. Fawcett, D. W. (1986) A Textbook of Histology, 11th ed. W. B. Saunders Co., Philadelphia, pp. 16- 19. Cane, A. K. and Spearman, R.I.C. (1969) The keratinized epithelium of the house-mouse (Mus musculus) tongue: its structure and histochemistry. Arch Oral Biol., 14: 829-841. Farbman, A. I. (1970) The dual pattern of kerati- nization in filiform papillae on rat tongue. J. Anat., 106: 233-242. Iwasaki, S. and Miyata, K. (1985) Light and trans- mission electron microscopic studies on the lingual dorsal epithelium of the musk shrew, Suncus muri- nus. Okajimas Folia Anat. Jpn., 62: 67-88. Krsti¢é, R. V. (1979) Ultrastructure of the Mamma- lian Cell. Springer-Verlag, Berlin, pp. 238-239. Sperry, D. G. and Wassersug, R. J. (1976) A pro- posed function for microridges on epithelial cells. Anat. Rec., 185: 253-258. ZOOLOGICAL SCIENCE 5: 69-76 (1988) © 1988 Zoological Society of Japan Myoglobin of the Shark Galeus nipponensis: Identification of the Exceptional Amino Acid Replacement at the Distal(E7) Position and Autoxidation of Its Oxy-form TOMOHIKO SUZUKI, REIKO MURAMATSU, TOMOMI KISAMORI and TAKAHIRO FURUKOHRI Department of Biology, Faculty of Science, Kochi Univesity, Kochi 780, Japan ABSTRACT—Myoglobin was isolated from red muscle of the shark Galeus nipponensis and its spectral properties were examined. The spectrum of Galeus oxymyoglobin (MbO,) was similar to those of mammalian myoglobins, while that of metmyoglobin was rather different. The partial amino acid sequence around E-helix region of Galeus myoglobin was determined with the aid of sequence homology. The distal (E7) histidine, which is widely conserved in mammalian hemoglobins and myoglobins, was replaced by glutamine in the globin of Galeus. The autoxiataion rate of Galeus MbO, was examined in 0.1 M buffer at 25°C over pH ranges of 4.5-10.5. Galeus MbO, was extremely unstable between pH 6 and 10.5, and the pH dependence of autoxidation was much smaller than that of mammalian MbO,. This property may be partly due to the absence of a distal (E7) histidine in Galeus myoglobin. INTRODUCTION The distal (E7) histidine is one of the most important residues and is widely conserved in vertebrate hemoglobins and myoglobins. This residue is capable of forming a hydrogen bond to the bound dioxygen and stabilizing it [1], as well as of protecting the FeO, bonding from easy access of solvent just like a gate to the heme pocket[2]. In vertebrate myoglobins so far sequenced, only five myoglobins have the exceptional amino acid replacement at position E7. Two are from Asian and African elephants, and the remaining three are from the sharks belonging to Triakididae; Mustelus antarcticus [3], Galeorhinus australis [4] and Galeorhinus japonicus [5]. In all cases, the distal (E7) histidine is replaced by glutamine. Therefore comparison of these myoglobins with usual myoglobins would provide some new in- formation on the role of the distal (E7) residue in a myoglobin molecule. Accepted July 8, 1987 Received June 22, 1987 In this paper, we report the isolation procedure, spectral characteristics, pH dependence for auto- xidation and partial amino acid sequence of myoglobin from the shark Galeus nipponensis (Scyliorhinidae). Galeus myoglobin has been shown to have the distal (E7) glutamine in place of histidine, and to have the unique properties for spectrum and autoxidation. MATERIALS AND METHODS Materials Sephadex G75 was a product of Pharmacia. DEAE-cellulose was purchased from Whatman (DE32, microgranular form). TPCK-treated tryp- sin and chymotrypsin were purchased from Worth- ington. The reagents for sequence determination were of Sequanal grade from Nakarai Chemicals Ltd. Myoglobin preparation Native oxymyoglobin (MbO,) and metmyoglo- 70 T. Suzuki, R. MuRAMATSU et al. bin (metMb) from Galeus nipponensis were 1iso- lated directly from red muscle according to our previous method [5, 6]. All procedures were carried out at low temperature (0-4°C) as far as possible. The water extract was fractionated with ammonium sulfate between 60 and 95% saturation at pH8.0. The crude myoglobin solution was passed through a Sephadex G75 column equili- brated with 5mM Tris-HCl buffer (pH 8.7) to separate myoglobin from hemoglobin. The myo- globin fraction was then applied to a DEAE- cellulose column (2X15 cm) equilibrated with 5 mM Tris-HCl buffer (pH 8.7), and eluted with a linear gradient of 15 mM (250 ml) to 100 mM (250 ml) Tris-HCl buffer (pH 8.5), to separate MbO; completely from metMb. The myoglobins thus obtained were dialyzed against 5mM_ Tris-HCl buffer (pH 8.5), and were kept at low temperature (0-4°C) until use. The myoglobin concentration was determined spectrophotometrically after con- version to cyanometmyoglobin, using extinction coefficient of 11.3mM~'cm™! at 540nm [7]. Absorption spectrum was recorded in a Hitachi 220A spectrophotometer. Partial amino acid sequence determination The major MbO, was selected for sequence determination. The methods of removal of heme, carboxymethylation of free cysteine and enzymatic digestion were the same as previously described [8, 9). The whole protein was digested with trypsin for 5 hr at 37°C. The tryptic peptides were purified by high-performance liquid chromatography (HPLC) (Hitachi 655) with a linear gradient of acetonitrile in 10mM ammonium acetate as solvent at room temperature. The column (2.6150 mm) was packed with Lichrosorb RP8 (Merck). The whole protein was also digested with chymotrypsin for 5 hr at 37°C. The chymotryptic peptides were first fractionated by HPLC (RP8), and each fraction (2 ml each) was purified further by HPLC (Cosmosil 5Cig-P, 4.6 150mm, Nakarai Chemicals Ltd.) with a linear gradient of acetonitrile in 0.1% trifluoroacetic acid (TFA). Peptides were routinely hydrolyzed with TFA/ HCI (1/2, v/v) containing 0.02% phenol at 170°C for 30 min in evacuated sealed tubes. Amino acid analysis was performed in a Hitachi 835-50 amino acid analyzer. The amino acid sequence was determined by the manual Edman method with modification [8]. Phenylthiohydantoin amino acid derivatives were identified by HPLC packed with Cosmosil 5PTH column (4.6250 mm, Nakarai Chemicals Ltd.) with isocratic elution. Autoxidation rate measurements The measurements were carried out in 0.1 M buffer over pH range of 4.5-10.5 at 25°C according to our standard method[10-12]. A 2 ml of solution containing 0.2 M appropriate buffer was placed in a test tube and incubated in a water bath at 25+0.1°C for 15 min. The reaction was started by adding 2 ml of fresh MbO, solution (60 “M) which had been incubated in water bath at 25+0.1°C for 5min, and the reaction mixture was quickly transferred to the quartz cell thermostated at 25+0.1°C. The changes in absorption spectrum from 450 to 650nm were recorded at measured intervals of time in a Hitachi 100-50 or Hitachi 220A spectrophotometer. The buffers used were acetate, 4-morpholinoethanesulfonic acid, phos- phate, Tris-HCl, cyclohexylaminoethanesulfonic acid and cyclohexylaminopropanesulfonic acid. RESULTS Myoglobin preparation and spectral properties We could isolate the native MbO, and metMb directly from red muscle of Galeus nipponensis, according to the same method as used in other shark myoglobins [5, 6]. From 50 g of red muscle, about 20 mg of myoglobin (7 mg in oxy-form and 13 mg in met-form) was obtained. The spectro- scopic properties of the isolated MbO, and metMb are compared in Tables 1 and 2, respectively, with those from other sources. The spectrum of Galeus MbO, was similar to those of mammalian MbO,s, while that of metMb was rather different. Partial amino acid sequence determination Figure 1 shows the HPLC pattern of the tryptic peptides of Galeus myoglobin. From the compari- son with the sequences of related shark myoglo- Myoglobin of the Shark Galeus nipponensis 71 TABLE 1. bins at pH 8.0 Absorption maxima, extinction coefficients and characteristic extinction ratios of oxymyoglo- Absorption maximum (nm) (extinction coefficient (mM ~'cm~')) Source Reference alpha/beta gamma/UV alpha beta gamma UV Sperm whale 10 581 543 418 280 1.08 3.52 (15.4) (14.3) (129) (36.6) Bovine 17 581 544 418 280 1.07 3.68 (15.5) (14.5) (134) (36.4) Heterodontus 12 578 542 418 278 1.02 3.73 (15.2) (14.9) (132) (35.4) Aplysia* 13 578 543 418 278 1.03 3.57 (13.8) (13.4) (120) (33.6) Dolabella* 11 578 543 418 274 1.02 3.67 (13.4) (13.1) (117) (31.9) Galeorhinus* 5 579 543 418 278 1.06 3.08 (15.7) (14.8) (127) (41.3) Galeus* This work 577 542 417 280 1.06 2.92 (15.6) (14.7) (124) (42.6) * These myoglobins are lacking the distal (E7) histidine. TABLE 2. Absorption maxima, extinction coefficients and characteristic extinction ratio of acid metmyoglobin Absorption maximum (nm) (extinction coefficient (mM~‘cm~')) Sources Reference gamma/CT Craleba CT gamma Sperm whale 18 635 505 409.5 16.6 (3.55) (9.47) (157) Horse 19 630 505 408 1527 (4.2) (10.2) (160) Heterodontus 12 630 500 407 17.1 (4.1) (9.49) (162) Aplysia* 19 640 505 400 7.6 (3.8) (13.1) (99) Dolabella* 12 640 506 402 8.2 (3.3) (11.9) (97.5) Galeorhinus* 12 640 505 398 8.6 (3.1) (12.2) (105) Galeus* This work 637 504 393 8.2 (3.2) (12.7) (104) * These proteins are lacking the distal (E7) histidine. bins, two peptides (T1 and 2) containing E-helix region were isolated. The chymotryptic peptides of Galeus myoglobin were first fractionated into 21 fractions, and from the 6th and 8th fractions two peptides (C1 and 2) were purified by rechroma- tography (Fig. 2). Amino acid compositions of the peptides are shown in Table 3. Amino acid sequence of the peptide was deter- mined by manual Edman degradation as shown in Figure 3. The continuity between the peptides T1 and T2 is supported by the peptide C1. The lysine residue at position 73 was determined from the amino acid composition of peptide T2. We could determine the sequence of 28 residues around the 72 TABLE 3. E-helix region T. Suzuki, R. Muramatsu et al. Amino acid compositions of tryptic and chymotryptic peptides containing the A.A. Tl T2 Cl C2 Asp 3.0 (3) 3.2 (3) 1.0 (1) 2.0 (2) Thr Ser 0.9 (1) Glu 1.1 (1) 1.1 (1) 1.1 (1) Gly 1.0 (1) 1.2 (1) Ala 2.0 (2) 3.0 (3) 2.0 (2) 1.1 (1) Cys Val 2.1 (3)* 1.4 (2)* 0.6 (1)* Met Ile 1.0 (1) 0.6 (1)* 0.6 (1)* Leu 2.0 (2) 2.1 (2) 1.0 (1) 1.2 (1) Tyr Phe Lys 2.0 (2) 2.0 (2) 1.0 (1) 1.1 (1) His Arg Pro Trp Total 12 16 8 8 Sequence 46-57 58-73 57-64 65-72 Yield (%) 50.2 47.6 41.4 5.8 The values in parentheses are the number of residues determined by sequencing * Due to Val-Val and/or Ie-Val bonds Fic. 1. A225 T1 it2 20 40 Time (min) 60 HPLC pattern of the tryptic peptides of Galeus myoglobin. The column was packed with Lichrosorb RP8 and equilibrated with 10mM ammonium acetate containing 2% acetonitrile. Myoglobin of the Shark Galeus nipponensis 73 E-helix region (positions 46 to 73). Autoxidation of Galeus MbO; The autoxidation rate of Galeus MbO, was examined in 0.1 M buffer at 25°C over pH range Cl A C2 B 60 = oe a 30 2 < a oO 2 0 20 40 Time (min) Fic. 2. Rechromatography of the 6th (A) and 8th (B) fractions of chymotryptic peptides of Galeus myog- lobin. The column (Cosmosil 5C,s-P) was equili- brated with 0.1% TFA containing 2% acetonitrile. 46 50 55 4.5-10.5. Under air-saturated conditions, the rate is given by —d[MbO,]dt=kop.[MbO2] [1] where k,,, represents the observed first-order rate constant at a given pH. Figure 4 shows a plot of log[k ps] versus pH for the autoxidation of Galeus MbO,, with that of sperm whale MbO, for comparison. The pH dependence of Galeus MbO, was quite different from that of sperm whale MbO,j in the acidic range of pH. DISCUSSION As shown in Table 1, the extinction coefficients and the extinction ratios of all the MbO,s com- pared were very similar as suggested previously [13], the alpha-peak being always higher than the beta peak and the absorbance ratios (alpha/beta ratio) ranging in 1.02-1.08. The presence or absence of the distal (E7) histidine which can form a hydrogen bond to the bound dioxygen [1] appears to give no effect on the spectral properties of MbO;. If the alpha/beta ratio reflects the electronic structure of the bound dioxygen in MbO, [14], the dioxygen in Galeus MbO, may be bound in a same way as in the case of mammalian MbO) . In contrast, the spectrum of acid metMb appeared to be strongly dependent on the presence or absence of the distal (E7) histidine, as shown in Table 2. The extinction ratio of gamma- to CT (charge transfer)-maximum (gamma/CT) provides a most sensitive and useful criterion for separating the two types of metMbs [12]. Table 2 clearly shows that the values of 15.7-17.1 are the ratios for the myoglobin with the distal (E7) histidine, while those of 7.6-8.6 are for the unusual myoglo- | 60 65 70 Ser Ile Ala Gin Leu Lys Asp Asn Ala Asp Leu Lys Ala Gin Ala Asp Val Val Leu Asn Ala Leu Gly Asn Ile Val Lys Lys 1 T2 FERRIER HHI JERE C1 C2 FRRIIIRIIRIRIBHIIIE 0 RIHIRIEIHIEIHFIIHIHIRHIRHHHI HIRE BRE Fic. 3. sequence of peptide was determined by manual Edman degradation (***). (E7) residue. Summary of data to establish the partial amino acid sequence around E-helix of Galeus myoglobin. The The arrow indicates the distal 74 bin without the distal histidine. Therefore, Galeus myoglobin with the value of 8.2 seemed to lack the distal (E7) histidine. This was confirmed by the Fic. 4. The log(k,,,) versus pH profile for the autox- idation of Galeus MbO, in 0.1M buffer at 25°C, with that of sperm whale MbO, [10] for compari- son. Myoglobin concentration, 30 uM. T. Suzuki, R. MuRAMATSU et al. partial amino acid sequence determination around E-helix region. The partial amino acid sequence of Galeus myoglobin is aligned with those of the myoglobins from the sharks Heterodontus portusjacksoni (Heterodontidae) [15], Heterodontus japonicus [12], Mustelus antarcticus (Triakididae) [3], Galeorhinus australis (Triakididae) [4], and Galeorhinus japonicus [5] as shown in Figure 5. Table 4 shows the number of amino acid differ- ences in Figure 5. In a limited sequence of 28 residues in Galeus myoglobin, 7 (25%), 9 (32%) and 15-16 (54-57%) are different from those in the corresponding positions in Galeorhinus, Mus- telus and Heterodontus myoglobins, respectively. In the six globins, 9 residues appear to be invariant. The phylogeny of these globin chains from Table 4 seems to be in good agreement with that from classical taxonomy. In contrast to the “primitive” sharks, Heter- odontus portusjacksoni and H. japonicus, the distal (E7) histidine of Galeus myoglobin is replaced by glutamine as in the cases of other “modern” sharks, Galeorhinus and Mustelus, as shown in Figure 5. On the other hand, the distal (E11) valine at position 63, which also affects on the oxygenation properties of mammalian hemoglobin and myoglobin, is conserved in all the shark 60 65 Fic. Galeus nipponensis Galeorhinus australis Galeorhinus japonicus Mustelus antarcticus Heterodontus portus jacksoni Heterodontus japonicus 46 50 55 Ser Ile Ala Gln Leu Lys Asp Asn Ala Asp Leu Lys Ala|Gin|Ala Asp Val Val Ser Leu Gly Glu Leu Lys Asp Thr Ala Asp Ile Lys Ala/GIn/Ala Asp Thr Val Ser Leu Gly Glu Leu Lys Asp Thr Ala Asp Ile Lys Ala|Gin/Ala Asp Thr Val Ser Leu Gly Glu Leu Gly Asp Thr Ala Ala Ile Lys Ala/GIn/Ala Asp Thr Val Pro Val Gin Gin Leu Gly Asn Asn Glu Asp Leu Arg Lys|His|Gly Val Thr Val Pro Val Glu Gln Leu Gly Asn Asn Glu Asp Leu Arg Lys|His|Gly Val Thr Val I de | Leu Asn Ala Leu Gly Asn Leu Lys Ala Leu Gly Asn Leu Arg Ala Leu Gly Asn Leu Ser Ala Leu Gly Asn Leu Arg Ala Leu Gly Asn Leu Arg Ala Leu Gly Asn * * * * * 70 73 Ile Val Lys Lys Ile Val Lys Lys Ile Val Lys Lys Tle Val Lys Lys Ile Leu Lys Gln Tle Leu Lys Gin * * — 5. Comparison of amino acid sequences (positions 46 to 73) of shark myoglobins. The distal (E7) residues are boxed. Asterisks indicate the invariable residue in the six globins. The helical segments (D and E) of “myoglobin fold” are also shown. TABLE 4. Sequence homologies (number of different residues) in positions 46 to 73 between shark myoglobins Galeus n. Galeorhinus a. Galeorhinus j. Mustelus a. Heterodontus p. Galeorhinus a. if Galeorhinus j. 7 1 Mustelus a. 9 3 3 Heterodontus p. 15 17 16 17 Heterodontus j. 16 16 15 16 1 Myoglobin of the Shark Galeus nipponensis 75 myoglobins. Recently it was shown that Aplysia [13]and Dolabella [11]MbOy,s, which lack the distal (E7) histidine, are extremely unstable and that the pH dependence of the stability is quite different from that of sperm whale MbO,. This unusual stability was attributed to the absence of a distal (E7) histidine [11, 13]. However it should be noted that there are many amino acid differences, other than the distal residue, between molluscan and mammalian myoglobins, with only 20-25% of the residues identical. Therefore it seems to be of great interest to examine the stability of Galeus MbO,, which is more related to mammalian globins and lacks the distal (E7) histidine, for further elucidation of the role of distal (E7) residue. Judging from the partial amino acid sequence shown in Figure 5, the sequence homolo- gy between Galeus and sperm whale myoglobins is about 50%. The pH dependence for autoxidation of Galeus MbO, is shown in Figure 4. Galeus MbO, was extremely unstable between pH 6 and 10.5, and the pH dependence was quite different from that of sperm whale MbO,. For example, Galeus MbO,)j is autoxidized 2.2, 25 and 88 times faster at pHs 4.5, 7.0 and 9.2, respectively. The most striking feature is the fact that unlike sperm whale MbO, the autoxidation rate of Galeus MbO;j is not enhanced with increase in the concentration of hydrogen ion under pH 6. We reported previously the unique stability properties of MbO;s from the sharks Proscyllium and Galeorhinus [6, 12]. Their pH dependences for autoxidation are very similar to that of Galeus myoglobin, and this fact implies that there is some structural analogy among the three shark myoglo- bins. As shown in Figure 5, Galeus and Galeorhi- nus myoglobins are lacking the distal (E7) histi- dine; the sequence of Proscyllium myoglobin is not known yet. On the other hand, since there are only 50% homologies in the sequences between the sharks and sperm whale myoglobins, it is difficult to attribute the difference in the pH dependence of autoxidation to a particular re- sidue. However the distal (E7) residue ought to be of primary importance, because the E7-histidine is the only residue capable of interacting directly with bound dioxygen and stabilizing it [1]. There- fore, our comparative studies on autoxidation of shark MbO,s without the distal (E7) histidine strongly support the idea that the distal (E7) histidine participates in the autoxidation reaction as a catalytic residue facilitating the movement of a catalytic proton [16]. If a myoglobin lacks the distal (E7) histidene, the autoxidation rate will not be accelerated under acidic pHs such as seen in Figure 4. Finally, it is interesing to note that all the vertebrate myoglobins with the exceptional re- placement at position E7 have glutamine at this position in place of histidine. From physiological and evolutional points of view, this must be discussed in near future. ACKNOWLEDGMENTS We thank Drs. O. Okamura and Y. Machida for identification and supply of the shark Galeus nip- ponensis. This paper is dedicated to Professor Shun-Ichi Umezawa in honor of his retirement from Kochi Uni- versity. REFERENCES 1 Phillips,S.E.V. and Schoenborn, B. P. (1981) Neutron diffraction reveals oxygen-histidine hy- drogen bond in oxymyoglobin. Nature, 292: 81-82. 2 Tucker, P.W., Phillips,S.E.V., Perutz, M.F., Houtchens, R. and Caughey, W. S. (1978) Structure of hemoglobins Zurich [His E7 (63) betaArg] and Sidney [Val E11 (67) beta—Ala] and the role of the distal residues in ligand bending. Proc. Natl. Acad. Sci. USA, 75: 1076-1080. 3 Fisher, W. K., Koureas, D. D. and Thompson, E. O. P. (1980) Myoglobins of cartilaginous fishes. IT. Isolation and amino acid sequence of myoglobin of the shark Mustelus antarcticus. Aust. J. Biol. Sci., 33: 153-167. 4 Fisher, W. K., Koureas,D.D. and Thompson, E. O. P. (1981) Myoglobins of cartilaginous fishes. IIT. Amino acid sequence of myoglobin of the shark Galeorhinus australis. Aust. J. Biol. Sci., 34: 5-10. 5 Suzuki, T., Suzuki, T.-N. and Yata, T. (1985) Shark myoglobins. II. Isolation, characterization and ami- no acid sequence of myoglobin from Galeorhinus japonicus. Aust. J. Biol. Sci., 38: 347-354. 6 Suzuki, T. and Kisamori, T. (1984) Shark myoglo- bins-I. Isolation and characterization of myoglobins from the sharks, Squalus japonicus and Proscyllium habereri. Comp. Biochem. Physiol., 78B: 163-166. 7 10 12 13 76 Drabkin, D.L. (1950) The distribution of the chromoproteins, hemoglobin, myoglobin, and chy- tochrome c, in the tissues of different species, and the relationship of the total content of each chro- moprotein to body mass. J. Biol. Chem., 182: 313- 333. Suzuki, T., Furukohri,T. and Gotoh, T. (1985) Subunit structure of extracellular hemoglobin from the polychaete Tylorrhynchus heterochaetus and amino acid sequence of the constituent polypeptide chain (IIC). J. Biol. Chem., 260: 3145-3154. Suzuki, T. and Gotoh, T. (1986) The complete amino acid sequence of giant multisubunit hemoglo- bin from the polychaete Tylorrhynchus heter- ochaetus. J. Biol. Chem., 261: 9257-9267. Suzuki, T. and Shikama, K. (1983) Stability prop- erties of sperm whale oxymyoglobin. Arch. Biochem. Biophys., 224: 695-699. Suzuki, T. (1986) Amino acid sequence of myoglo- bin from the mollusc Dolabella auricularia. J. Biol. Chem., 261: 3692-3699. Suzuki, T. (1987) Autoxidation of oxymyoglobin with the distal (E7) glutamine. Biochem. Biophys. Acta, 914: 170-176. Shikama, K. and Katagiri, T. (1984) Aplysia ox- 14 15 18 19 T. Suzuki, R. Muramatsu et al. ymyoglobin with an unusual stability property. J. Mol. Biol., 174: 697-704. Wang, C. M. and Briniger, W. S. (1979) A correla- tion of the visible and Soret spectra of dioxygen- and carbon monooxide-heme complexes and _five- coordinate heme complexes with the spectra of oxy-, carboxy-, and deoxyhemoglobins. Biochemistry, 18: 4960-4977. Fisher, W.K. and Thompson,E.O.P. (1979) Myoglobin of the shark Heterodontus portusjackso- ni; isolation and amino acid sequence. Aust. J. Biol. Sci., 32: 277-294. Sugawara, Y. and Shikama, K. (1980) Autoxidation of native oxymyoglobin. Thermodynamic analysis of the pH profile. Eur. J. Biochem., 110: 241-246. Gotoh, T. and Shikama, K. (1976) Autoxidation of native oxymyoglobin from bovine heart muscle. Arch. Biochem. Biophys., 163: 476-481. Hanania, G. I. H., Yeghiayan, A. and Cameron, B. F. (1966) Absorption spectra of sperm-whale fer- rimyoglobin. Biochem. J., 98: 189-192. Rossi-Fanelli, A. and Antonini, E. (1957) A new type of myoglobin isolated and crystallized from the muscles of Aplysiae. Biochimia, 22: 335-342. ZOOLOGICAL SCIENCE 5: 77-83 (1988) T Cell-Specific Antigen in Xenopus Identified with a Mouse Monoclonal Antibody: Biochemical Characterization and Species Distribution SABURO NAGATA Tokyo Metropolitan Institute of Gerontology, Sakaecho 35-2, Itabashiku, Tokyo 173, Japan ABSTRACT—A mouse monoclonal antibody (mAb) produced against Xenopus laevis thymocytes, named XT-1, recognizes a thymus-dependent (T) cell-specific antigen, provisionally designated XTLA-1. For biochemical characterization of the XTLA-1 antigen, the lysates of surface- radioiodinated thymocytes and splenocytes were immunoprecipitated with the XT-1 mAb, and analysed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and two-dimensional elec- trophoresis. The results indicate that the XT-1 mAb precipitates peptides with N-linked and other glycans, reduced form of that have an apparent molecular weight of approximately 120 KD and isoelectric point of pH 5.3-6.1. The species distribution study employing immunofluorescence showed that the XTLA-1 antigen was expressed on T cells from X. laevis, X. borealis and LG15 (X. laevisx X. gilli hybrid clone), but not on those from a frog (Rana nigromaculata), toad (Bufo japonicus), newt (Cynopus pyrrhogaster), teleost fish (Oryzias latipes), chicken, mouse and man. The © 1988 Zoological Society of Japan XTLA-1 antigen may provide a useful marker of Xenopus T cells. INTRODUCTION The immunobiology of the anuran amphibia, Xenopus, has recently been extensively studied and demonstrated that the immune system of this frog is remarkably like that of the man and experimental mammals. Thus, Xenopus, a widely used experimental animal in the developmental biology, may provide an alternative experimental model to the mouse in the study of developmental aspects of the immune system. Especially, studies with embryonic transplantation chimeras were informative to understand the process of early stem cell migration into the thymus and emigration of thymus-dependent (T) cells from the thymus [1- 4]. Informations on the differentiation of func- tionally distinct lymphocyte subpopulations are, however, relatively poor mainly because of the lack of reagents that enable the identification of lymphocyte subpopulations. Therefore, we have tried to produce reagents that can identify cell Accepted August 11, 1987 Received July 7, 1987 surface markers on Xenopus T cells and their subsets. In a series of previous papers [5-7], we de- scribed a specific cell surface markers on X. leavis T cells that can be identified by a mouse monoclo- nal antibody (mAb), XT-1. The studies on the tissue distribution [5], ontogeny, effect of early- larval thymectomy [6] and some functional aspects [7] of cells that have antigens recognized by this mAb suggest that the antigen may provide a marker for the cells in T lineage. The present paper describes the basic molecular nature and species distribution of this antigen, designated XTLA-1. MATERIALS AND METHODS Animals Xenopus laevis used in the present study were major _histocompatibility complex (MHC) homozygous, partially inbred J and K_ strain animals established in Dr. Katagiri’s laboratory of Hokkaido University [8, 9]. X. borealis were 78 S. NAGATA kindly provided by Dr. Tochinai of Hokkaido University and isogenic LG15 clone frogs (inter- species hybrids between X. laevis and X. gilli), originally established in Dr. Du Pasquier’s labora- tory at the Basel Institute for Immunology [10], were from Dr. Cohen of the University of Roches- ter. The thymectomy on day 5 postfertilization was performed by the method of Tochinai [8]. Japanese newts Cynopus pyrrhogaster, frogs Rana nigromaculata, toads Bufo japonicus, newly- hatched chickens, Japanese medaka (Teleost fish) and mice (C57BL/6, C3H and DBA strains) were purchased from commercial dealers. Antibody The mouse hybridoma producing XT-1 mAb was produced against J thymocytes as described previously [5]. The immunoglobulin G (IgG) fraction enriched for XT-1 mAb was isolated from the hybridoma ascites by an affinity chromatogra- phy on the column of protein A Sepharose CL-4B beads (Pharmacia). Radioiodination of cells Thymocyte and spleen cell suspensions in amphibian phosphate buffered saline (APBS; 100 mM NaCl, 10 mM sodium phosphate buffer, pH 7.2) were prepared from animals at least 5 months after the completion of metamorphosis. Spleno- cyte preparations were enriched for lymphocytes by depleting erythrocytes with sedimentation at 1xg. The cell suspensions containing 5X10’ viable cells in 500 41 APBS were added with 50 yl of 25 U/ml lactoperoxidase solution (Calbiochem), 50 wl of 5 U/ml glucose oxidase (Sigma), 1 mCi of carrier-free '*°I (Amersham) and 50d of 0.2 M glucose, and incubated at room temperature for 15 min with gentle agitation. The cells were washed with APBS containing 0.1% NaN3, and then incubated for 30 min on ice in 1 ml lysis buffer (0.5% Nonidet P-40, 10mM Tris-HCl, pH 7.2, 150 mM NaCl and 1mM_ phenylmethylsulfonyl fluoride). The lysates were cleared by centrifuga- tion at 8000 xg for 10 min. Immunoprecipitation and gel electrophoresis For immunoprecipitation, 100 l lysates in the Eppendorf tubes were precleared by incubation for 1 hr on ice with 20 ug normal mouse IgG and then for 30min with 25 “1 packed protein A- Sepharose CL-4B beads. The precleared samples were immunoprecipitated with 20 ~g mAb for 6- 16hr on ice, followed by protein A-Sepharose CL-4B beads. For one-dimensional SDS- polyacrylamide gel electrophoresis (SDS-PAGE), the samples were extracted by boiling the beads for 5 min in Laemmli’s SDS sample buffer and elec- trophoresed on 7% polyacrylamide gel with a discontinuous buffer system [11]. Two- dimensional electrophoresis for isoelectric point and molecular size was performed according to the method of O'Farrell et al. [12], and two- dimensional nonreducing and reducing SDS- PAGE according to the method of Allison et al. [13]. The electrophoresed gels were fixed with 10% methanol-10% acetic acid, dried on filter paper, and autoradiographed on Kodak 5R X-ray film with Cronex enhancer screen (Dupont) at —70°C. Glycosidase digestion Digestions of immunoprecipitates with endo-/- N-acetylglucosaminidase F (endo F), which re- moves N-linked glycans, were performed as de- scribed previously [14]. Briefly, immunoprecipi- tates on 20 l packed protein A beads were eluted by boiling for 5 min in 30 sl of 100 mM Tris buffer (pH 8.0) containing 1% SDS and 1% 2- mercaptoethanol (2-ME), and mixed with 0.25 U endo F (glycopeptidase F free; Boehringer Mann- heim Biochemica) in 100mM phosphate buffer (pH 6.1) containing 50 mM EDTA, 1% NP-40, 0.2% SDS and 1% 2-ME. The mixtures were incubated overnight at 37°C and then proteins were precipitated by adding 30 wl of 50% trichlor- oacetic acid (TCA). Immunofluorescence staining and flow cytometry Cells suspended in APBS containing 0.1% bovine serum albumin and 0.1% NaN 3 were incubated for 1 hr on ice with 2.5 ug/ml of XT-1 mAb, washed twice, and then incubated for 1 hr with the FITC-labeled anti-mouse IgG goat anti- body (TAGO, Inc.). Control samples were incu- bated with FITC-labeled reagent alone. The stained cells were observed under a fluorescence Xenopus T Cell Antigen 79 microscope with epiillumination or analysed with a EPICS-C cell sorter (Coulter Electronics) as described previously [5]. RESULTS To characterize the molecular nature of the XTLA-1 antigen, radioiodinated thymocyte and splenocyte lysates from J strain frogs were im- munoprecipitated with XT-1 mAb, and analysed by SDS-PAGE. Under nonreducing conditions, the immunoprecipitates from both thymocytes and splenocytes migrated as two bands of apparent molecular weight of 110 kilodalton (KD) and 120 KD (Fig. 1A). When the immunoprecipitated = die SE TS TS 150> ) me « 95 > 67> Fic. 1. materials were reduced with 2—ME and then electrophoresed, a single band of 120 KD could only be detected (Fig. 1B). Diagonal non-reducing and reducing SDS-PAGE of the same material revealed that the two peptide bands on non- reduced SDS-PAGE have the same mobility on reduced SDS-PAGE, forming two tandem spots corresponding to a molecular size of 120 KD on second dimension (Fig. 1C). To determine whether the immunoprecipitated peptides have N-linked glycans, immunoprecipitates were tre- ated with endo F and then analysed by SDS- PAGE. As seen in Figure 2, the apparent molecu- lar size of the 120KD material was slightly reduced, indicating the presence of N-linked gly- C NONREDUCED <— qaondqsay < Analysis of XTLA-1 antigen by one-dimensional SDS-PAGE under non-reducing (A) and reducing (B) conditions, and by two-dimensional non-reducing and reducing SDS-PAGE (C). Thymocytes (T) and splenocytes (S) from J strain X. laevis were radioiodinated, solubilized, and immunoprecipitated with XT-1 mAb followed by protein A-Sepharose CL-4B beads. For one- dimensional SDS-PAGE, immunoprecipitates were boiled in SDS sample buffer with (reduced) or without (nonreduced) 2—ME and electrophoresed on 7% polyacrylamide gels. For two-dimensional non-reducing and reducing SDS-PAGE, nonreduced immunoprecipitates from thymocyte lysates were run on 7% polyacrylamide tube gels for first dimension, gels were immersed in sample buffer containing 2—-ME and then electrophoresed for second dimension. visualized by autoradiography. Broken line in Fig. C represents diagonal. Electrophoresed gels were Molecular size (KD) standards were run on same gels and represented with arrow heads. 80 S. NAGATA 95 > 67 > Fic. 2. Effect of endo F treatment on XTLA-1 anti- gen immunoprecipitated from radioiodinated J strain thymocytes. Immunoprecipitates were incu- bated overnight at 37°C with (+) or without (—) endo F; proteins were precipitated with TCA and subjected to reduced SDS-PAGE. cans. For the further characterization of XTLA-1 antigen, immunoprecipitates from J thymocytes were analysed by two dimensional electrophoresis for isoelectric point and molecular size. As shown in Figure 3A, 120 KD peptide exhibited a charge heterogeneity near the acidic end of the gel (isoelectric point between pH 5.3-6.1). When the endo F-treated immunoprecipitates were analysed by two-dimensional electrophoresis under the same conditions, several spots with the same molecular size and the extensive charge heter- ogeneity (isoelectric point of pH 4.8-6.5) were identified (Fig. 3B). The distribution of XTLA-1 antigen in several amphibian and non-amphibian species was ex- amined by flow cytofluorometry with thymocytes and splenocytes after indirect immunofluorescence staining. Typical results of such analyses repre- sented in Figure 4 show that more than 95% thymocytes and 20%-35% splenocytes from three Xenopus (X. laevis, X. borealis and LG15) were positive for fluorescence. The fluorescence intensi- ty of positive cells in these histograms is similar, indicating the expression of approximately the same numbers of the determinants recognized by NEPHGE <— pH 8 7 6 5 v v v v A 150 > | om S i 95 > a > 67> @ B 150 > 95 > 67 P. Fic. 3. Two-dimensional nonequilibrium pH gel elec- trophoresis (NEPHGE) and SDS-PAGE of XTLA-1 antigen immunoprecipitated from J strain thymocytes. Endo F-treated (B) and -nontreated (A) immunoprecipitates were separated by NEPH- GE on first dimension, followed by second dimen- sion reduced SDS-PAGE. XT-1 mAb. As reported in the previous study on J strain X. laevis [5], early-larval thymectomy on these three toads resulted in depletion of fluores- cence-positive lymphocyte population in spleen. Similar immunofluorescence analyses revealed that cells from the newt (Cynopus pyrrhogaster), frog (Rana nigromaculata), toad (Bufo japonicus), human, mouse, chicken and fish (Japanese meda- ka) were all negative for the expression of the determinants (data for the last four animals were not shown). To examine whether the determinants identified by the flow cytofluorometry represent the same antigenic molecules, immunoprecipitation experi- Xenopus T Cell Antigen 81 XL/THY LG/THY RELATIVE CELL NUMBER XB/THY XB/SPL TX-XB/SPL a BJ/THY CP/THY RELATIVE FLUORESCENCE INTENSITY Fic. 4. Flow cytometry of thymocytes (THY) and splenocytes (SPL) from normal, and splenocytes from thymectomized (TX) J strain X. laevis (XL), LG15, X. borealis (B), Rana nigromaculata (RN), Bufo japonicus (BJ) and Cynopus pyrrhogaster (CP). Cells were stained by incubation with XT-1 mAb followed by FITC-labeled anti-mouse IgG goat antibody, and analysed with EPICS-C cell sorter. ments were performed on radioiodinated thymo- cytes from J and K strain X. laevis, LG15 and X. borealis. The results showed that the molecules precipitated with XT-1 mAb from all these frogs were peptides with an identical molecular size of 120 KD (Fig. 5). DISCUSSION The present study described basic molecular nature and species distribution of the XTLA-1 antigen recognized by the XT-1 mAb. From the results of SDS-PAGE and two-dimensional elec- trophoresis, the XTLA-1 antigen seems to be glycoprotein consisting of a single polypeptide, reduced form of that has an apparent molecular size of 120 KD and isoelectric point between pH 5.3-6.7. Endo F-treatment of the XTLA-1 anti- gen resulted in minor but distinct reduction of the apparent molecular size, indicating that the XTLA-1 molecule has N-linked glycans. In addition, the fact that the endo F-treated XTLA-1 antigen still showed an extensive charge heter- ogeneity as revealed by two-dimensional elec- trophoresis suggests the presence of additional (O-linked?) glycans. On SDS-PAGE, nonre- duced XTLA-1 antigen ran as two bands of apparent molecular size of 110 KD and 120 KD, both of these peptides have an identical mobility on reduced SDS-PAGE. They may reflect differ- ent forms of the 120 KD membrane peptide that has some 2-ME sensitive structures, such as intra-peptide chain disulfide bonds. Alternatively, the molecular conformation of the antigen might 82 S. NAGATA J KLGXB 67> Fic. 5. Reduced SDS-PAGE of XTLA-1 antigens im- munoprecipitated from thymocytes of J (J) and K (K) strain X. laevis, LG15 (LG) and X. borealis (XB). be partially damaged during the extraction and immunoprecipitation procedures. The XTLA-1 antigen could be detected on lymphocytes from three Xenopus (X. laevis, X. borealis and X. laevis x X. gilli hybrid LG15), but not on those from other amphibian and nonam- phibian species examined. In the previous paper [5], we failed to detect XTLA-1 antigen on spleen cells from K and A strain frogs by immunofluores- cence using hybridoma culture supernatants. But, as shown in the present paper, the high titer ascites or a purified IgG fraction consistently stains splenic T cells from these frogs, suggesting the use of culture supernatants could give misleading results. Immunoprecipitated antigens from the thymocyte lysates of these three toads seems to be an identical 120 KD peptide as analysed by SDS- PAGE. Since XTLA-1 positive spleen cells in these frogs were depleted by early-larval thymectomy, XTLA-1 antigen would be a T cell-specific cell surface antigen shared by these three Xenopus. Recent studies showed that X. borealis and LG series of cloned frogs provide excellent experimental models to study the de- velopmental and genetical aspects of the immune system [4, 15]. The XTLA-1 antigen as a T cell marker in these animals should prove useful for studying the T cell development and function. For some years, substantial efforts have been made with a minimum success to produce reagents that can identify cell surface markers specific to the functional lymphocyte subpopulations in Xenopus (reviewed in [16]). Thus, XTLA-1 antigen de- scribed in the present paper is the only known T cell marker, although surface Ig and MHC class II antigen have been established as markers of B cells and antigen-presenting cells [17, 18]. On the other hand, several T cell-specific antigens have recently been identified in teleost and amphibian species by using monoclonal or polyclonal antibodies [19- 21]. The molecular nature of these antigens has, however, not been characterized. Mansour and Cooper [19] suggested the presence of an equiva- lent of thy-1 antigen, a _ well-characterized mammalian T cell marker, in Rana pipiens. The present XTLA-1 antigen is, however, different from the thy—1-like antigen reported in R. pipiens in the aspects of tissue distribution [5] and molecu- lar nature. That is, as mammalian thy-1 antigen the thy—1-like antigen has molecular weight of less than 30 KD and distributes in the nervous tissues as well as on the T cell surface. Together with the results of species distribution studies, XTLA-1 antigen seems to have no direct relationship to the mammalian thy-1 antigen. ACKNOWLEDGMENT I thank Drs. Hirokawa and Maruyama, Tokyo Metro- politan Institute of Gerontology, for providing the use of their laboratory facilities, and Dr. Ishikawa for the help in gel electrophoresis. This work was supported in part by Grant No. 60440100 from the Japanese Ministry of Education, Science and Culture. REFERENCES 1 Tompkins, R., Volpe, E. P. and Reinschmidt, D. C. (1980) Origin of hemopoietic stem cells in amphib- ian ontogeny. In “Development and Differentiation of Vertebrate Lymphocytes”. Ed. by J. D. Horton, Elsevier/North-Holland, Amsterdam, pp. 25-34. 2 Flajnik, M. F., Horan, P. K. and Cohen, N. (1983) A flow cytometric analysis of the embryonic origin of lymphocytes in diploid/triploid chimeric Xenopus laevis. Dev. Biol., 104: 247-254. 3 Maeno, M., Todate, A. and Katagiri, C. (1985) The localization of precursor cells for larval and adult 10 11 12 13 Xenopus T Cell Antigen 83 hemopoietic cells in Xenopus laevis in two regions of embryos. Dev. Growth Differ., 27: 137-236. Maeno, M., Tochinai,S. and Katagiri, C. (1985) Differential participation of ventral and dorsolateral mesoderms in the hemopoiesis of Xenopus, as revealed in diploid-triploid or interspecific chimeras. Dev. Biol., 116: 503-508. Nagata, S. (1985) A cell surface marker of thymus- dependent lymphocytes in Xenopus laevis is identi- fiable by mouse monoclonal antibody. Eur. J. Immunol., 15: 837-841. Nagata, S. (1986) Development of T lymphocytes in Xenopus laevis: Appearance of the antigen recog- nized by an anti-thymocyte mouse monoclonal antibody. Dev. Biol., 114: 389-394. Nagata, S. (1986) T cell proliferative responses of Xenopus lymphocyte subpopulations separated on anti-thymocyte monoclonal antibody coupled to sepharose beads. Dev. Comp. Immunol., 10: 259- 264. Tochinai,S. and Katagiri, Ch. (1975) Complete abrogation of immune responses to skin allografts and rabbit erythrocytes in the early thymectomized Xenopus laevis. Dev. Growth Differ., 17: 383-394. Nakamura, K. (1985) Lethal graft-versus-host reac- tion induced by parental cells in the clawed frog, Xenopus laevis. Transplantation, 40: 393-397. Kobel, H. R. and Du Pasquier, L. (1975) Produc- tion of large clone of histocompatible fully identical clawed toads (Xenopus). Immunogenetics, 2: 87-91. Laemmli, U. K. (1970) Cleavage of structural pro- teins during the assembly of the head of the bacteriophage T4. Nature, 227: 680-685. O'Farrell, P. Z., Goodman, H. M. and O’Farrell, P. H. (1977) High resolution two dimensional elec- trophoresis of basic as well as acidic proteins. Cell, 12: 1133-1141. Allison, J. P., McIntyre, B.W. and_ Bloch, D. 14 15 16 17 18 19 20 21 (1982) Tumor specific antigen of murine T lympho- ma defined with monoclonal antibody. J. Immunol., 129: 2293-2300. Elder, J. H. and Alexander, S. (1982) Endo-N-/- acethylglucosaminidase F: endoglycosidase from Flabobacterium meningosepticum that cleaves both high-mamnnose and complex glycoproteins. Proc. Natl. Acad. Sci., USA., 79: 4540-4544. Bernard, C.C. A., Bordmann,G., Blomberg, B. and Du Pasquier, L. (1981) Genetic control of T helper cell function in the clawed toad Xenopus laevis. Eur. J. Immunol., 11: 151-155. Flajnik, M. F., Hsu, E., Kaufman,J.F. and Du Pasquier, L. (1987) Biochemistry, tissue distribution and ontogeny of surface molecules detected on Xenopus hemopoietic cells. In “Differentiation Antigen in Lymphohemopoietic Tissues”. Ed. by M. Miyasaka and Z. Trnka, Marcel Dekker, Inc., New York, pp. 387-419. Bleisher, P. A. and Cohen, N. (1981) Monoclonal anti-IgM can separate T cell from B cell proliferative responses in the frog, Xenopus laevis. J. Immunol., 127: 1549-1555. Kaufman, J. F., Flajnik, M. F., Du Pasquier, L. and Riegert, P. (1985) Xenopus MHC class II mole- cules. I. Identification and structural characteriza- tion. J. Immunol., 134: 3248-3257. Mansour, M. H. and Cooper, E. L. (1984) Purifica- tion and characterization of Rana pipiens brain thy- 1 glycoprotein. J. Immunol., 132: 2515-2523. Secombes, C.J., | Van Groningen, J.J.M. and Egbert, E. (1983) Separation of lymphocyte sub- populations in carp Cyprinus carpio L. by mono- clonal antibodies: Immunohistochemical studies. Immunology, 48: 165-175. Amenomori, A. and Sugiyama, K. (1986) Lympho- cyte markers in Rana catesbeiana detectable by mouse monoclonal antibodies. Zool. Sci., 3: 1010. ny be rosy by. a Wk pape 3 ' ¥* “ i L ZOOLOGICAL SCIENCE 5: 85-92 (1988) © 1988 Zoological Society of Japan Temperature-dependence in Reaggregation of Cells Dissociated from Sea Urchin Embryos with Different Seasonal Growth HirRosuKE FuyisAwA and SHONAN AMEmtIyA! Biological Institute, Faculty of Education, Saitama University, Urawa-shi, Saitama 338, and 'Misaki Marine Biological Station, University of Tokyo, Miura-shi, Kanagawa 238-02, Japan ABSTRACT—We investigated the temperature optimal for the aggregation of cells of sea urchin embryos as well as for morphological change of the aggregates into embryo-like organisms. The species of sea urchin used in the present work were Hemicentrotus pulcherrimus, Clypeaster japonicus and Anthocidaris crassispina which have different breeding seasons. In each species the range of temperature optimal for the aggregation of dissociated blastula cells coincided with the range of temperature optimal for their normal development and the ranges corresponded to those of the temperature of the sea water during their breeding seasons. Factors related to the species-specific temperature sensitivity of the embryonic cells were discussed. INTRODUCTION Since Giudice [1] opened the way to experi- ments on the reconstitution of blastula-like organ- isms using single isolated cells of sea urchin embryos, these cells have proved to be one of the most suitable materials for studies on the mecha- nism of cell adhesion during early development. Much work has been done to elucidate the processes occurring during reconstruction by the dissociated embryonic cells both electron- microscopically [2, 3] and biochemically [4-10]. The precise mechanism of cell adhesion during the reconstruction process is, however, far from being clearly understood as yet. The process does not seem to be as simple as could be explained only by the occurrence of certain aggregation substances on the cell surface or in the culture media; rather, it appears to be an intricate combination of cellular activities [2]. In order to gain more evidence to confirm this view, we have examined the physiolo- gical conditions necessary for the process. This paper reports the effects of temperature on the aggregation of dissociated embryonic cells of sea urchins followed by their reconstitution into blas- Accepted July 11, 1987 Received June 9, 1987 tula-like organisms. MATERIALS AND METHODS Sea urchins In this work we used three species of sea urchin, Hemicentrotus pulcherrimus, Clypeaster japonicus and Anthocidaris crassispina. These sea urchins have different breeding seasons: January through March for H. pulcherrimus, June through August for C. japonicus and A. crassispi- na. The sea urchins investigated in this experiment inhabit the tidal areas around the Misaki Marine Biological Station, Kanagawa Prefecture, on the Pacific coast of Japan (139°6' E, 35°2’ N. L.). Temperatures The temperatures of the sea water used for culturing sea urchin embryos and for incubating dissociated cells of sea urchin embryos were adjusted with an accuracy of +0.1 degree as read by a standard thermometer. Embryos Eggs and sperm were obtained by pipetting 0.55 M KCl into the body cavity of sea urchins. The eggs of H. pulcherrimus were fertilized and cultured in normal sea water at 12°C while those of C. japonicus and A. crassispina were treated at 20°C. Preparation of single isolated blastula cells After being washed with calcium-free 86 H. FuyIsAwA AND S. AMEMIYA sea water three times, blastulae were suspended in 0.44M_ sucrose-l1mM EDTA-10mM Tris-HCl buffer, pH 8.0, and dissociated into single cells by gentle pipetting according to the method of Giudice [1]. The temperatures of the calcium-free sea water and the buffer were the same as those used for incubation. Culture of single isolated blastula cells The single cells thus obtained were suspended in 25 ml of normal sea water at a density of 1.5 10° cells per ml. Each of the cell suspensions was incubated at various temperatures ranging from 2° to 35°C. Each suspension was gently stirred with a glass blade at 80 rpm. Measurement of aggregation of single isolated blastula cells At scheduled time intervals, an aliquot of each cell suspension was transferred to a hematocytometer plate and the total number of particles including single cells and cellular aggre- gates of various sizes was counted. The aggrega- tion rate was defined as an index (1—Np/Nt), where Nt is the total number of cells and Np that of particles in a constant volume of cell suspension. Since the total number of particles decreases as the aggregation proceeds to form larger-sized aggre- gates, this value was considered to reflect the aggregation rate. Drugs Cytochalasin B was _ purchased from Aldrich Chemical Co., Ltd. This antibiotic was dissolved in dimethylsulfoxide (DMSO) at a concentration of 5 mg/ml and stored at 4°C. Prior to incubation of the dissociated sea urchin embryonic cells, 25 1 of the solution or DMSO alone were added to 25 ml of normal sea water to give a final cytochalasin B concentration of 5 ng/ml of sea water. Colchicine was purchased from Wako Chemicals Co., Ltd. The drug was used at a concentration of 5 #M in normal sea water. RESULTS Effect of temperature on the aggregation of dissoci- ated blastula cells Sea urchin blastulae were successfully dissoci- ated into single cells by Giudice’s method [1]. The single cells prepared as described above started to 120 ow 2 0:5 Cc jo) 7 j~] o D = 0 0 30 60 90 120 Incubation time ( min ) Fic. 1. Time course changes in the aggregation rate upon culturing dissociated embryonic cells of Hemicentrotus pulcherrimus at temperatures of 2°C (open triangle), 10°C (solid triangle), 15°C (semi- solid triangle), 20°C (solid circle), 25°C (open cir- cle), and 30°C (semi-solid circle). Each point rep- resents the average of triplicate determinations and the vertical bar indicates the standard error of those values. aggregate immediately upon resuspension in nor- mal sea water. Figure 1 shows the time course of the aggregation rate of dissociated Hemicentrotus pulcherrimus blastulae cells at various tempera- tures. Aggregation indices is plotted against incubation time (min). An increase in the rate indicates progress of aggregation of single cells. Curves indicating the changes occurring in the value of the rate were different at different temperatures. The cells cultured at 25° and 30°C continued to aggregate within the first 50 min of incubation, but after this time the aggregated cells showed gradual redissociation. Differences in the rate at different temperatures were apparent after 60 min of incubation. The aggregation rates were high at 2°, 10° and 20°C. This result, shown in Figure 1, means that the optimal range of temperature for the aggregation Temperature-dependent Reaggregation of Embryonic Cells 87 of H. pulcherrimus blastula cells lies above 2°C and below 25°C. At 2°C, however, the cell-to-cell contact in aggregates of the dissociated blastula cells of H. pulcherrimus remained loose and no sign of further morphological change into blastula like organisms was observed even after 4 days of culture. The aggregates formed at temperatures ranging from 10° to 20°C changed morphologically within one day on the initial form showing loose inter- cellular contact and random cellular orientation into blastula-like aggregates with tight intercellular adhesion and regular cellular orientation. These aggregates ordinarily showed regeneration of cilia and the formation of a blastocoel, as reported by Giudice [1] and Amemiya [3]. The majority of aggregates were observed to develop into pluteus- like structures, as reported by Giudice [1]. The 1.0 q o—o—o—4 — . : 0 La { béaa O ow » 2 4 0.5 Cc (>) ca j=] iu) = 12) oO | | 0 0 30 60 90 120 Incubation time ( min ) Fic. 2. Time course changes in the aggregation rate upon culturing dissociated embryonic cells of Clypeaster japonicus at temperatures of 10°C (solid triangle), 20°C (solid circle), 30°C (semi-solid cir- cle), 35°C (open square) and 40°C (solid square). Each point is the average of triplicate determina- tions, the vertical bar being the standard error of those values. aggregates formed at 2°C did not undergo such morphological changes as those described above. Figures 2 and 3 show similar results observed for the dissociated blastula cells of C. japonicus and A. crassispina, respectively. The aggregation rate after 60 min of incubation were high at tempera- tures ranging from 10° to 30°C for embryos of these sea urchins, indicating that the optimal tempera- ture for aggregation of their dissociated cells lies within this range. Compared with the optimal range for cells of H. pulcherrimus, the optimal range for blastula cells of C. japonicus and A. crassispina was apparently higher than that for blastula cells of H. pulcherrimus. As for C. japonicus and A. crassispina, cell aggregates redis- sociated gradually after one day of culture at the lower temperature as well as at 10°C. Table 1 shows the ranges of temperature opti- @® pus) = Cc jo} oa a ta 12) Z 0 0 30 60 90 120 Incubation time ( min ) Fic. 3. Time course changes in the aggregation rate upon culturing dissociated embryonic cells of Anthocidaris crassispina at temperatures of 10°C (solid triangle), 15°C (semi-solid triangle), 20°C (solid circle), 25°C (open circle), 30°C (semi-solid circle), and 35°C (open square). Each point shows the average of triplicate determinations and the vertical bar depicts the standard error of those values. 88 H. FusyIsAwA AND S. AMEMIYA TABLE 1. Ranges of optimal temperature for development of the three species of sea urchin and the ranges of sea water temperature in their spawning seasons Species Range of optimal temperature for development Range of sea water temperature in the spawning seasons (C) Hemicentrotus pulcherrimus Clypeaster japonicus Anthocidaris crassispina mal for normal development of these three species of sea urchin. The upper and the lower limits of temperature optimal for the development as shown in Table 1 indicate that embryos cultured at a temperature above the upper limit and below the lower limit were unable to develop into normal plutei without the occurrence of a significant proportion of deformed individuals. We con- firmed that dissociated cells cultured at a tempera- ture above the upper limits were able to aggregate but that they were unable to maintain the state of 1.0 rate 0.5 Aggregation 0 30 60 90 120 150 Incubation time ( min ) Fic. 4. Capability of reaggregation of the embryonic cells of Clypeaster japonicus after lowering the temperature from 35°C to 20°C. The cells were incubated at 20°C after being heated at 35°C for 10 min (A), 30 min (a), and 60 min (2). Solid circle (@) shows the aggregation rate of the cells incu- bated continuously at 35°C. 4-23 10-17 17-29 19-25 16-29 19-27 aggregation. Cells redissociated at a temperature above the upper limit of the optimal range were able to reversibly reaggregate when the temperature was reduced to the optimal one as shown in Figure 4. Dissociated cells from the blastula of C. japonicus cultured at 35°C were able to reaggregate again when the temperature reverted back to the optimal one (20°C) within half an hour of incubation. Upon culturing the cells at this higher temperature for periods longer than one hour, however, the cells showed a tendency to lose the capacity to reaggregate, even when the temperature was changed back to the optimal one. In this experiment we were able to discriminate morphologically two phases in the process of aggregation. The first phase was one of mere aggregation with random loose intercellular con- tact, while the second phase induced compact aggregation with tight cell adhesion capable of forming an embryo-like structure. The first phase could be observed even at temperatures below or above the optimum, while the second phase only occurred at a temperature within the optimal range specific for each individual species. Effects of cytochalasin B and colchicine on the aggregation of dissociated blastula cells At 12°C, the optimal temperature for H. pul- cherrimus embryos, single cells dissociated from the sea urchin blastulae aggregated in the presence of cytochalasin B (Fig. 5). The aggregates effected by cytochalasin B, however, were unable to adhere tightly to form a compact spherical ones, and hence could not develop into blastula-like ones. These aggregates gradually redissociated after being incubated for longer than four hours and finally reverted back to single cells. The aggre- Temperature-dependent Reaggregation of Embryonic Cells 89 1.0 ~of-—o rate 0.5 Aggregation 0 1 2 3 4 5 6 7 8 19 Incubation time ( hour ) Fic. 5. Effect of cytochalasin B on the aggregation of embryonic cells dissociated from blastula of Hemi- cnetrotus pulcherrimus. Concentration of cytochar- asin B was 5 yg/ml of sea water. Dimethylsulphox- ide (DMSO) was also added at the concentration of 0.1% in the sea water. This concentration of DMSO was non-effective to the cell aggregation. ©: control. @: in the presence of cytocharasin B. Each point and vertical bar represents the average of triplicate determinations and the standard error of those values, respectively. gates themselves, however, were unable to show further morphological change into compact spher- ical aggregates and hence did not develop into a blastula-like form. In contrast, these aggregates redissociated gradually into single cells after in- cubation in the presence of the drug for more than four hours. Thus it was clarified that cytochalasin B affects only the second phase of aggregation, suggesting that microfilaments are only involved in the morphogenetic change occurring in the second phase of aggregation. The sea water added with cytochalasin B also contained DMSO at a concentration of 0.1%. However, this concentration of DMSO was shown not to produce any effect on these aggregation processes of dissociated cells. The first phase of aggregation was also un- affected by colchicine. The aggregates, however, were unable to undergo morphogenetic change into a compact form followed by development into a blastula-like form after being cultured for longer than four hours in the presence of colchicine. The aggregates remained in the initial state of loose intercellular contact throughout the culture period. This result suggests that the microtubule system is involved in the second phase of aggrega- tion. DISCUSSION In the present work we confirmed that the optimal range of temperatures for aggregation of dissociated cells of sea urchin embryos is species specifically determined. Only the cells cultured at an optimal temperature can form aggregates with tight cell adhesion, change into blastula-like organ- isms which regenerate cilia and blastocoel and develop further into pluteus-like ones. At temper- atures above the upper limit or below the lower limit of the range of temperature optimal for normal development, dissociated cells of sea urchin could aggregate, but the aggregates re- mained as loose-contacted ones without morpholo- gical changes into compacted spherical blastula- like organisms. Redissociation was usually observed after temporal aggregation when the cells were cultured at high temperature. Redissociation after temporal aggregation, though more gradual than that at high temperature, was also observed at 2°C in blastula of C. japonicus and A. crassispina whose breeding seasons are summer. We do not know why redissociation occurred after a brief time of incubation at higher temperatures. It is clear that the redissociation cannot be ascribed to low activity of the cytoskeleton since cytochalasin B could not cause such redissociation at brief time of incubation. A similar result has been reported in chick embryonic cells. Moscona [11] showed that the highest aggregation rate of retinal and of hepatic cells of chick embryos was achieved at 38°C, the temperature optimal for these embryos. He also noticed that at 15°C the formation of aggregates was prevented. Steinberg [12] also reported the prevention of aggregation by dissoci- ated epidermal cells of chick embryo at 6.5°C. They observed the inability of formation of aggre- gates into compacted spherical ones with tight cellular adhesion at low temperatures. Temporal 90 H. FusISAWA AND S. AMEMIYA formation of aggregates which remain in loose contact or redissociate after long incubation is thought to be possible at a low temperature since Curtis [13] reported that the cells dissociated from 5-day chick embryos were able to aggregate at 3°C although the aggregation rate was lowered. McClay and Baker [14] have already showed in dissociated neural retinal cells of chick embryo that the aggregation process is subdivided into at least two different phenomena: the reentering the population of cells and the adhesion itself. According to them both phenomena were sensitive to low temperature (4°C). We were also able to discriminate light microscopically two serial phases in the reaggregation in sea urchin embryonic cells. The first phase of aggregation is the formation of aggregates like clusters of single dissociated cells. In this phase cells adhere loosely and randomly to each other. Dissociated embryonic cells can form cluster-like loose aggregates in this first phase independently on thermal condition. The second phase is the formation of spherical compact aggre- gates with tight cell adhesion followed by the morphological change into blastula- and further pluteus-like organisms. The aggregates cannot develop into the second phase at a temperature outside the range optimal for normal embryogenesis. We confirmed that thermal condi- tion for the reaggregation into reconstruction of embryo-like organisms is almost the same as that for normal development. Difference of the breed- ing seasons of sea urchins among different species in the same locality is thus thought to be explained partly by the species specifically determined temperature sensitivities of their embryogenesis. In addition, we found that cells redissociated at a temperature over the upper limit of optimal temperature could reversibly reaggregate when the temperature was lowered to the optimal one. This reversibility, however, was lost after incubation longer than one hour at higher temperature. The reason why the cells redissociate finally at higher temperature is not certain. These results also imply that the process of cellular aggregation does not depend merely on interactions among cells through the so-called cementing substance and divalent cations such as calcium and/or magnesium [5, 6, 8] or aggregation promoting proteins [7, 9, 10]. The so-called cementing substance and/or divalent cations seems to be involved rather in the first phase of aggrega- tion of dissociated embryonic cells than the second phase, since the first phase of aggregation is independent of temperature while attainment of the second phase can be observed only in the range optimal for normal embryogenesis. The synthesis of some cell surface glycoproteins [4, 15, 16] has been reported to be essential for reassociation of dissociated embryonic cells. We think that the synthesis of such substances is necessary for the second phase of aggregation because the aggrega- tion was not affected but morphogenetic change of aggregates was arrested in the presence of some inhibitors of protein synthesis such as puromycin, ethionine [4] and cycloheximide (unpublished data). Stage-specific adhesion or aggregation has been reported in sea urchin embryos [9, 10] as well as in amphibian embryos [17]. The present results are concerned only with the embryos at the mesen- chymal blastula stage. We are now examining the aggregation of these sea urchin embryonic cells in other stages and the effect of temperature on the aggregation. The activity of some cytoskeletal systems of microfilaments and/or microtubules is responsible for the second phase of aggregates [3, 18]. Present work showed that cytochalasin B, as well as colchicine, completely inhibited the compaction of the aggregates and hence the following morpholo- gical change into blastula-like organisms, which was in accordance with the result of Weiss using Ehrlich ascites cells [19, 20]. Cytoskeletal activities are known to be also involved in lectin-mediated cellular aggregation [21]. Microvilli have been reported to be important for aggregation of embryonic cells of sea urchin [22, 23] and other cells [19, 20 24]. Microvilli are related to cyto- skeletal activities [25], therefore the morphogene- tic changes in the second phase, especially tight compaction of aggregates, may be aided with these cell surface structures. Electron microscopic inves- tigations are now in progress to test the involve- ment of these structures in the aggregation and compaction of dissocicated sea urchin embryonic cells. Fluidity of plasma membranes of aggrega- Temperature-dependent Reaggregation of Embryonic Cells 91 ting embryonic cells may be another important factor for the second phase of aggregation. The temperature dependence of aggregation, especial- ly of the second phase, seems to be well explained by this factor. Several pieces of direct evidence for the involvement of the fluidity in temperature- dependent cell agglutination have been obtained by modifying plasma membrane lipid composition or by alteration of proportion between saturated and unsaturated fatty acids [26-31]. Little evi- dence has been reported on species specific temperature dependent microfilament activity as yet. Therefore we think that possible candidate for explaining species specific temperature depend- ence of embryonic cell aggregation of these sea urchins with different breeding seasons is species- specific difference of plasma membrane fluidity of their embryonic cells. We are now attempting to confirm the possibility of relationship between the membrane property and the temperature depend- ence of aggregation. ACKNOWLEDGMENTS We wish to express our sincere thanks to Prof. H. Terayama and Prof. H. Katayama for their advice. We also thanks Dr. Y. Ichihara, Dr. S. Hattori, Dr. H. Kawasaki, Messrs. H. Kubo, M. Matsuda, S. Sone, T. Shinkai, M. Yoshikuni, Y. Ishida, Y.Kihira, T. Ishi- mori, Mrs. M. Toride, Misses T. Ohara, K. Ikeda, M. Ito, and S. Arai for their technical assistances. REFERENCES 1 Giudice, G. (1962) Restitution of whole larvae from disaggregated cells of sea urchin embryos. Dev. Biol., 5: 402-411. 2 Giudice, G. (1967) Electron microscopic study of the reaggregation of cells dissociated from sea urchin embryos. Dev. Biol., 5: 91-101. 3 Amemiya, S. (1971) Further studies on the rela- tionship between cilium formation and cell associa- tion in sea urchin embryos. J. Fac. Sci. Univ. Tokyo, IV., 12: 241-258. 4 Giudice, G. (1965) The mechanism of aggregation of embryonic sea urchin cells; a biochemical approach. Dev. Biol., 12: 233-247. 5 Kondo, K. and Sakai, H. (1971) Demonstration and preliminary characterization of reaggregation- promoting substances from embryonic sea urchin cells. Dev. Growth Differ., 13: 1-14. 6 20 21 Kondo, K. (1973) Cell-binding substances in sea urchin embryos. Dev. Growth Differ., 15: 201-216. Noll, H., Matranga, V., Cascino, D. and Vittorelli, G. (1979) Reconstitution of membranes and embryonic development in dissociated blastula cells of the sea urchin by reinsertion of aggregation- promoting membrane proteins extracted with buta- nol. Proc. Natl. Acad. Sci. U.S. A., 76: 288-292. Tonegawa, Y. (1973) Isolation and characterization of a particulate cell-aggregation factor from sea urchin embryos. Dev. Growth Differ., 14: 337-351. Oppenheimer, S. B. and Meyer, J. T. (1982) Isola- tion of species-specific and stage-specific adhesion promoting component by disaggregation of intact sea urchin embryo cells. Exp. Cell Res., 137: 472- 475. Oppenheimer, S. B. and Meyer, J. T. (1982) Car- bohydrate specificity of sea urchin blastula adhesion component. Exp. Cell Res., 139: 451-455. Moscona, A. (1961) Rotation-mediated histogenetic aggregation of dissociated cells. Exp. Cell Res., 22: 455-475. Steinberg, M. S. (1962) The role of temperature in the control of aggregation of dissociated embryonic cells. Exp. Cell Res., 28: 1-10. Curtis, A. S. G. (1963) Effect of pH and tempera- ture on cell reaggregation. Nature, 200: 1235-1236. McClay, D. R. and Baker, S. R. (1975) A kinetic study of embryonic cell adhesion. Dev. Biol., 43: 109-122. Grunz, H. (1969) Hemmung der Reaggregation dissoziierter Amphibienzellen durch Inhibitoren der RNS- und Protein Synthese. Wilhelm Roux’ Archiv., 163: 184-196. Moscona, A. (1961) Effect of temperature on adhesion to glass and histogenetic cohesion of dissociated cells. Nature, 190: 408-409. Suzuki, A. S., Ueno, T. and Matsusaka, T. (1986) Alteration of cell adhesion system in amphibian ectoderm cells during primary embryonic induction: changes in reaggregation pattern of induced neurec- toderm cells and ultrastructural features of the reaggregate. Roux’s Arch. Dev. Biol., 195: 85-91. Weiss, L. (1972) Studies on cellular adhesion in tissue culture. XII. Some effects of cytochalasins and colchicine. Exp. Cell Res., 74: 21-26. Weiss, L. (1967) Studies on cell deformability. III. Some effects of EDTA on sarcoma 37 cell. J. Cell Biol., 33: 341-347. Weiss, L. and Subjek,J.R. (1974) Interactions between the peripheries of Ehrlich ascites tumor cells as indicated by the binding of colloidal iron hydroxide particles. Int. J. Cancer, 13: 143-150. Smith, S. B. and Revel, J. P. (1972) Mapping of concanavalin A binding sites on the surface of several cell types. Dev. Biol., 27: 434-441. 22 23 24 23 26 27, 92 Spiegel, M. and Spiegel, E. (1979) The morphology and specificity of cell adhesion of echinoderm embryonic cells. Exp. Cell Res., 117: 261-268. Spiegel, M. and Spiegel, E. (1977) A _ scanning electron microscopical study of early sea urchin reaggregation. Exp. Cell Res., 108: 413-420. O’Neill, C. H. (1968) An association between viral transformation and forssman antigen detected by immune adherence in cultured BHK 21 cells. J. Cell Sci., 3: 405-422. Loor, F. (1976) Cell surface design. Nature, 264: 272-276. Wood, R. and Falch, J. (1974) Lipids of cultured hepatoma cells: II. Effect of media lipids on cellular phospholipids. Lipids, 8: 702-710. Jacobs, R. A. and Majerus, P. W. (1973) The reg- ulation of fatty acid synthesis in human skin fibro- blasts. J. Biol. Chem., 248: 8392-8401. H. FUJISAWA AND S. 28 29 30 31 AMEMIYA Wisnieski, B. J., Williams, R.E. and Fox, C.F. (1973) Manipulation of fatty acid composition in animal cells grown in culture. Proc. Natl. Acad. Sci. U.S. A., 72: 3669-3673. Williams, R. E., Wisnieski, B. J., Rittenhouse, H. G. and Fox, C. F. (1974) Utilization of fatty acid supplements by cultured animal cells. Biochemistry, 13: 1969-1977. Horwitz, A. F., Hatten,M.E. and Burger, M. M. (1974) Membrane fatty acid replacements and their effect on growth and lectin-induced agglutinability. Proc. Natl. Acad. Sci. U. S. A., 71: 3115-3119. Rittenhouse, H. G., Williams, R. E., Wisnieski, B. and Fox, C. F. (1974) Alterations of characteristic temperatures for lectin interactions in LM cells with altered lipid composition. Biochem. Biophys. Res. Commun., 58: 222-228. ZOOLOGICAL SCIENCE 5: 93-102 (1988) Probable Participation of Mitochondrial Ca** Transport in Calcification of Spicules and Morphogenesis in Sea Urchin Embryos Keiko MitsunaGA, YUKIO Fusino! and Ikuo YASUMASU7 Department of Biology, School of Education, Waseda University, 1-6-1 Nishiwaseda, Shinjuku-ku, Tokyo 160, and ‘Department of Pharmacology Teikyo University School of Medicine, 2-11-1 Kaga, Itabashi-ku, Tokyo 173, Japan ABSTRACT— In embryos of the sea urchin, Hemicentrotus pulcherrimus, kept with ruthenium red or 2, 4-dinitrophenol in a period between the mesenchyme blastula and the pluteus corresponding stage, formation of calcified spicules was blocked at their concentrations to inhibit *°Ca** uptake in isolated mitochondria. The ATP level slightly decreased in embryos kept with 2, 4-dinitrophenol but was hardly changed in those kept with ruthenium red. These compounds also inhibited spicule rod formation in cultured micromere-derived cells. This indicates that these compounds exert an effect directly on primary mesenchyme cells to block CaCO; deposition in the cells. Ca** uptake in mitochondria, known to be blocked by these compounds, probably participates in intracellular Ca?+t transport for spicule calcification. In embryos kept with these compounds, pluteus arm formation was also blocked, though quasi-normal archenterons were produced in embryos. Excess Ca?*+ concentra- tions in the cells due to inhibition of Ca** uptake in mitochondria may result in a blockage of pluteus © 1988 Zoological Society of Japan arm formation. INTRODUCTION Recently, we reported that Ca’* influx through Ca** channels across plasma membrane is made electrosilent in all cell lineages of sea urchin embryos by coupled transport of anions [1-3]. In spicule forming cells, such as mesenchyme cells in embryos, those isolated by the procedure of Harkey and Whiteley [4] and in cultured micro- mere-derived cells obtained according to the method of Okazaki [5], Ca** thus carried into cells is transported into skeletal vacuoles, in which spicules are produced [6], and utilized as a main mineral material for their production [7]. Main mineral component of spicules in sea urchin embryos is known to be CaCO; [7] and a huge amount of CaCO; is produced in these spicule forming cells. In these cells, the turnover rate of Accepted August 13, 1987 Received July 27, 1987 ? To whom requests of reprints should be addressed. Ca** in cytoplasm is certainly higher than in the other cell lineages. In these spicule forming cells, with a high rate of Ca’* turnover, as well as in the other cell lineages, cytosolic Ca?* level should be kept at an adequate concentration to maintain physiological activities of cell functions. It is well known that Ca** is one of important regulators of many cell functions. Probably, in the cells of sea urchin embryos, cytosolic free Ca’* level is not only regulated by its uptake and release in endoplasmic reticulum but also by Ca?* transport in mitochondria in a similar manner to the other cell types [8-13]. Indeed, in sea urchin eggs, inhibition of Ca*t uptake in mitochondria by an uncoupler of oxida- tive phosphorylation results in an elevation of cytosolic Ca** level [14], in the same manner as in the other type of cells [15, 16]. When Ca?* uptake in mitochondria is inhibited, cytosolic free Cat level may increase especially in spicule forming cells with a high rate of Ca?+ turnover in cyto- plasm. High level of cytosolic Ca7*, which may be 94 K. MitsunaGA, Y. FuJINO AND I. YASUMASU favorable for Ca** transport into skeletal vacuoles from cytoplasm in spicule forming cell, is expected to cause excess stimulation of Ca?* -dependent cell functions to result in an abnormal development of embryos. In the present study, we studied effects of ruthenium red, an inhibitor of ATP-dependent, H*-gradient mediated Ca** uptake in mitochon- dria [8, 17, 18], reported to elevate intracellular free Ca’* level [19, 20], and 2, 4-dinitrophenol (DNP), an uncoupler of oxidative phosphorylation causing discharge of H*-gradient to block uptake of Ca** in mitochondria, on morphogenesis espe- cially on the spicule formation in sea urchin embryos. MATERIALS AND METHODS Handling of gametes Matured animals were collected at Tsushima Island and Sagami Bay and kept in a temperature- controlled sea water tank until use. Gametes of the sea urchin, Hemicentrotus pulcherrimus, were obtained by injection of 0.5 M KCI into the body cavity. Eggs were washed with artificial sea water (ASW) three times and inseminated by adding an adequate amount of sperm. Then, the eggs were washed twice with ASW and cultured at 20°C with gentle stirring by a motor-driven propeller. Isolation and culture of micromeres Micromeres were isolated at the 16-cell stage as described by Kitajima and Matsuda [21] and cultured at 20°C in ASW containing 4% horse serum, according to the procedure described by Kitajima and Okazaki [22]. Treatment of embryos and micromere-derived cells with ruthenium red and DNP Stock solutions of ruthenium red (100 mM) and DNP (5mM) dissolved in distilled water were added to embryo culture 15 hr after fertilization at 20°C and to micromere-derived cell culture at 15 hr of culture. The embryos and the micromere- derived cells were kept for another 30 hr in the presence of these compounds and then the photo- graphs were taken. Less than 10 sl stock solutions of these chemicals were added to 2 ml culture. Addition of 10 J distilled water did not exert any effect on spicule formation and other morphogene- sis in the culture. In some cases, stock solutions of these compounds dissolved in ASW were used in place of those dissolved in distilled water. Calcium amount in spicules isolated from embryos Spicules were isolated from embryos according to the procedure described previously [2]. Calcium content in the spicule fraction was analyzed using an atomic-absorption spectrophotometer (Hitachi 170-S0A; Hitachi Ltd., Tokyo, Japan) as de- scribed previously [2]. Relative length of pluteus arm and archenteron The lengths of embryos along pluteus arms and along archenteron were measured on photographs of the embryos treated with ruthenium red or DNP. Length of pluteus arm was calculated by subtraction of the length along archenteron from that along the arm and relative value was express- ed as percentage of the length in normal embryos. The length of archenteron was also measured on photographs of the embryos. The relative length of archenteron in these treated embryos was also expressed as percentage of archenteron length in normal embryos. These lengths were measured on 30 embryos. The length of spicule in cultured micromere-derived cells was also measured on photographs obtained with a polarized-light micro- scope. When a spicule had branches, the sum of the length of branches was regarded as the length of this spicule. Spicule lengths in 30 clusters of micromere-derived cells in a culture were meas- ured and these values were expressed as san per cell. Ca’* uptake in mitochondria Mitochondrial fraction was obtained from embryos at 35hr after fertilization at 20°C (the prism stage). Embryos were collected by hand- driven centrifuge and were washed twice with ice cold 0.6M_ sucrose solution containing 5mM EDTA, 10mM MgCl, and 10 mM Tris-HCl, pH 7.2 (homogenizing medium). Embryo suspension in the homogenizing medium was transferred to glass homogenizer and homogenized in an ice bath Ruthenium Red and Embryogenesis a5 with a motor-driven Teflon pestle. The homoge- nate was centrifuged at 1,000xg for 10 min and resultant supernatant was again centrifuged at 8,000xg for 20min. The obtained pellet was washed once and suspended in the homogenizing medium without EDTA. This suspension was used as a crude mitochondria fraction. Reaction mix- ture in 200 l contained 5 uM CaCh, 5 #Ci/ml 4CaCh, 0.3M KCl, 5mM MgCh, 10 mM phos- phate buffer, pH 7.2, 10 mM Tris-HCl, pH 7.2, 1 mM ATP, 1mM ADP, 0.5 mM succinate, 5 ~«M cytochrome c, and 10 41 mitochondria suspension (about 5 mg protein ep./ml) in the homogenizing medium from which EDTA was omitted. The reaction was started at 20°C by adding mitochon- dria suspension. After 5 min incubation, it was terminated by adding 1 ml of ice cold 0.3 M KCl solution containing 5mM MgCl, 50 uM CaCl, and 10 mM Tris-HCl, pH 7.2. Then, this suspen- sion was filtered through Millipore _ filter (HAWP02500, Millipore Corp., Ma, USA) and the filter was washed three times in the cold with the medium used for termination of the reaction. The filter was dried in vacuo and analyzed for radioactivity by a liquid scintillation spectrometer (Aloka LSC 700, Aloka, Tokyo, Japan). *°Ca radioactivity, obtained following simultaneous additions to the reaction mixture of mitochondrial suspension and the solution with which the reac- tion was terminated, was subtracted from the values thus obtained. The rate of Ca?* uptake is expressed as cpm of “Ca radioactivity per mg protein of mitochondrial fraction per 5 min. Protein was determined by the method of Lowry et al. [23], using bovine serum albumin as the standard. ATP level in embryos The embryos, washed once with ice cold arti- ficial sea water and suspended in ice cold 0.6M sucrose solution, were centrifuged at 10,000 x g for 10min. The embryo pellet obtained was diluted with 5 volumes of ice cold 5% perchloric acid and homogenized by Polytron (PT10-35, Kinematica GmbH, Switzerland). To the homogenate of embryos, one tenth volumes of 1M triethanol- amine was added and the homogenate was centri- fuged for 30min at 15,000xg. The resultant precipitate was analyzed for protein and the supernatant was neutralized with saturated K,CO; in an ice-bath. After heavy white precipitate was centrifuged off, the supernatant was analyzed for ATP by the enzymatical methods [24], described in detail in a previous paper [25]. Chemicals Ruthenium red and cytochrome c were pur- chased from Sigma Chemical Co., Mo., USA. 2, 4-Dinitrophenol (DNP) was obtained from Kanto Chem. Co., Tokyo, Japan. Nicotinamideadenine dinucleotide phosphate, hexokinase, glucose-6- phosphate dehydrogenase (used for ATP deter- mination), Na,-ATP and ADP were the products of Boehringer Mannheim, Federal Republic of Germany. “CaCl, was purchased from the Radiochemical Centre, Amersham, Bucks, U. K. Artificial sea waters (ASW) were the products of Jamarin Laboratory, Osaka, Japan. RESULTS As shown in Figure 1, sea urchin embryos kept with ruthenium red at above 20 4M from the mesenchyme blastula stage (15 hr after fertiliza- tion at 20°C) became abnormal spherical ones (Fig. 1B-E) having no pluteus arms and quasi- normal archenteron, when control embryos (Fig. 1F) became well developed plutei (45 hr after fertilization). Embryos kept with ruthenium red were considerably dark, probably because of ruthenium red-stained cells. This compound is known to bind with polysaccharides on cell surface [26, 27]. In embryos kept with ruthenium red at 20 uM, at which formation of pluteus arms was completely blocked, irregular tri-radiate spicules were observed in embryos with a polarized light microscope (Fig. 1b). The spicules became smaller in relation to the concentration of ruthenium red (Fig. la-d) and were hardly observed in embryos kept with this compound at above 150 uM (Fig. le). In embryos kept with DNP at 100 and 200 «IM, pluteus arms (Fig. 2A, B) and spicules (Fig. 2a, b) were markedly shorter than in control embryos (Fig. 1F, f). DNP at 200 uM inhibited growth of spicule and pluteus arm more strongly than at 100 96 K. MitsunaGA, Y. FusINO AND I. YASUMASU Fic. 1. Embryos treated with ruthenium red. Ruthenium red was added to the culture medium 15 hr after fertilization at 20°C (the mesenchyme blastula stage). Embryos, kept with ruthenium red at 10 ~M (A), 20 HM (B), 40 uM (C), 100 uM (D) and 150 uM (E), were photographed 45 hr after fertilization (the pluteus corresponding stage). Normal pluteus is shown in F. Photographs marked with (A-F) and (a-f) are those observed with a light and a polarized-light microscope, respectively. Bar shows 100 pm. Fic.2._ Embryos treated with 2, 4-dinitrophenol (DNP). Treatment of embryos with DNP was performed in the same manner as described for the ruthenium red-treatment in the legends of Fig. 1. The embryos were treated with 100 uM (A) and 200 uM DNP (B). Spicules observed with a polar- ized-light microscope in the same embryos shown in A and B are shown in a and b, respectively. Bar shows 100 san. uM (Fig. 2). However, these were not completely blocked by DNP even at 1 mM (photograph not shown). Figure 3 shows the effects of ruthenium red (Fig. 3A) and DNP (Fig. 3B) on calcium deposition in spicules, growth of pluteus arms and of archenter- on. Embryos at the mesenchyme blastula stage (15 hr after fertilization), in which any spicules, pluteus arms and archenterons had not been formed yet, were cultured for another 30 hr at 20°C in the presence of ruthenium red or DNP. Then, the calcium amount in spicules, the length of pluteus arms and the size of archenteron were measured. Hence, these values indicate growth of these structures in embryos during developmental period between 15 and 45 hr after fertilization. As shown in Figure 3A, ruthenium red caused a slight decrease of calcium amount in the spicules at 10 uM and a marked decrease at concentrations above 50 uM. The calcium amount in spicules became close to zero at above 75 uM. The growth of pluteus arms was inhibited by ruthenium red at above 10 “M and was completely blocked at above 20 uM. Strong inhibition of pluteus arm formation by ruthenium red occurred at lower concentrations than those to cause complete inhibition of spicule Ruthenium Red and Embryogenesis 97 A — oO °o a a 200 — oO °o 50 Percent growth of spicule, pluteus arm and archenteron 0 500 Concentration (yM) 1000 Fic. 3. Effects of ruthenium red and DNP on calcium deposition in spicules, growth of pluteus arms and of archenteron. Ruthenium red (A) and DNP (B) were introduced to embryo culture 15 hr after ferti- lization, being kept at 20°C. Embryos were cul- tured for another 30 hr. Calcium content in spicule fraction, length of pluteus arms and of archenterons were measured as described in Materials and Methods. Percentages of calcium contents in spi- cules of embryos thus treated with these com- pounds to those of normal embryos, shown with open circles (©), are the means of three experi- ments. Calcium contents in spicules of control embryos obtained in three experiments were 2.42, 2.63 and 2.39 wmol/mg embryo protein, respective- ly. Percentages of length of pluteus arms and archenteron in thus treated embryos to those in control ones, shown with open triangles (A: pluteus arm) and open squares ((: archenteron), are the means of three experiments, in which those in 30 embryos are measured in all embryo cultures, re- spectively. Vertical bar indicates SE. Fifteen hr after fertilization, embryos did not have any spi- calcification. The size of archenteron hardly decreased at concentrations of ruthenium red lower than 75 4M and became considerably small at higher concentrations. Sensitivity of archenter- on formation to ruthenium red is markedly lower than that of spicule formation. DNP, at above 100 uM, caused a decrease in the calcium amount of spicules (Fig. 3B). The calcium amount of spicules in embryos treated with DNP at 200 “M was about 30% of the control embryos and did not decrease further even at 1 mM. Growth of pluteus arm was considerably inhibited by DNP at above 50 “M and was strongly blocked at above 200 uM (Fig. 3B). As far as examined, complete inhibition of pluteus arm formation and CaCO; deposition (spicule rods) was not obtained by DNP. The length of archenteron in the embryos kept with DNP at lower concentrations than 200 LM was almost the same as in the normal ones and became slightly short at above 500 uM (Fig. 3B). Inhibition of archenteron formation by DNP occurs at markedly higher concentrations than those to exert maximum inhibitory effect on spicule calcification and pluteus arm formation. As shown in Figure 4, calcified spicule formation in cultured micromere-derived cells was inhibited by ruthenium red (B) and DNP (C). Ruthenium red and DNP were added at 15hr of culture, at which spicule formation in the cultured cells had not been initiated yet. Photographs were taken at 45 hr of culture when well-developed spicules were observed in the control culture (A). Ruthenium red and DNP did not inhibit outgrowth of pseudopodial cables (shown with arrow heads in Fig. 4). These compounds seem to inhibit directly CaCO; production in the cables. As shown in Figure 5, the length of spicule became short in cultured micromere-derived cells at the concentra- tion ranges of ruthenium red and DNP to block calcified spicule formation in embryos. As shown in Figure 6, an increase in the length of spicule rod occurred in cultured micromere- derived cells in a period between 20 and 30 hr of culture and then, was followed by less steep cules, pluteus arms and archenterons. Hence, these values measured 45 hr after fertilization indi- cate growth of these embryonic organs in this period of development. 98 K. MitsuNAGA, Y. FuJINO AND I. YASUMASU increase in its length. This increase was instantly inhibited by adding 75 ~M ruthenium red at 15, 20, 30 and 40hr of culture. Even after cultured micromere-derived cells have been furnished with an ability to form spicule rods, these compounds are able to inhibit production of CaCO3. The same was observed using DNP (500 uM) in place of ruthenium red (data not shown). As shown in Figure 7, the rate of *Ca’* uptake in mitochondria isolated from embryos at the prism stage (35 hr after fertilization) decreased in the presence of these compounds. *Ca** uptake was evidently inhibited by DNP at above 50 uM and almost completely at above 200 uM (B). *Ca’t uptake in mitochondria was also inhibited by ruthenium red at the concentrations higher than 20 uM and complete inhibition was obtained at a me Fic. 4. Effect of ruthenium red and 2, 4-dinitrophenol (DNP) on calcified spicule formation in cultured micromere-derived cells. Ruthenium red (B, 25 uM) and DNP (C, 150 4M) were added to the culture at 15hr of culture at 20°C. Cells were photographed at 45 hr of culture. Control culture is shown in A. Arrow head indicates pseudopodial cable. Bar shows 50 um. = A B cz) E 50 es CT) 3 S (3) £& o 25 3 2 roy ” 5 2 ? Sr Ops. ea -! 190 9 500 1000 Concentration (uM) Fic. 5. Effect of ruthenium red and 2, 4-dinitrophenol (DNP) on growth of spicule rods in cultured micro- mere-derived cells. At 15 hr of culture, ruthenium red (A) and DNP (B) were added to the culture of isolated micromeres. Length of spicule rod (corre- sponding to its growth) was measured at 45 hr of culture as described in Materials and Methods. Each value shows the mean+SE of three different experiments. 60 | = We & 50 | yA & | V ~_-? 2° 40 3 SNe & 30 ro J 5 20 2 ~ A—h 5 10 pty re) e— 0 — 4 — A — AA — 0 10 20 30 40 50 Time in culture (hr) Fic. 6. Effect of ruthenium red on the growth of spi- cule rods in cultured micromere-derived cells. Ruthenium red at 75 uM was introduced at 15 (a), 20 (a), 30 (G) and 40 hr (m) of culture to these descendant cells of isolated micromeres. Arrows indicate the times of adding ruthenium red. Con- trol culture is shown by using solid circles (@). The values shown were means of 90 cell clusters mea- sured in different 3 cultures. Bar shows SE. Ruthenium Red and Embryogenesis 99 B % \ “°Ca** uptake in mitochondria (cpm x 10°/mg protein/5 min) fo) 50 Concentration (uM) 100 0 250 500 Fic. 7. Inhibition of Ca** uptake in mitochondria iso- lated from the embryos at the late gastrula stage by ruthenium red and 2, 4-dinitrophenol (DNP). Mitochondria fraction was obtained from embryos at the prism stage (35 hr after fertilization, being kept at 20°C). ‘°Ca?* uptake in mitochondria in the presence of ruthenium red (A) or DNP (B) was measured as described in Materials and Methods. above 50 uM (A). Concentration ranges of these compounds to inhibit Ca’* uptake in mitochon- dria were almost the same as those to inhibit calcified spicule formation in embryos and in cultured micromere derived cells. The ATP level in embryos at the late gastrula stage was 13.9+2.4 nmol/mg protein. The levels in embryos at the late gastrula corresponding stage, having been kept with 500 ~M DNP and 100 uM ruthenium red for 10 hr, were 10.5+4.1 and 14.1+4.0 nmol/mg protein, respectively. In the embryos kept with DNP, the ATP level is not so low as expected. The ATP level in DNP-treated embryos may be maintained by glycolysis system. The level in ruthenium red-treated cells was as high as in control embryos. These suggest that the inhibition by these compounds of calcified spicule formation, as well as growth of pluteus arm and archenteron, is not due to a shortage of ATP. DISCUSSION In the present study, it was found that ruthe- nium red inhibited spicule formation in cultured cells derived from micromeres of sea urchin embryos but did not block outgrowth of pseudopo- dial cables along which spicule rods were to be produced. These indicate that the inhibition of spicule rod formation by ruthenium red does not result from failure of outgrowth of pseudopodial cables. It was also observed that spicule rod formation was inhibited by ruthenium red even after spicule formation had been initiated. This indicates that spicule formation is inhibited by ruthenium red in the cells having been furnished with an ability to produce spicule rods. The inhibition of spicule rod formation by ruthenium red does not seem to result from possible blockage of mesenchyme cell differentiation into spicule forming cells but is probably ascribed to direct inhibition of several reactions in CaCO, produc- tion. Ruthenium red is an inhibitor of ATP- dependent, H*-gradient mediated Ca?* uptake in mitochondria [8, 17, 18]. Inhibition of spicule rod formation by this compound suggests that Ca’* uptake in mitochondria participates in spicule formation. Indeed, DNP, an uncoupler of oxida- tive phosphorylation to cause a discharge of H*-gradient in mitochondria, also inhibited spi- cule rod formation in a manner essentially similar to ruthenium red. These compounds inhibited *Ca?*+ uptake in mitochondria isolated from sea urchin embryos. The blockage of spicule rod formation occurred in almost the same concentra- tion ranges of ruthenium red and DNP for the inhibition of Ca’* uptake in mitochondria. These suggest that Ca** transport in mitochondria par- ticipates in spicule formation or CaCO; deposi- tion. DNP is also well known to make electron transport through mitochondrial respiratory chain uncoupled to oxidative phosphorylation, to cause a failure of ATP production in mitochondria. De- crease in ATP level certainly reduces the rates of ATP-dependent reactions in spicule formation, such as active ion transport across the membrane catalyzed by Cl’, HCO; -ATPase [28, 29] and H*, K*-ATPase [30]. In the embryos kept with DNP, as well as ruthenium red, however, the ATP level was almost the same as in control embryos. Failure of Ca** uptake caused by ruthenium red and DNP in mitochondria probably results in an 100 K. MitsuNAGA, Y. FUJINO AND I. YASUMASU increase in cytosolic Ca?* level. Also in sea urchin eggs, an uncoupler of oxidative phosphorylation has been found to cause a marked increase in cytosolic free Ca*t level [14]. Ca?* uptake through Ca** channels is reportedly high in its rate in cells with low cytosolic free Ca** [31]. Hence, a high level of cytosolic free Ca** probably causes a decrease in Ca?* influx into cells, with which Ca’*, a main material for the spicule rod forma- tion, is supplied from external sea water. Indeed, in spicule forming cells as well as in the other cells of sea urchin embryos, Ca?* uptake into cells is blocked by ruthenium red [2, 3], though this compound is known to exert any inhibitory effect on Ca?* channels. However, high level of cytoso- lic Ca?+, which is assumed to be rather favorable for Ca’* transport into skeletal vacuoles across their membrane, does not seem to be maintained, if Ca** is utilized for spicule rod formation. Thus, low rate of Ca** uptake through Ca** channels in spicule forming cells rather results from the inhibi- tion by these compounds of spicule formation. Uptake of Ca?+ in mitochondria, seems to be indispensable for Ca** supply to skeletal vacuoles across their membrane in spicule forming cells. Condensation of Ca** may occur in mitochondria to supply high concentration of Ca** into skeletal vacuoles. Ruthenium red is also known to inhibit Ca’*- ATPase in plasma membrane, a Ca** pump [32, 33]. If the inhibition of Ca’* pump in skeletal vacuoles by ruthenium red occurs in spicule forming cells, supply of Ca’* into skeletal vacuoles may be blocked to cause a failure of CaCO; deposition in the vacuoles. However, DNP, which does not inhibit this ATPase, con- siderably blocks spicule formation. These may be a reason why ruthenium red exerts stronger inhibitory effect on spicule formation than’ DNP, but it is unlikely that the inhibition of spicule formation by these compounds is solely due to inhibition of Ca?*+ transport by a Ca** pump in skeletal vacuole membrane. Ruthenium red also exerted strong inhibitory effects on morphogenesis other than spicule formation in embryos. In embryos kept with ruthenium red, pluteus arm formation was com- pletely inhibited at concentrations lower than those for complete blockage of spicule formation. Also in DNP treated embryos, pluteus arm forma- tion was inhibited more strongly than spicule formation, though their complete inhibition was not obtained at the concentrations of DNP ex- amined in the present study. These suggest that the blockage of pluteus arm formation by these compounds does not result from the failure of spicule rod formation. Ruthenium red and DNP, causing an increase in cytosolic Ca**, probably result in excess stimulation of Ca**-dependent enzymes. Excess stimulation of these enzymes may change cell functions to cause a failure of pluteus arm formation. It has also been found that embryos develop to abnormal ones with poor pluteus arms, no spicules and quasi-normal archenterons in the presence of Ca’* antagonists [1, 2, 28], which probably reduced cytosolic free Ca**+. Low cytosolic Ca?*+ probably prevents Ca**-dependent enzymes from the stimulation. It is likely that abnormal activities of Ca?t- dependent cell functions, either being excessively stimulated by ruthenium red and DNP or remain- ing unstimulated in the presence of Ca?* antagon- ists, fail to support morphogenesis such as pluteus arm formation. On the other hand, archenteron formation was hardly inhibited by ruthenium red and DNP. The same is found in embryos kept in the presence of Ca’** antagonists [1, 2, 28]. Archenteron is formed in blastocoel in which spicules are also produced. Spicule formation in cultured micromere-derived cells was inhibited by ruthenium red and DNP, as well as by Ca*t antagonists [28], at almost the same concentrations for its inhibition in embryos. Hence, concentrations of these compounds in blastocoel are assumed to be almost the same as in media surrounding embryos. Archenteron can be constructed with cells in which Ca”*-dependent cell functions are excessively stimulated or remain quite low in their activities, though the cells with adequate activities of Ca*t-dependent cell func- tions are necessary for the formation of pluteus arms and spicules. Contribution of Ca’*- dependent reactions to morphogenesis may be quite low in the formation of archenteron as compared to that in pluteus arm formation. One of Ca’*-dependent reactions to support the mor- Ruthenium Red and Embryogenesis 101 phogenesis may be protein kinase C. Several observations to support the assumption mentioned above will be reported elsewhere. ACKNOWLEDGMENTS This work was supported by Grants-in-Aid for Scien- tific Research (Nos. 60223028 and 61304009) from the Ministry of Education, Science and Culture, Japan. REFERENCES 1 Mitsunaga, K., Fujino, Y., Fujiwara, A. and Yasu- masu,I. (1984) Anion transport participates in spicule calcification of sea urchin embryos. Dev. Growth Differ., 26: 375, Abstract A33. 2 Yasumasu,I., Mitsunaga, K. and Fujino, Y. (1985) Mechanism for electrosilent Ca** transport to cause calcification of spicules in sea urchin embryos. Exp. Cell Res., 159: 80-90. 3 Fujino, Y., Mitsunaga, K. and Yasumasu, I. (1985) Inhibition of *°Ca?* uptake in the eggs and embryos of the sea urchin, Anthocidaris crassispina, by several calcium antagonists, anion transport inhibi- tor and chloride transport inhibitors. J. Exp. Zool., 235: 281-288. 4 Harkey,M. A. and Whiteley, A. H. (1980) Isola- tion, culture, and differentiation of echinoid primary mesenchyme cells. Wilhelm Roux Arch. Dev. Biol., 189: 111-122. 5 Okazaki, K. (1975) Spicule formation by isolated micromeres of the sea urchin embryo. Am. Zool., 15: 567-581. 6 Gibbins,J.R., Tilney,L.G. and Porter, K. R. (1969) Microtubules in the formation and develop- ment of the primary mesenchyme in Arbacia punc- tulata. 1. The distribution of microtubules. J. Cell Biol., 41: 201-226. 7 Inoue, S. and Okazaki, K. (1977) The development of an elaborate crystalline skeleton within the transparent embryo of the sea urchin provides a model for the shaping of hard mineralized tissues such as shell and bone. Sci. Am., 236: 82-92. 8 Moore,C.L. (1971) Specific inhibition of mitochondrial Ca?* transport by ruthenium red. Biochem. Biophys. Res. Commun., 42: 298-305. 9 Thorne,R.F.W. and Bygrave, F.L. (1974) The role of mitochondria in modifying the cellular ionic environment. Biochem. J., 144: 551-558. 10 Landry, Y. and Lehninger, A. L. (1976) Transport of calcium ions by Ehrlich ascites-tumor cells. Biochem. J., 158: 427-438. 11 Carafoli, E. and Crompton, M. (1978) The regula- tion of intracellular calcium by mitochondria. Ann. 12 14 15 16 17 19 20 21 22 23 24 25 26 27 N. Y. Acad. Sci., 307: 269-284. Rasmussen, H. and Gustin,M.C. (1978) Some aspects of the hormonal control of cellular calcium metabolism. Ann. N. Y. Acad. Sci., 307: 391-402. Carafoli, E. (1986) Membrane transport in the cellular homeostasis of calcium. J. Cardiovasc. Pharmacol., 8: suppl. 8, s3-s6. Eisen, A. and Reynolds, G. T. (1985) Source and sinks for the calcium released during fertilization of single sea urchin eggs. J. Cell Biol., 100: 1552-1527. Brinley, F. J., Tiffert, T., Scarpa, A. and Mullins, L. J. (1977) Intracellular calcium buffering capacity in isolated squid axons. J. Gen. Physiol., 70: 355- 384. Baker, P.F. and Singh, R. (1981) Influence of vanadate on calcium fluxes and net movement of calcium in intact squid axons. Biochim. Biophys. Acta, 646: 450-456. Vasington, F.D., Gazzotti,P., Tiozzo,R. and Carafoli, E. (1972) The effect of ruthenium red on Ca?* transport and respiration in rat liver mitochon- dria. Biochim. Biophys. Acta, 256: 43-54. Reed, K. C. and Bygrave, F. L. (1974) The inhibi- tion of mitochondrial calcium trasport by lanth- anides and ruthenium red. Biochem. J., 140: 143- 155. Baker, P. F., Knight,D.E. and Umbach, J. A. (1985) Calcium clamp of the intracellular environ- ment. Cell Calcium, 6: 5-14. Rigoni, F. and Deana, R. (1986) Ruthenium red inhibits the mitochondrial Ca** uptake in intact bovine spermatozoa and increases the cytoplasmic Ca** concentration. FEBS Lett., 198: 103-108. Kitajima, T and Matsuda, R. (1982) Specific protein synthesis of sea urchin micromeres during dif- ferentiation. Zool. Mag., 91: 200-205. Kitajima, T. and Okazaki, K. (1980) Spicule forma- tion in vitro by the descendants of precocious micromere formed at the eight-cell stage of sea urchin embryo. Dev. Growth Differ., 22: 265-279. Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193: 265-275. Lamprecht, W. and Trautschold, I. (1963) Adeno- sine-5’-triphosphate. Determination with hexoki- nase and glucose-6-phosphate dehydrogenase. In “Methods of Enzymatic Analysis”. Ed. by H. U. Bergmeyer, Acad. Press., N. Y., pp. 265-267. Yasumasu,I., Asami, K., Shoger, R.L. and Fu- jiwara, A. (1973) Glycolysis of sea urchin eggs. Exp. Cell Res., 80: 361-371. Gustafson, G. T. and Pihl, E. (1967) Histochemical application of ruthenium red in the study of mast cell ultrastructure. Acta Path. Microbiol. Scand., 69: 393-403. Gordon, M., Fraser,L.R. and Dandekar, P. V. 28 29 30 102 (1974) The effect of ruthenium red and concanava- lin A on the vitelline surface of fertilized and unfertilized rabbit ova. Anat. Rec., 181: 95-112. Mitsunaga, K., Makihara, R., Fujino, Y. and Yasu- masu, I. (1986) Inhibitory effects of ethacrynic acid, furosemide, and nifedipine on the calcification of spicules in cultures of micromeres isolated from sea-urchin eggs. Differentiation, 30: 197-204. Mitsunaga, K., Fujino, Y. and Yasumasu, I. (1986) Change in the activity of Cl”, HCO; -ATPase in microsome fraction during early development of the sea urchin, Hemicentrotus pulcherrimus. J. Biochem., 100: 1607-1615. Mitsunaga, K., Fujino, Y. and Yasumasu, I. (1987) Probable role of allylisothiocyanate-sensitive H*, 31 32 33 K. MitsunaAGA, Y. FUJINO AND I. YASUMASU K*-ATPase in spicule calcification in embryos of the sea urchin, Hemicentrotus pulcherrimus. Dev. Growth Differ., 29: 57-70. DiPolo, R. (1979) Calcium influx in internally dialyzed squid giant axons. J. Gen. Physiol., 73: 91- 113. Watson, E.L., Vincenzi,F.F. and Davis, P. W. (1971) Ca**-activated membrane ATPase: Selec- tive inhibition by ruthenium red. Biochim. Biophys. Acta, 249: 606-610. Tsuruo, T., lida, H., Tsukagoshi, S. and Sakurai, Y. (1981) Prevention of vinblastine-induced cytotoxicity by ruthenium red. Biochem. Pharma- col., 30: 213-216. ZOOLOGICAL SCIENCE 5: 103-107 (1988) © 1988 Zoological Society of Japan Spawning of Three Intraspecific Groups of the Ascidian, Halocynthia roretzi (Drasche), in the Wild, and Fertilization among Them TAKAHARU NUMAKUNAI, ZEN-ICHIRO Hosuino! and SHOoGO KAJIWARA Marine Biological Station, Tohoku University, Asamushi, Aomori 039-34, and ‘Department of Biology, Faculty of Education, Iwate University, Ueda, Morioka 020, Japan ABSTRACT— Spawning seasons and spawning times of three intra-specific groups of the ascidian, Halocynthia roretzi, were observed in the wild. They released gametes at the same time as in the laboratory. December. One of them, which spawns in April in the laboratory, spawned in the middle of This was caused by low sea water temperature in the laboratory. Experimentally, fertilization among the three Types was possible, and the resulting embryos gave rise to normal young adults. Experimental results and observations of spawning in the wild suggest the possibility of cross fertilization between Type A and Type C. INTRODUCTION We have reported of the ascidian, Halocynthia roretzi, that in Mutsu Bay there are three intraspe- cific groups (Types A, B and C), which have different spawning seasons and spawning times [1, 2]. The November, Type A spawning occurs in the morning, and the late October to mid-November, Type B spawning occurs in the evening, while the April, Type C group spawns at noon. We had previously investigated their spawning behavior under continuous light conditions, external charac- teristics and distribution in the bay, and allogeneic cellular reactions by blood cells [3-5]. During these investigations, we found that the most conspicuous difference was in spawning behavior under continuous light conditions, leading us to question whether these three types would spawn in the wild as they do in the laboratory. The survey of gamete release in the wild revealed that Types A and B have the same spawning season and time in the wild as they do in the laboratory, whereas Type C showed gamete release at a different season from that of the laboratory. Our observa- tions in the wild being: Type A animals begin to Accepted August 18, 1987 Received April 17, 1987 release their gametes near the end of Type B spawning season; Type C animals begin to spawn near the end of Type A spawning season, leading to a spawning season overlap. In addition, we discovered a location in the bay where Type C animals begin to spawn around noon, before Type A animals have completed their gamete discharge. These observations and the fact that many speci- mens are difficult to classify them into a definite type because of variations in their external charac- teristics [4] motivated us to investigate the degree of cross fertilization among these three groups. Here we report new observations on spawning and the possibility of cross fertilization among three groups in the wild. MATERIALS AND METHODS Spawning in the wild Since the spawning seasons in the laboratory varied a little from year to year, we extended our period of observation in the wild from 1981 to 1985. From August to November, adults of each Type were collected by scuba diving at several locations in the bay and were kept in an aquarium supplied with running sea water. After gamete release started by these laboratory animals, we 104 T. NUMAKUNAI, Z. HOSHINO AND S. KAJIWARA dived to observe the spawning of each Type in the wild. When gamete release was recorded in the wild, these animals were brought into the labora- tory to see if they would continue the discharge as in the wild. Fertilization and young adult formation To obtain gametes of the three Types at the same time, animals of Types A and B were kept at a low temperature (7°C) until Type C animals began to spawn. Some Type B animals were kept under continuous light at low temperatures to secure gametes in the morning [3]. After Type C animals began to spawn, the animals of Types A and B which had been kept at low temperatures were transferred to warm sea water (13°C) the day before the gametes were desired. After gamete discharge, the eggs were separated from the sperm suspension by filtration through nylon mesh (gauge =0.25 mm), and the sperm suspension was used for insemination. Fertilization was confirmed by observing the expansion of the perivitellin space. Developing embryos were kept at 13°C until they hatched out of the chorion. They were examined just before hatching to distinguish abnormal from normal embryos. Actively swim- ming larvae were selected and kept in plastic dishes to observe young adult formation. After attaching to the dishes and completing meta- morphosis, the young adults were kept in aquaria supplied with running sea water until July. To test for the possibility of fertilization between Types A Season Oct. Nov. and B in the wild, gametes of Type A discharged in the morning by natural spawning were kept at 13°C (sea water temperature in the wild) until evening when Type B animals started to release their gametes. Fertilization between them was checked as described above. This was possible because H. roretzi is strictly self sterile. RESULTS Spawning in the wild In late October, we observed some Type B animals attached to rocks at 10-20 meters, starting to release gametes at 4.00 p.m. In mid-November, Type A animals began to spawn at 9.30a.m. At that time, Type B animals were collected and brought into the laboratory where we found they continued to release a small quantity of gametes in the evening of the same day and for several days after. At 10.300 a.m. on November 21, 1984, we found three animals in the wild with typical Type C external characteristics discharging gametes, although no other Type C animals were spawning. These three animals were brought into the labora- tory for observation, where they continued to spawn in the morning for several days more, as did the Type A animals previously mentioned. By mid-December, some Type A animals were still spawning around noon, although gamete release was not as vigorous as at the peak of the spawning season. At one location in the bay, Type Morning 6 Type A Noon 12 i Evening 18 Fic. 1. Spawning seasons and spawning times in the wild. Type C begins to spawn in the middle of December, and the spawning season is prolonged to the following April by low sea water temperature. Spawning and Fertilization of H. roretzi TABLE 1. Fertilization among three Types Sperm Type A B (@ Egg Type A B C A B C A B C No. of eggs: Fertilized 335 342 334 351 361 344 356 332 365 Unfertilized 0 0 0 0 0 0 0 0 0 No. of larvae: Normal 202 269 295 300 308 292 285 274 207 Abnormal 83 73 39 51 53 52 vA 58 58 Fertilization among three Types was completely successful. No significant difference was observed between cross fertilization among three Types and fertilization within the same Type. TABLE 2. Fertilization between Type A and Type B Sperm Type A B A B Egg Type A B B A Insemination one 11.00 17.30 17.30 17.30 17.30 Fertilized 384 84 373 117 364 Unfertilized 0 301 0 254 0 Eggs of Type A animals discharged in the morning were fertilized by Type B sperm later in the day. C animals living in close proximity to Type A animals started releasing their gametes before Type A animals had stopped discharging theirs. These animals of Types A and C were brought into the laboratory and kept at the same temperature as in the wild. They all spawned as in the wild. Observations made of the spawning seasons and spawning times of Types A, B, and C animals are summarized in Figure 1. Fertilization, development and post-metamorphic growth The experiments were repeated several times; the results of a typical experiment are shown in Table 1. As shown in the table, all the eggs were successfully fertilized regardless of the combina- tion. Most embryos gave rise to normal larvae with active swimming movement, but some de- veloped into larvae with bent tails. No significant differences were observed when evaluating the results of cross fertilization among the three groups and fertilization within the same group (Table 1). Further cross fertilization experiments were made between Types A and B in mid-November. Eggs of Type A animals discharged in the morning were fertilized by Type B sperm latter in the day (Table 2). In this experiment, we discovered that Type A sperm could not fertilize Type B eggs, but when Type A eggs were inseminated with Type B sperm, all of the eggs were fertilized and de- veloped normally. The actively swimming larvae obtained were transferred to plastic dishes. Some larvae resorbed the tail on the water surface, which were dis- carded. All the larvae that were able to attach to the plastic dishes gave rise to young adults, and survived up to July, having grown to 1.8 mm. DISCUSSION In this study, it was confirmed by scuba diving that animals of three groups (Types A, B and C) discharged their gametes at the same time in the wild as they did in the laboratory, exception made for the three animals found in November, 1984, 106 T. NuMAKuNAI, Z. HOSHINO AND S. KAJIWARA with Type C characteristics and whose spawning was observed in the morning. However, a big difference of spawning season was found between Type C animals in the wild and in the laboratory. We previously reported that Type C animals in the laboratory began to spawn in April and that they could release their gametes at any time from January to March if they were kept in warm sea water [2, 3]. The apparent reason for the observed difference seemed to be the sea water tempera- ture. In the laboratory, surface sea water is supplied and its temperature is easily affected by air temperature. When we dived in_ mid- December, the sea water temperature at depths of 10 to 20 meters was 10°C and that of the water in the laboratory was 8C; the air temperature fluctuated approximately between S°C to —4°C. In this season, Type C animals fail to spawn in the laboratory, if they are not kept in warm sea water (higher than 10°C). By March, the sea water temperature had gradually decreased to 3°C. However, from mid-March, it gradually increased until it reached 10°C in April and Type C animals began to spawn. During January, February and March we collected Type C animals which had been reared by fishermen and found that their gonads varied from year to year, depending on the sea water temperature. In mid-January (1981), the gonads of Type C animals were empty, indicating the end of the spawning season. In early April of 1985 and 1986, we found that Type C animals had fully matured gonads and were able to release gametes when treated with warm sea water. Soon after Type C animals in the wild reach spawning season in mid-December, the sea water temperature begins to decrease. If it reaches too low a temperature before all the gametes have been released, the spawning season is prolonged until the following April when sea water tempera- ture gradually increases again. Experimentally, fertilization within these three groups (Types A, B and C) was possible and the embryos gave rise to normal young adults. Be- cause we found that there were many animals which were difficult to assign to a group, and that three animals with typical characteristics of Type C were releasing their gametes in the morning of November 21, 1984, the question of cross fertiliza- tion among the groups in the wild arises. The fertilization of Type A eggs by Type C sperm seems to be the most probable explanation, since we observed sperm release of Type C animals before Type A animals had ended gamete release. As shown in Table 2, Type A eggs released in the morning were completely fertilized by Type B sperm released in the evening. But in the wild, it seems plausible that Type A eggs have a better chance of being fertilized by Type A sperm than Type B sperm. It is interesting to note that genetically compatible Corella species are main- tained as discrete species by an 8 minute difference in spawning time [6]. If, however, our supposition is correct, it is strange that there are more animals that are difficult to classify as either Type B or Type C than there are animals difficult to confirm as either Type A or Type C. In order to clear up this problem, it is necessary to obtain mature animals by cross fertilization. In the laboratory, we have not yet been able to get fully mature adults from either the cross or the intra-group fertilizations. Recently, we have been successful in rearing them in the laboratory for about a year. It would seem that at least two years are necessary to harvest mature animals from embryos. In the near future, we are hoping that these young adults will provide offspring able to give us the information needed to formulate more precise conclusions. ACKNOWLEDGMENT We are grateful to Prof. C. Lambert for critical reading of manuscript. We also thank Mr. T. Mayama, Mr. S. Tamura and Mr. M. Washio for scuba diving with us. REFERENCES 1 Numakunai, T. and Hoshino, Z. (1973) Biology of the ascidian, Halocynthia roretzi (Drasche), in Mutsu Bay. I. Differences of spawning time and external features. Bull. Mar. Biol. Stat. Asamushi, Tohoku Univ., 14: 191-196. 2 Numakunai, T. and Hoshino, Z. (1974) Biology of the ascidian, Halocynthia roretzi (Drasche), in Mutsu Bay. II. One of the three Types which has the spawning season and time different from two others. Bull. Mar. Biol. Stat. Asamushi, Tohoku Univ., 15: 23-27. Spawning and Fertilization of H. roretzi 107 3 Numakunai, T. and Hoshino, Z. (1980) Periodic spawning of three Types of the ascidian, Halocynthia roretzi (Drasche), under continuous light conditions. J. Exp. Zool., 212: 381-387. Numakunai, T., Hoshino, Z. and Hori, R. (1981) Biology of the ascidian, Halocynthia roretzi (Dra- sche), in Mutsu Bay. III. Distribution and external characteristics of three Types. Annot. Zool. Japon., 54: 230-239. 5 Fuke, M. and Numakunai, T. (1982) Allogeneic cellular reactions between intra-specific types of a solitary ascidian, Halocynthia roretzi. Dev. Comp. Immunol., 6: 253-261. Lambert, G., Lambert, C. and Abbot, D. P. (1981) Corella species in the American Pacific Northwest: distinction of C. inflata Hunstman, 1912 from C. willmeriana Herdman, 1898 (Ascidiacea, Phlebo- branchia). Can. J. Zool., 59: 1493-1504. ' ipl iat hk 7 ' } 7 « iy -) es i ” e 7 Dy 3) ‘te 76 Be vohe eS we ae a , Wh eae : : - a . cy AX ' z y a a Crag nd a kel . Seglvonglh . ea ine ria 7 (se s 7 s on : 2 he “a, PON N yeu ~ ZOOLOGICAL SCIENCE 5: 109-118 (1988) © 1988 Zoological Society of Japan Normal Stages of Development in the Lamprey, Lampetra reissneri (Dybowski) YUTAKA TAHARA Department of Biology, Osaka Kyoiku University, Tennojiku, Osaka 543, Japan ABSTRACT— A series of normal stages of embryonic development in the lamprey, Lampetra reissneri (Dybowski) is presented. The developmental processes from the ovulated but unfertilized egg to the ammocoete larva just before burrowing are divided into 31 stages using as identifying criteria of the stages the externally visible changes such as the appearance of structural changes, the beginning of movements and the body size. The age is determined in terms of an average time at which the stages occur at a temperature of 15°C. The stages and developmental tempos of Lampetra reissneri are compared with those of Petromyzon marinus reported by Piavis. INTRODUCTION Studies on embryonic development of the Lam- prey have been attracting the attention of embryologists because of the phylogenetic position of this animal group. There is a good deal of literature on this subject. That consists mainly of descriptions of the separate phases in mor- phogenesis; for instance, the cleavage [1, 2], gastrulation [3-6] or the formation of the head [7, 8], the heart [9], the pronephros [10] and the blood vessels [13]. Around the middle of this century the Great Lakes fisheries suffered a threat of disaster which originated in the increase in the number of parasitic lamprey, Petromyzon marinus. Among the research programs aimed at controlling this menace, Piavis took charge of examining the effects of temperature on embryonic development in this species [11]. He also carried out a staging of embryonic development of the materials because the determination of standard stages was consi- dered to be essential to a better understanding of the results of the experiments [11, 12]. Piavis subdivides the developmental processes, from the ovulated but unfertilized egg to the first stage of ammocoete larva, into 19 stages. Accepted July 22, 1987 Received April 16, 1987 In Japan Hatta published a series of papers which dealt with the morphological descriptions on the development of one of the Japanese species of the Lamprey' [3-5, 9, 10, 13]. However, he did not attempt to establish the developmental stages of his materials. The present work deals with the staging of the embryonic development of Lam- petra reissneri (Dybowski) from the ovulated but unfertilized egg to the ammocoete larva just before burrowing. It intends to put this standard to use someday for the experimental studies of embryogenesis of the lamprey. MATERIALS AND METHODS The sexually matured adults were collected in one of the brooks which flow into Lake Utonai near Sapporo City in Hokkaido during the breed- ing seasons of 1981 and 1982. Fertilized eggs were obtained by artificial insemination. The testis was taken out from a male and sperm was kept in a dry form in a petri dish until used. Ovulated eggs were allowed to extrude on a sheet of filter paper which was laid on a petri dish and moistend with water. Fertilization was accomplished by diluting dry ‘In his 1891 paper [3], Hatta wrote his material was Petromyzon planeri or a variety of it, but in 1901 paper [13] Lampetra mitsukurii Hatta, and in others [4, 9, 10] only Petromyzon. 110 Y. TAHARA sperm in tap water and adding the sperm suspen- sion to eggs. Immediately after the addition of sperm, filtrated tap water was added and the filter paper was taken away. Embryos were reared in the petri dish kept at 15°C. The developmental stages were determined mainly by externally visible changes such as the appearance of structural changes, the beginning of muscle contraction and the body size (Figures 1- 4). The age was defined in terms of an average time at which the stages occurred at a temperature of 15°C. Internal changes were also observed histologically on sections of each stage of embryos. For histological preparations, embryos were fixed in Zenker acetic for 5 hr, serially sectioned at 7 ~m and stained with borax carmin and _ Pikro- Blauschwarz. DEVELOPMENTAL STAGES StageO0. Age Ohr. Piavis 0 Egg is creamy-white and oval in shape with narrow animal pole region. It is surrounded by vitelline coat to which sticky substance attaches. Size about 0.80.6 mm. Stage l. Age 0.07hr after fertilization. cone. Piavis 1 Animal pole region becomes flat and _peri- vitelline space appears. Extrusion of a polar cone at depressed polar region. Stage 2. Age 0.3 hr. Polar spot. Piavis 1 Polar cone is absorbed into egg cytoplasm leaving behind a polar spot. Egg becomes gradual- ly spherical. Size about 1.0 mm. Stage 3. Age 6.5 hr. Two cells. Piavis 2 Beginning of first cleavage. Cleavage furrow appears at animal pole and extends towards vegetal pole. Egg divides into two blastomeres with approximately equal size. Stage 4. Age 11.5 hr. Four cells. Piavis 3 Second cleavage furrow appears at animal pole. Division occurs meridionally producing four blas- tomeres with approximately equal size. Stage5. Age 15.5 hr. Eight cells. Piavis 4 Third cleavage begins. Cleavage furrow appears horizontally in general, meridionally in eggs of some batches. Blastomeres in animal hemisphere Ovulated, unfertilized egg. Polar are smaller than those in vegetal. Stage6. Age 20hr. Twelve to sixteen cells. Piavis 5 External: Fourth cleavage begins in animal hemisphere. Cleavage furrows appear meridional- ly or horizontally depending upon cleavage type in stage 5. Internal: Blastomeres arrange themselves in one cell thick. Segmentation cavity appears. Stage 7. Age 24hr. Twenty-four to thirty-two cells. Piavis 6 Fifth cleavage furrows appear in blastomeres of animal hemisphere. Stage 8. Age 28 hr. Morula. Piavis 7 Sixth cleavage begins in animal hemisphere. Stage9. Age 32hr. Early blastula. Piavis 8 External: Seventh cleavage begins in animal hemisphere. Animal half makes smooth surface, vegetal half being still rough. Internal: Segmentation cavity enlarges and blastocoel arises. The roof of blastocoel is com- posed of two cell layers and floor of three cell layers. Stage 10. Age 48hr. Mid blastula. Piavis 8 External: Blastocoel becomes visible from out- side through its thin and translucent roof. Internal: Blastocoel further expands, roof of which is composed of three cell layers and floor of five to seven cell layers. Stage ll. Age 58hr. Late blastula. Piavis 8 Roof of blastocoel becomes much thinner and transparent. Vegetal hemisphere makes smooth surface. Embryo becomes spherical in shape with smooth surface. Stage 12. Age 3 days. Dorsal cone; Gastrula I. Piavis 8 External: In majority of embryos one or two conical protuberances (dorsal cones) appear in the dorsal subequatorial region. Blastocoel becomes larger and expands below equator. A groove appears at boundary between thin roof of blasto- coel and thick vegetal yolk mass. Internal: Roof and lateral wall of blastocoel become two cell thick. A vertical slit arises between outer thin layer and inner yolk cell mass on dorsal side. Stage 13. Age 3'/, days. Brow-shape blastopore; Gastrula II. Piavis 9 Developmental Stages of Lamprey 111] External: A groove of brow-shape blastopore appears above the dorsal cone. Boundary groove bends towards animal pole on dorsal side. Internal: Roof and lateral wall of blastocoel become one cell thick. The vertical slit becomes deeper on dorsal side. It appears also on lateral and ventral sides. Stage 14. Age 3'/, days. Hemi-circular blasto- pore; Gastrula III. Piavis 9 External: Blastopore becomes hemi-circle, horseshoe or A-shape. Blastocoel becomes smal- ler than the preceeding stage. Boundary groove bends nearly vertical. Internal: Archenteron makes first appearance due to advancement of mesoderm invagination. Stagel5. Age 4 days. Elliptical blastopore; Gastrula IV. Piavis 9 External: The groove of blastopore becomes elliptical in shape. Blastocoel becomes much smaller, still visible in anterior part of embryo. Both anterior and posterior ends of embryo become taper. Internal: Archenteron elongates forming a narrow tube. Its roof is composed of mesodermal cells of two-cell thick. Stage 16. Age 4'/, days. Flat dorsal lip; Gastrula V. Piavis 9 External: Dorsal blastopore lip becomes flat. Blastocoel can not be seen from outside. Posterior end of embryo protrudes. Internal: Blastocoel becomes vestigial, still visible on antero-ventral side of embryos. Notochord begins to differentiate at the posterior part of archenteron roof. Ventral yolk cells are still uncovered. Stage 17. Age 5 days. Neural groove; Neurula I. Piavis 10 External: Formation of neural plate. A neural groove appears in the middle of the neural plate. Its length is about one fourth of circumference of embryo. Internal: Archenteron further elongates and reaches anterior end of embryo where it forms foregut. Yolk cells are mostly covered by epidermis. Stage 18. Age 5'/, days. Neural folds; Neurula II. Piavis 10 External: Elevation of neural folds on both sides of the neural groove. Neural groove elon- gates up to about one third of circumference of embryo. Internal: Neural anlage starts sinking under epidermis. Notochord begins to separate from somitic mesoderm in trunk region. Mesodermal sacs appear in archenteron of the head region. Stage 19. Age 6 days. Elevation of neural folds; Neurula III. Piavis 10 External: Neural folds further elevate and a deep neural groove is formed between them. It elongates up to about one half of circumference of embryo. Internal: Segmentation begins in somitic mesoderm. Endodermal cells cover the archenter- on roof in trunk region. Stage 20. Age 6'/, days. Neural rod; Neurula IV. Piavis 11 External: Neural folds contact and fuse in the dorsal midline. Anterior end of embryo begins to protrude. Swelling of foregut region. Body length is about 1.3 mm in crown-rump. Internal: Neural rod is formed and covered by epidermis. Notochord entirely separates from somitic mesoderm. Endodermal cells cover the archenteron roof in its total length. Stage 21. Age 7'/, days. Head protrusion I. Piavis 12 External: Head protrudes in front of yolk mass, being about 0.3 mm in length. Appearance of cheek-like swellings on both sides of head. Internal: Neural rod expands at its anterior- most part. The first visceral pouch attaches to epidermis. Appearance of the second visceral pouch. Liver diverticulum appears. Formation of proctodaeum. Anterior end of notochord sepa- rates from prechordal plate. The anteriormost myotomes separate from foregut wall and pre- chordal plate. Lateral plates are formed in trunk region. Stage 22. Age 8 days. Neural tube; Head protru- sion II. Piavis 13 External: Head further protrudes making an acute angle against yolk mass. It is about 0.4 mm in length, yet shorter than height of yolk mass. Proboscis-like sharpening of anterior end of head. A neck appears between cheek-like swellings and yolk mass. 112 Y. TAHARA Internal: A longitudinal slit appears in the nerve cord and the neural tube is formed. Mes- ectodermal cells increase in head region. Forma- tion of auditory placodes and infundibulum. Separation of the first myotome from the second. Lateral plate is formed in head region at the third myotome level. Stage 23. Age 9 days. protrusion III. Piavis 13 External: A stomodaeum appears as a longitu- dinal slit-like invagination. The cheek-like swell- ings fuse in the ventral midline. Formation of auditory pits. Head and neck elongate up to about 0.7 mm in length, nearly equal to height of yolk mass. Embryo becomes a commaz-like shape. Anus points forward. About 25 segments are formed in the somite. contraction. Internal: Formation of optic vesicles, optic stalks, lens placodes and ganglion placodes. The second visceral pouch attaches to epidermis. Formation of coelom. Differentiation of heart forming cells, sclerotome, dermatome and myo- blasts. Stage 24. Age 11 days. Piavis 13-14 External: Head and neck elongate up to about 1.0 mm in length. Appearance of a nasal pit. Through transparent skin the second to the fourth visceral pouches, endostyle, pericardial coelom, liver and pronephros become visible. Elevation of Stomodaeum; Head Somitic muscles. start Nasal pit; Hatching. dorsal fin. Embryos start hatching during this stage. Internal: Formation of lens cones and auditory vesicles. Fore-, mid- and hindbrain become dis- cernible due to formation of epiphysis and fold in cerebral commissure. The fourth visceral pouch appears. Formation of tubular heart, endocar- dium, epimyocardium, dorsal and ventral aorta, pronephric tubules and collecting ducts. Appear- ance of blood cells and primordial germ cells. Notochordal cells begin to vacuolize. Stage 25. Age 12 days. Heart beat; Tailbud I. Piavis 14 External: Heart starts beating. The stomo- daeum becomes a transverse slit-like invagination. Formation of the second to the sixth visceral pouch. Elongation of trunk. Tailbud makes first appearance. All embryos hatch out during this stage. Body length 3.5-4.0 mm. Internal: Formation of optic cups. Separation of lens vesicles from epidermis. Appearance of external naris and nasal cavity. Differentiation of white matter in central nervous system. Formation of the seventh visceral pouch. Folding of endo- style. Appearance of pronephric arteries. Stage 26. Age 16 days. Melanophores; Tailbud II. Piavis 15 External: Melanophores appear first in head, subsequently in trunk region. Appearance of hemoglobin in blood cells. Trunk becomes straight. Anus points ventralward. Body length 4.5-5.0 mm. Internal: Ectodermal layer of oral plate dis- appears. Formation of the fifth to the eighth visceral pouches and blood vessels in the first to the fourth gill arches. Obliteration of midgut lumen. Liver starts folding. Sinus venosus, auri- cle, ventricle and trunks arteriosus differentiate in the heart. Stage 27. Age 18 days. Eye spots; Tailbud III. Piavis 15 External: Pigmentation occurs in retinae. Up- per lip expands anteriorly and laterally. External naris shifts towards anterior. The eighth visceral pouch becomes visible. Trunk becomes straight. Larvae start swimming. Body length 5.5-6.0 mm. Internal: Formation of hypophysial sac. En- dodermal layer of oral plate disappears and mouth opens. A pair of velum is formed. Differentiation of gall bladder and bile duct in liver. Gill filaments begin to arise in anterior pairs of gill arches. Blood vessels appear in the fifth gill arch and in typhro- sole. Cilia appear in nephrostomes. Stage 28. Age 22 days. Velum beating. Piavis 16 External: Velum starts beating. -Upper lip further expands and oral hood is formed. Gill pores start contraction movement. Oral cirri appear. External naris further shifts antero- dorsally. Melanophores increase in number in trunk. Tip of the tail points backward. Anal tube elongates. Body length 7.0 mm. Internal: Lens vesicles become flat. Formation of definitive lumen throughout intestine. Gill filaments appear on six gill arches, from the first to the sixth. Blood vessels are formed in all eight Developmental Stages of Lamprey 113 12 Fic. 1. Before fertilization and Stages 1-12. 0, Ovulated, unfertilized egg; 1, Polar cone; 2, Polar spot; 3, Two-cell; 4, Four-cell; 5, Eight-cell; 6, Twelve to sixteen-cell; 7, Twenty-four to thirty-two-cell; 8, Morula; 9, Early blastula; 10, Mid blastula; 11, Late blastula; 12, Dorsal cone, Gastrula I. 0 and 1, View from lateral side; 2, View from lateral and slightly animal pole side; 3a—8a, View from animal pole; 3b and 4b, View from lateral and slightly animal pole side; 8b and 9-12, View from lateral side. Magnification 0 and 1, x30; 2-12, x25. 114 Y. TAHARA 13a 14a 13b 1db Sb 1éb 17a 18a 18b 19b \7e 18¢ : 19¢ - 20c Fic. 2. Stages 13-20. 13, Brow-shape blastopore, Gastrula II; 14, Hemi-circular blastopore, Gastrula III; 15, Elliptical blastopore, Gastrula IV; 16, Flat dorsal lip, Gastrula V; 17, Neural groove, Neurula I; 18, Neural folds, Neurula II; 19, Elevation of neural folds, Neurula III; 20, Neural rod, Neurula IV. 13a—16a, View from posterior, slightly ventral side; 13b-20b, View from left side; 17a—20a, View from dorsal side; 17c-20c, View from posterior side. Magnification 13-16, X19; 17-19, 25; 20, x23. Developmental Stages of Lamprey 115 Fic. 3. Stages 21-25. 21, Head protrusion I; 22, Neural tube, Head protrusion I; 23, Stomodaeum, Head protrusion III; 24, Nasal pit, Hatching; 25, Heart beat, Tailbud I. 21a—25a, View from left side; 21b, View from dorsal side; 21c and 22b-25b, View from ventral side. Magnification 21 and 22, X23; 23, X25; 24 and 25, x17. 116 Y. TAHARA 30b Fic. 4. Stages 26-30. 26, Melanophore, Tailbud II; 27, Eye spots, Tailbud III; 28, Velum movement; 29, Greenish bile; 30, completion of digestive tract, Earliest ammocoete larva. 26a-30a, View from left side; 26b-30b, View from ventral side. Magnification 26, X17; 27 and 28, X16; 29, x15; 30, «13.5. Developmental Stages of Lamprey 117 pairs of gill arches. Appearance of septum in endostyle. Irridescent pigment cells appear. Stage 29. Age 24 days. Greenish bile. Piavis 17 External: Greenish bile appears in gall blad- der. Oral hood expands. External naris arrives on dorsal side. Opening of all seven pairs of external gill pores. Irridescent pigment cells increase. Body length 8.0 mm. Internal: Appearance of _ cerebral spheres. Cilia appear in hypophysial sac. Gill filaments are formed on all eight pairs of gill arch. hemi- TABLE 1. in P. marinus Developmental phases Stage 30. Age 31 days. Completion of digestive tract; The earliest stage of ammocoete larvae. Piavis 18 External: Remnants of digested yolk are ex- truded from anus and formation of digestive tract completes. Oral cirri increase in number. Oesophagus becomes visible on left side due to torsion of liver and “stomach” (anterior intestine). Body length 9-9.5 mm. Internal: Appearance of cilia in dorsal ridge, hypopharyngeal groove, endostyle and oesopha- A comparison of subdivision of the stages in L. reissneri with that Subdivision of stages in L. reissneri P. marinus Stages Stages Ovulated, unfertilized egg 0 0 Fertilized, uncleaved egg 1-2 1 Cleavage 3-8 2-7 Blastulation 9-11 8 Gastrulation 12-16 8-9 Neurulation 17-20 10-11 Head protrusion 21-23 12-13 Hatching larva 24 13-14 Post-hatched larva 25-30 15-18 Total number of stages 31 19 TABLE 2. A comparison of the time after fertilization required to reach at the twelve stages in L. reissneri (at 15°C) with that in P. marinus (18.4°C) Stages L. reissneri P. marinus Two-cell (T3, P2)* 6.5 hr 2 hr Eight-cell (T5, P8) 15.5 hr 10 hr Morula (64-cell; T8, P7) 28 hr 19 hr Blastula (T10, P8) 48 hr 24 hr Gastrula (T13, P9) 78 hr 64 hr Neural plate (T17, P10) 5 days 4 days Head protrusion (T21, P12) 7'/, days 6 days Hatching (T24, P14) 11 days 10 days Melanophore (T26, P15) 16 days 13 days Eye spots (T27, P16) 18 days 15 days Gall bladder (T29, P17) 24 days 17+ days Completion of digestive tract 31 days 33 days (T30, P18) ee * T3 and P2 mean Tahara’s stage 3 and Piavis’s stage 2, respectively. 118 gus. Gill filaments further elongate. Formation of brush border on intestinal epithelium. Typhrosole protrudes into intestinal lumen. DISCUSSION Since the present work is aimed at serving the students of developmental biology of the lamprey embryos, the division of the early phases of embryogenesis are made much more detailed than the staging of Petromyzon marinus presented by Piavis [11, 12]. In Table 1 the staging in Lampetra reissneri is compared with that of P. marinus. Based on the present observations it may be concluded that the shape and size of the unferti- lized egg, the mode of cleavage, the aspects of structural and functional changes as well as the developmental tempos in L. reissneri bear a close resemblance to those in P. marinus. The time required to reach at the twelve stages in the both species of lamprey is compared in Table 2. It can be thought that the early development proceeds in an approximately equal tempo in the two species if one takes account of the difference in the tempera- ture at which the embryos are reared. ACKNOWLEDGMENT I wish to express my sincere gratitude to Professor Ch. Katagiri of Hokkaido University for his kind help in this study. I also thank Dr. T. Fujii of Ryukyu University for providing the lamprey embryos and also for collecting the adults. REFERENCES 1 Glaesner, L. (1910) Studien zur Entwicklungsge- schichte von Petromyzon fluviatilis. 1. Furchung Y. TAHARA 13 und Gastrulation. Zool. Jahrb. Abt. Anat., 29: 139-190. Veit, K. (1957) Einige Beobachtungen iiber die ersten Furchungsgeschritte bei Petromyzon planeri. Morphol. Jahrb., 98: 1-34. Hatta, S. (1891) On the formation of germinal layers in Petromyzon. J. Coll. Sci. Imp. Univ. Tokyo, 5: 129-147. Hatta, S. (1907) On the gastrulation in Petromyzon. J. Coll. Sci. Imp. Univ. Tokyo, 21: 1-44. Hatta,S. (1915) The fate of the peristomal mesoderm and the tail in Petromyzon. Annot. Zool. Japon., 9: 49-62. Selys-Longchamps, M. de (1910) Gastrulation et formation des feuillets chez Petromyzon planeri. Arch. Biol., 25: 1-75. Veit, O. (1939) Beitrage zur Kenntnis des Kopfes der Wirbeltiere. III. Beobachtungen zur Friihentwicklung des Kopfes von Petromyzon planeri. Morphol. Jahrb., 84: 86-107. Damas, H. (1944) Recherches sur la développment de Lampetra fluviatilis L. Contribution 4 l’étude de la céphalogénese des vertébrés. Arch. Biol., 55: 1- 284. Hatta, S. (1897) Contribution to the morphology of cyclostomata. I. On the formation of the heart in Petromyzon. J. Coll. Sci. Imp. Univ. Tokyo, 10: 225-237. Hatta, S. (1900) Contribution to the morphology of cyclostomata. II. The development of pronephros and segmental duct in Petromyzon. J. Coll. Sci. Imp. Univ. Tokyo, 13: 311-425. Piavis, G. W. (1961) Embryological stages in the sea lamprey and effect of temperature on develop- ment. Fishery Bull. Fish Wildl. Serv. U. S., 61: 111-143. Piavis, G. W. (1971) Embryology. In “The Biology of Lampreys”. Ed. by M. W. Hardisty and I. C. Potter, Academic Press, London, pp. 361-400. Hatta, S. (1901) Uber die Entwicklung des GefaB- systems des Neunauges, Lampetra mitsukurii Hat- ta. Zool. Jahrb. Abt. Anat., 44: 1-257. ZOOLOGICAL SCIENCE 5: 119-131 (1988) Analysis of Oral Replacement by Scanning Electron Microscopy and Immunofluorescence Microscopy in Tetrahymena thermophila during Conjugation Minoru TsuNEMOTO, Osamu Numata!, TosHIRO SUGA and YosHIO WATANABE!” Department of Biology, Joetsu University of Education, Joetsu, Niigata 943, ‘Institute of Biological Sciences, University of Tsukuba, Tsukuba, Ibaraki 305, and * Department of Biology, Ibaraki University, Mito, Ibaraki 310, Japan ABSTRACT—The process of oral replacement in Tetrahymena during conjugation has not so far been elucidated, mainly because the oral regions of a conjugating pair are hidden by the cell-cell junction in their proximate anterior which makes observation difficult. To analyze in detail the process of the oral replacement, a newly devised technique for deciliation and pair detachment was used in the present study in combination with scanning electron microscopy and immunofluorescence microscopy for tubulin. Oral replacement during conjugation was found to proceed synchronously, so that we could classify the process into 12 stages including 6 regressive and 6 reformation stages: Regression of the old oral apparatus occurred in the order of membranelles 1, 2, 3 (M1, M2, M3) and the undulating membrane (UM). After this, the paired cells approached a virtually astomatous state, and then, reformation of the oral apparatus at the pre-existing area started from the right ends of each membranelle. The regressive processes are considerably different from those of oral replacement observed in cells under amino acid deprivation, although the oral reorganization in amino acid-starved cells and that in conjugating cells are both called “oral replacement”. The discrepancy may be due to the difference in stage-specific cellular events between abortive division under amino acid deprivation © 1988 Zoological Society of Japan and conjugation. INTRODUCTION Studies on the serial morphological changes involved in oral neoformation in Tetrahymena have presented many important clues for under- standing the mechanisms of cell division [1-5] and synchrony induction [1-3, 6-8]. This is because oral neoformation can be considered to be a stage-specific indication of the cellular events underlying the metabolic changes, physiological changes and the changes of gene expression during the division cycle. During this cycle, the pre- existing anterior oral apparatus is also known to show stage-specific regressive and remodeling Accepted July 1, 1987 Received May 22, 1987 > To whom requests of reprints should be addressed. changes [3]. Other than these division cycle- dependent oral changes, much more extensive regression and reformation of the anterior oral apparatus has been known to occur under certain physiological conditions without being accompa- nied by cell division [3]. This phenomenon is referred to as “oral replacement”, and is observed, for example, in cells cultivated in an amino acid-free medium [6] and in cells during conjuga- tion [3]. In the former case, the process of oral replacement has been well studied by Frankel [6] and Frankel and Williams [3], using the synchro- nous rounding (abortive division) system de- veloped by us [9-11]. In the latter case, however, the process of oral replacement has not been made clear, because of the difficulty in studying the oral regions of the conjugating pairs which are hidden by the cell-cell junction. 120 M. TsunemotTo, O. Numata et al. We have previously shown that Tetrahymena intermediate filament protein (49K protein) [12- 14] plays a crucial role in the oral morphogenesis preceding binary fission in Tetrahymena [15]: Although the 49K protein is localized in the so-called posterior connectives of the functional oral apparatus, dissociation of the 49K protein from the oral apparatus occurs concurrently with the regression of the old oral apparatus which is prerequisite to the oncoming oral reorganization at the predividing stage, and reassociation of the 49K protein occurs a short time before the completion of the functional new and old oral apparatuses. The correspondence of oral structural change to the occurrence of the 49K protein urged us to investigate the localization of the protein in the oral apparatus of Tetrahymena cells during con- jugation. In a preliminary experiment, we observed that the 49K protein disappeared from the oral apparatus at an early phase of conjugation and reappeared in the apparatus after the separa- tion of conjugating pairs. Thus, Tetrahymena cells are assumed to undergo oral replacement (exten- sive regression and reformation of the pre-existing oral apparatus) during conjugation. In the present study, by using scanning electron microscopy (SEM) and immunofluorescence mi- croscopy for tubulin, we have attempted to observe the oral structures of the cells of conjugat- ing pairs detached artificially. We here describe the temporal sequence of oral replacement during conjugation, which differs from that of the oral replacement previously observed in cells cultivated in an amino acid-free medium (3, 6]. MATERIALS AND METHODS Tetrahymena thermophila mating types I and III of strain B were axenically cultivated at 26°C in a proteose-peptone medium containing 1% proteose peptone, 0.5% yeast extract and 0.87% dextrose [10]. Synchronous conjugation of Tetrahymena was induced by a modification of the method of Sugai and Hiwatashi [16] as follows. Cells of the T. thermophila mating types I and III were washed three times with an NKC solution containing 0.2% NaCl, 0.008% KCl, and 0.012% CaCl), and then resuspended in NKC at a concentration of approx- imately 410° cells/ml. Eight ml of the cell suspension was incubated at 26°C for 16-18 hr. Conjugation was elicited by mixing the cell cul- tures of the two mating types. After 1 hr at 26°C, mating cells were separated into individual cells again by a centrifugation at 100 g for 2 min at 0°C. The cells were resuspended in 4 ml of NKC and incubated at 26°C. The onset of pair formation is well synchronized in this treatment, so the time of the resuspension was considered to be the starting point (0 time) of conjugation. To obtain a popula- tion with a much higher conjugation rate, we gently diluted the mating mixture with NKC at 25 min and then carefully discarded the supernatant containing free swimming non-pairing cells, since the mating pairs tended to gather on the bottom of the culture dish. Thus, we were always able to obtain a high rate of synchronous conjugation with a pair rate of nearly 95%. For SEM, 3 ml-samples were withdrawn from the conjugation culture at desired intervals. First, deciliation of living cells was carried out as follows. Cells collected by a light centrifugation were suspended in 0.5-0.8 ml of 12.5 mM dibucain (pH 6.5) and stirred gently with a capillary for 1-2 min. The suspension containing deciliated and detached cells was quickly mixed with 1 ml of 2% glutaral- dehyde in 0.1M phosphate buffer (pH 7.0) for pre-fixation and immediately centrifuged. The cells in the pellet were then loosened and post- fixed with 2 ml of 2% glutaraldehyde in 0.1M phosphate buffer for 30 min. Next, the fixed cells were washed with the phosphate buffer and dehydrated in a series of acetone (50-70-90-99.5- 100%) and in isoamy] acetate, followed by drying at the critical point. Specimens were coated with gold before they were observed with a scanning electron microscope (JEOL T-100). By using this procedure, a considerable number of conjugating pairs could be separated throughout the conjuga- tion process. Therefore, we mainly observed the oral structures of such detached cells. A rabbit antiserum specific for Tetrahymena 49 K protein used in this paper was the same as that used in our previous papers [14, 15]. For im- munofluorescence staining materials, Nonidet P- 40-extracted conjugating pairs were prepared as Oral Replacement in Conjugating Tetrahymena 121 described by Goodenough [17] and air-dried on slides. These slides were treated with anti-49K protein antiserum (diluted 1:160 in phosphate- buffered saline, PBS) and then rhodamine- conjugated goat anti-rabbit IgG (diluted 1: 200 in PBS, Miles). The macro- and micronuclei were stained with 0.5 ug/ml of DAPI (4, 6-diamidino-2- phenylindole dihydrochloride) . Polyclonal anti-tubulin antiserum was produced in a male rabbit using highly purified tubulin prepared from ciliary axoneme by _ two- dimensional polyacrylamide gel electrophoresis [18]. Affinity purification of this antiserum was performed by the method of Talin ef al. [19]. The specificity of the antiserum was shown by the immunoblotting technique of Towbin et al. [20]. Deciliated conjugating pairs were air-dried on slides. The slides were fixed with 2.5% paraform- aldehyde in PBS for 5 min at room temperature, and subsequently with acetone for 30sec at —20°C. After washing in PBS, the slides were treated with 0.1M glycine in PBS for 30 min at room temperature. For staining, these slides were treated with anti-tubulin antiserum (diluted 1:100 in PBS) and then fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit IgG. The staining of the nuclei was done in the same way as with the anti-49K protein. RESULTS Relation among oral immunofluorescence for 49K protein, food uptake activity and formation of conjugating pairs To assess the occurrence of oral replacement during conjugation, we investigated the localiza- tion of 49K protein within cells during conjugation by using the indirect fluorescent antibody tech- nique. As shown in Figures 1-3, t-shaped oral fluorescence suddenly disappeared as the forma- tion of pairs proceeded at the early phase of conjugation. During the middle stage of conjuga- tion, immunofluorescence for 49K protein showed very interesting distribution relevant to the be- havior of germ nuclei as reported in our previous paper [21]. At the late phase of conjugation, r- shaped oral immunofluorescence reappeared after paired cells became naturally detached from each other (Figs. 4-6). Next we investigated the percentages of cells having t-shaped oral fluorescence and those of cells capable of uptaking carbon particles as a function of time after the start of conjugation. As shown in Figure 7, at the early phase of conjuga- tion, curves representing both the percentages of cells with oral fluorescence and of cells with food uptake activity decreased simultaneously and sharply in a mirror-image fashion as the percent- age of pair formation increased. In contrast to this, at the late phase of conjugation, curves representing both the ratios of cells with oral fluorescence and of cells with food uptake activity increased again nearly correspondingly with the decrease of paired cell ratio. The result strongly suggests that oral replacement occurs synchro- nously during conjugation. Consequently, we tried to analyze the oral structural changes in detail by SEM. Scanning electron microscopic observation of se- quential changes in oral structure during conjuga- tion We observed various kinds of SEM images of the oral apparatus in the course of synchronous conjugation. These images were classified into 12 stages according to both the temporal sequence of their occurrence and the characteristics of the oral structures. Stages 1-6 are regressive stages and stages 7-12 are reformation ones (The durations of each stage are shown in the abscissa of Fig. 7). The characteristics of each oral replacement stage are schematically shown using diagrams (Fig. 8). Each diagram was produced from several SEM images of the same stage. For reference, we also show typical SEM images as representatives of each stage (Figs. 9-20). Although the oral replacement process during conjugation is complex and in- volves many sequential and parallel events, the following descriptions can be made for each stage. Stage l: At the time of mixture of com- plementary mating types, the oral apparatuses of the cells are intact and functional (designated as stage 1) and are essentially the same as those of interphase cells during exponential growth: The oral apparatus consists of complete membranelles 122 M. TsunemotTo, O. Numata et al. Fics. 1-6. Immunofluorescence localization of 49K protein in the oral apparatus of cells at an early phase and late phase of conjugation. Fic. 1. A cell in costimulation stage just after mixing two cultures of complementary mating-type cells. Note that the t-shaped region of the oral apparatus fluoresces. Fic. 2. A pair-forming cell at very early stage of conjugation. Oral fluorescence is still evident. Fic. 3. A pair taken from a conjugating culture at 1.5 hr after mixing. Note that the rt -shaped fluorescence disappears from the conjugating pair. Fic. 4. A conjugating pair at the late stage of conjugation. Oral fluorescence has not yet reappeared. Fic. 5. Exconjugants immediately after detachment from conjugating pairs. Oral fluorescence is still obscure. Fic. 6. An exconjugant at the end of conjugation. Note the obvious zt -shaped fluorescence in the oral apparatus. a, immunofluorescence image of a selected cell or conjugating pair; b, its phase contrast image; c, its nuclear state as revealed by DAPI staining. b and c are cited in aid of judging the stage of each cell shown in a. All 1400. Oral Replacement in Conjugating Tetrahymena 123 fertili- nuclear meiosis , zation differentiation 100 & o a i = = 50 3 1S) ) O 5 10 15 20 25 30 stage hours after mixing Fic. 7. Oral immunofluorescence for 49K protein and food uptake activity in Tetrahymena thermophila during conjugation. Ordinate, percentages of cells or pairs having t -shaped oral immunofluorescence (jj), percentages of cells or pairs capable of uptaking carbon particles when the cells were incubated in 2% India ink for 5 min (a---a), or percentages of conjugating pairs (@——@); abscissa, time (hr) after mixing the cells of complementary mating types. On the upper part of the figure, durations of germ nuclear events are shown for reference. As described later, oral replacement was observed during conjugation. The oral morphological changes involved in oral replacement were classified into 12 stages. The duration of each stage (from stage 1 to stage 12) is shown with a bar and stage number in the lower abscissa for the sake of convenience. 1, 2 and 3 (M1, M2 and M3), the undulating membrane (UM), the oral rib, and the buccal cavity (Figs. 8 and 9). Stage 2: About 1.5 hr after the cultures were mixed, a sign of regression (resorption) in the oral apparatus began to appear. Several ciliary stubs of the first and second rows of M1 and of the first row of M2 (see legend for Fig. 8) disappeared from the animal’s left side (Figs. 8 and 10). Stage 3: About 4 hr after the cultures were mixed, disappearance of ciliary stubs from the left sides of M1 and M2 proceeded further than in stage 2. At this stage, the striation of oral rib became somewhat unclear (Figs. 8 and 11). Stage 4: About 6 hr after the cultures were mixed, the oral regions of the conjugating pairs became smaller in size. Half of the ciliary stubs of M3 disappeared from its left side. In addition, the disappearance of ciliary stubs of M1 and M2 proceeded still further. At this stage, the striation of the oral rib became unclear and the buccal cavity became smaller and shallower as compared to stages 1-3 (Figs. 8 and 12) . Stage 5: About 7-8 hr after the cultures were mixed, the buccal cavity became more shallower than in stage 4, so that the oral region itself became dish-like in shape. At this stage, only 12- 18 ciliary stubs remained in the three mem- branelles, and the ciliary stubs of the UM dis- appeared progressively (Figs. 8 and 13). Stage 6: About 8-12 hr after the cultures were mixed, the oral region became its smallest in size, being nearly completely flat, and its surface was smooth as if it was covered with a membrane. At this stage, the UM ciliary stubs were absent and only about 10-12 ciliary stubs presumably consist- ing of sculpturing portions remained in the approx- imately triangle-shaped oral region (Fig. 8). Fig- ure 14 shows the early phase of this stage. Stage 7: This oral stage was observed in con- 124 M. TsuNEmoto, O. Numata et al. stage 10 stage 11 stage 12 Fic. 8. Schematic figures from SEM images of 12 stages of the oral replacement in Tetrahymena thermophila during conjugation. Closed circles represent ciliary stubs and open circles represent basal body-like bulges. The upper and lower directions of all diagrams correspond to the animal’s anterior and posterior, respectively. The reader’s left corresponds to the animal’s right (in the text, the direction from the animal’s side is used). Membranelles 1, 2, 3 (M1, M2, M3), undulating membrane (UM), oral rib (OR) are shown in the diagram of stage 1. The ciliary stub rows within the membranelles are here numbered in an anterior-to-posterior order for the sake of convenience. Unciliated basal bodies were difficult to observe in our preparation, so the UM was drawn as a single row. jugating cells 12-17hr after the cultures were show signs of entering into the reforming process: mixed. Although no significant structure was The oral region began to sink inside again and the observed inside the oral region except for the circumference of the oral region (especially at the presence of a few cilia, the oral image seemed to __ right side) bulged (Figs. 8 and 15). Oral Replacement in Conjugating Tetrahymena 125 Fics. 9-20. SEM micrographs of oral replacement in Tetrahymena thermophila during conjugation. Each micrograph is presented as a representative of each stage of oral replacement illustrated in Fig. 8. Correspondence of micrographs to the stages of oral replacement is as follows. Fig. 9, stage 1; Fig. 10, stage 2; Fig. 11, stage 3; Fig. 12, stage 4; Fig. 13, stage 5; Fig. 14, stage 6; Fig. 15, stage 7; Fig. 16, stage 8; Fig. 17, stage 9; Fig. 18, stage 10; Fig. 19, stage 11; Fig. 20, stage 12. For the structural features of respective stages, see Results. Magnifications of all micrographs are the same, and the bar in Fig. 9 indicates 1 «am. 126 M. TsunNemoto, O. Numata et al. Stage $8: About 14-19hr after the cultures were mixed, reorganization of M1 and M2 began to occur from the right sides of each membranelle. In these membranelles, rearrangements of ciliary stubs proceeded in the order of the third, second and first rows. At this stage, the depth of the buccal cavity was nearly the same with the width of the oral apparatus. A few ciliary stubs were observed in the region corresponding to the right end of the incipient M3. In the bulged oral peripheral zone corresponding to the presumptive UM or its posterior area, we sometimes observed small bulges gathering, suggesting the presence of unarranged and unciliated basal body distribution as in usual stomatogenesis (Figs. 8 and 16). Stage9: About 17-22hr after the cultures were mixed, organization of M3 proceeded in addition to that of M1 and M2. Alignment of the ciliary stubs in the UM was seemingly completed at this stage (Figs. 8 and 17). Stage 10: About 22-24hr after the cultures were mixed, M3 and M2 were reorganized nearly completely, but the left side of the first row of M2 was sometimes incomplete. As for M1, ciliary stub rearrangement at the left side remained incom- plete, especially in the first and second rows of this membranelle. So-called sculpturing was not clear- ly recognizable yet. At this stage, the UM was completely formed and the oral size became larger, being comparable to the functional oral apparatus, but striation of the oral rib was still somewhat unclear (Figs. 8 and 18). Stage 11: This oral stage was observed in the detached exconjugants 24-28 hr after the cultures were mixed. Reorganization of M1, M2, and M3 and sculpturing of the right parts of each mem- branelle were completely finished, and the buccal cavity deepened more than in stage 10. Striation of the oral rib became clear partly between the surface and a flat stair present inside the buccal cavity (Figs. 8 and 19). Stage 12: The oral rib of this stage was fully striated and the oral pharynx was well formed. Thus, all replacement processes of the oral region expected in an exconjugant were completed (Figs. 8 and 20). The oral structure of this stage is, therefore, essentially the same as that of stage 1. However, the numbers of ciliary stubs of the left side of M1 tended to be reduced as compared with those of stage 1, presumably due to the influence of long-term starvation. Sequential changes of oral structures during conjugation as revealed by anti-tubulin immunofluorescence To back up the results obtained by SEM observation, we investigated the localization of basal bodies in the oral apparatus at various stages of synchronous conjugation, using indirect immunofluorescence microscopy for anti-tubulin antiserum. The specificity of the antiserum was analyzed by immunoblotting technique (Fig. 21). Two immunofluorescence bands at a molecular weight of about 55K were observed, proving that the antiserum recognized a- and /#-tubulin specif- ically among the proteins contained in Tetra- hymena cilia. Furthermore, the relationship be- tween sequential changes in oral structure and Fic. 21. Specificity of anti-tubulin antiserum. Tetra- hymena cilia (a) and total cells (b) were elec- trophoresed on a sodium dodecyl sulfate (SDS) polyacrylamide gel and blotted onto a nitrocellulose strip. The strip was subjected to immunoreaction with anti-tubulin serum followed by reaction with fluorescein-labeled 2nd antibody. Coomassie Blue stain (left) and the corresponding immunoblot (right) of SDS-polyacrylamide gels are presented in pairs. Oral Replacement in Conjugating Tetrahymena 127 nuclear events during conjugation was revealed by fixed cells double-stained with FITC-conjugated antibody and DAPI (Figs. 22-32). At stage 1, the organization of basal bodies was intact in the oral apparatus (Fig.22a). We observed food vacuole formation at this stage (Fig. 22d). Stages 2 and 3 correspond to meiotic prophase. During these stages, several basal bodies in M1 and M2 disappeared from the left to the right in the order of first, second and third rows (Figs. 23a and 24a) . At stage 4, the two main nuclear events, meiosis and fertilization, occurred. At this stage, the regression of basal bodies of M1 and M2 proceeded further and the basal bodies of UM disappeared progressively from the lower left to the upper right (Fig. 25a). At stage 5, the fertilization nuclei underwent two mitotic divi- sions. At this stage, SEM observation revealed that only 12-18 ciliary stubs remained in the three membranelles, and that ciliary stubs of UM dis- appeared progressively (Figs. 8 and 13). On the other hand, immunofluorescence microscopic observation showed that many basal bodies of the three membranelles and UM still remained (Fig. 26a). These results imply that the basal bodies of three membranelles and UM separated from the cell membrane and sank into cytoplasm. At stage 6, the mitotic products of the fertilization nucleus differentiated into new micro- and macronuclei (Fig. 27c, 27d). At this stage, the ciliary stubs of the UM and M3 were absent (Fig. 14) and only about 20-25 basal bodies remained in M1 and M2 (Fig. 27a). At stage 7, an oral primordium was first observed between the somatic ciliary rows (Fig. 28a). At stage 8, reorganization of M1 and M2 began to occur from the right sides of each membranelle and unarranged basal bodies were observed in the region corresponding to the incipient M3 (Fig. 29a). At stage 9, the both M1 and M2 coincidently extended to the left and a small M3 appeared. A double row of the basal bodies of the UM organized itself along the right margin of the three membranelles (Figs. 29a and 30a). At stage 10, the basal bodies of M1, M2 and M3 were arranged into a rectangular formation within each membranelle. At this stage, the UM was nearly completely formed and the oral size became larger (Figs. 18 and 31a). At the stages 11 and 12, reorganization of M1, M2, and M3 and the sculpturing of the right parts of each membranelle were completely finished (Fig. 2a). During the stages 7 and 8, each conjugant contained two new macronuclei and two new micronuclei. At stage 9, the exconjugants of a pair separated from each other. At the stages 10 through 12, the cells recommenced food uptake. DISCUSSION In Tetrahymena, the oral primordium is known to develop at an equatorial site of the cell in usual exponential growth and to become the anterior new oral apparatus of the posterior division product. The process has been extensively studied by using the silver impregnation method in either or both light microscopy [1, 3-5] and SEM [22- 25]. However, under certain conditions [6, 26-30], oral replacement is known to occur. One of the conditions eliciting oral replacement is the condi- tion for inducing conjugation [3]. In this regard, it is only known that the oral apparatus of a conjugating pair in Tetrahymena remains intact at the very early phase of conjugation, but the exconjugants are astomatous [3]. However, details of how the old oral apparatuses are resorbed and the new ones later appear have not so far been elucidated. In Paramecium tetraurelia, oral re- placement during conjugation seems to be sus- tained under the long-term control of micronuclear genes [31, 32]. More than 20 years ago, we found that Tetrahymena cells exhibited synchronous rounding and abortive synchronous division, by subjecting amino acid-starved cells to the ordinary heat treatment for synchronization [9-11]. By using this system, Frankel [6] demonstrated that the synchronous rounding cells performed oral replacement synchronously. According to the detailed observation on the oral replacement by Frankel and Williams [3] and Frankel [6], (i) resorption of the old oral apparatus occurs in the order of UM, M3, M2 and M1, and (ii) concurrent- ly with the resorption of the old oral apparatus, oral neoformation occurs in the order of M1, M2, M3 and UM. The latter neoformation sequence is shown to be much the same as the sequence 8 Anti-tublin Antibody M. TsuNEmoto, O. Numata et al. Oral Replacement in Conjugating Tetrahymena 129 Anti-tublin Antibody DAPI Food Vacuole Fics. 22-32. Relations among the alignment of basal bodies, the nuclear events and food vacuole formation during conjugation. a, immunofluorescence localization of tubulin in the oral region. Each basal body could be seen by using this method. b, immunofluorescence localization of tubulin in the whole cell. Cells in a and b are the same except for the ones in Fig. 22 and Fig. 32. c, DAPI-stained nuclei of the same cell shown in b. d, food uptake activity and nuclear events of cells and pairs corresponding to the stages of a. At each point, an aliquot of culture was mixed with an equal volume of 1% India ink and incubated for 5 min at 26°C. The cells were fixed with an equal volume of 5% formaldehyde in PBS and stained with 10 “g/ml DAPI. Since cells seen in Fig. 22d, Fig. 31d and Fig. 32d had taken India ink into the food vacuoles, the food vacuoles were observed as black particles in the cells. In the left sides of photos a, b, c, d, stage numbers classified are given. 130 M. TsuNEmoto, O. Numata eft al. observed in oral neoformation during division, except that only the forming site is different. The above-mentioned sequences are different in some points from the sequences of oral replace- ment during conjugation demonstrated in the present paper (Figs. 8-32) . The major differences found in the process of conjugation are: (i) Resorption of the old oral apparatus occurs in the order of M1, M2, M3 and UM; (ii) Resorption and neoformation of the oral apparatus occur not concurrently but rather independently in the order of resorption and neoformation; (iii) There exists a nearly astomatous state, although a very small oral area is recognized without significant surface struc- tures. On the other hand, neoformation occurs in nearly the same order as that observed in synchro- nous rounding. These differences are presumably due to a difference in cellular events between conjugation and abortive division. Thus, it is noteworthy that, although oral regression and reformation observed in conjugation and in abor- tive division are both called oral replacement, the detailed processes differ in some points from each other. To our knowledge, this paper is the first report describing in detail the oral morphological changes during conjugation in Tetrahymena. Concerning the biological function of Tetra- hymena intermediate filament protein, 49K pro- tein, we previously emphasized its role in the oral morphogenesis in binary fission cells [15]. The same is true for the oral morphological change during conjugation (Figs. 1-7). The intermediate filament protein is dissociated from the oral apparatus at an early phase of conjugation, which may trigger the regression of the old oral appa- ratus. Then, the protein plays important roles in various germ nuclear events, such as meiosis, selection of functional meiotic products, transfer of pronucleus, and zygote formation [32]. After this, the protein reassociates with the immature oral apparatus, and may cause the oral apparatus to become a functional one (Figs. 4-7). It follows from these findings that the 49K protein is indis- pensable for the various processes during conjuga- tion and is multifunctional. 1 10 13 14 REFERENCES Frankel, J. (1962) The effects of heat, cold, and p-fluorophenylalanine on morphogenesis in synchro- nized Tetrahymena pyriformis GL. C. R. Trav. Lab. Carlsberg, 33: 1-52. Frankel, J. (1965) The effect of nucleic acid an- tagonists on cell division and oral organelle develop- ment in Tetrahymena pyriformis. J. Exp. Zool., 159: 113-148. Frankel, J. and Williams, N.E. (1973) Cortical development in Tetrahymena. In “Biology of Tet- rahymena”. Ed. by A. M. Elliott, Dowden, Hutch- inson, and Ross, Stroudsburg, Pennsylvania, pp. 375-409. Holz, G. G., Jr., Scherbaum, O. H. and Williams, N. E. (1957) The arrest of mitosis and stomatogene- sis during temperature-induction of synchronous division in Tetrahymena pyriformis, mating type 1, variety 1. Exp. Cell Res., 13: 618-621. Williams, N. E. and Frankel, J. (1973) Regulation of microtubules in Tetrahymena. I. Electron micro- scopy of oral replacement. J. Cell Biol., 56: 441- 457. Frankel, J. (1970) The synchronization of oral development without cell division in Tetrahymena pyriformis GL-C. J. Exp. Zool., 173: 79-100. Frankel, J. (1964) Cortical morphogenesis and syn- chronization in Tetrahymena pyriformis GL. Exp. Cell Res., 35: 349-360. Frankel, J. (1964) The effects of high temperatures on the pattern of oral development in Tetrahymena pyriformis GL. J. Exp. Zool., 155: 403-436. Tamura,S., Toyoshima, Y. and Watanabe, Y. (1966) Mechanism of temperature-induced syn- chrony in Tetrahymena pyriformis. Analysis of the leading cause of synchronization. Jpn. J. Med. Sci. Biol., 19: 85-96. Watanabe, Y. (1963) Some factors necessary to produce division conditions in Tetrahymena pyri- formis. Jpn. J. Med. Sci. Biol., 16: 107-124. Watanabe, Y. (1971) Mechanism of synchrony in- duction. I. Some features of synchronous rounding in Tetrahymena pyriformis. Exp. Cell Res., 68: 431- 436. Numata,O., Yasuda,T., Hirabayashi,T. and Watanabe, Y. (1980) A new fiber-forming protein from Tetrahymena pyriformis. Exp. Cell Res., 129: 223-230. Numata, O., Yasuda, T., Ohnishi, K. and Wata- nabe, Y. (1980) Jn vitro filament formation of a new fiber-forming protein from Tetrahymena pyriformis. J. Biochem., 88: 1505-1514. Numata, O. and Watanabe, Y. (1982) In vitro assembly and disassembly of 14-nm filament from Tetrahymena puriformis. The protein component of 15 16 17 18 19 20 21 22 23 Oral Replacement in Conjugating Tetrahymena 14-nm filament is 49,000-dalton Biochem., 91: 1563-1573. Numata, O., Hirono, M. and Watanabe, Y. (1983) Involvement of Tetrahymena intermediate filament protein, a 49K protein, in the oral morphogenesis. Exp. Cell Res., 148: 207-220. Sugai, T. and Hiwatashi, K. (1974) Cytologic and autoradiographic studies of the micronucleus at meiotic prophase in Tetrahymena pyriformis. J. Protozool., 21: 542-548. Goodenough, U. W. (1983) Motile detergent- extracted cells of Tetrahymena and Chlamydomo- nas. J. Cell Biol., 96: 1610-1621. Hirabayashi, T., Tamura, R., Mitsui, I. and Wata- nabe, Y. (1983) Investigation of actin in Tetrahyme- na cells. A comparison with skeletal muscle actin by a devised two-dimentional gel electrophoresis method. J. Biochem., 93: 461-468. Talin, J.C., Olmsted, J.B. and Goldman, R. D. (1983) A rapid procedure for preparing fluorescein- labeled specific antibodies from whole antiserum: Its use in analyzing cytoskeletal architecture. J. Cell Biol., 97: 1277-1282. Towbin, H., Staehelin, T. and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacryl- amide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci. U.S. A., 76: 4350-4354. Numata, O., Sugai, T. and Watanabe, Y. (1985) Control of germ cell nuclear behaviour at fertiliza- tion by Tetrahymena intermediate filament protein. Nature, 314: 192-194. Buhse, H. E., Jr., Stamler, S.J. and Corliss, J. O. (1973) Analysis of stomatogenesis by scanning electron microscopy in Tetrahymena pyriformis strain W during synchronous cell division. Trans. Am. Microsc. Soc., 92: 95-105. Williams, N. E. and Bakowska, J. (1982) Scanning electron microscopy of cytoskeletal elements in the protein. J. 24 25 26 Zi 28 29 30 31 32 131 oral apparatus of Tetrahymena. J. Protozool., 29: 382-389. Bakowska, J., Nelsen, E. M. and Frankel, J. (1982) Development of the ciliary pattern of the oral apparatus of Tetrahymena thermophila. J. Pro- tozool., 29: 366-382. Jerka-Dziadosz,M. and Frankel,J. (1979) A mutant of Tetrahymena thermophila with a partial mirror image duplication of cell surface pattern. I. Analysis of the phenotype. J. Embryol. Exp. Mor- phol., 49: 167-202. Albach, R. A. and Corliss, J.O. (1959) Regenera- tion in Tetrahymena pyriformis. Trans. Am. Mi- crosc. Soc., 78: 276-284. Buhse, H. E., Jr. (1966) Oral morphogenesis during transformation from microstome to macrostome and macrostome to microstome in Tetrahymena vorax strain V> type S. Trans. Am. Microsc. Soc., 85: 305—- 313. Buhse, H. E., Jr. (1967) Microstome-macrostome transformation in Tetrahymena vorax strain V2 type S induced by a transforming principle, stomatin. J. Protozool., 14: 608-613. Frankel, J. (1964) Morphogenesis and division in chains of Tetrahymena pyriformis GL. J. Protozool., 11: 514-526. Roque, M., de Puytorac, P. and Savoie, A. (1970) Charactéristiques morphologiques et biologiques de Tetrahymena bergeri sp. nov., cilié hyménostome tetrahyménien. Protistologica, 6: 343-351. Ng, S. F. and Newman, A. (1984) The role of the micronucleus in stomatogenesis in sexual reproduc- tion of Paramecium tetraurelia: conjugation of ami- cronucleates. Protistologica, 20: 517-523. Ng, S. F. and Newman, A. (1985) The macronu- clear anlage does not play an essential role in stomatogenesis in conjugation in Paramecium tetraurelia. Protistologica, 21: 391-398. £3" 55 x \ a vs ’ eS ah 4 — A i oe a ae Q { | Moa ‘i a — a ——- ~ : 7 iW * Dr Ra = ——e ZOOLOGICAL SCIENCE 5: 133-137 (1988) © 1988 Zoological Society of Japan Comparative Effects in vitro of Myxine, Squalus, Avian and Mammalian Insulins on DNA-Synthesis in 3T3 Mouse Fibroblasts WILHELM ENGSTROM, EVA DAFGARD and STURE FALKMER Department of Tumour Pathology, Karolinska Institutet, Karolinska Hospital, S-104 01 Stockholm, Sweden ABSTRACT—Five kinds of natural insulins, namely those obtained from the Atlantic hagfish, the spiny dogfish, turkey, pig, and man, failed to stimulate quiescent serum-starved 3T3 fibroblasts to re-enter the cell cycle. However, after addition of epidermal growth factor (EGF) four of the insulins enhanced its mitogenic effect. Turkey insulin was found to be the most potent, followed in this respect by the porcine and human insulins. Dogfish insulin only exerted a limited, but nevertheless significant, effect on the EGF-treated quiescent cells. In contrast, no permissive effect was observed by the hagfish insulin. In trying to explain these differences in the growth promoting effects of the five insulins as due to their amino-acid sequences in the receptor-binding regions of their molecules, it was concluded that any sequential arrangement in the B22-B26 positions seems not to be a prerequisite to exert a permissive effect on EGF-induced DNA-replication in mammalian fibroblasts. INTRODUCTION The evolution of hormones and growth factors has resulted in so-called polypeptide hormone families where the amino acid sequences are so similar that it is reasonable to assume that a joint ancestor molecule exists for all members [1]. The insulin family consists of the hormone relaxin and the growth factors NGF (nerve growth factor) and the IGFs (insulin like growth factors) [2, 3]. A comparison of their amino-acid sequences and their X-ray crystallographic appearance suggests that the ability to attain an insulin-like tertiary molecular structure—the so called insulin fold —has been retained [3]. The insulin and IGF molecules have also retained a few surface residues in common which mediate their interactions with insulin and type 1 receptors. In contrast, their ability to form dimers, hexamers, bind antibodies and interact with type II receptors has not been universally retained in evolution [4]. From an evolutionary point of view it has been proposed Accepted July 23, 1987 Received July 10, 1987 that insulin and its relatives stem from a common precursor molecule which was encoded for by a gene that underwent duplication about 600 million years ago [1]. The subsequent gene duplication which led to diversion into insulin, relaxin, NGF, IGF I and IGF II may have taken place about 300 million years ago, when mammals appeared on earth [5]. In the present study we have compared the effects of insulins representing different stages in evolution on DNA-synthesis in a target cell line of fibroblastic origin (viz Swiss mouse 3T3 cells; cf.[6]) . When these cells are starved to quiescence in low serum, they cannot be stimulated to resume proliferation by addition of human or porcine insulin only [7]. However, when 3T3 cells are concomitantly exposed to another mitogenic fac- tor, viz EGF (epidermal growth factor), insulin exerts a substantial permissive effect on DNA- synthesis [8]. We report in this study that insulins from different species differ in their capacity to enhance the mitogenic effect of EGF. Differences in their amino acid sequences in the receptor- binding region cannot explain these differences in biological activity between insulins from different phyla and species. 134 W. Encstr6m, E. DaFGARD AND S. FALKMER MATERIAL AND METHODS Growth factors and hormones EGF was purchased from Collaborative Re- search Inc.(Waltham, MA., USA). Porcine and human insulins were purchased from Sigma Co. (Stockholm, Sweden). Turkey insulin was a kind gift from Dr. Steve Wood, Department of Crystal- lography, Birkbeck College, Malet Street, Lon- don, U. K.: Dogfish and hagfish insulins were prepared in our laboratory (cf. [9, 10]). Cell culture Swiss mouse 3T3 cells, obtained from Flow laboratories, Stockholm, Sweden, were main- tained in monolayer cultures in NUNC 10 ml tissue culture bottles. The stock cultures were grown in humidified 5% CO,/95% air mixture in Dulbecco’s modified Eagle’s medium (DMEM), | sup- plemented with 10% (v/v) foetal calf serum, 50 units penicillin/ml and 50 yg streptomycin/ml and never allowed to reach confluency. For transfer, the cells were treated with 0.25% trypsin (w/v) in Tris-buffered saline, containing 0.5 mM EDTA for 2-3 min at 37°C. The stock cultures were passaged by seeding 3,000 cells/cm? culture bottle area and transferring them every third day. For experi- mental purpose, 3,000-4,000 cells/cm? were seeded in Petri dishes containing a glass coverslip in the bottom in DMEM and 10% (v/v) serum. Autoradiography DNA-synthesis was estimated by incorporating 0.5 “Ci 7H-thymidine (Amersham; 56 mCi/mmol) per ml medium into TCA-unprecipitable material. At the end of each experiment, the cultures were washed twice in 0.9% NaCl (w/v) solution, fixed in 95% ethanol for at least 24 hr and maintained in an air dried state until autoradiography was per- formed. Before the film (Kodak AR10 stripping film) was applied, the cells were treated in 5% (w/v) TCA at 4°C for 5 min and then washed for 20 min in running tap water to remove non- incorporated thymidine. After 7 days’ exposure, the autoradiograms were developed with Kodak D19 developer (4 min at 18°C) , briefly rinsed in water, fixed in Kodak acid X-ray fixative with hardener (5 min at 18°C) , washed in running cold tap water for 20min and routinely stained in haematoxylin and eosin. The percentage of cells that had initiated DNA-synthesis during the ex- perimental period was determined by counting at least 500 cells in the light microscope. RESULTS AND DISCUSSION Figure 1 summarizes the effects of insulin on DNA-synthesis in quiescent serum starved 3T3- fibroblasts grown in sparse cultures. It was found that none of the five insulins obtained from the Atlantic hagfish (Myxine glutinosa), the spiny dogfish (Squalus acanthias), turkey, pig and man exerted any significant effects on DNA-synthesis in 50 = ug insulin/ml Fic. 1. Effect of five kinds of natural insulins on DNA- synthesis in quiescent serum starved mouse 3T3 fibroblasts in vitro. The cells were starved to quiescence for 48 hr in 0.1% serum. DMEM sup- plemented with various concentrations of Myxine (@), Squalus (*), turkey (™), porcine (a), or human (@) insulin, was added. Cells, continuously ex- posed to 0.1% serum, were used as an internal control. The cells were maintained in the ex- perimental media for 24 hr in the presence of 0.5 pCi 3H-thymidine per ml medium. The proportion of cells that had initiated DNA-synthesis was deter- mined by autoradiography. The results clearly demonstrate that neither of the five insulins stimu- lated DNA-synthesis in quiescent 3T3 cells when added as the sole macromolecular supplement. Growth Factor Potency of Various Insulins 135 50 25 % cells yg insulin/ml Fic. 2. The permissive effect of five kinds of natural insulins (as described in legend to Fig. 1) on EGF- stimulated quiescent 3T3 fibroblasts. Experiments were performed as shown in Fig. 1 but now the insulins were added together with 10 ng EGF/ml medium. The results show that turkey, porcine, human and Squalus insulins exerted a permissive effect on EGF-treated quiescent 3T3 fibroblasts. In contrast Myxine insulin failed to augment the stim- ulatory effects of EGF in this situation. this target cell type. This is in marked contrast to human IGF I or IGF II that both increase the proportion of *H-thymidine-labelled cells up to three times under similar experimental conditions (Dafgard et al., unpublished data) . Figure 2 illustrates the permissive effects of insulins of the same five species as in Figure 1 on DNA-synthesis in serum-starved cells concomitantly exposed to TABLE 1. another growth factor, viz EGF. The results show that the maximum mitogenic effect was achieved by adding turkey insulin together with EGF. A lower, but nevertheless substantial, permissive effect was achieved by porcine or human insulin. When Squalus insulin was added, the percentage of labelled cells was approximately half of that achieved by EGF and turkey insulin. In contrast, Myxine insulin only exerted a minor and statistical- ly insignificant effects in combination with EGF. The inability of insulin alone to induce DNA- replication in mouse 3T3 fibroblasts has previously been reported [8, 11]. However, mammalian insulin, added together with EGF, is known to enhance the mitogenic effect on these cells many- fold [7, 8, 11]. Furthermore, it has been demon- strated that Swiss mouse 3T3 cells, unlike for instance human glia cells, require a multiple set of growth factors and hormones for proliferation under serum-free conditions [12]; in a pilot study we observed that some kinds of natural insulin seemed to exert a permissive rather than a primarily mitogenic effect [13]. A similar permis- sive effect of Squalus insulin on platelet derived growth factor-stimulated 3T3 fibroblasts was re- ported by Bajaj [14]. This lack of effect of insulin on mitogenesis in 3T3 fibroblasts is somewhat puzzling since all 3T3-clones hitherto examined express functional insulin receptors [15]. We have found that insulins from four different species that exhibit an identical amino acid sequence in the B22-B26 receptor binding region, differ complete- ly in their permissive effect on DNA-synthesis in EGF-treated quiescent 3T3-cells (Table 1). Even Comparison of the amino acid sequences of insulin and insulin like growth factors in positions B22-B26 which comprises the insulin receptor binding region B-chain 22 23 24 25 26 Human Arg Gly Phe Phe Tyr Porcine Turkey Squalus Lys Tyr Myxine hIGF I Tyr Phe hIGF II Tyr Phe Data from [3]. 136 though differences in binding characteristics have been reported [16], it is conceivable that binding of insulin to the insulin receptor is not conditional for progression towards S-phase [8]. Nor can the substitution of B25-Phe with B25-Tyr, as in Squalus insulin, fully explain these differences in biological activity, since it evoked a lesser effect than for instance IGF I and IGF IT on EGF-treated 3T3-cells (Dafgard et al., unpublished data). It is tempting to propose that the permissive effect of insulin on EGF-treated 3T3-fibroblasts involves an initial binding of insulin to some other membrane receptor. One possible candidate is the type 1 receptor [17] which in many respects resembles the insulin receptor [18]. Consideration of the affinity of insulin to the type 1 receptor is, however, more difficult in the absence of extensive binding data. Notwithstanding, it has been demonstrated that Squalus, turkey and porcine insulins displace radiolabelled IGF I from target cells in a dose- dependent manner (Dafgard and Rees, unpub- lished data). It is, therefore, conceivable that the observed effects on mitogenesis are more depen- dent on binding to the type 1 receptor than to the insulin receptor. The type 2 receptor is probably not involved in the mitogenic signalling since insulin cannot bind—even with low affinity—to this receptor [19]. The exact relationship between EGF and insulin action is not known, but a similar synergistic effect between these two growth factors has been observed in human embryonic corneal stromal cells [20], indicating that this phenomenon is not unique to the 3T3 fibroblastic cell line. It remains, however, to be shown how other biolo- gical effects on the cell cycle, as effects on cellular enlargement, are mastered by insulin from diffe- rent phyla [21]. These experiments are currently in progress. ACKNOWLEDGMENTS This study was generously supported by the Swedish Medical Research Council (Project No. 12X-718), the Swedish Diabetes Association, the Stockholm Cancer Society and the Robert Lundberg memorial fund. E. D. was the recipient of a British Council Scholarship at Department of Crystallography, Birkbeck College, London. 10 11 12 13 W. Encstr6OM, E. DAFGARD AND S. FALKMER REFERENCES Blundell, T. L. and Humbel, R. (1980) Pancreatic hormone families. Nature, 287: 781-787. Rindeknecht, E. and Humbel, R. (1976) Polypeptides with non-suppressable insulin like and cell growth promoting activities in human serum. Isolation, chemical characterization and some biolo- gical properties of forms I and II. Proc. Natl. Acad. Sci. USA, 73: 2365-2369. Dafgard, E., Bajaj, M., Honegger, A. M., Pitts, J., Wood, S. and Blundell, T. L. (1985) The conforma- tion of insulin like growth factors. J. Cell. Sci., Suppl. 3: 53-64. Rechler, M. M. and Nissley, S. P. (1985) The na- ture and regulation of the receptors for insulin like growth factors. Ann. Rev. Physiol., 47: 425-442. Froesch, E.R. and Zapf,J. (1985) Insulin like growth factors and insulin. Comparative aspects. Diabetologia, 28: 485-493. Larsson, O., Dafgard, E., Engstr6m, W. and Zet- terberg, A. (1986) Immediate effects of serum depletion on dissociation between growth in cell size and cell division in proliferating cells. J. Cell. Physiol., 127: 267-273. Zetterberg, A., Engstrém,W. and Larsson, O. (1982) Growth activation of resting cells. Ann. N. Y. Acad. Sci., 397: 130-147. Zetterberg, A., Engstrom,W. and Dafgard, E. (1984) The relative effects of different types of growth factors on DNA-replication, mitosis and cellular enlargement. Cytometry, 5: 368-375. Bajaj, M., Blundell, T. L., Pitts, J. E., Wood, S. P., Tatnell, M. A., Falkmer, S., Emdin, S. O., Gowan, L. K., Crow, H., Schwabe, C., Wollmer, A. and Strassburger, W. (1983) Dogfish insulin. Primary structure, conformation and biological properties of an elasmobranchial insulin. Eur. J. Biochem., 135: 535-542. Emdin, S. O., Steiner, D. F., Chan, S. J. and Falk- mer, S. (1985) Hagfish insulin; Evolution of insulin. In “Evolutionary Biology of Primitive Fishes”. Ed. by R.E. Foreman, A. Gorbman, J.M. Dodd and R. Olsson, Plenum Press, New York, pp. 363-370. Jimenez de Asua, L., O’Farrel, M., Clingan, D. and Rudland, P. S. (1976) Temporal sequences of hor- monal interactions during the prereplicative phase of quiescent cultured 3T3 fibroblasts. Proc. Natl. Acad. Sci. USA, 74: 3845-3849. Shipley, G. D. and Ham, R. G. (1983) Multiplica- tion of Swiss 3T3 cells in a serum free medium. Exp. Cell Res., 146: 249-260. Falkmer, S., Dafgard, E. and Engstrém, W. (1986) Phylogeny of insulin - primitive insulins and the cell cycle. Chemica Scripta, 26B: 209-212. 14 15 16 17 Growth Factor Potency of Various Insulins Bajaj, M. (1984) Ph. D.-thesis, Birkbeck College, London. Murphy, R. F., Powers, S., Verderame, M., Can- tor,C.R. and _ Pollack,R. (1982) Flow cytofluorometric analysis of insulin binding and internalization by Swiss 3T3-cells. Cytometry, 2: 402-406. Emdin, S.O., Sonne, O. and Gliemann, J. (1980) Hagfish insulin; The discrepancy between binding affinity and biological activity. Diabetes, 29: 301- 303. Massague, J. and Czech, M. P. (1982) The subunit structures of two distinct receptors for insulin like growth factors I and II and their relationship with the insulin receptor. J. Biol. Chem., 257: 5048- 5055. 18 19 20 21 137 Gammeltoft, S. (1984) Insulin receptors. Physiol. Rev., 64: 1321-1377. Czech, M. P., Massague, J., Yu, K., Oppenheimer, C. L. and Mottola, C. (1984) Subunit structures and actions of the receptors for insulin and the insulin like growth factors. In “The Importance of Islets of Langerhans for Modern Endocrinology”. Ed. by K. Ederlin and I. Scholholt, Raven Press, New York, pp. 41-53. Hyldahl, L. (1986) Studies on the human embryonic cornea. Ph. D.-thesis, Karolinska Institutet, Stock- holm. Falkmer, S., Dafgard, E., El-Salhy, M., Engstrom, W., Grimelius, L. and Zetterberg, A.(1985) Phy- logenetical aspects on islet hormone families. Pep- tides, 6: 315-320. ¥ ye 1 ‘ n ; ‘ abe f i : s ¢ ‘ 1 ; in \ we 3 fii ” hip ' “ aia J noma “eS ba . ney - - % he 1? ~ * - \ : as ao 5 i | 1 - > ‘ a i i m1 ty i 7 — _ ™ “ tl 1 * ' ca i i 4 ay hy ri eri ZOOLOGICAL SCIENCE 5: 139-143 (1988) Effects of Long-term Progesterone Treatment on Synchronized Ovulation in Guinea Pigs Hipeco UEpA, TADASHI Kosaka and Kazuaki W. TAKAHASHI! Toxicology Division, Institute of Environmental Toxicology, Suzuki-cho 2-772, Kodaira-shi, Tokyo 187, and ‘Department of Laboratory Animal Science, Nippon Veterinary and Zootechnical College, Kyonan-cho 1-7-1, Musashino-shi, Tokyo 180, Japan ABSTRACT—Long-term effects of progesterone implants at different stages of the estrous cycle on synchronized ovulation were studied in guinea pigs. Females were subcutaneously implanted with Silastic tubing which contained crystalline progesterone for 7, 14, and 21 days (groups A, B and C) on day 0, 5, 10, and 15 of vaginal opening. The mean number of days to vaginal opening and leucocytic influx into vaginal smears in groups B and C were significantly less than those of group A (P<0.05 and 0.01, respectively). The length of leucocytic influx in the groups treated with progesterone took mostly 6 days, the mode day, following removal of the progesterone; number of animals showing results on this mode day ranged from 38% (9/24) in group A to 79% (19/24) in groups B and C. No significant differences in the mean number of days to vaginal opening or leucocytic influx were seen among subgroups given progesterone implants at different stages of the estrous cycle. Ovulation was observed in all animals sacrificed on the day of leucocytic influx into the vaginal smear. These findings indicate that long-term implantation of progesterone tubing, greater than 14 days, given at any estrous stage to female guinea pigs induces the synchronized ovulation within 5-6 days after the removal of © 1988 Zoological Society of Japan progesterone tubing. INTRODUCTION Reproductive efficiency of guinea pigs for ex- perimental animals is less than that of another kind of rodents; length of estrous cycles and gestation period in guinea pigs are 3 to 4 times longer and litter size 1/2 to 1/4 times less than those of rats, mice or hamsters. It is difficult to obtain the guinea pigs at a uniformed age or quality, which is needed for performing the toxicological study of chemicals or larger sized study at a scheduled time. So, synchronizing the estrous cycles of guinea pigs is one direction for researching how to control a scheduled supply with good uniformity. Progester- one treatment for synchronization of the estrous cycle has been widely used. In rats [1, 2], hamsters [3-5], and guinea pigs [6], a single injection of progesterone has caused either postponed or advanced ovulation depending on what stage of Accepted August 19, 1987 Received April 24, 1987 the estrous cycle it was administered. The effect of long-term progesterone treatment on the estrous cycle has been studied in sheep [7], cattle [8] and swine [9, 10]. Woody et al. [11] reported that in guinea pigs, multiple injections of progesterone dissolved in oil for 6 days reduced the average length of estrous cycles. Recently, continuous progesterone treatment by subcutaneous implants instead of multiple injections of progesterone dissolved in oil has been used because of its ability to maintain a constant steroid supply to target organ [12]. We investigated, therefore, how many days were needed for receiving progesterone implants and/or what stage of the estrous cycle these implants were given at to be effective in inducing synchronized ovulation following influx of leucocytes into vaginal smears in guinea pigs. MATERIALS AND METHODS Adult female guinea pigs of the Hartley strain at 3-4 months of age (weight 575-925 g) were used. 140 They were provided with commercial pellets (GB- 1: Funabashi Farm Co., Ltd.) and tap water ad libitum. Room temperature was maintained at 20- 24°C, relative humidity at 45-65 %, photoperiod at 14-hr light and 10-hr darkness (light on at 5:00 a.m. and off at 7:00 p.m.) and ventilation at 12 times an hour. Vaginal closure membranes and vaginal smears were examined once a day until removal of the progesterone implant, and twice a day (morning and evening) thereafter. Vaginal opening was determined to be positive when the vaginal membrane was fully ruptured. The first day of vaginal opening was designated as day 0 of vaginal opening. The influx of leucocytes into the vaginal smear after the appearance of cornified cells in the sample was observed under a light H. Uepba, T. KoSAKA AND K. W. TAKAHASHI sacrificed in order to examine the state of ovula- tion. The number of ova ovulated was confirmed by direct observation of ova in the oviduct and uterus and by the appearance of the corpora lutea; color, size, and elevation from the ovarian surface were noted. The average duration of estrous cycle recorded before the beginning of this experiment was 17.5 days (range: 14-21 days). Females received a subcutaneous implant (Silas- tic tubing, 1.0cm long, 0.4cm i.d.) of pure crystalline progesterone (Sigma Co. Ltd., Lot No. 73F-0198) for 7 days (group A), 14 days (group B), or 21 days (group C). The control group (group D) received empty Silastic tubing. Each group was divided further into four subgroups according to the stage at which the progesterone microscope. At this time, the animals were implant was to be received. Subcutaneous im- TABLE 1. Effects of progesterone on vaginal opening and leucocytic influx into the vaginal smear Treatment Days to vaeutal Days to leucocytic No. of opening after influx into the vaginal Group Implantation ey of estrous cycle animals removal of smear after removal period when Silastic tube examined Silastic tube of Silastic tube (days) was inserted (Mean+S.D.) (Mean+S.D.) 0 6 5.8+1.6 8.8+1.5 5 6 5.5+1.4 7.742.0 A q 10 6 4.0+0 6.3+40.8 15 6 4.54+2.1 7.542.6 Average 5.0+1.6 7.6+1.9 0 6 3.8+1.5 5.8+0.4 5 6 3.54+1.2 6.0+0 B 14 10 6 4.2+0.8 5.8+0.4 15 6 4.3+0.5 5.5+0.5 Average 4.0+1.0*, # 5.8+0.4**, + 0 6 4.0+0 5.7+0.5 5 6 3.8+0.4 6.0+0 C 21 10 6 3.7+1.0 6.0+0.6 15 6 4.3+0.8 5.8+0.4 Average 4.0+0.7*, + 5.9+0.4**, # 0 3 14.0+1.7 16.3+2.1 5 3 OTIS 11.34+2.1 D 21 10 3 3.342.9 4.7+3.2 (Control) 15 3 4.3+6.7 6.3+6.7 Average 7.8+5.5 9.7+5.9 *, **: Significantly different from group A at the 5% and 1% levels of probability, respectively. #: Significantly different from group D at the 5% level of probability. Synchronized Ovulation in Guinea Pigs 141 plants were given as follows: subgroup a, on day 0; subgroup b, on day 5; subgroup c, on day 10; subgroup d, on day 15 of vaginal opening. Remov- al of Silastic tubing was carried out at 16:00 hr- 18: 00 hr under ether anesthesia. Results were analyzed statistically using the Mann-Whitney U test and Kruskal-Wallis H test. RESULTS Vaginal opening was not observed in any ani- mals treated with progesterone implants during the implantation period. After removal of the Silastic progesterone tubing, the mean days to vaginal opening in the groups treated for 14 and 21 days was significantly less than that in the control group and the group exposed for 7 days (P<0.05, Table 1). The mean days to leucocytic influx into vaginal smears following removal of progesterone tubing was 7.6+1.9 days in group A, 5.8+0.4 days in group B, 5.9+0.4 days in group C, and 9.7+5.9 days in group D (Table 1). The mean length of time to leucocytic influx following vaginal opening in the groups treated for 14 and 21 days (groups B and C) were also significantly reduced compared with the control group (group D, P<0.05) and the group treated for 7 days (group A, P<0.01). No significant differences in the mean length of time to vaginal opening and leucocytic influx were found among subgroups treated with progesterone on day 0, 5, 10, and 15 of vaginal opening. Figure 1 shows the variation in the day of leucocytic influx into vaginal smears after the progesterone removal. The length of time to leucocytic influx in the control group (group D) varied widely, while those in the progesterone treated groups took mostly 6 days, the mode day, after Silastic tubing removal. Among the groups treated with progesterone, the number of animals showing results on this mode day (at 6 days) ranged from 38% in group A (9/24) to 79% in groups B and C (19/24). Fresh ova in the oviduct were seen in all animals sacrificed on the day of leucocytic influx into the vaginal smear with the exception of one of 24 females in group A and two of 12 females in group D, whose ova were, however, seen in the uterus (Table 2). The mean number of ova seen in No. of oO - animals 10 QO: day 0 of implantation (subgroup a), D day 10 (c), [fj : day 15 (a : day 5 (b) a Group A Fe] fed M El Sas id ests Ga Gea 61 7] 6] 9 [io[ 13] 12) 13] 14125 219 0 fF: SER Er Te ee ae Group B M _E|M E|M E|M E|M E|M E|M E|M E[M E[M E[M E|M E] 7[ 8] 9 [io[ii[i2[13] 24 [15 [16 [17] 18] Group C ° M_E[M E[M E[M Las Ti6 [17 [ 18} M E[M E[M E[M E[M E|M E|M E|M E[M E|M E[M E[M EM E[M E 1 273 ,4[s Te[7][e] 9 fio fir [212 [13] 14 Bra Group D 1 fa nel Ea el ° 7G oe UE eee eG it 3f4 [se [7] els fio fir fi2 [23 fas fis i Lie 273 Days after removal of the tubing Fic. 1. Day of leucocytic influx into the vaginal smear after removal of the progesterone-filled (Groups A-C) or empty (Group D) tubing in the guinea pigs. M: 08: 00-09: 00 hr, E: 17: 00-18: 00 hr oviduct was not significantly different among any of the groups. DISCUSSION The influx of leucocytes into the vaginal smear following the appearance of fully cornified cells has been considered to be the end of vaginal estrus [13], and an indication of ovulation [14, 15]. In these studies, all animals sacrificed on the first day of leucocytic smear showed ovulation. Our data confirmed that this smear pattern was a sign of ovulation in cyclic female guinea pigs. As the agent for the synchronization or the alteration of estrous cycle, progesterone adminis- tration has been extensively used in many species of animals [7-10]. Single or multiple injections of progesterone given at different stage of the estrous cycle in guinea pigs induced prolongation of ovulation depending on the stage of administration [6, 16]. In guinea pigs, the timing of ovulation 142 TABLE 2. H. UepA, T. KOSAKA AND K. W. TAKAHASHI Results of ovulation test at the day of leucocytic influx into the vaginal smear after removal of the progesterone-filled (Groups A-C) or empty (Group D) tubing in the guinea pigs No. of animals Treatment showing ovulation é No. Ht No. of ee roup Implantation Day of estrous cycle aeauae nig inwendue period het Silastic tube °*#mined Oe Hae (Ranee) (days) Was Waeeried in oviduct in uterus 0 6 6 4.2 (3-5) A 7 5 6 5 1 4.0 (2-5) 10 6 6 4.0 (3-5) 15 6 6 4.0 (3-5) 0 6 6 4.2 (4-5) a 4 5 6 6 4.3 (3-5) 10 6 6 3.7 (3-4) 15 6 6 4.3 (3-6) 0 6 6 4.5 (4-5) C ‘1 5 6 6 4.2 (3-5) 10 6 6 4.3 (4-5) 15 6 6 4.3 (4-5) 0 3 1 2 40(4) > ot 5 3 3 3.3 (3-4) 10 3 3 : = (Control) nye 15 3 3 4.0 (3-5) after progesterone injection seems to be regulated by the stage of the estrous cycle at which the progesterone was administered. Induction of ovulation in guinea pigs by other hormones re- quires injection at some fixed time during the estrous cycle [17-21]. However, the present data revealed that effect of long-term progesterone treatment on inducing the synchronized ovulation following leucocytic influx was independent on the stage of the estrous cycle at which progesterone implants were given. Here, we have shown that long-term progester- one treatment for 14 and 21 days was effective in inducing ovulation within 5-6 days after the removal of the progesterone implant. Tso and Tam [22] reported that approximately 5 days elapse between onset of luteolysis and ovulation in guinea pigs. Perhaps the removal of the progester- one implants provides the same function as the onset of luteolysis which in turn halted the suppression of gonadotropins caused by chronic progesterone treatment, resulting in synchronized ovulation within 5-6 days. These findings indicate, therefore, that long-term implantation of proges- terone tubing, greater than 14 days, given at any estrous stage of female guinea pigs induces the synchronized ovulation within 5-6 days after the removal of progesterone tubing. Present study resulting in the synchronized ovulation in female guinea pigs gives first step to control a scheduled supply of animals with good uniformity. However, the following step has been remained: whether the animals synchronously ovulated using this method have normal reproduc- tive activities, i. e. copulation, pregnancy, parturi- tion, and lactation. ACKNOWLEDGMENT The authors are grateful to Dr. Y. Shirasu, Toxicology Division, Institute of Environmental Toxicology, Tokyo, for his valuable advice and suggestions during this study. 10 11 12 Synchronized Ovulation in Guinea Pigs REFERENCES Everett, J. W. (1944) Evidence in the normal albino rat that progesterone facilitates ovulation and cor- pus luteum formation. Endocrinology, 34: 136-137. Everett, J. W. (1948) Progesterone and estrogen in the experimental control of ovulation time and other features of the estrous cycle in the rat. Endocrinolo- gy, 43: 389-405. Leuter,L. A., Ciaccio,L. A. and Lisk, R. D. (1970) Progesterone: Regulation of estrous cycle, ovulation and estrous behavior in the golden ham- ster. Endocrinology, 86: 1287-1297. Reuter, L. A. and Lisk, R. D. (1973) A biphasic effect of progesterone on ovulation in the hamster. Fed. Proc., 32: 230. Greenwald, G. S. (1977) Exogenous progesterone: Influence on ovulation and hormone levels in the cyclic hamster. J. Endocrinol., 73: 151-155. Joslyn, W. D., Wallen, K. and Goy, R. W. (1976) Advancement of ovulation in the guinea-pig with exogenous progesterone and related effects on length of the oestrous cycle and life span of the corpus luteum. J. Endocrinol., 70: 275-283. O’Mary,C.C., Pope,A.L. and Casida, L. E. (1950) The use of progesterone in the synchroniza- tion of the estrual periods in a group of ewes and the effect on their subsequent lambing records. J. Anim. Sci., 9: 499-503. Christian, R.E. and Casida,L.E. (1948) The effects of progesterone in altering the estrous cycle of the cow. J. Anim. Sci., 7: 540. Ulberg, L.C., Grummer, R. H. and Casida, L. E. (1951) The effects of progesterone upon ovarian function in gilts. J. Anim. Sci., 10: 665-671. Baker,L.N., Ulberg,R.H. and Casida, L. E. (1954) Inhibition of heat by progesterone and its effect on subsequent fertility in gilts. J. Anim. Sci., 13: 648-657. Woody, C. O., First, N. L. and Pope, A. L. (1967) Effect of exogenous progesterone on estrous cycle length. J. Anim. Sci., 26: 139-141. Biegon, A., Parsons,B., Krey,L.C., Kamel, F. 13 14 15 16 20 21 22 143 and McEwen, B. S. (1983) Behavioral and neuroen- docrine effects of long-term progesterone treatment in the rat. Neuroendocrinology, 37: 332-335. Donovan, B. T. and Lockhart, A. N. (1972) Light and the timing of ovulation in the guinea-pig. J. Reprod. Fertil., 30: 207-211. Young, W.C., Myers, H.I. and Dempsey, E. W. (1933) Some data from a correlated anatomical, physiological and behavioristic study of the repro- ductive cycle in the female guinea pig. Am. J. Physiol., 105: 393-398. Hermreck, A. S. and Greenwald, G. S. (1964) The effects of unilateral ovariectomy on_ follicular maturation in the guinea pig. Anat. Rec., 148: 171- 176. Ginther, O. J. (1967) Length of estrous cycle and size of corpus luteum in guinea pigs and sheep treated with progesterone at different days of the estrous cycle. Am. J. Vet. Res., 30: 1975-1978. Reed, M. and Hounslow, W. F. (1971) Induction of ovulation in the guinea pig. J. Endocrinol., 49: 203- 211. Donovan, B.T. and Lockhart, A.N. (1972) Gonadal hormones and the control of ovulation in the guinea pig. J. Endocrinol., 55: 599-607. Rawson, J.M.R., Galey,C.1., Weinberg, L. C. and Hodgson, B. J. (1979) Effect of gonadotropins on follicular development, ovulation, and atresia in the mature guinea pig. Hormone Res., 10: 25-36. Terranova, P. F. and Greenwald, G. S. (1981) In- creased ovulation rate in the cyclic guinea pig after a single injection of an antiserum to LH. J. Reprod. Fertil., 61: 37-42. Garza, F., Shaban,M.A. and Terranova, P. F. (1984) Luteinizing hormone increases the number of ova shed in the cyclic hamster and guinea pig. J. Endorinol., 101: 289-298. Tso, E.C. and Tam, W.H. (1977) The effect of continuous treatment with prostaglandin F-2a on oestrous cycle length and corpus luteum regression in hysterectomized guinea-pigs. J. Reprod. Fertil., 50: 335-336. ZOOLOGICAL SCIENCE 5: 145-152 (1988) Neuroendocrine Regulation of the Development of Seasonal Morphs in the Asian Comma Butterfly, Polygonia c-aureum L.: Difference in Activity of Summer-morph-producing Hormone from Brain-extracts of the Long-day and Short-day Pupae KaTsuHIKo ENDo, TADAKATSU Masaki and Kanyt Kumacar' Environmental Biology Laboratory, Biological Institute, Faculty of Science, and ‘Biological Institute, Faculty of Liberal Arts, Yamaguchi University, Yamaguchi 753, Japan ABSTRACT— Seasonal morphs of the butterfly, Polygonia c-aureum L., were shown to be deter- mined by a factor producing summer morphs (SMPH). The factor, which was extracted with 2% NaCl from the brains of Polygonia pupae, but unsuccessful with acetone or 80% ethanol, was thought to be a peptide hormone. The factor present in the 2%-NaCl extracts was precipitated by ammonium sulfate at 80% saturation. The summer-morph-producing activity was evaluated by injecting extracts containing a sufficient amount of the factor into the abdomen of 0-day-old pupae of autumn-morph producers. The recipients showed a dose-dependence in response to the factor (manifestation of characteristics of summer morphs). The factor was present in the pupal brains of both summer-morph and autumn-morph producers (LD- and SD-pupae). The quantity of SMPH present in the brains of 0-day-old pupae of autumn-morph producers seemed to be larger than in those of summer-morph producers of the same age. Sexual differences in the quantity of the factor also seemed to exist in the summer-morph producers. Furthermore, the SMPH-activity was detected from the brain-extracts of © 1988 Zoological Society of Japan Papilio xuthus, Lycaena phlaeas daimio and Bombyx mori. INTRODUCTION The Asian comma butterfly, Polygonia c- aureum L., exhibits seasonal dimorphism, i. e. summer and autumn morphs (Fig. 1). The season- al-morph development is governed by photoperiod and temperature during the larval stage as has been reported in previous papers [1, 2]. The physiological mechanism underlying the photoperiodic control of seasonal-morph deter- mination was shown to involve a factor producing summer morphs (SMPH). The neurosecretory cells responsible for the factor are present in the pars intercerebralis of the brain. The factor is then conveyed along the axons and released into the hemolymph from the corpora cardiaca and/or corpora allata in the early pupal stage [3, 4]. The present study was designed to establish Accepted July 3, 1987 Received November 20, 1986 methods of extraction of the factor showing SMPH-activity and of bioassay using Polygonia pupae. The study was extended to see whether or not the quantity of the factor varies according to the sex of the donors or to photoperiodic condi- tions under which the donors developed from the egg stage. Subsequently, some preliminary experi- ments were carried out to assess whether or not the factor showing SMPH-activity is present in the brain of other lepidopteran insects. MATERIALS AND METHODS Animals Eggs and larvae of the butter- flies, P. c-aureum, Papilio xuthus and Lycaena phlaeas daimio, and those of the silkmoth, Bom- byx mori, were held in two kinds of transparent plastic containers (@ 9X5 cm? or 19135 cm?) and were exposed to either a long-day photoperiod alternating 16-hr light and 8-hr dark periods (16L-8D) or a short-day photoperiod of 8L-16D 146 K. ENpo, T. MASAKI AND K. KUMAGAI Summer Autumn Fic. 1. Female and male butterflies of summer and autumn morphs. Butterflies in the upper row show the wing patterns of the dorsal side, whereas those in the lower row show the wing patterns of the ventral side. at 20°C and 25°C. The larvae of P. c-aureum were fed on leaves of Humulus japonicus, whereas those of P. xuthus, L. phlaeas daimio and B. mori were fed on leaves of Fagara ailanthoides, Rumex acetosa and Morus tiliaefolia, respectively. The rearing containers were placed in a cabinet with temperatures of either 20°C or 25°C and were illuminated by two 20-W white fluorescent tubes, which were controlled by a 24-hr time-switch. During the light period, the light-intensity was about 500 lux. Under long-day conditions at 20°C and 25°C, larvae and pupae of P. c-aureum all developed into summer morphs, whereas under short-day condi- tions, they all developed into autumn morphs without exception. Larvae and pupae developed from the egg stage under long-day conditions are referred to hereafter as LD-larvae and LD-pupae, whereas those de- veloped under short-day conditions are referred to as SD-larvae and SD-pupae. Extraction of SMPH Brains were ob- tained from 0-day-old Polygonia pupae (4-12 hr after larval-pupal ecdysis), and pharate pupae of Papilio and Lycaena, by dissection in saline (0.9% NaCl). Brains of the silkmoth, B. mori, were also obtained in the same manner. One hundred brains from each species were grouped and stored at —85°C. Each 100-brain sample was homogenized in acetone (500 yl X2) with a Teflon homogenizer at ice-bath temperature and was dried to powder under reduced pressure at room temperature (about 25°C). Then the sample was washed in 80% ethanol (100 sl x2) and extracted with 2% NaCl (50 1x3) at ice-bath temperature. At each step insoluble materials were separated by centrifuga- tion at 12,000 xg for 30 min at 5°C. A supernatant of 2% NaCl was added with 84 mg of ammonium sulfate (80% saturation) to precipitate the factor and the precipitate was dissolved in distilled water to give an extract of 2%-NaCl. Washings with acetone and 80% ethanol were dried under re- Summer-Morph-Producing Hormone 147 duced pressure at room-temperature and the residues were redissolved in saline (acetone and 80%-ethanol extracts). Bioassay of SMPH-activity Five sl of the sample containing the extract of 1- to 20-brain equivalents was injected into the abdomen of grade 0 grade 1 grade 2 Q-day-old female Polygonia SD-pupae (4-12 hr after larval-pupal ecdysis). In controls, the 0-day- old female Polygonia SD-pupae were injected with saline containing 2.8 mg/5 wl of ammonium sulfate (40% saturation) or distilled water. The injection was made through the ventro- Fic. 2. Wing patterns of the female butterflies of each grade of summer-morphs (grades 0-4), upon which bioassay of summer-morph-producing hormone was performed. The butterflies arranged in the upper row show the wing patterns of the dorsal side, whereas those in the lower row show the wing patterns of the ventral side. TABLE 1. Criteria for seasonal-morph classification on the basis of ventral-side wing color Grades Morphs Characteristics 0 autumn Ventral sides of the wings are mostly covered with dark-brown scales and dark-yellow scales are present only on the basal region of the wings. 1 autumn Ventral sides of the wings are mostly covered with dark-brown scales, but dark yellow scales appear in the peripheral regions of the wings. 2, intermediate A thick/dark-brown stripe remains on the ventral sides of the wings, but the other regions are mostly covered with light-brown scales. 3 summer Ventral sides of the wings are mostly covered with dark-yellow scales, but a thick/brown stripe remains. 4 summer The thick stripe becomes light brown and the other regions are mostly covered with dark-yellow scales. 148 K. ENpo, T. MASAKI AND K. KUMAGAI lateral/intersegmental region between the 6th and the 7th abdominal segments. On the day of emergence, the female butterflies were examined for the characteristics of summer morphs and classified into one of grades 0-4. An average grade score (AGS) for summer morphs was obtained from the response of 6-20 insects, the classification being based on a gradient of the color of the ventral side of the wings (Fig. 2, Table 1). Female butterflies of grade 0 and 1 were regarded as autumn morphs, those of grade 3 and 4 were regarded as summer morphs, and there were also intermediates of grade 2. RESULTS SMPH-activity in the brain-extracts of 0-day-old LD- and 0-day-old SD-pupae of P. c-aureum Brain-extracts of 2% NaCl, acetone and 80% ethanol were provided from 0-day-old LD- and 0-day-old SD-pupae, and 5d, containing the extract of 10-brain equivalents, were injected into the abdomen of 0-day-old female SD-pupae. For the control groups, the injection was made with either saline containing ammonium sulfate of 40% saturation (ca. 2.8 mg/S d) or distilled water. When 2%-NaCl extracts of the brains of LD- pupae were applied to 0-day-old female SD-pupae, the majority (24 out of 30) of the recipients responded and developed into summer or in- termediate morphs. On the other hand, the recipient SD-pupae injected with acetone or 80%- ethanol brain-extract developed into autumn morphs of grade 0, as did the untreated controls and SD-pupae treated with either saline containing 2.8 mg/S 4 of ammonium sulfate or distilled water (Table 2). Similar results were also obtained by the injec- tion with brain-extracts of 0-day-old SD-pupae. All SD-pupae receiving the 2%-NaCl extract of the brains of SD-pupae developed into summer morphs (grade 3 and 4). They recorded an AGS of 3.9 (average grade score for summer morph). They were judged as having more eminent charac- teristics of summer morphs than those injected with 2%-NaCl brain-extracts of LD-pupae (AGS 2.3). In contrast, the recipients of either acetone or 80%-ethanol brain-extracts of 0-day-old LD- pupae developed into autumn morphs of grade 0 (Table 2). The results indicated that a factor showing the SMPH-activity is present in the brains of both 0-day-old SD- and 0-day-old LD-pupae of P. TABLE 2. Effects of the brain-extracts of Polygonia pupae on the development of seasonal morphs No. of butterflies classified Source of extracts into grades: and extractants Ne: AGS 0 1 2, 3 4 Control pupae Untreated 20 20 0 0 0 0 0.0 Distilled water 20 20 0 0 0 0 0.0 Saline (0.9% NaCl) 20 20 0 0 0 0 0.0 Saline containing 20 20 0 0 0 0 0.0 2.8 mg of (NH4)2SO4 LD-pupae 2% NaCl 30 12 3 2.1 Acetone 7 0 0 0 0 0.0 80% ethanol 22 22 0 0 0 0 0.0 SD-pupae 2% NaCl 21 0 0 0 2 19 3.9 Acetone 19 19 0 0 0 0 0.0 80% ethanol 27 23 4 0 0 0 0.1 Each recipient injected with an extract of 10-brain equivalents. Summer-Morph-Producing Hormone 149 c-aureum. The factor could be extracted with 2% NaCl and precipitated by adding ammonium sul- fate at 80% saturation. However, the factor could not be extracted with acetone or 80% ethanol. Dose-dependence of SMPH-activity in the 2%- NaCl extracts of the brains of LD- and SD-pupae (0-day-old) Brain-extracts were made with 2% NaCl from 0-day-old LD- and 0-day-old SD-pupae and 5 wl were injected into the abdomen of 0-day-old female SD-pupae. During these injections, the SD-pupae of each of the groups received a different dose of brain-extract (0- to 20-brain equivalents). As is summarized in Figure 3, the recipient SD-pupae showed dose-dependent responses in AGS with brain-extracts of both LD- and SD- pupae. The dose eliciting a half response (AGS 2) was obtained by an extract containing 3-brain equivalents of SD-pupae. For a full response (AGS 4), the recipients required doses of extract larger than 10-brain equivalents. When injecting brain-extract of LD-pupae, the recipients required approximately 2-fold larger doses than with ex- tracts of SD-pupae; the dose eliciting a half response (AGS 2) was obtained by an extract containing 5-brain equivalents of LD-pupae, but a dose of 20-brain equivalents was still insufficient for the induction of a full response (AGS 4). Ww Oo = 01 3 5 10 20 Number of brains injected Fic. 3. Dose-response curves of SMPH obtained in the brain-extracts of 0-day-old SD- (solid circles) and 0-day-old LD-pupae (open circles). Each point depicts the AGS score (average grade score for summer morphs) with a standard error (thin verti- cal line) of about 20 females. The results indicate that the quantity of factor in the brains of autumn-morph-producers (0-day-old SD-pupae) is approximately 2-fold larger than the factor in the brains of summer-morph-producers of the same age. However, the SMPH-rich brain of the autumn-morph-producer (0-day-old SD-pupa) does not seem to contain a sufficient amount of factor to produce a summer-morph butterfly (grade 4) from the pupa of an autumn-morph- producer (SD-pupa). Sexual differences in the SMPH-activity of the brain-extracts of LD- and SD-pupae (0-day-old) To clarify whether or not a sexual difference is present in the SMPH-activity of brain-extracts of LD- and SD-pupae, brain-extracts were made with 2% NaCl from 0-day-old male and 0-day-old female SD-pupae and precipitated by adding ammonium sulfate to 80% saturation. The precipi- tate was suspended in distilled water and 5 wl containing the precipitate of 5-brain equivalents was injected into the abdomen of 0-day-old female SD-pupae. Brain-extracts were also made from male and female LD-pupae of the same age and 5-brain equivalents were injected into female SD-pupae in the same manner. The brain-extract of female SD-pupae was found to show approximately the same relative SMPH-activity as the brain-extract of male SD- pupae. The recipient SD-pupae injected with extract of either male or female SD-pupae de- veloped into summer-morph and _ intermediate- morph butterflies in addition to a few autumn- morph ones. They recorded AGS scores of 1.6 (male extract) and 1.5 (female extract) (Table 3). Sexual differences in SMPH-activity were pres- ent in the brain-extracts of 0-day-old LD-pupae. The extract made from the brains of male LD- pupae showed a high SMPH-activity (AGS 1.5), almost corresponding to the activity of SMPH-rich extracts of the brains of SD-pupae (AGS 1.6 in male extract and AGS 1.5 in female extract). In contrast, the SMPH-activity of the brain-extract of female LD-pupae (AGS 0.6) was significantly lower than the SMPH-activity obtained with the brain-extract of male LD-pupae (Table 3). The results indicate that a large amount of factor showing SMPH-activity is present in the brains of 150 K. ENpo, T. MASAKI AND K. KUMAGAI TABLE 3. short-day pupae Sexual differences in the SMPH-activity of the brain-extracts of Polygonia long-day and No. of butterflies classified into grades: Source of extracts No. AGS 0 1 2 3 4 LD-pupae males 13 Z 5 6 0 0 1.3 females 14 6 8 0 0 0 0.6 SD-pupae males 10 2 2 5 1 0 1.5 females 14 0 a 5 2 0 1.6 Each recipient was injected with an extract of 5-brain equivalents. both male and female SD-pupae (0-day-old). In the brains of LD-pupae (0-day-old), however, sexual differences exist with regard to the quantity of factor; the amount present in the brains of 0-day-old male LD-pupae is approximately twice the amount present in the brains of female LD-pupae of the same age. SMPH-activity in brain-extracts of other lepidopter- an species To assess whether or not factor showing SMPH- activity is present in the brains of other lepidopter- an insects, brains were obtained from pharate pupae of two species of butterflies, P. xuthus and L. phlaeas daimio. The brains were homogenized in acetone, washed in 80% ethanol, extracted with 2% NaCl, and precipitated by adding ammonium sulfate to 80% saturation. Then the precipitate or residue of evaporated washings (acetone and 80% ethanol) was dissolved in saline and 5 J containing the precipitate (or residue) of 20-brain equivalents was injected into the abdomen of female SD- pupae (0-day-old). Extracts were also made from the brains of diapause-egg producers (LD) of the silkmoth, B. mori, and 20-brain equivalents were injected into the abdomen of 0-day-old SD-pupae in the same manner. Each 2%-NaCl extract made from the brains of Papilio pharate-pupae and Lycaena pharate-pupae showed SMPH-activity (AGS 0.9 and 0.6) when the extract was injected into the abdomen of Polygonia female SD-pupae (0-day-old). The majority of the recipient pupae injected with 2%-NaCl brain-extracts of pharate pupae of either Papilio or Lycaena produced butterflies having some characteristics of summer morphs. Howev- TasLe 4. SMPH-activity in the brain-extracts of three species of lepidopteran insects No. of butterflies classified into grades: Source of extracts and extractants No. AGS 0 1 2 3 4 Papilio xuthus (pharate pupae) 2% NaCl 10 3 2) 2 0 0 0.9 80% ethanol 10 10 0 0 0 0 0.0 Lycaena phlaeas daimio (pharate pupae) 2% NaCl 10 1 9 0 0 0 0.9 80% ethanol 10 10 0 0 0 0 0.0 Bombyx mori (adults) 2% NaCl 9 0 0 2 3 4 3.2 80% ethanol 6 2 3 1 0 0 0.8 ees SS ee eee Recipient SD-pupae injected with an acetone brain-extract all developed into autumn morphs of grade 0. Summer-Morph-Producing Hormone 151 er, the recipient SD-pupae injected with acetone extract or 80%-ethanol extract of the brains of Papilio or Lycaena pharate pupae did not show any response and developed into autumn morphs of grade 0 (Table 4). Positive data were obtained by injection of two of the extracts (2% NaCl and 80% ethanol) made from the brains of the silkmoth, B. mori. The recipient SD-pupae injected with 2% NaCl brain- extract developed into summer or intermediate morphs, whereas those injected with 80%-ethanol extract developed into butterflies having some characteristics of summer morphs (grades 1 and 2). They recorded AGS scores of 3.2 (2%-NaCl extract) and 0.9 (80%-ethanol extract), respective- ly, but the recipients of the acetone brain-extract all developed into autumn morphs of grade 0. The results indicate that a factor showing the same SMPH-activity as the cerebral factor of P. c-aureum is present in the brains of three species of lepidopteran insects (P. xuthus, L. phlaeas daimio and B. mori). The factor can be extracted with 2% NaCl and precipitated by adding ammonium sul- fate to 80% saturation as observed in the case of the factor of P. c-aureum. The SMPH-activity present in the brains of the silk-moth, B. mori, is approximately six times greater than the activity present in the brains of Papilio (or Lycaena) pharate pupae and is almost comparable to the activity of SMPH-rich brains of 0-day-old SD- pupae. DISCUSSION Summer-morph butterflies of P. c-aureum were shown to be produced by a neurosecretory factor (SMPH) which is secreted in the early pupal stage in summer-morph producers [3, 4]. The factor was present in the brains of both 0-day-old LD- and 0-day-old SD-pupae in this butterfly. It is thought to be a peptide hormone, since it was extracted with 2% NaCl but not with acetone or 80% ethanol. This theory is further supported by the fact that it was precipitated by raising the concen- tration of ammonium sulfate to 80% saturation. Although we failed to evaluate the SMPH- activity remaining in the aqueous part of the ammonium-sulfate precipitation (80% saturation), it appears to be negligible. One half of the SMPH-activity originally present in the brain- extracts (2% NaCl) was precipitated by raising the saturation of ammonium sulfate from 50% to 65% (unpublished data). When the factor showing SMPH-activity is deprived from hemolymph in the early pupal stage, Polygonia pupae are thought to develop into autumn-morph butterflies [1, 3, 4]. It has been determined that SMPH-deprivation occurs in SD- pupae and the deprived state is thought to be achieved mainly by suppressing the secretion of the factor. On the other hand, in LD-insects, both production and secretion of the factor may be enhanced by long days. A large amount of factor is thought to be conveyed along axons, secreted into hemolymph hours following larval pupal ecdysis and it is believed that SMPH-activity remaining in the brains of 0-day-old LD-pupae becomes lower than one-twentieth the SMPH-activity required for summer-morph development. In addition, sexual differences were present in the SMPH-activity of the brain-extracts of 0-day- old LD-pupae (Table 3). However, in LD-pupae, both male and female brains produce an equally large amount of the factor and are thought to secrete it into hemolymph until the hemolymph- titer of the factor satisfies an essential level for summer-morph development. Therefore, all LD- pupae succeed in developing into summer morphs. In contrast, P. c-aureum has an apparent sexual dimorphism in autumn morphs. It may be that females having darker colored wings than those of males as autumn morphs require an exposure to a higher titer of the factor than the males in the development of _ light-color-winged summer morphs. The high-titer flux is thought to be achieved by females secreting a larger quantity of the factor than males. In Papilio and Lycaena, mechanisms underlying the photoperiodic control of seasonal morphs —spring and summer morphs—were also shown to involve a neuroendocrine factor (SMPH?) which is secreted from the brains of summer- morph producers in the pharate-pupal stage [5, 6]. The factor (?) was extractable with 2% NaCl from the brains of Papilio and Lycaena pharate- pupae, but not with acetone and 80%-ethanol, as 152 K. Enpo, T. Masaki AND K. KUMAGAI observed in the factor of P. c-aureum. The brains of the silkmoth, B. mori, were found to contain factor showing SMPH-activity in P. c-aureum and was able to be extracted with 2% NaCl and 80% ethanol. The brain-extract of the silkmoth was estimated to have a SMPH-activity six-times high- er than the brain-extracts of Papilio (or Lycaena) pharate-pupae (Table 4). We were unable to provide any evidence as to whether or not the factor extracted from the brains of Papilio and Lycaena pharate-pupae plays an essential role in the summer-morph development of these insects. However, we have concluded that a factor (peptide hormone?) showing the same effects as the Polygonia SMPH is present in the brains of several other lepidopteran insects and that the quantity may vary depending on the insect species from which the brains are obtained. ACKNOWLEDGMENT The authors wish to express their sincere gratitude to Professor A. Okajima and to Professor Y. Chiba of Yamaguchi University for advice and valuable sugges- tions during the course of this work. This work was supported in part by a grant from the Ministry of Education, Science and Culture of Japan (No. 6154052). REFERENCES 1 Fukuda, S. and Endo, K. (1966) Hormonal control of the development of seasonal forms in the butterfly, Polygonia c-aureum L. Proc. Japan Acad., 42: 1082- 1087. 2 Hidaka, T. and Takahashi, H. (1967) Temperature condition and maternal effect as modifying factor in the photoperiodic control of seasonal forms in Poly- gonia c-aureum (Lepidoptera, Nymphalidae). Annot. Zool. Japon., 40: 200-204. 3 Endo, K. (1972) Activation of corpora allata in relation to ovarian maturation in the seasonal forms of the butterfly, Polygonia c-aureum L. Dev. Growth Differ., 14: 263-274. 4 Endo, K. (1984) Neuroendocrine regulation of the development of seasonal forms of the Asian comma butterfly, Polygonia c-aureum L. Dev. Growth Dif- fer., 26: 217-222. 5 Endo, K. and Funatsu, S. (1985) Hormonal control of seasonal morph determination in the swallowtail butterfly, Papilio xuthus L. (Lepidoptera: Papilioni- dae). J. Insect Physiol., 31: 669-674. 6 Endo, K. and Kamata, Y. (1985) Hormonal control of seasonal-morph determination in the small copper butterfly, Lycaena phlaeas daimio Seitz. J. Insect Physiol., 31: 701-706. ZOOLOGICAL SCIENCE 5: 153-157 (1988) © 1988 Zoological Society of Japan Sexual Maturation in Female Wild Mice: Combined Effect of Adults’ Urinary Chemosignals and Minimum Time of Exposure to Stimulus Substances for Bringing the Effects SUBHASH C. PANDEY and SHEO D. PANDEY Department of Zoology, Christ Church College, Kanpur, India ABSTRACT— Young females exposed to urine of adult males attained puberty earlier than those exposed to urine of adult females. The puberty accelerating property of the male urine was masked when mixed with the urine of females either in ratio of 1:1 or 2:1. Further, puberty acceleration was initiated only when the young females were exposed to male urine for a minimum of 6 days. However, the puberty delay in subject females was not apparent by 6 days but was effectuated only after a long exposure of 9 days to female urine INTRODUCTION The age of puberty in mammals is thought to be determined by genetic factors and is specific for a particular species [1]. In addition to hereditary control over the onset of puberty, other factors such as nutrition and social stimuli also play an important role in scheduling the onset of puberty within the range of ages limited by heredity [2]. First oestrus is accelerated in young female mice caged with adult males relative to control females housed alone [3]. This effect is due to a male urinary pheromone [4] which acts synergistically with social cues associated with the physical presence of the male [5]. Several workers have replicated much of the accelerating effect by exposing the females to the urine collected from adult males [6-8]. In contrast to the ability of the adult male to accelerate sexual maturation in juvenile female mice, a pheromone of female origin is known to delay the maturation in the same sex [9]. McIntosh and Drickamer [10] reported that the delay of puberty in young female mice also occurs by exposure to urine collected from grouped adult females. Urine from socially isolated females is without effect on the time of puberty. However, Accepted August 1, 1987 Received April 22, 1987 bladder urine of females invariably contains the maturation-delaying chemosignal irrespective of social condition [10]. Our laboratory studies have revealed that puberty in young female wild mice is also suscepti- ble to social influences similar to its laboratory cousin. The causative factors are contained in urine of the adult individuals. The puberty- delaying chemosignal is released in the urine of females after housing them together for 15 days and a daily exposure of subject females to chemo- signal, at least for 2 hr is needed for effective delay of puberty to occur [11]. The purpose of this study was to determine: (i) what is the resultant effect, if both puberty accelerating and delaying chemosig- nals are mixed, and (ii) how long the young females must be treated with these chemosignals for bringing about their respective effects. MATERIALS AND METHODS The experiments were carried out on wild mice, Mus musculus domesticus. The animals used in these experiments were wild trapped as the species usually avoids breeding in the laboratory. Mice weighing 12-16g were used as adults while females of 4-5 g body weight were employed as youngs in the experiments. All animals were maintained in the laboratory at 28-34°C tempera- ture with natural light-dark hours and fed on a diet 154 S.C. PANDEY AND S. D. PANDEY consisting of soaked Bengal gram (Cicer arieti- num), boiled rice and milk. Water was supplied ad libitum. Young females were housed individually in galvanised steel cages, 34 x 18 x 14 cm, and were exposed to urine of adults as described in experi- ments I and II. The adult females were grouped in colony cages (303030 cm) at a density of 10 mice/cage for 30 days before urine collection. The males remained isolated in steel cages for the same period. The urine was collected by placing the mice over a petridish and gently squeezing the abdomen, and diluted in distilled water, 1:9. A drop (0.5 ml) of diluted urine was applied on external nares of the subject females with the help of a small paintbrush twice daily at 9.00 and 17.00 hr. Young subject females were examined daily from the start of experiment until the occurrence of vaginal perforation. Starting on the day of vaginal perforation, a vaginal lavage was taken daily and examined under microscope to deter- mine the stage of the oestrous cycle, using the criteria of Bronson et al. [12] and Rugh [13]. The vaginal smears were examined until the occurrence of first oestrus. The data were analysed by one way analysis of variance. Experiment I To see the combined effect of male and female urinary chemosignals on puberty, 25 individually housed young females weighing 4.4+0.28 g were randomly divided into 5 groups (5 mice/group). They were painted daily on their external nares TABLE 1. with water (control, group I) or urine from males (group II) or urine from females (group III) or mixed urines from males and females (1:1, group IV; 2:1, group V). The treatments continued till the occurrence of first vaginal oestrus. Experiment II In this experiment minimum time of exposure to stimulus substances for bringing the effects on puberty was observed. The experiment consisted of two parts. In one part, young females were exposed to male urine while in other part they were exposed to female urine. In each part there were 5 groups, each having 5 randomly selected young mice weighing 4.2+0.23g. Females in group I were treated with water for 12 days (control). The urine treatment was given for 3 days (group II), 6 days (group III), 9 days (group IV) or 12 days (group V). Vaginal smears were examined from all subjects until the occurrence of first vaginal oestrus. RESULTS Experiment I Sexual maturation in young females as assessed by the time taken in occurrence of first vaginal oestrus was delayed by exposure to mixed urines in the same manner as if they were exposed to urine of adult females. Ages of first oestrus observed in groups III, IV and V were not significantly different with each other (C. D.=2.08). Onset of Effect of exposure to urines collected from adult individuals or their mixtures on sexual maturation of young females (n=5/group) Mean time (in days) taken for first Group Treatment vaginal oestrus to occur I Painted with water (control) 29.2 +0.66 | II Painted with urine from adult males 21.2 +0.63 | III Painted with urine from adult females 36.2+0.58 | IV Painted with male-female urine (1:1) 35.2+0.77 Vv Painted with male-female urine (2:1) 34.6+0.46_ F=80.9** d.f.=4, 20 C.D. at 5% =2.08 Means connected with same vertical line are at par at 5% level of significance. Urinary Chemosignals and Sexual Maturation 155 TaBLe 2. Exposure of young females (n=5/group) to stimulus substances for different duration and its effect on pubertal onset Mean time (in days) taken for the occurrence of first vaginal oestrus in subject females exposed to water Group Treatment Se inhe hon Male donors Female donors I 12 days exposure to 30.2+0.52 29.2+0.78— water (control) a II 3 days exposure to urine 29.0+0.63_ 30.4+0.69_ Ill 6 days exposure to urine 21.6+0.83 | 32.4+0.92 - IV 9 days exposure to urine 20.2+0.51 35.8+0.52— Vv 12 days exposure to urine 20.6+0.45_ 36.4+0.60_ F=49.3** F=16.2** d.f.=4, 20 d.f.=4, 20 C.D. at 5% =2.05 C.D. at 5% =2.34 Means connected with same vertical line are puberty in females exposed to urine of adult males was significantly earlier (P<0.01) than those ex- posed to urine of adult females. The puberty accelerating effect of male urine was completely masked by female urine (even when the latter constituted only one third of the total mixed urine) as the young females exposed to mixed urines attained sexual maturity significantly later (P<0.01) than the control females (Table 1). Experiment II The time taken for the onset of puberty in females treated with male urine for 3 days was not significantly different from that of control (group I). By contrast, the first oestrus in females treated with male urine for 6 days occurred significantly earlier than control or 3 day-treated females. There was no significant difference in onset of puberty (i.e., occurrence of first oestrus) among females of groups III, IV and V treated with male urine for 6, 9 and 12 days respectively (Table 2). In the second part of the experiment, the subjects treated with female urine for 3 or 6 days matured at about the same time at which the control ones. The occurrence of first oestrus was significantly dalayed in females of group IV (9 days treatment). No further delay in puberty was observed by increasing the day of treatment (group V). at par at 5% level of significance. DISCUSSION Following conclusions are derived from the foregoing pair of experiments: (1) The maturation- delaying chemosignal of female origin can override the male urinary factor causing puberty accelera- tion in young females. (2) Young females require exposure of, at least, 6 days to male urine or 9 days to female urine for puberty acceleration or delay to occur. Investigations of the factors causing acceleration or delay in puberty in females have been focussed on laboratory strains of mice [3, 4, 7, 14]. Experiments conducted in our laboratory on wild mice have revealed that onset of puberty in females is a labile phenomenon in this species which is regulated by adults’ chemosignal. Bron- son and Maruniak [5] have reported that in addition to chemical stimuli, tactile cues also play some role in male-induced acceleration in young females. In an attempt to isolate the active fraction of urine accelerating puberty, it was shown that the substance is androgen dependent, heat labile and apparently associated with the protein fraction of the urine [15]. Jemiolo et al. [16] have recently synthesized and tested the analogs (2-(sec-butyl)-4, 5-dihydrothiazole and de- hydro-exobrevicomin) of the male urinary factors involved in the Whitten effect in mice. How the acceleratory effect of male urine is suppressed by 156 S.C. PANDEY AND S. D. PANDEY female urine is not clear; the finding clearly indicate the high potency of female chemosignals in regulating puberty in juvenile females. It is to emphasize here that, though the male chemosignal is not effective over delaying substance of female origin, the former evokes the puberty accelerating process in prepubertal females by comparatively shorter exposure (Table 2). Stimulatory and inhibitory pheromonal in- fluences on puberty in juveniles living in popula- tion can not be examined separately because they are exposed to urine deposited by both males and females. One signal could overwhelm the other or two could be balanced in their action. Drickamer [17] reported that the inhibitory effect of the urine from grouped females takes precedence over urine source(s) that accelerates puberty. The sensitivity of juvenile females to the delay chemosignal from grouped females even in the presence of accelera- tory signals from the males suggests that retarda- tion of puberty may play an important role in modulating population growth. Massey and Van- denbergh [18] while working on natural popula- tions of mice, found that urine collected from females in dense population delayed puberty in test females; urine collected from females in sparse population failed to retard pubertal onset. Drickamer [19] has reported that 4-7 days of treatment with urine containing the delay chemo- signal is required for puberty delay to occur in laboratory mice. The wild Mus varies with its laboratory cousin as it requires a longer treatment, at least, of 9 days for puberty retardation. Howev- er, the puberty accelerating chemosignal of male origin is effective only by 6 days treatment. It seems that at low population density (when de- laying chemosignal does not operate) male chemo- signal accelerates the pubertal onset in juvenile females and promotes the population growth. When the density is increased, the females start retarding the sexual maturation in juvenile females. This phenomenon does provide a natural force for dispersal of individuals in a natural population. The shorter exposure time to male chemosignal for bringing the effect facilitates the quick propagation of the species. ACKNOWLEDGMENT Authors are grateful to the Department of Science and Technology, India for financial support and Council of Scientific and Industrial Research, New Delhi for a Senior Research Fellowship to SCP. REFERENCES 1 Stone, C. P. and Barker, R. C. (1940) Change of the age of puberty in albino rats by selective mating. Proc. Soc. Exp. Biol. Med., 44: 48-50. 2 WVandenbergh, J. G. (1983) Social factors controll- ing puberty in the female mouse. In “Hormones and Behaviour in Higher Vertebrates”. Ed. by J. Balth- azart, E. Prove and R. Gilles, Springer-Verlag, Berlin, pp. 342-349. 3 Vandenbergh, J. G. (1967) Effect of the presence of a male on the sexual maturation of female mice. Endocrinology, 81: 345-349. 4 Vandenbergh, J. G. (1969) Male odour accelerates female sexual maturation in mice. Endocrinology, 84: 658-660. 5 Bronson, F. H. and Maruniak, J. A. (1975) Male- induced puberty in female mice : evidence for a synergistic action of social cues. Biol. Reprod., 13: 94-98. 6 Cowley, J.J. and Wise, D. R. (1972) Some effects of mouse urine on neonatal growth and reproduc- tion. Anim. Behav., 20: 499-506. 7 Colby, D. R. and Vandenbergh, J. G. (1974) Reg- ulatory effects of urinary pheromones on puberty in the mouse. Biol. Reprod., 11: 268-279. 8 Drickamer,L.C. and Murphy,R. X. (1978) Female mouse maturation: effects of excreted and bladder urine from juvenile and adult males. Dev. Psychobiol., 11: 63-72. 9 Vandenbergh, J.G., Drickamer, L. C. and Colby, D. R. (1972) Social and dietary factors in the sexual maturation of female mice. J. Reprod. Fertil., 28: 397-405. 10 McIntosh, T. K. and Drickamer, L. C. (1977) Ex- creted urine, bladder urine and the delay of sexual maturation in female house mice. Anim. Behav., 25: 999-1004. 11 Pandey,S.C. and Pandey, S.D. (1986) Stimulus exposure time and period of grouping of donors required for the release of pheromonal cues delaying puberty in young female wild mice. Zool. Sci., 3: 687-690. 12 Bronson, F.H., Dagg,C.P. and Snell, G. D. (1966) Reproduction. In “Biology of Laboratory Mouse”. Ed. by E. L. Green, McGraw Hill Book Co., New York, 2nd ed., pp. 187-204. 13 Rugh, R. (1968) The Mouse; its Reproduction and 14 15 16 Urinary Chemosignals and Sexual Maturation Development. Burgess Publ. Co., Minneapolis. Drickamer, L.C. (1974) Sexual maturation of female house mice: Social inhibition. Dev. Psycho- biol. , 7: 257-265. Vandenbergh, J. G., Whitsett, J. M. and Lombardi, J.R. (1975) Partial isolation of a pheromone accelerating puberty in female mice. J. Reprod. Fertil., 43: 515-523. Jemiolo, B., Harvey, S. and Novotny, M. (1986). Promotion of the Whitten effect in female mice by synthetic analogs of male mouse urinary consti- tuents. Proc. Natl. Acad. Sci. USA., 83: 4576-4579. 17 18 19 157 Drickamer, L. C. (1982) Acceleration and delay of first vaginal oestrus in female mice by urinary chemosignals: dose levels and mixing urine treat- ment sources. Anim. Behav., 30: 456-460. Massey, A. and Vandenbergh, J. G. (1980) Puberty delay by a urinary cue from female house mice in feral populations. Science, 209: 821-822. Drickamer, L. C. (1977) Delay of sexual maturation in female house mice by exposure to grouped females or urine from grouped females. J. Reprod. Fertil., 51: 77-81. > ZOOLOGICAL SCIENCE 5: 159-164 (1988) © 1988 Zoological Society of Japan Sexual Interference in the Alpine Newt, 7riturus alpestris (Amphibia, Urodela, Salamandridae) Pau A. VERRELL! Department of Biology, The Open University, Milton Keynes, U.K. ABSTRACT—This paper describes sexual interference (a form of intermale competition) in the alpine newt, Triturus alpestris. When a male encounters a female already engaged in courtship, he may interfere with the courting male’s attempts to inseminate her. The rival displays to the female and may inseminate her himself. The courting male may respond to sexual interference by displaying to the rival and leading him away from the female; this behavior can be interpreted as sexual defense. Females can be multiply inseminated as a consequence of sexual interference, although they frequent- ly flee from competitively interacting males. Sexual interference in T. alpestris appears to be less complex than that of the congeneric smooth newt, T. vulgaris. INTRODUCTION Competition between males for access to females is ubiquitous in the animal kingdom, and is an important selective agent in the evolution of both morphological and behavioral characters of males [1, 2]. The nature of such competitive interactions within any one species depends, at least in part, on its mode of fertilization. In the majority of urodele amphibians, fertilization is internal but sperm transfer is indirect; a sperma- tophore is usually deposited on the substrate during the latter stages of courtship [3]. In many urodele species, the male does not physically sequester the female during the spermatophore deposition and transfer stage of courtship, and the pair are thus susceptible to attention from other males at this time. If a male encounters a pair already engaged in courtship, he may interfere with the efforts of the courting male to inseminate the female. Such sexual interference has been described in a number of urodele species [4-7] and, in many, the interfering male mimics be- havior normally shown by a sexually responsive Accepted July 21, 1987 Received June 27, 1987 ' Present address: Allee Laboratory of Animal Be- havior, Department of Biology, University of Chica- go, 940 East 57th Street, Chicago, Illinois 60637, U.S.A. female as the courting male initiates the latter stages of courtship. The interferer may, for example, stimulate the tail of the courter tactually, causing the latter to deposit a spermatophore which will not be picked up by a female. The interferer may then deposit a spermatophore of his own, with which the female can become insemi- nated. By adopting this strategy of sexual interfer- ence by female mimicry, the interfering male may obtain an insemination without first investing in a period of courtship display [4-7]. Newts of the European genus Triturus exhibit courtship behavior in which there is very little physical contact between partners [8]. In the smooth newt, T. vulgaris, sexual interference by female mimicry is commonly seen between males in the laboratory [7]. It also occurs in natural populations during those parts of the breeding season when sexually responsive females are fewest in number [9, 10]. Field observations indicate that similar intermale behavior occurs in the crested newt, 7. cristatus [11; Verrell, unpubl. data]. Sexual interference in the alpine newt, T. alpestris, is described in the present paper, and is compared with that exhibited by congeneric newts. MATERIALS AND METHODS The alpine newts used in this study were taken from two introduced populations in southern 160 P. A. VERRELL England (one in Sussex, the other in North- amptonshire). The species occurs naturally in the northwestern part of mainland Europe [12]. Five males and five females were obtained from each population, and all of these newts remained in breeding condition during the course of this study (i.e. the males had well-developed secondary sexual characters and the females were gravid with eggs). In the laboratory, the newts were maintained in single-sex aquaria measuring 32X30 x30cm. Food was provided ad libitum, and consisted of Tubifex and chopped earthworms (Lumbricus). The water in these aquaria was neither filtered nor aerated, and ranged in temperature from 15 to 20°C. The photoperiod to which the newts were exposed was made to track the natural (Milton Keynes) light-dark cycle. All observations were made during spring 1986 in an aquarium measuring 32 X30 x 30 cm, floored with gravel overlaid with fine sand. The water in the aquarium was neither filtered nor aerated, and ranged in temperature from 18 to 22°C. A female and two males were placed in the aquarium at approximately 17:00 hr, and all in- teractions between the three individuals were recorded on videotape (Panasonic video tape recorder VTR-NV-8030) for a continuous period of about 14 hr, using a timelapse facility (recording interval 0.18 sec, as against 0.02 sec for normal speed). A total of 15 trios were recorded in this way, the trios consisting of unique combinations of individuals drawn from the available pool of newts. In addition, five pairs were placed together and their behavior recorded (as described above) in order to gain a first-hand impression of the normative courtship of this species. RESULTS Sexual behavior between single males and females A brief account is given, for comparison, of the sexual behavior of single male and female pairs of T. alpestris (for more detailed descriptions, see [8, 13-15]). The courtship of 7. alpestris differs markedly from that of congeneric newts. After a period of “orientation”, during which the male attempts to assume a position in front of the female, there follows a long period of “static display”. During this period, the male remains stationary in front of the female and “fans” his tail towards her. Fan- ning is a single, stereotyped tail movement which produces water currents which probably stimulate the female tactually and olfactorily, carrying secre- tions produced by glands in the cloaca of the male. Provided the female does not flee during static display, the male then turns away from her and initiates spermatophore deposition and transfer behavior. The male “creeps” in front of the female, quivering his tail, and after a few seconds, deposits a spermatophore on the substrate in front of her. Although the female often nudges at the base of the male’s tail with her snout to elicit spermatophore deposition, such tactile stimulation seems less necessary in T. alpestris than it is in other newts (Verrell, personal observation). The male then “creeps-on” away from the female for a distance of about one body-length and turns to block her path as she follows him. This ends the first sequence of the courtship encounter. The sperm mass may or may not be picked up in the cloaca of the female at this time. The male may then creep again, thus initiating a second sequence of spermatophore deposition and transfer be- havior. This usually occurs without the male reverting to an intervening period of display. As many as three spermatophores may be deposited during a single courtship encounter (i.e., three sequences may be completed). Sexual interference Fifty three discrete sexual encounters involving interactions between males were extracted from the videotape records for the 15 trios whose behavior was recorded (mean+1SD number of encounters per trio was 3.5+1.7). In all of these encounters, one male, hereafter called “the court- er”, began to court the female. The other male was thus given the status of “the rival”. Close inspection of the videotape records revealed seven basic behavior patterns which occur in the context of competitive sexual encounters. These are as follows: 1) Rival approaches pair: the rival moves Sexual Interference in the Alpine Newt 161 towards a male and female already engaged in courtship. His first response on contact is usually to nudge either or both of the courting newts with his snout. 2) Rival displays to female: the rival fans his tail in the direction of the female. 3) Rival displays to courter: the rival fans his tail in the direction of the courter; either of the males may initiate display, which may be mutual. 4) Rival creeps: by creeping in front of the female, the rival initiates spermatophore deposi- tion and transfer behavior. 5) Female stays: the female remains with the rival, following him as he moves into creep-on. 6) Female leaves: the female moves away from the two males. Her action appears to be deliber- ate, and she remains on the substrate some distance from the males. Moving away of this type is not immediately followed by the female ascend- ing to the water surface in order to breathe. 7) Males leave: the courter and rival move away from the female, fanning towards one another as they go. Figure 1 summarizes the frequency with which rival males were observed to proceed from one stage of a competitive sexual encounter to another. Of the 53 approaches made by rivals towards courting pairs, 40 (75.5%) were made after the female had completed one prior courtship se- quence with the courter, 9 (17%) after she had completed two sequences and 4 (7.5%) after the al rival approaches pair 7 |: 3 rival displays ae rival displays 2 to to po fs > Fival female J males leave i | ee female stays Fic. 1. Sequence diagram showing the frequency of behavioral transitions during competitive sexual en- counters between trios consisting of two male and one female Triturus alpestris. See text for further information and for detailed descriptions of the behavior patterns involved. completion of three prior sequences. Forty out of the 53 (75.5%) encounters terminated when either the female or males left after a period of fanning display. In only 11 (21%) encounters did the rival male creep in front of the female, and in only 4 of these did the female remain with the rival during creep-on. In none of these encounters did the female nudge the rival’s tail before he entered creep-on, suggesting that if he deposited a sper- matophore, he did so in the absence of tactile stimulation. Unfortunately, the optical resolution of the video equipment used in this study was not sufficiently high to record the presence or absence of a spermatophore on the substrate. Therefore, it is concluded that only 4 (7.5%) of the 53 competi- tive sexual encounters observed potentially could have resulted in the insemination of the female by the rival male. The rival appeared to assess the behavior of the courter during an encounter, as judged by the behavior he exhibited towards a courting pair. Consider the courtship behavior of 7. alpestris as consisting of two phases: static display and sper- matophore deposition and transfer (creep and creep-on). There are thus two types of behavior available to each male in a trio. In 31 encounters in which the courter was displaying, the rival responded with his own display in 28 (90%) and creeping in front of the female in 3 (10%). In 22 encounters in which the courter was creeping, the rival also crept in 8 (36%) and displayed in 14 (64%). The rival male was significantly more likely to display himself when the courter was also displaying (¥?=20.16, P<0.001), but not more likely to creep himself when the courter was also creeping (¥°=1.64, P>0.1). Sexual defense In the face of sexual interference of the form described above, the courting male is expected to retaliate with behavior patterns whose function is to defend the female [4, 5]. Subtle alterations in the temporal patterning of the courter’s behavior were not monitored during this study, although they may play an improtant role in sexual defense [7]. More obvious behavioral responses which might function in the context of sexual defense were observed. First, 28% of all competitive 162 P. A. VERRELL encounters terminated when the courter and rival moved away from the female, displaying to one another (see Fig. 1). Of the 15 encounters in which this occurred, the courter initiated the move in 11 (73%) cases. Secondly, during one encounter, the courter appeared to lead the female away from the rival by moving backwards away from her as he displayed: she followed him. This action closely resembled “retreat display” seen in the smooth newt (see Discussion). DISCUSSION When a male alpine newt encounters a female already engaged in courtship, he may interfere with the efforts of the courting male to inseminate his partner. As shown in Figure 1, the rival usually initiates such interference by displaying to the female, which may then lead him to display to the courter. Most competitive sexual encounters break down at this point, either because the female moves away from the males or because the males leave the female. If the female remains close to the rival as he displays, he may initiate spermato- phore deposition and transfer behavior. From a position close to the female, the rival creeps, apparently waiting for her to touch his tail with her snout. Even if such stimulation is not provided, a spermatophore may be deposited. Because sper- matophores could not be seen on the substrate, definite instances of deposition and pick up could not be determined. However, as judged by the behavior of the rival, it is concluded that the probability of insemination by the rival as a consequence of sexual interference was no greater than 7.5%. Sexual interference in T. alpestris can thus be considered a “side-payment” conditional mating strategy [16-18]. Individuals pursue a primary, high-gain strategy of courtship, but will accept lower gains by adopting the subsidiary strategy of sexual interference should opportuni- ties for the latter arise (see also [7]). Female 7. alpestris had always completed at least one sequence of courtship with the courter prior to the approach of the rival. It is thus likely that sexual interference will sometimes lead to the insemination of the female by both of the males attending her (first the courter, then the rival). Rafinski [19] found that a high proportion (17 out of 18) of female alpine newts collected in the field laid clutches which showed multiple paternity (as determined by electrophoretic analysis). Multiple insemination due to sexual interference may be responsible for at least some instances of multiple paternity in this species. Multiple insemination may also lead to sperm competition, another manifestation of competition between males. Sperm competition has been demonstrated only in one species of plethodontid salamander [20], but is likely to occur in many species, including newts [21]. Sexual interference in T. alpestris differs markedly from that observed in the smooth newt, T. vulgaris, the only congener for which detailed descriptive data are available. In the latter species, a rival male seldom expends time and energy in displaying to a female already engaged in courtship [7], and usually interferes just as the courter initiates creep. By interposing himself between the partners, the rival nudges the courter’s tail (female mimicry), causing him to deposit a profitless spermatophore, and then creeps himself [7]. In T. alpestris, most competitive sexual encounters be- gin with the rival displaying to the female (Fig. 1). If the encounter continues long enough for the courter to creep, the rival does not mimic female behavior by nudging his tail before creeping himself. These two major differences suggest that the sexual interference behavior of T. vulgaris is more derived than that of T. alpestris; the former involves less expenditure of energy by the rival and occurs at a time when the courter is “locked” into the relatively stereotyped behavior which follows spermatophore deposition [7, 22]. In addition, the observed probability of successful insemination by a rival is 20% in T. vulgaris [7], compared with a lower predicted maximum of 7.5% for T. alpestris. Another reason why the sexual interference behavior of T. alpestris seems less derived than that of 7. vulgaris concerns the apparent absence of well-defined sexually defensive behavior in the former species. Such behavior functions to defend the courting male against the deleterious effects of interference [4, 5]; interference and defense are expected to coevolve as a competitive arms race [23]. In 7. vulgaris, the most obvious response of a Sexual Interference in the Alpine Newt 163 courting male to the threat of interference is to increase the duration of the “retreat display” stage of courtship. Verrell [7] interpreted this increase as an attempt to draw the female away from the rival male. Retreat display is absent from the repertoire of T. alpestris, and aside from one instance of behavior that could be interpreted as leading, the only other behavior that could be considered as defensive in function is male-male display. As discussed above, most instances of male-male display were initiated by the courter and resulted in both males moving away from the female. In the laboratory, this ended the courtship encounter, but in the field the courter may return to the female having driven the rival away. Field observations of mating behavior in natural alpine newt populations are needed to test this hypoth- esis. One important similarity between T. alpestris and T. vulgaris concerns the response of the courting female to sexual interference. In the latter species, the female often flees as soon as the rival approaches and is thus lost to both males [7]. This seems to be as much a function of the density of males in her vicinity as of the behavior of those males [24]. A similar aversion is apparent in female T. alpestris; in only 11 (20.75%) of the 53 encounters did the female stay long enough for the rival to creep, and in no cases did she stay long enough for the rival to creep more than once. Comparative analysis of the sexual behavior of newts in the genus Triturus led Halliday [8] to propose a tentative phylogeny of this taxon. The single tail display (fan) and absence of retreat display in T. alpestris suggest that its courtship is closer to that of the ancestral type than are the more derived courtship behavior patterns of other, well-studied species (7. vulgaris, T. helveticus and T. cristatus). I suggest that the rather unstructured nature of sexual interference in T. alpestris, as described in this paper, is further evidence that the behavior of this species is less derived than that of its congeners. Detailed studies of competitive interactions between male newts in other Triturus species are needed to provide a proper test of this hypothesis. ACKNOWLEDGMENTS This work was supported financially by grants from the Open University Research Committee and the National Science Foundation (BSR 8506766). I am most grateful to Trevor Beebee and Kenneth Blackwell for kindly lending me the alpine newts. I also thank Lynne Houck, Norah McCabe, Steve Arnold, Tim Halliday and Chris Raxworthy for helpful discussion and comments on the manuscript. REFERENCES 1 Darwin, C. (1871) The Descent of Man and Selec- tion in Relation to Sex. John Murray, London. 2 Wilson, E. O. (1975) Sociobiology: The New Synth- esis. Belknap Press, Harvard, Massachusetts. 3 Salthe,S.N. (1967) Courtship patterns and the phylogeny of the urodeles. Copeia, 1967: 100-117. 4 Arnold,S.J. (1976) Sexual behavior, sexual in- terference and sexual defense in the salamanders Ambystoma maculatum, Ambystoma tigrinum and Plethodon jordani. Z. Tierpsychol., 42: 247-300. 5 Arnold, S.J. (1977) The evolution of courtship behavior in New World salamanders with some comments on Old World salamandrids. In “The Reproductive Biology of Amphibians”. Ed. by D. H. Taylor and S.I. Guttman, Plenum Press, New York, pp. 141-183. 6 Verrell, P. A. (1983) The influence of the ambient sex ratio and intermale competition on the sexual behavior of the red-spotted newt, Notophthalmus viridescens (Amphibia: Urodela: Salamandridae). Behav. Ecol. Sociobiol., 13: 307-313. 7 Verrell, P. A. (1984) Sexual interference and sexual defense in the smooth newt, Triturus vulgaris (Amphibia, Urodela, Salamandridae). Z. Tier- psychol., 66: 242-254. 8 Halliday, T. R. (1977) The courtship of European newts. An evolutionary perspective. In “The Repro- ductive Biology of Amphibians”. Ed. by D.H. Taylor and S.I. Guttman, Plenum Press, New York, pp. 185-232. 9 Verrell, P. and Halliday, T. (1985) Reproductive dynamics of a population of smooth newts, Triturus vulgaris, in southern England. Herpetologica, 41: 386-395. 10 Verrell, P. A. and McCabe, N.R. Field observa- tions of the sexual behaviour of the smooth newt, Triturus vulgaris vulgaris (Amphibia: Salamandri- dae). J. Zool., London. (In press) 11 Zuiderwijk, A. and Sparreboom, M. (1986) Territo- rial behaviour in crested newt Triturus cristatus and marbled newt T. marmoratus (Amphibia, Urodela). Bij. Dierkunde, 56: 205-213. 12 15 16 17 18 164 Steward, J. W. (1969) The Tailed Amphibians of Europe. David and Charles, Newton Abbot, Eng- land. Finkler, W. (1923) Analytical studies on the factors causing sexual display in the mountain newt (Tritur- us alpestris). Proc. R. Soc. Lond., Ser. B, 95: 356- 364. Meissner, K., Rohler, E. and Rohler, L. (1983) Zur Balz des Bergmolches, Triturus alpestris: 1. Aqu. Terr., 6: 210-213. Meissner, K., Rohler, E. and Rohler, L. (1983) Zur Balz des Bergmolches, Triturus alpestris: 2. Aqu. Terr., 7: 245-248. Dawkins, R. (1980) Good strategy or evolutionarily stable strategy? In “Sociobiology: Beyond Nature/ Nurture?”. Ed. by G. W. Barlow and J. Silverman, Westview Press, Colorado, pp. 331-367. Davies, N. B. (1982) Alternative strategies and competition for scarce resources. In “Current Prob- lems in Sociobiology”. Ed. by King’s College Sociobiology Group, Cambridge Univ. Press, Cam- bridge, pp. 363-380. Dunbar, R. I. M. (1982) Intraspecific variations in 19 20 21 22 23 24 P. A. VERRELL mating strategy. In “Perspectives in Ethology, Vol. 5”. Ed. by P.P.G. Bateson and P. H. Klopfer, Plenum Press, New York, pp. 385-431. Rafinski, J. (1981) Multiple paternity in natural populations of the alpine newt, Triturus alpestris (Laur.). Amphibia-Reptilia, 2: 282. Houck, L. D., Tilley, S. G. and Arnold, S. J. (1985) Sperm competition in a plethodontid salamander: preliminary results. J. Herpetol., 19: 420-423. Halliday, T. R. and Verrell, P. A. (1984) Sperm competition in amphibians. In “Sperm Competition and the Evolution of Animal Mating Systems”. Ed. by R.L. Smith, Academic Press, New York, pp. 487-508. Halliday, T. R. (1974) Sexual behavior of the smooth newt, T. vulgaris. J. Herpetol., 8: 277-292. Dawkins, R. and Krebs, J. R. (1979) Arms races between and within species. Proc. R. Soc. Lond., Ser. B, 205: 489-511. Verrell, P. A. (1984) Responses to different densi- ties of males in the smooth newt, Triturus vulgaris: “one at a time, please”. J. Herpetol., 18: 482-484. ZOOLOGICAL SCIENCE 5: 165-178 (1988) © 1988 Zoological Society of Japan Land Hermit Crabs from the Ryukyus, Japan, with a Description of a New Species from the Philippines (Crustacea, Decapoda, Coenobitidae) YuKIO NAKASONE Biological Laboratory, College of Education, University of the Ryukyus, Okinawa 903-01, Japan ABSTRACT—Six species of land hermit crabs are now known from Japan. brevimanus and C. violascens are recorded from Japan for the first time. Of them Coenobita C. purpureus and C. violascens hitherto synonymized with C. perlatus and C. cavipes, respectively, are valid. These species are redescribed and discussed in more detail. C. pseudorugosus is described and illustrated as a new species on the basis of the specimens from the Philippines. INTRODUCTION In the Indo-West Pacific region, the genus Coenobita is represented by ten valid species [1- 6]: C. rugosus H. Milne Edwards, 1837; C. purpureus Stimpson, 1858; C. perlatus H. Milne Edwards, 1837; C. cavipes Stimpson, 1858; C. violascens Heller, 1862; C. brevimanus Dana, 1852; C. scaevola (Forskal, 1775); C. spinosus H. Milne Edwards, 1837; C. carnescens Dana, 1852; C. longitarsis De Man, 1902. Among these spe- cies, C. purpureus and C. violascens have hitherto been treated as the synonym of C. perlatus and C. cavipes, respectively. In our recent study, howev- er, comparison of the specimens from the Ryukyus reveals that they are valid species; Miyake [7] had already separated C. purpureus from C. perlatus. The specimens from the Ryukyus were collected from the Miyako and the Yaeyama Islands during the ecological and distributional studies of land hermit crabs in Okinawa Prefecture, except C. perlatus, although this species has also been reported from Kuroshima, the Yaeyama Islands by Miyake [7]. The specimens collected from Cebu I., the Philippines were also examined and revealed to belong to an undescribed species of the genus Accepted August 19, 1987 Received June 12, 1987 Coenobita. However, this species has been not reported from the Ryukyus. The objectives of the present paper are to provide information on land hermit crab species in Japan, and to resurrect the synonym of some species, as well as to describe a new species of Coenobita. Coenobita pseudorugosus n. sp. (Figs. 1A—H and 2) Material examined: Holotype, male (SL=Shield Length, 12.37mm). Paratypes, 15 males (SL= 7.29-12.13 mm), 22 non-ovigerous females (SL= 5.63-10.75 mm), Cebu I., the Philippines, Apr. 30, 1986, T. Higa leg. Diagnosis: Rostrum small and triangular. Ocu- lar acicle broad basally, triangular and terminating in a small spine. Antennular basal segment with very produced laminar portion proximally and vertical margin of its lamina making an obtuse angle with upper margin of segment. Palm of left cheliped with an oblique series of seven to ten up-standing laminar teeth on upper part of outer surface; lower margin of propodus nearly straight in distal half and not four-cornered in an external form. Outer surfaces of dactylus and propodus of left third leg flat, smooth and separated from dorsal surface by a well-marked longitudinal crest. Right coxa of fifth legs in male produced into an 166 Y. NAKASONE Fic. 1. Coenobita pseudorugosus n. sp., male, paratype (SL=11.3 mm). A, shield and some cephalic appen- dages, dorsal view; B, shield and antennal segments, lateral view; C, left cheliped; D, chela and carpus of right cheliped; E, left third leg, lateral view; F, basal segment of antennule; G, flagella of antenna; H, sternite and coxae of male fifth legs. Scale bars indicate 10 mm for A, C, E, 5 mm for B, D, H, and 2.5 mm for F and G. Six Known and a New Species of Land Hermit Crabs 167 elongate tube, always longer than left one. Description: Shield usually longer than broad, narrower anteriorly; anterior margin between ros- trum and lateral projections concave; rostrum small and triangular; dorsal surface with scattered granules on anterior and lateral portions; and lateral margins setose. Ocular acicle broad basally, triangular and terminating in a small spine. Ocular peduncle compressed, reaching nearly to two-thirds the length of ultimate segment of antennal peduncle. Antennular basal segment with very produced laminar portion proximally and vertical margin of its lamina making an obtuse angle with upper margin of segment; small flagellum of antennule reaching nearly to one-half length of large one. Antennal acicle fused with second segment of its peduncle. In left chelipeds (Fig. 2) palm with an oblique series of seven to ten up-standing laminar teeth on upper part of outer surface; lower margin of propodus nearly straight in distal half and not four-cornered in an external form; palm with scattered round granules in addition to oblique teeth on outer surface, numerous especially on its lower portions; both fingers also with numerous round granules on outer surfaces; inner surface of palm strongly elevated in middle part and covered with large scale-like tubercles; movable finger with corneous-tipped granules on inner surface. In Fic. 2. Coenobita pesudorugosus n. sp., male, para- type (SL=11.3 mm). Enlargement of chela of the same cheliped as Fig. 1C. Scale 5mm. small cheliped fingers and palm with corneous- tipped granules and setae on outer surfaces. In left third leg (Fig. 1E) outer surfaces of dactylus and propodus flat, smooth and separated from dorsal surface by a well-marked longitudinal crest; walking legs except left third leg with setae on lower margin of each segment. In male coxae of fifth legs of both sides produced ventrally, unequal, and right coxa produced into an elongate tube, always longer than left one; its tube turning to the left and curved ventrally. Coloration: Small individuals with a broad dark brown transverse band at anterior one-third of shield and two longitudinal stripes of the same color on posterior portion. Large individuals sometimes with two dark brown patches behind anterior margin of shield. Side walls of shield with a dark brown transverse band on anterior part. Ventral surface of ocular peduncle dark brown. Palm of left cheliped with a longitudinal white stripe on middle portion and the other part dark-brownish. Dactyli of left second and third legs each with a dark brown patch at proximal part. Propodus of left third leg with a broad dark brown band on middle portion; propodi of other legs with a white band at distal one-fourth, other area dark-brownish. Carpi of first and third legs with a longitudinal dark brown stripe; meri with dark brown ring distally. Distribution: Known only from the type locality. Remarks: C. pseudorugosus is most closely related to C. rugosus, but is distinguishable from the latter by the following characters. In C. pseudorugosus the lower margin of the propodus of the large cheliped is nearly straight for the distal half and the palm is, therefore, not four-cornered in an external form, while the palm of rugosus has an obtuse corner and is thus four-cornered in an external form; the palm of the present species is dark-brownish, lacking a distinct large patch of dark brown on the outer surface, but that of rugosus has a distinct large patch of dark brown. In all the specimens of the male, the right sexual tube is distinctly longer than the left one, while ina great number of the specimens of rugosus ex- amined, the right sexual tube is almost equal in length to the left, or the right is slightly longer than the left. 168 Y. NAKASONE Material examined: I have treated a large Coenobita rugosus H. Milne Edwards number of individuals during the ecological and (Figs. 3A-G and 9A) distributional studies of the species in Okinawa Coenobita rugosa H. Milne Edwards [2], p. 241. Prefecture. Coenobita rugosus: Alcock [9], p. 143, pl. 14, fig. 3, 3a; Distribution: Widely distributed in the Indo- Ball and Haig [15], p. 89; Miyake [7], p. 115, pl. 39, West Pacific region. In Japan this species is fig. 1; Yu [12], p. 61, pl. 1, fig. D. recorded from Chichijima and Anijima Islands in Wy OY NS Orc \* ee Sy Fic. 3. Coenobita rugosus H. Milne Edwards, male. A, shield and antennal segments, lateral view; B, chela and carpus of left cheliped; C, chela of right cheliped; D, basal segment of antennule; E, flagella of antenna; F, left third leg; G, sternite and coxae of male fifth legs. Scales 10 mm for B and F, 5 mm for A, C and G, 3 mm for D, 2.5 mm for E. Six Known and a New Species of Land Hermit Crabs 169 the Bonin Islands [19], Amami-Ohshima, Oki- noerabujima and Yoronjima Islands in the Amami Islands [20] and from each island of Okinawa Prefecture except Kitadaitojima Island [17, 18, PANE Remarks: This species is very abundant in Okinawa Prefecture. A great number of indi- viduals are transported from Okinawa to Japan mainland as “pets” together with C. purpureus and C. cavipes. It is known that the transportation have been started since 1934 [22]. Coenobita purpureus Stimpson (Figs. 4A—-F and 9B) Coenobita purpurea Stimpson [4], p. 83; Stimpson [23], p. 198. Coenobita perlata var. purpurea: Bouvier [24], pp. 148- 150. Coenobita purpureus: Miyake [7], p. 221. Material examined: 1 have examined a large number of individuals during the ecological and distributional studies of this species in Okinawa Island. Fic. 4. Coenobita purpureus Stimpson, male. A, shield and antennal segments, lateral view; B, chela and carpus of left cheliped; C, left third leg; D, basal segment of antennule; E, flagella of antennule; F, sternite and coxae of male fifth legs. Scales 10 mm for A-C and F, 5mm for D and E. 170 Y. NAKASONE Description: Shield narrower anteriorly, strong- ly swollen just behind front; dorsal surface with numerous scattered granules on anterior, posterior and lateral portions; lateral margins of shield with setae. Ocular peduncle compressed and reaching near- ly to median part of ultimate segment of antennal peduncle. Small flagellum of antennule reaching nearly to median part of large one. Antennal acicle fused with second segment of its peduncle. Palm of left cheliped with an oblique series of up-standing four to five laminar teeth on upper part of outer surface; upper portion of palm with numerous scattered granules other than oblique teeth on outer surface, but lower portion with small granules and nearly smooth; both fingers with scattered granules on outer surface. In left third leg outer surfaces of dactylus and propodus smooth, separated from dorsal surfaces by a well-marked longitudinal crest; outer surface of propodus not flat, but slightly swollen in midline. In male coxae of fifth legs of both sides produced ventrally, unequal and right coxa produced into an elongate tube, usually longer than left one; its tube turning to the left and curved ventrally. Coloration: Small individuals generally of cream color and large ones of purple color. Distribution: In Japan this species is recorded from Chichijima, Hahajima, Anijima, Hirajima, Mukaijima and Kitaiwojima Islands in the Bonin Islands [19], the Ohsumi Peninsula and Biro I. in southern Kyushu, Tanegajima, Yakujima, Naka- nojima and Takarajima Islands in the Tokara Islands, Amami-Ohshima, Kakeromajima, Kikai- jima, Tokunoshima, Okinoerabujima and Yoron- jima Islands in the Amami Islands [20], and from many islands in the Ryukyu Islands [17, 18, 21]. Remarks: The present species has been treated as the synonym of C. perlatus H. Milne Edwards by some authors since Henderson [25]. However, Bouvier [24] recognized this species as a variety of C. perlatus, but Terao [11] treated it as the synonym of C. rugosus. Miyake [7] resurrected C. purpureus from C. perlatus in the key to the Japanese species of Coenobita, but he did not give the colored illustration of the species. This species is easily separated from C. perlatus by the colora- tion, the shape of the palm of the large chela, the shape of dactylus and propodus of the left third leg, and by the shorter right sexual tube. This species grows to a large size as in C. perlatus. This species is very abundant in the Amami and the Okinawa Islands [17, 20, 21], but it is not so abundant in the southern Ryukyus [18]. I have treated and observed a great number of specimens during the ecological investigations of the present species [21]. C. purpureus is without doubt a valid species. Coenobita perlatus H. Milne Edwards (Figs. SA-F and 9C) Coenobita perlata H. Milne Edwards [2], p. 242. Coenobita perlata: Fize and Seréne [10], p. 24, fig. 3C, fig. 4; Yaldwyn and Wodzicki [13], p. 12. Coenobita perlatus: Alcock [9], p. 145, pl. 14, fig. 2, 2a; Miyake [7], p. 115, pl. 39, fig. 2; Haig [16], p. 124. Material examined: Male (SL=24.74mm), Kitaiwojima I., the Bonin Islands, Aug. 1986, Kimura Johnson leg. Distribution: Widely distributed in the Indo- West Pacific region. In Japan this species is known from Chichijima, Kitaiwojima and Minamitorishi- ma Islands in the Bonin Islands [19] and from Kuroshima Island in the Yaeyama Islands [7]. Remarks: This species was first reported from Japan by Miyake [7]. His specimen was a single male and collected from Kuroshima Island by Dr. Imafuku in 1979 (Miyake, pers. commun.), but the specimens of this species were not collected since 1979. I had an opportunity to examine a male specimen from Kitaiwojima Island. According to Yaldwyn and Wodzicki [13], the juvenile specimen of 8 mm in the carapace length is creamy-white in general color and had red bands on the carpi of the chelipeds and walking legs, but the juvenile specimens of the same size of C. purpureus have no red bands. The question is whether C. perlatus reported by De Haan [26] from Kagoshima (=Satsuma) and Ryukyu is a genuine perlatus species, because C. perlatus had never been found anywhere in Kagoshima and Okinawa Prefectures after 1979 [17, 18, 20]. Six Known and a New Species of Land Hermit Crabs 171 Fic. 5. Coenobita perlatus H. Milne Edwards, male. A, shield and antennal segments, lateral view; B, chela and carpus of left cheliped; C, left third leg; D, basal segment of antennule; E, flagella of antennule; F, sternite and coxae of male fifth legs. Scales 10 mm for A-D and F, 5mm for E. Coenobita cavipes Stimpson (Figs. 6A-F and 9D) Coenobita cavipes Stimpson [4], p. 245; Stimpson [23], p. 200. Coenobita cavipes: Alcock [9], p. 146, pl. 14, fig. 1; Fize and Seréne [10], p. 30, fig. 3B, fig. SA-C, pl. I, 4, 6. Material examined: 1 have examined a number of individuals during the ecological studies of this species. Distribution: Widely distributed in the Indo- West Pacific region. In Japan this species is known from Chichijima and Anijima Islands in the Bonin Islands [19], Okinoerabujima and Yoronjima Is- lands in the Amami Islands [20] and from many islands in the Ryukyu Islands except Akajima, Ikemajima and Nanbokudaitojima Islands [17, 18]. Remarks: This species has been confused with C. violascens Heller by some authors. The type locality was Loo Choo (Ryukyu) [4] and Bouvier 172 Y. NAKASONE Fic. 6. Ses oR Coenobita cavipes Stimpson, male. A, shield and antennal segments, lateral view; B, chela and carpus of left cheliped; C, left third leg; D, basal segment of antennule; E, flagella of antennule; F, sternite and coxae of male fifth legs. Scales 10 mm for A-C, 5mm for D-F. [24] and Fize and Seréne [10] gave one locality name “Chine” in the distribution. This name “Chine” is a dialect and is now called “Kin”, which is located in the central part of Okinawa mainland. Coenobita violascens Heller (Figs. 7A-F and 9B) Coenobita violascens Heller [6], p. 524; Heller [27], p. 82, pl. 7, fig. 1. Material examined: Two males (SL=4.16, 19.66mm), 3 females (SL=6.81-16.09 mm), Hosozaki, Kohamajima I., the Yaeyama Islands, Aug. 4, 1986, K.Shimamura leg.; female (SL=17.38 mm), IkemajimalI., the Miyako Is- lands, Sept. 12, 1986, M. Toyama leg.; male (SL=14.02 mm), 4 females (SL=14.19-19.14 mm), estuary of Shiira river, Iriomotejima I., the Yaeyama Islands, Jan. 16, 1987, Y. Nakasone leg. ; male (SL=13.43 mm), 3 females (SL=9.51-14.11 mm), Cebu I., the Philippines, Apr. 30, 1986, T. Higa leg. Description: Shield narrower anteriorly and slightly convex behind front; dorsal surface with scattered granules and punctations; tip of antero- lateral margin of shield produced into a spinule which is white in the distal half. Ocular acicle long and pointed. Ocular pedun- cle compressed, reaching almost to median part of ultimate segment of antennal peduncle. Small Six Known and a New Species of Land Hermit Crabs 173 Fic. 7. Coenobita violascens Heller, male. A, shield and antennal segments, lateral view; B, chela and carpus of left cheliped; C, left third leg; D, basal segment of antennule; E, flagella of antennule; F, sternite and coxae of male fifth legs. Scales 10 mm for A-C, 5mm for D-F. flagellum of antennule small, shorter and not reaching to basal part of aesthetasc pad of large flagellum. Antennal acicle fused with second segment of its peduncle. In left cheliped palm without an oblique series of up-standing teeth on upper part of outer surface; upper half portion of palm with numerous scat- tered granules, and lower half few granules, nearly smooth and with a distinct large patch of dark brown on outer surface; lower margin of palm straight or concave in middle portion and its proximal part (proximal lower margin near carpus) strongly produced into a lobe-like projection; both fingers violascent. In left third leg outer surface of propodus nearly smooth and separated from dorsal surface by a well-marked longitudinal crest; inner margin of propodus strongly projecting inwards, and con- cave; a longitudinal ridge on ventral surface of propodus very small, indistinct. Second and third legs with numerous tufts of long stiff setae. In male, coxae of fifth legs of both sides thick and short. A sternal protuberance between both coxae very small. Coloration: Whole body except abdomen violascent, but showing light lavender to dark violet by individuals. Distribution: Nicobar Islands; Cebu I., the 174 Y. NAKASONE Philippines. In Japan this species is known from Ikemajima Island in the Miyako Islands and Ishigakijima, Kohamajima, Taketomijima, Iriomotejima, Yonagunijima Islands in the Yaeyama Islands [17, 18]. Remarks: This species is distinguished from C. cavipes by having the straight or concave lower margin of the palm of the left chela, by the violascent fingers of the left chela, by having the strongly projecting inner margin and the conspic- uously concave inner surface of the left third leg propodus, by having a very small sternal protuber- ance between the coxae of the fifth legs of both sides, by the violascent body coloration, and by inhabiting mainly bay and estuary. In the juveniles the dactyli and propodi of the second and third legs are red brownish, and the ocular peduncle has a longitudinal dark brown stripe on the ventral surface and the other part is red brownish. Their juveniles are abundantly found near mangrove forest in Ishigakijima I., the Yaeyama Islands and they are easily distinguished from the juveniles of C. cavipes by the coloration of both the walking legs and the ocular peduncles and by the morpho- logical difference of the propodus of the left third leg. This species has been treated as the synonym of C. compressus and C. cavipes by Miers [28], De Man [5] and Fize and Seréne [10], but it seems that C. violascens is distinctly a separate species. The specimens from the Philippines are easily distin- guished from C. cavipes. Coenobita brevimanus Dana (Figs. 8A—G and 9F) Coenobita clypeata var. brevimanus Dana [3], p. 473; Dana [8], pl. 30, fig. 4b. Coenobita clypeata: Alcock [9], p. 142, pl. 15, fig. 1, la; Fize and Seréne [10], p. 7, fig. 1A-C, pl. I, 1. Coenobita hilgendorfi Terao [11], p. 338; Yu [12], p. 61, pl. 1, fig. C. Coenobita brevimanus: Rathbun [14], p.314; Ball and Haig [15], p. 88; Haig [16], p. 124. Coenobita brevimana: Yaldwyn and Wodzicki [13], p. ll. Material examined: Three males (SL=15.37- 18.15 mm), female (SL=15.89 mm), Ishigakijima I., the Yaeyama Islands, Dec. 4, 1985, Y. Naka- sone leg. Male (SL=21.99 mm), Ishigakijima I., Sept. 11, 1986, M. Toyama leg. Two males (SL=13.58, 17.95 mm), Ikemajima I., the Miyako Islands, July 30, 1986, H. Iraha leg. Female (SL=17.07 mm), ovigerous female (SL=16.42 mm). Ikemajima I., Sept. 12, 1986, M. Toyama leg. Distribution: Widely distributed in the Indo- West Pacific region. In Japan this species is now known from many islands in the Miyako and the Yaeyama Islands except Irabujima and Shimoji- jima Islands [17, 18]. Remarks: Until 1955, Coenobita clypeatus and C. hilgendorfi had been used as the scientific name of this species by many authors except some ones. Rathbun [14] used the name of C. brevimanus for specimens from the Indo-West Pacific for the first time. In Japan, some individuals of this species were for the first time found within a forest near seashore in Ishigakijima Island on 4 December, 1985 and also collected from Ikemajima Island. This species is now distributed only in the southern Ryukyus. The animals were active within the forest in the daytime and they produced sound such as “Kukku, Kukku...” when they are caught. They mainly used Turbo (Marmarostoma) argyro- stomus (Linné) as their host shell and one oviger- ous female was collected in early September. KEY TO THE SEVEN SPECIES OF COENOBITA I. Antennal acicle fused with second segment of its peduncle; ocular peduncle strongly com- pressed; a brush of hairs on inner upper margins of palms of both chelipeds. A. An oblique series of up-standing laminar teeth on upper part of outer surface of left palm; right coxa of fifth legs of male narrow- er than left one, tubular. a. Outer surface of propodus of left third leg separated from dorsal surface by a well- marked longitudinal crest. 1. Palm of left cheliped four-cornered in an external form; outer surface of propodus of left third leg flat; right coxa almost equal in length to left one, or slightly longer than 175 Six Known and a New Species of Land Hermit Crabs ) ¢ we y: 44 / Vi m t / i Wr} i /y Ky) j Ya iy | hy iy” i Coenobita brevimanus Dana, male. A, shield and antennal segments, lateral view; B, chela and carpus of left cheliped; C, left third leg; D, chela of right cheliped; E, basal segment of antennule; F, flagella of Fic. 8. antenna; G, sternite and coxae of male fifth legs. Scales 10 mm for A-E, 5 mm for F and G. 176 Y. NAKASONE Fic. 9. Comparison of external forms of chelae. A, Coenobita rugosus; B, C. purpureus; C, C. perlatus; D, C. cavipes; E, C. violascens; F, C. brevimanus. Scales 10 mm for A-F. LCEt cca sgonsiensmnatensonategareneseenseaanece rugosus Palm of left cheliped not four-cornered in an external form; outer surface of prop- odus of left third leg flat; right coxa produced into an elongate tube, always longer than left one ...pseudorugosus n. sp. Outer surface of propodus of left third leg not flat, slightly swollen in midline; coxae of fifth legs of both sides produced ventral- ly, unequal and right coxa produced into an elongate tube, usually longer than left ONE Ra ssteeensttar wereaianarteranue sos purpureus b. Outer surface of propodus of left third leg not separated from dorsal surface by a Six Known and a New Species of Land Hermit Crabs 17 longitudinal crest. 4. Outer surface of propodus of left third leg convex and with scattered white tubercles; right coxa produced into a long curved tUDEs teeconrereceecesh-ceteneeeuntesbeuensas perlatus B. No oblique series of up-standing laminar teeth on upper part of outer surface of left palm; coxae of fifth legs of both sides thick and short. 5. Lower margin of left palm with an obtuse corner in middle portion; inner margin of propodus of left third leg not strongly projecting inwards and inner surface almost flat; a longitudinal ridge on ventral surface of propodus well-developed, dis- tinct; a sternal protuberance between both coxae of fifth legs large ...............- cavipes 6. Lower margin of left palm straight or concave in middle portion; inner margin of propodus of left third leg strongly project- ing inwards and inner surface strongly concave; a longitudinal ridge on ventral surface of propodus very small, indistinct; a sternal protuberance between both coxae of fifth legs very small .............. violascens II. Antennal acicle not fused with second seg- ment of its peduncle; ocular peduncle not compressed; a brush of hairs on inner upper margin of palm of right cheliped only Piiaieaslela attaje Clauss)aia stelle aisieie eeiaieyeleiomicclnicinnes brevimanus ACKNOWLEDGMENTS I am very grateful to Dr. S. Shokita, Department of Marine Sciences, University of the Ryukyus, Mr. M. Toyama, Cultural Administration Section, Okinawa Prefectural Office of Education, and Mr. K. Shimamura, Yaeyama Senior High School, for providing specimens for the present study. I am indebted to Dr. S. Miyake, Emeritus Professor of Kyushu University, for valuable advice and encouragement. Most of this work was made by a support of Agency for Culture Affairs, the Ministry of Education, Science and Culture of Japan. REFERENCES 1 Lewinsohn, Ch. (1969) Die Anomuren des Roten Meeres (Crustacea Decapoda: Paguridea, Galatheidea, Hippidea). Zoologische Verhandeling- 10 11 13 14 en, Leiden, 104: 1-213, pls. 1-2. Milne Edwards, H. (1837) Histoire naturelle des Crustacés, comprenant l’anatomie, la physiologie et la classification de ces animaux, Vol. 2. Roret, Paris, pp. 1-531. Dana, J. D. (1852) Crustacea. United States Ex- ploring Expedition during the years 1838, 1839, 1840, 1841, 1842..., Vol. 13, Part 1. C. Sherman, Philadelphia, pp. 1-685. Stimpson, W. (1858) Prodromus descriptionis ani- malium evertebratorum...VII. Crustacea Anomura. Proc. Acad. nat. Sci. Philadelphia, 10: 225-252. Man, J.G. De (1902) Die von Herrn Professor Kikenthal im Indischen Archipel gesammelten De- capoden und Stomatopoden. Abh. Senckenb. naturf. Ges., 25: 467-929, pls. 19-27. Heller, C. (1862) Neue Crustaceen gesammelt wahrend der Weltumsegelung der K. K. Fregatte Novara. Zweiter vorlaufiger Bericht. Verhandl. k. k. zool. -bot. Gesellsch. Wien, 12: 519-528. Miyake, S. (1982) Japanese Crustacean Decapods and Stomatopods in Color. Vol. I. Macrura, Ano- mura and Stomatopoda. Hoikusha Publishing Co., Ltd., Osaka, pp.1-261. (In Japanese). Dana, J. D. (1855) Crustacea, Atlas. United States Exploring Expedition during the years 1838, 1839, 1840, 1841, 1842...., Vol. 14. C. Sherman, Phil- adelphia, pp. 1-27, pls. 1-96. Alcock, A. (1905) Catalogue of the Indian Decapod Crustacea in the Collection of the Indian Museum. Part I. Anomura. Fasciculus I. Pagurides. Indian Museum, Calcutta. pp. i-xi, 1-197, pls. 1-16. Fize, A. and Seréne, R. (1955) Les Pagures du Viétnam. Institut Océanographique, Nhatrang, note 45. pp. i-ix, 1-228, pls. 1-6. Terao, A. (1913) A catalogue of hermit-crabs found in Japan (Paguridea excluding Lithodidae), with descriptions of four new species. Annot. Zool. Japon., 88: 355-391. Yu, H. P. (1985) Notes on the land hermit-crabs (Crustacea, Decapoda, Coenobitidae) from Lan-yu Island in the Southern Taiwan. J. Taiwan Museum, 38: 59-64, pl. 1. Yaldwyn, J.C. and Wodzicki, K. (1979) System- atics and ecology of the land crabs (Decapoda: Coenobitidae, Grapsidae and Gecarcinidae) of the Tokelau Islands, Central Pacific. Atoll Res. Bull., 235: 1-53, figs. 1-6. Rathbun, M. J. (1910) Decapod crustaceans col- lected in Dutch East India and elsewhere by Mr. Thomas Barbour in 1906-1907. Bull. Mus. comp. Zool., Harvard, 52: 305-317, pls. 1-6. Ball, E. E. Jr. and Haig, J. (1972) Hermit crabs from Eastern New Guinea. Pacific Science, 26: 87— 107. Haig, J. (1984) Land and freshwater crabs of the 19 20 Zt 178 Seychelles and neighbouring islands. In “Biogeogra- phy and Ecology of the Seychelles Islands”. Ed. by D. R. Stoddart, Dr W. Junk Publishers, pp. 123- 139. Toyama, M. and Kurozumi, T. (1987) Geographical distribution of the genus Coenobita in Okinawa Prefecture. In “A Report on the Distribution and Ecology of Land Hermit Crabs in Okinawa Prefec- ture”. Okinawa Prefectural School Board, pp. 200- 203. (In Japanese). Shimamura, K. (1987) Ecological studies of land hermit crabs in the Yaeyama Islands. In “A Report on the Distribution and Ecology of Land Hermit Crabs in Okinawa Prefecture.” Okinawa Prefectural School Board, pp.61-118. (In Japanese). Suganuma, H., Tachikawa,H. and Masuda, M. (1987) Report on the habitats of the land hermit crabs in the Bonin Islands. Tokyo Metropolitan School Board, pp. 1-98. (In Japanese). Saisho, T. and Suzuki, H. (1987) An urgent study on the distribution and ecology of land hermit crabs, genus Coenobita, in Kagoshima Prefecture. Kagoshima Prefectural School Board, pp. 1-64. (In Japanese). Nakasone, Y. (1987) Ecological studies of land hermit crabs in the southern part of Okinawa Island. In “A Report on the Distribution and Ecology of Land Hermit Crabs in Okinawa Prefecture”. Okina- wa Prefectural School Board, pp. 16-60. (In Japanese). Y. NAKASONE 22 23 24 25 26 27 28 Toyama, M. (1987) Gathering land hermit crabs as resources. In “A Report on the Distribution and Ecology of Land Hermit Crabs in Okinawa Prefec- ture”. Okinawa Prefectural School Board, pp. 219- 224. (In Japanese). Stimpson, W. (1907) Report on the Crustacea (Brachyura and Anomura) collected by the North Pacific Exploring Expedition, 1853-1856. Smithson. misc. Collns., 49: 1-240, pls. 1-26. Bouvier, E. L. (1890) Révision des Cénobites du Muséum. Bull. Soc. philom. Paris, 2: 143-150. Henderson, J. R. (1888) Report on the Anomura collected by H. M.S. Challenger during the years 1873-76. Report on the scientific results of the voyage of H.M.S. Challenger during the years 1873-76... Zoology, Vol. 27. pp. i-ix, 1-221, pls. 1- 21. Haan, W. De (1849) Crustacea. In “Fauna Japoni- ca”. Ed. by P. F. von Siebold, Lugduni Batavorum, pp. 197-243. Heller, C. (1865) Crustaceen. In “Reise der 6ster- reichischen Fregatte “Novara” um die Erde, in den Jahren 1857, 1858, 1859, unter den Befehlen des Commodors B. von Wiillerstorf-Urbair, Zool., Vol. 2. pp. 1-280, pls. 1-25. Miers, E. J. (1880) Crustacea Anomura and Mac- rura (except Penaeidea). On a collection of Crus- tacea from the Malaysian Region. Part III. Ann. Mag. nat. Hist., (5) 5: 370-384. pls. 14, 15. ZOOLOGICAL SCIENCE 5: 179-182 (1988) [COMMUNICATION] © 1988 Zoological Society of Japan Direct Evidence for Axopodial Fusion Preceding Cell-to-cell Contact in a Heliozoan Echinosphaerium Takako Nisut!, Makoto KospayasHr’, Mami Isomura, Hipeki IsHipa* and YOSHINOBU SHIGENAKA Laboratories of Physiology and Cell Biology, Faculty of Integrated Arts and Sciences, Hiroshima University, Hiroshima 730, Japan ABSTRACT—In a heliozoan Echinosphaerium, the electrical interactions between two organisms during the process of cell fusion were studied physiologically. The electron-microscopical observations were also made to examine the connection of axopodia of the two organ- isms. Results showed that the two organisms interacted electrically through their axopodia before the two cell bodies began to fuse. INTRODUCTION The large heliozoan, Echinosphaerium, fre- quently shows a characteristic type of cell fusion, termed plasmogamy, which is not preceded by the encystation and thus considered as an asexual phenomenon [1-3]. Vollet and Roth [3] have observed that the cell fusion can be induced only if the treatments are used causing large amounts of new cell-surface membrane to be formed in the presence of divalent cations. In a Japanese heliozoan strain Echinosphaerium akamae, the cell fusion was found to occur between two approaching organisms even at nor- mal cultural conditions [4]. The fusion process has then been studied in relation to the axopodial degradation [5] and the cell membrane fluidity [6]. The timing and process of the fusion of cell Accepted August 6, 1987 Received July 3, 1987 ' Present address: Department of Physiology, School of Medicine, Kagoshima University, Kagoshima 890, Japan. * Present address: Department of Biology, Faculty of Science, Shimane University, Matsue 690, Japan. > To whom reprints should be requested. membranes, however, have not been investigated in detail. Electrophysiologically, the activities of proto- zoan cell membranes have been studied in many species in relation to the excitability, conduction and the movement of cilia or flagella [7, 8], but very few investigations have been done on the membrane characteristics during the cell fusion of two organisms. In the present study, the electrical connections between two organisms during the process of cell fusion have been pursued physiologically, and the electron-microscopical observations have also been made especially on the axopodia of just- approaching organisms. Both physiological and morphological studies have shown that the axopo- dial fusion precedes the fusion of cell bodies themselves. MATERIALS AND METHODS The large multinucleated heliozoan, Echino- sphaerium akamae [9], was cultured at 20+1°C in 0.01% Knop_ solution containing 0.24mM Ca(NO3)o, 0.14 mM KNO3, 0.058 mM MgSO, and 0.11 mM KH,PO,. One or two grains of boiled wheat were added to each 15 ml of the culture medium. Small ciliates and flagellates such as Tetrahymena, Colpidium and Chilomonas were also added to the medium as food. Subculturing was carried out at 2-week intervals. Prior to physiological experiments, the organ- isms were transferred from the culture medium to 180 T. Niso1, M. Kopayasuli et al. the physiological saline solution in an experimental chamber. The saline solution was composed of 1 mM KCl, 1 mM CaCl, 1 mM MgCl, and 5 mM tris (hydroxymethyl) aminomethane, the pH being ad- justed to 7.2 by HCl. The experimental chamber was set up on the stage of an inverted microscope through which observation and photographing of the organisms were made. When a pair of organisms came close to each other, a microelec- trode, filled with 3 M KCI and having resistances of 15 to 30 MQ, was inserted into each organism for recording membrane potential. The heliozoan Echinosphaerium showed a re- markable hyperpolarization, 15 to 20 mV in ampli- tude and several hundreds msec in duration, correlated with the contraction of contractile vacuole, termed H-CV [10]. The electrotonic spread of H-CV as well as of an electric potential caused by a current injection into one of the paired organisms was employed for determining of the degree of electric coupling between the two organisms. The electrical data were stored in a data recorder (Sony, DFR 3515) and displayed on an ink-writing recorder (Nihon-Kohden, RJG 4024). For electron microscopy, the two just- approaching organisms were fixed with glutaral- dehyde and osmium tetroxide fixatives by the method of Shigenaka et al. [11]. The fixed samples were then rinsed briefly, dehydrated, and embed- ded into a low viscosity embedding medium [12]. Ultra-thin sections were prepared with a glass knife loaded onto a Porter-Blum ultra-microtome (type MT-1) and stained with 3% uranyl acetate in 50% ethanol for 10 min and Reynolds’ lead citrate stain [13] for 3 min. The sections were observed with a transmission electron microscope (JEOL, JEM-100S) operating at 80 or 100 kV and photo- graphed using electron-microscopic films (Fuji, type FG). RESULTS AND DISCUSSION According to Shigenaka and Kaneda [5], the process of cell fusion in heliozoans can be divided into the following four stages; stage A before the tip-to-tip contact of axopodia, stage B until the contact of axopodial tip to the partner’s cell Fic. 1. Two organisms of Echinosphaerium akamae and records of their membrane potential. Upper photograph shows a pair of organisms at the early stage C of cell fusion. Arrows indicate the contrac- tile vacuoles. A and B in the lower tracings show the membrane potential recorded from the indi- viduals A and B of the upper photograph, respec- tively. An arrow shows the onset of H-CV. surface, stage C until the cell-to-cell contact, and stage D until the completion of fusion of the cell bodies. In a pair of just-approaching organisms at the stage A, any electrical signals in one organism were not detected in its partner. At the stage B, the electrotonic potential elicited by current ap- plication or H-CV in one organism was also recorded in the paired cell. On the contrary, action potentials, which developed spontaneously or by current injection, rarely conducted to the partner even at the late stage B. At the stage C, however, the conducted action potentials origin- ated in one organism could also be recorded in the partner’s cell membrane. Upper photograph of Figure 1 shows a pair of organisms at the early stage C, in which some axopodia touched to the partner’s cell surface while the two cell bodies are still about 120 ~m apart from each other. A and B in the lower Axopodial Fusion in Heliozoa 181 tracings of Figure 1 show the membrane potential recorded from the organisms A and B of the upper photograph, respectively. It is obvious that an H- CV produced in organism B was electrotonically spread to A. The coupling ratio, which is a ratio of the amplitude of spread H-CV in organism A to that of H-CV in B, was about 0.2. An action potential appeared after H-CV and spontaneous action potentials conducted or not case by case. In the cases shown at the right in Figure 1, a spontaneous action potential elicited in A con- ducted to B. The conduction velocity, calculated from the distance between two electrodes divided by the time interval between two action potentials, was about 3 mm/sec. On the contrary, an action potential elicited in B did not conduct to A with leaving a slow small depolarization. Summarizing these electro-physiological results, the coupling ratio was 0.1 to 0.3 at the stage B and early stage C, 0.4 to 0.5 at the late stage C, and 1.0 at the stage D. The ratio 1.0 means that the two cells are isopotential. At the late stage C and stage D, almost all action potentials elicited in one cell conducted to its partner. These results show that Fic. 2. Electron micrographs of just-attaching (a) and fusing (b) axopodia from a pair of individuals. b Note that the coiling direction of axonemal microtubules in the two adjacent axopodia is different from each other. 37,000. 182 both the electrotonic spread of potential responses and the conduction of action potentials can occur through the membrane of axopodia before two cell bodies begin to fuse. We examined a great number of ultra-thin sections of just-approaching organisms, especially at the early stage C. Two adjacent axopodia were seen as if they were just-attaching and fusing as shown in Figure 2. It is necessary to determine if these axopodia were of the same individual or the two different ones. We found that the coiling direction of axonemal microtubules was always the same in every axopodia of one organism, i.e., it was clockwise if the axonemal microtubules were viewed from the axopodial base to the tip. As shown in Figure 2, in the two just-attaching or fusing axopodia, the coiling direction was different from each other. This indicates that they were derived from the two different individuals, respec- tively. The present study has revealed the degree of electrical coupling between two organisms during the cell fusion and has shown that the two organisms have electrical interactions through their axopodia before the two cell bodies begin to fuse. Observations through electron microscope have presented a strong evidence supporting the T. Niso1, M. KosBayAsuli et al. physiological findings. 12 13 REFERENCES Barrett, S. M. (1958) J. Protozool., 5: 205-209. Vollet, J. J., Roth, L. E. and Davidson, M. (1972) J. Cell Biol., 55: 269a. Vollet, J. J. and Roth, L. E. (1974) Cytobiologie, 9: 249-262. Shigenaka, Y., Ogura, T. and Maruoka, T. (1976) Zool. Mag., 85: 65-69. Shigenaka, Y. and Kaneda, M. (1979) Annot. Zool. Japon., 52: 28-39. Shigenaka, Y., Maruoka,T., Toyohara, A. and Suzaki. T. (1979) Annot. Zool. Japon., 52: 163- 178. Naitoh, Y. (1982) In “Electrical Conduction and Behavior in ‘Simple’ Invertebrates”. Ed. by G. A. B. Shelton, Clarendon Press, Oxford, pp. 1-48. Podesta, R. B. (1982) Membrane Physiology of Invertebrates, Marcel Dekker, Inc., New York. Shigenaka, Y., Watanabe, K. and Suzaki, T. (1980) Annot. Zool. Japon., 53: 103-119. Nishi, T., Kobayashi, M. and Shigenaka, Y. (1986) J. Exp. Zool., 239: 175-182. Shigenaka, Y., Roth, L. E. and Pihlaja, D. J. (1971) J. Cell Sci., 8: 127-152. Spurr, A. R. (1969) J. Ultrastruct. Res., 26: 31-43. Reynolds, E. S. (1963) J. Cell Biol., 17: 208-212. ZOOLOGICAL SCIENCE 5: 183-186 (1988) [COMMUNICATION] © 1988 Zoological Society of Japan An Established Marine Fish Cell Line with High Plating Efficiency Noriaki EpiraNi and TosHiyuki Kuso! Biological Laboratory, North Shore College, Atsugi 243, and ‘Biological Laboratory, Sophia University, Tokyo 102, Japan ABSTRACT—A fibroblast-like cell line with a high plating efficiency was established from the fins of a marine teleost fish, Sebastiscus marmoratus. The cells, designated SMF, have been subcultured over 300 pas- sages. SMF cells have grown well in TC-medium 199 supplemented with 2.28 g/l NaCl and 15% FBS at 25 °C. The modal chromosome number was 99, while that of primary cells was 48. Plating efficiency of SMF cells was over 70% when 200-400 cells were inoculated in a 25 cm? flask with a tight cap. A fish cell line with a high plating efficiency is a very useful material for in vitro studies of the effects of radiations, chemical mutagens and other environmental agents on fish cells. Since the report of Clem et al. [1] on a cell line derived from fins of the yellow-striped grunt, a number of fish cell lines have been established. However, the cell lines derived from marine teleost fish have so far been compared in number with that from fresh water teleost fish [2, 3]. The plating efficiencies of several fish cell lines have been reported to be less than 10% [2, 4], but Shima et al. [5] have established a cell line derived from the goldfish with a plating efficiency of approximately 20%. Plating efficiency in the fish cell lines is however much lower than in the mammalian cell lines. The purpose of the present study was to obtain a cell line with a high plating efficiency from marine teleost fish for studying the effects of several chemical mutagens contained in sea water. Accepted June 24, 1987 Received May 27, 1987 We have established a cell line from a scorpion fish, Sebastiscus marmoratus, with a_ plating efficiency of over 70% and was designated SMF. Some characteristics of clones obtained from SMF cells will be reported in a subsequent paper. MATERIALS AND METHODS The scorpion fish, Sebastiscus marmoratus, caught in Sagami Bay, Japan on 28 February 1983, was used in the present study. The individual, from which the cell line was established, was a female and 20 cm in total length. All of the fins were excised just above the muscled areas in the field. They were placed in chilled calcium and magnecium free marine phos- phate buffered saline containing 0.197M NaCl (M-PBS_) supplemented with 10% calf serum, 400 IU penicillin/ml (Meiji Seika, Tokyo, Japan), 200 ug streptomycin/ml (Meiji Seika) and 10 IU nystatin/ml (GIBCO). After 24 hr, these fins were immersed in Dakin’s solution for 3 min and then washed three times with saline containing antibio- tics for 10 min each. The washed fins were minced into small pieces and all the fins from a single fish were transferred to a flask containing 30 ml of 0.25% trypsin (1:250, Difco) solution in M- PBS”. After stirring at low speed with a small magnetic bar at 25°C for 15 min, the trypsin solution was discarded and replaced with another aliquot. The fluid was harvested 20 min later and more trypsin solution was added to the tissue. This procedure was repeated three times. At each harvest, calf serum was added to it to give a 184 N. EsBITANI AND T. Kuso concentration of 10% and stored in ice bath. The harvests were filtered through lens papers and centrifuged at 1,000 rpm for 5 min at 10°C. The supernatant was decanted and the cells were resuspended in culture medium. The cells were inoculated at the concentration of 300 cells in 1 mm? of a Corning 75 cm? culture flask with a tight cap and kept at 25°C. The culture medium used in the present study was TC-199 (GIBCO) supplemented with 2.28 yg NaCl/ml, fetal bovine serum (FBS) (GIBCO), 200 IU penicillin/ml, 100 g streptomycin/ml and 5 IU nystatin/ml. The medium was changed every third or fourth day by renewal of one-half of the old medium. Subculture was carried out by routine procedure. Chromosome preparations were made, em- ploying the air-drying technique. After the metaphases were arrested with colchicine, cells were harvested by trypsinization. Cells were treated with potassium chloride (0.075 M) for 15 min, fixed soon thereafter with methanol-acetic acid (3:1) and air dried. chromosomes were stained with 6% Giemsa (in 1/15 M phosphate buffer) solution. Plating efficiency was estimated by the following procedure. From 25 to 400 cells were inoculated in a Corning 25cm? culture flask with a tight cap containing 5 ml of the conditioned medium and kept at 25°C. After incubation for 14 days, cells were fixed with methanol and stained with 6% Giemsa solution. The number of colonies, con- taining more than 50 cells were counted. RESULTS AND DISCUSSION Morphology and growth The cells from the fins of Sebastiscus marmoratus continued to proliferate for more than 1,000 days with 300 passages. The population doubling time was about 40 hr. The morphology of primary cells and SMF cells at 203 PDN is shown in Figure 1. Although primary cells were fibroblast-like and arranged regularly, SMF cells were crisscrossed. The optimal concentration of FBS sup- plemented in the culture medium for cell growth was tested at 25°C. In each flask, 200 cells per mm? were inoculated. The results are presented in Ze Sie Led h aS E SB NSS? Fic. 1. Phase-contrast photomicrograph of primary cells (A) and SMF cells at 203 PDN (B). 50. 1 2 3 TIME AFTER INOCULATION (DAYS) Fic. 2. Effect of concentration of fetal bovine serum on cell growth at 25°C. Figure 2. The best growth was obtained at a concentration of 20% FBS. Hardly any difference could be demonstrated between the results at 15% and 20% FBS. The growth at concentrations of 5 and 10% FBS was delayed considerably. For determining the optimal temperature for cell growth, the flasks containing 200 cells per mm? were incubated at 15, 20, 25 and 30°C in the medium with 15% FBS. The results are shown in Figure 3. At 25°C, the growth of cells was extremely high and cells increased to almost 650 per mm? at the third day of culture. Growths of cells at other temperatures were lower. These results suggest that 25°C may be physiologically optimum for marine fish cells. An Established Marine Fish Cell Line 185 o ro) o m2 /m CELLS 400 OF NUMBER 0 1 2 2 TIME AFTER INOCULATION (DAYS) Fic. 3. Growth curves of SMF cells at 15, 20, 25 and 30°C. NUMBER OF CELLS /mm? (x10?) - = ~ al o a Oo > Fic. 4. Growth curve of SMF cells at 25°C. ée > i) = ge 4? es aa Lo "y ° = Bam dlehey Joncas < Wore Jenene Ny ets Nyt rostdy sh avs #. . 3 Qeee ) -— 3 ose asort A oS 2 aN ee Fic. 5. Photomicrograph of chromosomes of a primary cell (A) and a SMF cell at 203 PDN (B). 400 Taking these results into consideration, SMF cells were continued to culture at 25°C in the medium containing 15% FBS. The growth curve of SMF cells is shown in Figure 4. The saturation density of SMF cells was 1.9 10° cells per mm? and the contact inhibition of cell growth worked well. For long-term storage of SMF cells, 3-4 10° cells were suspended in one milliliter of fresh culture medium supplemented with 12% DMSO and cryopreserved at —196°C. The viability of the thawed cells stored for 10 months was around 60%. Distribution of chromosome number Chro- mosomes are rather small in size. Figure 5 (A and B) show the chromosomes of a primary cell and a SMF cell at 203 PDN. The distribution of chromosome number of primary cells and SMF cells at 146 and 203 PDNs is summarized in Figure 6. The modal chromosome number of primary cells was 2n=48 and the karyotype was comprised of one pair of metacentric and 23 pairs of acrocentric chromosomes. This result is consistent with that of ff} A> = — Ha | Fic. 6. Distribution of chromosome numbers in pri- mary cells and SMF cells at 146 and 203 PDN. 5 aa SS nc tad Pte OF oe IN roe ie Fic. 7. Number of colonies and plating efficiency of SMF cells at 237 PDN. Results are given as mean + standard deviation. 186 N. EBITANI AND T. KuBo Nishikawa et al. [6]. The chromosome number of SMF cells differed from those of primary cells. At 146 PDN, the major distribution of chromo- some numbers of SMF cells ranged from 96 to 98. Their modal number was 96 and more than 60% of the cells had 4N number. Their karyotype was composed of 4 metacentric and 92 acrocentric chromosomes. At 203 PDN, SMF cells showed a wide range in the major distribution of chromosome numbers, which were from 87 to 106. Their modal number was 99 and cells with 96 and 102 chromosomes were predominant. Their karyotype was com- posed of 4 metacentric and acrocentric chromo- somes. In view of these results, the transition in chromosome distribution from a diploid mode to a subtetraploid mode is likely to take place through a tetraploid phase. The doubling of chromosome set occurs probably in the early passages by endomitosis and then the loss or gain of acrocentric chromosomes is assumed to have taken place progressively. Wolf and Quimby [7] stated that the fish cell lines for which chromosome number had been determined were all heteroploid and attainment of the potential for indefinite subculturing was usually accompanied by alterna- tion in heteroploid chromosome constitution. Colony formation and_ plating — efficiency The colony forming ability of SMF cells was tested by counting the number of colonies on the 14th day after inoculation. The results were presented in Figure 7. When 200-400 cells per flask were inoculated, the plating efficiencies of SMF cells were 80, 83 and 72% at 233, 237 and 248 PDNs respectively. Only a very few reports have been made in the colony forming ability of fish cell lines. On the cell line derived from a goldfish, Suyama and Etoh [4] reported that the plating efficiency was around 10%, when 2,000 or more cells per dish were inoculated. Shima ef al. [5] reported that the plating efficiency of cell lines derived from the same species was about 20% at 183 or more PDNs. As for marine teleost fish cell lines, Clem et al. [1] reported that the plating efficiency of the cell line derived from the yellow-striped grunt was less than 1%. In comparison with these results, SMF cells cultured in the conditioned medium have a much higher plating efficiency. This high plating efficien- cy makes it possible to obtain clones from somatic cells of the marine teleost fishes. ACKNOWLEDGMENTS The authors thank Prof. S. Nadamitsu of Hiroshima Women’s University for his invaluable advice during the conduct of the present experiments. REFERENCES 1 Clem, L. W., Moewus, L. and Sigel, M.M. (1961) Proc. Soc. Exp. Biol. Med., 108: 762-766. 2 Wolf, K. and Quimby, M. C. (1969) In “Fish Physiol- ogy”. Ed. by W.S.Hoar and D.J. Randall, Academic Press, New York and London, Vol. 3, pp. 253-305. 3 Fryer,J.L., Yusha, A. and Pilcher, K.S. (1965) Ann. N. Y. Acad. Sci., 126: 566-586. 4 Suyama, I. and Etoh, H. (1979) Zool. Mag. (Tokyo), 88: 321-324. 5 Shima, A., Nikaido, O., Shinohara, S. and Egami, N. (1980) Exp. Gerontol., 15: 305-314. 6 Nishikawa, S., Honda, M. and Wakatsuki, A. (1977) J. Shimonoseki Univ. Fish. (Japan), 25: 187-191. 7 Wolf, K. and Quimby, M.C. (1962) Science, 135: 1065-1066. ZOOLOGICAL SCIENCE 5: 187-189 (1988) [COMMUNICATION] © 1988 Zoological Society of Japan Female Heterogametic Sex-determination in Xenopus laevis as Reconfirmed by Repeated Diploid Gynogenesis TosHiHiro NAKAMURA! ‘Zoological Institute, Faculty of Science, Hokkaido University, Sapporo 060, Japan ABSTRACT—Diploid gynogenetic production of offspring was repeated for 4 generations in Xenopus laevis. Analyses of the resulting sex phenotypes in these progenies supported the hypotheses that (a) the female is heterogametic (ZW), (b) a recombination between sex-determination genes and the centromere occurs with high frequency, and (c) homogametic (WW) females are highly fertile. INTRODUCTION Although the South African clawed frog Xeno- pus laevis lacks sex-chromosome dimorphism, the female heterogamety (ZW) of this animal has been demonstrated on the basis of mating experiments with sex-reversed frogs [1-4]. Mating of sex- reversed genetic males with normal males gives rise to all male offspring [2]. Conversely, mating of sex-reversed genetic females with either normal females or sex-reversed genetic males yields offspring with the sex ratio of 3 (female): 1 (male) or of 1:1, respectively [3]. These results are explained on the basis of the presence of both heterogametic (ZW) and homogametic (WW, “super”) genotypes in the female phenotype, while phenotypic males are homogametic (ZZ). The viability and fertility of homogametic super- females have recently been demonstrated in stu- dies on the progeny of first diploid gynogenesis [5]. The behavior of sex-determination genes was Accepted June 17, 1987 Received April 27, 1987 " Present address: Nippon Institute for Biological Sci- ence, 2221-1 Shinmachi, Ohme, Tokyo 198, Japan. examined during our efforts to establish a histo- compatible Xenopus laevis colony by repeated gynogenetic propagation [6, 7], as described below. MATERIALS AND METHODS The females of Xenopus laevis (Daudin) were induced to ovulate by injecting human chorionic gonadotropin (Gonatropin; Teikoku Zoki, Tokyo) into the dorsal lymph sac. For mating experi- ments, inbred J strain males [8] were used. Gynogenetic diploid animals were obtained as described previously [6, 9]. Briefly, eggs were inseminated by UV-irradiated (5,400-7,200 erg/ mm”) J sperm, followed by a cold shock for 15 min at 2°C between 12 and 30 min after insemination for suppressing the second polar body emission. Previous analyses [8] proved that under these conditions the participation of sperm genome is totally eliminated and the resulting normally de- veloping individuals are all diploid. The lack of sperm-derived genome was also confirmed by an acute response against J skin grafts in metamor- phosed animals [6, 7]. Larvae were fed boiled alfalfa leaf powder or mashed green peas, and metamorphosed animals were fed chopped liver or a commercial fish meal (Oriental Yeast Ind., Tokyo), three to six times a week. All animals were reared in aquaria at 23+0.5°C. The sex of progeny was identified morphologi- cally or functionally as early as one year after fertilization. 188 T. NAKAMURA RESULTS AND DISCUSSION Gynogenetic F, progeny produced from four randomly selected outbred females (A-D in Table 1) comprised 92.0 to 100% females. Starting from three gynogen F, mothers (Al, Bl and B2), gynogenetic propagation was repeated to produce F,-F, progeny. As summarized in Table 1, all F,- F, offspring from the Al mother were females, whereas F, and F; progeny from B line mothers included a small proportion (<16.7%) of males. These results may be explained in the context of a female heterogametic system operated by one locus per genome [4], as follows. Upon gynogene- tic propagation, if there is no recombination between a putative sex-determinant locus and the centromere, only homogametic females (WW) and males (ZZ) will be obtained [cf., 6, 10, 11]. However, recombination at the pertinent locus should give rise to heterogametic (ZW) females, yielding females and males in a ratio dependent on the rate of recombination. In this respect, the occurrence of 92.0 or 92.3% females in the gynogenetic F, generation (offspring of A and C in Table 1) is not much different from the rate of TABLE 1. 83.3% (ZW+WW) expected theoretically on the basis of the maximum frequency of recombination (66.7% ZW) in the pertinent locus in gynogenetic offspring [11, 12]. The hypothesis that there is a relatively high frequency of recombination of the sex-determination gene is tenable in view of the appearance of males in gynogens F; and F; derived from putatively heterogametic (ZW) mothers (Bl and B2 in Table 1). The rate of the appearance of females (83.3-94.7%) in gynogens F,-F; in this study is quite similar to the previously reported values of 80.7% [5] and 86.0% [13] obtained in F, gynogens from outbred animals. Gynogenetic progeny of line A frogs was charac- teristic in that all 286 F,-F, frogs were consistently female (Table 1), suggesting that they were homo- gametic (WW). To prove their homogamety, three randomly selected F; frogs (A111, A112 and A113) and an Fy, frog (A1111) were mated with J males. More than 500 (JxXline A) hybrids thus obtained were all females (Table 2). The heter- ogamety (ZW) of these females was supported by the result that a high proportion of males appeared when a (J x A111) hybrid frog was mated with a J male (Table 2). In contrast to the fertility of The number of offspring and the ratio of females in gynogenetically produced F,-F, frogs Number of offspring Generation Mother* female (%) male A 23 ( 92.0) 2 F B 4 (100.0) 0 ; Cc 12 ( 92.3) 1 D 6 (100.0) 0 Al 120 (100.0) 0 F, Bl 18 ( 94.7) 1 B2 8 ( 88.9) 1 All 46 (100.0) 0 F; Bll 5 ( 83.3) B12 3 (100.0) 0 Alll 49 (100.0) 0 F, A112 37 (100.0) 0 A113 34 (100.0) 0 a, A-D, different outbred females; numbers, the lineage relationship of A and B progeny, e.g., B11 and B12 are progeny of B1. Sex-determination in Gynogenetic Xenopus 189 TABLE 2. The number and the ratio of females and males in the offspring produced by mating of gynogenetic progeny with J males Number of offspring Mother Father female (%) male (%) A111? J 178 (100.0) 0 (0) A112 J 151 (100.0) 0 (0) A113 J 78 (100.0) 0 (0) Allll J 106 (100.0) 0 (0) (JxA111)° J 32 ( 58.2) 23 (41.8) a, A111-A113 indicate F; gynogen derived from F, gynogen Al1; Al111, F, gynogen derived from F; gynogen A111. b, one female progeny derived from mating of A111 with one J male. putative homogametic (WW) gynogenetic diploid females, all females produced by refrigerating W eggs after fertilization with unirradiated sperm were sterile, apparently due to their triploid (ZWW or WWW) state (data not shown). In conclusion, our results not only support the previous notion that the female of Xenopus laevis is heterogametic (ZW), but also show that the recombination between this gene(s) and the cen- tromere occurs in rather high frequency. Fur- thermore, consistent with Colombelli et al. [5], stable and highly fertile lines of WW super-females are easily obtained by gynogenetic propagation. Although it has been shown that the sex- determination locus is not linked with the major histocompatibility complex [7], it would be worth- while to determine the linkage between the sex- determining genes and already identified genes in this species [13]. ACKNOWLEDGMENT This study was supported in part by Grant-in-Aid for Scientific Research No. 60440100 from the Ministry of Education, Science and Culture of Japan. The author expresses appreciation to Drs. Ch. Katagiri and S. Tochinai for their guidance and help in preparing the manuscript, and to T. Enami and H. Ohinata for their collaboration in producing gynogens. REFERENCES Chang, C. Y. and Witchi, E. (1955) Proc. Soc. Exp. Biol. Med., 89: 150-152. 2 Chang, C. Y. and Witchi, E. (1956) Proc. Soc. Exp. Biol. Med., 93: 140-144. 3. Mikamo, K. and Witchi, E. (1963) Genetics, 48: 1411-1421. 4 Mikamo, K. and Witchi, E. (1966) Cytogenetics, 5: 1-19. 5 Colombelli, B., Thiebaud, Ch. H. and Muller, W. P. (1984) Mol. Gen. Genet., 194: 57-59. 6 Nakamura, T., Kawahara,H. and Katagiri, Ch. (1985) Zool. Sci., 2: 71-79. 7 Nakamura, T., Sekizawa, A., Fujii, T. and Katagiri, Ch. (1986) Immunogenetics, 23: 181-186. 8 Katagiri, Ch. (1978) Dev. Comp. Immunol., 2: 5- 14. 9 Kawahara, H. (1978) Dev. Growth Differ. , 20: 227- 236. 10 Nace, G. W., Richards, C. M. and Asher, J. H., Jr. (1970) Genetics, 66: 349-368. 11 Asher, J. H., Jr. (1970) Genetics, 66: 370-391. 12 Volpe, E. P. and Dasgupta, S. (1962) J. Exp. Zool., 151: 287-302. 13. Reinschmidt, D., Friedman, J., Hauth, J., Ratner, E., Cohen, M., Miller, M., Krotoski, D. and Tomp- kins. R. (1985) J. Hered., 76: 345-347. ae “) ; sees - sce i® " ZOOLOGICAL SCIENCE 5: 191-195 (1988) [COMMUNICATION] © 1988 Zoological Society of Japan Medaka Hatching Enzyme Consists of Two Kinds of Proteases which Act Cooperatively SHIGEKI YASUMASU, IcHIRO IUCHI and KENJIRO YAMAGAMI Life Science Institute, Sophia University, Kioicho, Chiyoda-ku, Tokyo 102, Japan ABSTRACT — Intensive fractionation of hatching liquid of the medaka, Oryzias latipes, by repeating Toyopearl HW-5S0 Superfine gel filtration chromatography in alka- line buffers gave rise to five fractions of proteases ultimately. In terms of some characteristics of the enzymes such as the choriolytic activity relative to the proteolytic activity, and the chromatographic behavior expressed as the ratio of the effluent volume for each fraction to the total column volume (V,/V,), the en- zymes of these five fractions could be classified into two groups; one was a protease with high choriolytic activity and V./V, of 0.80-0.83 (high choriolytic enzyme, HCE), and the other was a protease with low choriolytic activity and V./V, of 0.47-0.48 (low choriolytic enzyme, LCE). When both enzymes were combined together, they showed a marked synergistic choriolytic activity. This fact strongly suggests that HCE and LCE participate in actual choriolysis cooperatively. The hatching enzyme of various animal species is secreted from the hatching embryo into the perivitelline space, participates in breakdown or solubilization of an egg envelope, and is released into the medium after the egg envelope became soluble [1-4]. Physical heterogeneity or polymorphism of the hatching enzyme(s) has been reported so far, implying that the enzyme with a similar activity in the hatching liquid is separated into some fractions representing different molecular weights or differ- ent electric charges on some fractionation proce- dures such as gel filtration chromatography or zone electrophoresis. Such heterogeneity has been Accepted August 11, 1987 Received June 27, 1987 ascribed to a probable difference in the state of association of one and the same enzyme with some concomitant heterologous substances [5-9]. However, it is dubious, at present, whether the hatching enzyme is a single enzyme or an enzyme system in which multiple enzymes with different properties participate in egg envelope digestion. It was suggested that the association of the enzyme with any heterologous substances might be canceled by repeating gel filtration column chro- matography in an alkaline buffer [6]. The present report describes the presence of two different types of proteases in the hatching liquid of medaka as revealed by repeating Toyopearl gel filtration chromatography in an alkaline buffer system. Both enzymes are considered to be responsible for natural choriolysis (egg envelope digestion), as a synergistic solubilization of chorion (egg envelope) occurred when both enzymes were applied together to the isolated chorion. MATERIALS AND METHODS The fertilized eggs of the orange-red variety of the medaka, Oryzias latipes, were used as mate- rials. The embryos were cultured in the medium (10mM NaCl-1 mM NaHCOs) containing penicil- lin G (K-salt, 100 units/ml, Meiji Seika Co.) and streptomycin sulfate (100 «g/ml, Meiji Seika Co.) until the time of hatching. The culture medium was refreshed every other day. The hatching liquid, the culture medium in which the larvae hatched out, was filtered and used as the enzyme source. 192 S. YASUMASU, I. luCHI AND K. YAMAGAMI Determination of enzyme activity Through- out the experiment, the enzyme activity was expressed in terms of two kinds of activities, 1. e., proteolytic activity (P.A.) and choriolytic activity (C.A.). P.A. was determined in the standard method using casein (Hammarsten, Merck) as substrate and expressed in terms of the increase in absorbance of the supernatant of the deproteinized reaction mixture at 280 nm (AOD»,0) [10]. C.A. was determined following two different methods. One was the turbidimetric method using fine fragments of chorion as substrate and the activity was expressed as increase of transmission at 610 nm of the reaction mixture (AT¢19) as reported previously [11]. The other was a method by which the amount of peptides solubilized from chorion fragments was determined: Twenty mg of coarse fragments of chorion was incubated with the enzyme in 1 ml of reaction mixture (50 mM Tris -HCI-10mM NaCl, pH7.5) at 30°C with con- tinuous shaking. After the reaction was stopped by adding 1 ml of 30mM ethylenediamine tet- raacetate (EDTA) [10], the amount of the solubil- ized peptides in the supernatant was determined in terms of the absorbance at 280 nm (AOD2g9). Column chromatography The hatching li- quid derived from 10,000-15,000 hatching larvae was added by solid ammonium sulfate to 60% saturation, allowed to stand for 1hr and then centrifuged at 10,000 rpm for 20 min to sediment the precipitates. The precipitates were solubilized by suspending in about 10 ml of S mM Tris: HCI-5 mM NaCl (pH 8.5) and dialyzing against the same buffer for 1-2 hr. The clear solution thus obtained was applied onto a Toyopearl HW-50 Superfine (Toyo Soda) column (1.5X91 or 2.678 cm) equilibrated with the same buffer and eluted at a flow rate of 14 ml/hr or 24 ml/hr. At this step, four peaks of proteolytic activity, Pa (1), Pa (2), Pa (3) and Pa (4), were obtained. Further fractionation of Pa (1), and the combination of the other three peaks, Pa (2, 3, 4) was performed by repeating the ammonium sulfate precipitation, followed by Toyopearl HW-50 Superfine gel filtration chroma- tography in 50 mM NaHCO;-Na,CO; (pH 10.2). All the procedures described above were carried out at 0-4°C. The elution pattern of protein was depicted on the basis of absorbance at 280 nm. RESULTS AND DISCUSSION Toyopearl HW-50 Superfine column chroma- tography of the ammonium sulfate precipitate of the hatching liquid in a slightly alkaline buffer (pH 8.5) produced four proteolytically active peaks, Pa (1), Pa (2), Pa (3) and Pa (4) (Fig. 1-a). Among them, the void volume peak, Pa (1), contained a large amount of chorion digest. The second peak was characterized by its having very low choriolytic activity as compared with the proteolytic activity. The rechromatography of Pa (1) on the same column in an alkaline buffer (pH 10.2) shifted the effluent position of the activity backward and gave rise to two proteolytically active fractions as shown in Figure 1-b; one was an irregular peak of activity, Pa (1)-1, and the other was a single sharp peak, Pa (1)-2. Pa (1)-1 was found to include two proteases: When Pa (1)-1 was rechromatographed in the alkaline buffer (pH 10.2) as before, it was separated again into two sharp peaks of proteolytic activity, Pa (1)-1-1 and Pa (1)-1-2 (Fig. 1-c). Pa (2, 3, 4) also could be separated by rechroma- tography on the Toyopearl HW-50 Superfine column in the alkaline buffer (pH 10.2) into two fractions, each of which represented a single sharp peak of proteolytic activity (Fig. 1-d). After all, the intensive fractionation of the hatching liquid by repeating Toyopearl HW-50 Superfine column chromatography in the alkaline buffer gave rise to five sharp peaks of protrolytic activity, Pa (1)-1-1, Pa (1)-1-2, Pa (1)-2, Pa (2, 3, 4)-1 and Pa (2, 3, 4)-2, each of which could not be fractionated further. These five fractions could be classified into two groups in terms of the choriolytic activity relative to the proteolytic activ- ity and their chromatographic behavior expressed as V,/V;, where V, is the effluent volume for each fraction and V, is the total column volume [12]. As shown in Table 1, the enzymes in Pa (1)-1-1 and Pa (2, 3, 4)-1 are considered to belong to one group, as they are characterized by low choriolytic activity as compared with the proteolytic activity (CA/PA=0.63-1.21) and by a low value of V./V, (0.47-0.48). The enzymes in Pa (1)-1-2, Pa (1)-2 and Pa (2, 3, 4)-2 seem to belong to the other group, as they have very high choriolytic activity (CA/PA =30.47-48.05) and exhibit higher V./V, Hatching Enzyme of Medaka 193 03 bis | a ret oe he oh 1 | ! | $ 2 Pale (a) | = = = | esl. lo: 2 ¢@ 0 g eee ee | | | 05; la 5 | | to) te) : Very, Paizaa)-2 = Ponni-2 b Ee h (06 =~ k0 a if = \ s & 4 = = 2 0 HO i é &| € Ne 4 © 3 | | " 4 2 ny \ < Pee Fe a |e a toa € POE co ° I WPS \ < OF Hl g|o as 5 oy E 4 Paizaai-] i te) Mt t) as 10 \. Cc n 10 > flo 7 9 ' f] 24 = = E g 8 | 3s (a) fe) g£/ 4 as i) 05 10 V, ay, Fic. 1. Fractionation of proteolytic enzymes in the hatching liquid of medaka by Toyopearl HW-SO0 Superfine gel filtration column chromatography. (a) Ammonium sulfate (60% saturation) precipitate of the hatching liquid (16,000 larvae eq.) was fractionated at pH 8.5 (5mM Tris-HCI-5 mM NaCl). (b) Pa(1) in (a) was collected, added by ammonium sulfate (60% saturation) and the precipitate was collected, dissolved and chromatographed at pH 10.2 (50 mM NaHCO;3-Na,CO3). (c) Pa (1)-1 in (b) was collected, precipitated by ammonium sulfate (60% saturation) and rechromatographed at pH 10.2. (d) Pa (2), Pa(3) and Pa (4) in (a) were combined and added by ammonium sulfate (60% saturation) to obtain precipitates. Pa (2, 3, 4) thus obtained was chromatographed at pH 10.2. Column size and the volume of each tube were 2.6 x78 cm and 8.6ml, respectively in (a) and (b), and 1.5x91lcm and 3.4ml, respectively in (c) and (d). ——: Protein amount, g—4m: Proteolytic activity determined using casein as substrate [cf. 10], G---(: Choriolytic activity determined by turbidimetry [cf. 11]. 194 S. YAsuMASU, I. TUCHI AND K. YAMAGAMI TABLE 1. Choriolytic activity relative to proteolytic activity and chromatographic behavior of various fractions of proteases obtained from the hatching liquid of medaka by repeated Toyopearl HW-50 gel filtration chromatography Fraction Proteolytic activity (ml)* (A r026)5 mini30°C) (AOD 90/30 min/30°C) — C-A./P.A. V/IVi Paty 35) 0.14 0.223 0.63 0.47 Pa ao) 6.99 0.197 35.48 0.81 Paty 50) 8.50 0.279 30.47 0.80 Pae13} ae 0.25 0.206 121 0.48 Pa 0} oe 16.00 0.333 48.05 0.83 C.A./P.A. is expressed simply in terms of the ratio of the choriolytic activity to the proteolytic activity. fraction used as the enzyme solution. Paiz3.41-] 20 + Paiz34i-2 Pai23.41-2 05 incubation time (min) Fic. 2. Cooperative choriolytic action of two kinds of proteolytic enzymes fractionated from the hatching liquid of medaka. Pa (2, 3, 4)-1 and Pa (2, 3, 4)-2 refer to LCE and HCE, respectively. The chorioly- tic activity (C. A.) was determined by the second method described in Materials and Methods. The amounts of both enzymes used were approximately equivalent in terms of their proteolytic activity to- ward casein (AOD>»./30°C/30 min=0.4). Arrow- head indicates the time when Pa (2, 3, 4)-1 was added to the incubated mixture of chorion frag- ments and Pa (2, 3, 4)-72. value (0.80—0.83) in common. Thus, the proteoly- tic enzymes in the hatching liquid of medaka seem to be classified ultimately into two types of Ve/V, refers to retention of each enzyme fraction to the column. * Volume of the enzymes; low choriolytic enzyme (LCE) and high choriolytic enzyme (HCE). It is considered that a large part of Pa (2) and Pa (4) in Figure l-a correspond to LCE and HCE, respectively, from the values of V./V,. Moreover, it is highly probable that LCE and/or HCE in the form complexed with some other substances are eluted as Pa (1) and Pa (3). As shown in Figure 2, when LCE of Pa (2, 3, 4)- 1 and HCE of Pa (2, 3, 4)-2 were applied together to the chorion, a marked synergistic choriolysis occurred, while LCE, if applied singly, exerted no significant choriolytic activity. Essentially the same results were obtained when Pa (1)-1-1 was used for LCE, and either Pa (1)-2 or Pa (1)-1-2 was used for HCE. Thus, making a distinction among the enzymes in those fractions was possible also from their roles in choriolysis. Recently both LCE and HCE have been obtained in homogeneous forms and found to be distinct from each other in some physical chemical characteris- tics. Their biochemical properties will be reported in the following papers. Proteolytic enzyme in the hatching liquid of medaka has been fractionated by Sephadex G-75 column chromatography (pH 8.5) into two frac- tions, PI and PII, [6, 11] and they seem to correspond to Pa (1) and Pa (2, 3, 4), respectively, in the present experiment. Each of Pa (1) and Pa (2, 3, 4) includes both LCE and HCE. Therefore, Hatching Enzyme of Medaka 195 a previous view that the polymorphic hatching enzyme of medaka is a single and the same enzyme [6] should be corrected according to the present results, which strongly suggest that the hatching enzyme of medaka is an enzyme system composed of two distinct enzymes, HCE and LCE, acting cooperatively. In the present experiment, a pre- vious presumption that a proteolytic enzyme with- out clearing (choriolytic) activity might be present in the hatching liquid of medaka [11] was verified. ACKNOWLEDGMENT This work was partly supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture, Japan. REFERENCES 1 Bourdin, J. (1926) C. R. Soc. Biol., 95: 1242-1243. Ishida, J. (1936) Annot. Zool. Japon., 15: 453-457. Cooper, K. W. (1936) Proc. Natl. Acad. Sci., USA., 22: 433-434. Yamamoto, M., Iuchi, I. and Yamagami, K. (1979) Dev. Biol., 68: 162-174. Barrett, D., Edwards, B.F., Wood,D.B. and Lane, D. J. (1971) Arch. Biochem. Biophys., 143: 261-268. Yamagami, K. (1975) J. Exp. Zool., 192: 127-132. Ogawa, N. and Ohi, Y. (1968) Zool. Mag. (Tokyo), 77: 151-156. Ohi, Y. and Ogawa, N. (1970) Zool. Mag. (Tokyo), 79: 17-18. Schoots, A. F. M., Sackers, R. J., Overkamp, P. S. G. and Denucé, J. M. (1983) J. Exp. Zool., 226: 93-100. Yamagami, K. (1973) Comp. Biochem. Physiol., 46B: 603-616. Yamagami, K. (1972) Dev. Biol., 29: 343-348. Ackers, G. K. (1975) In “The Proteins”. Ed. by H. Neurath and R.L. Hill, Academic Press, New York, Vol. 1, pp. 1-94. se ad ZOOLOGICAL SCIENCE 5: 197-200 (1988) [COMMUNICATION] © 1988 Zoological Society of Japan Corpuscles of Stannius of Clarias batrachus in Response to 1, 25 Dihydroxyvitamin D; Administration AJAI K. SRIVASTAV and SHYAM P. SRIVASTAV Department of Zoology, University of Gorakhpur, Gorakhpur-273 009, India ABSTRACT— Administration of 1, 25 dihydroxyvitamin D; (1, 10 or 50U/100g bw) to the catfish, Clarias batrachus for 10 days activated the corpuscles of Stan- nius. INTRODUCTION The corpuscles of Stannius (CS) secrete hypocal- cemic factor(s). This is evident by a rise in serum calcium level after removal of CS which is cor- rected by administration of CS extract [1]. A PTH-like substance (parathyrin) has recently been localized immunocytochemically in the eel CS [2]. Lafeber et al. [3] have reported parathyroid hormone-like effects of rainbow trout Stannius products on bone resorption of embryonic mouse calvaria. Although vitamin D; is abundantly present in fish liver, its role in calcium homeostasis has been emphatically denied [4]. Recently, it has been reported that vitamin D3 and its metabolites induce hypercalcemia in fishes [5-8]. Moreover, Srivastav et al. [9] have also reported hyperactivity of CS after vitamin D; treatment. In the present study we have investigated the effects of 1,25 dihydroxyvitamin D3; (1, 25 (OH)D3) on the CS of the freshwater catfish, Clarias batrachus. MATERIALS AND METHODS Adult male specimens of Clarias batrachus (21- 25 cm, 65-90 g) were collected locally and acclima- Accepted August 18, 1987 Received April 21, 1987 tized to the laboratory conditions for one week prior to use. They were then divided into four groups (A, B, C and D). Fish of all the groups were kept in identical all-glass aquaria, each containing eight liters of tap water (renewed daily). Only six fish were kept in each aquarium and they were not fed during the experiment. Fish from all the groups (A-D) were given daily intramuscular injections of the following treat- ments for 10 days: Group A: 0.1 ml/100 g bw of vehicle (ethanol) Group B: 1 U of 1, 25 (OH)2D;/100 g bw Group C: 10 U of 1, 25 (OH),D;3/100 g bw Group D: 50 U of 1, 25 (OH)2D;/100 g bw In all cases, the injection volume was 0.1 ml/100 g bw. Six fish from each group were anesthetized with MS 222 (Sandoz Ltd., Basle) 2 hr after the last injection on the Ist, 3rd, 5th and 10th day of the experiment. Blood samples were collected by the sectioning of caudal peduncle. Blood samples from the non-treated (normal) specimens were also taken before the start of the experiment. After clotting of the blood the sera were separated by centrifugation at 3500 rpm and analysed for serum calcium level according to Trinder’s [10] method. After the collection of blood samples, the CS along with the adjoining portion of the kidney, were extirpated from the fish and fixed in Bouin’s fluid and Zenker’s formol. After routine proces- sing in graded series of alcohols and clearing in xylene, tissues were embedded in paraffin. Serial sections were cut at 4-6 ~m and stained with 198 A. K. SRIVASTAV AND S. P. SRIVASTAV Fic. 1. CS of vehicle-treated fish showing the arrangement of corpuscular cells along septa (S). AF 400. Fic. 2. CS of fish from group B displaying decreased stainability of corpuscular cells after 10 days of treatment. AF x400. Fic. 3. CS of fish from group C exhibiting dilatation of sinusoids (S) after 5 days of treatment. HE x 400. Fic. 4. CS of fish from group D depicting empty spaces (ES), cell debris (D) and loss of cellular arrangement (arrows) after 10 days of treatment. AF 400. CS Response to 1, 25 (OH).D; 199 hematoxylin-eosin (HE) and aldehyde fuchsin (AF). The nuclear diameter was measured with the aid of an ocular micrometer. Each nucleus was measured along its long and short axes and mean value was calculated. From each group 300 nuclei were measured (fifty nuclei from each specimen) at every interval. Differences in the serum calcium level and nuclear diameter among different groups were analysed by Student’s r-test. RESULTS The changes in the serum calcium levels and corpuscular cell nuclear size of groups A-D at various intervals have been summarised in Table L In normal specimens, CS are enveloped by a thin connective tissue sheath from which a number of septa extend into the gland. The corpuscular cells are arranged along these septa (Fig. 1). Each cell possesses a distinct nucleus and homogenous cytoplasm; however, the cell boundaries are not distinct. There is only one cell type. When stained with HE, the cytoplasm of corpuscular cells is eosin-positive. However, the granules are not discernible by this stain. After AF staining, cytoplasm of corpuscular cells exhibits positive response and displays many coarse AF-positive TABLE 1. groups (A-D) after various intervals granules which are densely aggregated around the nucleus. In group B (1 U/100 g bw of 1, 25 (OH)2D3), there is no change in the corpuscular cells until day 5. On day 10, the nuclear size increases (Table 1) and there is decreased stainability of corpuscular cells (Fig. 2). In group C (10 U/100 g bw of 1, 25 (OH)2D3), the corpuscular cells show a progressive increase in nuclear size from day 5 to day 10 (Table 1). There is a progressive sinusoidal dilatation and decreased stainability of corpuscular cells from day 5 onwards (Fig. 3). In group D (50 U/100 g bw of 1, 25 (OH)2D3) there is an increase in nuclear size (Table 1), decreased stainability of corpuscular cells and dilatation of sinusoids on day 3. These responses are exaggerated on day 5. Also, there is noticed degeneration of certain cells. The nuclear size of intact cells shows a further increase (Table 1). On day 10, the AF stainability of corpuscular cells increases. The degenerative changes are at their peak — the cells loose their arrangement, the emp- ty spaces and cell debris become conspicuous (Fig. 4). DISCUSSION The present study reveals that the activity of CS is affected quite perceivably on treatment with Serum calcium (mg/100 ml) and nuclear size (4m) of corpuscular cells of different Days Group A Group B Group C Group D 1 Serum Ca 10.42+0.11 10.32 +0.23 10.68 +0.14 10.76 +0.23 Nuclear size 4.32 +0.04 4.29+0.03 4.36+0.02 4.34+0.06 3 Serum Ca 10.36+0.21 10.94+0.17 11.30+0.24? 11.92+0.27° Nuclear size 4.29+0.03 4.33+0.03 4.42+0.06 4.48 +0.06° 5 Serum Ca 10.28 +0.19 11.60+0.13° 13.48+0.319 14.06+0.21¢ Nuclear size 4.34+0.04 4.46+0.05 4.85+0.049 4.96 +0.034 10 Serum Ca 10.32+0.21 11.82 +0.25° 13.08+0.279 13.56+0.329 Nuclear size 4.28 +0.05 4.75 +0.06° 5.02 +0.05¢ 5.19+0.044 Each value represents mean+S.E. of six specimens. a, b, c and d indicate significant responses compared to group A: P<0.05, <0.02, <0.01 and <0.001, respectively. In normal fish, serum level of calcium was 10.25 +0.16 mg/100 ml and nuclear size of corpuscular cells was 4.21+0.03 sm. 200 different doses of 1, 25 (OH) D3 which is express- ed by the decreased stainability and nuclear hypertrophy of corpuscular cells and sinusoidal dilatation. These changes have been considered as indications of the hyperactivity of the gland [9, 11]. The hyperactivity of the corpuscular cells suggests an increased synthesis and release of hypocalcemic factor(s) to combat the elevated serum calcium levels caused by 1,25 (OH)2D3 treatment. The degeneration of the corpuscular cells can be attributed to their overactivation in response to perpetual hypercalcemic challenge. Degeneration due to hyperactivity has also been reported by Hiroi [12] and Srivastav et al. [9]. Although Lopez et al. [13] failed to get any change in CS activity in eels following 1,25 (OH),D3 treatment, the present study clearly indicates that this metabolite affects the activity of CS in C. batrachus. The failure of Lopez et al. [13] in not observing any change in CS could be explained as they have sacrificed their specimens 24 hr after the last injection of 1, 25 (OH)2D; (in their experiments only two injections of 1, 25 (OH),D3 were given at 0 and 48 hr and the fish were killed 72 hr after the first injection). It may be possible that by this time the changes in CS may have been recovered. ACKNOWLEDGMENTS One of us (AKS) is thankful to Sandoz Ltd., Basle for generous gift of MS 222 and U. G. C., New Delhi for financial assistance. 10 11 A. K. SRIVASTAV AND S. P. SRIVASTAV REFERENCES Kenyon, C. J., Chester Jones, I. and Dixon, R. N. B. (1980) Gen. Comp. Endocrinol., 41: 531-538. Lopez, E., Tisserand-Jochem, E. M., Eyvuem, A., Milet, C., Hillyard, C., Lallier, F., Vidal, B. and MacIntyre, I. (1984) Gen. Comp. Endocrinol., 53: 28-36. Lafeber, F.P.J.G., Schaefer, H.I.M.B., Herr- mann-Erlee, M. P. M. and Wendelaar Bonga, S. E. (1986) Endocrinology, 119: 2249-2255. MacIntyre, I., Colston, K. W., Evans, I. M., Lopez, E., Macauley, S. J., Peignoux-Deville, J., Spanos, E. and Szelke, M. (1976) Clin. Endocrinol., Suppl. 5: 85. Swarup, K. and Srivastav, S. P. (1982) Gen. Comp. Endocrinol., 46: 271-274. Srivastav, A. K. (1983) J. Fish Biol., 23: 301-303. Swarup, K., Norman, A. W., Srivastav, A. K. and Srivastav, S. P. (1984) Comp. Biochem. Physiol., 78B: 553-555. Fenwick, J.C., Smith, K., Smith, J. and Flik, G. (1984) Gen. Comp. Endocrinol., 55: 398-404. Srivastav, S. P., Swarup, K. and Srivastav, A. K. (1985) Cell. Mol. Biol., 31: 1-5. Trinder, P. (1960) Analyst, 85: 889-894. Olivereau, M. and Olivereau, J. (1978) Cell Tissue Res., 186: 81-96. Hiroi, O. (1970) Bull. Fac. Fish. Hokkaido Univ., 21: 179-192. Lopez, E., Peignoux-Deville, J., Lallier, F., Col- ston, K. W. and MacIntyre, I. (1977) Calcif. Tissue Res., Suppl. 22: 19-23. ZOOLOGICAL SCIENCE 5: 201-203 (1988) [COMMUNICATION] © 1988 Zoological Society of Japan Translocation of *Ca from the Endolymphatic Sacs to the Bone in Rana nigromaculata Masako Fusimort!, YuIcHI SASAYAMA and CHITARU OGURO” Department of Biology, Faculty of Science, Toyoma University, Toyama 930, Japan ABSTRACT—In the frog, Rana nigromaculata, intra- arterially administered *°Ca was incorporated into all tissues and organs within 2 hr. The endolymphatic sacs showed the highest level of incorporation, followed by bones. After 4 days, bone was the only tissue that showed significant increase in *°Ca deposition in com- parison with the situation after 2 hr. In all other tissues and organs including the endolymphatic sacs, the values of incorporation were significantly lower than those after 2hr. From these results, it was concluded that the endolymphatic sacs serve as a temporary depot tissue for calcium, which then moves chronically to the bones. INTRODUCTION The ultimobranchial gland (UB) of the frog contains calcitonin which evokes hypocalcemia and hypophosphatemia in rats [1-3]. This is supported by the facts that in several anuran species, substances in the ultimobranchial cells crossreact with anticalcitonin antisera [4-6]. It has been reported that in Rana nigromaculata, removal of the UB markedly reduces the calcium content in the endolymphatic sacs [7]. Moreover, administration of calcitonin promotes an uptake of “SCa into the endolymphatic sacs, but produces no effect on the calcium kinetics in other tissues and organs [7]. Thus, calcitonin seems to be a promoter of calcium storage in the endolymphatic sacs in frogs. The present report describes the fate of “Ca Accepted August 6, 1987 Received July 15, 1987 ' Present address: Fukumitsu Senior High School, Toyama 939-16, Japan. * To whom reprints should be addressed. incorporated into the endolymphatic sacs at var- ious times after administration. MATERIALS AND METHODS Adult males of Rana nigromaculata, 15-40 g bw, were used in the present study. They were collected in the suburbs of Toyama City during spring and summer. The frogs were kept in tapwater (Ca 1.7, Mg 0.6, Na1.0, K 0.2 mg/100 ml) which had been allowed to stand for a long period. The sciatic artery of anesthetized frogs was cannulated using polyethylene tubing (PE 10, Clay Adams) for the administration of “Ca and calcito- nin (synthetic salmon calcitonin, dissolved in 0.6% NaCl adjusted to pH 4.6 with HCl). Each frog received calcitonin (50 mU/50 1/10 g bw) via the cannula. Thirty min later, Ca (8 peCi/50 4/10 g bw) was administered through the same cannula. Five frogs were killed 2 hr after- wards and 30 different tissues and/or organs were dissected out and *°Ca radioactivity in each was measured. Another 5 frogs were killed 4 days afterwards and treated in the same way as the 2-hr frogs. RESULTS The results obtained are shown in Figure 1. This figure shows the levels of “Ca activity in the endolymphatic sacs, femur, vertebrae, cartilage, skin and some other soft tissues and organs. Levels in some soft tissues and organs which 202 M. Fusimori, Y. SASAYAMA AND C. OGuRO 2 HOURS AFTER [__]4 DAYS AFTER 45Cq (cpm/mg) . q | tea fa EF GH IJK AiB C.D Fic. 1. “Ca activity in various tissues and organs of Rana nigromaculata 2hr and 4 days after intra- arterial administration of “Ca. A, Skin (ventral). B, Skin (dorsal). C, Kidney. D, Muscle. E. Stomach. F, Gallbladder. G, Liver. H, Endolym- phatic sacs. I, Femur. J, Vertebrae. K, Cartilage. showed a similar tendency to the values shown in the figure are excluded. Two hours after *°Ca administration, all samples examined showed a certain level of *°Ca activity. The endolymphatic sacs showed the highest activ- ity (15,530 cpm/mg tissue), followed by the femur and vertebrae (4,848 and 4,351 cpm/mg tissue, in that order). Skin (dorsal and ventral) and cartilage showed fairly high values, whereas “°Ca activity in the other samples was very low. Four days after *Ca administration, radioactiv- ity values in all tissues and organs, including the skin and cartilage, had decreased significantly except for the bones and gallbladder. However, the increased activity level in the gallbladder was not statistically significant. The activity levels in the endolymphatic sacs, femur and vertebrae were 10,643, 7,695 and 6,346 cpm/mg tissue, respective- ly. The increases in the levels for the femur (P<0.001) and vertebrae (P<0.001) and the decrease in the level for the endolymphatic sacs (P<0.05) were significantly different from the values obtained 2 hr after “Ca administration. Thus, the only portion for which “Ca activity was increased after 4 days was the bones, in comparison with the level 2 hr after Ca adminis- tration. DISCUSSION It has been previously reported that in Rana nigromaculata, ultimobranchialectomy (UBX) brings about a decrease in calcium content only in the endolymphatic sacs, whereas other organs show no response [7]. Futhermore, salmon calcito- nin promotes *°Ca incorporation into the endolym- phatic sacs but not into other tissues [7]. From these results and other indirect evidence [8, 9], it was concluded that the function of the UB is to promote calcium incorporation into the endolym- phatic sacs through calcitonin secretion. However, nothing has been known about the long-term fate of “Ca incorporated into the endolymphatic sacs. “Ca activity was decreased in the majority of tissues and organs after 4 days in comparison with the situation after 2 hr. Moreover, it was remark- able that *°Ca incorporated into the endolymphatic sacs was also decreased as in the other soft tissues and organs. It seems, therefore, that the function of calcitonin is rather transient and not long- lasting. The only portion in which *°Ca activity increased was bone. The increase in activity in the femur was 63% and that in the vertebrae was 46%. These levels are in marked contrast to the decrease observed in the endolymphatic sacs. These results clearly show that *Ca incorporated into the endolymphatic sacs under the specific influence of calcitonin was translocated into the bones within several days. It has previously been reported that the en- dolymphatic sacs accumulate a calcium salt which is utilized for skeletal ossification during meta- morphosis when the larvae do not feed [10, 11]. It has been suggested that the UB of frog larvae promotes the accumulation of calcium in the Translocation of **Ca from the endolymphatic sacs 203 endolymphatic sacs at this stage [12]. It is concluded from the present results that in 3 adult anurans one of the functions of the endolym- phatic sacs is to serve as a temporary depot tissue for calcium, facilitating subsequent calcium supply 5 to the bones. ACKNOWLEDGMENTS The present study was supported in part by Grants-in- Aid for Scientific Research from the Ministry of Educa- tion, Science and Culture of Japan (No.60440006) and 8 the Tamura Foundation for the Encouragement of Science and Technology. 9 REFERENCES 1 Oguro, C., Nagai, K.-I., Tarui, H. and Sasayama, 11 Y. (1981) Comp. Biochem. Physiol., 68A: 95-97. 2 Oguro,C. and Sasayama, Y. (1985) In “Current 12 Trends in Comparative Endocrinology”. Ed. by B. Loft and W. N. Holmes, Hong Kong Univ. Press, Hong Kong, pp. 839-841. Oguro,C., Nogawa,H., Nagai, K.-I. and Sasayama, Y. (1986) Zool. Sci., 3: 663-668. Van Noorden, S. and Pearse, A. G. E. (1971) His- tochemie, 26: 95-97. Treilhou-Lahille, F., Jullienne, A., Aziz, M., Beaumont, A. and Moukhtar, M.S. (1984) Gen. Comp. Endocrinol., 53: 241-251. Sasayama, Y., Oguro, C., Yui, R. and Kambegawa, A. (1984) Zool. Sci., 1: 755-758. Oguro, C., Fujimori, M. and Sasayama, Y. (1984) Zool. Sci., 1: 82-88. Robertson, D. R. (1969) Gen. Comp. Endocrinol., 12: 479-490. Robertson, D. R. (1972) In “Calcium, Parathyroid Hormone and the Calcitonins”. Ed. by R. V. Talmage and P. L. Munson, Excerpta Medica, Amsterdam, pp. 21-28. Guardabassi, A. (1960) Z. Zellforsch., 51: 278-282. Pilkington, J.P. and Simkiss, K. (1966) J. Exp. Biol., 45: 329-341. Robertson, D. R. (1971) Gen. Comp. Endocrinol., 16: 329-341. > 7 , i wor * ao tele -. - £ o R Vx: - ’ re ¢t . , bel : Foy of Spence) VE ioe | Suge et) Te | We a! ; 4 y : ¥ yy Re ae hp ar % an j matt Mer ret Loe OUR ag am te - ie * Yoh ue Gh ee } ‘ ’ ‘ | f t nae 7 ‘ | i) y" i x ir ==. \ a GQ). PRUE AR Te yi 53 eeta PT ZOOLOGICAL SCIENCE 5: 205-208 (1988) [COMMUNICATION] © 1988 Zoological Society of Japan Morphological Observations of the Large Intestine in the Common Vole, Microtus arvalis Pallas HasIME AMASAKI, MASAYUKI DaiGco, and NorIFUMI Mecuro! Department of Veterinary Anatomy, Nippon Veterinary and Zootechnical College, 1-7-1 Kyonan-cho, Musashino-shi, Tokyo 180, and ‘Itoh Equine Clinic, 2-2 Katsushima, Shinagawa-ku, Tokyo 140, Japan ABSTRACT—In forty common voles, Microtus arvalis Pallas, the morphology and physiology of the large intestine (without the rectum) was observed and analyzed. As a result, we found the large intestine to consist of five segments: 1) the caecum, 2) the cranial region of the proximal colon, 3) the caudal region of the proximal colon, 4) the medial colon and 5) the distal colon. These were determined according to the mesenterial attaching pattern, the arterial supplying system and the mucous epithelial surface structure. The level of total volatile fatty acid (VFA) and level of water consistency in the digestive contents of the anatomically defined segments may correspond with digestive and absorptive functions. INTRODUCTION Previously, investigators have examined the vole, Microtus montebelli, which has a compound stomach (with the forestomach) and a well- developed large intestine (with the caecum and colon) for fermentative chambers [1-5]. These functions in the vole are similar to those of the forestomach in ruminant animals [5, 6]. Recently, a morphological investigation was also conducted on the caecum and proximal colon of the vole, Microtus agrestis [7]. This paper presents some morphological and physiological observations of the large intestine (without the rectum) in the common vole, Micro- tus arvalis Pallas. Accepted July 1, 1987 Received April 17, 1987 MATERIALS AND METHODS Forty mature common voles, Microtus arvalis Pallas, were used for morphological and physio- logical observation. These voles were given to us from the Department of Veterinary Physiological Chemistry of the Nippon Veterinary and Zootech- nical College. Fifteen were used for macro- anatomical observations of the mesenterial attaching pattern and the vascular supplying sys- tem, and five for scanning electron microscopical (SEM) observation of the epithelial surface struc- ture. Each specimen used for SEM investigation was fixed in Zamboni-solution. After fixation, they were dried to the critical point (HITACHI, HCO-1). They were coated with gold using a vacuum evaporating ion coater (NEVA, FTM-112), and observed by SEM (JSM-S25, Mk 2). The remaining 20 animals were used for phys- iological analysis of total volatile fatty acid (VFA) levels, which consisted of acetic acid, propionic acid and butyric acid, and the level of water consistency in the digestive contents in each part of the large intestine. The level of total VFA of the digestive contents was determined using gas- chromatography (SHIMAzU, GC-6AM), and the level of water consistency of the digestive contents was determined using the drying chamber. RESULTS AND DISCUSSION Snipes [7, 8] reported that the large intestine of 206 H. Amasaki, M. Daico AND N. MEGURO Microtus agrestis can be defined as three segments, _ result was observed. Namely, part M1 consisted of M1, M2 and M3. In our investigation on the the caecum and proximal colon covered with the common vole, Microtus arvalis Pallas, the same __caecal mesentery, which continued from the ileo- 7. of, ( Mt l lt Jejunum & Ileum Ae Na: t . ag 2S WA Neeson satis amare ; LONE INCE gS : 4 : Sawigu seers = i EF AACE, i A ARM peh ® si, as — ES, pia SN NI pln, Naperenn POTENT NA ae e ger Fic. 1. Schematic diagram of the division in the large intestine (without the rectum) according to the mesenterial attaching pattern. A: Proximal part of the small intestinal mesentery covering over the duodenum and the proximal jejunum. B: Distal part of the small intestinal mesentery covering over the distal jejunum and the ileum. C(M1): Proximal part of the large intestinal mesentery covering over the caecum and the proximal colon. D(M2): Medial part of large intestinal mesentery covering over the medial colon. E(M3): Distal part of large intestinal mesentery covering over the distal colon and the rectum. Fic. 2. Schematic diagram of the vascular supplying system in the intestine. A. M. Cr.: Cranial mesentery artery. A.M. Ca.: Caudal mesentery artery. A. I. C.: Ileo-colic artery. Aa. I.: Ileal arteries. Fic. 3. Mucous epithelial surface in part S1. Well developed mucous folds are observed on the epithelial surface. Bar: 100 yan. Fic. 4. Mucous epithelial surface in part $2. V-shaped mucous folds are made by many continuous papillae. Bar: 100 yam. Fic. 5. Mucous epithelial surface in part S3. V-shaped mucous folds are disappearing caudally. Bar: 100 um. Fic. 6. Mucous epithelial surface in part S4. The longitudinal folds appear on the epithelial surface. Bar: 100 ym. Large Intestine in Common Vole caecal mesentery. Part M2 was the medial colon covered with the colic mesentery. Part M3 was the distal colon covered with the caudal intestinal mesentery, which was attached to the dorsal abdominal wall (Fig. 1). In Microtus agrestis, Snipes [7, 8] described the colon as being divided into four segments (Al, A2, A3 and A4) according to the vascular supplying system. In our investigation, part Al was the most proximal colon from the ileo-caecal orifice to the centripetal gyri. The colic branch of the ileo-caecal 7 V.F.A.-Consistency 100 PIP P29 Pse Pa PS COLON Fic. 7. 207 artery, which was the branch from the cranial intestinal artery, dispersed into part Al. Part A2 was the caudal region of the proximal colon from the centripetal gyri to the cranial colic flexure. The right colic artery, which was the branch from the cranial intestinal artery, dispersed into part A2. Part A3 was the medial colon. The medial colic artery, which was the branch from the cranial intestinal artery, dispersed into part A3. Part A4 was the distal colon. The caudal intestinal artery dispersed into part A4. Meanwhile, the caecal Water-Consistency 100 0 Pie RZ IPS Ph “PS COLON Schematic diagram of the division of the large intestine (without the rectum) in the common vole, Microtus arvalis Pallas, according to our anatomical investigations. P1: Caecum, which consists of parts M1, Al and S1. P2: Proximal colon (1), which consists of parts M1, Al, S1 and S2. P3: Proximal colon (2), which consists of parts M2, A2, S2 and $3. P4: Medial colon, which consists of parts M2, A3 and S4. PS: Distal colon, which consists of parts M3, A4 and S4. Fic. 8. The level of total VFA and the level of water consistency of the digestive contents in the anatomically defined segments of the large intestine. Mean with standard error. (N=10; mmole/dl for the level of VFA, % for the level of water consistency). 208 H. AmasaklI, M. Daico AND N. MEGURO branch of the ileo-caecal artery, which was the branch of the cranial intestinal artery, dispersed into the caecum (Fig. 2). The caecum added to part Al. As for the epithelial surface structure observed by SEM, the colon was divided into four segments (S1, S2, S3 and S4). The epithelial surface structure of part S1 had well-developed mucous folds, which had similar structures to those of the caecum (Fig.3). In part $2, V-shaped mucous folds consisted of continual small papillae (Fig. 4). In part S3, these folds gradually disappeared approaching the caudal region (Fig. 5). In part S4, the longitudinal folds appeared on the epithelial surface (Fig. 6). These findings were nearly the same as those reported by Behmann [1]. In view of the above findings, the large intestine (without the rectum) may be divided into five segments by mesenterial attaching pattern and vascular supplying system. The epithelial surface structure was included into above divided seg- ments of the large intestine. Part 1 is the caecum which consists of parts M1, Al and S1. Part 2 is the cranial region of the proximal colon which consists of parts M1, Al, S1 and S2. Part 3 is the caudal region of the proximal colon which consists of parts M2, A2, S2 and S3. Part 4 is the medial colon which consists of parts M2, A3, and S4. Part 5 is the distal colon which consists of parts M3, A4 and S4 (Fig. 7). Analyses of the levels of total VFA and water consistency were conducted on the digestive con- tents in each segment of the large intestine. In part 1, the level of total VFA in the digestive contents was higher than in all the other parts. The level of VFA tended to decrease going from part 1 to part 4. In part 4, the level of the VFA was about 1/3 of that of part 1 (Fig. 8). The VFA may be produced and absorbed well in the caecum and the proximal colon (part 1 and part 2). However, the level of VFA increased in part 5. Meanwhile, the level of water consistency in the digestive contents merely tended to decrease caudally (Fig. 8). High levels of VFA in part 5 may actually be the reason for the produce of VFA in the fecal mass. The peripheral zone of the fecal mass might be dried. VFA may be producing at the core zone of the fecal mass. It is possible that VFA is increased at the fecal mass of the distal colon (part 5). The fecal mass began formation in the medial colon, and were completely formed in the distal colon. REFERENCES 1 Behmann, H. (1973) Z. wiss. Zool., 186: 173-298. 2 Golley, F. B. (1960) J. Mammal., 41: 89-99. 3 Kajigaya, H. and Goto, N. (1980) J. Mammal. Soc. Jpn., 8: 171-180. 4 Kurohmaru,M., Nishida,T. and Mochizuki, K. (1981) Jpn. J. Vet. Sci., 43: 887-899. 5 Obara, Y. and Goto., N. (1980) Jpn. J. Zootech. Sci., 51: 393-396. 6 Sugawara, M. (1982) Jpn. J. Zootech. Sci., 53: 400- 405. 7 Snipes, R. L. (1979) Anat. Embryol., 157: 181-203. 8 Snipes, R. L. (1979) Anat. Embryol., 157: 329-346. 209 INSTRUCTIONS TO AUTHORS ZOOLOCIGAL SCIENCE publishes contri- butions, written in English, in the form of (1) Reviews, (2) Articles, and (3) Communications of material requiring prompt publication. A Review is usually invited by the Editiors. Those who submit reviews should consult with the Editor-in- Chief or the Managing Editor in advance. Articles of less than 6 printed pages and Communications less than 3 printed pages will be published free of charge. Charges will be made for extra pages (7,000 yen/page). A Communication cannot ex- ceed 4 printed pages. No charge will be imposed for invited reviews up to 15 printed pages. No free reprints of Articles and Communications are avail- able. To the author of an invited review 50 reprints are provided gratis. Submission of papers from nonmembers of the Society is welcome. However, page charges (7,000 yen/page) will be made to nonmembers. A. SUBMISSION OF MANUSCRIPT The manuscript should be submitted in tripli- cate, one original and two copies, each including all illustrations. Rough copies of line drawings and graphs may accompany the manuscript copies, but the two copies of continuous-tone prints (photo- micrographs, etc.) should be as informative as the original. The manuscript should be sent to: Dr. SENCHIRO KAWASHIMA, Managing Editor, ZOOLOGICAL SCIENCE, Zoological Institute, Faculty of Science, Hiroshima University, 1-1-89 Higashisenda-machi, Naka-ku, Hiroshima 730, Japan. B. CONDITIONS All manuscripts are subjected to editorial re- view. A manuscript which has been published or of which a substantial portion has been published elsewhere will not be accepted. It is the author’s responsibility to obtain permission to reproduce illustrations, tables, etc. from other publications. Accepted papers become the permanent property of ZOOLOGICAL SCIENCE and may not be reproduced by any means, in whole or in part, without the written consent of both the Zoological Society of Japan and the author(s) of the article in question. C. ORGANIZATION OF MANUSCRIPT The desirable style of the organization of an original paper is as follows: (1) Title, Author(s) and Affiliation (2) Abstract (3) Introduction (4) Materials and Methods (5) Results (6) Discussion (7) Acknowledgments (8) References (9) Tables (10) Illustrations and Legends (11) Footnotes. The author is not obliged to adhere rigidly to this organization. He or she may modify the style when such modification makes the presentation clearer and more effective. In a Communication, combination of some of these sections is recom- mended. There is no restriction on the style of review articles. D. FORM OF MANUSCRIPT Manuscripts should be typewritten and double spaced throughout on one side of white type- writing paper with 2.5cm margins on all sides. Abstract not exceeding 250 words, tables, figure legends and footnotes should be typed on separate sheets. All manuscript sheets must be numbered successively. The use of footnotes to the text is not recommended. 1. Title page The first page of manuscript should contain title, authors’ names and addresses of university or institution, abbreviated form of title (40 characters or less, including spaces), and name and address for correspondence. Authors with different affilia- tions should be identified by the use of the same superscript on name and affiliation. If one or more of the authors has changed his or her address since the work was carried out, the present address(es) to be published should be indicated in a footnote. In addition, a sub-field of submitted papers to be used as heading of an issue may be indicated in the first page. Authors are encouraged to choose one of the following: physiology, cell biology, molecu- lar biology, genetics, immunology, biochemistry, 210 developmental biology, reproductive biology, endocrinology, behavior biology, ecology, phy- logeny, taxonomy, or others (specify). However, the Editors are responsible for the choice and arrangement of headings. 2. Introduction This section should clearly describe the objec- tives of the study, and provide enough background information to make it clear why the study was undertaken. Lenghty reviews of past literature are discouraged. 3. Materials and Methods This section should provide the reader with all the information that will make it possible to repeat the work. For modification of published methodol- ogy, only the modification needs to be described with reference to the source of the method. 4. Results Results should be presented referring to tables and figures, without discussion. 5. Discussion The Discussion should include a concise state- ment of the principal findings, a discussion of the validity of the observations, a discussion of the findings in the light of other published work dealing with the same subject, and a discussion of the significance of the work. Redundant repetition of material in Introduction and Results, and extensive discussion of the literature are discouraged. 6. Statistical analysis Statistical analysis of the data using appropriate methods is mandatory and the method(s) used must be cited. 7. References References should be cited in the text by an Arabic numeral in square parentheses and listed at the end of the paper in numerical order. For example: 1 Takewaki, K. (1931) Oestrus cycle of female rat in parabiotic union with male. J. Fac. Sci. Imp. Univ. Tokyo, Sec. IV, 2: 353-356. 2 Shima, A., Ikenaga, M., Nikaido, O., Takabe, H. and Egami, N. (1981) Photoreac- tivation of ultraviolet light-induced damage in cultured fish cells as revealed by increased colony forming ability and decreased content of pyrimidine dimers. Photochem. Photo- biol., 33: 313-316. 3 Hubel,O. and Wiesel, T.N. (1986) The functional architecture of the striated cortex. In “Physiological and Biochemical Aspects of Nervous Integration”. Ed. by F. D. Carlson, Prentice-Hall, New Jersey, pp. 153-161. 4 Campbell, R. C. (1974) Statistics for Biolo- gists. Cambridge Univ. Press, London, 2nd ed., pp. 59-61. Titles of cited papers may be omitted in Reviews and Communications. The source of reference should be given following the commonly accepted abbreviations for journal titles (e,g., refer to ‘International List of Periodical Title Abbrevia- tions’). The use of “in preparation”, “submitted for publication” or “personal communication” is not allowed in the reference list. “Unpublished data” and “Personal communication” should appear parenthetically following the name(s) in the text. Text citations to references with three or more authors should be styled as, e.g., Everett et al. [7]. E. ABBREVIATIONS Abbreviations of measurement units, quantity units, chemical names and other technical terms in the body of the paper should be used after they are defined clearly in the place they first appear in the text. However, abbreviations that would be rec- ognized by scientists outside the author’s field may be used without definition, such as SD, SE, DNA, RNA, ATP, ADP, AMP, EDTA, UV, and CoA. The metric system should be used for all measurements, and metric abbreviations (Table) Table. Abbreviations for units of measure which may be used without definition Length: km, m, cm, mm, san, nm, pm, etc. Area: km?, m?, cm’, mm? pm?, nm?, pm”, etc. Volume: km, m?, cm?, mm’, san’, nm?, pm’, kl, liter (always spellout), ml, ,d, nl, etc. Weight: kg, g, mg, wg, ng, pg, etc. Concentration: M, mM, uM, nM, %, g/l, mg/l, ug/l, etc. Time: hr, min, sec, msec, sec, etc. Other units: A, W, C, atm, cal, kcal, R, Ci, cpm, dB, v, Hz, Ix, Xg, rpm, S, J, IU, etc. should, in general, be expressed in lower case without periods. F. PREPARATIONS OF TABLES Tables should only include essential data needed to show important points in the text. Each table should be typed on a separate sheet of paper and must have an explanatory title and _ sufficient explanatory material. All tables should be referred to in the text, and their approximate position indicated in the margin of manuscript. G. PREPARATION OF ILLUSTRATIONS All figures should be appropriately lettered and labelled with letters and numbers that will be at least 1.5 mm high in the final reproduction. Note the conventions for abbreviations used in the journal so that usage in illustrations and text is consistent. All figures should be referred to in the text and numbered consecutively (Fig. 1, Fig. 2, etc.). The figures must be identified on the reverse side with the author’s name, the figure number and the orientation of the figure (top and bottom). The preferred location of the figures should be indi- cated in the margin of the manuscipt. Illustrations that are substandard will be returned, delaying publication. Illustrations in color may be pub- lished at the author’s expense. 1. Line drawings and graphs Original artwork of high quality, glossy prints mounted on appropriate mounting card (less than 25 X38 cm) should be submitted for reproduction. Author(s) may indicate size preference by making on the back of figures, such as “Do not reduce”, “Two-column width” (no wider than 14.5 cm), or “One-column width” (no wider than 7 cm). Lines must be dark and sharply drawn. Solid black, white, or bold designs should be used for histo- grams. Xerox or any other copying mean may be used for the two review copies. 211 2. Continuous-tone prints Three sets of continuous-tone prints (photo- micrographs, etc.) must be submitted. One set for reproduction should be mounted on appropriate mounting card, and the other two for reviewers may be unmounted prints. Xerox or similar copies of photomicrographs are not acceptable for review purposes. The continuous-tone prints should be submitted preferably at the exact magnification which is to be used in the published papers and trimmed to conform to the page size (in no case should it exceed 14.520 cm). Press-on numbers should be applied to the lower right corner of individual prints. Letters (a, b, c, etc.) should be used for multiple parts of a single figure. If important structures will be covered by use of the lower right corner, identification may be applied in the lower left corner. Reproduction of color photographs will have to be approved by the Editors. The extra costs of color reproduction will be charged to the authors. 3. Figure legends Each figure should be accompanied by a title and an explanatory legend. The legends for several figures may be typed on the same sheet of paper. Sufficient detail should be given in the legend to make it intelligible without reference to the text. H. PROOF AND REPRINTS A galley proof and reprint order will be sent to the submitting author. The first proofreading is the author’s responsibility, and the proof should be returned within 72 hours from the date of receipt (by air mail from outside Japan). The minimum quantity for a reprint order is fifty. Manuscript, tables and illustrations will be discarded after the editorial use unless their return is requested when the manuscript is accepted for publication. 212 ERRATUM In the Contents on the back-cover of Vol. 4, No. 5, 1987, the authors’ names of the article entitled ‘Geotactic behavior in Paramecium caudatum. II. Geotaxis assay in a population of the specimens’ were incorrectly printed. ‘Taneda, K.” should read ‘Taneda, K., S. Miyata and A. Shiota’. Published by 7} 4 Gf P Gi & eronbation the Japanese Society of Developmental Biologists The journal is devoted to the publication of original papers dealing with any aspects of developmental phenomena in all kinds of organisms, including plants and micro-organisms. Papers in any of the following fields will be considered: developmental genetics, growth, morphogenesis, cellular kinetics, fertilization, cell division, dormancy, germination, metamorphosis, regeneration and pathogenesis, at the biochemical, biophysical and analytically morphological levels ; reports on techniques applicable to the above fields. At times reviews on subjects selected by the editors will be published. Brief complete papers will be accepted, but not preliminary reports. Members of the Society receive the Journal free of charge. Subscription by institutions is also welcome. Papers in Vol. 30, No. 1. (Feburary, 1988) 1. REVIEW: A. Fustwara, K.Supon and I. Yasumasu: Activation of sea urchin eggs by halothane and its inhibition by dantrolene. 2. §. Pine and R. FLICKINGER: Base composition of poly (A*) nuclear RNA of frog embryo and friend erythroleukemia cells. 3. Y.Maeda: Changes of the endocytotic activities during the cell cycle of Dictyostelium cells. 4. H. Shimizu, N. Noro and R. Matsupa: Micromere differentiation in the sea urchin embryo: Expression of primary mesenchyme cell specific antigen during development. 5. R. Matsuda, T. Kitajima, H.Ohinata, Y. Katoh and T. HIGASHINAKAGAWA : Micromere differentiation in the sea urchin embryo: Two-dimensional gel electrophoretic analysis of newly synthesized proteins. 6. D.E. Chandler and V. D. VacquieR: Phorbol myristate acetate induces the phosphorylation of plasma membrane-associated proteins in sea urchin eggs. 7. K. Ueno, Y. Hiramoto, S. Hayashi and H. Konpou: Introduction and expression of recom- binant @-galactosidase genes in cleavage stage mouse embryos. 8. M. Okamoto: Inhibition of lens regeneration by nickel subsulfide in the Japanese common newt, Cynops pyrrhogaster. 9. M. Yamashita: A fine structural study of the fertilization process of the jellyfish Cladonema uchidai. 10. P. Guerrier, I. Néant and P. CLEDON: Urea-induced meiosis reinitation in oocytes of the starfish Marthasterias glacialis. Development, Growth and Differentiation (ISSN 0012-1592) is published bimonthly by The Japanese Society of Developmental Biologists, Department of Biology, School of Education, Waseda University, Tokyo 160, Japan. 1988: Volume 30. Annual subscription U. S. $110.00 including air speed delivery except Japan. Application to mail at second class postage rate is pending at Jamaica, NY 11431, U.S.A. Outside Japan: Send subscription orders and notices of change of address to Academic Press, Inc., Journal Subscription Fulfillment Department, 6277 Sea Harbor Drive, Orlando, FL 32887, U.S. A. Send notices of change of address at least 6-8 weeks in advance. Please include both old and new addresses. U.S. A. POSTMASTER: Send changes of address to Development, Growth and Differentiation, Academic Press. Inc., Journal Subscription Fulfillment Department, 6277 Sea Harbor Drive, Orlando, FL 32887, U.S. A. In Japan: Send nonmember subscription orders and notices of change of address to Business Center for Academic Societies Japan, 16-3, Hongo 6-chome, Bunkyo-ku, Tokyo 113, Japan. Send inquiries about membership to Business Center for Academic Societies Japan,4—16, Yayoi 2-chome, Bunkyo-ku, Tokyo 113, Japan. Air freight and mailing in the U.S. A. by Publications Expediting, Inc., 200 Meacham Avenue, Elmont, NY 11003, U.S.A. THE ULTIMATE NAME IN MICROMANIPULATION NARISHIGE OUR NEW MODELS MO-102 and MO-103 MAKE PRECISION MICROMANIPULATION SO EASY! SOME FEATU « The mani (Photo: by courtesy of Olympus Optical CO., LTD.) RES of MO-102 and MO-103: pulator head is so small that it can be mounted directly on the microscope stage. There is no need for a bulky stand. * Hydraulic remote control ensures totally vibration-free operation. * 3-D movements achieved with a single joystick. | NARISHIGE Micromanipulators Microelectrode pullers Stereotaxic instruments NARISHIGE SCIENTIFIC INSTRUMENT LABORATORY CO.,LTD. 4-9-28, Kasuya, Setagaya-ku, Tokyo 157 JAPAN Telephone: 03-308-8233 Telex: NARISHG J27781 (Contents continued from back cover) munofluorescence microscopy in Tetrahyme- na thermophila during conjugation Yasumasu, S., I.Iuchi and K. Yamagami: Medaka hatching enzyme consists of two kinds of proteases which act cooperatively (COMMUNICATION) Endocrinology Engstrom, W., E. Dafgard and _ S. Falkmer: Comparative effects in vitro of Myxine, Squalus, avian and mammalian insulins on DNA-synthesis in 3T3 mouse fibroblasts Ueda, H., T. Kosaka and K. W. Takahashi: Effects of long-term progesterone treatment on synchronized ovulation in guinea pigs Endo, K., T.Masaki and K. Kumagai: Neuroendocrine regulation of the develop- ment of seasonal morphs in the Asian com- ma butterfly, Polygonia c-aureum L.: Differ- ence in activity of summer-morph-producing hormone from brain-extracts of the long-day and short-day pupae Srivastav, A. K. and S. P. Srivastav: Corpus- cles of Stannius of Clarias batrachus in re- sponse to 1, 25 dihydroxyvitamin D3 admin- istration (COMMUNICATION) Fujimori, M., Y.Sasayama and C. Oguro: Translocation of “Ca from the endolympha- tic sacs to the bone in Rana nigromaculata (COMMUNICATION) _.................. 201 Morphology Amasaki,H., M.Daigo and WN. Meguro: Morphological observations of the large in- testine in the common vole, Microtus arvalis Pallas (COMMUNICATION) _........... 205 Behavior Biology Pandey,S.C. and S.D. Pandey: Sexual maturation in female wild mice: Combined effect of adults’ urinary chemosignals and minimum time of exposure to stimulus sub- stances for bringing the effects ........... 153 Verrell, P. A.: Sexual interference in the alpine newt, Triturus alpestris (Amphibia, Urodela, Salamandridae) ................ 159 Taxonomy Nakasone, Y.: Land hermit crabs from the Ryukyus, Japan, with a description of a new species from the Philippines (Crustacea, De- capoda, Coenobitidae) ................... 165 Instructions to Authors .................... 209 Erratum ees ese eee ce eee ree roses 212 ZOOLOGICAL SCIENCE VOLUME 5 NUMBER 1 FEBRUARY 1988 CONTENTS REVIEWS nipponensis: Identification of the excep- de Pomerai, D.I.: The _ transdifferentiation tional amino acid replacement at the distal of neural retina into lens in vitro. ......... 1 (E7) position and autoxidation of its oxy- Tsuneki, K.: The neurohypophysis of cyclo- FOFM avecthadesdoge ves bee eee eee 69 stomes as a primitive hypothalamic center of ; VEItEDrateS a. ccncacnacdes ons anesmsinn mame wine 21 Genetics Nakamura, T.: Female heterogametic sex- ORIGINAL PAPERS determination in Xenopus laevis as recon- Physiology firmed by repeated diploid gynogenesis Doman mirinerenita tye ater ofsee at (COMMUNICATION) ................5- 187 adapting light in vertebrate photoreceptors Immunology Pie Aree osatan sae etaadeenamea rage uae 33 Nagata, S.: T cell-specific antigen in Xenopus Takei, Y., J.Okubo and _ K. Yamaguchi: identified with a mouse monoclonal anti- Effects of cellular dehydration on drinking body: Biochemical characterization and spe- and plasma angiotensin II level in the eel, cies distribution ....-2....cu: 5) eee 77 Anguilla japonica ........ 0... eee eee eee 43 Nishi, T., M. Kobayashi, M. Isomura, H. Ishi- Developmental Biology da and Y. Shigenaka: Direct evidence for Fujisawa, H. andS. Amemiya: Temperature- axopodial fusion preceding cell-to-cell con- dependence in reaggregation of cells dissoci- tact in a heliozoan Echinosphaerium (COM- ated from sea urchin embryos with different MUNICATION) ...........0...2.e0ee ees 179 seasonal growth .............-e seen eee eee 85 Cell Biology Zama, N. and H. Katow: A method of quan- titative analysis of cell migration using a computerized time-lapse videomicroscopy oy Pe Fe eer eee en ere renee een 53 Iwasaki,S. K.Miyata and K. Kobayashi: Fine structure of the filiform papillar epithe- lium from the tongue of the frog, Rana nigromaculata Ebitani, N. and T.Kubo: An _ established marine fish cell line with high plating efficien- cy (COMMUNICATION) Biochemistry Suzuki, T., R. Muramatsu, T. Kisamori and T. Furukohri: Myoglobin of the shark Galeus Mitsunaga, K., Y.Fujino and I. Yasumasu: Probable participation of mitochondrial Ca’* transport in calcification of spicules and morphogenesis in sea urchin embryos Numakunai, T., Z. Hoshino and S. Kajiwara: Spawning of three intraspecific groups of the ascidian, Halocynthia roretzi (Drasche), in the wild, and fertilization among them... 103 Tahara, U.: Normal stages of development in the lamprey, Lampetra reissneri (Dybowski) Tsunemoto, M., O. Numata, T. Sugai and Y. Watanabe: Analysis of oral replacement by scanning electron microscopy and im- (Contents continued on inside back cover) INDEXED IN: Current Contents/LS and AB & ES, Science Citation Index, ISI Online Database, CABS Database Issued on February 15 Printed by Daigaku Printing Co., Ltd., Hiroshima, Japan aL. ee: Z Bb4 Vi ZOOLOGICAL ~ SCIENCE An International Jou ZOOLOGICAL SCIENCE The Official Journal of the Zoological Society of Japan Editor-in-Chief: The Zoological Society of Japan: Hideshi Kobayashi (Tokyo) Toshin-building, Hongo 2-27-2, Bunkyo-ku, Managing Editor: ok Tokyo 113, Japan. Tel. (03) 814-5675 Seiichiro Kawashima _ (Hiroshima) Officers: Assistant Editors: President: Nobuo Egami (Tsukuba) Takeo Machida (Hiroshima) Secretary: Hideo Namiki (Tokyo) Sumio Takahashi (Hiroshima) Treasurer: Tadakazu Ohoka_ (Tokyo) Kazuyoshi Tsutsui (Hiroshima) Librarian: Shun-Ichi Uéno (Tokyo) Editorial Board: Howard A. Bern (Berkeley) Walter Bock (New York) Aubrey Gorbman (Seattle) Horst Grunz (Essen) Robert B. Hill (Kingston) Yukio Hiramoto (Chiba) Susumu Ishii (Tokyo) Yukiaki Kuroda (Mishima) Koscak Maruyama (Chiba) Roger Milkman (Iowa City) Hiromichi Morita (Fukuoka) Kazuo Moriwaki (Mishima) Tokindo S. Okada (Okazaki) | Andreas Oksche (Giessen) Hidemi Sato (Nagoya) Hiroshi Watanabe (Shimoda) | Mayumi Yamada (Sapporo) Ryuzo Yanagimachi (Honolulu) ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and communications in the broad field of Zoology. The journal was founded in 1984 as a result of unification of Zoological Magazine (1888-1983) and Annotationes Zoologicae Japonenses (1897-1983), the former official journals of the Zoological Society of Japan. ZOOLOGICAL SCIENCE appears bimonthly. An annual volume consists of six numbers of more than 1000 pages including an issue containing abstracts of papers presented at the annual meeting of the Zoological Society of Japan. MANUSCRIPTS OFFERED FOR CONSIDERATION AND CORRESPONDENCE CONCERN- ING EDITORIAL MATTERS should be sent to: Dr. Seiichiro KAWASHIMA, Managing Editor, Zoological Science, Zoological Institute, Faculty of Science, Hiroshima University, 1-1-89 Higashisenda-machi, Naka-ku, Hiroshima 730, Japan, in accordance with the instructions to authors which appear in the first issue of each volume. Copies of instructions to authors will be sent upon request. SUBSCRIPTIONS. ZOOLOGICAL SCIENCE is distributed free of charge to the members, both domestic and foreign, of the Zoological Society of Japan. To non-member subscribers within Japan, it is distributed by Business Center for Academic Societies Japan, 6-16—3 Hongo, Bunkyo-ku, Tokyo 113. Subscriptions outside Japan should be ordered from the sole agent, VSP, Utrechtseweg 62, 3704 HE Zeist (postal address: P.O. Box 346, 3700 AH Zeist), The Netherlands. Subscription rates will be provided on request to these agents. New subscriptions and renewals begin with the first issue of the current volume. All rights reserved. No part of this publication may be reproduced or stored in a retrieval system in any form or by any means, without permission in writing from the copyright holder. © Copyright 1988, The Zoological Society of Japan Publication of Zoological Science has been supported in part by a Grant-in-Aid for Scientific Publication from the Ministry of Education, Science and Culture, Japan. ZOOLOGICAL SCIENCE 5: 213-215 (1988) © 1988 Zoological Society of Japan OBITUARY Kiyoshi Takewaki (1905-1988) Emeritus Professor Kiyoshi Takewaki of the University of Tokyo died at the age of 82 on the 16th of January 1988 after a long period of unconsciousness. His death was due to bronchial pneumonia originating in a traffic accident on the 12th of April 1987. Until the time of accident, Professor Takewaki had been healthy and vigorous. We feel great regret at the loss of this distinguished zoologist as well as an eminent endocrinologist. All that is left to us is to cherish his memory. Kiyoshi Takewaki was born on the Ist of March 1905 at Toyama City in the middle of the Honshu on the Japan Sea. As a primary school boy, he was watching a water-scorpion in the evening of a summer day, and was deeply impressed by this beautiful scene of nature. He wrote later about this moment of beauty in a book “Mizukamakiri wa tobu (A Water-Scorpion Flies)”. This esthetic impression exerted a long-lasting influence on his study and taste. He loved arts and crafts, mainly pictures, and occasionally he himself painted scenes from nature. In research, he maintained his naturalist’s mind, being always modest before nature as well as in evaluation of his own results. In 1922, he entered the Fourth High School, Science Course, at Kanazawa, and finished the Course after 3 years. Before leaving high school, he told his teacher about his choice of future occupation as zoologist. The teacher was greatly astonished at hearing this unexpected idea, because Takewaki always had the best record in the school. After a while, the teacher just said “It’s impossible to live that way”. In 1925, he was admitted to the Zoological Institute of the Faculty of Science at the Imperial University of Tokyo. During his university student period, he wrote four short papers based on his observations of insects, although his full-fledged studies did not begin until after his graduation in 1928. Here, Emeritus Professor Naohide Yatsu of the Imperial University of Tokyo should be mentioned in order to understand why Takewaki began to study endocrinology. Yatsu went to the United States the 214 OBITUARY year after graduation from the Imperial University (1900) to enter graduate school at Columbia University. He was strongly impressed there by Prof. Jacques Loeb’s lectures, and recognized the importance of experimental zoology. After he obtained Ph. D. in 1905, Yatsu moved to the Stazione Zoologica of Naples, then came back to the Imperial University as lecturer (1907), and became associate professor (1909). Yatsu made efforts to develop experimental zoology, but met with stout resistance to his attempted innovation. So, he moved as professor of the Department of Anatomy at Keio University in 1919. However, he was finally called back to be professor at the Imperial University in 1922, where again he began to develop experimental zoology. During a period of “Sturm und Drang” in the Imperial University of Tokyo, Yatsu introduced the technique of parabiosis into experimental zoology in an attempt to elucidate the balance of sex hormones in rats (1916). These efforts marked the commencement of experimental endocrinology in Japan. In 1928, Takewaki was added as assistant (instructor) at Yatsu’s laboratory and first encountered rats and mice as the experimental animals which he used for more than 50 years thereafter. His first study on rats was to examine the state of various blood cells following gonadectomy (1929), but the second study on parabiosis between intact and gonadectomized rats became the origin of his life work on the endocrinology of reproduction (1931). In 1933-1935, his subject of study shifted to the transplanted testis and ovary in intact, senile, unilaterally or bilaterally gonadectomized and cryptorchidized rats. In this period he also published two papers on the state of the testis transplanted into intact and gonadectomized lizards, a pioneer study in comparative endocrinology. He also studied changes in the mouse adrenal glands following treatments with gonadotropic extract from human pregnancy urine. The influences of castration, hysterectomy, pregnancy, pseudopregnancy and testis-implantation on the ovary and adrenal cortex were then examined in rats and mice (1936-1940). He obtained the degree of D. Sc. in 1936, and became an associate professor in 1938. At this time, he used not only rats but also other species of animals for experiments. He examined changes in the kidney and genital tract after removal of gonads and hypophysis in the snake, Natrix tigrina tigrina. He also studied the relationship between gonads and sex character in the isopod crustacean, Armadillidium vulgare and hormonal control of the molting in the canary in cooperation with Hideshi Kobayashi (1941-1947). Takewaki became professor in 1947. As the 1940’s were a severely difficult period for scientists in Japan, it was inevitable for Takewaki to reduce his work on mammals. After 1949, however, he again took up enthusiastically his studies in rats and mice, especially on the relationship between gonads and hypophysis. During this period of his research renaissance, he spent several months (in 1954) in Prof. C. R. Moore’s laboratory at University of Chicago. In 1955-1961, his interests were concentrated on the negative and positive feedback mechanisms between hypothalamo-hypophyseal and gonaaat sys- tems. For elucidation of these mechanisms he examined changes in intrasplenic ovarian grafts in gonadectomized rats. Neonatal treatments of rats with steroid hormones were carried out in connection with the study of sex differentation. In cooperation with his associates, he also studied the relationship between gonads and adrenals in rats, the actions of sinus gland hormones in shrimps, and the hypophyseal control of reproductive functions in fishes. The Third International Symposium on Comparative Endocrinology was held from June 5 to 11 in 1961 at Oiso, Japan. As the chairman of the international organizing committee, Takewaki led this symposium to a highly successful conclusion, resulting in an immeasurably strong impact on young zoologists studying the general and comparative endocrinology in Japan, especially since this synpo- sium was one of the earliest international meetings held in Japan after the war. In 1962-1965, he studied the permanent alteration of hypothalamo-hypophysio-gonadal system in persistent-estrous or -diestrous rats treated neonatally with sex hormones. He also worked on the genesis and nature of testicular tumors in rats. Thus, Takewaki succeeded in advancing the experimental endocrinology he had inherited from Yatsu, contributing greatly to our understanding of the hormonal control of structure and function of OBITUARY 215 reproductive organs. He also carried out comparative studies of reproductive phenomena in several animal species, laying the stage for the later flourishing of comparative endocrinology in Japan. Takewaki was a man of action. Until 70 years of age, he came to the laboratory before 8 a.m. every day, even on Sundays, and began his day by taking care of his animals. His speed of lecturing was very fast in order to convey as much knowledge as possible to students in a limited amount of time. He also rapidly reviewed the many manuscripts sent to him by investigators. Takewaki retired from the University of Tokyo in 1965, but never wanted to spend his days in retirement from research. He continued his studies of sex phenomena in rats for the next five years, as professor at Tokyo Women’s University. In 1970, he was invited to be a professor at Kawasaki Medical College in Kurashiki City, Okayama Prefecture, where he worked energetically with Dr. Yasuhiko Ohta, mainly on the uterine response to hormones under various physiological conditions in rats. He was elected a member of the Japan Academy in 1975, and expended great efforts in editing its Proceedings to the last. After he retired from the Medical College in 1976, he came back to Tokyo, and enjoyed ‘travelling, reviewing papers, appreciating various works of art, giving lectures at Atomi Women’s University (as a part-time lecturer), collecting insects around his cottage near Mt. Kurohime, Nagano Prefecture, for the rest of his life. Thus, Professor Kiyoshi Takewaki continued his efforts as an investigator until the end of his active life. His academic publications totalled more than 200, and his students and followers more than 50. He was a rather reserved professor, who provided a powerful stimulating influence on the research activity of his students by his deeds rather than by his words. We can expect that Kiyoshi Takewaki’s unflagging pioneer’s spirit that originated with Yatsu will continue to be sustained by the many zoologists who follow him. Nosoru TAKASUGI Department of Biology Yokohama City University “4 ~~. : 5 Ga ’ § us hf ' 14 4 —- : : juin, N, ‘ : : u - ; une ' r i = athe 2 Sh : i - ~ \ y Sa : a rar : x a : be i Lo 7 1 ce - — co = = 7 ‘ i fauule oarenll ee | pills (ise ZOOLOGICAL SCIENCE 5: 217-265 (1988) © 1988 Zoological Society of Japan REVIEW Female Reproduction in Malacostracan Crustacea JEAN-JACQUES Meusy and GENEVIEVE G. PAYEN Laboratoire de Physiologie de la Reproduction, Equipe Neuroendocrinologie des Crustacés, Université Pierre et Marie Curie et C. N. R. S., UA 040555 4, place Jussieu, Batiment A, F-75252 Paris Cedex 05, France CONTENTS IntroductiGN 63:4 aiethabacsens ceo cen and sabe cee eleens na mink ahs dravh conte. cutttandes w metesamadelanmalounah ders 218 I. Early steps of oogenesis and previtellogenesis ............. 0.000 cece cece cence ees 219 A. Chronology and cytology: sss. nccices cece cates te to snedse Made tsaeeaweng oe tes esines 219 Bs Re pull ett Oasis ores ctae ts est scepter aa e sapeepe aoe erty ce anions encrabe Cresonciarehes Sesreves Siesevors 220 1. Mechanisms of oocyte differentiation ........... 0.0... c cece eee eee 220 2. Maintenance of the germinative zone ............ 0.0 cece cece e eee tenn eee 221 3. Oocyte growth until puberty ss: csi sotsee timer aes io weeatesnseseseetinmecguane DOL Ie s Vitellogenesisi: snadhatretd cagiins ahs aaaisantnede cop watstcae aioe ahs ote Naan atelier sicieeerein Dee LOL As. --VitelOgenesisS! PLOCESS cis 56ers, seiiatteie Sake crs furusetttoiers sists Qualité atessr a) sie ag uetdle w ajtuels eras ayereaals 221 1. General considerations .......... 0... nee e tenn eens 221 2e/ SOTISIVOL VItCHOSENING Whee. cae actia sa pee cea gular sie h aekina sits metemeisiere seca eee 3. Vitellogenin uptake by vitellogenic ovaries ............ 0.0.0 c cece eee 223 a) Transformation and role of the follicle envelope ....................0.0.00005 223 b) Vitellogenic oocyte and endocytosis mechanism ..................0 0c eee eee ee 224 4. From vitellogenin to vitellin: a processing? .......... 0.0... c cece eee eee eee eee 224 5. Vitellogenin synthesis and vitellogenin level in haemolymph as means for monitoring vitellogenesis .............. 00.0 cece cece eee eens 226 6. Timing of the reproductive cycle: duration and relation with the: molting cycle: scc.cteev eacstauceded send eon eoutes Go eae tamede aegiaenegemess 227 B:. Vitellogenesis control cicscis natina ea ce iadst wh aie seem camaro nd aa emetd anmenieed ais 229 1. Inhibitory control by VIH (Vitellogenesis Inhibiting Hormone) .................. 229 a) The X organ-sinus gland complex ............ 0.0.0 cece eee eect eee nes 229 Eyestalked species Eyestalkless species Db) “Wa ysiok Acton ss iemiact hose ate Me mame aeeavecs oteimnatte date bene wav onwe chan gated 230 Control of vitellogenin synthesis Control of vitellogenin uptake by the oocytes c) Extraction and purification of VIH .............0 000s ccc ccc eee n eens 231 @)» ‘Datestidatavon VI ez. ss 08 ho .c4 cies ocd aregeaa arseats c isinine desisns Co edure aarmtiemee® 231 2. Stimulatory control! ss.cchache Gongeniwur end western oh tea da ekens eave tu mandtaedion 233 a) Neurohumoral factors .......... 000.0 e tence eee nees 234 b) Vitellogenin Stimulating Ovarian Hormone (VSOH) ..................0.005- 234 C) PE CA YStErO1ds Beg tices caceaseereocnetrt ape x ctmictay hens occa yd ai eee See le oe ee 234 A) s WUVEN OLS cai care eas create ee reetn cae ase ae evden edie bes ace Do Mae apie 235 e) Ovary-stimulating factor from males ............0 2.0.0.0 ccc cece eee eee ees 237 Received October 30, 1987 218 J.-J. MEUSY AND G. G. PAYEN 3. Effect of other:substances: «i. ..cccnnsieasssn Se ee dee uteoenes eka ous seenenenee 237 a) Androgenic: hormone. o..crcc. eae ecco deurene da saraind os eb ansaeusqeyolnea dearer 237 b): Vertebrate hormones: 22 eck dered asad oa beet Dew anys aeons nena e nate aee 237 c) “Queen-substance” of honeybees ............. 0.0. e cece cece cece eee een ees 238 4. Environmental factors and vitellogenesis ........... 0... c cece cece eee e eee ees 238 a) Light ‘and temperature : + cicoccce0d.on hades ooh MS Aue antes cake date Melee aee 238 Ib): Other factors: ‘Ss.2cienesaci vied seed netvere: ve eae einvacienein ehenerec avers aus apevars ci ehe ee nterneee aero 238 5. Parasitism and vitellogenesis. oc oie: aceiciave oizck siin ara a ers ayere ae. apdeve a erie aiaiaaisaren pelle 240 HI -Oocyte maturation ye arisciscctoacrrsitte cane tesa arclacstuaides gunteaiactre acini eiieiie ae ee eee 240 EV OVA OM ana porecesici ng hie acai 2g eln catevel a Rede nyse SG coe Sie lately Sfove Gancde elapse ater eee hace eva a etane eee 242 V6 OOCYTE ACH VATION. 05.65 son. eie ole nestled ative sid bunsasacisbeys Riazese Ahetzan eunrPaicie aie eibes avout gieielane Fiv-a eon cep ne 243 A. Meiosis reinitiation of metaphase I-arrested oocytes ............. 00. cece eee eee eee 243 B. Fertilization potential o.c0.6 sig eudie wmode ote oepete ale set Gelet ane deters Seats cee ee 243 C.. Cortical reaction 63.8 f55 bis eb verses oa Ho nemnouloes od coemioemstene seetoua manent meer 244 VI. (OVIPOSINON. 5.59 n2nimeotiiin Sena ten soa eens ne Glee seh ae aaa uae ano uae a aucuemey eee 245 VII. Sex characteristics associated to sperm storage, mating and egg incubation .................. 245 A.. Genitalduct? sesiiesctsinedan sig seinen ctonieaealeairiays cinta corse Mahdaigiecis male lac mie mee Meter 246 B. External sex characteristics: ec csssccceasn os credence sf oe scian spoedmasinwe es soeeamamacar 246 Ie. (Differentiation asc f eisdiestenek cee astec tina eras ileventheuscieneuautyelnd cum ous calelanaserennrreteeree 246 2. (Regulation of development’ cicc.:. oli loss. cceroudlesiavocess eidisveiaue sveseinieiavsie-oveyaleis eiendioheds ei cjaxsreleters 248 VII. Sex recognition and mating behavior .... 2.0.02... 0.00. c cece cence tenet nent aes 249 References: euicaed ne ste adulee nes swags foulas bats ons ouamesemwarelepts toes mesdins pe neck baie otee ee 250 regulation need to be known not only for the goal INTRODUCTION As in most organisms, the series of complex events that render the female germ cell of crusta- ceans capable of conjugation with the spermato- zoon evolves over a long period which extends from the time of oogonial differentiation to the final maturation of the oocyte. It is now well-known that the study of female reproduction falls into five main areas: 1) the mechanism of ovarian differentiation, 2) the se- quence of morphological steps leading to vitel- logenesis, oocyte maturation and activation, 3) the endocrine regulation of the onset, completion and maintenance of these different steps, 4) the in- fluence of external factors such as photoperiod, temperature, ionic concentration of sea water on the female gametogenesis, 5) the events that follow mate selection and allow a specific response of the oocyte surface to the spermatozoon for a successful fertilization. Then, the normal growth of the embryo is ensured during incubation, a period of the life span of the female that is associated with the development of external sex characteristics and the secretion of pleopod tegumental glands in some species. We must point out that the major phenomena that characterize crustacean oogenesis and its of basic research but also for the benefit of the aquaculture field. Achieving control of reproduc- tion is often identified as a major problem that prevents the potential of shrimp farming from becoming a profitable industry. Thus, for their economical interest, a number of research pro- grams are now devoted to the Decapoda, one of the three most-studied Malacostraca orders beside the Amphipoda and Isopoda. A few general features related to the knowledge of malacostracan reproduction are recalled hereafter: 1. With the exception of Oxyrhyncha crabs which become sexually mature after a terminal molt, as the majority of insects, most malacostra- can Crustacea continue to molt after puberty. 2. In peracarids and natantian decapods, spawning is obligatorily preceded by a molt and, during the period of genital activity, the number of spawnings varies according to the species. 3. Except for penaeid prawns that are free- spawners, malacostracans incubate their eggs. As a matter of fact, the time interval between each spawning is always longer than the time of egg incubation. 4. In species in which the development com- prises larval stages, an ovigerous female carries Female Reproduction in Malacostracan Crustacea 219 around several thousands of small sized-eggs (a- bout 250 «m of diameter in the blue crab), whereas in species with abbreviated (direct) development a hundred or less eggs of larger size (about 3500 ~m of diameter in the European crayfish) are incu- bated. Indeed, the loss of eggs due to predators or unfavorable environmental conditions as particu- larly encountered by the eggs of penaeids, is considerably reduced in malacostracans which have an abbreviated development and spawn in protected areas. In this paper, we have tried to recapitulate the known processes undergone by the oocyte in order to acquire the capacity to generate the species. As in most animals, the sequence of major morpho- logical transformations that occur during crusta- cean oogenesis includes the differentiation and the evolution of oogonia into primary oocytes that undergo previtellogenesis, vitellogenesis and meiotic maturation. Then, activation follows ferti- lization and spawning, physiological events that permit the completion of gametogenesis. We have included the main features of the sex characteristics associated to the evolution of the female gamete as well as the regulation of these morphological events by hormonal and environ- mental factors. At last, the adjacent aspects related to sex recognition and mating behavior are briefly surveyed. It must be noticed that the most intensely investigated aspect of oogenesis concerns the vitellogenesis. The term “vitellogenesis” will be employed in this review in the way it is the most frequently found in the literature referring to the reproductive physiology of egg laying animals, i.e., as synonymous with “secondary vitellogenesis” [1]. We shall see (Section II, A, 1) that this conspic- uous event corresponds to a combination of extra- and intra-oocytic yolk production. Therefore, all previous steps called “previtellogenesis” and “primary vitellogenesis” by Dhainaut and De Leersnyder [1], as well as Charniaux-Cotton [2, 3] and Zerbib [4] can be grouped into a previtel- logenic phase. Moreover, it is necessary to be aware that the term “maturation” often found in aquaculture publications with the meaning of “ovarian growth” must be avoided for it corre- sponds to the resumption of meiosis following vitellogenesis (cf. Section HI). I. EARLY STEPS OF OOGENESIS AND PREVITELLOGENESIS A. Chronology and cytology The early steps of malacostracan oocyte growth were chiefly studied in the amphipod Orchestia gammarella and in few decapods from ultrastruc- tural observations [1, 2, 4, 5; reviews in 3, 6, 7]. The undifferentiated gonad of young genetic females forms the germinative zone of the ovary. This structure that resembles a network in which each gonium is completely surrounded by mesodermal cells [8] persists the whole life of the female. Oogonial mitoses take place exclusively in the germinative zone [9]. In gonochoristic decapods, differentiation of the ovary is characterized by a precocious functioning of the germinative zone, i.e., by a precocious initiation of oogenesis as compared with sperma- togenesis. Therefore, in the European crayfish, Pontastacus leptodactylus leptodactylus, oogenesis begins during the third postembryonic stage, while at the seventh stage the testes contain gonia not yet engaged in spermatogenesis. A similar delay between male and female gametogenesis also occurs in crabs and in the penaeid shrimp Penaeus japonicus [10-12]. In Talitridae amphipods, there seems to be a slight precocity in oogenesis in comparison with spermatogenesis. Such a precocity is clearly visible in males whose testes display an ovarian region. Some oogonia leave continually the germinative zone [13] by a mechanism as yet unknown and rapidly enter prophase of the first meiotic division up to diakinesis. Then, they become primary oocytes with condensed chromosomes which appear in synaptonemal complexes at the ultra- structural level. The decondensation of chromo- somes is accompanied by marked cytoplasmic changes such as an accumulation of free ribosomes and the differentiation of a rough endoplasmic reticulum (RER). These phenomena characterize the beginning of the _ previtellogenesis. Mesodermal tissue forms around each oocyte a follicle multilayered epithelium. Follicle cells are 220 J.-J. MEUSY AND G. G. PAYEN connected with one another by desmosome-like cell junctions and are themselves holded by hemidesmosomes on the basal lamina [14]. When endogenous glycoproteins accumulate in the numerous RER vesicles, oocytes carry out the “endogenous vitellogenesis” [1]. Simultaneously, the oocytes acquire a vitelline envelope and their surface becomes irregular with the formation of short microvilli and a few micropinocytotic vesi- cles. Oocytes grow continuously until they reach a diameter typical for the species. Oogenesis stops at the end of this step in young females and during genital rest in puberal females [15, 16]. As a general rule, female genital puberty is realized when a one layered-epithelium surrounds for the first time each fully grown previtellogenic oocyte (cf. Section II, A). B. Regulation 1. Mechanisms of oocyte differentiation and onset of oogenesis The hypothesis of a spontaneous ovarian dif- ferentiation, or ovarian autodifferentiation, of the gonadal rudiment in the absence of diffusing androgenic hormone (AH) was stated for the first time by Charniaux-Cotton [17]. This hypothesis was based on the observation of a precocious development of an anterior ovarian region before the onset of spermatogenesis in the gonads of males of Orchestia mediterranea, Talitridae amphi- pods (rudimentary hermaphroditism). Gonia of the anterior region are less subjected to the AH —the androgenic glands (AG) are located posteriorly — and differentiate into oocytes that acquire follicle cells and grow until the end of previtellogenesis. Posteriorly to this region, gonia give rise to the various stages of spermatogenesis. Experimental proofs of an ovarian autodifferentia- tion in amphipods were then obtained in O. montagui after ablation of the AG by Charniaux- Cotton and Ginsburger-Vogel [18] and in Talitrus saltator after implantation of testes into males of O. gammarella from which AG have been re- moved [19]. In O. gammarella, the transformation of testes into ovaries is possible if the testes are protected from the action of AH, before the onset of spermatogenesis that occurs at the second intermolt. This is accomplished by implantation of gonads from young males into females. If im- planted before the beginning of spermatogenesis, the young testes can develop into ovaries; if implanted later, the testes do not transform but instead degenerate [20]. Ovarian autodifferentiation has been also dem- onstrated in the oniscoid isopod Helleria brevicor- nis following implantation of undifferentiated gonads deprived of AG rudiments [21]. Among decapods, the proterandric _her- maphroditic shrimps, such as Pandalus borealis and Lysmata seticaudata, give a good proof of the ovarian autodifferentiation (review in [22]). Thus, when the AG degenerate at the time of sex- reversal, or after their ablation (andrectomy) during the male phase, oogenesis spreads into the gonad. Another proof of ovarian autodifferentia- tion has been obtained in the shrimp Macro- brachium rosenbergii [23]. In young males in which the gonads contain only gonia, andrectomy is followed by differentiation of normal ovaries with oocytes in previtellogenesis and development of oviduct rudiments. Several natural data confirm the inherent tendency of gonia in genetic females or males to effect oogenesis. Thus, female gametogenesis appears not only in the gonad rudiment of young males of several Talitridae, as O. mediterranea and O. cavimana [24], but also in the testes of mature males of gonochoristic decapods. For instance, the testes of the crayfish Pontastacus leptodactylus leptodactylus, exhibit oogenesis of variable intensi- ty during genital rest. At this time, when the AG are very small, spermatogenesis stops and oocytes sometimes appear in different parts of the testes [11]. Thus, oogenesis may occur in some testicular acini as soon as the gonia receive an insufficient quantity of AH, or none at all. It results in the formation of normal primary follicles but this oogenesis always stops at the end of previtel- logenesis (cf. Section I, B, 3, a). To summarize, ovarian differentiation of the gonadal rudiment is an autodifferentiation. It concerns oogenesis from gonia to the end of previtellogenesis and takes place spontaneously in the absence of any hormone, female or male. This Female Reproduction in Malacostracan Crustacea 221 proves the “emerging inherent tendency to de- velop into an ovary” as in mammals ([25], p. 39). Initiation of oogenesis does not appear to be controlled by a neurohormone. In crabs, it is not accelerated by removal of eyestalks from larvae and from very young females, whereas a neurohor- mone from eyestalks regulates the initiation of spermatogenesis, through its moderating control of the AG [26, 27]. 2. Maintenance of the germinative zone The germinative zone of the ovary, in contrast to that of the testis, does not require the presence of a neurohormone for its maintenance. Thus, in the shrimps Palaemon serratus and Crangon crangon, after cauterization of the median zone of the protocerebron or culture of isolated ovaries, the gametogenic activity of the germinative zone persists [28-30]. Likewise, sacculinid rhizocepha- lans, through contact and at some distance, cause the destruction of neurosecretory regions of the host crabs in both sexes. However, the germina- tive zone of parasitized female crabs is not modified, while the one of parasitized males degenerates [31, 32]. 3. Oocyte growth until vitellogenesis As already mentioned, oogenesis up to the end of previtellogenesis is a continuous phenomenon. Some studies have shown that the continuous phase of oogenesis is regulated by a moderating neurohormone. In the juvenile freshwater crab Eriocheir sinensis, ablation of the eyestalks brings on an increase in the synthesis of DNA in the germinal cells that is expressed by an increase in the number of oogonial mitoses and in the number of oocytes entering into prophase of meiosis [33]. Molting hormone interferes little in oogonial mitoses and in previtellogenesis, as is shown by the ablation of Y-organs from the juvenile and pre- puberal crabs [34]. Thus, small quantities of ecdysone which remain in serum after the ablation seem sufficient to allow a normal oogenesis in these destalked crabs. On the other hand, when Y-organ and eyestalks are removed simultaneously, most of the previtel- logenic oocytes degenerate. It appears that re- peated injections of 20 OH-ecdysone are necessary to restore a normal previtellogenic growth in the destalked crabs [34]. Ablation of eyestalks in 1-year-old female Para- telphusa_ hydrodromous, during their post- oviposition period, seems to accelerate previtel- logenesis, leading to an early vitellogenesis. At that time, the ovaries normally show empty folli- cles. This result is given as an argument in favor of an inhibitory control of previtellogenesis by eye- stalks [35]. The protein synthesizing capacity of previtel- logenic ovaries of Uca pugilator has been tested for 24hr in vitro. In presence of neuroendocrine tissues such as eyestalk or thoracic ganglion, as well as cyclic AMP (10-°M), the rate of incor- poration of radioactive leucine into protein by the ovary is inhibited [36]. Since cyclic AMP appears to mimic eyestalk tissue (well-known to have an inhibitory effect on oocyte growth, as recalled in Section I], B, 1), the decrease in protein sythesis is attributed by the authors to “changes in cyclic nucleotide levels”. No clear interpretation con- cerns the inhibitory effect of the thoracic ganglion. The lack of a specific component from the medium would explain this unexpected effect occurring instead of the stimulation observed in vivo. Il. VITELLOGENESIS A. Vitellogenesis process 1. General considerations Vitellogenesis is the step of the crustacean reproduction during which oocytes accumulate a large amount of yolk, especially — but perhaps not exclusively — by internalization of an extraovarian precursor named vitellogenin. It affects synchro- nously all the elder oocytes, i.e., all the oocytes which have reached the end of previtellogenesis. Such a process is common to many groups other than Crustacea particularly in insects, amphibians, fishes and birds. From several aspects, vitellogene- sis is a very important step of the female reproduc- tion: —Most of the endocrine controls on reproduc- tion known at the present time apply on vitel- logenesis. —Contrary to previtellogenesis, vitellogenesis is 222 J.-J. MEUSY AND G. G. PAYEN not a continuous process: it is inhibited during non-breeding season and, for some species, in artificial conditions. So, it is easy to understand that the aquaculture services play a special atten- tion to the control of vitellogenesis mechanisms. —The ability to carry out vitellogenesis is the criterium on which the puberty concept was built in female crustaceans. Crustacean vitellogenin is a high molecular weight protein associated with lipidic, glucidic and carotenoid prosthetic groups to which very few studies have been devoted. When vitellogenesis takes place, the presence of carotenoids linked to vitellogenin brings on a bright color of the ovary in most species. So, it is quite easy to know, without dissecting the animals, whether vitellogenesis has begun in the species whose exoskeleton is transparent, such as most prawns and shrimps. 2. Origin of vitellogenin The existence in the haemolymph of vitellogenic females of a “female-specific-protein” — the early name for vitellogenin when the physiological significance of this protein was not firmly stated — was reported for the first time in Crustacea by Frentz [37] in the crab Carcinus maenas. This observation and the role of vitellogenin in the vitellogenesis process as the major precursor of yolk was confirmed in several other species during the following years [38-43 for review]. The site of vitellogenin synthesis in Crustacea was known quite lately and the question is not yet completely elucidated. The first hypotheses con- cerned the hepatopancreas [44] and were sup- ported by the early observations on the transit of carotenoid pigments from this organ to the ovaries (e.g., [45]). However, nothing suggested that the proteinic part of vitellogenin has the same origin than the carotenes associated with it. Kerr [46] cultured different tissues and organs from the crab Callinectes sapidus — muscle, heart, hepatopancreas, total haemolymph and serum - in the presence of ‘“C-leucine and analyzed the protein released in the medium by column chroma- tography and electrophoresis. The author found some suggestion for the hemocytes as site of vitellogenin synthesis but the results did not seem conclusive. In Uca pugilator and Libinia emarginata Wolin et al. [47] reported a complete immunochemical identity between vitellogenin and a protein from the hepatopancreatic extract. | Nevertheless, haemolymph could have contaminated the extract. Recently, Paulus and Laufer [48], using immuno- histochemical technics for the study of the crabs Libinia emarginata and Carcinus maenas, localized vitellogenin in hepatopancreatic specialized cells they called vitellogenocytes. According to the authors, these cells are contained in small haemal sinuses between the hepatopancreatic tubules and can be found in association with some other tissues, especially connective tissue. They may be similar to the adipocytes which are considered by other authors as the site of vitellogenin synthesis ([63], see further). Lui et al. [49-51] incubated ovaries of the crayfish Procambarus sp. and of the crab Pachy- grapsus crassipes in a *H-leucine medium or a mixture of tritiated amino acids up to 48 hr. After denaturation and electrophoresis, they demon- strated that radioactivity was present in the main polypetide subunits of Procambarus vitellogenin and in all the three subunits of that of Pachygrap- sus; so they concluded that the ovaries are the source of vitellogenin. Unfortunately, the authors have not cultured other organs than ovaries as controls. Using similar methods to those of Lui et al., Eastman-Reks and Fingerman [52] drawn the same conclusion about the ovary of the crab Uca pugilator. As the preceeding authors, they did not attempt to incubate other tissues. In the kuruma prawn, Penaeus japonicus, Yano and Chinzei [53] reported also that “ovary is the site of vitellogenin synthesis”. These authors incubated ovaries and hepatopancreas — but no fat body - in Ringer solution containing labeled amino acids. Protein synthesized by the ovary and precipitated with anti-vitellin serum was shown by electrophoresis and fluorography to consist of two polypeptides corresponding to the components of vitellogenin. No immunoreactive material was found in the hepatopancreas and its incubation medium. Some ultrastructural studies of the vitellogenic oocyte gave indications in favor of both intra- and extraoocytic sources of yolk (in the spider crab, Female Reproduction in Malacostracan Crustacea 223 Libinia emarginata, the isopod, Oniscus asellus, the terrestrial hermit crab, Coenobita clypeatus, [54-56]). In an amphipod Crustacea, Orchestia gam- marella, Junéra et al. [57] showed that vitellogenin synthesis did not stop immediately after bilateral ovarietomy, as would be the case if the ovaries were the site — or, more precisely, the exclusive site — of vitellogenin synthesis: it only stopped 5 to 8 days after the operation for reasons which will be discussed further on (cf. Section H, B, 2, b). In 1980, Picaud and Souty using double diffusion technique and autoradiography, demonstrated that fat body from Porcellio dilatatus incubated with a '4C-leucine medium synthesized vitel- logenin. Junéra and Croisille [58] and Croisille and Junéra [59] made the same inference in O. gammarella but, in this species, the subepidermal adipose tissue only seems to be involved in vitellogenin synthesis. In the shrimp Palaemon serratus, Meusy et al. [60] demonstrated also by immunohistochemistry the presence of vitel- logenin in the same structures of the vitellogenic females; the hepatopancreas was not labeled. Similar results were obtained later in two other decapods, the penaeids Penaeus japonicus [61] and Parapenaeus longirostris [62]. The adipocytes from O. gammarella display ultrastructural modifications when vitellogenin synthesis takes place [63]. They acquire a well- developed rough endoplasmic reticulum, the space in the cell occupied by /-glycogen and lipid droplets significantly reduces and the vitellogenin is detectable in dense bodies by the peroxidase- antiperoxidase method. Furthermore, when vitel- logenin synthesis stops following a total ovariec- tomy, the adipocytes acquire the ultrastructural features of non-vitellogenic or male adipocytes [64]. Though some of these studies are seemingly contradictory, it should be noted that, in some insects — Drosophila and few others —, not only fat body but also follicle cells of the ovary are able to synthesize vitellogenin [65-67]; such a possibility of a double origin of vitellogenic material may exist also in Crustacea or in some orders of Crustacea. Moreover, the results about “vitel- logenocytes” associated with the hepatopancreas [48] and those about “adipocytes” may not be inconsistent, since these two cellular types would be homologous. Incubation of fat body, ovary and hepatopancreas is a very hazardous method since the maintenance of the integrity of these tissues/ organs during the process is not reliable. Particu- larly with the hepatopancreas, the release of proteolytic enzymes into the incubation medium is difficult to avoid. So, an attractive approach of this problem would be to look for messenger RNA coding for vitellogenin in these various tissues/ organs. 3. Vitellogenin uptake by vitellogenic ovaries a) Transformations and role of the follicle en- velope At the onset of vitellogenesis, each oocyte is surrounded by a follicle envelope which comes from a permanent tissue: the follicle tissue from the eggs which have been laid is utilized again for setting up the new follicles [3, 68-71], except in the isopod, Idotea balthica basteri, in which it seems to degenerate just prior to oviposition [72]. At the beginning of vitellogenesis, a tubular network has been observed in the cells of the follicle envelope of four species of Palaemonidae (Palaemon adspersus, Macrobrachium rosenbergii [70, 73, 74], Palaemonetes_ varians and Palaemon_ serratus (Jugan, unpublished)). These tubules, character- ized by a diameter of 0.15 ~m, are bound by a single membrane and enclose a granular electron- dense material. They connect up all the extracellu- lar compartments: haemolymph, intercellular spaces, space between the oocytes and the follicle epithelium. After incubation of ovaries in a peroxidase containing medium, the diaminobenzi- dine reaction product was seen in the tubular network and in all these compartments. Peroxi- dase penetrated also into the vitelline membrane and in some pinocytotic vesicles of the oocytes [73, 74]. The tubular network regresses at the end of vitellogenesis [70, 75]. This structure which has been also described in copepods [76], makes easier the passage of substances from haemolymph to vitellogenic oocytes. In Idotea balthica basteri, where no tubular network has been described, some features seem to play the same role as in Palaemonidae shrimps: 224 J.-J. MEUSY AND G. G. PAYEN the cells of the follicle envelope acquire oocyte oriented villi, tight junctions appear between follicle villi and oocyte microvilli, and the spaces between follicle cells become very wide [72]. Beside its role of interface, the follicle envelope has an endocrine function. Charniaux-Cotton [77, 78] has demonstrated in females of O. gammarella that the vitellogenic ovary controls a secondary sexual characteristic which is temporary and appears during vitellogenesis: the long ovigerous setae on oostegites, the role of which is connected with incubation. The follicle cells are presumably the source of this ovarian hormone, though this has not yet been established (cf. Section VII, B, 1 and 2). b) Vitellogenic mechanism At the beginning of vitellogenesis, microvilli develop towards the follicle cells [54, 55, 79, 80]. In M. rosenbergii, some of them have been described penetrating deeply in tubules of the follicle cells [73, 74]. A glycocalyx, or cell coat, covers the external surface of the microvilli. In addition to microvilli, macrovilli have been also observed in an amphipod, O. gammarella [80] and in some decapods ([81]; Lysmata seticaudata, Zerbib, unpublished). These micro- and macrovil- li increase considerably the oocyte surface and probably its exchange ability. Endocytotic vesicles, 100-140 nm in diameter, appearing at the surface of the cortical ooplasm, have been described in many species. In some of them, their content seems to be drained towards yolk spheres by a network of microcanalicules, 45- 60 nm in diameter in Orchestia gammarella (Fig. 1) and in the crayfishes Astacus astacus and A. leptodactylus, [4, 81]. It has been demonstrated, by incubating oocytes in a horseradish-peroxidase containing medium [82, 83] or by using fluores- ceine isothiocyanate conjugated vitellogenin [47] or tritiated vitellogenin [378], that these structures are related to an endocytotic — and not exocytotic ~ process. Recently, Jugan and Soyez [84] conju- gated vitellin of Macrobrachium rosenbergii with colloidal gold and observed a labelling at the surface of the microvilli, on endocytotic vesicles and yolk spheres. Jugan [75], working on M. rosenbergii demonstrated that vitellogenin inter- oocyte and — endocytosis nalization in Crustacea is a receptor mediated process. The receptors have a high affinity (Kp=3.5X10~%) and are very numerous (about 10'° receptors per oocyte). The yolk spheres grow in size by fusing together and are pushed towards the medullar ooplasm by those more recently formed. At the end of vitellogenesis, they take a polyhedric shape and measure up to about 40 um (“yolk platelets”). It has been shown that yolk spheres contain a lipo-glyco-carotenoproteic material. Lipid drop- lets have been also observed during vitellogenesis, but the origin of their content still remains undetermined (cf. comments in: [6], pp. 472-473). The proteins and lipids represent the major enrich- ment of the ovaries during vitellogenesis ([85] cited in [6]). The sequestration of vitellogenin by the oocytes is a specific feature of vitellogenesis. Nevertheless, a rough endoplasmic reticulum is still present during this phase and it seems likely that in- traoocyte synthesis of proteinaceous material con- tinues [4, 86-88]. Moreover, transfer of nuclear material to the ooplasm, as a possible prelude to protein synthesis, has been reported [89, 90]. As suggested by Adiyodi and Subramoniam [6], the relative emphasis on autosynthesis and _ heter- osynthesis probably varies with species. Some time before the end of vitellogenesis, microvilli (and also macrovilli in the species where they are present) regress and the endocytotic phenomena disappear (in the isopod Jdotea bal- thica, [72] and in M. rosenbergii, (Jugan, personal communication)). The oocytes are overloaded with yolk spheres and lipid droplets, except in the cortical and perinuclear ooplasm. Cortical vesi- cles appear and seem to be related to the formation of the fertilization envelope (in O. gammarella [91]). For Goudeau and Lachaise working on Carcinus maenas [92], these cortical granules would originate from the “endogenous yolk” (cf. Section V, C). 4. From vitellogenin to vitellin: a processing? When vitellogenin, previously termed “female specific protein”, enters the oocytes, it is usually named vitellin (or lipovitellin). With a historical regard, it seems that these two different names Female Reproduction in Malacostracan Crustacea 225 Fic. 1. referred principally to the two compartments, haemolymph and oocytes, where these substances were found. Little was known about the chemical structure of vitellogenin and vitellin respectively. The “female specific protein” of the haemolymph, 1.e., vitellogenin, was initially char- acterized as an electrophoretically slow moving protein [37]. In the following years, the relation of vitellogenin to vitellogenesis was firmly established (cf. for review [93]). The presence of associated lipids (Sudan Black staining), carbohydrates (PAS positiveness) and carotenoids (pigment extraction and absorption spectrum study) was demonstrated and vitellogenin was consequently identified as a lipo-glyco-carotenoprotein (e.g., the early works: Active endocytosis during vitellogenesis in the amphipod, Orchestia gammarella, and relationship between the oocyte and the follicle cells (FC) (courtesy of C. Zerbib). ev: endocytotic vesicles; FC: follicle cells; mc: microcanalicules; mt: microtubules; Mv: macrovilli; mv: microvilli; YB: yolk body. in the crabs Paratelphusa hydrodromous [39], Carcinus maenas [94] and Callinectes sapidus [42]). The carotenoids give a bright color — varying according to the species — to the vitellogenin and vitellin and, consequently, to the vitellogenic oocytes. They are provided by the food and are not synthesized by the animal itself. It seems probable that they have a screening function against light (review in [95]; [96]). The molecular weight (MW) of the vitellogenin in Crustacea was reported in the amphipod O. gammarella: 397 +27 kD [57], and in the isopod Porcellio dilatatus: 315+54 kD [97]. Vitellin, the major constituant of yolk, is also a lipo-glyco-carotenoprotein. It contains between 28 226 J.-J. MEUSY AND G. G. PAYEN and 35% of lipids [98, 99] and about 4.8% of sugars [100]. On the basis of double-diffusion tests or related techniques, no immunological differ- ence between vitellogenin and vitellin has ever been demonstrated in any species [42, 47, 99, 101- 106]. The MW of vitellin is not very different from that of vitellogenin in the species where both have been determined [57, 97, 107]. The amino acid composition of the vitellin of some species has been established ([50, 51, 99, 107-109]; the results are compared in the review [6]), but no compari- son with vitellogenin is available; so, these data bring no indication on a possible processing. Treatment of the vitellin by denaturing agents revealed in several species the presence of two polypeptide subunits with close MW of about 100 kD (Palaemon adspersus, Uca pugilator, Homarus gammarus, Macrobrachium rosenbergii [52, 96, 110, 111]; Penaeus japonicus, MW not determined [53]) or less (Parapenaeus longirostris, 45 and 66 kD, [106]). In the prawn, Macrobrachium rosen- bergii, the vitellogenin and the vitellin have been both studied and exhibited the same two subunits of 84 and 92.2kD MW [111]. In some other species, numerous fractions have been visualized (O. gammarella, Procambarus sp., Squilla mantis, Penaeus japonicus [50, 107, 112-114], but it seems probable that only few of them are native polypeptide subunits. Although the possibility of a processing of the vitellogenin when, or after, entering the oocytes has been considered, especially with regard to the proteinic part of this yolk precursor, few informa- tions are yet available. It is likely that vitellogenin and vitellin, if not identical, are very closely related substances. 5. Vitellogenin synthesis and vitellogenin level in haemolymph as means for monitoring vitellogenesis A first attempt to know whether there is a close relation between the vitellogenesis process and the vitellogenin metabolism was carried out by inject- ing tritiated leucine to vitellogenic females of Orchestia gammarella at various steps of the reproductive cycle [115, 116]. The diagram (Fig. 2a) shows that the amount of radiolabeled vitel- logenin in the haemolymph is growing from the beginning to the 3/4 of the cycle, though endocyto- sis is maximal during this period (except at the very beginning of the cycle). This amount falls down during the last quarter of the cycle, though endocytosis, as seen by electron microscopy in several species (Idotea balthica basteri [72], Palaemonetes varians and Macrobrachium rosen- bergii (Soyez and Jugan, personal communica- tion)), become negligible at this period. This result has been confirmed by in vitro incorporation of '*C-leucine by the fat body of an isopod, Idothea balthica basteri (Fig. 2b)[117]. In addition to this statement, a diurnal rhythm of vitellogenin release was observed in vivo in another isopod, Porcellio dilatatus [118]. Other haemolymphatic and ovarian proteins seem also to be subjected to circadian variations [119-121]. In the lobster, Homarus americanus, where the reproductive cycle is not easy to study because it lasts about one year or more, it has been shown that the level of circulating vitellogenin, as meas- ured by electrophoregram scanning, “is always highest well prior the maximum accumulation of yolk in the oocytes, and the levels dropped off markedly prior to oviposition” [105]. In the freshwater prawn, Macrobrachium rosen- bergii, whose reproductive and molting cycles are short (about 3 weeks) and concomitant, as those of O. gammarella, an ELISA titration of circulating vitellogenin has shown that the vitellogenin level, very low at stages A and B, increases during stage C, i.e., during the period of intense uptake of vitellogenin by the oocytes, remains at a high level during stages Do-D, and fall down thereafter, though vitellogenesis is not still achieved (Fig. 3) ({111]; Derelle and Meusy, unpublished data). At the end of the molting/reproductive cycles, before and just after exuviation, the vitellogenin level is very low again. It will go up after oviposition if a new vitellogenesis takes place again. It is notewor- thy that during the period of rapid decrease of the vitellogenin level, the vitellogenic oocytes display no more endocytosis but, nevertheless, their vitel- lin content increases up to oviposition. This observation is an indication for a vitellin synthesis by the oocytes themselves. These studies, carried out on various species, firmly established that vitellogenin synthesis and Female Reproduction in Malacostracan Crustacea 227 CPM x 10°5/3 1 hl/6h A tBe Cc, Cp: Ge Do vitellogenin (CPM x10 5/100 ig proteins) ABC, Cy C3 Fic. 2. exuviation spawning Di ‘abc D>» A B molting cycle b Dg Dyy-y-Do 4 A molting cycle (days) (a) Vitellogenin synthesis in the amphipod, Orchestia gammarella, during the molting cycle. Six hours before sampling of the haemolymph, the animals received an injection of 2.5 “Ci of *H-leucine. The radioactivity of the vitellogenin was determined after separation by polyacrylamide gel electrophoresis of the serum proteins and corresponds to 3 sd of haemolymph (from [116, 347]). (b) Relationship between the incorporation rates of '*C-leucine by incubated fat bodies and the ovarian cycle (or molting cycle) of the isopod, /dotea balthicu basteri (from [117]). EX: exuviation. vitellogenesis are closely correlated. Some impor- tant aspects of the mechanisms of control begin now to be elucidated. 6. Timing of the reproductive cycle: duration and relation with the molting cycle The duration of the female reproductive cycle in malacostracan Crustacea generally reduces when temperature increases and is very different from one species to another. For instance, it lasts about 3-4 weeks in the amphipod Orchestia gammarella, reared at the laboratory temperature, and in the prawn Macrobrachium rosenbergii at 27°C (observations made in our laboratory), several Ve) oO oO WW Ww _ J.-J. MEUSY AND G. G. PAYEN 160 140 120 100 total proteins (mg/ml) CR «—exuviation 8 Do ODI’ n=11 > ib Cc Cc Cc c molting cycle 228 8 ° 7 = 6 ~ >?) E 5 Y) = 4 i— o elke — 2 os =, 4. 1 O (@) |} A B G-3 C-6 G=11 C-19 aM = = bes " WW l " " WW c c (= c Cc Cc Fic. 3. Variations of circulating vitellogenin and total protein titres during the molting cycle in vitellogenic female prawns, Macrobrachium rosenbergii, of the same size. The titres of vitellogenin were determined by indirect ELISA and total proteins by Lowry’s method (Derelle and Meusy, unpublished data). Bars: standard error of the mean; n: number of animals for each molting stage. months in many species and about one or two years in the lobster, Homarus americanus [105, 21. Various features can be found concerning the relation between vitellogenesis and molting cycle. In some species, vitellogenesis takes place during one intermolt and egg laying occurs just after the exuviation (for instance, in the amphipod, O. gammarella [123], the isopods, Porcellio dilatatus and Idotea balthica [124, 125], the decapods, Lysmata seticaudata and M. rosenbergii {16, 69]). In some other species, vitellogenesis can take place during more than one molting cycle, according to the season (for instance, in the decapods, Palaemon serratus and Athyaephyra desmaresti (126; 127]). It is noteworthy that the molting cycle generally lasts a longer time during the reproductive season than during the genital resting period, because vitellogenesis lengthens the cycle [123, 126]. In all these above malacostracans, egg laying takes place just after the exuviation. This is not the case of the crab, Carcinus maenas, whose vitellogenesis occurs only during the intermolt stage C, and which lays its eggs before premolt stages (Do to Dz), i.e., a long time before the exuviation [128]. In the crab, Uca pugilator, Webb [129] gives the following sequence of events: vitellogenesis — oviposition — incubation — hatch- ing — molt. In the stone crab, Menippe mercenaria, several spawnings may occur within a single intermolt [130]. A very particular feature is that of few malacos- tracans which do not molt their whole life and become pubescent after their last molt, called “puberty molt” (cf. Section II, B, 2, d). In conclusion, the relationship between vitellogenesis and molting exhibits in Crustacea many different Female Reproduction in Malacostracan Crustacea 229 features and seems to have supported a long and divergent evolution. B. Vitellogenesis control 1. Inhibitory control by VIH (Vitellogenesis In- hibiting Hormone) The first control to be known was inhibitory and its source is located in the central nervous system (cf. Fig. 4 for schematic representation of the main endocrine controls of vitellogenesis). a) The X organ-sinus gland complex Eyestalked species The works of Han- strom, who discovered neurosecretory cells in the eyestalks of some species of stomatopods and decapods — the X organ or “Hanstr6m’s organ” — and a connected neurohaemal organ — the sinus gland [131-133] -, marked the beginning of the EYESTALK EXTERNAL FACTORS MANDIBULAR ORGAN JUVENOIDS ? NERYOUS SYSTEM ECDYSTEROIDS Fic. 4. VITELLOGENIN SYNTHESIS SITE modern studies on crustacean endocrinology. At this time, the concept of neurosecretory cells was new: it was brought out only few years ago by Ernst Scharrer [134] from the observation of the hypothalamo-hypophyseal complex in Teleostei. X organ is contained in the medulla terminalis of the optic lobes (protocerebrum), and consists of perikarya whose axons end in the sinus gland. The sinus gland, opalescent looking, is not really a gland but a neurohaemal organ. It stores and releases by exocytosis materials mainly from the X organ and contains no cell, except glial cells ({135- 138]; review in [139]). The role of the X organ-sinus gland complex was demonstrated by Panouse [140, 141], in the shrimp Palaemon serratus. This author observed that eyestalk ablation induces an acceleration of the molting cycle and a rapid growth of the ovary. He did not specify what stage of oogenesis was OOGONIAL MITOSES VITELLOGENIN VITELEGSENESIS Schematic representation of the main endocrine controls of oogenesis in malacostracans. VIH: vitellogenesis inhibiting hormone, VSH: vitellogenesis stimulating hormone; VSOH: vitellogenin stimulat- ing ovarian hormone. 230 J.-J. MEUSY AND G. G. PAYEN specially affected by this ablation and he thought that the two effects, on molting and ovogenesis, could result from the suppression of the same hormone (“anti-auxinic effect of the eyestalk hormone”). The results of Panouse’s experiments, classically referred as “the Panouse effect”, found their applying in aquaculture. In some species, the breeders do not carry out vitellogenesis in artificial conditions: a unilateral eyestalk ablation is usually practiced to trigger vitellogenesis [142]. A bilater- al ablation is not required and has some disadvan- tages varying with species: for instance, it may shorten the life of the female and/or bring about some abnormalities of vitellogenesis [143, 347]. An extensive bibliography of the early works on the anatomy and physiological functions of the X organ-sinus gland complex is given in the book of Gabe [144] and a more recent review has been produced by Chaigneau [139]. It is noteworthy that eyestalk ablation often promotes either vitellogenesis or molting, accord- ing to the species, state of the ovaries, age of the animals, and temperature. The idea arises of a molting-vitellogenesis antagonism, though the real mechanism of this antagonism, hormonal or meta- bolic, remained unknown [35, 145, 146]. This hypothesis was backed up by the observation of the females of some Oxynrhyncha which molt during a limited part of their life and whose reproduction begins after the last molt and the degeneration of the Y-organ (molt organ): Pisa tetraodon, Libinia emarginata [147, 148]. While many other hormonal effects of the X organ-sinus gland complex were discovered and studied, i.e., on glucidic and lipidic metabolism, water balance, and chromatophores, most of the authors thought that vitellogenesis and molting are controlled by two distinct hormones, the Molt Inhibiting Hormone, MIH [149], and the Ovary Inhibiting Hormone, OIH [150]. The early ultra- structural observations of the sinus gland were in agreement with the hypothesis of several hor- mones (for review [139]), though the typing of the neurosecretion granules only took into account morphological criteria. It is clear enough that the number of granule types cannot be directly related to the number of alleged hormones. It is now established that the “Ovary Inhibiting Hormone” acts mainly on vitellogenesis and is responsible for the sexual rest. So, the name of “Vitellogenesis Inhibiting Hormone” (VIH), proposed by Char- niaux-Cotton and Touir [16], seems more ade- quate and precise. Eyestalkless species The whitish and opalescent aspect of the sinus gland makes it quite easy to identify the gland in the vicinity of the optic lobes in eyestalkless species (review in [139]). In contrast, the identification of a structure homolo- gous to X organ is much more difficult. In the isopods which have no medulla termina- lis, connections between neurosecretory cells of the brain and the sinus gland were found in Porcellio dilatatus [151], but the search for an X organ equivalent was mainly carried out by elec- tive destruction of parts of the protocerebrum and optic lobes. Most authors located the source of VIH in the median part of protocerebrum (in Idothea balthica and Ligia oceanica (125, 152, 153]). In the amphipod, Orchestia gammarella, electro- coagulation of the antero-median part of the protocerebrum prevents the onset of vitellogenesis and this zone can be considered as stimulatory [116]. A VIH or a VIH-like substance seems to be secreted by some other part of the brain: a supernumerary brain grafted into females of this species inhibits vitellogenesis [154]. According to the author, the graft, which is deprived of external influences, would secrete continously the inhibit- ing hormone. The concerned neurosecretory cells remain to be found in this order. b) Ways of action Control of vitellogenin synthesis When the concept of vitellogenin as the haemolymph precur- sor of vitellin became established, it appeared likely that the inhibitory action of VIH on vitel- logenesis could act via the control of vitellogenin synthesis. Frentz [37] and Shade and Shivers [83] reported indications favorable to this hypothesis. Meusy ef al. [60], injecting tritiated leucine to female shrimps, Palaemon serratus, showed that the ablation of eyestalks triggers vitellogenin synthesis. This result was confirmed and extended by in vitro experiments in the isopod, Porcellio dilatatus: extracts of sinus glands from non- vitellogenic females display a direct inhibitory Female Reproduction in Malacostracan Crustacea 21 effect on vitellogenin synthesis by the fat tissue [155]. Control of vitellogenin uptake by the oocytes Unpublished observations on the amphipod, O. gammarella, by Meusy and Junéra suggested that the vitellogenin uptake might be hormonally controlled. Females do not usually lay eggs if mating has not occurred, for instance, in the absence of male. In this circumstance, a resorption of the non-laid oocytes is observed and, conse- quently, a very large amount of vitellin is detected in the haemolymph [101, 102]. Though vitellin could be used for a new vitellogenesis, in place of vitellogenin, as it has been proved by injecting radiolabeled vitellin in a vitellogenic female (Meusy, unpublished data), it happened that some of these females enter in the resting period, especially if the experiment was carried out at the end of autumn or at the beginning of winter. Similar observations were conducted on the prawn, Macrobrachium rosenbergii, when females were experimentally prevented from egg laying. Direct evidence for a hormonal control of vitellogenin uptake by the oocytes of the prawn, M. rosenbergii, has been related by Jugan and Soyez [84]: a sinus gland extract inhibited the binding of colloidal-gold labeled vitellin on oocyte microvilli (Fig. 5). In preliminary studies using peroxidase-labeled vitellin, Jugan [75] reported that the affinity of VIH for the receptors to vitellin would be higher than that of the vitellin itself. c) Extraction and purification of VIH Though some other eyestalk hormones have been isolated in the seventies [156-158], the first attempt at purification of VIH was published only in 1981 by Bomirski et al. [159]. In a preliminary study, Klek-Kawinska and Bomirski [160] realized aqueous extracts of eyestalks of the shrimp, Crangon crangon, and tested their activity on destalked females of the same species. They found that the hormone is apparently absent during the early part of the breeding season. Later on, Bomirski et al. [159] dialysed, boiled and filtrated on Sephadex G-25 gel the eyestalk extracts from Cancer magister before testing them on destalked females of Crangon crangon. They concluded that VIH - they called GIH, i.e., Gonad Inhibiting Hormone -, is heat stable, dialyzable and has a molecular weight of about 2000 Daltons. The thermostability was confirmed in the spiny lobster, Panulirus argus [146]. Quackenbush and Herrn- kind [161], after extraction in phosphate buffer, pH 6.8, separated VIH and other peptides from the eyestalks of the spiny lobster, using Sephadex G-25 gel and bioassayed the fractions in eyestalk- less female fiddler crabs, Uca pugilator. According to these authors, this neuropeptide has an appar- ent molecular weight near 5kD and is different from the Molt Inhibiting Hormone, MIH, which did not induce gonadal inhibition. In a recent abstract, Quackenbush and Keeley mentioned a lighter MW for the GIH-VIH of the shrimp Penaeus vannamei: 3.3 kD [162]. More recently, Soyez et al. [163] extracted proteic material from isolated sinus glands of the lobster, Homarus americanus, with 0.1 N hydro- chloric acid and purified the active factor by a two step reversed phase high performance liquid chro- matography procedure. A bioassay, operated on destalked females of the shrimp, Palaemonetes varians, and an SDS-urea polyacrylamide gel electrophoresis revealed the presence of a single active peptide with a molecular weight between 7 and 8 kD. Some other peptides of similar molecu- lar weight and with closely related elution time were partially characterized. Their amino-acid composition exhibits broad similarities (Soyez et al., unpublished data). d) Latest data on VIH Recently, Meusy et al. [164] demonstrated that VIH from the lobster, H. americanus, is not strictly species specific from immunochemical criteria: the antibodies raised against the purified VIH from H. americanus crossreact in direct ELISA with sinus gland extracts from some other species (shrimps: varians and Palaemon serratus; prawn: Macrobrachium rosen- bergii; crab: Carcinus maenas) and not with that from several others (prawns: Penaeus vanamei and P. monodon,; crayfishes: Astacus leptodactylus and Orconectes limosus; spiny lobster: Jasus paulen- sis). Immunocytochemical studies of the sinus gland of H. americanus, using the same antibodies and colloidal gold labelling, revealed that VIH is mainly localized in electron dense granules of medium size, 110-185 nm in diameter (Fig. 6). Palaemonetes 232. J.-J. MEUSY AND G. G. PAYEN Fic. 5. Endocytosis in the prawn, Macrobrachium rosenbergii, as studied by incubation of vitellogenic oocytes in a medium containing colloidal gold conjugated vitellin (a, b). The effect on endocytosis of a sinus gland extract is shown (c). Microvilli (mv), endocytotic vesicles (ev) and yolk bodies (YB) are labeled. No significant labeling is observed in the presence of a sinus gland extract (courtesy of P. Jugan and D. Soyez). Female Reproduction in Malacostracan Crustacea 233 Similar studies with an antiserum raised against the Crustacean Hyperglycemic Hormone (CHH) [165] have shown that this hormone, chemically related to VIH, is contained chiefly in large granules (170- 260 nm) (Meusy et a/., unpublished results). So, the axonal endings, and consequently the neurosecretory perykaria, seem _ specialized, though the number of granule types recognized is below that of the neurohormones yet known. It is likely that the criteria, mainly morphological, used for the typology of the secretory granules may not be satisfactory. 2. Stimulatory control Many examples of hormonal antagonisms avail- able in other groups, especially in mammals, suggested the possible occurrence of a vitellogene- sis stimulating system in Crustacea. Moreover, following the opinion of some authors, the variable effect of eyestalk ablation on vitellogenesis, gener- Fic. 6. Immunocytochemical (colloidal gold) staining of VIH in the sinus gland of the lobster, Homarus america- nus, using a mouse serum against H. americanus VIH as primary antibody. The labeling is located on neurosecretory granules of medium size (A) (110-185 nm in diameter). The larger granules (B) (170-260 nm) are not labeled (from [164]). 234 J.-J. MEUSY AND G. G. PAYEN ally stimulatory but dependent on the sexual condition of the female, the species and the environmental circumstances, seems to credit the hypothesis of an antagonistic control. a) Neurohumoral factors Though secretory cells have been described initially in the thoracic ganglia of crabs [166, 167], Otsu [168, 169] gave the first indications of a stimulatory control of vitellogenesis by substances issued from these structures: he observed a preco- cious development of the ovaries in the crab, Potamon dehaani, after implantation of thoracic ganglia. This result was confirmed in some other decapods [170-173]. Boiled aqueous extracts of thoracic ganglia from the fiddler crab, Uca pugilator, stimulated vitel- logenesis in both intact and destalked crabs [36]. Takayanagi et al. [174] demonstrated in vivo and in vitro that aqueous extracts from not only thoracic ganglia but also brain have a positive effect on vitellogenesis in oocytes of the shrimp Paratya compressa. In the amphipod, O. gammarella, where the role of the thoracic ganglia has not been investigated until now, Blanchet-Tournier ef al. [116] demonstrated that the antero-median part of the protocerebrum is stimulatory. To conclude, the existence of an aqueous- soluble substance, secreted by nervous cells and having a stimulatory effect on vitellogenesis, seems established. | However, the nature of this substance — perhaps a peptide -, its precise origin and the mechanism of its action remain to be studied. b) Vitellogenin Stimulating Ovarian Hormone (VSOH) As already mentioned (cf. Section I, B, 1), the ovary of Crustacea develops itself, i.e., its dif- ferentiation is not hormonally controlled [175, 176]. On the contrary, the testis - and the male secondary characters -— are induced by the androgenic hormone secreted by the androgenic glands whose development is genetically induced [175]. If the testis is protected against the action of the androgenic hormone before the onset of sperma- togenesis, it develops into an ovary (cf. Section I, B, 1). But the surgical suppression of the androgenic glands in pubescent males is generally followed by the arrest of spermatogenesis and the degeneration of the testes only: vitellogenin syn- thesis, as well as oogenesis, do not take place. It has been demonstrated in O. gammarella that the implantation of an ovary is necessary for triggering vitellogenin synthesis [177]. On the other hand, the ovariectomy in vitel- logenic females of O. gammarella is followed by the arrest of vitellogenin synthesis [177] and the fat body acquires the same features as the fat body of males and non-vitellogenic females [64]. This effect, considered alone, could be eventually explained by a feed-back regulation mechanism, as suggested by Picaud and Souty [178] for similar results obtained in females of the isopod, P. dilatatus. But the results of the preceding experi- ments performed on males of O. gammarella plead in favor of an ovary hormone. It might be possible that VSOH is the same hormone as the ovarian hormone controlling the ovigerous setae ([77, 78]; cf. Section VII, B). In the isopod, Armadillidium vulgare, vitellogenin synthesis is not ovary de- pendent [179]. Up to now, no other study has been carried out on VSOH which seems to play a similar role to that of estradiol-17£ in egg laying vertebrates. c) Ecdysteroids The Y-organs are responsible for molting [180] by secreting a-ecdysone, which is hydroxylated to the active hormone, 20 OH-ecdysone, also called B-ecdysone, ecdysterone or 208-hydroxyecdysone [181-185, 345]. Except the early works [186-190], several stud- ies have shown that vitellogenesis cannot take place after Y-ectomy in the isopods, Idotea balth- ica, Porcellio dilatatus and Armadillidium vulgare [125, 191, 193], and in the amphipod, Orchestia gammarella [192]. Nevertheless, the relationship between 20 OH-ecdysone secretion and _ vitel- logenesis is not easy to define. It has been demonstrated by radioimmunoassay that a high peak of ecdysteroids occurs in the haemolymph of various species before exuviation, during a short time of stage D; (or D,-D,) of the molting cycle (in the crab Carcinus maenas, in O. gammarella, in the shrimp, Palaemon serratus [194-197] and in the prawn, Macrobrachium rosenbergii, Derelle and Meusy, unpublished data). Vitellogenesis and Female Reproduction in Malacostracan Crustacea 235 vitellogenin synthesis have begun a long time before this short increase of ecdysteroid level in haemolymph and cannot be directly related to this phenomenon (Fig. 7a and 7b). Moreover, molting and reproduction cycles are not synchronous in several Crustacea. The extreme instance is that of oxyrhynch crabs whose Y-organs degenerate in males as well as in females and enter a terminal anecdysis after the puberty molt [198, 199]. Further data on the effect of molting hormone on vitellogenesis have been brought on by studies on vitellogenin. Meusy ef al. [192] have demon- strated that Y-ectomy in O. gammarella is fol- lowed by a decrease of the vitellogenin synthesis. In the isopod, Porcellio dilatatus [200], a decrease of the amount of the circulating vitellogenin was observed after Y-ectomy and this effect has been compensated by 20 OH-ecdysone injection to the animals. But administration of 20 OH-ecdysone to non-operated females of O. gammarella failed to trigger or stimulate the vitellogenin synthesis [201]. Furthermore, molting hormone is not necessary for an in vitro synthesis of vitellogenin by the fat body from female [202] or even male P. dilatatus [203], though an in vivo stimulatory effect has been reported in this species [200]. A stimula- tory effect has been also reported on ovarian protein synthesis [348]. So, it is unlikely that the molting hormone plays a specific stimulatory effect on the vitellogenin synthesis and the vitellogenesis. Numerous studies carried out on insects seem to credit 20 OH- ecdysone with a stimulatory effect on several metabolisms, but not specifically on the vitel- logenin synthesis which is controlled by juvenile hormone (the haematophagic insects, where 20 OH-ecdysone triggers vitellogenin synthesis after a blood meal, seem to be a particular feature). The function and destiny of the ecdysteroids found in the ovaries of O. gammarella at the end of the vitellogenesis [196] and in the ovaries of Carcinus maenas, especially ponasterone A [204, 205], are still undetermined. d) Juvenoids Some authors have speculated that the juvenile hormone, JH, which regulates metamorphosis and gametogenesis in insects might also play a role in the physiology of crustaceans. Four approaches to this topic were carried out by: 1) injecting juvenile hormone or analogs; 2) observing some structural similarities of the mandibular organs of Crustacea with the corpora allata of insects and steroid- producing cells; 3) implanting these mandibular organs in experimental animals; 4) identifying sesquiterpenoid compounds in haemolymph and mandibular organs. Several authors have observed a chemosterilant effect of JH-I (on Orchestia gammarella [68]) or juvenile hormone analogs (on the mud crab, Rhithropanopeus harrisii [206], and on the imma- ture spider crab, Libinia emarginata [207]). In these experiments, the addition of hormone in- creased the current haemolymphatic level to a supraphysiological state which might have toxic effects on the ovary. Similar results have been reported in insects ({208], p. 247), though the corpora allata, source of juvenile hormone in insects, are necessary for vitellogenin synthesis and vitellogenesis in most species [209]. The presence of corpora allata has never been pointed out in Crustacea but endocrine organs located in the vicinity of each mandible have been described by Le Roux [210] who postulated that these organs might have an endocrine function related to oogenesis. The ultrastructural features of the so-called mandibular organs showed analo- gies with steroid-producing cells [211-213] and corpora allata of insects [214]. The mandibular organs are controlled by the eyestalks, probably by a hormone from the sinus glands: they become hypertrophied after eyestalk ablation [211, 215]. Their involvement in vitel- logenesis has been suggested in Libinia emargina- ta: mandibular organs from adult male spider crab were able to induce vitellogenesis when implanted in immature females [216]. After a preliminary work [217] in which the authors detected a juvenile hormone activity in two decapods, Laufer ef al. [218] demonstrated the in vitro secretion of methylfarnesoate by the This compound is structurally and biologically related mandibular organs of Libinia emarginata. to JH-III, as a major product, and a very small amount of JH-III (1000 times less than methyl- farnesoate). After eyestalk ablation, the secretion of methylfarnesoate was enhanced by at least two 236 J.-J. MEUSY AND G. G. PAYEN a 45 8 a ¢ 35 ee = 6€ = © S sE ie s = 4c S o @ 20 a pur) o a c 3 > 15 = ov . = = 5) 8S 2> 10 3 x< 5 2 y 0 0 A B C-3 C6 C-I1 C-19 Do DI’ DIT DIT 02 A molting cycle 150 = ce own lor -5) (oy =] — + +3 100 ae oO nes Te) es i> vo o—) stage, i.e., approximately 4 hr before exuvia- tion. The divalent chromosomes that are not yet organized in a metaphase plate become visible at the oocyte surface, only 1-2 hr before the exuvia- tion. They lay in a nucleoplasmic region devoid of nuclear envelope. The first meiotic spindle can be seen at the time of exuviation. The oocytes remain blocked at this stage of metaphase I until spawning. To determine the exact stimulus that governs meiosis resumption, experimental studies have been conducted in O. gammarella and P. serratus [268, 269]. In O. gammarella, if exuviation is advanced by injection of 20 OH-ecdysone or de- Female Reproduction in Malacostracan Crustacea 241 Fi. 8. prawn Palaemon serratus (redrawn from [270]). Diagram showing the evolution of the nuclear apparatus during maturation and activation processes in the A, B, C, Do, Dy, Di-, Dy”, Dz: stages of the molt cycle; cch: condensed chromosomes (ch); dnu: dissociation of nucleolus; Ex: exuviation or maturation molt; eBl: end of the first blocking stage in prophase 1; Bll: second blocking in metaphase 1 (M1); fa: filamentous apparatus; fm: fertilization membrane; GV: germinal vesicle; GVBD.: germinal vesicle breakdown; ne: nuclear envelope; nu: nucleolus; 1‘ PB, 2" PB: first, second polar bodies; pvs: perivitelline space; T1, T2: telophase 1 and 2; OPn: female pronucleus; y: yolk body. layed by cauterization of the median zone of the protocerebrum, the two phenomena remain simul- taneous only if the oocytes have reached a certain size (about 500 um of diameter). Furthermore, if exuviation is blocked by Y-ectomy, no maturation In P. serratus meiotic reinitiation of prophase I blocked — oocytes is triggered if imma- ture oocytes are incubated in presence of either 20 OH-ecdysone (10~°M), or ponasterone A, or occurs. ionophore A 23187 (5 uM in a normal or Ca’* free seawater milieu). These results suggest that ster- oids are involved in meiotic maturation. To our knowledge the only comparable data in the other arthropod concern the insect Locusta migratoria [271]. Moreover, in the prawn, the treatment with ionophore indicates that steroid inducers may act via intracellular calcium. It is tempting to correlate Clédon’s results with those that mention a high 242 concentration of ecdysteroids in the ovaries of the shore crab Carcinus maenas at the end of vitel- logenesis (10-°M compared to 10~°M in the haemolymph), as well as in the eggs immediately after egg laying [204, 272]. If there exists a relationship between ecdysteroids and the resump- tion of meiosis, do steroids primarily at the level of oocyte membrance induce a cascade of events similar to those known in amphibian oocyte [273], or only after their entrance into the ooplasm where they accumulate? yoy Fic. 9. 2 J.-J. MEUSY AND G. G. PAYEN IV. OVULATION The process by which oocytes are expelled from the ovarian environment (ovarian spawning) has been rarely studied in crustaceans and must be distinguished as a separate process from oviposi- tion that is the release of oocytes or eggs in the external milieu. The only available description of ovulation has been carried out by Fauvel [274] in the prawn Macrobrachium rosenbergii. In this species, it occurs after ecdysis when the follicle epithelium retracts at the periphery of the ovary, i.e., when Poon \) (| Ee a t d Phases of ovulation in the prawn, Macrobrachium rosenbergii (adapted from [274]). a. At the beginning of ovulation the follicle epithelium retracts from the vitellogenic oocytes (arrow) located near the oviduct (Ovd). b. Course of ovulation. Retracted mesodermal tissue forms crests (Cr) between the ovulated oocytes (OO) at the periphery of the ovary (Ov). c. and d. End of ovulation. The follicle tissue occupies the empty space left by the spawned oocytes. MT: mesodermal (follicle) tissue; OE: oviducal epithelium; OW: ovarian wall; PoC: perioviducal cavity; VF: vitellogenic follicle; GZ: germinative zone. Female Reproduction in Malacostracan Crustacea 243 the follicle envelope separates from the oocyte (Fig. 9). The retraction begins in an ovarian zone close to the oviduct. Then, crests of retracted mesodermal tissue are formed between the ovu- lated oocytes. At last, the follicle tissue occupies the empty space left by the spawned oocytes. This tissue, which always develops in continuity with the epithelium of the oviduct and remains in the peripheral region of the gonad, is used again for a new folliculogenesis. Although evidence of a direct hormonal intervention in ovulation has not yet been reported, we must mention that Matsu- moto [275] described an increased neurosecretory cell activity associated with ovulation in the crabs Potamon, Sesarma, Neptunus and Chionocetes. It is not sure that ovulation is used by the author with the above meaning. V. OOCYTE ACTIVATION Oocyte activation allows the completion of meiosis. It is characterized by the release of the second meiotic block which follows exuviation. It leads to both the extrusion of the polar bodies and the elaboration of the fertilization membrane. At last, the female pronucleus is formed, ready to fuse with the male pronucleus to make a zygote. Among the sequence of morphological events that characterize fertilization and lead to the spermatozoon-oocyte association, we shall limit our interest to two aspects: 1) the fertilization potential, 2) the cortical reaction, i.e., the re- sponse of the oocyte plasma membrane to the spermatozoon penetration. Therefore, we do not describe the initial events of the gamete contacts and particularly the acrosome reaction that fits better in a review on male reproduction. Indeed, a number of well-documented papers dealing with this topic concern the decapods which show the particularity to have non-motile spermatozoa (cf. e.g., [276-281]). A. Meiosis reinitiation of metaphase I - arrested oocytes It is now well-known that oocytes of amphipods and several decapods are at the first meiotic metaphase at the time of spawning and meiosis resumes soon thereafter [266-269, 281-289]. Anaphase stage takes place in the spawned eggs. For a long time it has been uncertain whether spawning or fertilization triggers meiosis to com- plete. In order to elucidate this question, different experiments were undertaken on Palaemon serra- tus. It thus appears that the release of the second meiotic block can be obtained in vitro, in presence of an excess of extracellular calcium (10 to 30 mM), or KCI (60 mM), or ionophore A 23187 (5 uM) [269]. As for the resumption of the first meiotic block, the stimulation by A 23187 requires the presence of Ca*t. In addition, experimental fertilization performed in vitro indicates that in P. serratus, fertilization is responsible for the second meiotic resumption [269, 270]. However, other works carried out on the same prawn have shown that meiosis resumes when the egg comes into contact with seawater, independently of fertiliza- tion [288]. Investigations concerning a possible ionic control of activation have led to the conclu- sion that the presence of external Mg’*, but not the external Ca’*, is required for resumption of metaphase I in P. serratus oocytes. It is the change from the low Mg?* environment of the ovary (10 mM) to the high Mg** of seawater (> 15 mM) that stimulates meiosis to resume. Therefore, activa- tion occurs at spawning and does not require fertilization [289]. No indication yet concerns a possible role of extracellular Mg’* for the sper- matozoon-oocyte fusion. An electrophysiological study completes the above results. It states that an increased oocyte membrane permeability to K* occurs at spawning in P. serratus. It is not dependent on fertilization but depends on the increase in external Mg’* concentration at spawning. In other words, at spawning, the hyperpolarization of the oocyte membrane to K* only occurs in presence of a sufficient external Mg** concentration [289]. B. Fertilization potential In contrast to various animal groups in which the electrical characteristics of oocytes at different steps of their development, including fertilization, have been described (cf. reviews [290, 291]), the electrical response to fertilization was evidenced quite recently in malacostracans. The investiga- tions concern the crabs Carcinus maenas and Maia 244 J.-J. MEUSY AND G. G. PAYEN squinado, and the lobster Homarus gammarus [287, 290, 291]. They show that the fertilization potential consists of a sustained hyperpolarization of the egg membrane (from —32 to —62 mV in the crabs). In these decapods, in vitro insemination revealed a sperm-triggered increase in the ionic permeability of the egg membrane which becomes selective for K*, whereas before insemination it was predominantly selective for Cl”. This instan- taneous shift that constitutes the fertilization potential seems to be promoted by a rise in cytoplasmic-free Ca** that might mediate the hyperpolarization. It occurs concurrently with the second meiotic reinitiation in the metaphase I- arrested oocytes. It must be pointed out that under natural conditions, the early events in crab fertilization take place internally in the female genital duct and sometimes in the lumen of the ovary [128, 280, 282, 292, 293], whereas the lobster oocytes are fertilized in the external environment [294]. It thus appears that the electrical response of the oocyte to fertilization may reflect a general property of reptantian Decapoda. As already pointed out (cf. Section V, A) in the prawn Palaemon serratus in which fertilization is external (the extruded oocytes pass over the sperma- tophore previously deposited by the male at mating), a similar increase in K* conductance of the oocyte membrane takes place at spawning. This increase is not dependent on fertilization, but depends on an increase in external Mg** concen- tration at spawning [289]. Until now, the egg’s electrical response to fertilization remains to be explored in Palaemon. C. Cortica! reaction The cortical reaction can be defined as one of the anatomical responses (besides the formation of the fertilization cone and the elaboration of the first polar body) of the oocyte developing an hyperpolarization response after insemination. However, this phenomenon may be also initiated after exposure to sea water. The morphological events that occur in the cortex of eggs, i.e., the exocytosis of cortical granules into the perivitelline space and the transformation of the plasma membrane were investigated by means of scanning and transmis- sion electron microscopy in the penaeid shrimps Penaeus aztecus and P. setiferus, and the shore crab Carcinus maenas [295-297]. A brief descrip- tion of the cortical granules has been also reported in the amphipod Orchestia gammarella [4, 91]. We shall examine the cortical reaction process respec- tively in these three models, although it must be known that the elaboration of the fertilization envelope was described in detail in cirriped eggs [298]. During the cortical reaction, early fertilized C. maenas eggs maintained under in vitro conditions appear to release successively: 1) a fine granular material that accumulates in about 15 min on the inner face of the vitelline envelope [296] and, 2) a massive amount of ring-shaped elements which coalesce to give rise to a new thick coating underlying the vitelline envelope and represents most of the fertilization envelope. This phe- nomenon lasts about 7-8hr. The ring-shaped elements come from egg cortical vesicles. It was established by Goudeau [297] that these elements and their enclosing vesicles originate in the endo- plasmic reticulum from which they are released by direct endocytosis. The author considers that the ring-shaped elements are precursors common to the cortical exudate and to the endogenous yolk (cf. Section II, A). The cortical reaction in the eggs of penaeids is unique with respect to: 1) the size of the cortical specializations (rods that are around 40 «m length for an egg diameter of about 270 um), 2) the rapid expulsion and dissipation of these elements in response to sea water, 3) the decrease in the egg volume after the reaction. The rods are always located perpendicular to the oolemma and com- posed of numerous tightly packed fibrillar struc- tures. Each cortical rod lies within a partially membrane bound crypt and is separated from the external environment by a thin coat that complete- ly surrounds the egg. As the rods are expelled in the sea water, a corona forms around the oocyte and then quickly dissipates. Simultaneously an extensive membrane vesiculation associated with cortical rod crypts become apparent around the entire egg surface and later forms a homogenous jelly. Mg** ions and a protease dependence of this jelly release have been demonstrated in Penaeus Female Reproduction in Malacostracan Crustacea 245 and in another penaeid, Sicyonia [279, 281, 299]. In O. gammarella, cortical granules have been observed towards the end of vitellogenesis when the oocytes are no longer attached to the follicle tissue due to the retraction of the macro- and microvilli. At this period, the vitelline envelope becomes thick. The cortical granules are oval- shaped and measure 0.2-0.3 um of mean dia- meter. They become visible when the microcanals and pinocytotic vesicles disappear. They display lamellae alternatively lucent and electron dense and are bound by an outer membrane. Glycopro- teins have been detected. After fertilization, the cortical reaction consists of two steps. The first shows a fusion of the cortical granule membrane with the egg plasmic membrane, leading to the release of granule contents into the perivitelline space. During the second step, the vitelline envelope is elevated off the surface of the oocyte and acquires on its inner face an opacity that rapidly extends to the outer face. This opacity is concomitant with important modifications of the vitelline envelope that becomes the fertilization envelope. The origin of the cortical granules and the duration of the cortical reaction remain to be studied in this peracarid. VI. OVIPOSITION Generally, oviposition takes place when the environmental conditions are favorable for embryonic development. Thus, according to the geographical distribution of the species, oviposi- tion is spread over a season or restricted to some months (cf. for review [235, 300]). In peracarids and some natantians, this phenomenon is usually preceded by a molt, whereas it is confined to intermolt in many brachyurans. A few results concern the existence of a control of oviposition by an eyestalk factor. They have been obtained from diverse eyestalk-ablated de- capods: —lIn juvenile Carcinus maenas the operation leads to a precocious vitellogenesis and spawning may follow but, since the puberal form of the external sex characteristics is not completed, the eggs do not remain attached to the pleopods [301]. According to the author, a factor linked to the presence of eyestalks would be involved in the development of the female external sex character- istics. —Postmolt crabs, Menippe mercenaria, spawn precociously without undergoing accelerated o- varian development [130]. This may be an argu- ment in favor of the fact that spawning and ovarian development would be controlled by different eyestalk hormones. —However, in juvenile prawns Penaeus japoni- cus, oviposition never follows the accelerated vitellogenesis because the oocytes degenerate (Laubier and Bizot-Espiard, personal communica- tion). Similarly, in the crabs Rhithropanopeus harrisii and Paratelphusa hydrodromous eyestalks would be necessary to the process of oviposition at time of the reproductive period [302, 303]. Moreover, the would be released several days before spawning. Oviposition-inducing hormone Environmental factors such as water tempera- ture and photoperiod, probably channeled through the neuroendocrine system, seem also to affect oviposition, as has been shown in the crayfish Orconectes virilis [248, 304] and the crab Pachy- grapsus marmoratus [285]. Synchronization of oviposition with specific tidal phases has been also reported in the stomatopods Gonodactylus zacae and G. falcatus. [305]. Spawning postures have been described for the spider crab Chionocetes opilio and the spiny lobster Panulirus homarus [306, 307]. In contrast to insects in which oviposition is a fully-studied event beginning shortly after mating (cf. review [308]), more thorough investigation is needed in crustaceans because this process occurs either after or before mating, depending on the considered species (cf. Section V, B). VII. SEX CHARACTERISTICS ASSOCIATED TO SPERM STORAGE, MATING AND EGG INCUBATION Among the female characteristics related to sperm storage, mating and egg incubation, one can distinguish specialized regions of the genital duct of species in which fertilization occurs internally and structural modifications of some body seg- ments, as well as of different appendages. 246 J.-J. MEUSY AND G. G. PAYEN A. Genital duct The most original feature of the genital duct of some malacostracans is the spermatheca (seminal receptacle), a specialized area that receives the spermatophores in which spermatozoa are stored. The presence of spermatheca is essentially known in brachyurans. Study of the genital apparatus morphogenesis carried out in the crab Rhithropa- nopeus harrisii [11] reveals that, beginning at the fourth postlarval stage, one can determine in the female genital duct three distinct regions from the gonad to the sexual orifice, or vulva: an oviduct, a spermatheca and, distally, a vagina. This is in agreement with Hartnoll’s description [309]. The oviducts shorten when the spermatheca develop. In puberal crabs, the cuticle lines the walls of the spermatheca. A scanning electron microscopy examination of the luminal wall of the genital duct of Carcinus maenas reveals that, at the level of the spermatheca, the epithelium and the cuticular covering form numerous parallel folds and dif- ferentiate two lateral pouches filled with stored spermatozoa [293]. Such an anatomical pattern explains sperm retention after mating for several successive reproductive periods separated by molts [310]. In addition, Anilkumar and Adiyodi [311] reported a cyclic synthetic activity of the sper- matheca epithelial cells of the crab Paratelphusa hydrodromous in relation to the reproductive cycle. The oviducal epithelium of the shore crab differentiates during the reproductive premolt period two distinct secretory zones that are sup- posed to be involved in the release of a sexual pheromone attracting the male for copulation. Another interpretation of Anghelou-Spiliotis and Goudeau [293] is that the secretions could have also a lytic function on the spermatophore walls. At last, when the oocyte is ready to be spawned, these substances could be involved in the modifica- tion of the chemical composition of the vitelline envelope. B. External sex characteristics The external sex characteristics only present in females and involved in specific functions are either permanent or temporary. The permanent characteristics generally develop in the juvenile females. They are: 1) the shape of the last thoracic sternite which bears an external seminal receptacu- lum in caridean shrimps or a thelycum in penaeids, 2) the oostegites in some peracarids. The tempora- ry characteristics include: 1) the sexual setae used for pairing, 2) the oostegites in some isopods, 3) the ovigerous setae (oosetae) of amphipods and decapods, 4) the brood chamber of Caridae. These last three characteristics are associated with the incubation of embryos. 1. Differentiation As an example of permanent female characteris- tics we have chosen to describe the development of the oostegites that has been well-studied in the amphipod O. gammarella [123]. Oostegites appear as small outgrowths on the internal face of the coxae of the second gnathopod and pereopods 3, 4, and 5. At that time, the ovaries contain follicles with previtellogenic oocytes. The oostegites de- velop further at each molt and form the brood pouch or marsupium in the puberal female. Dur- ing development of the oostegites in the young females, trichogenic matrices are set up. They form short setae (0.02 mm in length). Almost all amphipods possess four pairs of oostegites; howev- er, this number can vary, as in Caprellidea, where only two pairs of oostegites are born by segments 3 and 4 [312]. In O. gammarella, during the reproductive season, a vitellogenesis occurs during each molt cycle, and spawning after each ecdysis. During stage D of the molt cycle, the trichogenic matrices form long setae (0.8 mm in length) which appear at ecdysis. These setae border the ooste- gites and ensure a good closing of the brood chamber. They are temporary sex characteristics associated with the incubation of embryos. They are replaced by short setae during the intermolt cycles without vitellogenesis [123]. Females of isopods also possess oostegites. These sexual appendages are permanent charac- teristics in Ligia oceanica, Helleria brevicornis and the aquatic species Asellus aquaticus, Sphaeroma serratum and Idotea balthica [125, 152, 313-315]. They acquire their functional form only at molts followed by egg laying and again take their non-functional form at the period of genital rest. Female Reproduction in Malacostracan Crustacea 247 Their functional form can be considered as a temporary characteristic related to egg incubation. However this is not the case in the aquatic isopod, Idotea balthica, in which the functional form persists throughout the life span of the female. In Oniscidea, except Ligiidae and Tylidae, the pres- ence of oostegites constitutes a temporary charac- teristic which only appears at time of the molts followed by egg laying and disappears at time of genital rest. In J. balthica, another temporary characteristic concerns the sexual setae born on the internal surface of the second pereopods which disappear at the molt preceding the first egg laying [125, 316]. The morphological modifications leading to the formation of the brood chamber in Caridea have been particularly studied in some Palaemonidae [317], Atyaephyra desmaresti [318] and Macro- brachium rosenbergii [23]. However, no correla- tion with the developing ovaries has been noted. The brood chamber is not permanent in some Palaemonidae. It disappears during genital rest. Indeed, coxopodites and sternites take again the juvenile form. No study concerns the trichogenic matrices. The brood chamber is formed by broadening of the sternites and lengthening of the coxopodites of the first three somites of the pleon. Basipodites enlarge and display a groove-like shape, with the concavity turned to the rear. An effective closing of the brood chamber is ensured by long plumose setae arranged in two rows on each basipodite. The eggs are attached to the ovigerous setae (oosetae) which are located on the internal edge of the basipodites. The newly laid eggs are guided to the seminal receptaculum by means of long setae both on the coxa of the pereopods 3, 4 and 5, and around the gonopores. In female crabs, four pairs of unsegmented and hairless biramous pleopods develop from the third crab stage, on the second to the fifth abdominal segments. Segmentation of the endopodites, and exopodites generally precedes by one stage the appearance of tufts of hairs on the endopodites and of a setiferous fringe on the endopodites [11, 319]. In the crab Pachygrapsus marmoratus, as in caridean shrimps, oosetae develop on the pleopods and along the edge of the abdominal sternites at the molts that are followed by egg laying; the oosetae disappear when the ovary is at rest [320]. These temporary characteristics appear for the first time during the molt preceding the first egg laying. In some species, such as Carcinus maenas, the abdomen acquires the female form (enlarging, curving inwards, and hairs on pleopods) at the puberty molt that occurs one or several molts before the first egg laying (301, 321]. The external characteristics acquire their definitive shape only at the final puberty molt in brachyurans that undergo a limited number of molts as the Majidae [322, 323], the Leucosiidae, and the Portunidae of the genus Callinectes [324, 325]. In the Majidae Acanthonyx lunulatus and Libinia, the first egg laying occurs immediately or sometime after the puberty molt [326, 327]. Indeed, in these two species, the first vitellogenesis begins respectively during the course of the last intermolt cycle and after puberty. The mechanism that allows a newly laid egg to be attached to ovigerous setae has been studied in the crab Carcinus maenas [128, 328]. It involves the formation of a funiculus that originates from the two superimposed vitelline envelopes [92, 286]. Examination of the structure of the funiculus and of the morphological features of its binding to maternal egg-carrying setae revealed that the tip of the funiculus is coiled around the setae without adjunction of any additional attachment sub- stance. Four concentric envelopes which are successively secreted by the ectodermal embryonic cells underneath the fertilization envelope have been detected during the embryo development. It is noteworthy that ponasterone A, an ecdysteroid present in high concentration, would be involved in the deposition of the embryonic envelopes [328]. In some species, special secretions of tegumental glands of the ventral abdomen seem to be used for attachment of eggs. For example, Mason [329] indicates that the oviposition posture leads to the formation of a water-filled cavity into which the eggs of the crayfish, Pacifastacus leniusculus trow- bridgii, pass brushing across the glandular areas. Moreover, it has been noted in another crayfish, Austropotamobius pallipes, that the activity of these glands is possibly linked to egg growth [330]. 248 J.-J. MEUSY AND G. G. PAYEN At last, these glands would have also a role in dissolution of the spermatophore wall and trans- port of spermatozoa [329]. An ultrastructural analysis of the pleopod tegumental gland in the lobster, Homarus america- nus, was recently carried out [331]. It completes a previous study in the same species by Aiken and Waddy [332]. The pleopods of both male and female lobsters contain rosette type glands. However, they are most abundant in females with well-developed ovaries. Two types of secretion seem to be produced continually. They would be involved in the hardening of the cuticle after molting and also in condensation and hardening of the outer egg coat during egg attachment to the ovigerous setae. 2. Regulation of development An ovarian control of external female character- istics has been demonstrated in several peracarids. Such a control mechanism has been reported in various reviews [123, 333-336]. In decapods, attempts at surgical removal of the ovaries have not so far been successful and there is only an indirect proof of the existence of ovarian hor- mone(s) [337]. In O. gammarella, the ovaries control the permanent and temporary characteristics. The control of oostegites (permanent characteristic) has been studied by implantation of a young or fully-developed ovary into a male from which AG have been removed. In both circumstances, ooste- gites appear at the first or second postoperative molt. The follicle cells of previtellogenic oocytes seem to be the source of the ovarian hormone responsible for the formation of oostegites. Since this hormone is secreted throughout the life span of the female, Charniaux-Cotton and Payen [336] proposed to call it “Permanent Ovarian Hormone” (POH). The induction of oostegites by POH is irreversible: it persists in castrated females. In some isopods, as /dotea balthica and Ligia oceanica, the oostegites differentiate in young females without the mediation of a hormone. In castrated females, the marsupium develops nor- mally [125, 152, 316]. In addition, a marsupium develops in andrectomized males of /. balthica although the testes are not reversed into ovaries. The experimental results obtained in O. gam- marella and few isopods (Armadillidium vulgare, Porcellionides pruinosus and Porcellio laevis) [77, 338-340] show that the ovary secretes during vitellogenesis a hormone controlling the formation of the temporary external characteristics. Char- niaux-Cotton and Payen [336] have called it “Temporary Ovarian Hormone” (TOH). In O. gammarella, ovariectomy during the re- productive period is followed by replacement of ovigerous setae by juvenile setae [77]. Likewise, when a vitellogenic ovary is implanted into an andrectomized male, the induced oostegites ac- quire ovigerous setae. The follicle cells of the vitellogenic oocytes seem to be the source of TOH. The formation of ovigerous setae requires also the presence of molting hormone. During genital rest, when 20 OH-ecdysone acts alone, juvenile setae are formed. If the quantity of ovarian hormone does not attain a certain threshold, as after a partial ovariectomy, the elongation of the trichogenic matrices is only partial and, as a result, the length of the setae is intermediate. When molting and ovarian hormones are present, the matrices stretch out extensively and form oviger- ous setae. In Armadillidium vulgare and Porcellio dilata- tus, the oostegites (temporary characteristic) never appear in ovariectomized females [338, 341, 342]. Reimplantation of a small ovarian portion induces the formation of an incomplete marsupium. In decapods, in the absence of successful ovar- iectomies and implantations, the control of perma- nent female characteristics is not known. Tempo- rary external characteristics appear to be control- led by vitellogenic ovaries, as in peracarids. Implantation of portions of early vitellogenic ovary into AG ablated male freshwater prawns Macrob- rachium rosenbergii [337] results in the induction of female breeding characteristics as ovigerous and Ovipositing setae and brood chamber. Indeed, these characteristics develop during vitellogenesis and disappear when the ovaries are resting. These relationships between the morphogenesis of the temporary characteristics and the course of vitel- logenesis remain to be precised. However, it is worthy to note that the existence of an ovarian hormone controlling the external sexual character- Female Reproduction in Malacostracan Crustacea 249 istics was suggested in the forties by some authors who studied the effects of parasitic or X-ray castration on the shrimp Leander (Palaemon) serratus [343]. Furthermore, at the same time, evidence of a correlation between ovarian and tegumental gland development was noted in the shrimp Crangon crangon and in the crayfishes Cambarus virilis and C. rusticus [304, 344]. VIII. SEX RECOGNITION AND MATING BEHAVIOR As in nearly all phyla, recognition and attraction of a sexual partner to promote successful mating depends in malacostracans on a_ broadcast of identifying behaviors. The display patterns are mainly acoustic, visual, olfactory, tactile and chemical signals. Most of them have been re- viewed separately or in their whole in some significant papers (cf. [235, 350—353]). It must be pointed out that the diverse com- munication systems seem adapted to the various inhabited spatial localization. Thus, after the first experimental demonstration in the female crab Portunus sanguinolentus [379], crustacean sex pheromones inducing precopulatory behavior in the male have been detected in several aquatic species. (cf. for review [353]). Their stimulating effects predominate in small forms such as amphi- pods (cf. [354-356]), as well as in natantian and reptantian decapods which include large forms (shrimps, lobsters, crabs, etc.) (e.g., [357-360]). A synergistic effect between the pheromone re- leased by the female, olfactory and visual stimuli of either or both sexes is also sometimes required for registering a positive response [361]. At last, among terrestrial and semi-terrestrial species, chemical cues emitted by females appear to be less important than visual, tactile and(or) acoustic signals from males [362, 363]. We have limited the scope of this section to recall the conditions that enable females to be- come attractive and receptive to males. The releasing and the possible producing site(s) of sex pheromones, as well as their nature and their target organs in the male are also briefly examined. In addition to local climatic conditions that restrict the copulatory period, the female’s attrac- tiveness determining behavioral responses of the males is generally linked to its physiological state, i.e., its molting and ovarian development stages. In some Brachyura (Cancridae and Portunidae), some Astacidea and peracarids, a female is able to mate only when it is soft, shortly after ecdysis [355, 364, 365], while in other Brachyura (Majidae, Xanthidae and Gecarcinidae) mating involves in- termolt (hard-shelled) females [364, 366-368]. Detailed informations concerning the ovarian de- velopmental stage of females at the time of mating are lacking. In Gammarus duebenii, Hartnoll and Smith [356] mention that “ovarian ripeness” is one prerequisite of the female’s attractiveness and that “ovarian condition and molt stage have a synergis- tic effect”. Similar data were reported by Ducruet [355] in two other gammarids. However, unpub- lished works by Vilotte and Fontaine (cited in [353]) indicate that in Scyllarus arctus copulation occurs during previtellogenesis, while in Carcinus maenas, a female can remain attractive after exuviation (until A>), when ovaries are engaged in vitellogenesis. Takayanagi er al. [371] identified an ovary-stimulating pheromone that would be re- leased by male organs such as the testis or the vas deferens in the freshwater shrimp, Paratya com- pressa. In some crabs, we have already mentioned that sperm is stored for prolonged periods and can fertilize subsequent spawnings (cf. Section VII, A and for review [235] p. 224). When a female attracts a male prior to its molt, as in C. maenas, presumably because this attrac- tion is initiated pheromonally, the male usually guards her in a precopulatory embrace until she molts and copulation is possible [369]. In Homarus americanus, as in the crayfish Austropotamobius pallipes, the passive, or cooperative, reaction of the female when encountered by a male is con- sidered as a premating behavior [357, 365]. Like- wise, in Cardisoma armatum, a semi-terrestrial crab, it is thought that courtship reduces aggressive tendencies in the female [368], but the cues used in sex recognition have not been investigated. According to Hartnoll and Smith [372], there is no evidence for the production of a male stimulat- ing pheromone in the urine of courted premolt female crab, Cancer pagurus. However, behavior- al studies indicate that sex pheromone emission is 250 J.-J. MEUSY AND G. G. PAYEN often associated with urine release from the female antennal gland (cf. reviews [350, 353, 359, 360]). The site of pheromone production is still not well known: Kamiguchi [373] described a sternal gland in female, Palaemon paucidens, and Bauchau [353] discovered in C. maenas an ectodermic gland more developed in females than in males. It opens in the ureter in the vicinity of the nephropore and seems well-suited to a pheromone release into urine. Information on the chemical nature of crustacean sex pheromones is scarce. Bauchau [353] reported that their molecular weight is ranging from 1000 to 10,000 daltons, according to the species. There are inconclusive data concerning ecdysone (or its derivative), serotonin or peptide as sex attractants ({374, 380], for review [353]). Chemoreceptor sensilla (aesthetascs) on the outer flagellum of antennules are involved in the detection of sex pheromones in several decapods ([353, 358, 375], for review [350]. It has been recently suggested that a hormonal modulation of pheromone- triggered courtship display behavior would exist in the blue crab, Callinectes sapidus [376]. ACKNOWLEDGMENT We are grateful to Madeleine Martin and Daniel Soyez for their friendly assistance. REFERENCES 1 Dhainaut,A. and DeLeersnyder,M. (1976) Etude cytochimique et ultrastructurale de l’évolu- tion ovocytaire du crabe Eriocheir sinensis. 1. Ovogenése naturelle. Arch. Biol., 87: 261-282. Charniaux-Cotton, H. (1978) L’ovogenése, la vitellogénine et leur contrdle chez le Crustacé Amphipode Orchestia gammarellus (Pallas). Com- paraison avec d’autres Malacostracés. Arch. Zool. exp. Gén., 119: 365-397. 3 Charniaux-Cotton, H. (1980) Experimental studies of reproduction in Malacostraca Crustaceans. De- scription of vitellogenesis and of its endocrine control. In “Advances in Invertebrate Reproduc- tion, 2”. Ed. by W. H. Clark, Jr. and T. S. Adams, Elsevier, North Holland, pp. 177-186. 4 Zerbib, C. (1980) Ultrastructural observation on oogenesis in the Crustacea Amphipoda Orchestia gammarellus (Pallas). Tissue and Cell, 12: 47-62. 5 Komm, B. S. and Hinsch, G. W. (1987) Oogenesis in the terrestrial hermit crab Coenobita clypeatus ine) 10 13 15 16 17 (Decapoda, Anomura). II. Vitellogenesis. J. Mor- phol., 192: 269-277. Adiyodi, R.G. and Subramoniam,T. (1983) Arthropoda-Crustacea. In “Reproductive Biology of Invertebrates. Vol. 1: Oogenesis, Oviposition, and Oosorption”. Ed. by K. G. and R. G. Adiyodi, J. Wiley & Sons, Ltd., Chichester, New York, Brisbane, Toronto, Singapore, pp. 443-495. Charniaux-Cotton,H. and Payen,G. (1988) Crustacean reproduction. In “Invertebrate Endo- crinology, Vol. 2”. Ed. by R. G. H. Downer and H. Laufer, Alan R. Liss, Inc., New York, pp. 279- 304. Meusy, J.-J. (1986) Ultrastructure de la zone germinative et des gonies du testicule et de l’ovaire d’Orchestia gammarella P. (Crustacé Amphipode). Ann. Sci. Nat., Zool. Biol. Anim., 10: 101-116. Meusy, J.- J. (1963) La gamétogenése d’Orchestia gammarella Pallas, Crustacé Amphipode. Bull. Soc. Zool. Fr., 88: 197-220. Payen, G. G. (1973) Etude descriptive des princi- pales étapes de la morphogenese sexuelle chez un Crustacé Décapode a développement condensé, l’Ecrevisse Pontastacus leptodactylus leptodactylus (Eschscholtz, 1823). Ann. Embryol. Morphog., 6: 179-206. Payen, G. G. (1974) Morphogenése sexuelle de quelques Brachyoures (Cyclométopes) au cours du développement embryonnaire, larvaire et postlar- vaire. Bull. Mus. Nat. Zool., 209: 201-262. Laubier, A., Chim, L. and Payen,G.G. (1985) Morphogenése sexuelle et régulation hormonale de l’activité génitale chez la crevette Penaeus japoni- cus en élevage. IFREMER, Actes de Colloques, n'1: 195-206. Soyez,D. (1974) Etude comparée de lactivité prévitellogénétique pendant les saisons de repos génital et d’activité sexuelle chez le Crustacé Amphipode Orchestia gammarellus (Pallas). C. R. Acad. Sci., Sér. D, 278: 1867-1870. Rateau, J. G. and Zerbib, C. (1978) Etude ultra- structurale des follicules ovocytaires chez le Crusta- cé Amphipode Orchestia gammarellus (Pallas). C. R. Acad. Sci. Paris, Sér. D, 286: 65-68. Weitzman, M. C. (1966) Oogenesis in the tropical land crab Gecarcinus lateralis (Freminville). Z. Zellforsch., 75: 109-119. Charniaux-Cotton, H. and Touir, A. (1973) Con- trdle de la prévitellogenése et de la vitellogenése chez la crevette hermaphrodite Lysmata seticauda- ta Risso. C. R. Acad. Sci. Paris, Sér. D, 276: 2717- 2720. Charniaux-Cotton, H. (1959) Etude comparée du développement post-embryonnaire de l’appareil génital et de la glande androgéne chez Orchestia gammarella et Orchestia mediterranea (Crustacés 18 19 20 21 22 23 24 ZS 26 27 28 Female Reproduction in Malacostracan Crustacea 251 Amphipodes). Autodifférenciation ovarienne. Bull. Soc. Zool. Fr., 84: 105-115. Charniaux-Cotton, H. and Ginsburger-Vogel, T. (1962) Preuve expérimentale de l’autodifférencia- tion ovarienne chez Orchestia montagui Audouin (Crustacé Amphipode). C. R. Acad. Sci. Paris, 254: 2836-2838. Charniaux-Cotton, H. (1963) Démonstration ex- périmentale de la sécrétion d’hormone femelle par le testicule inversé en ovaire de Talitrus saltator (Crustacé Amphipode). Considération sur la géné- tique et l’endocrinologie sexuelle des Crustacés supérieurs. C. R. Acad. Sci. Paris, 256: 4088-4091. Hort-Legrand,C., Berreur-Bonnenfant, J. and Ginsburger-Vogel, T. (1973) Inversion ex- périmentale du testicule en ovaire par greffe dans la cavité péricardiale de femelles adultes, chez Orchestia gammarella Pallas (Crustacé Amphi- pode). C. R. Acad. Sci., Sér. D, 276: 1891-1894. Juchault, P. and Legrand, J.-J. (1964) Mise en évidence d’un inducteur sexuel male distinct de VPhormone adulte et contribution a l’étude de lautodifférenciation ovarienne chez l’Oniscoide Helleria brevicornis. C. R. Acad. Sci. Paris, 258: 2416-2419. Charniaux-Cotton, H. (1975) Hermaphroditism and gynandromorphism in malacostracan Crus- tacea. In “Intersexuality in the Animal Kingdom”. Ed. by R. Reinboth, Springer-Verlag, Berlin and New York, pp. 91-105. Nagamine, C. and Knight, A. W. (1980) Develop- ment, maturation, and function of some sexually dimorphic structures of the Malaysian prawn Mac- robrachium rosenbergii (de Man) (Decapoda, Palaemonidae). Crustaceana, 392: 141-152. Graf, F. (1958) Développement post-embry- onnaire des gonades et des glandes androgénes d’Orchestia cavimana (Heller) Crustacé Amphi- pode. Bull. Soc. Sci. Nancy, 17: 223-261. Ohno, S. (1979) Major Sex-Determining Genes, Springer-Verlag, Berlin and New York. Payen, G., Costlow, J. D. and Charniaux-Cotton, H. (1967) Recherches sur le rdle de la neurosécré- tion dans la différenciation sexuelle du Crabe Callinectes sapidus Rathbun. C. R. Acad. Sci., Sér. D, 264: 2148-2151. Payen, G., Costlow, J. D. and Charniaux-Cotton, H. (1969) Mise en évidence expérimentale de Pindépendance de la réalisation du sexe chez le Crabe Rhithropanopeus harrisii (Gould) a l’égard du complexe neurosécréteur organe de Hanstr6m- glande du sinus. C. R. Acad. Sci. Paris, Sér. D, 269: 1878-1881. Touir, A. (1977) Donées nouvelles concernant lendocrinologie sexuelle des Crustacés Décapodes Natantia hermaphrodites et gonochoriques. I. 29 30 31 32 33 34 35 36 37 38 Maintien des glandes androgénes et role de ces glandes dans le contréle des gamétogenéses et des caractéres sexuels externes males. Bull. Soc. Zool. Fr., 102: 375-400. Touir, A. (1977) Données nouvelles concernant lendocrinologie sexuelle des Crustacés Décapodes Natantia hermaphrodites et gonochoriques. I. Maintien des gonies et évolution des gaméto- genéses in vivo et in vitro. C. R. Acad. Sci. Paris, Sér. D, 284: 2515-2518. Touir, A. (1977) Données nouvelles concernant Yendocrinologie sexuelle des Crustacés Décapodes Natantia hermaphrodites et gonochoriques. II. Mise en €vidence d’un controle neurohormonal du maintien de l'appareil génital male et des glandes androgénes exercé par le protocérébron médian. C. R. Acad. Sci. Paris, Sér. D, 285: 539-542. Rubiliani, C. and Payen, G. G. (1979) Modalités de la destruction des régions neurosécrétrices des Crabes Carcinus maenas (L.) et C. mediterraneus Czerniavsky infestés par la sacculine. Gen. Comp. Endocrinol., 38: 215-228. Rubiliani-Durozoi, M., Rubiliani,C. and Payen, G.G. (1980) Déroulement des gamétogenéses chez les Crabes Carcinus maenas (L.) et C. mediterraneus Czerniavsky parasités par la Saccu- line. Int. J. Invertebr. Reprod., 2: 107-120. De Leersnyder, M. and Dhainaut, A. (1977) Ac- tion de l’épédonculation sur les premiéres étapes de Vovogenése d’Eriocheir sinensis H. Milne- Edwards (Crustacé Décapode Brachyoure). Etude cytologique et autoradiographique. Arch. Zool. Exp. Gen., 118: 335-348. De Leersnyder, M., Dhainaut A. and Porcheron, P. (1981) Influence de l’ablation des organes Y sur Povogenése du crabe Eriocheir sinensis dans les condition naturelles et aprés épédonculation. Gen. Comp. Endocrinol., 43: 157-169. Kurup, K.N. P. and Adiyodi, R. G. (1981) The programming of somatic growth and reproduction in the crab Paratelphusa hydrodromous (Herbst). Int. J. Invertebr. Reprod., 3: 27-39. Eastman-Reks, S. B. and Fingerman, M. (1984) Effects of neuroendocrine tissue and cyclic AMP on ovarian growth in vivo and in vitro in the fiddler crab, Uca pugilator. Comp. Biochem. Physiol., 79A: 679-684. Frentz, R. (1960) Contribution a l’étude biochimi- que du milieu intérieur de Carcinus maenas Linné. Bull. Soc. Sci. Nancy, Nile série, 19: 1-176. Descouturelle, G. and Frentz, R. (1967) Etude par électrophorése et immunoélectrophorése des pro- téines d’extraits totaux et de Il’hémolymphe d’Atyaephyra desmaresti Millet. Influence de l’abla- tion des pédoncules oculaires. Bull. Biol. Sci. Nancy, 6: 259-270. 252 39 40 4] 43 44 45 46 47 48 49 50 Adiyodi, R. G. (1968) Protein metabolism in rela- tion to reproduction and moulting in the crab, Paratelphusa hydrodromous (Herbst): Part I. Elec- trophoretic studies on the mode of utilization of soluble proteins during vitellogenesis. Indian J. Exp. Biol., 6: 144-147. Besse, G. and Mocquard, J.-P. (1968) Etude par électrophorése des quantités relatives des protéines de ’hémolymphe d’individus normaux et de femel- les castrées chez deux Crustacés Isopodes: Porcel- lio dilatatus Brandt et Ligia oceanica. C. R. Acad. Sci. Paris, Sér. D, 267: 2017-2019. Meusy, J.-J., Charniaux-Cotton, H. and Croisille, Y. (1969) Etude par électrophoreése chez Orchestia gammarella (Pallas) et Orchestia mediterranea Cos- ta (Crustacés Amphipodes) des protéines de ’hé- molymphe: comparaison entre les males, les femel- les et les intersexués. C. R. Acad. Sci. Paris, Sér. D, 269: 741-743. Kerr, M. S. (1969) The hemolymph proteins of the blue crab, Callinectes sapidus. 11. A lipoprotein serologically identical to oocyte lipovitellin. Dev. Biol., 20: 1-17. Gibert, J. (1972) Synthése bibliographique des recherches électrophorétiques sur les protéines des Crustacés. Ann. Biol., 11: 305-327. Ceccaldi, H. J. and Martin, J.-L. (1969) Evolution des protéines de l’hémolymphe chez Carcinus maenas L. durant lovogenese. C. R. Soc. Biol., 163: 2638-2641. Abeloos, M. and Fisher, E. (1926) Sur lorigine et les migrations des pigments carotinoides chez les Crustacés. C. R. Soc. Biol., 95: 383-384. Kerr, M. S. (1968) Protein synthesis by the hemo- cytes of Callinectes sapidus: a study of in vitro incorporation of *C-leucine. J. Cell Biol., 39: 72a— 73a. Wolin, E.M., Laufer,H.and Albertini, D. F. (1973) Uptake of the yolk protein, lipovitellin, by developing crustacean oocytes. Dev. Biol., 35: 160-170. Paulus, J. E. and Laufer, H. (1987) Vitellogeno- cytes in the hepatopancreas of Carcinus maenas and Libinia emarginata (Decapoda brachyura). Int. J. Invertebr. Reprod. Dev., 11: 29-44. Lui,C.W., Sage,B.A. and O’Connor, J. D. (1974) Biosynthesis of lipovitellin by the crusta- cean ovary. J. Exp. Zool., 188: 289-296. Lui, C. W. and O’Connor, J. D. (1976) Biosyn- thesis of lipovitellin by the crustacean ovary. II. Characterization of and in vitro incorporation of amino acids into the purified subunits. J. Exp. Zool., 195: 41-51. Lui, C. W. and O’Connor, J.D. (1977) Biosyn- thesis of lipovitellin by the crustacean ovary. III. The incorporation of labeled amino acids into the nN im) 53 35 56 a7 58 60 61 62 J.-J. MEUSY AND G. G. PAYEN purified lipovitellin of the crab Pachygrapsus cras- sipes. J. Exp. Zool., 199: 105-108. Eastman-Reks, S. B. and Fingerman, M. (1985) In vitro synthesis of vitellin by the ovary of the fiddler crab, Uca pugilator. J. Exp. Zool., 233: 111-116. Yano, I. and Chinzei, Y. (1987) Ovary is the site of vitellogenin synthesis in kuruma prawn, Penaeus japonicus. Comp. Biochem. Physiol., 86B: 213- 218. Hinsch,G. W. and Cone,M. V. (1969) Ultra- structural observations of vitellogenesis in the spider crab, Libinia emarginata L. J. Cell Biol., 40: 336-342. Beams, H. and Kessel, R. G. (1980) Ultrastructure and vitellogenesis in the oocyte of the Crustacean, Oniscus asellus. J. Submicrosc. Cytol., 12: 17-27. Komm, B. S. and Hinsch, G. W. (1985) Oogenesis in the terrestrial hermit crab Coenobita clypeatus (Decapoda, Anomura). I. Previtellogenic oocytes. J. Morphol., 183: 219-224. Junéra, H., Zerbib, C., Martin, M. and Meusy, J.-J. (1977) Evidence for control of vitellogenin synthesis by an ovarian hormone in Orchestia gammarella (Pallas), Crustacea; Amphipoda. Gen. Comp. Endocrinol., 31: 457-462. Junéra, H. and Croisille, Y. (1980) Recherche du lieu de synthése de la vitellogénine chez le Crustacé Amphipode Orchestia gammarella (Pallas). Mise en évidence d’une activation de la synthése protéi- que dans le tissu adipeux sous-épidermique des femelles en vitellogenése secondaire. C. R. Acad. Sci. Paris, Sér. D, 290: 703-706. Croisille, Y. and Junéra, H. (1980) Recherches du lieu de synthése de la vitellogénine chez le Crustacé Amphipode Orchestia gammarella (Pallas). Démonstration, a l’aide d’anticorps spécifiques, de la présence de vitellogénine dans le tissu adipeux sous-épidermique des femelles en vitellogenése secondaire. C. R. Acad. Sci. Paris, Sér. D, 290: 1487-1490. Meusy, J.-J., Junéra, H., Clédon, P. and Martin, M. (1983) La vitellogénine chez un Crustacé Décapode Natantia, Palaemon serratus Pennant. Mise en évidence, comparaison immunologique avec les vitellines, site de synthese et rdle des pédoncules oculaires. Reprod. Nutr. Dév., 23: 625-640. Vazquez-Boucard, C. (1985) Identification pré- liminaire du tissu adipeux chez le Crustacé Déca- pode Penaeus japonicus Bate, a l’aide d’anticorps antilipovitelline. C. R. Acad. Sci. Paris, Sér. III, 300: 95-97. Tom, M., Goren, M. and Ovadia, M. (1987) Loca- lization of the vitellin and its possible precursors in various organs of Parapenaeus longirostris (Crus- tacea, Decapoda, Penaeidae). Int. J. Invertebr. 63 64 66 67 68 69 70 71 72 73 74 Female Reproduction in Malacostracan Crustacea Reprod. Dev., 12: 1-12. Meusy, J.-J., Zerbib, C., Dacheux, F. and Dubois, M.P. (1983) Subcellular localization of vitel- logenin in Crustacean adipocytes by the unlabelled antibody enzyme method. Tissue and Cell, 15: 301-310. Zerbib, C. and Meusy, J. -J. (1983) Electron mi- croscopic observations of the subepidermal fat body changes following ovariectomy in Orchestia gammarellus (Pallas) (Crustacea: Amphipoda). Int. J. Invertebr. Reprod., 6: 123-127. Wyss-Huber, M. and Liischer, M. (1975) Protein synthesis in “fat body” and ovary of the physogas- tric queen of Macrotermes subhyalinus. J. Insect Physiol., 21: 1697-1704. Gutzeit, H.O. (1980) Yolk synthesis in ovarian follicles of Drosophila. Wihelm Roux’s Arch., 189: 221-224. Brennan, M.D., Weiner, A. J., Goralski, T. J. and Mahowald, A. P. (1982) The follicle cells are a major site of vitellogenin synthesis in Drosophila melanogaster. Dev. Biol., 89: 225-236. Charniaux-Cotton, H. (1974) Données nouvelles concernant la vitellogenése des Crustacés Malacos- tracés obtenues chez l’Amphipode Orchestia gam- marellus (Pallas): folliculogenése a partir d’un tissu permanent; action du busulfan; action inhibitrice de Phormone juvénile. C. R. Acad. Sci. Paris, Sér. D, 279: 563-566. Fauvel, C. (1981) Etude de l’ovaire de la Crevette d’eau douce Macrobrachium rosenbergii (de Man) au cours du cycle de reproduction. Premiére description de la folliculogenése secondaire chez un Crustacé Décapode. C. R. Acad. Sci. Paris, Sér. IIT, 292: 547-552. Arcier, J.-M. and Bréhelin, M. (1982) Etude histo- logique et ultrastructurale du tissu folliculaire au cours des cycles de développement ovarien chez Palaemon adspersus (Rathke, 1837). Arch. Biol., 93: 79-97. Charniaux-Cotton, H. (1985) Vitellogenesis and its control in Malacostracan Crustacea. Am. Zool., 25:197-206. Souty, C. (1980) Electron microscopic study of follicle cell development during vitellogenesis in the marine crustacean Isopoda, Idotea balthica basteri. Reprod. Nutr. Dév., 20: 653-663. Jugan, P. and Zerbib, C. (1984) Follicle cell tubu- lar system in the prawn Macrobrachium rosenber- gil. A route for exchanges between haemolymph and vitellogenic oocytes. Biol. Cell, 51: 395-398. Meusy, J.-J., Jugan, P. and Zerbib, C. (1985) Les cellules folliculaires secondaires de la crevette Macrobrachium rosenbergii (Décapode Palaemoni- dé): mise en évidence d’un réseau tubulaire impli- qué dans les échanges entre I’hémolymphe et [B) 76 77 78 79 80 81 83 84 85 86 87 233 lovocyte. IFREMER, Actes de Colloques, n°1: 207-216. Jugan, P. (1985) Régulation de la croissance ovocytaire chez le Crustacé Macrobrachium rosen- bergii (de Man). Démonstration d’une endocytose par récepteurs et approche du mode d’action de la neurohormone inhibitrice de la vitellogenése. These Doctorat, Univ. P. et M. Curie (Paris 6). Blades-Eckelbarger, P.I. and Youngbluth, M. J. (1984) The ultrastructure of oogenesis and yolk formation in Labidocera aestiva (Copepoda: Cala- noida). J. Morphol., 179: 33-46. Charniaux-Cotton, H. (1953) Etude du détermi- nisme des caractéres sexuels secondaires par castra- tion chirurgicale et implantation d’ovaire chez un Crustacé Amphipode (Orchestia gammarella). C. R. Acad. Sci. Paris, 236: 141-143. Charniaux-Cotton, H. (1958) Contrdéle hormonal de la différenciation du sexe et de la reproduction chez les Crustacés supérieurs. Bull. Soc. Zool. Fr., 83: 314-336. Beams, H. W. and Kessel, R. G. (1963) Electron microscope studies on developing crayfish oocytes with special references to the origin of yolk. J. Cell. Biol., 18: 621-650. Zerbib, C. (1973) Contribution a l’étude ultra- structurale de lovocyte chez le crustacé Amphi- pode Orchestia gammarellus Pallas. C. R. Acad. Sci. Paris, Sér. D, 277: 1209-1212. Zerbib, C. (1979) Etude ultrastructurale de l’ovo- cyte en vitellogenése chez les Ecrevisses Astacus astacus et A. leptodactylus. Int. J. Invertebr. Reprod., 1: 289-295. Zerbib, C. (1977) Endocytose ovocytaire chez le Crustacé Amphipode Orchestia gammarellus (Pal- las). Démonstration par la peroxydase. C. R. Acad. Sci. Paris, Sér. D, 284: 757-760. Schade, M. L. and Shivers, R. R. (1980) Structural modification of the surface and cytoplasm of oocytes during vitellogenesis in the lobster, Homarus americanus. An electron microscope- protein tracer study. J. Morphol., 163: 13-26. Jugan, P. and Soyez, D. (1985) Démonstration in vitro de l’inhibition de l’endocytose ovocytaire par un extrait de glandes du sinus chez la Crevette Macrobrachium rosenbergii. C. R. Acad. Paris, Sér. III, 300: 705-709. Anilkumar, G. (1980) Reproductive physiology of female crustaceans. Ph. D. Thesis, Calicut Uni- versity. Kessel, R. G. (1986) Mechanism of protein yolk synthesis and deposition in Crustacean oocytes. Z. Zellforsch. mikrosk. Anat., 89: 17-38. Ganion, L. R. and Kessel, R. G. (1972) Intracellu- lar synthesis, transport and packaging of pro- teinaceous yolk in oocytes of Orconectes immunis. 254 88 89 90 91 93 94 95 96 97 98 99 100 101 J. Cell Biol., 52: 420-437. Eurenius, L. (1973) An electron microscope study on the developing oocytes of the crab Cancer pagurus L. with special reference to yolk formation (Crustacea). Z. Morphol. Tiere, 75: 243-254. Adiyodi, R. G. (1969) Protein metabolism in rela- tion to reproduction on moulting in the crab, Paratelphusa hydrodromous. Part II. RNA activ- ity and protein yolk biosynthesis during normal vitellogenesis and under conditions of acute inani- tion. Indian J. Exp. Biol., 7: 13-16. Hinsch, G. W. (1970) Possible role of intranuclear membranes in nuclear-cytoplasmic exchange in spider crab oocytes. J. Cell Biol., 47: 531-535. Zerbib, C. (1975) Premiéres observations de gra- nules corticaux dans l’ovocyte d’un Crustacé, l’am- phipode Orchestia gammarellus (Pallas). C. R. Acad. Sci. Paris, Sér. D, 281: 1345-1347. Goudeau, M. and Lachaise, F. (1980) ”Endoge- nous yolk” as the precursor of a possible fertiliza- tion envelope in a crab (Carcinus maenas). Tissue and Cell, 12: 503-512. Meusy, J.-J. (1980) Vitellogenin, the extraovarian precursor of the protein yolk in Crustacea: a review. Reprod. Nutr. Dév., 20: 1-21. Ceccaldi, H. J. (1967) Transport des pigments caroténoides dans l’hémolymphe de Carcinus maenas Linné. C. R. Acad. Sci. Paris, Sér. D, 161: 1105-1110. Ghidalia, W. (1985) Structural and_ biological aspects of pigments. In “The Biology of Crustacea, Vol. 9”. Ed. by D.E. Bliss and L. H. Mantel, Academic Press, New York, pp. 301-394. Zagalsky, P. F. (1985) A study of the astaxanthin- lipovitellin, ovoverdin, isolated from the ovaries of the lobster Homarus gammarus (L.). Comp. Biochem. Physiol., 80B, 589-597. Picaud, J.-L. (1978) Contribution a l’étude des propriétés physico-chimiques des protéines spéci- fiques femelles de Porcellio dilatatus Brandt (Crus- tacé Isopode, Oniscoide). C. R. Soc. Biol., 172: 299-303. Wallace, R. A., Walker, S.L. and Hauschka, P. V. (1967) Crustacean lipovitellin. Isolation and characterization of the major high-density lipopro- tein from the eggs of Decapods. Biochemistry, 6: 1582-1590. Fyffe, W. E. and O’Connor, J. D. (1974) Charac- terization and quantification of a crustacean lipo- vitellin. Comp. Biochem. Physiol., 47B: 851-867. Ceccaldi, H. J., Cheesman, D. F. and Zagalsky, P. F. (1966) Quelques propriétés et caractéristiques de lovoverdine. C. R. Soc. Biol., 160: 587-590. Meusy, J.-J. (1972) La gamétogenése et la fraction protéique de l’hémolymphe spécifique du sexe femelle chez quelques Crustacés supérieurs: étude 102 103 104 105 106 107 108 109 110 111 112 J.-J. MEUSY AND G. G. PAYEN descriptive et role des glandes androgénes. Thése Doctorat d’Etat, Univ. Paris 6, AO CNRS n° 6583, pp. 1-165. Croisille, Y., Junéra, H., Meusy, J.-J. and Char- niaux-Cotton, H. (1974) The female-specific pro- tein (vitellogenic protein) in Crustacea with par- ticular reference to Orchestia gammarella (Amphi- poda). Am. Zool., 14: 1219-1228. Arcier, J.-M. and Tournamille, J. (1974) Recher- ches immunochimiques sur les protéines de vitel- logenése chez Palaemon adspersus (Rathke, 1837) (Crustacés Décapodes Natantia). C. R. Acad. Sci. Paris, Sér. D, 278: 495-498. Picaud, J.-L. (1978) Parentés antigéniques des protéines spécifiques femelles chez quelques Crus- tacés Isopodes. C. R. Soc. Biol., 172: 320-324. Byard, E. H. and Aiken, D. E. (1984) The rela- tionship between molting, reproduction, and a female-specific protein in the lobster, Homarus americanus. Comp. Biochem. Physiol., 77A: 749- 751. Tom,M., Goren,M. and Ovadia,M. (1987) Purification and partial characterization of vitellin from the ovaries of Parapenaeus longirostris (Crus- tacea, Decapoda, Penaeidae). Comp. Biochem. Physiol., 87B: 17-23. Vazquez-Boucard, C., Ceccaldi, H. J., Benyamin, Y. and Roustan, C. (1986) Identification, purifica- tion, et caractérisation de la lipovitelline chez un crustacé Décapode Natantia Penaeus japonicus (Bate). J. Exp. Biol. Ecol., 97: 37-50. Ceccaldi, H. J., Daumas, R. and Zagalsky, P. F. (1967) Comparaison des compositions en acides aminés des caroténo-lipoprotéines provenant d’ovaires de trois crustacés et d’un mollusque marin. C. R. Soc. Biol., 161: 1111-1113. Zagalsky, P. F. (1972) Comparative studies on the amino-acid compositions of some _ carotenoid- containing lipoglycoproteins and a glycoprotein from the eggs and ovaries of certain aquatic invertebrates. Comp. Biochem. Physiol., 41B: 385-395. Arcier, J.-M. and Bonami, J. R. (1979) Contribu- tion a l’étude des lipovitellines chez Palaemon adspersus (Rathke, 1837). Arch. Int. Physiol. Biochim., 87: 471-484. Derelle, E., Grosclaude, J., Meusy, J.-J. and Mar- tin, M. (1986) ELISA titration of vitellogenin in the freshwater prawn, Macrobrachium rosenbergii, with monoclonal antibody. Comp. Biochem. Phy- siol., 35B: 1-4. Meusy, J.-J. and Junéra, H. (1979) Analyse préli- minaire de la composition en sous-unités polypepti- diques de la vitellogénine et des lipovitellines du Crustacé Amphipode Orchestia gammarella (Pal- las) C. R. Acad. Sci. Paris, Sér. D, 288: 1415-1418. 113 114 115 116 117 118 119 120 121 122 123 124 Female Reproduction in Malacostracan Crustacea Junéra, H. and Meusy, J.-J. (1982) Vitellogenin and lipovitellins in Orchestia gammarellus (Pallas) (Crustacea, Amphipoda); labelling of subunits after in vivo administration of *H-leucine. Ex- perientia, 38: 252-253. Marzari, R., Ferrero, E., Mosco, A. and Savoini, A. (1986) Immunological characterization of the vitellogenic proteins in Squilla mantis hemolymph (Crustacea, Stomatopoda). Exp. Biol., 45: 75-80. Meusy, J.-J., Junéra, H. and Croisille, Y. (1974) Données sur la synthése de la fraction protéique femelle chez Orchestia gammarella (Crustacé Amphipode), au cours de l’intermue et chez les femelles en repos sexuel. C. R. Acad. Sci. Paris, Sér. D, 279: 587-590. Blanchet-Tournier, M.-F., Meusy, J -J. and Junéra, H. (1980) Mue et vitellogenése chez le Crustacé Amphipode Orchestia gammarella (Pal- las). Etude des effects de la destruction de la région antéro-médiane du protocérébron sur la synthese de la vitellogénine. C. R. Acad. Sci. Paris, Sér. D, 291: 829-832. Souty, C. and Picaud, J.-L. (1981) Vitellogenin synthesis in the fat body of the marine crustacean Isopoda, Idotea balthica basteri, during vitellogene- sis. Reprod. Nutr. Dév., 21: 95-101. Gohar,M., Souty,C. and Picaud, J.-L. (1982) Variations nycthémérales du taux des protéines hémolymphatiques et de la libération de vitellogé- nine chez Porcellio dilatatus Brandt (Crustacé, Isopode). Bull. Soc. Ecophysiol., 7: 79-85. Souty, C. (1983) Démonstration d’une alternance des pics de synthése des fractions, respectivement endogéne et exogéne, du vitellus protéique chez POniscoide Porcellio dilatatus Brandt. C. R. Acad. Sci. Paris, Sér. HI, 296: 221-223. Nakagawa, H. and Ceccaldi, H. J. (1985) Circa- dian variations of haemolymph lipoprotein of Palaemon serratus. Biochem. Syst. Ecol., 13: 345- 348. Vazquez-Boucard,C.G., Moureau,C.E. and Ceccaldi, H. J. (1985) Etude préliminaire des variations circadiennes des protéines de Il’hémoly- mphe de Penaeus japonicus Bate. J. Exp. Mar. Biol. Ecol., 85: 123-133. Aiken, D. E. and Waddy, S. L. (1976) Reproduc- tive biology. In “The Biology and Management of Lobsters”, Vol. 2. Ed. by J. S. Cobb and B. F. Phillips, Academic Press, New York, pp. 215-276. Charniaux-Cotton, H. (1975) Croissance, régé- nération et déterminisme endocrinien des carac- teres sexuels d’Orchestia gammarella (Pallas), Crustacé Amphipode. Ann. Sci. Nat. Zool. Biol. Anim., 19: 411-559. Legrand, J.-J. (1958) Induction de la maturité ovarienne et de la mue parturiale par la féconda- 126 127 128 129 130 131 132 136 137 138 255 tion chez l’Oniscoide Porcellio dilatatus. C. R. Acad. Sci. Paris, Sér. D, 247: 754-757. Reidenbach, J.-M. (1971) Les mécanismes en- docriniens dans le contréle de la différenciation du sexe, la physiologie sexuelle et la mue chez le Crustacé Isopode marin: /dotea balthica (Pallas). Thése Doctorat d’Etat, Univ. Nancy 1, AO CNRS n°: 4874, pp. 1-335. Drach, P. (1955) Systeme endocrinien pédoncu- laire, durée d’intermue et vitellogenése chez Lean- der serratus (Pennant), Crustacé Décapode. C. R. Soc. Biol., 149: 2079-2083. Descouturelle, G. (1978) Influence de l’ablation des pédoncules oculaires sur la longévité, l’évolu- tion ovarienne, et la durée du cycle d’intermue chez la Crevette d’eau douce Atyaephyra des- maresti Millet 1831. Etude des facteurs tempéra- ture, saison et sexualité. Arch. Zool. exp. gén., 119: 433-445. Cheung, T.S. (1966) The interrelations among three hormonal-controlled characters in the adult female shore crab, Carcinus maenas (L.). Biol. Bull., 130: 59-66. Webb, M. (1977) Eyestalk regulation of molt and vitellogenesis in Uca pugilator. Biol. Bull., 153: 630-642. Cheung, T. S. (1969) The environmental and hor- monal control of growth and reproduction in the adult female stone crab, Menippe mercenaria (Say). Biol. Bull., 136: 327-346. Hanstrom, B. (1931) Neue Untersuchungen tber Sinnesorgane und Nervensystem der Crustaceen. I. Z. Morphol. Okol. Tiere, 23: 80-286. Hanstrom, B. (1933) Neue Untersuchungen tber Sinnesorgane und Nervensystem der Crustaceen. II. Zool. Jhb. Anat., 56: 387-520. Hanstr6m, B. (1939) Hormones in Invertebrates, Oxford Univ. Press, London and New York. Scharrer, E. (1928) Die Lichtempfindlichtkeit blin- der Elritzen (Untersuchungen tiber das Zwi- schenhirn der Fische). Z. vergl. Physiol., 7: 1-38. Hodge, M. H. and Chapman, G. B. (1958) Some observations on the fine structure of the sinus gland of a land crab, Gecarcinus lateralis. J. Biophys. Biochem. Cytol., 4: 571-574. Fingerman,M. and Aoto,T. (1959) The neurosecretory system of the dwarf crayfish, Cam- barellus shufeldti, revealed by electron and light microscopy. Trans. Am. Microsc. Soc., 78: 305- 317. Bunt, A.H. and Ashby, E. A. (1967) Ultra- structural study of the sinus gland in the crayfish, Procambarus clarkii. Gen. Comp. Endocrinol., 9: 334-342. Meusy, J.-J. (1968) Précisions nouvelles sur l’ul- trastructure de la glande du sinus d’un Crustacé 256 139 140 141 142 144 145 146 147 148 149 Décapode Brachyoure, Carcinus maenas L. Bull. Soc. Zool. Fr., 93: 291-299. Chaigneau, J. (1983) Neurohemal organs in Crus- tacea. In “Neurohemal Organs in Arthropods”. Ed. by A. P. Gupta, Charles C Thomas, Spring- field, pp. 53-89. Panouse, J. B. (1943) Influence de l’ablation des pédoncules oculaires sur la croissance de l’ovaire chez la crevette Leander serratus. C. R. Acad. Sci. Paris, 217: 553-555. Panouse, J. B. (1946) Recherches sur les phé- noménes humoraux chez les Crustacés. L’adapta- tion chromatique et la croissance ovarienne chez la Crevette Leander serratus. Ann. Inst. Oceanogr. Monaco, 23: 65-147. Primavera, H. and Borlongan, E. (1978) Ovarian rematuration of ablated sugpo prawn Penaeus monodon Fabricius. Ann. Biol. Anim. Bioch. Biophys., 18: 1067-1072. Anilkumar, G. and Adiyodi, K. G. (1980) Ovarian growth, induced by eyestalk ablation during the prebreeding season, is not normal in the crab, Paratelphusa hydrodromous. Int. J. Invertebr. Reprod., 2: 95-105. Gabe, M. (1966) Neurosecretion. International Series of Monographs in Pure and Applied Biolo- gy, Zoology Division, Vol. 28, Pergamon Press. Armitage, K.B., Buikema, A.L.,Jr. and Wil- lems, N. J. (1973) The effect of photoperiod on organic constituents and molting of the crayfish Orconectes nais (Faxon). Comp. Biochem. Phy- siol., 44A: 431-456. Quackenbush, L. S. and Herrnkind, W. F. (1981) Regulation of molt and gonadal development in the spiny lobster Panulirus argus (Crustacea: Palinur- idae): effect of eyestalk ablation. Comp. Biochem. Physiol., 69A: 523-527. Vernet-Cornubert, G. (1960) Influence de l’abla- tion des pédonculaires sur la mue, la ponte et les caractéres sexuels externes de Pisa tetraodon Pen- nant. Bull. Inst. Océanogr. Monaco, 1186: 1-24. Hinsch, G. W. (1972) Some factors controlling reproduction in the spider crab, Libinia emargina- ta, Biol. Bull., 143: 358-366. Passano, L. M. (1960) Molting and its control. In “The Physiology of Crustacea, Vol. 1”. Ed. by T. H. Waterman, Academic Press, New York and London, pp. 473-536. Carlisle, D. B. (1953) Studies on Lysmata seti- caudata Risso (Crustacea Decapoda). V. The ovarian inhibiting hormone and the hormonal inhibition of sex-reversal. Pubbl. Staz. Zool. Napo- li, 24: 355-372. Martin, G. (1982) Etude ultrastructurale de la régénération de la glande du sinus chez l’oniscoide Porcellio dilatatus Brandt; données com- 153 155 157 158 159 160 161 162 J.-J. MEUSY AND G. G. PAYEN plémentaires sur l’origine des terminaisons de cet organe neurohémal. J. Physiol., Paris, 78: 558- 565. Besse, G., Juchault, P., Legrand, J.-J. and Moc- quard, J.-P. (1969) Contribution a l’étude de la physiologie sexuelle femelle de Ligia oceanica L. (Crustacé Oniscoide). Différenciation des oosté- gites et contrdle neurohumoral de la maturation ovarienne. C. R. Acad. Sci. Paris., Sér. D, 269: 733-736. Mocquard, J.-P., Besse, G., Juchault, P., Legrand, J.-J., Maissiat, J. and Noulin, G. (1971) Contribu- tion a analyse du contr6le neurohumoral de la croissance, de la mue et de la physiologie sexuelle male et femelle chez l’Oniscoide Ligia oceanica L. (Crustacé Isopode). Ann. Embryol. Morphol., 4: 45-63. Blanchet-Tournier, M.-F. (1987) Mise en évidence dune activité neurohormonale inhibitrice de la vitellogenése chez l'amphipode Orchestia gam- marella. Can. J. Zool., 65: 1917-1922. Gohar, M., Souty-Grosset,C., Martin,G. and Juchault, P. (1984) Mise en évidence d’une inhibi- tion de la synthése de la vitellogénine par un facteur neurohumoral (V.I.H.) chez le Crustacé Isopode terrestre Porcellio dilatatus Brandt. C. R. Acad. Sci. Paris, Sér. II, 299: 785-787. Ferlund, P. (1974) Structure of the red-pigment- concentrating hormone of the shrimp, Pandalus borealis. Biochim. Biophys. Acta, 371: 304-311. Kleinholz, L. H. (1976) Crustacean neurosecretory hormones and physiological specificity. Am. Zool., 16: 151-156. Huberman, A., Arechiga,H., Cimet, A., De La Rosa, J. and Aramburo, C. (1979) Isolation and purification of a neurodepressing hormone from the eyestalks of Procambarus bouvieri (Ortmann). Eur. J. Biochem., 99: 203-208. Bomirski, A., Arendarczyk, M., Kawinska, E. and Kleinholz, L. H. (1981) Partial characterization of Crustacean gonad-inhibiting hormone. Int. J. In- vertebr. Roprod., 3: 213-219. Klek-Kawinska, E. and Bomirski, A. (1975) O- vary-inhibiting hormone activity in shrimp (Cran- gon crangon) eyestalks during the annual reproduc- tive cycle. Gen. Comp. Endocrinol., 25: 9-13. Quackenbush, L. S. and Herrnkind, W. F. (1983) Partial characterization of eyestalk hormones con- trolling molt and gonadal development in the spiny lobster Panulirus argus. J. Crust. Biol., 3: 34-44. Quackenbush,L.S. and Keeley,L.L. (1986) Vitellogenesis in the shrimp, Penaeus vannamei. Am. Zool., 26: 56A. Soyez,D., Van Deijnen, J.E. and Martin, M. (1987) Isolation and characterization of a vitel- logenesis inhibiting factor from sinus glands of the 164 165 166 167 168 169 170 171 172 173 174 175 176 177 Female Reproduction in Malacostracan Crustacea lobster, Homarus americanus. J. Exp. Zool., 244: 479-484. Meusy, J.-J., Martin, M., Soyez, D., van Deijnen, J. E. and Gallo, J.-M. (1987) Immunochemical and immunocytochemical studies of the crustacean Vitellogenesis Inhibiting Hormone (VIH). Gen. Comp. Endocrinol., 67: 333-341. Gorgels-Kallen, J.L. and VanHerp,F. (1981) Localization of Crustacean Hyperglycemic Hor- mone (CHH) in the X-organ sinus gland complex in the eyestalk of the crayfish, Astacus leptodactylus (Nordmann, 1842). J. Morphol., 170: 347-355. Enami, M. (1951) The source and activities of two chromatophorotropic hormones in crabs of the genus Sesarma. II. Histology of incretory elements. Biol. Bull., 101: 241-258. Matsumoto, K. (1958) Morphological studies on the neurosecretion in crabs. Biol. J. Okayama Univ., 4: 103-176. Otsu, T. (1960) Precocious development of the ovaries in the crab, Potamon dehaani, following implantation of the thoracic ganglion. Annot. Zool. Japon., 33: 90-96. Otsu, T. (1963) Bihormonal control of sexual cycle in the freshwater crab, Potamon dehaani. Embryologia, 8: 1-20. Gomez, R. (1965) Acceleration of development of gonads by implantation of brain in the crab, Paratelphusa hydrodromous. Naturwissenschaften, 9: 216. Gomez, R. and Nayar, K. K. (1965) Certain en- docrine influences in the reproduction of the crab, Paratelphusa hydrodromous. Zool. Jb. Abt. Phy- siol., 71: 694-701. Hinsch, G. and Bennet, D. C. (1979) Vitellogene- sis stimulated by thoracic ganglion implants into destalked immature spider crabs, Libinia emargi- nata. Tissue and Cell, 11: 345-351. Babu, O., Shyamasudari, E.T. and Rao, K. H. (1980) Correlative changes in neurosecretion and ovarian growth after bilateral ablation of eyestalks in the crab, Menippe rumphii. Indian J. Exp. Biol., 18: 265-268. Takayanagi, H., Yamamoto, Y. and Takeda, N. (1986) An ovary-stimulating factor in the shrimp, Paratya compressa. J. Exp. Zool., 240: 203-209. Charniaux-Cotton, H. (1960) Sex Determination. In “The Physiology of Crustacea, Vol. 1”. Ed. by T. H. Waterman, Academic Press, New York and London, pp. 411-447. Charniaux-Cotton, H. (1962) Preuve — ex- périmentale de l’autodifférenciation ovarienne chez Orchestia montagui Audouin (Crustacé Amphipode). C. R. Acad. Sci. Paris, 254: 2836- 2838. Junéra,H., Martin,M., Solari, A. and Meusy, 178 179 180 181 183 184 185 186 187 188 189 190 257) J.-J. (1977) Détermination du poids moléculaire de la vitellogénine et des lipovitellines d’Orchestia gammarella, Crustacé Amphipode. C. R. Acad. Sci. Paris, Sér. D, 285: 909-912. Picaud, J.-L. and Souty,C. (1981) Approche quantitative de l’influence de l’ovariectomie sur la synthése de la vitellogénine chez Porcellio dilatatus Brandt (Crustacé, Isopode). C. R. Acad. Sci. Paris, Sér. HI, 293: 479-482. Souty-Grosset, C. and Juchault, P. (1987) Etude de la synthése de la vitellogénine chez les males intersexués d’Armadillidium vulgare Latreille (Crustacé Isopode Oniscoide): Comparaison avec les males et les femelles intactes ou ovariectomi- sées. Gen. Comp. Endocrinol., 66: 163-170. Echalier, G. (1959) L’organe Y et le déterminisme de la croissance et de la mue chez Carcinus maenas (L.), Crustacé Décapode. Ann. Sci. Nat. Zool., 1: 1-59. Kater, S. B. and Spaziani, E. (1971) Incorporation of cholesterol '*C into crab cholesterol pools and ecdysones as a function of the molting cycle. Am. Zool., 11: 672. Willig, A. and Keller, R. (1976) Biosynthesis of a- and f-ecdysone by the crayfish Orconectes limosus in vivo and by its Y-organs in vitro. Experientia, 32: 936-937. Chang, E.S., Sage,B.A. and O’Connor, J. D. (1976) The qualitative and quantitative determina- tions of ecdysones in tissues of the crab, Pachy- grapsus crassipes, following molt induction. Gen. Comp. Endocrinol., 30: 21-33. Chang, E.S. and O’Connor, J. D. (1977) Secre- tion of a- and f-ecdysone by crab Y-organs in vitro. Proc. Natl. Acad. Sci., USA, 74: 615-618. Keller, R. and Schmid, E. (1979) Jn vitro secretion of ecdysteroids by the Y-organs and lack of secretion by mandibular organs of the crayfish following molt induction. J. Comp. Physiol., 130: 347-353. Arvy, L., Echalier, G. and Gabe, M. (1954) Mod- ification de la gonade de Carcinides maenas L. (Crustacé Décapode) aprés ablation bilatérale de Porgane Y. C. R. Acad. Sci. Paris, 239: 1853-1855. Arvy,L., Echalier,G. and Gabe,M. (1956) Organe Y et gonade chez Carcinides (Carcinus) maenas L. Ann. Sci. Nat. Zool., 18: 263-267. Demeusy, N. (1959) Pédoncules oculaires, glandes de mue et appareil génital chez Carcinus maenas L. C. R. Acad. Sci. Paris, 248: 2652-2654. Demeusy, N. (1962) Observations sur la matura- tion ovarienne du crabe Carcinus maenas L. aprés ablation des glandes de mue chez les femelles adultes. Premiers résultats. C. R. Acad. Sci. Paris, 255: 3062-3064. Demeusy, N. (1962) Réle de la glande de mue dans 258 191 192 193 194 195 197 198 199 200 Pévolution ovarienne du crabe Carcinus maenas Linné. Cah. Biol. Mar., 3: 37-56. Besse, G. and Maissiat, J. (1971) Action de la glande de mue sur la vitellogenése du Crustacé Isopode Porcellio dilatatus (Brandt). C. R. Acad. Sci. Paris, Sér. D, 273: 1975-1978. Meusy, J.-J., Blanchet,M.-F. and Junéra, H. (1977) Mue et vitellogenése chez le Crustacé Amphipode Orchestia gammarella Pallas. Il. Etude de la synthése de la vitellogénine (“fraction protéi- que femelle” de ’hémolymphe) apres destruction des organes Y. Gen. Comp. Endocrinol., 33: 35- 40. Suzuki, S. (1986) Effect of Y-organ ablation on oocyte growth in the terrestrial isopod, Armadilli- dium vulgare. Biol. Bull., 170: 350-355. Andrieux, N., Porcheron, P., Berreur- Bonnenfant, J. and Dray, F. (1976) Détermination du taux d’ecdysone au cours du cycle d’intermue chez le crabe Carcinus maenas. Comparaison entre individus sains et parasités par Sacculina carcini. C. R. Acad. Sci. Paris, Sér. D, 283: 1429-1432. Lachaise, F., Lagueux,M. Feyereisen,R. and Hoffmann, J. A. (1976) Métabolisme de l’ecdy- sone au cours du développement de Carcinus maenas (Brachyura, Decapoda). C. R. Acad. Sci. Paris, Sér. D, 282: 943-946. Blanchet, M.-F., Porcheron,P. and_ Dray, F. (1979) Variations du taux des ecdystéroides au cours des cycles de mue et de vitellogenése chez le Crustacé Amphipode Orchestia gammarellus. Int. J. Invertebr. Reprod., 1: 133-139. Baldaia, L., Porcheron, P., Coimbra, J. and Cas- sier,P. (1984) Ecdysteroids in the shrimp Palaemon serratus: relations with molt cycle. Gen. Comp. Endocrinol., 55: 437-443. Carlisle, D. B. (1957) On the hormonal inhibition of moulting in decapod Crustacea. II. The terminal anecdysis in crabs. J. Mar. Biol. Assoc. U. K., 36: 291-307. Chaix, J.-C., Trilles, J.-P. and Vernet, G. (1976) Dégénérescence de l’organe Y chez les males pubéres d’Acanthonyx lunulatus (Risso) (Crus- tacea, Oxyrhyncha). C. R. Acad. Sci. Paris, Sér. D, 283: 523-525. Souty,C., Besse,G. and Picaud, J.-L. (1982) Stimulation par l’ecdysone du taux hémolymphati- que de la vitellogénine chez le Crustacé Isopode terrestre, Porcellio dilatatus Brandt. C. R. Acad. Sci. Paris, Sér. III, 294: 1057-1059. Blanchet, M.-F., Junéra,H. and Meusy, J.-J. (1975) Mue et vitellogenése chez Orchestia gam- marella Pallas, Crustacé Amphipode: étude de la synthése de la fraction protéique femelle apres introduction d’ecdystérone. Experientia, 31: 865- 867. 202 203 204 205 206 208 209 210 211 212 213 214 215 J.-J. MEUSY AND G. G. PAYEN Picaud, J.-L. and Souty, C. (1980) Démonstration par immunoautoradiographie de la synthése de la vitellogénine par le tissu adipeux de Porcellio dilatatus Brandt (Crustacé, Isopode). C. R. Acad. Sci. Paris, Sér. D, 290: 1019-1021. Gohar, M. and Souty, C. (1983) Mise en évidence in vitro d’une synthése et d’une libération de la vitellogénine dans le tissu adipeux male de Porcel- lio dilatatus Brandt (Crustacé Isopode terrestre). C. R. Acad. Sci. Paris, Sér. D, 297: 145-148. Lachaise, F. and Hoffmann, J. (1982) Ecdysteroids and embryonic development in the shore crab, Carcinus maenas. Hoppe-Seyler’s Z. Physiol. Chem., 363: 1059-1067. Lachaise, F., Goudeau, M., Hetru, C., Kappler, C. and Hoffmann, J. A. (1981) Ecdysteroids and ovarian development in the shore crab, Carcinus maenas. Hoppe-Seyler’s Z. Physiol. Chem., 362: 521-529. Payen, G. and Costlow, J. D. (1977) Effects of a juvenile hormone mimic on male and female gametogenesis of the mud-crab, Rhithropanopeus harristi (Gould) (Brachyura: Xanthidae). Biol. Bull., 152: 199-208. Hinsch, G. W. (1981) Effects of juvenile hormone mimics on the ovary in the immature spider crab, Libinia emarginata. Int. J. Invertebr. Reprod., 3: 237-244. Slama, K., Romanuk, M. and Sorm, F. (1974) In- sect hormones and bioanalogues. Springer-Verlag, Wien and New York, pp. 1-477. Engelmann, F. (1983) Vitellogenesis controlled by juvenile hormone. In “Endocrinology of Insects, Vol. 1”. Ed. by R. G. H. Downer and H. Laufer, Alan R. Liss Inc., New York, pp. 259-270. Le Roux, A. (1968) Description d’organes man- dibulaires nouveaux chez les Crustacés Décapodes. C. R. Acad. Sci. Paris, Sér. D, 266: 1414-1417. Bazin, F. (1976) Mise en évidence des caractéres cytologiques des glandes stéroidogénes dans les glandes mandibulaires et les glandes Y du crabe Carcinus maenas (L.) normal et épédonculé. C. R. Acad. Sci. Paris, Sér. D, 282: 739-741. Yudin, A. I., Diener, R. A., Clark, W. H., Jr. and Chang, E. S. (1980) Mandibular gland of the blue crab, Callinectes sapidus. Biol. Bull., 159: 760-772. Taketomi, Y. and Kawano, Y. (1985) Ultrastruc- ture of the mandibular organ of the shrimp, Penaeus japonicus, in untreated and experimental- ly manipulated individuals. Cell. Biol. Int. Rep., 9: 1069-1074. Byard, E. H. and Shivers, R. R. (1975) The man- dibular organ of the lobster, Homarus americanus. Cell Tissue Res., 162: 13-22. Le Roux, A. (1983) Réactions de l’organe man- dibulaire a l’ablation des pédoncules oculaires chez 216 217 218 219 220 221 222 223 224 225 226 Female Reproduction in Malacostracan Crustacea les larves et les juvéniles de Palaemonetes varians (Leach). C. R. Acad. Sci. Paris, Sér. III, 296: 697-— 700. Hinsch, G. W. (1980) Effect of mandibular organ implants upon the spider crab ovary. Trans. Am. Microsc. Soc., 99: 317-322. Schneiderman, H. A. and Gilbert, L.I. (1958) Substances with juvenile hormone activity in Crus- tacea and other invertebrates. Biol. Bull., Woods Hole, Mass., 115: 530-535. Laufer, H., Landau, M., Borst, D. and Homola, E. (1986) The synthesis and regulation of methyl- farnesoate, a new juvenile hormone for crustacean reproduction. In “Advances in Invertebrate Repro- duction 4”. Ed. by M. Porchet, J.-C. Andries and A. Dhainaut, Elsevier, North Holland, pp. 135- 143. Laufer, H., Borst, D., Baker, F. C., Carrasco, C., Sinkus,M., Reuter,C.C., Tsai,L.W. and Schooley, D. A. (1987) Identification of a juvenile hormone-like compound in a crustacean. Science, 235: 202-205. Couch, E. F., Hagino,N. and Lee, J. W. (1987) Changes in estradiol and progesterone im- munoreactivity in tissues of the lobster, Homarus americanus, with developing and immature ova- ries. Comp. Biochem. Physiol., 87A: 765-770. Takayanagi, H., Yamamoto, Y. and Takeda, N. (1986) Ovary-stimulating pheromone in the fresh- water shrimp, Paratya compressa. J. Exp. Zool., 240: 397-400. Jassem, W., Juchault,P. and Mocquard, J.-P. (1982) Déterminisme de la reproduction saison- niére des femelles d’Armadillidium vulgare Latr. (Crustacé, Isopode, Oniscoide). V. Rdle du male dans le cycle de reproduction des femelles (induc- tion et durée de la période reproduction). Ann. Sci. Nat. Zool., Paris, 4: 195-201. Besse, G., Juchault, P., Legrand, J.-J. and Moc- quard, J.-P. (1970) Modification de lélectro- phorégramme de Il’hémolymphe des Oniscoides (Crustacés Isopodes) par action de Il’hormone androgéne. C. R. Acad. Sci. Paris, Sér. D, 270: 3276-3279. Meusy, J.-J., Junéra, H. and Croisille, Y. (1971) Recherche de la “fraction protéique femelle” dans l’hémolymphe des femelles d’Orchestia gammarella Pallas ayant subi une inversion expérimentale du sexe. C. R. Acad. Sci. Paris, Sér. D, 273: 592-594. Bomirski, A. and Klek-Kawinska, E. (1976) Stim- ulation of oogenesis in the sand shrimp, Crangon crangon, by human gonadotrophin. Gen. Comp. Endocrinol., 30: 239-242. Souty, C. and Picaud, J.-L. (1984) Effet de l’injec- tion d’une gonadotropine humaine sur la synthése et la libération de la vitellogénine par le tissu 20 230 234 237 238 239 259 adipeux du Crustacé Isopode marin /dotea balthica basterii Audouin. Gen. Comp. Endocrinol., 54: 418-421. Kulkarni, G. K., Nagabhushanam, R. and Joshi, P.K. (1979) Effect of progesterone on ovarian maturation in a marine penaeid prawn Para- penaeopsis hardwickii (Miers, 1878). Indian. J. Exp. Biol., 17: 986-987. Donahue, J. K. (1948) Fluorimetric and biological determination of estrogens in the eggs of the American lobster (Homarus americanus). Proc. Soc. Exp. Biol. Med., 69: 179-181. Donahue, J. K. (1952) Studies on ecdysis in the American lobster (Homarus Americanus). I. The lobster egg as a source of estrogenic hormone. State of Maine Dept. of Sea and Shore Fisheries, Res. Bull., no. 8. Donahue, J. K. (1957) Chromatographic iden- tification of lobster estrogen. State of Maine Dept. of Sea and Shore Fisheries, Res. Bull., no. 28. Jeng,S.S., Wan, W.C.-M. and Chang, C. F. (1978) Existence of an oestrogen-like compound in the ovary of the shrimp Parapenaeus fissurus. Gen. Comp. Endocrinol., 36: 211-214. Ollevier, F., De Clerck, D., Diederik, H. and De Loof, A. (1986) Identification of nonecdysteroid steroids in hemolymph of both male and female Astacus leptodactylus (Crustacea) by gas chroma- tography-mass spectrometry. Gen. Comp. Endo- crinol., 61: 214-228. Doane, W. W. (1973) Role of hormones in insect development. In “Developmental Systems: In- sects”. Ed. by S. J. Counce and C. H. Waddington, Academic Press, London and New York, p. 436. Carlisle, D.B. and Butler,C.G. (1956) The “queen substance” of honeybees and the ovary- inhibiting hormone of crustaceans. Nature, 177: 276-277. Sastry, A. N. (1983) Ecological aspects of repro- duction. In “The Bioloby of Crustacea. Vol. 8”. Ed. by F.J.Vernberg and W.B. Vernberg, Academic Press, New York, pp. 179-270. Segerstrale,S.G. (1970) Light control of the reproductive cycle of Pontoporeia affinis Lindstr6m (Crustacea Amphipoda). J. Exp. Mar. Biol. Ecol., 5: 272-275. Steele, V. J., Steele, DD. H. and MacPherson, B. R. (1977) The effect of photoperiod on the reproductive cycle of Gammarus setosus Demen- tieva, 1931. Crustaceana, suppl. 4: 58-63. De March, B. G. E. (1977) The effects of photo- period and temperature on the induction and termination of reproductive resting stage in the freshwater amphipod Hyalella azteca (Saussure). Can. J. Zool., 55: 1595-1600. Steele, V. J. (1981) The effect of photoperiod on 260 240 242 243 245 246 247 248 the reproductive cycle of Gammarus lawrencianus Bousfield. J. Exp. Mar. Biol. Ecol., 53: 1-7. Steele, V.J. and Steele, D.H. (1986) The in- fluence of photoperiod on the timing of reproduc- tive cycles in Gammarus species (Crustacea, Amphipoda). Am. Zool., 26: 459-467. De March, B. G. E. (1982) Decreased day length and light intensity as factors inducing reproduction in Gammarus lacustris lacustris Sars. Can J. Zool., 60: 2962-2965. McQueen, D. J. and Steel, C. G. (1980) The role of photoperiod and temperature in the initiation of reproduction in the terrestrial isopod Oniscus asellus Linnaeus. Can. J. Zool., 58: 235-240. Steel, C. G. (1980) Mechanisms of coordination between moulting and reproduction in the terres- trial Isopod Crustacea. Biol. Bull., 159: 206-218. Juchault, P., Jassem,W. and Mocquard, J.-P. (1982) Déterminisme de la reproduction saison- niere des femelles d’Armadillidium vulgare Latr. (Crustacé, Isopode, Oniscoide). VI. Mise en évi- dence dune photopériode critique permettant I’en- trée en reproduction; modalités du maintien en reproduction. Ann. Sci. Nat. Zool., Paris, 4: 203- 210. Laubier-Bonichon, A. (1975) Induction de la maturation sexuelle et ponte chez la crevette Penaeus japonicus Bate en millieu contrdlé. C. R. Acad. Sci. Paris, Sér. D, 281: 2013-2016. Laubier-Bonichon, A. (1978) Ecophysiologie de la reproduction chez la crevette Penaeus japonicus. Trois années d’expérience en milieu contrdlé. Oceanol. Acta, 1: 135-150. Stephens, G. C. (1952) Mechanism regulating the reproductive cycle in the crayfish, Cambarus. 1. The female cycle. Physiol. Zool., 25: 70-83. Aiken, D. E. (1969) Ovarian maturation and egg- laying in the crayfish Orconectes virilis: influence of temperature and photoperiod. Can. J. Zool., 47: 931-935. Suko, T. (1958) Studies on the development of the crayfish. V. The histological changes of the de- velopmental ovaries influenced by the condition of darkness. Sci. Rep. Saitama Univ., Sér. B, 3: 67- 78. Lowe, M. E. (1961) The female reproductive cycle of the crayfish Cambarellus shufeldti: the influence of environmental factors. Tulane Stud. Zool., 8: 157-176. Kracht, D. (1972) Application de conditions hiver- nales et estivales, épédonculation et action d’une ecdystérone chez lEcrevisse femelle adulte Orconectes limosus R., pendant la période d’anec- dysis: conséquence sur la ponte et sur la mue. C. R. Acad. Sci. Paris, Sér. D, 275: 1677-1680. Rice, P.R. and Armitage, K. B. (1974) The in- 259 261 262 263 264 J.-J. MEUSY AND G. G. PAYEN fluence of photoperiod on processes associated with molting and reproduction in the crayfish Orconectes nais (Faxon). Comp. Biochem. Phy- siol., 47A: 243-259. Nelson, K. (1986) Photoperiod and reproduction in lobsters (Homarus), Am. Zool., 26: 447-457. Pradeille-Rouquette, M. (1976) Role de la photo- période dans la fonction de reproduction des femelles du crabe Pachygrapsus marmoratus. C. R. Acad. Sci. Paris, Sér. D, 282: 199-201. Nagabhushanam, R. and Farooqui, U. M. (1981) Photoperiodic stimulation of ovary and testis in the immature marine crab, Scylla serrata Forskal. Indian J. Mar. Sci., 10: 396-398. Lagardere, J.-P. (1982) Effets du bruit sur le métabolisme, la croissance et la reproduction de Crangon crangon (L.) (Décapode, Natantia). C. R. Acad. Sci. Paris, Sér. III, 294: 425-428. Von Hagen, H. O. (1962) Freilandstudien zur sex- ual und fortplanzungs Biologie von Uca tangeri in Anadalusien. Z. Morphol. Oekol. Tiere, 51: 611- 725: Von Hagen, H. O. (1970) Adaptations to the spe- cial intertidal level of habitat in Ocypodid crabs (Decapoda, Brachyura). Forma Functio, 2: 361- 413. Zuker, N. (1976) Behavioral rhythms in the fiddler crab Uca terpsichores. In “Biological Rhythms in the Marine Environment”. Ed. by P.J. De Coursey., Univ. of South Carolina, Carolina Press, Columbia, pp. 145-159. Pérez, C. (1933) Processus de résorption phago- cytaire des ovocytes dans l’ovaire chez les Macro- podia sacculinées. C. R. Soc. Biol. Fr., 112: 1049- 1051. Veillet, A. (1945) Recherches sur le parasitisme des Crabes et des Galathées par les Rhizocéphales et les Epicarides. Ann. Inst. Oceanogr. Paris, 22: 193-341. Cornubert, G. (1952) Influence de la Sacculine Sacculina carcini Thompson sur le Crabe Pachyg- rapsus marmoratus Fabricius. C. R. Acad. Sci. Paris, 234: 1218-1220. Payen,G.G. (1975) Effets masculinisants des glandes androgénes implantées chez la femelle pubére pédonculectomisée de Rhithropanopeus harrisii (Gould) (Crustacé Décapode Brachyoure). C. R. Acad. Sci. Paris, Sér. D, 280: 111-115. Rubiliani, C., Rubiliani-Durozoi, M. and Payen, G. G. (1980) Effets de la Sacculine sur les gonades, les glandes androgénes et le systeme nerveux central des Crabes Carcinus maenas (L.) et C. mediterraneus Czerniavsky. Bull. Soc. Zool. Fr., 105: 95-100. Andrieux, N., De Freschville, J. and Herberts, C. (1986) Etude des protéines vitellines de ’hémo- 266 267 268 269 270 271 272 PA) 274 215 276 Pig 278 Female Reproduction in Malacostracan Crustacea lymphe et de l’ovaire chez le Crustacé Décapode Carcinus maenas; incidence du parasite Sacculina carcini (Crustacé Rhizocéphale). Can. J. Zool., 64: 2279-2287. Leroux, M.-L. (1933) Recherches sur la sexualité des Gammariens. Bull. Biol. Fr. Berg., suppl. 16: 24-46. Campbell-Parmentier, F. (1963) Vitellogenése, maturation des ovocytes, accouplement et ponte en relation avec l’intermue chez Orchestia gammarella (Pallas), Crustacé, Amphipode, Talitridae. Bull. Soc. Zool. Fr., 88: 474-488. Mathieu-Capderou, C. (1980) Relation entre la maturation ovocytaire et l’exuviation chez le Crus- tacé Amphipode Orchestia gammarellus (Pallas). C. R. Acad. Sci. Paris, Sér. D, 290: 1495-1498. Clédon, P. (1985) Analyse cytologique et ex- périmentale de la maturation et de l’activation de Povocyte de la Crevette Palaemon serratus (Crusta- cé Décapode Natantia). C. R. Acad. Sci. Paris, Sér. D, 301: 317-322. Clédon, P. (1986) Study on oocyte maturation and activation of the common prawn Palaemon serratus (Pennant): Relationship between oocyte matura- tion and the molt cycle cytological aspects. Gamete Res., 13: 353-362. Lanot, R. Thiebold, J., Lagueux, M., Goltzene, F. and Hoffmann, J. A. (1987) Involvement of ecdy- sone in the control of meiotic reinitiation in oocytes of Locusta migratoria (Insecta, Orthoptera). Dev. Biol., 121: 174-181. Lachaise, F. and Hoffmann, J. (1977) Ecdysone et développement ovarien chez un Décapode, Carci- nus maenas. C. R. Acad. Sci. Paris, Sér. D, 285: 701-704. Maller, J. L. (1985) Oocyte maturation in amphib- iens. In “Developmental Biology, A Comprehen- sive Synthesis. Vol. 1 Oogenesis. Ed. by L. W. Browder, Plenum Press, New York, London, pp. 289-311. Fauvel, C. (1983) L’ovaire de Macrobrachium resenbergii (De Man) (Crustacé Décapode) au moment de la ponte. Description de l’ovulation. C. R. Acad. Sci. Paris, Sér. II, 296: 1053-1058. Matsumoto, K. (1958) Neurosecretion in the tho- racic ganglion of the crab, Eriocheir japonicus. Biol. Bull., 106: 60-68. Brown, G. D. (1966) Ultrastructural studies of sperm morphology and sperm-egg interaction in the decapod Callinectes sapidus. J. Ultrastruct. Res., 14: 425-440. Hinsch, G. W. (1971) Penetration of the oocyte envelope by spermatozoa in the spider crab. J. Ultrastruct. Res., 35: 86-97. Talbot, P. and Chanmanon, P. (1980) Morpholo- gical features of the acrosome reaction of lobster 280 281 iw) lo.2} i) 284 iw) oo Nn 286 287 288 290 291 261 (Homarus) sperm and the role of the reaction in generating forward sperm movement. J. Ultra- struct. Res., 70: 287-297. Lynn, J. W. and Clark, W. H. (1983) A morpholo- gical examination of sperm-egg interaction in the freshwater prawn, Macrobrachium rosenbergii. Biol. Bull., 164: 446-458. Goudeau, M. (1982) Fertilization in a crab. I. Early events in the ovary, and cytological aspects of the acrosome reaction and gamete contacts. Tissue and Cell, 14: 97-111. Clark, W.H., Yudin, A.I., Griffin, F.J. and Shigekawa, K. (1984) The control of gamete activation and fertilization in the marine Penaeidae Sicyonia ingentis. In “Advances in Invertebrate Reporduction. Vol. 3”. Ed. by W. Engels, W. H. Clark, Jr., A. Fischer, P. J. W. Olive and D.F. Went, Elsevier, North Holland, pp. 459-472. Binford, R. (1913) The sperm cells and the process of fertilization in the crab, Menippe mercenaria. J. Morphol., 24: 147-201. Bloch, F. (1935) Contribution a l’étude des ga- metes et de la fécondation chez les Crustacés Décapodes. Trav. Station Zool. Wimereux, 12: 183-279. Hoestlandt, H. (1948) Recherches sur la biologie de JlEriocheir sinensis en France (Crustacé Brachyoure). Ann. Inst. Oceanogr., 24: 1-116. Pradeille-Rouquette, M. (1976) Etude de la fonc- tion de reproduction chez les femelles du crabe Pachygrapsus marmoratus (F.) et de différents facteurs qui lui sont liés. Cah. Biol. Mar., 17: 387- 403. Goudeau, M. and Lachaise, F. (1980) Fine struc- ture and secretion of the capsule enclosing embryo in a crab (Carcinus maenas). Tissue and Cell, 12: 287-308. Goudeau, H., Kubisz, P. and Goudeau, M. (1984) Mise en évidence du potentiel de fécondation chez les Crustacés Décapodes Brachyoures Carcinus maenas et Maia squinado. C. R. Acad. Sci. Paris, Sér. III, 299: 167-172. Goudeau, M. and Goudeau, H. (1986) The re- sumption of meiotic maturation of the oocyte of the prawn Palaemon serratus is regulated by an in- crease in extracellular Mg*?* during spawning. Dev. Biol., 118: 361-370. Goudeau, M. and Goudeau, H. (1986) External Mg’* is required for hyperpolarization to occur in ovulated oocytes of the prawn Palaemon serratus. Dev. Biol., 118: 371-378. Goudeau, H. and Goudeau, M. (1985) Fertiliza- tion in crabs: IV. The fertilization potential consists of a sustained egg membrane hyperpolarization. Gamete Res., 11: 1-17. Goudeau, H. and Goudeau, M. (1986) Electrical 262 293 nN \o nN 296 297 298 299 300 301 302 303 304 305 and morphological responses of the lobster egg to fertilization. Dev. Biol., 114: 325-335. Spalding, J. F. (1942) The nature and formation of the spermatophore and sperm-plug in Carcinus maenas. Q. J. Microsc. Sci., 83: 399-422. Anghelou-Spiliotis, A. and Goudeau, M. (1982) Analyse au microscope électronique a balayage de la paroi limitant la lumiére du conduit génital de la femelle adulte de Carcinus maenas (L), Crustacé Décapode Brachyoure. C. R. Acad. Sci. Paris, Sér. Il], 294: 617-622. Herrick, F. H. (1909) Natural history of the amer- ican lobster. Bull. U. S. Bur. Fish., 29: 149-408. Clark, W. H., Lynn, J. W., Yudin, A. I. and Per- syn, H.O. (1980) Morphology of the cortical reaction in the eggs of Penaeus aztecus. Biol. Bull., 158: 175-186. Goudeau, M. and Becker, J. (1982) Fertilization in a crab. II. Cytological aspects of the cortical reaction and fertilization envelope elaboration. Tissue and Cell, 14: 273-282. Goudeau, M. (1984) Fertilization in a crab. III. Cytodifferentiation of vesicles enclosing ring- shaped elements involved in the cortical reaction. Gamete Res., 9: 409-424. Klepal, W., Barnes,H. and Barnes, M. (1979) Studies on the reproduction of cirripedes. VII. The formation and fine structure of the fertilization membrane and egg case. J. Exp. Mar. Biol. Ecol., 36: 53-78. Clark, W. H., Lynn, J. W. and Yudin, A. I. (1975) The cortical reaction in the egg of the Penaeid shrimp. Int. Symp. on Reproductive Physiology of Invertebrates, 1: 21. Adiyodi, R. G. (1985) Reproduction and its con- trol. In “The Biology of Crustacea. Vol. 9”. Ed. by D.E. Bliss and L.H. Mantel, Academic Press, New York, pp. 147-215. Demeusy, N. (1958) Recherches sur la mue de puberté du Décapode Brachyoure Carcinus maenas Linné. Arch. Zool. Exp. Gen., 95: 253-491. Bomirski, A. and Klek, E. (1974) Action of eye- stalk on the ovary in Rhithropanopeus harristi and Crangon crangon (Crustacea Decapoda). Mar. Biol., 24: 329-337. Anilkumar, G. and Adiyodi, K. G. (1985) The role of eyestalk hormones in vitellogenesis during the breeding season in the crab, Paratelphusa hydro- dromous (Herbst). Biol. Bull., 169: 689-695. Stephens, G.C. (1952) The control of cement gland development in the crayfish, Cambarus. Biol. Bull., 103: 242-258. Reaka, M. L. (1976) Lunar and tidal periodicity of molting and reproduction in stomatopod Crus- tacea: A selfish herd hypothesis. Biol. Bull., 150: 468-490. 306 307 308 309 310 311 312 313 314 315 316 317 318 319 J.-J. MEUSY AND G. G. PAYEN Watson, J. (1970) Maturity, mating, and egg-laying in the spider crab, Chionocetes opilio. J. Fish. Res. Board Can., 27: 1607-1616. Berry, P. F. (1970) Mating behavior, oviposition, and fertilization in the spiny lobster Panulirus homarus (Linnaeus). Invest. Rep. Oceanogr. Res. Inst. Durban, 24: 1-16. Engelmann, F. (1984) Reproduction in insects. In “Ecological Entomology”. Ed. by C. B. Huffaker and R.L. Rabb, J. Wiley & Sons Inc., North- Holland, pp. 113-147. Hartnoll, R. G. (1968) Morphology of the genital ducts in female crabs. J. Linn. Soc. (Zool), 47: 279-300. Cheung, T.S. (1968) Trans-molt retention of sperm in the female stone crab, Menippe mercenar- ia (Say). Crustaceana, 15: 117-120. Anilkumar, G. and Adiyodi, K. G. (1977) Sper- matheca of the freshwater crab, Paratelphusa hydrodromous (Herbst) in relation to the ovarian cycle. In “Advances in Invertebrate Reproduction, Vol. 1”. Ed. by K. G. Adiyodi and R. G. Adiyodi, Peralam-Kenoth, Kerala, pp. 269-274. Peyrot, S. and Trilles, J.-P. (1964) Recherches sur la sexualité et la glande androgéne de Caprella aequilibra Say (Amphipode, Caprellidae). Bull. Inst. Océanogr. Monaco, 63: 1-28. Mead, F. (1963) Sur l’existence d’une cavité in- cubatrice complexe chez l’Isopode terrestre Heller- ia brevicornis Ebner. C. R. Acad. Sci. Paris, 257: 775-777. Balesdent, M.-L. (1964) Recherches sur la sexuali- té et le déterminisme des caractéres sexuels d’Asel- lus aquaticus Linné (Crustacé Isopode). Thése Doctorat d’Etat, Univ. Nancy, pp. 1-231. Daguerre de Hureaux, N. (1966) Etude ex- périmentale du rdle de quelques formations endo- crines cérébrales chez Sphaeroma serratum (Crus- tacé Isopode). Thése Doctorat d’Etat, Univ. Bor- deaux, AO CNRS n°1036, pp. 1-115. Tinturier-Hamelin, E. (1962) Sur un caractére de ’Idotea balthica (Pallas) (Isopode Valvifére) con- sidéré jusqu’alors comme un caractére sexuel secondaire male. Arch. Zool. Exp. Gen., 101: 54- 58. Sollaud, E. (1923). Recherches sur l’embryogénie des Crustacés Décapodes de la sous-famille des Palaemonidae. Bull. Biol. Fr. Belg., 5: 1-234. Descouturelle,G. (1971) Différenciation des caractéres sexuels femelles chez la Crevette d’eau douce Atyaéphyra desmaresti Millet (Crustacea, Decapoda, Natantia). C. R. Soc. Biol., 169: 1412- 1416. Shen, C. J. (1935) An investigation of the post larval development of the shore crab, Carcinus maenas, with special reference to the external 320 321 322, 323 324 325 326 327 328 329 330 331 333 334 Female Reproduction in Malacostracan Crustacea secondary sexual characters. Proc. Zool. Soc. London, 1-33. Vernet-Cornubert, G. (1958) Recherches sur la sexualité du crabe Pachygrapsus marmoratus Fabr. Arch. Zool. Exp. Gén., 96: 101-276. Veillet, A. (1945) Recherches sur le parasitisme des Crabes et des Galathées par les Rhizocéphales et les Epicarides. Ann. Inst. Paris, 22: 193-341. Teissier, G. (1933) Etude de la croissance de quelques variants sexuels chez Macropodia rostra- ta. Bull. Biol. Fr. Belg., 67: 401-444. Teissier, G. (1935) Croissance des variants sexuels chez Maia squinado L. Trav. Stn. Biol. Roscoff, 13: 93-130. Churchill, E. P. (1919) Life history of the blue crab. Bull. Bur. Fish. Wash., 36: 93-134. Van Engel, W. A. (1958) The blue crab and its fishery in the Chesapeake Bay. Pt. I. Reproduc- tion, early development, growth and migration. U.S. Fish. Wild. Serv., Comml. Fish. Rev., 20: 6- 17. Chaix, J.-C. (1979) Le cycle biologique et quelques aspects de la reproduction du Crabe Oxyrhynque Acanthonyx lunulatus (Risso, 1816) (Crustacea, Decapoda, Oxyrhyncha). Téthys, 9: 17-22. Hinsch, G. W. (1968) Reproductive behavior in the spider crab, Libinia emarginata (L.). Biol. Bull., 135: 273-278. Goudeau, M. and Lachaise, F. (1983) Structure of the egg funiculus and deposition of embryonic envelope in a crab. Tissue and Cell, 15: 47-62. Mason, J.C. (1970) Egg-laying in the western North American crayfish, Pacifastacus trowbridgii (Stimpson) (Decapoda, Astacidae). Crustaceana, 19: 37-44. Thomas, W. J. and Crawley, E. (1975) The glair glands and oosetae of Austropotamobius pallipes (Lereboullet). Experientia, 31: 183-185. Johnson, B. and Talbot, P. (1987) Ultrastructural analysis of the pleopod tegumental glands in male and female lobsters, Homarus americanus. J. Crust. Biol., 7: 288-301. Aiken, D. E. and Waddy,S.L. (1982) Cement gland development, ovary maturation, and repro- ductive cycles in the American lobster Homarus americanus. J. Crust. Biol., 2: 315-327. Charniaux-Cotton, H. (1965) Hormonal control of sex differentiation in invertebrates. In “Orga- nogenesis”. Ed. by R. De Haan and H. Ursprung, Holt, Rinehart and Winston, New York, pp. 701- 740. Charniaux-Cotton, H. (1972) Recherches récentes sur la différentiation sexuelle et l’activité génitale chez divers crustacés supérieurs. In “Hormones et Différenciation Sexuelle chez Invertébrés”. Ed. by E. Wolff, Gordon and Breach, New York, pp. 335 338 339 341 342 w & Ww 345 346 347 128-178. Legrand, J.-J. and Juchault, P. (1972) Le contréle humoral de la sexualité chez les Crustacés Isopodes gonochoriques. In “Hormones et Différenciation Sexuelle chez les Invertébrés”. Ed. by E. Wolff, Gordon and Breach, New York, pp. 179-213. Charniaux-Cotton, H. and Payen G. (1985) Sexual differentiation. In “The Biology of Crustacea, Vol. 9”. Ed. by D. E. Bliss and L. H. Mantel, Academic Press, New York, pp. 217-295. Nagamine, C. and Knight, A. W. (1987) Induction of female breeding characteristics by ovarian tissue implants in androgenic gland ablated male fresh- water prawns Macrobrachium rosenbergii (de Man) (Decapoda, Palaemonidae). Int. J. Invertebr. Re- prod. Dev., 11: 225-234. Legrand, J.-J. (1955) Role endocrinien de l’ovaire dans la différenciation des oostégites chez les Crustacés Isopodes terrestres. C. R. Acad. Sci. Paris, 241: 1083-1087. Luca, V. and Patane, L. (1961) Ulteriori dati sul determinisme dei caratteri sessuali secondari negli Isopodi Oniscoidei (Porcellio laevis Latr.). Atti della Accad. Giaenia, Sc. Nat. Catania, 13: 1-19. Shimoizumi, M. (1964) Effects of ovariectomy on the oostegite formation in the Isopod Crustacean, Porcellio pruinosus. J. Gakugei, Tokushima Univ., 14: 1-8. Takewaki, K. and Nakamura, N. (1944) The effects of gonadectomy on the sex characters of Armadillidium vulgare, an isopod crustacean. J. Fac. Sci. Imp. Univ. Tokyo, 4: 368-382. Besse, G., Juchault, P. and Legrand, J.-J. (1968) Etude expérimentale du déterminisme des carac- teres sexuels externes chez des Porcellio dilatatus Brandt intersexués (Crustacés Isopodes terrestres). C. R. Soc. Biol., 162: 2196-2200. Callan, H. G. (1940) The effects of castration by parasites and X-rays on the secondary sex charac- teristics of prawns (Leander sp.). J. Exp. Biol., 17: 168-179. Lloyd, A. J. and Yonge, C. M. (1940) Correlation between egg-carrying setae and cement glands in decapod Crustacea. Nature, 146: 334. Bollenbacher, W. E. and O’Connor, J. D. (1973) Production of an ecdysone by crustacean Y-organs in vitro. Am. Soc. Zool., 13: 1274. Lisk, R. D. (1961) Estradiol-17/ in the eggs of the American lobster, Homarus americanus. Can. J. Biochem. Physiol., 39: 659-663. Dhainaut, A. and DeLeersnyder,M. (1976) Etude cytochimique et ultrastructurale de l’évolu- tion ovocytaire du crabe Eriocheir sinensis. II. Ovogenése apres ablation des pédoncules ocu- laires. Arch. Biol., 87: 283-302. Gohar, M. and Souty, C. (1984) Action temporelle 264 349 350 Ww Nn i) Ww nN & 355 356 358 359 360 361 d’ecdystéroides sur la synthése protéique ovarienne in vitro chez le Crustacé isopode terrestre Porcellio dilatatus (Brandt). Reprod. Nutr. Dév., 24: 137- 145. Meusy, J.-J. and Charniaux-Cotton, H. (1984) En- docrine control of vitellogenesis in Malacostraca Crustaceans. In “Advances in Invertebrate Repro- duction. Vol. 3”. Ed. by W. Engels, W. H. Clark, Jr., A. Fischer, P.J.W. Olive and D. F. Went, Elsevier, North Holland, pp. 231-241. Dunham, P. J. (1978) Sex pheromones in Crus- tacea. Biol. Rev., 53: 555-583. Salmon, M. (1983) Courtship, mating systems, and sexual selection in decapods. In “Studies in Adap- tion, The Behavior of Higher Crustacea”. Ed. by S. Rebach and D. Dunham, John Wiley and Sons Inc., pp. 143-169. Salmon, M. and Hyatt, G. W. (1983) Communica- tion. In “The Biology of Crustacea. Vol. 7”. Ed. by F.J. Vernberg and W.B. Vernberg, Academic Press, New York, pp. 1-40. Bauchau, A. (1986) Sex pheromones in Crustacea. In “Advances in Invertebrate Reproduction. Vol. 4”. Ed. by M. Porchet, J.-C. Andries and A. Dhainaut, Elsevier Science Publishers, North Hol- land, pp. 337-343. Dahl, E., Emmanuelson,H. and von Mecklen- burg, C. (1970) Pheromone reception in the males of the amphipod Gammarus duebeni Lilljeborg. Oikos, 31: 42-47. Ducruet, J. (1973) Comportement sexuel spéci- fique et interspécifique chez les gammares du groupe pulex (Crustacés Amphipodes). C. R. Acad. Sci. Paris, Sér. D, 276: 1037-1039. Hartnoll, R. G. and Smith, S.M. (1980) An ex- perimental study of sex discrimination and pair formation in Gammarus duebenii (Amphipoda). Crustaceana, 38: 254-264. Atema, J. and Engstrom, D. G. (1971) Sex pher- omone in the lobster, Homarus americanus. Na- ture, London, 232: 261-263. Kamiguchi, Y. (1972) Mating behavior in the freshwater prawn, Palaemon paucidens. A study of the sex pheromone and its effect on males. J. Fac. Sci., Hokkaido Univ., Ser. VI, Zool., 18: 347-355. Eales, A. J. (1974) Sex pheromone in the shore crab Carcinus maenas, and the site of its release from females. Mar. Behav. Physiol., 2: 345-355. Gleeson, R. A. (1980) Pheromone communication in the reproductive behavior of the blue crab, Callinectes sapidus. Mar. Behav. Physiol., 7: 119- 134. Teytaud, A. R. (1971) The laboratory studies of sex recognition in the blue crab, Callinectes sapidus Rathbun. Sea Grant Tech. Bull., Univ. Miami Sea Grant Prog., 15: 1-62. 362 363 364 365 366 367 368 369 370 371 373 374 375 376 377 J.-J. MEUSY AND G. G. PAYEN Salmon, M. and Atsaides, S. (1968) Visual and acoustic signalling during courtship of fiddler crabs (genus Uca). Am. Zool., 8: 623-639. Bliss, D. E. (1979) From sea to tree: Saga of land crab. Am. Zool., 8: 355-392. Hartnoll, R. G. (1969) Mating in the Brachyura. Crustaceana, 16: 161-181. Ingle, R. W. and Thomas, W. (1974) Mating and spawning of the crayfish Austropotamobius pallipes (Crustacea: Astacidae). J. Zool., London, 173: 525-538. Berry, P. F. and Hartnoll, R. G. (1970) Mating in captivity of the spider crab Pleistacantha moseleui (Miers) (Decapoda, Majidae). Crustaceana, 19: 214-215. Swartz, R. C. (1976) Agonistic and sexual be- havior of the Xanthid crab, Neopanope syi. Che- sapeake Science, 17: 24-34. Ameyaw-Akumfi, C. (1987) Mating in the lagoon crab Cardisoma armatum Herklots. J. Crust. Biol., 7: 433-436. Berrill, M. and Arsenault, M. (1982) Mating be- havior of the green shore crab Carcinus maenas. Bull. Mar. Sci., Miami, 32: 632-638. Ducruet, J. (1982) Effets de l’ecdystérone sur la formation de couples et sur la vitellogenése chez Gammarus pulex (L) et Gammarus fossarum Koch (Crustacés Amphipodes). Pol. Arch. Hydrobiol., 29: 307-317. Takayanagi, H., Yamamoto, Y. and Takeda, N. (1986) Ovary-stimulating pheromone in the fresh- water shrimp, Paratya compressa. J. Exp. Zool., 240: 201-207. Hartnoll,R.G. and Smith, S. M. formation in the edible crab Brachyura). Crustaceana, 36: 23-28. Kamiguchi, Y. (1972) A histological study of the “sternal gland” in the female freshwater prawn, Palaemon paucidens, a possible site of origin of the sex pheromone. J. Fac. Sci. Hokkaido Univ., Ser. VI, Zool., 18: 356-365. Seifert, P. (1982) Studies on the sex pheromone of the shore crab, Carcinus maenas, with special regard to ecdysone secretion. Ophelia, 21: 147- 158. Gleeson, R. (1982) Morphological and behavioral identification of the sensory structures mediating pheromone reception in the blue crab, Callinectes sapidus. Biol. Bull., 163: 162-171. Gleeson, R. A., Adams, M. A. and Smith, A. B. (1987) Hormonal modulation of pheromone- mediated behavior in a crustacean. Biol. Bull., 172: 1-9. Jassem, W., Mocquard, J.-P. and Juchault, P. (1982) Déterminisme de la reproduction saison- niére des femelles d’Armadillidium vulgare Latr. (1979) Pair (Dacapoda, 378 Female Reproduction in Malacostracan Crustacea (Crustac, Isopode, Oniscoide). IV. Contribution a la connaissance de la perception du signal photo- périodique induisant l’entrée en reproduction: mode de discrimination entre le jour et la nuit; longueurs d’onde actives. Ann. Sci. Nat. Zool., Paris, 4: 85-90. Zerbib, C. and Mustel, J.-J. (1984) Incorporation de la vitellogénine tritiée dans les ovocytes du 379 380 265 Crustacé Amphipode Orchestia gammarellus (Pal- las). Int. J. Invertebr. Reprod. Dev., 7: 63-68. Ryan, E. P. (1966) Pheromone: evidence in a decapod crustacean. Science, 151: 340-341. Gleeson, R. A., Adams, M. A. and Smith, A. B. (1984) Characterization of a sex pheromone in the blue crab, Callinectes sapidus: crustecdysone studies. J. Chem. Ecol., 10: 913-921. —> === e oni ane ¢ of me Oia fp ; - i) oh)" scabs Bae tg =) oe shag iihante tiki eer, i) Pte : : a tera, wy x 7 : * - bale Tal : 0) Ts ae ‘i se y Wes Oo e oem 4 ee al ae * = i os by Unsere ae i —e ta he , al j 7 ti 4 ZOOLOGICAL SCIENCE 5: 267-280 (1988) © 1988 Zoological Society of Japan REVIEW Control of Prolactin and Growth Hormone Secretion in Teleost Fishes RICHARD S. NISHIOKA, KEVIN M. KELLEY and Howarp A. BERN Department of Zoology and Cancer Research Laboratory, University of California, Berkeley, CA 94720, U.S.A. INTRODUCTION In teleost fish, hypothalamic fibers terminate in proximity to adenohypophysial (AH) cells of the pituitary, thereby circumventing a functional me- dian eminence and a hypophysial portal system, structures considered to be both ancient and conservative in vertebrate evolution (see [1-3]). Thus, unlike the situation in most other verte- brates, teleosts have a more or less direct neural control of pituitary function. Although claims for a median eminence-portal system have been made for some teleostean species (see [4, 5]), the existence of such a system is generally considered absent from teleosts [2, 3, 6-8]. Two general classes of hypothalamic nerve fibers innervate the pituitary of all teleosts investigated to date: type A fibers, containing “elementary neurosecretory granules” (115-170 nm granule di- ameter) and generally considered to be pep- tidergic, and type B fibers, containing “large granulated vesicles” (45-95 nm granule diameter) and generally considered to be aminergic. A possible third type of hypothalamic fiber, however, is observed in Oreochromis mossambicus (tilapia) [9]; these “type C” fibers are similar to type A fibers, but are characterized by granules of in- termediate size (90-145nm granule diameter) with limiting membranes usually separated from the granule core. Type A fibers, because they are commonly observed leading into the neurohy- Received December 8, 1987 pophysis (NH) and pars intermedia (PI) from the preoptic nuclei (PON), are believed to be con- cerned primarily with the secretion of neurohy- pophysial octapeptides and the control of the PI, whereas type B fibers, because they are commonly observed emanating from the nucleus lateralis tuberis (NLT) and ending in synaptic contact with the secretory cells of the pars distalis (PD), are believed to be more closely involved in the control of the PD [2, 6]. Type C fibers of tilapia are observed regularly in the hypothalamus and neurohypophysis, but no physiological function can be attributed to these fibers on the basis of morphological criteria alone [9]. More recently, however, physiological evidence of important pep- tidergic control of the PD has modified these early views; thus, type A (and possibly in tilapia, type C) control of PD function is considered in this review. The focus of this review is the control of prolactin (PRL) and growth hormone (GH) secre- tion in teleosts, with primary emphasis on the tilapia; therefore, the hypothalamus and other parts of the brain will not be extensively discussed except as they relate directly to the material presented herein (for earlier reviews and articles of neural control of teleost adenohypophysial func- tion, see [2—4, 6, 10-13]). Briefly, many different hypothalamic nuclei exist in teleosts, a number of which may be important in the control of AH function. Two nuclei previously mentioned have been particularly well studied: the PON and the NLT. The PON occurs in tilapia as layers of neurons alternating with layers of axonal processes 268 R. S. NisH1oka, K. M. KELLEY AND H. A. BERN lying above the optic chiasma; in general, the cells of the PON are type A and can be divided into two groups, a ventral pars parvocellularis and a dorsal pars magnocellularis. Axonal processes from the PON converge mid-ventrally, progress along the NLT region into the pituitary, and become a part of the NH [2, 11, 13, 14]. The NLT, on the other hand, is usually present in teleosts as a pair of structures lying adjacent to the anterior and lateral regions of the base of the infundibular stalk from which, most commonly, type B fibers run in a main tract alongside type A (PON) fibers into the NH [1, 2, 10, 11, 14]. ANATOMICAL ASPECTS The degree of directness of hypothalamic in- nervation of pituitary cells varies in the different regions of the pituitary, as well as among teleosts. In the case of the tilapia rostral pars distalis (RPD), hypothalamic fibers do not leave the NH but instead terminate on an adjacent basement membrane (see [15, 16]). Similar observations have been made on Fundulus heteroclitus [17] and on Carassius auratus [18, 19]. As a consequence of this anatomical arrangement, chemical informa- tion from the brain must pass through the base- ment membrane, a layer of stellate cell processes, and also the adrenocorticotropic hormone (ACTH) cells surrounding the neurohypophysial processes [15] (Fig. 1). Because of their direct contact with PRL cells, stellate cells and their processes may be involved directly in information transfer; perhaps stellate cells modulate movement of neurohormonal factors from the neurohypophy- sial nerve endings to the PRL cells located several cell layers away. The stellate cells may be equivalent to the tanycytes that link the third ventricle with the NH [20]. The role of the intervening layer of ACTH cells in information transfer, if any, is unknown, but a paracrine role for one or more proopiomelanocorticotropin pro- ducts is possible (Fig. 1). The anatomical basis for control of GH secre- Preoptic Nucleus (PON) ACTH and/or STELLATE CELLS PRL CELLS UROPHYSIS Fic. 1. OSMOTIC PRESSURE Lateral Tuberal Nucleus (NLT) GLYCOPROTEIN- SECRETING and/or STELLATE CELLS ? GH CELLS SEX | CORTICAL STEROIDS Pathways for the control of PRL and GH secretion in tilapia. Thick lines (arrows) denote important pathways of control and thin lines indicate pathways of less impor- tance. pathways. Question marks denote possible “paracrine” influences and less certain UI is a CRF-like peptide with possible influence on ACTH cells. UII shares an important tripeptide sequence with somatostatin and inhibits PRL but not GH secretion. Prolactin and Growth Hormone in Teleosts 269 tion in tilapia parallels that for PRL. Unlike the indirect innervation seen in PRL cells, however, ultrastructural observations of the proximal pars distalis (PPD) indicate direct aminergic and pep- tidergic innervation of the GH cells by hypothala- mic fibers [9] (Fig. 1). Fiber types A and B (and C?) are in close proximity to GH cells and often show synaptoid contacts; in addition, some appear to be in direct apposition to the cells of the PPD and PI [21]. Patterns of innervation similar to tilapia’s, but with expected species differences, have been de- scribed for the goldfish (C. auratus) [18, 19], the killifish (F. heteroclitus) [17], the medaka (Oryzias latipes) [22], the molly (Poecilia latipinna) (20, 23], the longjawed mudsucker (Gillichthys mirabilis) [15], and the mullet (Mugil cephalus) [24]. In less advanced teleosts, such as eels and salmonids, there are noticeable differences in the innervation of the AH. In the European eel (Anguilla anguilla), the AH and NH are separated by a basement membrane as in other teleosts, but the penetration of the NH into the AH is not so extensive as is observed in the tilapia, the molly, and other more advanced teleosts, especially in the region of the PPD [2]. Furthermore, control of the PD is reported to be at least partially exerted via hypothalamic nerve endings on hypothalamic arteries in the rostral NH leading to the PD in the Atlantic salmon, Salmo salar [25] and brook trout, Salvelinus fontinalis [4]. Thus, although teleosts have in common a similar scheme for communication between the brain and the pituitary secretory cells (via direct innervation), there is much variation in the scheme among members of the vast taxonomic grouping termed “Teleostei”. HYPOTHALAMIC CONTROL Early studies of pituitary transplantations that suggested control of PRL secretion in teleosts by a hypothalamic PRL release-inhibiting factor (PIF) (see [5, 10, 21, 26-29]), and other investigations that demonstrated a direct hypothalamic innerva- tion of the teleost PD [2, 9, 15, 24, 30], prompted interest in the nature of hypophysiotropic factors operating in the hypothalamo-hypophysial area in teleosts [31]. Indeed, many factors have now been implicated in the control of teleost PD function. Aminergic factors The role of catecholamines, especially dopamine (DA), in the control of pituitary function in vertebrate tetrapods is well documented; in fact, DA is considered by many to be the vertebrate PIF (cf. [32]). In teleosts, however, the nature of the innervation of the pituitary by aminergic fibers has been unclear. Initial studies described type B fibers either directly innervating the PRL cells in G. mirabilis [15, 33] and in P. latipinna [20] or ending on the adjacent membrane of the NH in tilapia [15]. However, when the formaldehyde- induced fluorescence technique for monoamines (a definitive method for the identification of cate- cholaminergic material [34]) is applied to the pituitary of tilapia [9] and G. mirabilis [35], none of the type B fibers initially observed in the NH fluoresces positively. Thus, in tilapia, G. mirabilis, and possibly other species, catecholamines appear to be absent from the pituitary. In C. auratus, however, using *H-DA, type B fibers have been demonstrated going to and in close apposition to the basement membrane [36], a result supported by studies using a specific antibody against DA [37]. DA antiserum has also been used on the cyprinodont, Cynolebias whitei, to demonstrate DA-ergic fibers in the NH [38]. Furthermore, Halpern-Sebold ef al. [39] have detected tyrosine hydroxylase-immunoreactivity in the hypothala- mus and pituitary of Xiphophorus maculatus, particularly in the PRL cells. Serotonin, on the other hand, although not observed in fibers in the pituitary of P. latipinna using autoradiographic techniques [40], has not been demonstrated in the pituitary of C. whitei using serotonin antiserum [38]. Studies of the actions of aminergic factors on PRL secretion using physiological and pharmaco- logical agents in vitro have demonstrated effects of DA, serotonin and adrenalins in various teleost species. An inhibitory effect of DA on PRL release is reported in the tilapia [41, 42], P. latipinna [23, 43], G. mirabilis [41], C. auratus [44], A. anguilla [45, 46], and Salmo gairdneri [47, 48]. In addition, endogenous levels of catechol- 270 R. S. NIsHIOoKA, K. M. KELLEY AND H. A. BERN amine presumably act to inhibit PRL secretion from the pituitary gland autotransplanted into the anterior chamber of the eye of P. latipinna [49]. Pharmacological agents that affect dopaminergic systems, as well as DA precursors, have been tested in various teleosts and have the expected actions based on mammalian studies; these agents (ergocryptine, 6-hydroxydopamine, L-dopa, reser- pine, and various receptor antagonists) and their actions in teleosts are listed in Table 1. Serotonin, on the other hand, is stimulatory to PRL secretion in vitro in S. gairdneri [48] and in A. japonica [50]; in addition, injection of 5-hydroxytryptophan, a serotonin precursor, appears to activate PRL cells in the same species [48, 51]. Furthermore, the intraperitoneal injection of pargyline, a mono- amine oxidase inhibitor, results in elevated brain serotonin concomitant with increased pituitary TABLE 1. PRL levels in C. auratus [52]. In contrast, parachlorophenylalanine, a tryptophan hydroxy- lase inhibitor, reduces hypothalamic serotonin content in A. anguilla [53] and in S. gairdneri [54]. The few studies on the role of adrenalins in the control of teleost PRL secretion are apparently contradictory. In C. auratus, epinephrine and norepinephrine increase adenylate cyclase activity in PD homogenates [55]. In P. latipinna, however, phenylephrine (a-adrenergic agonist) and _ iso- proterenol (f-adrenergic agonist) inhibit PRL se- cretion in vitro; furthermore, the adrenergic block- ing agents phentolamine and propranolol have no direct effect on PRL secretion, but oppose DA inhibition of PRL secretion in vitro [43]. Aminergic control of GH secretion in teleosts has received relatively little attention. In contrast to its inhibitory effect on PRL secretion, DA Factors affecting the dopaminergic control of prolactin secretion Species Dopamine L-Dopa Ergo- cryptine Receptor 6-HODA antagonist Reserpine Anguilla anguilla 45* — 140 141 — 142 Gillichthys mirabilis 41 — 143 Heteropneustes fossilis 144 Mugil platanus 145 Poecilia latipinna 43 = 146 — 147 Salmo gairdneri 47 = 48 — — Oreochromis mossambicus 41 _ 42 — 148 Xiphophorus helleri 147 * Numbers under species names are references. Prolactin and Growth Hormone in Teleosts 271 appears to be stimulatory to GH secretion in C. auratus; by using various agents administered intraperitoneally and intraventricularly, Chang et al. [56] have shown that DA may act centrally to stimulate GH secretion. Wigham et al. [43], however, report an in vitro inhibition of secretion of the putative GH in P. latipinna by DA, although this inhibition is not opposed by the specific DA antagonist 3, 4-dimethylphenylethylamine. Nor- epinephrine, on the other hand, may act directly to inhibit GH secretion in C. auratus [56]. In P. latipinna, however, the adrenergic blocker pro- pranolol inhibits GH secretion, whereas phentol- amine and other adrenergic pharmacological agents have no in vitro effect [43]. Peptidergic agents The demonstration of a lack of aminergic fluorescence in the hypophysial area of tilapia and G. mirabilis raises questions about the nature of the PIF (and/or other factors) in these and other teleosts (see [57]). More recently, a role for somatostatin (SRIF) as an inhibitor of PRL secre- tion in teleosts has been investigated. By im- munocytochemical techniques, SRIF-like material is observed in the brain of the catfish (Hetero- pneustes fossilis) [58] and S. gairdneri [59], and can be further localized to the hypothalamus and NH near the PD in F. heteroclitus, G. mirabilis [60], tilapia [60, 61], P. latipinna (62, 63], and in several other freshwater and seawater teleosts [64-67]. In addition, Batten [63] suggests that the SRIF- immunoreactive (IR) fibers in the pituitary of P. latipinna appear to correspond to a particular class of type A fibers: “type A2” identified by electron microscopy (EM) [20]. A similar situation exists in tilapia, where the SRIF-IR fibers in the NH appear to correspond well with the distribution of type C fibers as detected by EM. In addition, immunocytochemical studies of the hypothalamo-hypophysial system following acute changes in environmental salinity offer further evidence for the role of SRIF in control of PRL secretion in tilapia [68]. Preliminary observations indicate that shortterm (up to 3 hr) transfer from SW to FW results in increased SRIF-IR in the cell bodies of the PON and decreased SRIF-IR in the neurohypophysial processes penetrating the RPD. These observations suggest that transport of SRIF from the PON is inhibited in FW. The reciprocal transfer (FW to SW), on the other hand, is seen to deplete SRIF-IR in the cell bodies of the PON and increase SRIF-IR in the neurohypophysial fibers; furthermore, in SW-transferred tilapia, SRIF-IR is more prominent in the NH, and fibers containing SRIF-IR appear to penetrate the RPD more deeply than in FW-transferred tilapia. SRIF may be the significant PIF in tilapia based on its potent activity in vitro. In tilapia, SRIF is inhibitory to PRL secretion in vitro [42, 69, 70], and this inhibitory effect is at least partially independent of any effects of SRIF on PRL synthesis [71]. In P. latipinna, SRIF inhibits total and newly-synthesized PRL secretion in vitro [72] and reduces synthetic activity of the PRL cells as detected by EM [73]. Coupled with observations on salmonids and eels that indicate less penetration of SRIF-IR hypothalamic fibers into the pituitary [66] as compared with more advanced teleosts, it is interesting that SRIF does not inhibit PRL secretion in vitro in S. gairdneri [48]. GH secretion is similarly inhibited by SRIF in vitro in Anguilla japonica [50]. In tilapia, SRIF inhibits GH secretion in vitro [70, 74], even in the presence of cortisol which stimulates GH secretion [75]. Helms et al. [76] have shown that SRIF is a more effective inhibitor of GH secretion in smaller (ca. 60g) than in larger (ca. 120-180 g) tilapia. Using a perifusion system, Marchant et al. [77] have shown that SRIF (and SRIF-28) inhibit GH secretion in C. auratus; in the same study, howev- er, it was found that catfish pancreatic SRIF—22 has no effect on GH secretion. Furthermore, injection of SRIF lowers plasma GH levels in C. auratus [78] as well as in Oncorhynchus kisutch [79]. A role for thyrotropin-releasing hormone (TRH) in the hypothalamic control of teleost PRL and GH secretion is supported by recent evidence. In S. salar sebago, TRH-IR is observed in the brain and pituitary [80]. TRH stimulates PRL secretion in vitro in A. japonica [SO]. In P. latipinna, TRH stimulates PRL secretion in vitro, even in a hyperosmotic medium [72], and evidence of increased synthetic activity is observed by EM 212: R.S. NISHIOKA, K. M. KELLEY AND H. A. BERN [73]. Similarly, Prunet and Gonnet [81] have shown that TRH stimulates PRL secretion in vitro in a dose-dependent manner in S. gairdneri. Barry and Grau [82] have observed a stimulation of PRL secretion in vitro by TRH from pituitaries pre- treated with 17/-estradiol. James and Wigham [48] have not observed an in vitro effect of TRH on PRL secretion in S. gairdneri; however, P. Prunet and F. Gonnet (personal communication) have shown that the addition of a protease inhibitor (e.g., Bacitracin) to the incubation medium pre- vents breakdown of TRH, allowing TRH to stimulate PRL secretion in a dose-related fashion in the same species. The in vitro effects of vasoactive intestinal polypeptide (VIP) and peptide histidine isoleucine (PHI) have recently been examined in tilapia [83]. VIP and PHI are potent stimulators of PRL secretion in tetrapods in vivo and in vitro (cf. [84]), including in an amphibian [85]. In tilapia, howev- er, the first non-tetrapod species in which VIP and PHI have been tested, both VIP and PHI inhibit PRL secretion in vitro; these two peptides have no effect, however, on GH secretion. Preliminary immunocytochemical observations suggest a mod- erate amount of VIP-IR, but no definitive PHI-IR, in the hypothalamo-hypophysial area of tilapia (R. S. Nishioka et al., unpublished; see [83]). Such a discordant pattern of secretion between fish and mammal is not readily explainable, and studies are being conducted presently to gain some under- standing of this phenomenon. In tilapia, other peptide factors have been investigated for their in vitro effect on PRL and/or GH secretion. Urotensin H (UII), a dodec- apeptide showing sequence similarity to SRIF (sharing an important tripeptide, Phe-Trp-Lys), inhibits PRL secretion, but does not have any significant effect on GH secretion [69, 70]. Re- cently, Kewish et al. (personal communication) have determined that growth hormone-releasing hormone (GHRH) stimulates GH release. Peptidergic factors of potential importance in the control of teleost PRL and GH secretion are suggested by some immunocytochemical and other studies. Studies on gonadotropin-releasing hor- mone (GnRH; LHRH) in teleosts, for example, suggest a potentially important role for GnRH. In S. gairdneri, immunoreactive GnRH is observed within telencephalic perikarya and in fibers passing to the pituitary stalk [86], whereas in P. /atipinna, GnRH-IR fibers are observed in the NH leading toward and contacting the gonadotropes in the PPD [63] where GH cells are nearby. Similarly, Kah et al. [87] have detected GnRH-IR fibers going to the PPD of C. auratus. These anatomical studies of GnRH have prompted our interest in the potential effect of GnRH on tilapia PRL and GH secretion. Furthermore, inasmuch as gonadotro- pin-associated peptide (GAP), a segment of the GnRH precursor, belongs to the same family of peptides as VIP and PHI [84], we have begun investigation of the effects of this peptide (and some of its cleavage products) using our in vitro system (J. Planas et al., unpublished). Corticotro- pin-releasing factor (CRF), as well, is observed in areas suggestive of a possible relation to PRL or GH secretion. In C. auratus and Cyprinus carpio, CRF-IR perikarya are observed in the PON and PVN, with fibers from them leading to the pitui- tary and ending anterior to the ACTH cells of the RPD [67]; the possibility of the intermingled PRL cells of the RPD being affected by such an innervation is suggested. Similar distributions of CRF-IR have been observed in S. gairdneri [88, 89] and P. latipinna [63]. Combined with the fact that some teleosts are known to secrete PRL while under stress (P. Prunet and M. Avella, personal communication), these CRF immunocytochemical data make further in vitro work desirable. Furth- ermore, urotensin I, a CRF-like peptide from the fish caudal neurosecretory system, and possibly also associated with the family of substances that includes VIP and PHI [84], is another conceivably important peptide factor, as there are claims for its presence in the brain in Catostomus commersoni [90]. Finally, in P. latipinna, fibers immunoreac- tive for the two neurohypophysial octapeptides, arginine vasotocin and isotocin, originate from separate preoptic perikarya and end near all AH cell types (except PRL cells) [63]; the possible interaction between neurohypophysial octa- peptides which stimulate ACTH secretion in vitro in C. auratus [91], and the secretion of PRL and GH, is unknown in teleosts. Thus, several peptide factors of potential importance in the control of Prolactin and Growth Hormone in Teleosts 273 teleost PRL and GH secretion merit further study. EXTRAHYPOTHALAMIC CONTROL In addition to hypothalamic factors, various extrahypothalamic factors also appear to control PRL and GH secretion in teleosts (see [3, 92-94]). These factors, which include medium osmotic pressure, sex steroids, corticosteroids and thyroid hormones, also appear to modulate the control of PRL and GH secretion. Osmotic factors In a pioneering study, Pickford and Phillips [95] demonstrated the important role of PRL in FW osmoregulation; subsequent investigations on numerous species of teleosts have substantiated the importance of PRL in FW adaptation (see [28, 92, 96-98]). There is general consensus that low osmotic pressure, characteristic of FW, is an important factor stimulating PRL secretion in euryhaline teleosts. Osmotic pressure and PRL secretion in vivo and in vitro are inversely related (Table 2). The salmonids present an exceptional group regarding the control of PRL secretion by osmotic pressure. In some salmonid species, it is difficult to demonstrate a direct relationship between osmotic pressure and PRL cell activity in vitro, but, in other salmonid species, the typical inverse rela- tionship exists between osmotic pressure and PRL cell activity (Table 3). It has been suggested that this disparity may be due to the differences of the stage of development of the individuals used in a particular study [99]. Upon scrutiny of the studies listed in Table 3, there appears to be a tendency for decreased ability to regulate PRL secretion with advances in development; PRL cells of alevins, parr and smolts are responsive, whereas those of mature fish of some species are less responsive. In addition, there appear to be differences in responses in vivo and in vitro . For example, PRL secretion in vitro from the PD of adult O. keta is unresponsive to variations in medium osmotic pressure [100], whereas plasma levels of PRL respond to changes in osmotic pressure in vivo [101]. These studies suggest that some factor in intact O. keta (possibly hypothalamic) is absent in vitro. On the other hand, Cook and van Over- TABLE 2. Fishes showing inverse relationship between PRL cell activity and medium osmotic pressure and/or ion concentration Species Reference Anguilla anguilla A. japonica Aphanius dispar Carassius auratus Cichlasoma _ biocellatum Fundulus heteroclitus Gasterosteus aculeatus Gillichthys mirabilis Lebistes sp. Mugil sp. Oreochromis mossambicus Oryzias latipes Platichthys stellatus Poecilia sp. Tilapia sp. Xiphophorus sp. 105, 107, 108, 109, 149, 150 50, 111 151, 152 18, 111, 153 154 95, 155, 156 157, 158 21, 41, 159 160 30, 161 49, 97, 119, 126, 148, 162, 163, 164 111, 165, 166 159 20, 49, 107, 150, 168, 169, 170 118 171, 172, 173, 174, 175, 176, 177, 178, 179 274 R. S. NisHioka, K. M. KELLEY AND H. A. BERN TABLE 3. Prolactin cell activity at low osmotic pressure in salmonids based on various criteria Species Age Response Criterion Reference Oncorhynchus keta adult + RIA (plasma) 101 (chum salmon) adult — incubation 100 O. kisutch alevin + cytology 180 (coho salmon) fry + cytology 181 smolt + cytology 99 “yearling” + RIA (plasma & RPD) 182, 183 parr/smolt + incubation 182 Oncorhynchus nerka smolt cytology 114 (sockeye salmon) smolt “te RIA (plasma) 103 adult cytology 102 Salmo gairdneri adult — incubation 108 (rainbow trout) “yearling” + RIA (plasma) 184 “yearling” —(?) incubation 185 beeke [102] and McKeown and Leatherland [103] found no in vivo responsiveness in adult O. nerka subjected to different osmotic environments. Thus, although it may be generally true that euryhaline teleosts can respond to osmotic pres- sure changes by altering PRL secretion, there are some exceptions to this rule among salmonids. Further study of this group is needed, with special attention to various life stages. The catadromous A. anguilla [104-110] and A. japonica [100, 111] respond to lower osmotic pressure in vivo and in vitro by increasing PRL secretion at any stage of development. The anadromous Gasterosteus aculeatus has reduced PRL cell activity in vivo when exposed to FW containing Ca** and Mg** equal to that of SW [112]. Seasonal differences in PRL cell secretory cycle have been reported by Lam and Leatherland [113]. GH secretion, on the other hand, has received less attention regarding the influence of osmotic pressure. In two salmonid species, O. nerka [114] and O. keta [115], and in P. latipinna [116], there are no detectable changes in GH secretion in vitro in response to osmotic pressure changes of the medium. In S. gairdneri, however, large shortterm increases in the sodium content of the ambient medium inhibit GH secretion [108]. Similarly, in A. anguilla, high sodium medium inhibits and low sodium medium stimulates GH secretion in vitro [108]. These results are supported by cytological studies of the GH cells in A. anguilla [104, 110, 117] and A. japonica [100, 111]. Furthermore, in two other tilapia species, Tilapia grahami and T. alcalica, the GH cells appear more active in fish acclimated to FW than in fish from African “soda” lakes [118]. However, Zambrano et al. [119] reported no ultrastructural changes in the GH cells of 20-30 g tilapia after transfer from SW to FW. In contrast, Helms et al. [76] have observed increased GH secretion in vitro in response to increased osmotic pressure in tilapia weighing ca. 60g, but not in larger fish (ca. 120 g). The fact that GH has been shown to promote hypoosmoregulatory abil- ity in salmonids (see [120] for references) suggests a possible role for GH in SW osmoregulation in these fish. At present, no generalization on the control of GH secretion by osmotic pressure is possible. Hormones Extrahypothalamic factors other than osmotic pressure have also been implicated in the control of PRL and GH secretion. Prolactin itself, by injection or as a result of uninhibited release from transplanted pituitaries, may have an inhibitory effect on in situ PRL cells (see [28]). Cortisol, which is believed to be a SW-adapting hormone in some teleosts, inhibits PRL secretion in vitro [49] and stimulates GH secretion in vitro [75, 76] in tilapia (Fig. 1). In contrast, cortisol is without effect on in vitro PRL secretion in S. gairdneri [48]. Prolactin and Growth Hormone in Teleosts 275 D, L-thyroxine inhibits GH release in vitro in A. anguilla [108], although triiodothyronine does not have any effects on in vitro GH secretion in tilapia ([75]; B.Kewish ef al., personal com- munication). 17/-Estradiol stimulates PRL syn- thesis [49] and promotes stimulation of secretion of PRL by TRH in tilapia [82]. Estradiol also stimulates PRL secretion in A. japonica [100]. In C. auratus, treatment of females in vivo with synthetic estrogen (ethinylestradiol) [111] or with 17f-estradiol [121] causes an increase in GH cell activity as detected by EM. Similarly, Young and Ball [122] found GH cells in P. latipinna strongly activated by 17/-estradiol treatment. B. Kewish et al. (personal communication) have recently tested methyltestosterone in male tilapia, and found no effect of this steroid in vitro. y-Amino-n-butyric acid has also been tested in tilapia by Wigham et al. [49], and no effect on PRL secretion in vitro was observed. Intracellular mediators Although increased PRL secretion was once ascribed to a direct effect of low environmental Cat * rather than to low osmotic pressure and/or Na? concentration in G. aculeatus by Wendelaar Bonga [112] and in tilapia by Wendelaar Bonga and van der Meij [123, 124], these authors subse- quently proposed that their earlier conclusions of stimulation of PRL secretion by low cation levels in vivo may have resulted from an indirect effect of Cat+* concentration on gill permeability [125]. Furthermore, other workers have observed that PRL secretion in vitro is independent of physio- logical Cat * concentration in tilapia [69, 97, 126], in P. latipinna [116], in O. kisutch [127] and in S. gairdneri (L. R. Johnston and T. Wigham, person- al communication), provided a minimal level is present. MacDonald and McKeown [127] have found an optimal concentration of Ca‘? for promoting PRL secretion in vitro in O. kisutch, and this concentration is roughly equal to the physiological levels of Cat * found in the plasma of O. nerka [128]. Evidence for a role of Ca‘ * as an intracellular mediator of teleost PRL secretion is provided by various in vitro studies. For example, Taraskevich and Douglas [129] report that spontaneous secre- ++ tion of PRL from the RPD of the teleost Alosa pseudoharengus is associated with action potentials partly mediated by extracellular Ca**. Similarly, exposure of tilapia PRL cells to a depolarizing concentration of K* elevates PRL secretion, presumably through the opening of Ca‘ * chan- nels [130]. Further work on tilapia is in accord with these data: exposure of PRL cells to the Ca** ionophore A23187 elicits increased PRL secretion, even in the presence of SRIF [69], and exposure of PRL cells to D600 (an organic Ca‘ * channel blocker), blocks K*t-induced PRL eleva- tion [130]. Interestingly, D600 does not have much effect on hypoosmotic medium-induced PRL secretion, suggesting that the effects of osmotic pressure on PRL release may be mediated by mechanisms different from those operating during chronic depolarization [130]. The mode of action of Ca in intracellular mechanisms is unclear, although a few studies suggest some particular roles for Ca**. An action of Cat* distal to its influx through a voltage- regulated channel may occur, as the use of Co’ *, a competitive inhibitor of Ca** in various cal- cium-mediated processes [131], suppresses PRL secretion induced by hypoosmotic medium in tilapia [130], as well as by K*-depolarizing medi- um (N.H. Richman and E. G. Grau, personal communication). On the other hand, in vitro exposure of the pituitary to chlorpromazine, a drug that may act on a Ca‘? gate involving phosphatidyl inositol turnover [132-134] and that stimulates PRL secretion in mammals [135], simul- taneously stimulates *H-PRL secretion and de- creases “Cat * accumulation in the coho salmon pituitary [136]. Post-receptor mechanisms utilizing the second messenger cAMP have been studied in teleost PRL cells by various groups. In tilapia, PRL secretion in vitro is stimulated by treatment with dibutyryl cyclic AMP (db-cAMP) alone, with 3-isobutyl-1-methylxanthine (IBMX, a phospho- diesterase inhibitor) alone, or with a combination of the two [69]. In S. gairdneri, db-cAMP stimulates both synthesis and release of PRL in vitro [47] and, in O. kisutch, db-cAMP stimulates synthesis, but not secretion, of PRL in vitro [136]. Similarly, L. M. H. Helms et al. (personal com- ++ 276 R. S. NisHioka, K. M. KELLEY AND H. A. BERN munication) have stimulated PRL release with forskolin (adenylate cyclase stimulator) in tilapia, and L.R.Johnston and T.Wigham (personal communication) have demonstrated in S. gairdneri a db-cAMP stimulation of PRL synthesis, but not release, as well as a forskolin stimulation of PRL secretion and cAMP production. Recently, in preliminary studies on the striped bass (Morone saxatilis) in our laboratory, 8-bromo-cAMP, IBMX, and forskolin have proven to be equally stimulatory on PRL secretion in vitro (R.S. Nishioka et al., unpublished). A relationship between Ca* * and the adenylate cyclase-cAMP system, thought to possibly operate via Ca‘ */calmodulin-induced cAMP formation [137-139], has not been studied directly in tele- osts. However, MacDonald and McKeown [136] report a net “Ca** accumulation in O. kisutch RPD tissue treated with db-cAMP, and Grau et al. [69] can reverse SRIF inhibition of PRL secretion by treatment with A23187, a calcium ionophore; thus, these studies suggest a relationship between Ca** and cAMP, possibly via calmodulin, merit- ing further investigation. PERSPECTIVES As stated earlier, neuronal processes from the hypothalamus do not directly contact the PRL cells in tilapia. Generally, a layer of ACTH and stellate cell processes is interposed. It is possible that both these cell types play a “paracrine” role in stimulus transmission (Fig. 1). Since stellate cells with slender processes are closely apposed to PRL cells throughout the RPD of freshwater tilapia, their paracrine involvement could be substantiai. In- deed, stellate cells with prominent processes stand out among the condensed PRL cells (under maximal inhibition) in seawater-adapted tilapia. Stellate cells and gonadotropic/thyrotropic cells could play a similar role in regard to GH secretion (Fig. 1). The diversity of factors controlling PRL and GH secretion and release in tilapia and other teleosts is apparent in this review. It is possible that some of these molecules may have coincidental structural similarities in minor amino acid sequences and/or spatial configuration that may “fit” a particular receptor. If this be true, then a substance foreign to the organism may cause a response in a fashion indistinguishable from that of a bonafide agent. Currently, many substances have been found to have inhibitory or stimulatory activity, but it is unlikely that all of these compounds originate from the hypothalamus of a teleost. On the other hand, redundancy of control may be built into these systems to allow separate maintenance of osmoregulatory, reproductive, and other physio- logical pathways. The continued utilization of isolated RPD and PPD of teleosts in general for in vitro studies offers many advantages. PRL cells in the RPD and GH cells in the PPD are segregated into nearly homogeneous masses which are easy to dissect and convenient for use in incubation (or perifusion). These cell masses can be dissociated to provide uniform cell populations for a variety of studies. In addition, the secretory activity of PRL cells and some GH cells can be manipulated simply by altering the tonicity of the medium, thus facilitat- ing the analysis of inhibitory and stimulatory control. ACKNOWLEDGMENTS We thank Professor Tetsuya Hirano and Dr. E. Gordon Grau for their helpful critique of this manu- script. Ms. L. M. H. Helms and Drs. P. Prunet, N. H. Richman and T. Wigham generously provided unpub- lished information for inclusion in this review. Support by the National Science Foundation (DCB 84-05249), by NOAA, National Sea Grant Program, Department of Commerce, under grant number NA80AA-D-00120, through the California Sea Grant College Program, and by the California State Resources Agency, project number R/F-101, has been essential to the research from our laboratory reported in this survey. The U.S. Government is authorized to reproduce and distribute for governmental purposes. REFERENCES 1 Knowles, F.G. W. and Vollrath, L. (1966) Phil. Trans. R. Soc. London, B, 250: 329-342. 2 Holmes, R. L. and Ball, J. N. (1974) The Pituitary Gland: A Comparative Account. Cambridge Univ. Press, London, pp. 1-397. 3 Ball, J.N. (1981) Gen. Comp. Endocrinol., 44: 135-170. 10 11 12 13 14 15 16 17 24 Prolactin and Growth Hormone in Teleosts Henderson, N. E. (1969) nol., 12: 148-153. Peter, R. E. (1973) Am. Zool., 13: 743-755. Zambrano, D. (1972) Gen. Comp. Endocrinol., Suppl. 3: 21-31. Zambrano, D., Nishioka, R.S. and Bern, H. A. (1972) In “Brain-Endocrine Interaction. Median Eminence: Structure and Function”. Ed. by K. M. Knigge, D. E. Scott and A. Weindl, Karger, Basel, pp. 50-66. Schreibman, M. P., Leatherland,J.F. and Mc- Keown, B. A. (1973) Am. Zool., 13: 719-742. Bern, H. A., Nishioka, R.S. and Nagahama, Y. (1974) In “Recherches Biologiques Contempor- aines”. Ed. by L.Arvy, Imprimerie Vagner, Nancy, pp. 170-194. Ball, J. N. and Baker, B. I. (1969) In “Fish Phys- iology”, Vol. 2. Ed. by W.S. Hoar and D. J. Randall, Academic Press, New York, pp. 1-110. Perks, A. M. (1969) In “Fish Physiology”, Vol. 2. Ed. by W.S. Hoar and D. J. Randall, Academic Press, New York and London, pp. 111-205. Peter, R. E. (1970) Gen. Comp. Endocrinol., 14: 334-356. Dodd, J. M., Follett, B. K. and Sharp, P. J. (1971) In “Advances in Comparative Physiology and Biochemistry”. Ed. by O. Lowenstein, Academic Press, New York, Vol. 4, pp. 113-223. Haider, S. and Sathyanesan, A. G. (1972) Anat., 81: 202-208. Bern, H. A., Zambrano, D. and Nishioka, R. S. (1971) In “Memoirs of the Society for Endocrinol- ogy”, No. 19. Ed. by H. Heller and K. Lederis, Cambridge Univ. Press, London and New York, pp. 817-822. Sage, M. and Bern, H. A. (1971) Int. Rev. Cytol., 31: 339-376. Betchaku, T. and Douglas, W. W. (1980) Anat. Rec., 198: 595-609. Leatherland, J. F. (1972) Can. J. Zool., 50: 835- 844. Kaul, S. and Vollrath, L. (1974) Cell Tissue Res., 154: 231-249. Batten, T. F. C. and Ball, J. N. (1977) Cell Tissue Res., 185: 409-433. Nagahama, Y., Nishioka, R.S. and Bern, H. A. (1974) Gen. Comp. Endocrinol., 22: 21-34. Aoki, K. and Umeura, H. (1970) Japon., 17: 45-56. Batten, T. F. C. and Ball, J. N. (1984) In “Current Trends in Comparative Endocrinology”, Vol. 1. Ed. by B. Lofts and W. N. Holmes, Hong Kong Univ. Press, Hong Kong, pp. 69-72. Abraham, M. (1974) Gen. Gomp. Endocrinol., 24: 121-132. Gen. Comp. Endocri- Acta Endocrinol. 25 26 39 40 41 47 48 PHA) Fridberg, G. and Ekengren, B. (1977) Can. J. Zool., 55: 1284-1296. Ball, J. N. (1969) In “Fish Physiology”, Vol. 2. Ed. by W. S. Hoar and D. J. Randall, Academic Press, New York, pp. 207-240. Ball, J. N., Baker, B. I., Olivereau, M. and Peter, R. E. (1972) Gen. Comp. Endocrinol., Suppl. 3: 11-21. Ensor, D. M. (1978) Comparative Endocrinology of Prolactin. Chapman & Hall, London. Ball, J. N., Batten, T. F. C. and Young, G. (1980) In “Hormones, Adaptation and Evolution”. Ed. by S. Ishii, T. Hirano and M. Wada, Japan Sci. Soc. Press, Tokyo/Springer-Verlag, Berlin, pp. 45-56. Abraham, M. (1971) Gen. Comp. Endocrinol., 17: 334-350. Crim, J. W., Dickhoff, W. W. and Gorbman, A. (1978) Am. Zool., 18: 411-424. Ben-Jonathan, N. (1985) Endocrine Rev., 6: 564— 589. Zambrano, D. (1970) 516. Falck, B. (1962) Acta Physiol. Scand., 56, Suppl. 197: 1-25. Swanson, D., Nishioka, R.S. and Bern, H. A. (1975) Acta Zool. Stockholm, 56: 225-237. Kah, O., Dubourg, P., Chambolle, P. and Calas, A. (1984) Cell Tissue Res., 238: 621-626. Kah, O., Dubourg, P., Onteniente, B., Geffard, M. and Calas, A. (1986) Cell Tissue Res., 244: 577-582. Ruiter, J. M., Steinbusch, H. W. M. and Joosten, H. W. J. (1984) Neurosci. Letters, Suppl. 18: S96. Halpern-Sebold, L. R., Margolis-Kazan, H., Schriebman, M. P. and Joh, T. H. (1985) Proc. Soc. Exp. Biol. Med., 178: 486-489. Groves, D. and Batten, T. F. C. (1985) Cell Tissue Res., 240: 489-492. Nagahama, Y., Nishioka, R. S., Bern, H. A. and Gunther, R. L. (1975) Gen. Comp. Endocrinol., 25: 166-188. Wigham, T., Nishioka, R.S. and Bern, H. A. (1977) Gen. Comp. Endocrinol., 32: 120-131. Wigham, T., Ball, J. N. and Ingleton, P. M. (1975) J. Comp. Physiol., B 104: 87-96. Peter, R. E. and McKeown, B. A. (1975) Gen. Comp. Endocrinol., 25: 153-165. Olivereau, M. (1975) Gen. Comp. Endocrinol., 26: 550-561. Olivereau,M., Aimar,C. and Olivereau, J. M. (1980) Cell Tissue Res., 208: 389-404. McKeown, B. A., Jenks,B.G. and van Over- beeke, A. P. (1980) Comp. Biochem. Physiol., 65B: 705-709. James, V. A. and Wigham, T. (1984) Gen. Comp. Z. Zellforsch., 110: 496- 278 49 50 51 52 53 54 55 56 58 59 60 61 62 63 64 65 66 67 68 69 70 Endocrinol., 56: 231-239. Wigham, T. and Ball, J. N. (1977) Gen. Comp. Endocrinol., 31: 148-153. Suzuki, R., Kishida, M. Zool. Sci., 2: 974. Olivereau, M. and Olivereau, J. (1979) Cell Tissue Res., 196: 397-408. Olcese, J. M., Hall, T. R., Figueroa, H. R. and de Vlaming, V. L. (1979) Gen. Comp. Endocrinol., 38: 309-313. Olivereau, M. (1978) Cell Tissue Res., 191: 93-99. Olcese, J., Figueroa, H., Hall, T. R., Yurgens, P., Kiebzak, G., Meyer, R. and de Vlaming, V. (1981) Gen. Comp. Endocrinol., 43: 462-466. Deery, D. J. (1975) Gen. Comp. Endocrinol., 25: 395-399. Chang, J.P., Marchant, T.A., Cook, A. F., Nahorniak, C. S. and Peter, R. E.(1985) Neuro- endocrinology, 40: 463-470. Nishioka, R.S., Bern,H.A., Kelley, K. M., Grau, E. G., Rivas, R. and Young, G. (1986) First Int. Congr. Neuroendocrinol., p. 65. Vale, W., Ling, N., Rivier, J., Villareal, J., Rivier, C. and Brown, M. (1976) Metabolism, 35: 1491- 1494. Dubois, M. P., Billard, R., Breton, B. and Peter, R. E. (1979) Gen. Comp. Endocrinol., 37: 220- 232. Grau, E.G., Nishioka, R.S., Young,G. and Bern, H. A. (1985) Gen. Comp. Endocrinol., 59: 350-357. King, J. A. and Millar, R. P. (1979) Endocrinol- ogy, 105: 1322-1329. Batten, T. F.C., Groves, D. J., Ball, J. N., Cam- bré, M.L., Verdonck, W.A., Ollevier, F. and Vandesande, F. (1985) Cell Tissue Res., 242: 115- 125. Batten, T. F. C. (1986) Cell Tissue Res., 244: 661- 672. Kah, O., Chambolle, P., Dubourg, P. and Dubois, M. P. (1982) C.R. Acad. Sci., 294: 519-524. Vigh-Teichman, K., Vigh, B., Korf, H.W. and Oksche, A. (1983) Cell Tissue Res., 233: 319-334. Olivereau, M., Ollevier, F., Wandesande, F. and Olivereau, J. (1984) Cell Tissue Res., 238: 289- 296. Olivereau, M., Ollevier, F., Wandesande, F. and Verdonck, W. (1984) Cell Tissue Res., 237: 379- 382. Nishioka, R. S. (1987) Ph. D. Dissertation, Univ. Tokyo, Japan. Grau, E.G., Nishioka,R.S. and Bern, H. A. (1982) Endocrinology, 110: 910-915. Rivas, R.J., Nishioka, R.S. and Bern, H. A. (1986) Gen. Comp. Endocrinol., 63: 245-251. and Hirano, T. (1985) 71 AZ 76 77 85 86 87 R. S. NisHioka, K. M. KELLEY AND H. A. BERN Grau, E.G., Ford, C.-A., Helms, L. M. H., Shi- moda, S. K. and Cooke, I. M. (1987) Gen. Comp. Endocrinol., 65: 12-18. Wigham, T. and Batten, T. F.C. (1984) Gen. Comp. Endocrinol., 55: 444-449, Batten, T. F.C. and Wigham,T. (1984) Cell Tissue Res., 237: 595-603. Fryer,J.N., Nishioka,R.S. and Bern, H. A. (1979) Gen. Comp. Endocrinol., 39: 244-246. Nishioka, R.S., Grau,E.G. and Bern, H. A. (1985) Gen. Comp. Endocrinol., 60: 90-94. Helms, L.M.H., Grau,E.G., Shimoda, S. K., Nishioka, R.S. and Bern,H. A. (1987) Gen. Comp. Endocrinol., 65: 48-55. Marchant, T. A., Fraser,R.A., Andrews, P. C. and Peter, R. E. (1987) Reg. Peptides, 17: 41-52. Cook, A. F. and Peter, R. E. (1984) Gen. Comp. Endocrinol., 54: 109-113. Sweeting, R. M. and McKeown, B. A. (1986) Can. J. Zool., 64: 2062-2063. Jackson, I. M. D. and Reichlin, S. (1974) Endocri- nology, 95: 854-862. Prunet, P. and Gonnet, F. (1986) First Int. Congr. Neuroendocrinol., p. 114. Barry, T. P. and Grau, E. G. (1986) Gen. Comp. Endocrinol., 62: 306-314. Kelley, K. M., Nishioka, R.S. and Bern, H. A. (1988) Gen. Comp. Endocrinol., (in press). Said, S. I. (1986) J. Endocrinol. Invest., 9: 191- 200. Koiwai, K., Kikuyama,S., Seki, T. and Yanai- hara, N. (1986) Gen. Comp. Endocrinol., 64: 254- 259. Goos, H. J. T. and Murathanoglu, O. (1977) Cell Tissue Res., 181: 163-168. Kah, O., Breton,B., Dulka,J.G., Nunez- Rodriguez, J., Peter, R. E., Corrigan, A., Rivier, J.E. and Vale, W. W. (1986) Cell Tissue Res., 244: 327-337. Ollevier, F. and Verdonck, W. (1984) Gen. Comp. Endocrinol., 54: 433. Verdonck, W. and Ollevier, F. (1984) Gen. Comp. Endocrinol., 53: 436. Yulis, C. R., Lederis, K., Wong, K. L. and Fisher, A. W. F. (1986) Peptides, 7: 79-86. Fryer, J., Lederis, K. and Rivier, J. (1985) Reg. Peptides, 11: 11-15. Clarke, W.C. and Bern, H. A. (1980) In “Hor- monal Proteins and Peptides”, Vol. 8. Ed. by C. H. Li, Academic Press, New York, pp. 105-197. Bern, H. A. (1983) Am. Zool., 23: 663-671. Specker, J. L., Ingleton, P.M. and Bern, H. A. (1984) Frontiers and Perspectives in Prolactin Secretion: A Miultidisciplinary Approach. Academic Press, New York, pp. 17-30 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 Prolactin and Growth Hormone in Teleosts Pickford, G. E. and Phillips, J. G. (1959) Science, 130: 454-455. Hirano, T. and Mayer-Gostan, N. (1978) In “Com- parative Endocrinology”. Ed. by P. J. Gaillard and H. H. Boer, Elsevier/North Holland, Amsterdam, pp. 209-212. Nicoll, C.S., Wilson, S. W., Nishioka R.S. and Bern, H. A. (1981) Gen. Comp. Endocrinol., 44: 365-373. Loretz,C. A. and Bern, H. A. (1982) Neuroen- docrinology, 35: 292-304. Nagahama, Y., Clarke, W.C. and Hoar, W. S. (1977) Can. J. Zool., 55: 183-198. Suzuki, R. and Hirano, T. (1987) Proc. 1st Congr. Asia and Oceania Soc. Comp. Endocrinol., pp. 335-336. Hirano, T. (1985) Aquaculture, 45: 375-394. Cook, H. and van Overbeeke, A. P. (1969) Can. J. Zool., 47: 937-941. McKeown, B. A. and Leatherland, J. F. (1973) Z. Zellforsch., 140: 459-471. Olivereau, M. (1967) Z. Zellforsch., 80: 264-285. Aler, O. M. (1971) Acta Zool., 52: 145-156. Olivereau, M., Lemoine, A. M. and Dimovska, A. (1971) Ann. Endocrinol., 32: 271. Ingleton, P. M., Baker, B. I. and Ball, J. N. (1973) J. Comp. Physiol., 87: 317-328. Baker, B. I. and Ingleton, P. M. (1975) J. Comp. Physiol., 100: 269-282. Benjamin, M. and Baker, B. I. (1978) Cell Tissue Res., 191: 161-170. Hall, T. R. and Chadwick, A. (1978) Gen. Comp. Endocrinol., 36: 388-395. Nagahama, Y. (1973) Mem. Fac. Fish. Hokkaido Univ., 21: 1-63. Wendelaar Bonga, S. E. (1978) Gen. Comp. En- docrinol., 34: 265-275. Lam, T. J. and Leatherland, J. F. (1969) Comp. Endocrinol., 12: 385-387. Leatherland, J. I. and McKeown, B. A. (1974) J. Comp. Physiol., 89: 215-226. Suzuki, R., Kishida, M., Ogasawara, T., Hasega- wa, S. and Hirano, T. (1987) Gen. Comp. Endo- crinol., 68: 76-81. Batten, T. F. C., Young, G. and Ball, J. N. (1983) Gen. Comp. Endocrinol., 51: 113-130. Olivereau, M. and Ball, J. N. (1970) Mem. Soc. Endocrinol., 18: 57-85. Leatherland, J.F., Ball,J.N. and Hyder, M. (1974) Cell Tissue Res., 149: 245-266. Zambrano, D., Clarke, W.C., Hajek, A., Sage, M. and Bern, H. A. (1974) Acta Zool., 55:202- 216. Bolton, J. P., Collie, N. L., Kawauchi, H. Hirano, T. (1987) J. Endocrinol., 112: 63-68. Gen. and 121 122 123 129 130 142 143 144 145 146 147 279 Kaul, S. and Vollrath, L. (1974) Cell Tissue Res., 154: 211-230. Young, G. and Ball, J. N. (1979) J. Endocrinol., 83: 30P. Wendelaar Bonga, S.E. and van der Meij, J.C. A. (1980) Gen. Comp. Endocrinol., 40: 391-401. Wendelaar Bonga, S.E. and van der Meij, J.C. A. (1981) Gen. Comp. Endocrinol., 43: 432-442. Wendelaar Bonga, S. E., Lowik, C. J. M. and van der Meij, J.C. A. (1983) Gen. Comp. Endocri- nol., 52: 222-231. Grau, E.G., Nishioka,R.S. and Bern, H. A. (1981) Gen. Comp. Endocrinol., 45: 406-408. MacDonald, D. J. and McKeown, B. A. (1983) Can. J. Zool., 61: 682-684. Copp, D. H. and Ma, S. W. Y. (1981) In “Hor- monal Regulation of Calcium Metabolism”. Ed. by D.V.Cohn, R. V. Talmage and J. L. Matheus, Excerpta Medica, New York, pp. 298-306. Taraskevich, P.S. and Douglas, W. W. Nature, 276: 832-834. Grau, E. G., Shimoda, S. K., Ford, C.-A., Helms, L. M. H., Cooke, I. M. and Pang, P. K. T. (1986) Endocrinology, 119: 2848-2855. Thaw, C. N., Geras Raaka, E. and Gershengorn, M. C. (1984) Am. J. Physiol., 247: C150. Thompson, (1980) The Regulation of Membrane Lipid Metabolism. CRC Press, Boca Raton, pp. 1- 218. Michell, R. H. (1982) Nature, 296: 492-493. Axelrod, J. and Hirata, F. (1982) Trends Pharma- col., 3: 156-158. Heindel, A. A. and Clement-Cormier, (1981) Endocrinology, 108: 310-317. MacDonald, D. J. and McKeown, B. A. Comp. Biochem. Physiol., 80A: 553-557. Sweennen, L. and Denef, C. (1982) Endocrinolo- gy, 111: 398-405. Merritt, J. E., MacNeil,S., Tomlinson,S. and Brown, B. L. (1983) Endocrinology, 98: 423-429. Schettini, G., Cronin, M. J. and Macleod, R. M. (1983) Endocrinology, 112: 1801-1807. Olivereau, M. (1971) Z. Zellforsch., 121: 232-243. Olivereau, M. and Lemoine, A. M. (1973) C. R. Acad. Sci. Paris, 276D: 1325-1327. Olivereau, M. and Lemoine, A. M. (1973) C. R. Acad. Sci. Paris, 276D: 1883-1886. Sage, M. (1970) J. Exp. Zool., 173: 121-127. Sundararaj, B. I. and Nayyar, S. K. (1969) Gen. Comp. Endocrinol., Suppl. 2: 69-80. Zambrano, D. (1975) Cell Tissue Res., 162: 551- 563. Wigham, T. and Ball, J. N. (1976) J. Comp. Phys- iol., 110: 135-143. McKeown, B. A. (1972) Experientia, 28: 675-676. (1978) Yn (1985) 280 148 149 163 165 166 167 Zambrano, D., Clarke, W.C., Hawkins, E. F., Sage, M. and Bern, H. A. (1973/74) Neuroendo- crinology, 13: 284-298. Olivereau, M. and Lemoine, A. M. Comp. Physiol., 79: 411-422. Olivereau, M. (1966) Am. Zool., 6: 598. Abraham, M., Dinari-Lavie, V. and Lotan, R. (1977) Cell Tissue Res., 179: 317-330. Abraham, M. and Lotan, R. (1974) Cell Tissue Res., 149: 267-274. Leatherland, J. F. (1972) Can. J. Zool., 50: 835- 844. Mattheij, J. A. M., Stroband, H. W. J. and King- ma, F. J. (1971) Z. Zellforsch., 118: 113-126. Ball, J. N. and Pickford, G. E. (1964) Anat. Rec., 148: 358. Emmart, E. W., Pickford, G. E. and Wilhelmi, A. E. (1966) Gen. Comp. Endocrinol., 7: 571-583. Leatherland, J. F. (1970) J. Endocrinol., 48: xxxi- XXXIL. Leatherland, J. F. (1970) Z. Zellforsch., 104: 301- 317. Nagahama, Y., Nishioka, R.S. and Bern, H. A. (1973) Z. Zellforsch., 136: 153-157. Ichikawa, T., Kobayashi, H., Zimmermann, P. and Miller, V. (1973) Z. Zellforsch., 141: 161- 179. Olivereau, M. (1968) Z. Zellforsch., 87: 545-561. Dharmamba, M. and Nishioka, R. S. (1968) Gen. Comp. Endocrinol., 10: 409-420. Clarke, W. C. (1973) Can. J. Zool., 51: 687-695. Wendelaar Bonga, S. E., Flik,G., Lowik, C. W. (1972) J. G.M. and vanEys,G.J.J.M. (1985) Gen. Comp. Endocrinol., 57: 352-359. Nagahama, Y. and Yamamoto, K. (1971) Bull. Jpn. Soc. Sci. Fish., 37: 691-698. Utida, S., Hatai, S., Hirano, T. and Kamemoto, F. (1971) Gen. Comp. Endocrinol., 16: 566-573. Ball, J. N. and Olivereau, M. (1965) Am. Zool., 5: 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 R. S. NisHiokA, K. M. KELLEY AND H. A. BERN 232-233. Hopkins, C. R. (1969) Tissue and Cell, 1: 653-672. Ensor, D. M. and Ball, J. N. (1968) Gen. Comp. Endocrinol., 11: 104-110. Ball, J. N. and Ingleton, P. M. (1973) Gen. Comp. Endocrinol., 20: 312-325. Holtzman, S. and Schreibman, M. P. (1972) J. Exp. Zool., 180: 187-196. Schreibman, M.P. and Kallman, K. D. Gen. Comp. Endocrinol., 6: 144-155. Holtzman, S. and Schreibman, M. P. (1971) Am. Zool., 11: 653-654. Sage, M. (1968) Gen. Comp. Endocrinol., 10: 70- 74. Holtzman, S. (1973) Ph. D. Dissertation, City Univ. N. Y. Sage, M. (1966) J. Endocrinol., 34: ix—x. Holtzman, S. and Schreibman, M. P. (1975) Gen. Comp. Endocrinol., 25: 447-455. Holtzman, S. and Schreibman, M. P. (1975) Gen. Comp. Endocrinol., 25: 456-461. Holtzman, S. and Schreibman, M. P. (1976) Gen. Comp. Endocrinol., 30: 234-237. Leatherland, J. F. and Lin, L. (1975) Zool., 53: 297-310. McKeown, B. A. and Hazlett, C. A. (1975) Can. J. Zool., 53: 1195-1200. Brewer, K. J. and McKeown, B. A. Comp. Physiol., 140: 217-225. McKeown, B. A. and Brewer, K.J. (1978) In “Comparative Endocrinology”. Ed. by P. J. Gail- lard and H.H. Boer, Elsevier/North-Holland Biomedical Press, Amsterdam, pp. 357-362. Prunet, P., Boeuf,G. and MHoudebine, L. M. (1985) J. Exp. Zool., 235: 187-196. Gonnet, F., Prunet, P., Tonon, M. C., Dubourg, P., Kah, O. and Vaudry, H. (1988) Gen. Comp. Endocrinol., 69: 252-261. (1966) Can. J. (1980) J. ZOOLOGICAL SCIENCE 5: 281-290 (1988) A Possible Sugar Receptor Protein Found in the Labellum of the Blowfly, Phormia regina MaMIKO OZAKI Department of Biology, College of General Education, Kobe University, Nada, Kobe 657, Japan ABSTRACT— Previous physiological studies have suggested the existence of at least three functionally separated receptor sites in the labellar sugar receptor of the fly, called the pyranose (P site), the furanose (F site) and the third sites (T site), and that starch acts as a competitive inhbitor for the P site. I detected in this work a new candidate protein for the P site in the labellar extract by affinity electrophoresis with starch. The dissociation constant of the candidate protein-starch complex was estimated to be 0.7%, a value consistent with the electrophysiological estimate of the inhibition constant for starch on the sugar response. The stimulus sugars for the P site bound to the candidate protein in competition with starch. The dissociation constants of the candidate protein-sugar com- plexes were highly correlated with the electrophysiological constants defined as the sugar concentra- tions which give rise to half maximal responses. However, the stimulus sugars for the F site did not compete with starch for the candidate protein. The candidate protein was water insoluble and appeared to be located in the distal process and the cell body, but not in the axon, of the labellar © 1988 Zoological Society of Japan chemosensory cell. INTRODUCTION Since the pioneering work of Dastoli and Price [1], repeated attempts have been made to identify sweet-taste receptor molecules in both vertebrates and invertebrates. However, no taste receptor proteins have been definitively identified yet, and biochemical studies on the sense of taste have met with only limited success. In the fly, many behavioral and electrophysiolog- ical studies have suggested that the sugar receptor cell is sensitive to a broad spectrum of chemicals and that it has multiple receptor sites. At least three functionally separate receptor sites have been documented in the sugar receptor cell of the fleshfly [2-4]. They are the pyranose site (P site) sensitive to sucrose, maltose, D-glucose, L-fucose, etc., the furanose site (F site) sensitive to D- fructose, D-fucose, etc., and the third site (T site) sensitive to aliphatic carboxylate anions. As a working hypothesis, the P site was proposed to be identical with an a-glucosidase [5], but some inconsistencies have been reported for the hypoth- Accepted September 2, 1987 Received December 8, 1986 esis [6]. I recently found that starch acts as a competitive inhibitor on the P site [7]*. Therefore, it was possible to detect a new candidate protein for the P site by affinity electrophoresis with starch. In this paper, I report on the identification of the candidate protein based on its affinity for starch or stimulus sugars and its location in the chemosensory cell. MATERIALS AND METHODS Sample preparation Five to seven-day-old blowflies, Phormia regina, reared at 20-25°C and fed with 0.1 M sucrose were used. Living flies were anesthesized by cooling on ice, and the labella were cut at the distal end of the proboscis. Two hundred labella so obtained were frozen with a small volume of liquid nitrogen in a hand mortar. After the frozen labella were homogenized for 20min, the homogenate was suspended in 0.1 ml of sample buffer (4.65 mM sodium barbiturate-HCl, 2% Triton X-100, 10% * Paper presented by the author in her maiden name, Hara. 282 M. Ozaki glycerol, pH 6.8) and incubated at 4°C for 1 hr. The suspension was centrifuged at 3,000 rpm at 4°C for 10 min, and the supernatant was used as the sample extract. The extract of the proboscis without labellum was prepared in the same way to serve as control. To examine the water soluble fraction, the labellar homogenate was suspended in distilled water, incubated at 4°C for 1 hr, and centrifuged at 50,000rpm at 4°C for lhr. The supernatant was then lyophilized and dissolved in the same buffer. In addition, the labellum was separated into “Jabellar content” and the “labellar integument” following the method of Amakawa et al. [8]. Labella were gently sonicated, 150-200 at a time, in 5 ml of distilled water with a Tominaga model UR-150P ultrasonicator (25 kw output) on ice for 5 min. Subsequently, 500 “labellar integuments”, or cuticle lobes, were collected and washed twice with distilled water. The extract was prepared in the same way as the whole labellum extract. The “labellar contents”, which were washed out from the 500 labella into distilled water during sonica- tion, were collected by lyophilization. They were then homogenized with 0.1 ml of sample buffer in a glass homogenizer on ice for 10 min and incu- bated at 4°C for 1 hr. The supernatant obtained following centrifugation at 3,000 rpm for 10 min was used as the sample extract. Electron microscopy For the scanning electron microscopy, whole labella or “labellar integuments” were fixed in 3% glutaraldehyde at 20°C for 1lhr, dehydrated through ethanol series and isoamylacetate, dried at the critical point of CO, with a Hitachi model HPC-2 dryer, and coated with gold by spattering with an Eiko model IB-3 ion-coater. They were then observed with Hitachi model S—430 scanning electron microscope. For the transmission electron microscopy, whole labella or “labellar integuments” were fixed in 3% glutaraldehyde at 20°C for 1 hr and 2% osmium tetroxide on ice for lhr, dehydrated through ethanol series and propylene oxide, and embedded in Epon 812 resin. Sections were cut with a Porter-Blum model MT-1 ultramicrotome, dou- ble-stained with lead acetate and uranyl acetate and observed with a Hitachi model H-300 electron microscope. Affinity electrophoresis As many differnt kinds of proteins are present in the blowfly labellum, it was difficult to detect a minor protein such as the sugar receptor protein in one-dimensional electrophoresis. Therefore, two- dimensional polyacrylamide gel electrophoresis, in which an affinity ligand was added to the running gel in the first dimension, was adopted. The gel system was similar to the Ornstein and Davis’s stacking system [9, 10], except that Triton X-100 (2% at the final concentration) was added to the stacking (4.5% acrylamide) and running gels (7.5% acrylamide), and barbiturate buffers (stack- ing gel buffer: 9.3mM sodium barbiturate-HCl, pH 6.7; running buffer: 91.1 mM sodium barbitu- rate-HCl, pH8.9; electrode buffer: 41.1mM sodium barbiturate-glycine, pH 8.3) were used instead of Tris buffers, which inhibit sugar re- sponses [11]. Proteins were detected by the silver staining method of Oakley et al. [12]. The first dimensional electrophoresis was carried out with 10-20 yl of the sample extract at 3 mA for 100 min in a disc gel (2 mm in diameter, 130 mm in length). The gel was removed into sample buffer, shaken at room temperature (20-25°C) for 1 hr, and loaded onto a slab gel (130115 x1 mm) for the second dimensional electrophoresis, carried out at 30 mA for 250 min. The composition of gels and buffers used in the two electrophoretic runs were similar except for the starch added in the first dimension. To evaluate the affinity of sugars for the candidate protein, stimulus sugar was added to the running gel, together with starch, in the first dimensional electrophoresis. During electrophoreses, the gel temperature measured through the glass tube or plate was 21+2°C, and the pH of the running gel was 9.4 immediately after the run. Calculation of dissociation constant Starch is a large polysaccharide molecule with no electric charge, which, when complexed with a protein, greatly retards the mobility of the protein during electrophoresis. Therefore, if the mobility of the protein, m,, decreases to m in the presence of starch of concentration [J], the protein-starch Sugar Receptor Protein in the Blowfly 283 interaction can be expressed by the following equation [13]: m/m=1+{[I]/K; (1) where K; is the dissociation constant of protein- starch complex. Thus, the plot of m,/m against [/] yields a straight line with [/] intercept at —K;. Here the dissociation constant is expressed in % wiv because starch is polydisperse and therefore does not lend itself to the molarity measure. On the other hand, stimulus sugars for the sugar receptor are much smaller than starch and have no electric charge so that the mobility of the protein- sugar complex is nearly the same as that of the free protein. Therefore, if the mobility, m, of a protein in the presence of starch becomes m’ in the presence of both sugar (concentration, [S]) and starch (concentration, [/]), the protein-sugar in- teraction is given by the following equation [14]: m'(m,—m’)=K(1+[S\V/Ka)/[] (2) where Kj, is the dissociation constant of protein- sugar complex. Thus, the plot of m’/(m,—m’) against [S] gives a straight line intercepting the [5S] axis at —K,. From Eqs. (1) and (2) it may be seen that if m’ is equal to m, Kg is infinitely large. This means that the protein-sugar interaction through the starch binding center is negligible. That is, the protein migrates as if the gel contained only starch. Assay of a-glucosidase The a-glucosidase activity in the disc gel was assayed with p-nitrophenyl a-D-glucopyranoside (a-PNPG) [15]. After the electrophoresis in the first dimension, the disc gel was removed from the glass tube, and the running gel part was divided equally into 19 pieces (5 mm length). Each piece was incubated in the reaction mixture containing 10 mM a-PNPG in 0.5 ml of 0.1 M sodium citrate buffer (pH 6.0). After incubation at 27°C for 1 hr, the reaction was stopped by adding 2 ml of 0.5 M Tris-HCl (pH 9.0), and the absorbance of liber- ated p-nitrophenol was measured at 410 nm with the Shimadzu model UV-202 recording spec- trophotometer. Chemicals Sucrose, maltose, D-glucose, D-xylose, L- sorbose and D-fructose were purchased from Wako Pure Chemicals, Osaka, Japan. D- and L-fucose were obtained from Nakarai Chemicals, Ltd., Osaka Japan. RESULTS Affinity for starch When two-dimensional electrophoresis of the labellar extract was carried out without starch, all proteins whose mobilities are different from each other migrated into a diagonal line on the slab gel, because each individual protein in the labellar extract showed the same mobility in the first dimension as in the second (Fig. la). When the electrophoresis was carried out with starch in the first dimension, however, a single spot was repro- ducibly found separated from the diagonal line. Figure 1b, c and d show that the mobility of this protein in the first dimensiom decreases with increasing concentration of starch in the running gel. In this way, a protein with affinity for starch was easily detected using this system. The values of m, and m for the protein were directly measured on each gel (see Fig. 2b), and the ratio of mobilities, m,/m, was plotted against the concentration of starch, [/], to estimate the dis- sociation constant of the protein-starch complex, K;, to be 0.7% (Fig. 2). The value is consistent with an electrophysiological estimate of the inhibi- tion constant for starch at the P site of the sugar receptor at 22+2°C, i.e. around 0.6% [7]. The responsiveness of the fly to sugars is stable over a pH range of 3 to 10 [16]. Thus, these elec- trophoretical and electrophysiological estimates are comparable with each other. Affinity for sugars L- and D-fucose stimulate the P and the F sites, respectively [2], though they are neither metabo- lized in the blowfly [17] nor bind to the membrane- bound a-glucosidase in the blowfly labellum [6]. When electrophoresis was carried out in the presence of both D-fucose and starch, a protein spot was detected almost in the same position (Fig. 3c) as in the presence of starch alone (Fig. 3a). When L-fucose was added instead of D-fucose, 284 M. Ozaki Fic. 1. Two dimensional affinity electrophoresis of the labellar extract: (a), 0%; (b), 0.5%; (c), 1%; (d), 2% starch in the first dimension. In Figs. 1 and 3, the origin of electrophoresis is the upper left-hand corner, and the proteins migrated toward the right and the bottom, respectively, in the first and the second dimensional electrophoreses (see Fig. 2b). The candidate protein spot is indicated by an arrow in each figure. however, this spot was detected closer to the diagonal line (Fig. 3b). These results indicated that D-fucose did not compete with starch for the protein but L-fucose did. Thus, this protein is a possible sugar receptor molecule for the P site but not for the F site. Moreover, this new candidate protein very likely is different from a-glucosidase, since L-fucose does not bind to a-glucosidase [6]. Figure 4 shows the plot of the ratio m’/(m,—m’) against the concentration [S] of L- (a) and D- fucose (b), respectively. The calculated dissocia- tion constant of the candidate protein-L-fucose complex was 254 mM, while that of the candidate protein-D-fucose complex was infinite. Some other stimulus sugars for the P site were also examined, and the dissociation constants of the complexes between the candidate protein and these sugars are listed in Table 1. These dissocia- tion constants were calculated using Eq. (2) in Materials and Methods. For sucrose, maltose, D-glucose and D-fructose, the dissociation con- stants, Ky, were compared with the mid-point concentrations, K,, defined as the concentration of stimulus which gives rise to half maximal elec- trophysiological responses. As seen in Table 1, the dissociation constant was 4-5 fold larger than the mid-point concentration for sucrose, maltose or D-glucose but infinitely larger than that for D- fructose. Localization of the candidate protein The water soluble fraction of labellum was Sugar Receptor Protein in the Blowfly 285 RELATIVE MOBILITY. -05 0 05 10 15 20 CONCENTRATION OF STARCH, II] %/o b—> Fic. 2. (a) Determination of the dissociation constant of the candidate protein-starch complex by plotting the reciprocal of the relative mobility of the candi- date protein, m,/m, against the concentration of starch, [J]. (b) Illustration for the measurement of mandm,. In the presence of starch, the candidate protein migrates to the position p. The position to which it would have migrated in the absence of starch, q, is estimated by extending the line op horizontally to the diagonal line defined by the protein stain (dotted area). When the protein stain appeared smeared in the central portion, the di- agonal line was defined as that line from the origin of the electrophoresis to some distinguishable spots near the leading front of protein migration. The mobilities, m and m,, are proportional to the dis- tance op and oq, respectively. examined first for the presence of the candidate ; : : Fic. 3. Comparison of affinity for the candidate pro- protein, but the candidate protein was not de- tein between L- and D-fucose. (a) same elec- tected. The extract of the proboscis from which trophoretic pattern shown in Fig. Id presented as the labellum was removed also yielded negative no sugar control; (b) 0.2M L-fucose plus 2% results. These results suggested that the candidate starch, (c) 0.2M D-fucose plus 2% starch in the : : first dimension. protein was a labellum specific, membrane-bound ii protein. I attempted to determine in what part of 286 M. OzakI -0.2 -0.1 0 0.1 0.2 0.3 04 CONCENTRATION OF L-FUCOSE, M 0.2 0 0.1 CONCENTRATION OF D-FUCOSE, M Fic. 4. Determination of the dissociation constant of the candidate protein-sugar complex by plotting m'/(m,-m’) against the concentration of L-fucose (a) and D-fucose (b). the sensory cell the candidate protein is located. Although the sensillum tip is especially rich in receptor membranes, they are too thin to cut and to collect. Therefore, I attempted to isolate the receptor membranes by sonication. Sonication separated the “labellar integument”, which con- tained the receptor membranes, from the “labellar content”, which consisted of the sensory cell bodies, labial nerve, supportive cells and connec- tive tissues. In the intact labellum, the cell bodies of 4 chemosensory and a mechanosensory cells are surrounded by supportive cells at the base of each chemosensillum, and their axons extend into the labial nerve. In the “labellar integument”, howev- er, all these structures were completely removed, exposing the inside surface of the cuticle (Fig. Sa, c). Nevertheless, the chemosensilla were still attached to “labellar integument” preserving the membrane fragments in the inner lumen (Fig. 5d). These membrane fragments were thought to be derived from the distal processes of the chem- osensory cells. In both the “labellar content” and the “labellar integument”, the candidate protein was detected as a spot separated from the diagonal line, similar to the spot seen in Figure 1. Thus, the candidate protein seemed to be located in both the cell body and the distal process of the sensory cell. a-Glucosidase To examine the hypothesis that an a-glucosidase is the P site molecule, I compared the mobility of the a-glucosidase with that of the newly detected candidate protein in disc gel electrophoresis in the presence of varying concentration of starch. The a-glucosidase activity was always found at 30-35 mm from the origin regardless of the presence of starch, while the mobility of the new candidate protein decreased with increasing concentrations of starch (Fig. 6). TABLE 1. Dissociation constant of candiate protein-sugar complex Stimulu Site Kg+S.D. Test K,+S.D.” Test sugar specificity” (mM) No. (mM) No.” sucrose P 104+ 25 4 21+10 25 maltose P 110+20 2 26+6 15 L-fucose P 254+21 5 D-glucose P 360425 4 83427 8 L-sorbose P 575 +98 3 D-xylose P 1504 +596 2 D-fructose F oo 4 53417 14 D-fucose F foe) 5 1): Shimada (1974), 2): Hara (1983) co, Calculated value is more than 10M. Sugar Receptor Protein in the Blowfly 287 Fic. 5. Electron micrographs of “labellar integument” and intact labellum. (a) and (b), scanning electron micrographs of the internal appearance of labellar lobes of a “labellar integuments” and an intact labellum, respectively; (c), scanning electron micrograph of the bases of chemosensilla seen from the inside of a “labellar integument”. An attached sensillum can be seen through the crack in the cuticle; (d) and (e), cross sections of chemosensilla of the “labellar integument” and the intact labellum, respectively. Bars indicate 100 4m (a, b), 10 wm (c), and 1 pm (d, e). 288 M. OzakI 0%. starch 0 A 0.5%. starch A 1 1%. starch 0 A Ze starch ABSORBANCE AT 410 nm Ora 5 10 DISTANCE FROM ORIGIN, cm Fic. 6. Localization of a-glucosidase activity in the disc gel used in the first dimension under varying con- centrations of starch. The ordinate indicates the a-glucosidase activity determined by the absorbance at 410 nm, and the abscissa shows the distance from the origin in the running gel. The location of the candidate protein in the disc gel was estimated by two dimensional electrophoresis (arrow heads). DISCUSSION Affinity electrophoresis It has been thought that, in the taste receptor, the receptor-stimulus interaction is too weak and the quantity of the receptor protein is too small to detect or isolate the receptor protein. In the labellar sugar receptor of the fly, the P site-sucrose interaction is comparatively strong, but the dis- sociation consant of the P site-sucrose complex is still more than 0.01 M, according to Morita [18]. Such a weak interaction cannot be detected direct- ly by any means other than affinity electrophoresis. For the application of affinity electrophoresis, however, it is imperative that affinity ligand is water-soluble and immobile in the polyacrylamide gel. Starch becomes water-soluble when heated. Its molecular weight, estimated to be 10-100 times larger than that of the candidate protein, was expected to be large enough for it to be immobile in 7.5% polyacrylamide gel. The result shown in Figure 2 satisfied the linear relationship between m,/m and [J] demanded by Eq. (1), strongly suggesting that starch when complexed with the candidate protein was indeed immobile in the polyacrylamide gel. Thus, starch is a satisfactory affinity ligand and has already been used in affinity electrophoresis for phosphorylase [13, 14] or am- ylase [19]. As in the experiments on phosphorylase or amylase, the P site-starch interaction was strong enough to detect the P site molecule. Further- more, the two dimensional electrophoresis tech- nique applied here was very useful in isolating the protein of interest among many different kinds of protein. As for the quantity of the receptor protein, Hansen and Wieczorek [20], on the basis of semi-quantitative calculation, estimated that 107’ or 107 '° g of labellar sugar receptor protein can be obtained from 1,000 flies. The silver staining method of Oakley et al. [12] is sufficiently sensitive to detect protein density of 10~''-10~ '° g/mm? on a gel plate. Since the candidate protein extracted from 20-40 flies made a visible spot about 2 mm diameter on the gel, at least 10°-°g of the candidate protein should be obtainable from 1,000 flies, consistent with the estimate of Hansen and Wieczorek [20]. Receptor-sugar interaction The receptor-stimulus interaction is thought to be the primary process in the chemosensory transduction mechanism. The existence of the specific receptor molecule has not always been accepted in the case of salt, water or bitter taste reception [21]. In the case of sugar or amino acid reception, however, it is generally accepted that a specific receptor molecule mediates the receptor function [20, 21]. The electrophysiological analysis of the sugar receptor cell of the fly, in particular, is well developed, and the relation between the dissociation constant of the receptor-sugar com- Sugar Receptor Protein in the Blowfly 289 plex, Kz, and the mid-point concentration, Ky, is described [18] as follows: Ky=K,(sg/G +1), where s is the total receptor site; g, the conduct- ance per activated site; G, the conductance across the receptor membrane in the resting state. Ap- plying the electrophysiological data on the recov- ery process of the sucrose response to the above equation, Ninomiya et al. [23] obtained sg/G=4 or Ky=5K, . This is in good agreement with my result, i.e., Ky=4K, (Table 1). These results may be inter- preted in terms of amplification at the conductance level. That is, in the primary process of the chemosensory transduction in the sugar receptor of the fly, sucrose, maltose, D-glucose, etc., bind to the P site molecule, consisting of the candidate protein, according to their individual affinities, but the conductance change across the sugar receptor membrane uniformly gives 4 to 5-fold amplifica- tion of the sensitivity. Table 1 shows the calculated dissociation con- stants, assuming the simple case that one molecule of starch or sugar binds to the P site molecule. Therefore, the calculated value for D-glucose is probably an overestimate because two glucose molecules are thought to bind to each P site molecule [24]. Location of the sugar receptor protein The candidate protein was found in the “labellar content” and the “labellar integument” but not in the proboscis from which the labellum had been cut off. This suggested that the candidate protein is located in the cell body and the distal process but not in the axon of the labellar chemosensory cell. The P site molecule is probably synthesized in the cell body of the sugar receptor cell, transported to the distal process and concentrated in the receptor region at the tip. Recently, we suggested that the P site is located not only at the tip but also in the intermediate length of the distal process [23] and that a considerable number of receptor molecules for the P site exist in regions other than the receptor region of the sugar receptor cell [25]. Comparison with a-glucosidase Many behavioral and electrophysiological stud- ies have documented the stereospecificity of the P site for stimulus sugar [2, 7, 26], and it has been suggested that three successive equatorial hydroxyl groups in the chair form of the pyranose ring are essential for stimulation at the P site regardless of their position. Actually, all sugars which bind to the candidate protein have this essential structure and are stimulatory at the P site. As for starch, its glucopyranose residues do not have this essential structure, but it can compete with the stimulus sugars for the P site [7]. Although all stimuli which have been estimated behaviorally or electrophys- iologically could not be investigated in the present work, the candidate protein showed binding spec- ificity for several sugars and starch similar to that exhibited by the P site in behavioral or electro- physiological studies. However, L-fucose and starch do not bind to the labellar a-glucosidase. Thus, I conclude that this candidate protein is different from the a-glucosidase, and that it is rather likely to be the P site receptor molecule. ACKNOWLEDGMENTS I thank Dr. Taisaku Amakawa for letting me carry out this work in his laboratory as a research student. My thanks are also due to Professor Hiromichi Morita and Professor Tomiyuki Hara for the use of electron micro- scopes in their laboratories and Dr. Koichi Ozaki for his help with scanning electron microscopy. I also thank Professor Tsutomu Inoue for his invaluable advice on affinity electrophoresis, Dr. Fumio Hayashi for his general discussion and Professor W. L. Pak of Purdue University for kindly correcting and improving my English. REFERENCES 1 Dastoli, F. R. and Price, S. (1966) Sweet-sensitive protein from bovine taste buds: Isolation and assay. Science, 154: 905-907. Shimada, I., Shiraishi, A., Kijima, H. and Morita, H. (1974) Separation of two receptor sites in a single labellar sugar receptor of the fleshfly by the treat- ment with p-chloromercuribenzoate. J. Insect Phys- iol., 20: 605-621. 3 Shimada, I. (1975) Two receptor sites and their relation to amino acid stimulation in the labellar ie) 14 290 sugar receptor of the fleshfly. J. Insect Physiol., 21: 1675-1680. Shimada, I. and Isono, K. (1978) The specific receptor site for the aliphatic carboxylate anion in the labellar sugar receptor of the fleshfly. J. Insect Physiol., 24: 807-811. Hansen. K. (1969) The mechanism of insect sugar reception, a biochemical investigation. In “Olfaction and Taste, Vol. 3”. Ed. by C. Pfaffmann, Rockefel- ler Univ. Press, New York, pp. 382-391. Nakashima, M., Enomoto, K., Kijima,H. and Morita, H. (1982) Discrepancy between the affini- ties of certain inhibitors for the membrane-bound a-glucosidases and those for the sugar receptor of the flies. J. Insect Physiol., 12: 579-585. Hara, M. (1983) Competition of polysaccharides with sugars for the pyranose and the furanose sites in the labellar sugar receptor cell of the blowfly, Phormia regina. J. Insect Physiol., 29: 113-118. Amakawa, T., Kijima, H. and Morita, H. (1975) Insoluble-glucosidase: Possible pyranose site of the sugar receptor of the labellar hairs of the blowfly, Phormia regina. J. Insect Physiol., 21: 1419-1425. Ornstein, L. (1964) Disc electrophoresis. I. Back- ground and theory. Ann. N. Y. Acad. Sci., 121: 321-349. Davis, B. J. (1964) Disc electrophoresis. II. Method and application to human serum proteins. Ann. N. Y. Acad. Sci., 121: 404-427. Kawabata, K., Kijima, H., Shiraishi, A. and Mori- ta, H. (1973) a-Glucosidase isozymes and the label- lar sugar receptor of the fly. J. Insect Physiol., 19: 337-348. Oakley, B.R., Kirsch,D.R. and Morris, N.R. (1980) A simplified ultrasensitive silver stain for detecting proteins in polyacrylamide gels. Anal. Biochem., 105: 361-363. Takeo, K. and Nakamura, S. (1972) Dissociation constants of glucan phosphorylase of rabbit tissues studied by polyacrylamide gel disc electrophoresis. Arch. Biochem. Biophys., 153: 1-7. Takeo, K. and Nakamura, S. (1978) Dissociation constants of glucan phosphorylases to origo- and polyglucans studied by the affinity electrophoresis. In “Affinity Chromatography”. Ed. by O. Hoff- mann-Ostenhof, M. Breitenbach, F. Koller, D. Kraft and O. Scheiner, Pergamon Press, Oxford and New York, pp. 67-70. M. Ozaki 15 21 22 23 24 26 Halvorson, H. and Ellias, L. (1958) The purification and properties of an a-glucosidase of Saccharomyces italicus. Biochem. Biophys. Acta, 30: 28-40. Shiraishi, A. and Morita, H. (1969) The effects of pH on the labellar sugar receptor of the fleshfly. J. Gen. Physiol., 53: 450-470. Hassett, C. C., Dethier, V. G. and Gans, J. (1950) A comparison of nutritive values and_ taste thresholds of carbohydrates for the blowfly. Biol. Bull., 99: 446-453. Morita, H. (1969) Electrical signs of taste receptor activity. In “Olfaction and Taste, Vol. 3”. Ed. by C. Pfaffmann, Rockefeller Univ. Press, New York, pp. 370-381. Inoue, T. (1978) Affinity electrophoresis of a- amylase. Physico-Chem. Biol., 22: 135-138. Hansen, K. and Wieczorek, H. (1981) Biochemical aspects of sugar reception in insect. In “Biochemis- try of Taste and Olfaction”. Ed. by R. H. Cagan and M. R. Kare, Academic Press, New York, pp. 139- 162. Kurihara, K., Miyake, M. and Yoshii, K. (1981) Molecular mechanisms of transduction in chem- oreception. In “Biochemistry of Taste and Olfac- tion”. Ed. by R.H.Cagan and M.R. Kare, Academic Press, New York, pp. 249-285. Cagan, R. H. (1981) Recognition of taste stimuli at the initial binding interaction. In “Biochemistry of Taste and Olfaction”. Ed. by R. H. Cagan and M. R. Kare, Academic Press, New York, pp. 175-203. Ninomiya, M., Ozaki, M., Kashihara, Y. and Mori- ta, H. (1986) Destruction and reorganization of the receptor membrane in labellar chemosensory cells of the blowfly: Recovery of response to sugars after destruction. J. Gen. Physiol., 87: 1003-1016. Morita, H. and Shiraishi, A. (1968) Stimulation of the labellar sugar receptor of the fleshfly by mono- and disaccharides. J. Gen. Physiol., 52: 559-583. Ozaki, M., Ninomiya, M., Kashihara, Y. and Mori- ta, H. (1986) Destruction and reorganization of the receptor membrane in labellar chemosensory cells of the blowfly: Long lasting latent action of colchicine. J. Gen. Physiol., 87: 533-549. Hanamori, T., Shiraishi, A., Kijima, H. and Mori- ta, H. (1974) Structure of effective monosaccharides in stimulation of the sugar receptor of the fly. Chem. Senses Flavour, 1: 147-166. ZOOLOGICAL SCIENCE 5: 291-298 (1988) Auditory Evoked Potentials Dynamically Related to Sleep-Waking States in Unrestrained Rats YASUHISA OKANO, Epuarp Davip!, Kazuki HonDA and SHosirRo INOUE? Insitute for Medical and Dental Engineering, Tokyo Medical and Dental University, Kanda-Surugadai 2-3-1, Tokyo 101, Japan and ‘Department of Physiology, University of Witten-Herdecke, D-5804 Herdecke, Federal Republic of Germany ABSTRACT— Auditory evoked potentials (AEPs) to continuous click stimuli delivered at 1-s intervals were bipolarly recorded between the frontal and the fronto-parietal cortex in freely behaving rats throughout the first 3 hours of the light and the dark period. Dynamic changes in the middle and late latency components of AEPs were serially analyzed during slow wave sleep (SWS), paradoxical sleep (PS) and wakefulness (W). The sleep-waking stages affected greatly the latency of the first and second negative (N, and N;) and positive (P; and P,) waves. Especially during SWS, N, and P, dynamically changed: the deeper SWS, as evidenced by an elevated delta activity, was accompanied by the longer latency and the higher amplitude. The peak-to-peak voltage difference was maximal when delta-sleep occurred. During PS, AEPs remained quite stable, exhibiting a steady level of N, and P; amplitudes and no fluctuation of their latencies. During W, N, tended to decrease in amplitude and sometimes © 1988 Zoological Society of Japan disappeared due to habituation to the stimuli. latency and amplitude of the middle and late AEP components. Significant circadian variations were found in the Thus, the state-dependent and time-of-day-dependent characteristics of AEPs might be utilized as a good indicator for sleep-vigilance scoring. INTRODUCTION An electroencephalogram (EEG) is universally adopted as an objective index for scoring the sleep-waking stages, which requires time- consuming analysis and a large storage space. Auditory evoked potentials (AEPs) can also pro- vide an objective evaluation of the sleep-waking state in a more concise form. A definite change in the latency and/or amplitude of AEPs is known to occur during sleep-waking cycles in humans [1- 11], cats [12-16] and rats [17, 18]. In the late components of AEPs, markedly increased ampli- tudes are observable during slow wave sleep (SWS) in animals or non-REM sleep in humans, although the early components equivalent to Accepted August 20, 1987 Received July 18, 1987 2 To whom requests of reprints should be addressed. brainstem evoked responses are considerably stable in humans [6, 19]. However, in these studies, attention was mainly focussed on the analysis and comparison of AEPs at a certain steady state of sleep-waking stages. Since the conscious level of wakefulness (W) and the stages of sleep dynamically change as a function of time, a continuous time-course analysis of AEPs is of special interest. As far as the present authors know, Molnar et al. [15] noted that the peak latency of the latest negative component of AEPs shifts forward during the rapid eye move- ment phase of paradoxical sleep (PS) in cats; whereas Ujszaszi and Halasz [8] observed that the latency and amplitude of late components of AEPs show a considerable variety during stage 2 non- REM sleep in humans. The present paper deals with the first systematic approach to a dynamic aspect of AEP variations, both minute-to-minute and day-to-night, with special reference to the 292 sleep-waking stages, 1.e. SWS, PS and W, in freely behaving rats. Since the rat is a multiphasic sleeper and a night-active animal, approximately two thirds of total sleep time are distributed in the light period under a 12-h light and 12-h dark schedule [20]. The duration of SWS and PS episodes is longer in the light period than in the dark period, while that of W episodes is shorter in the light period [21]. Hence the question arises as to whether circadian variation in the AEP components exist between the early phases of the light and the dark period. The experimental facts dealt with here indicate that this is really the case. The preliminary results are published elsewhere in abstract form [22-24]. MATERIALS AND METHODS Seventeen male rats of the Sprague-Dawley strain, raised in our closed colony on a 12-h light and 12-h dark schedule (light period: 08: 00-20: 00h) under a constant air-conditioned environ- ment of 25+1°C and 60+6% relative humidity with free access to rat chow and water, were used. At the age of 60-90 days, animals weighing 300- 450 g were anesthetized with sodium pentobarbital (50 mg/kg i.p.), placed on a stereotaxic apparatus and permanently implanted with three cortical electrodes for EEG and AEP recording, two nuchal electrodes for electromyogram (EMG) recording, and a silver plate on the skull as a reference electrode. The EEG-AEP electrodes were located on the surface of the frontal cortex (1.8 mm lateral to the central suture and 4.5 mm anterior to the bregma) and of the fronto-parietal cortex (3.7 mm lateral and 1.5 mm posterior, as above). EEG and AEP were bipolarly recorded between the two electrodes. The surgical proce- dure was the same as described in a previous paper [25]. The rats were individually housed in a special cylindrical cage which enabled continuous mon- itoring of EEG and EMG, and continuous audi- tory stimulation (Fig. 1). A slip-ring fixed above the cage guaranteed the free movement of the rats. Each cage was placed in a sound-proof, electro- magnetically shielded chamber under the same environmental conditions as above. A week was 3 Y. Okano, E. Davin et al. EEG |Signal AEP | processor Slip-ring EEG EEG Speaker —- | = Computer Polygraph EMG EMG |system re | Click generator Experimetal system. For details, see text. Sound-proof chamber Fic. 1. allowed for recovery from surgery before the experiment. Then EEG and EMG were poly- graphically recorded, and SWS, PS and W were visually scored according to the routinized criteria [20]. After observing a steady circadian rhythmic- ity in sleep-waking amounts, auditory stimuli consisting of clicks were delivered at 1-s intervals through a speaker placed above the cage. The clicks were 0.1-ms square wave pulses, the inten- sity of which was adjusted to 70 dB above the noise level on the floor of the cage. Click stimuli were continuously given to the rats either between 08: 00 h and 11: 00 h or between 20: 00 h and 23: 00 h. Under continuous recordings of EEG and EMG, AEPs were collected from two out of the three cortical electrodes for averaging. Fifty AEPs were averaged at one time by a signal processor (7T17, NEC San-ei), stored in a floppy disk and simul- taneously recorded on a plotter in the following 10 seconds. Hence each averaged AEP was recorded at l-min intervals. Averaged AEPs were then analyzed with reference to the EEG-EMG defined sleep-waking stages and to the delta activity (0.5- 3.5 Hz) of the EEG records which was filtered and integrated at 1-min intervals. For an analysis of circadian variations, AEPs recorded at the definite occurrence of SWS, PS and W in the 1-h period of either 08: 00-09: 00h or 20: 00-21: 00h were respectively compared and statistically analyzed by Student’s f-test. RESULTS AEP. components Typical AEPs during SWS, PS and W in freely State-dependent Auditory Responses 293 SWws No Ny N3 50 pV P, ae PS LM PY oe Ww 6) 100 200 300 Time (ms) Fic. 2. A typical example of averaged AEPs with the definition of peaks during SWS, PS and W. The arrow indicates click stimuli. behaving rats are shown in Figure 2. The wave- forms were largely in accordance with those described in previous studies [17, 18]. The middle and late components of AEPs were composed of several positive and negative deflections in their waveforms. The peaks were designated as the first negative (N,) and positive (P,) waves, the second negative (N>) and positive (P) waves, and so forth, according to their polarity and their se- quence order. The waveform, the latency of peaks, and the peak-to-peak amplitude changed dynamically and were largely dependent on sleep- waking stages, as described below. AEPs during SWS In the typical waveform of an AEP during SWS, the first peaks, N, and P,, clearly appeared within 30 ms after the onset of click stimuli. The N,-P, amplitude was considerably small, never exceeding 10 nV. Large N> and P> deflections then followed around 50 ms and 120 ms after the onset of click stimuli, respectively. Hence the difference be- tween these peaks became very large, sometimes exceeding 80 «V, which was far greater than that during PS and W. Subsequent peaks (N3, P3, Ny and P,) were observable in a latency range from 110 to 200 ms (Fig. 2). The waveform of AEPs varied dynamically during the course of SWS (Fig. 3, right). Deep SWS, as evidenced by the elevated occurrence of delta activity, was characterized by a prolongation of N> and P; latencies and a profound increase in their amplitude. In contrast, the N> latency and amplitude declined in accordance with the reduction of delta activity (Fig. 3, left). AEPs during PS During PS, definite rises and falls in amplitude occurred three times within 100 ms after click stimuli (N,;—N3 and P,;—P3, Fig. 2), and some rats exhibited 4th deflections (N4 and P,). All peak-to- peak amplitudes were smaller than those of SWS, never exceeding 30 ~V. Wave components were not clearly distinguishable later than 100 ms after stimulation. AEP components were characterized by their stability during the course of PS (Fig. 4). The latency of N> and P> showed little fluctuation. The N>-P> amplitude remained at a steady level. Sometimes, the duration of PS episodes was too short for a time-series analysis of successive AEPs, which were averaged at 1-min intervals. AEPs during W In the waking state, N,; and P, appeared approximately 20 ms and 40 ms after click stimuli, respectively (Fig. 2). nents, N5 and P3, occurred in the following 50 ms. No clear waveform was observable after then. Peak-to-peak amplitudes varied considerably, ranging from 5 to 40 «V. The N> and P> tended to decrease their amplitude and often almost dis- appeared (Fig. 5, right). Television monitoring revealed that, during such a change, rats some- times displayed definite behavior such as eating, drinking and grooming. N> and P; waves usually reappeared during the course of the same W episode. The N, and P, latency was relatively stable (Fig. 5). Subsequent wave compo- Time-of-day-dependent variation in AEPs AEFPs in the early phase of the light period were compared to those of the dark period. Since 294 Y. OKANO, E. Davin et al. rol AMP (mv) , (pV) 01 760 LATENCY (ms) 40 60 80 100 TIME (min) Fic. 3. A typical example of dynamic changes in the amplitude and latency of N, component of AEPs during SWS. The corresponding AEPs are shown at the right side. AMP (ordinate) means the difference between the reference N, amplitude at the initial SWS (indicated by an open circle) and that of the other AEPs. Increments and decrements in absolute values are expressed by arrows directed upward and downward, respectively. Sleep-waking stages are shown by an oblique column, in which SWS, PS (shown only in Fig. 4) and W are represented by black, dotted and white sections, respectively. The initial and final time of day is indicated for the main episodes. Integrated delta activity (6 ) is shown at the left side. LATENCY (ms) Click TIME ( min) Fic. 4. A typical example of dynamic changes in the amplitude and latency of N, component of AEPs during PS. AMP (ordinate) means the difference between the reference N> amplitude at the initial PS (indicated by an open circle) and that of the other AEPs. For further explanations, see the legend of Fig. 3. N \o nN State-dependent Auditory Responses LATENCY (ms) 40 60 80 100 $e TIME ( min) Fic.5. A typical example of dynamic changes in the amplitude and latency of the P5 component of AEPs during W. AMP (ordinate) means the difference between the reference P, amplitude at the initial W (indicated by open circle) and that of the other AEPs. For further explanations, see the legend of Fig. 3. PEAK-TO-PEAK AMPLITUDE (pV) : 1 ——— 0) 40 80 120 160 0 40 80 120 160 LATENCY (ms) LATENCY (ms) Fic. 6. The peak-to-peak amplitude and latency of AEP components as a function of the light-dark period and sleep-waking stages in two different rats A and B. Values are mean+SEM during the early light period (08: 00-09: 00h, A: n=10 for SWS; n=3 for PS; n=24 for W, B: n=18 for SWS; n=4 for PS; n=7 for W) and the early dark period (20: 00-21: 00h, A: n=16 for SWS; n=3 for PS; n=18 for W, B: n=11 for SWS; n=3 for PS; n=15 for W). Circles, squares and triangles indicate respectively SWS, PS and W, in which white and black signs mean respectively the light and the dark period. Peak-to-peak amplitudes are expressed in absolute values and arrranged in sequence from left to right: the N,-P,; amplitude, the P,-N, amplitude, the N,-P, amplitude and the P,;-N; amplitude (shown only for PS in rat A). Asterisks indicate that the peak-to-peak amplitude difference between the light and the dark period was statistically significant at P<0.05 (*) and P<0.01 (**). A symbol (f) indicates that the latency difference between the light and the dark period was statistically significant at P<0.05S. 296 Y. OKANO, E. Davin et al. individual variations were so large, no clear difference was found in average values from the results pooled for all animals. However, if data were individually processed for all AEPs during each state, a significant circadian variation was obtained from most of the rats. Figure 6 illustrates typical examples, in which the peak-to-peak ampli- tude and latency of AEP components showed significant differences between the light and the dark period. During SWS, the differences in each peak-to- peak amplitude were apparent. In most cases, the P,—-N> amplitude significantly differed between the two periods. The voltage difference ranged from 3 to 10 nV. The later components showed usually a larger difference which was insignificant due to considerable variations. The latency of Ps and N3 during nocturnal SWS was largely delayed by 5-20 ms in comparison with that during diurnal SWS. However, no statistical significance was detected because of large variations. During PS, the circadian difference in the AEP parameters was rather small and statistically insig- nificant except for that of some components (for examples, Fig. 6A: the N,-P, amplitude signif- icantly differed by 14 uV; Fig. 6B: the N, latency significantly differed by 8 ms). During W, the peak-to-peak amplitude of most AEP components significantly differed between the light and dark periods. The difference ranged from 2 to 15 nV. However, their latency exhibited little difference. DISCUSSION The middle and late components of AEPs in freely behaving rats exhibited state-dependent changes. variations occurred in the waveform of AEPs It was found that minute-to-minute during the course of W. In addition, our time series analysis first demonstrated that the latency and amplitude of the AEP components, especially during SWS, varied dynamically with close relation to the fluctuations in the EEG delta activity. Since the delta activity is regarded as a reliable indicator of deep sleep, it is likely that the AEP parameters might specifically indicate the time-course changes in the state of sleep-wakefulness. The AEP activity in our rats was most promi- nent during the deep sleep stage. This was comparable to the previous reports in which auditory stimuli of a low frequency, up to 10 Hz, were given to cats [12-16] and rats [17,18]. It seems likely that in these animals, auditory inputs at a relatively low frequency may easily provoke a larger amplitude and a longer latency in the middle and late AEP components of AEPs during SWS than during PS and W. The mechanism involved in this change remains unknown. There are several speculations. Firstly, Weitzman and Kremen [8] concluded that AEPs during sleep represents summed K complexes elicited from the auditory stimulation. Secondly, the chronic twitches of the middle ear muscles during PS are responsible for a reduction of auditory inputs, which eventually causes a reduc- tion in the amplitude of AEPs [12, 13]. This assumption, however, is not confirmed by a later study [17]. Thirdly, a state-dependent change in body temperature may be considered. It is re- ported that heating and cooling of the body can respectively shorten and enlarge the latency and amplitude of auditory brainstem responses [26]. Since the early phase of an SWS episode generally accompanies a fall in body temperature [27], the slower middle and late latencies of AEPs during SWS might be accounted for by the lowered temperature, which accompanies a delayed synap- tic transmission. Finally, the blockade of sensory information in neural circuits during W may be considered. Attention causes changes in the latency of late components of AEPs [28]. This was largely due to the habituation provoked by the inhibitory mechanism or the gating to the monoto- nous stimulation. Therefore, changes in attentive levels might reflect the amplitude fluctuations. In contrast, during SWS no attentive activity exists and the slow wave generator mechanism is pre- dominant. This may result in an enlarged AEP waveform. At present no information is available to determine which possibility is most plausible. In human studies, however, the results are conflicting; some investigators report a similar increase in peak amplitudes and a prolongation of their latencies in deep sleep stages [6, 8-11], whereas others report little changes depending on State-dependent Auditory Responses 297 the sleep-waking stages (see [29]); most authors refer to a reduced AEP activity during PS, while Ornitz et al. [5] demonstrated an enlarged wave- form during PS. Furthermore, if auditory stimuli are given at sufficiently high frequency, up to 60 Hz, the amplitude changes are smaller during sleep than during waking [3, 4]. The present study first detected the diurnal and nocturnal differences in the amplitude and latency of the middle and late AEP components. In this connection, Hanada and Kawamura [30] noted that a clear circadian variation occurs in the amplitude of an early component of evoked potentials caused by electrical stimulation of the optic tract in rats. These facts may indicate the existence of a time-of-day-dependent change in the sleep-waking state. However, the existence of large individual variations in these AEP para- meters may indicate that the differences between the light and dark periods reflected not only the rhythmicity derived from the circadian oscillator but also the variations in the behavioral situations in each rat, since an attentive behavior during W or an elevated delta activity during SWS might easily modify the AEP parameters. Apart from the above discussions, the close correlation between the AEP parameters and the sleep-waking stages, especially during SWS, may suggest that the time-consuming sleep scoring based on polysomnography could be replaced or supplemented by the time-course analysis of AEPs. The latter technique seems to be simpler, more conventional and easier to define the dynam- ic state changes. Hence the AEP dynamics could be applied not only to the evaluation of sleepiness [31] but also to an objective sleep-vigilance scoring. ACKNOWLEDGMENTS E. David joined this study during his stay in Tokyo, which was sponsored by the Japan Society for the Promotion of Science, Tokyo, and the Deutscher Akademischer Austauschdienst, Bonn. S. Inoué partial- ly prepared the manuscript during his stay in Herdecke, which was sponsored by the Alexander von Humboldt- Stiftung, Bonn. Both authors wish to express their gratitude for the sponsorship. This research was sup- ported in part by Grant-in-Aid for Special Project Research of Endogenous Neuroactive Substances from the Ministry of Education, Science and Culture of Japan (No. 60126002) to S. I. REFERENCES 1 Buchsbaum, M., Gillin, J.C. and Pfefferbaum, A. (1975) Effect of sleep stage and stimulus intensity on auditory evoked responses. Psychophysiology, 12: 707-712. Erwin, R. and Buchwald, J. S. (1986) Midlatency auditory evoked responses: differential effects of sleep in the human. Electroencephalogr. Clin. Neurophysiol., 65: 383-392. 3 Linden, R.D., Campbell, K.B., Hamel, G. and Picton, T. W. (1985) Human auditory steady state evoked potentials during sleep. Ear Hearing, 6: 167-174. 4 Linden, R. D., Picton, T. W., Campbell, K. B. and Hamel, G. (1985) Thresholds for auditory steady state evoked potentials during sleep. In “Sleep ’84”. Ed. by W.P. Koella, E. Ruther and H. Schulz, Gustav Fischer Verlag, Stuttgart, pp. 228-231. 5 Ornitz, E.M., Ritvo, E. R., Carr, E. M., Paniman, L.M. and Walter, R. D. (1967) The variablity of the auditory evoked response during sleep and dreaming in children and adults. Electroencepha- logr. Clin. Neurophysiol., 22: 514-524. 6 Osterhammel, P. A., Shallop, J. K. and Terkildsen, K. (1985) The effect of sleep on the auditory brainstem response (ABR) and the middle latency response (MLR). Scand. Audiol., 14: 47-50. 7 Ujszaszi, J. and Halasz, P. (1986) Late component variants of single auditory evoked responses during NREM sleep stage 2 in man. Electroencephalogr. Clin. Neurophysiol., 64: 260-268. 8 Weitzam, E. D. and Kremen, H. (1965) Auditory evoked responses during different stages of sleep in man. Electroencephalogr. Clin. Neurophysiol., 18: 65-70. 9 Williams, H.L., Marlock,H.C., Marlock, J. V. and Lubin, A. (1964) Auditory evoked responses and EEG stages of sleep. Ann. New York Acad. Sci., 112: 172-179. 10 Willams, H. L., Tepas,D.1I. and Marlock, H. C., Jr. (1962) Evoked responses to clicks and elec- troencephalographic stages of sleep in man. Science, 138: 685-686. 11 Yokoyama, T., Ryu, H., Uemura, K., Miyamoto, T., Yamamoto, T., Furuya, Y. and Nakajima, S. (1986) Middle latency auditory evoked potentials in awake and asleep. Rinsho Noha, 28: 9-44. 12. Baust, W., Berlucchi,G. and Moruzzi,G. (1964) Changes in the auditory input in wakefulness and during the synchronized and desynchronized stages of sleep. Arch. Ital. Biol., 102: 657-674. in) 16 20 21 298 Berlucchi,G., Munson,J.B. and _ Rizzolatti, G. (1967) Changes in click-evoked responses in the auditory system and the cerebellum of free-moving cats during sleep and waking. Arch. Ital. Biol., 105: 118-135. Chen, B. M. and Buchwald, J. S. (1986) Midlatency anditory evoked responses: differential effects of sleep in the cat. Electroencephalogr. Clin. Neurophysiol., 65: 373-382. Molnar, M., Kormos,G. and Csépe, V. (1986) Laminar analysis of intracortical auditory evoked potentials during the wakefulness-sleep cycle in the cat. Int. J. Psychophysiol., 3: 171-182. Winters, W. D. (1964) Comparison of the average cortical and subcortical evoked response to clicks during various stages of wakefulness, slow wave sleep and rhombencephalic sleep. Electroencepha- logr. Clin. Neurophysiol., 17: 234-245. Hall, R. D. and Borbély, A. A. (1970) Acoustically evoked potentials in the rat during sleep and waking. Exp. Brain Res., 11: 93-110. Knight, R. T., Brailosky, S., Scabini, D. and Simp- son, G. V. (1985) Surface auditory evoked poten- tials in the unrestrained rat: component definition. Electroencephalogr. Clin. Neurophysiol., 61: 430- 439. Campbell, K. B. and Bartoli, E. A. (1986) Human auditory evoked potentials during natural sleep: the early components. Electroencephalogr. Clin. Neu- rophysiol., 65: 142-149. Honda, K. and Inoué, S. (1981) Effects of sleep- promoting substance on sleep-waking patterns of male rats. Reports Inst. Med. Dent. Eng., 15: 115- 123. Honda, K., Komoda, Y., Nishida, S., Nagasaki, H., Higashi, A., Uchizono, K. and Inoué,S. (1984) Uridine as an active component of sleep-promoting 23 PH) 28 29 30 31 Y. OKANO, E. Davin et al. substance: its effects on the nocturnal sleep in rats. Neurosci. Res., 1: 243-252. David, E., Inoué,S., Okano, Y., Honda, K. and Pfotenhauer, M. (1986) Click-evozierte Potentiale im Schlaf- und Wachzustand; ein Vergleich zwischen Ratte und Mensch. Biomed. Tech., 31: suppl. 146. Inoué, S., Okano, Y., Honda, K. and David, E. (1986) Auditory evoked potentials specific to sleep- waking stages in unrestrained rats. Abstracts of 3rd Int. Evoked Potentials Symp., p. 125. Okano, Y., Honda, K., David, E. and Inoué, S. (1986) Cortical auditory evoked potentials specific to the sleep-waking stages in unrestrained rats. Abstracts of Ist Int. Congr. Neuroethol., p. 71. Honda, K. and Inoué, S. (1978) Establishment of a bioassay method for the sleep-promoting substance. Reports Inst. Med. Dent. Eng., 12: 81-85. Marsch, R. R., Yamane, H. and Potsic, W. P. (1984) Auditory brain-stem response and temperature: relationship in the guinea pig. Electroencephalogr. Clin. Neurophysiol., 57: 289-293. Obal, F., Jr. (1984) Thermoregulation and sleep. Exp. Brain Res., Suppl. 8: 157-172. Polich, J. (1986) Attention, probability, and task demands as determinants of P300 latency from auditory stimuli. Electroencephalogr. Clin. Neuro- physiol., 63: 251-259. Mendel, M. I. (1980) Clinical use of primary cortical responses. Audiology, 19: 1-15. Hanada, Y. and Kawamura, H. (1984) Circadian rhythms in synaptic excitability of the dorsal lateral geniculate nucleus in the rat. Int. J. Neurosci., 22: 253-262. Broughton, R. (1982) Performance and evoked potential measures of various states of daytime sleepiness. Sleep, 5: 135-146. ZOOLOGICAL SCIENCE 5: 299-309 (1988) The Anatomy and Physiology of the Stomatogastric Nervous System of Squilla. Il. The Cardiac System KENRO TAZAKI Biological Laboratory, Nara University of Education, Takabatake, Nara 630, Japan ABSTRACT—A subsystem of the stomatogastric ganglion (STG) of the mantis shrimp, Squilla oratoria, which controls movements of the cardiac stomach was studied. Muscles of the cardiac stomach and their motor nerves leaving the STG are described. The paired anterior ventricular nerves and paired lateral ventricular nerves (lvn), which exit the STG, contain the motor axons that control muscular contractions of the cardiac stomach. One pair of superior oesophageal nerves (son) also carry motor axons. The lvn and son give rise to several peripheral nerves which innervate 8 identified muscles. Spontaneous motor activity recorded from the lvn and son (termed the cardiac cycle) has a long cycle period (ca. 10 sec) and consists of alternate firing of bursts by two motor neuron groups. Cardiac cycles similar to spontaneous ones are triggered by stimulation of afferent fibers entering the commissural ganglia. Four constrictor and 3 dilator motor neurons in the STG are identified. They are involved in sequential movements of the dorsal and lateral gastric walls of the cardiac stomach. The function of the cardiac cycle is described with reference to the maceration and transfer of food. It is compared to that of the posterior cardiac plate and pyloric cycle (pcp-pyloric cycle) which has been described previously. Comparisons are made of the cardiac systems in STG of stomatopods and © 1988 Zoological Society of Japan decapods. INTRODUCTION The stomatogastric ganglion (STG) of decapods, a small, semi-autonomous nervous system, has been extensively studied as a model for the generation of patterned motor outputs [1, 2]. In the STG both individual cellular properties and neural networks can be analyzed. The lobster STG contains just over 25 identifiable motor neurons which innervate identified muscles of the cardiac and pyloric stomachs [3-5]. It produces cyclic patterned motor outputs having two distinct rhythms responsible for the sequential muscular contractions of the gastric and pyloric regions of the stomach, respectively [6, 7]. The gastric and pyloric cycles are modulated by the supra- oesophageal and commissural ganglia (CG) [8- 10]. The cardiac dilator and constrictor neurons are identified in the lobster STG and oesophageal ganglion (OG), and their motor activity modulates Accepted August 26, 1987 Received May 22, 1987 the pyloric cycle or coordinates it with the gastric cycle [11, 12]. The stomach in stomatopods is subdivided into the cardiac stomach, the posterior cardiac plate (pep) and the pyloric stomach [14, 15]. The cardiac stomach is not equipped with a chewing apparatus such as the gastric mill ossicles of decapods. Most of the ossicles of the stomach form the pcp and pyloric systems of sieves and channels through which digested food moves. The muscles in the stomach and their sequential move- ments have been described by Kunze [15]. The anatomy and physiology of the stomatogas- tric nervous system of Squilla have been found to be similar to those of decapods [16]. Its gross anatomy was first described by Police [13]. The stomatogastric nervous system in stomatopods is basically similar to that in decapods. While the STG in decapods sends several unpaired and paired motor nerves to control the gastric mill and pyloric stomach, the paired lateral ventricular nerves (lvn) which leave the STG in stomatopods carry principal motor axons to innervate the 300 K. TAZAKI muscles of the pep and pyloric stomach. Charac- teristics of the cyclic motor outputs observed in the lvn have been described previously in relation to the channel-opening function of the pcp and pyloric systems. The patterned motor outputs have been termed the pcp-pyloric cycle. The pep-pyloric cycle of stomatopods is apparently homologous to the pyloric cycle in decapods although the neural circuits for its pattern genera- tion remain to be analyzed. The pcp-pyloric cycle is also modulated by input fibers of the superior and inferior oesophageal nerves (son, ion). Car- diac motor activity controlling movements of the cardiac stomach is rarely observed to occur spon- taneously in the semi-isolated preparation, but can be triggered or primed by stimulation of the son input fibers. The present paper deals with the anatomy and physiology of the cardiac system of the stomato- gastric nervous system of the mantis shrimp, Squilla oratoria. Muscles of the cardiac stomach and their motor nerves are described. The cyclic motor activity of the STG is shown to explain the sequential muscular contractions which control the functional movements of the cardiac stomach. Relations between the cardiac and pcp-pyloric systems primed by stimulation of the son are analyzed. MATERIALS AND METHODS The stomatogastric nervous system of a stomato- pod, Squilla oratoria, was employed in this study. Animals were held in tanks of recirculated sea water at the temperature of 20°C. About 200 specimens were used for anatomical and physio- logical observations. Methods of dissection and recording were the same as those described in the previous paper [16]. Cardiac stomach The cardiac stomach, which is a large cuticular sac serving to store ingested food, contains two pairs of ossicles [15]. The anteroventral cardiac ossicles (avc) lie along the anterior lateral margins of the ventral gastric wall (Fig. 1A). They act as supports for muscle attachment. The small pos- terior lateral cardiac ossicles lie in the posterior cardiac stomach Icm1 Fic. 1. A: Diagrammatic lateral view of musculature in the cardiac stomach. avc, anteroventral cardiac ossicle: pep, posterior cardiac plate: dacm, dorsal anterior cardiac muscle: dcm, dorsal cardiac mus- cle: dicm1 and dlem2, dorsolateral cardiac muscle: Icm1-Icm3, longitudinal cardiac muscle: pcm, pos- terior cardiac muscle: plem, posterior lateral car- diac muscle. B: Diagrammatic lateral view of the stoma- togastric nervous system. The diagram shows peripheral nerve courses of the lvn and son. CG, commissural ganglion: OG, oesophageal ganglion: STG, stomatogastric ganglion: coc, circum- oesophageal connective: com, commissure: ion, in- ferior oesophageal nerve: ivn, inferior ventricular nerve: mbn, mandibular nerve: son, superior oesophageal nerve: stn, stomatogastric nerve: avn, anterior ventricular nerve: lvn, lateral ventricular nerve: d-Ivn, dorsal lateral ventricular nerve: m- lvn, median lateral ventricular nerve: v—lvn, ventral lateral ventricular nerve: pln, posterior lateral nerve: cnl—cn4, cardiac constrictor nerve: dnl- dn4, cardiac dilator nerve. lateral gastric wall anterior to the pcp (not shown), though their function is unknown. These two ossicles do not play a role of masticating ingested food. Kunze [15] has classified the muscles in the cardiac stomach into three groups: cardiac mus- Stomatogastric Nervous System. II 301 cles, longitudinal muscles and cardiac floor mus- cles. All of these muscles have been named, and the sequence of movements of the cardiac stomach after ingestion has been described. However, the functional anatomy of the muscles of the cardiac stomach has not been described in detail, and their origins and insertions remain obscure. The stom- ach muscles are divided into two groups: extrinsic muscles originating on the inner side of the exoskeleton and inserting on the stomach wall or ossicles, and intrinsic muscles having both attach- ments on the stomach wall itself [5]. Contractions of these individual muscles constrict or dilate different regions of the gastric wall. In this paper, the muscles in the cardiac stomach, which are innervated by motor nerves leaving the STG, have been identified (see Anatomy section). Preparation Two types of preparations were used in this study. The semi-intact preparation was made in order to record the spontaneous cardiac motor activity. After cutting away most of the append- ages in the cephalothoracic region, the animal was immobilized, ventral side up, in the preparation box filled with saline. The ventral and lateral regions of carapace anterior to the mandibles were removed by cutting the extrinsic muscles thus exposing the cardiac stomach and the stomatogas- tric nervous system. The cardiac stomach moved rhythmically for two hours or more under these conditions. The lvn and son, which include motor axons controlling movements of the cardiac stom- ach, were drawn into suction electrodes. This preparation was also used to observe how the extrinsic and intrinsic muscles move the gastric wall. The semi-isolated preparation, which consists of the gastric wall with the muscles, STG, OG and associated nerves, was employed to observe how motor nerves innervate identified muscles. Iden- tification of the cardiac bursting units of the peripheral nerves was also made in this prepara- tion: the son input fibers were stimulated by brief pulses with moderate intensity and frequency to activate rhythmic motor outputs from the STG. The semi-isolated preparation, including the CG, was used to study stimulus-induced cardiac motor activity. The son input fibers were activated by stimulation of the afferent fibers of a mandibular nerve which enter the CG. The saline used here was modified from that for Squilla developed by Watanabe et al. [17]. It had the following composition (in mmol/l): Na, 450; K, 15; Ca, 10; Mg, 20; Cl, 525; N-2-hydroxyethyl- piperazine-N’-2-ethanesulfonic acid (HEPES), 2. The saline was adjusted to pH7.6. All experi- ments were done at room temperature (20-26°C). RESULTS Anatomy Muscles The muscles of the cardiac stom- ach are shown diagrammatically in Figure 1A. Some of these muscles have been named by Kunze [15]. Some have not been previously identified, and are described later. All of these muscles are bilaterally symmetrical. Four cardiac muscles are associated with medial movements of the dorsolateral gastric wall. The (Icm1) attaches to the dorsal gastric wall near the pcp and intrinsic longitudinal cardiac muscle 1 runs to the anterior region of the cardiac stomach. It serves to constrict the dorsolateral gastric wall. Two extrinsic muscles, which are antagonists of the Icm1, are the dorsal cardiac muscle (dem) and the dorsolateral cardiac muscle | (dlcm1). The dem originates on the dorsal carapace and inserts on the dorsal gastric wall. The dlcm1 originates on the lateral carapace and inserts on the dorsolateral gastric wall ventral to the Iem1l. They dilate the gastric wall. The unpaired dorsal anterior cardiac muscle (dacm) is also an antagonist of the Icm1. The dacm originates on the anterior ventral carapace, and inserts on the anterior extremity of the cardiac stomach. Its contraction pulls the anterior gastric wall forward. Six cardiac muscles are associated with medial movements of the lateral gastric wall which is medially folded to form the lateral cardiac fold (Icf). The intrinsic longitudinal muscles 2 and 3 (Icm2, Icm3) attach to the Icf. The Icm2 originates on the gastric wall at the lateral side of the pcp, and runs anteriorly. The Icm3 originates on the gastric wall at the ventral side of the pcp, and runs 302 K. TAzakI anteriorly parallel to the lem2. The Iem2 and Icm3 move the lateral gastric wall medially. The antagonists of the lcm2 and Icm3 are the dorso- lateral cardiac muscles 1 and 2 (dlcm1, dlcm2), respectively. The extrinsic dlcm2 originates on the lateral carapace where the dlcm1 attaches, and inserts on the ventrolateral gastric wall. The lcm2 and Icm3 are situated between the dicm1 and dicm2. The intrinsic posterior cardiac muscle (pem) attaches to the lateral gastric wall at the ventral side of the pcp. The pcm serves to constrict the posterior lateral gastric wall in front of the pcp. The posterior lateral cardiac muscle (plem) origi- nates on the posterior lateral carapace, and inserts on the lateral gastric wall near the pcp. The plcm is an antagonist of the pcm. Kunze [15] has described dacm, dem, dlem, pcm and plem but not Icm1, Icm2 and Icm3. Besides these identified cardiac muscles, 2 intrinsic and 5 extrinsic muscles insert on the ventral gastric wall (not shown). Innervation The stomatogastric nervous system and peripheral motor nerves are illustrated in Figure 1B. Most of the nerves are bilaterally paired. The anterior ventricular nerve (avn) leaves the STG a short distance posteriorly, carrying motor axons to the dacm. The lateral ventricular nerve (lvn) leaves the anterior end of the STG, and runs posteriorly along the Icm1 on the gastric wall, branching extensively at the region just anterior to the pcp. The dorsal, median and ventral lvn (d—lvn, m-lvn, v—lvn) supplies motor nerves to the muscles of the pep and pyloric stomach [16]. The posterior lateral nerve (pln) connects the lvn with the son emerging from the CG which resides on the circumoesophageal connective (coc). The lvn and son divide peripherally into two functional groups of motor nerves: cardiac con- strictor nerves (cn) and cardiac dilator nerves (dn). Two branches emerging from the lvn run dorsally to innervate the anterior and posterior dcm. These nerves are named the dnl because they innervate the same muscle. The cnl1 branches from the lvn anterior to the pcp to innervate the lcm1. The pin gives rise to several branches. The dn2 is a long branch, traveling anteriorly to innervate the dicm1. The cn2 and cn3 are short branches carrying motor axons to the Icm2 and Icm3, respectively. The dn4 is a small branch which innervates the plem. The son leaves the stomato- gastric nerve (stn) posterior to the STG, and runs within the dlcm2, sending out the dn3 to it. The cn4 branches from the son near the pln to innervate the pcm. The OG sends out several nerves. The ion connects the OG with the CG, and it branches These branches innervate intrinsic and extrinsic muscles which insert on the ventral gastric wall. They have not been examined in this study which is concerned with the motor activity of the STG relevant to the control of the muscles in the cardiac stomach. extensively on the ventral gastric wall. Physiology Cardiac cycle The dorsal and lateral gas- tric walls are moved by the sequential contractions of identified muscles. The patterned outputs that control movements of the cardiac stomach are provided by the motor axons which run in the avn, lvn and son. Recordings of spontaneous motor activity were made simultaneously from the Ivn and son in semi-intact preparations. Movements of the cardiac stomach were observed by monitor- ing the motor activity. Spontaneous firing patterns of STG motor activity recorded from the lvn and son are shown in Figure 2. The cyclic motor outputs consisted of long-lasting bursts of alter- nately firing constrictor and dilator units (Fig. 2A). Bursts repeated at variable frequencies averaging about 5/min. The duration of bursts in the two units ranged from 3 to 15sec. The impulse frequency in the constrictor units ranged from 30 to 120 Hz, and in the dilator units from 20 to 80 Hz. The initial units in the bursts leading to constriction of the gastric wall occurred simul- taneously in the lvn and son. They were followed by dilator impulses. These patterned outputs are termed the cardiac cycle. There were variations of the motor activity of two units in these output nerves. In Figure 2B-D alternate silent periods in dilator and constrictor units are observed in the lvn or son. Such variations of the motor pattern seemed to be related to functional movements of the cardiac stomach (see later). Spontaneous cardiac cycles were rarely seen even in the semi- isolated preparation which included the CG and Stomatogastric Nervous System. II 303 2° sec Fic. 2. Cyclic motor activity of the STG related to movements of the cardiac stomach. Spontaneous bursts were simultaneously recorded from the lvn and son. Initial bursts occurred in constrictor units, and were followed by those of dilator units. A: All constrictor and dilator burst units fired alternately. B-D: Either dilator or constrictor units were silent. In B and C displacements of the lvn trace are due to muscular contractions. OG as well as the STG with associated nerves (similar to a combined preparation in the lobster: [1]). The avn was too small to permit monitoring impulses from it, but contraction of the dacm could be seen during bursting of dilator units. The bursting pattern contributes to the sequen- tial muscular contractions of the cardiac stomach. The burst units of the lvn and son observed by en passant recording correlate with those of their peripheral nerves invading various identified mus- cles. An example of such recordings is shown in Figure 3. The cardiac constrictor and dilator units are designated as CA and CD, respectively. The bursts of the Ivn consist of two constrictor units. The bursting units of CAl and CA2 are prop- agated to the cnl and cn2, respectively (Fig. 3-A1, 2). Identification of these two units can be made by using the peripheral recording electrode for electrical stimulation of the axon to elicit an antidromic impulse in the lvn. The son also contains motor axons of two constrictor neurons. CD3 200 msec Fic. 3. Bursting units in peripheral nerves correspond- ing to those of the lvn or son. En passant record- ings were made from the lvn or son at the region anterior to its ramification. The peripheral bursting units were recorded from nerve branches invading identified muscles. A: Burst firing of cardiac con- strictor neurons (CA1-CA4). B: Burst firing of cardiac dilator neurons (CD1—CD3) (see text). The long-lasting bursting unit of CA3 travels to the cn3 (Fig. 3-A3), and the short bursting unit of CA4 to the cn4 (Fig. 3-A4). The CA2 and CA3 neurons appear to discharge for a longer period than the CAI and CA4 neurons. For the dilator units, the burst of large impulses in the Ivn, which is attributed to CD1, is propagated to the dnl (Fig. 3-Bl). The dn2 contains two motor axons: one originates from the CD1 neuron, and the other from the CD2 neuron (Fig. 3—B2). Large impulses of CD2 unit in the son travel to the dn3 (Fig. 3-B3). The burst of small impulses of CD3 unit is propagated from the son to the dn4 (Fig. 3-B4). The size of impulses of the cardiac constrictor and dilator neurons extracellularly recorded from the lvn and son is relatively consistent: CD1 and CD2 are larger than CA, and CD3 is the smallest (Fig. 304 K. TAZAKI 3-B1, 2). The bursting patterns of motor outputs such as shown in Figure 2 are associated with movements of the cardiac stomach. The dorsolateral gastric wall is moved medially by contractions of the Icm1 and Icm2. The corresponding motor pattern is shown in Figure 2D. Two constrictor (CA1, CA2) and 2 dilator (CD1, CD2) neurons contribute to this movement. Medial movements of the poste- rior lateral gastric wall in front of the pcp are commanded by the CA3 and CA4 neurons, each causing contractions of the Ilem3 and pcm. The motor pattern is shown in Figure 2C. During these two types of movements, ingested foods may be moved backward or forward in the cardiac stom- ach. Besides these movements, the dorsolateral and posterior lateral gastric walls are moved medially. The motor patterns shown in Figure 2A and B control these movements. The CA1 and CA2 neurons command constriction of the dorso- lateral gastric wall, while the CA3 and CA4 neurons command constrictions of the lateral and posterior lateral gastric walls. During sequences of all these types of movements, ingested foods would be macerated into fine particles. Other types of patterned motor outputs are shown in Figure 5. These output patterns may control the pumping of digestive juices or the transfer of digested food between the cardiac and pyloric stomachs (see later). Table 1 summarizes the present observations on motor neurons, location of axon, muscles inner- vated and their functions. Although the precise number of neurons in the STG could not be determined, there seemed to be at least 7 motor neurons involved in the generation of the cardiac cycle. Priming effect of input fibers via CG The son inputs induced dual effects on the motor activity of the STG: one input accelerated the pep-pyloric cycle, and the other activated the cardiac cycle [16]. Cardiac cycles such as those seen during spontaneous activity (Fig. 1A) could be activated in the semi-isolated preparation by stimulating afferent fibers of the mandibular nerve with moderate intensity at 50 Hz, the highest frequency tested in this series of experiments (Fig. 4). Such stimulation activated the son input fibers via the CG. The pcep-pyloric cycle, which consisted of pyloric (PY) and pyloric dilator (PD) bursting units, repeated at a rate of about 2 Hz before stimulation. It disappeared gradually during stim- ulation, while the long-lasting cardiac cycle units were activated. The size of the PD and CD1 impulses is almost the same in this record. They can be distinguished by burst firing in the son which occurs simultaneously with the PD burst (see squares in Fig. 4). Initially, repetitive firing of the CD1 and CD2 neurons occurred in the Ivn and son. It was followed by bursting of the CA1 and CA3 neurons. Bursting of the PY units in the lvn appeared to be accelerated during the high- frequency discharges of the CD1 neuron, and inhibited during the activity of the CA neurons TaBLE 1. Motor neurons of the cardiac cycle in the STG Neuron en mente Funetion Cardiac constrictor constricts the gastric wall CAI lvn, cnl Icm1 CA2 Ivn, cn2 Icm2 CA3 son, cn3 Icem3 CA4 son, cn4 pem Cardiac dilator dilates the gastric wall CD1 lvn, dnl dem Ivn, dn2 dicm1 CD2 son, dn2 dicm1 son, dn3 dicm2 CD3 son, dn4 plem Stomatogastric Nervous System. II 305 PY PD Py CDI CAI Fic. 4. Cardiac cycle triggered by son input fibers activated by repetitive stimulation of afferent fibers of the mandibular nerve entering the CG with moderate intensity at 50 Hz (marked by the underline). Records A-B are continuous recordings. A long-lasting cardiac cycle was activated during the stimulation. The CD1 and CD2 neurons fired initially, and thereafter the CAl and CA3 neurons fired. PY, pyloric units: PD, pyloric dilator units. Bursts of PD occurred simultaneously with those in the son (see squares). A (Fig. 4A). Bursting of the CA neuron in the son was preceded by that in the lvn. Both continued after the cessation of stimulation (Fig. 4B). The priming effect was dependent on the stimulus frequency: the cardiac cycle became more obvious ay with increasing frequency. This observation sug- gests that some inputs from the CG may be B necessary for activation of the cardiac cycle pattern ‘ generator. Wn Interaction of cardiac and __ pcp-pyloric cycles The function of the cardiac stomach is CHR aS to macerate food and to transfer digested particles | . to the pyloric stomach. The pcp-pyloric cycle Samant: the opening and ae ean C \ mae we ion | po 3 in the pcp and pyloric stomach [16]. Functional Ivn cena meres — tis interactions between the cardiac and pcp-pyloric Dey cycles were studied in the semi-isolated prepara- tion. Modulation of the cardiac cycle by the pep-pyloric cycle units was examined after cycling 300 msec had been triggered by stimulation of son input — Fig. 5. Modulation of cardiac cycle units by pcep- fibers. An example is shown in Figure 5. Bursting pyloric cycle units. A, B: Bursts of CAl and CA4 of the CA1 and CA4 neurons was suppressed by neurons were suppressed during bursting of PY bursting of the PY neurons (Fig. 5A, B). The CA neuron. C: Bursts of PCP and PD neurons sup- <= pressed bursting of CAl and CA2 neurons. Burst- and PY neurons fired alternately. The PY neurons ee: ae ‘ ‘ ing of CA3 neuron continued during bursting of provide the motor commands for movements of PCP and PD neurons. Dots indicate individual the ampulla in the pyloric stomach which forms the stimuli given to the son. 306 K. TazakI channels leading to the midgut [16]. The CAI and CA4 neurons control contractions of the Icm1 and pcm, respectively. The results, therefore, indicate that constrictions of the dorsolateral and posterior lateral gastric walls near the pcp may not be occurring during constriction of the ampulla. In Figure 5C bursting of the CAl and CA2 neurons recorded from the lvn was suppressed by that of the pcp neurons | and 2 (PCP1, PCP2) and PD neurons in the pcp-pyloric cycle. The PCP and PD neurons provide the motor commands for opening the pcp and cardio-pyloric channels. Bursting of the CA3 neuron, as recorded from the son, continued without inhibition from the PCP and PD neurons. The observations suggest that the CA neurons, which command constriction of the dorsolateral gastric wall of the cardiac stomach, may be inactivated during the opening of the pcp and pyloric channels to allow the transfer of suspensions of food particles from the cardiac to the pyloric stomach. DISCUSSION The stomatogastric nervous system in decapods has been widely used for research on the genera- tion of rhythmic motor output pattern [1, 2]. We have introduced the stomatopod preparation, which lacks the gastric system found in the STG of decapods, to comparative neurophysiology of this nervous system [16]. In the previous paper, the pep-pyloric system of Squilla was analyzed for comparison with the pyloric system in decapods, Panulirus and Homarus. In the present paper, the anatomy and physiology of the cardiac system in the STG of Squilla has been described. The cardiac ossicles located on the cardiac sac are not a chewing system like gastric mill ossicles in dec- apods [14, 15]. The food, partially masticated by the mandibles, is macerated by muscular contrac- tions of the cardiac stomach with the aid of digestive juices that are pumped forward by the pep and pyloric stomach [15, 16]. Four intrinsic muscles and 5 extrinsic cardiac muscles that constrict or dilate the dorsal and lateral gastric walls have been identified. All of these identified muscles are innervated by motor axons from the constrictor and dilator neurons in the STG. Innervation It is of interest to compare innervation of the muscles in the cardiac stomachs of stomatopods and decapods. Maynard and Dando [5] have described the detailed anatomy of muscles and nerves of the stomach in several decapods. In Squilla the lvn is the principal nerve carrying the motor axons which innervate the muscles of the pep and pyloric stomach [16]. It was found that the lvn and, in addition, son are the principal nerves which contain motor axons both of the cardiac constrictor and dilator neurons of the STG. They divide peripherally into constrictor and dilator nerves supplying 8 identified muscles. One extrin- sic muscle is doubly innervated. In Panulirus, on the other hand, the dorsal ventricular nerve contains most of motor axons from the STG which innervate the muscles of the gastric mill and pyloric stomach [1]. The anterior median nerve and median ventricular nerve (mvn) carry the motor axons of the cardiac constrictor neurons (AM and IC neurons) of the STG which innervate the muscles of the cardiac stomach and the muscles of the ventral cardiac ossicles (cv2) [1, 5, 11]. The motor innervation provided by the cardiac con- strictor neurons in Panulirus is relatively simple, and differs from that in Squilla. In Panulirus, the mvn also contains a motor axon from the cardiac dilator neuron (VD neuron) which innervates the muscles of the ventral cardiac ossicles (cv1). The cvl and cv2 muscles move the ventral gastric wall of the cardiac stomach in Panulirus: the one is homologous to the plem in Squilla, and the other to the pcm. Four nerves carry motor axons of 2 cardiac dilator neurons (CD neurons) in Panulirus [11, 12]. Motor nerves branch from these nerves to innervate 7 extrinsic muscles: 5 muscles are inner- vated by the CD1, 4 muscles by the CD2, and thus 2 muscles receive a double innervation. The peripheral courses of the CD neurons in Panulirus appear complex compared to those in Squilla. Cardiac cycle Cardiac cycling occurred spontaneously in STG of semi-intact preparations (Fig. 2). The most notable characteristics of the cardiac cycle pattern, compared to the pcp-pyloric cycle pattern, are that Stomatogastric Nervous System. II 307 the burst duration is very long (over 10 times), the cycle rate is low, and the pattern is composed of only two alternately bursting groups. Although cardiac cycles did not occur spontaneously in the semi-isolated preparation, they could be produced by activation of the son inputs. These were activated via the CG by stimulation of afferent fibers (Fig. 4). While the cardiac cycle is brought on by inputs from a higher center, the pcp-pyloric cycle is intrinsic to the STG, although it is modifiable [16]. The mechanism of cyclic pattern generation may be different for the two cycles. The cardiac cycle is generated by two groups of neurons. Four CA neurons and 3 CD neurons could be identified in the STG from the peripheral recordings (Fig. 3), but interactions present among them were not analyzed. Comparison with the motor neurons of the cardiac system between stomatopods and dec- apods has been made as follows. In Panulirus, they have been described by Moulins and Vedel [11]. The CA1, CA2 and CA3 neurons are homologous to the AM neuron in Panulirus. These neurons control movements of the intrinsic muscles spread over the gastric wall. The CA4 neuron may be functionally homologous to the IC neuron in Panulirus. Both command constriction of the posterior lateral gastric wall. The CD3 neuron may be homologous to the VD neuron in Panulirus: the one is an antagonist of the CA4 neuron, and the other of the IC neuron. The VD and IC neurons are involved in the pyloric cycle [7]. The function of these neurons appears to be related to movements of both the cardiac and the pyloric stomachs, although their role in the pyloric cycle is unknown [6, 11]. Comparison of the phase relationships of the pcp-pyloric cycle in Squilla with those of the pyloric cycle in Panulirus suggests that the PCP1 (or PCP2) and PCP3 neurons, which control cardiac plate muscles, are homologous to the VD and IC neurons, respectively [16]. It is unknown whether the CA4 and CD3 neurons are involved in the pcp-pyloric cycle in controlling movements of these muscles. The CD1 and CD2 neurons in Squilla are certainly homologous to those in Panulirus. Panulirus exhibit complex functional properties The two CD neurons in that contribute to the organization of the neural network taking part in motor commands [11, 12]. The cell body of CD1 is located in the OG, and that of CD2 in the STG. Furthermore, their peripheral nerve distributions are quite complex. Such complexity has not been observed in the 3 CD neurons of Squilla. Their cell bodies are located in the STG. The functional movements of the cardiac stom- ach commanded by the cardiac cycle can now be described, although movements of the ventral gastric wall remain to be analyzed. Three of the four types of movements of the gastric wall described by Kunze [15] have been examined in this study. Four variations of spontaneous cardiac cycling are associated with them. The motor patterns shown in Figure 2D command medial movements of the dorsal and lateral gastric walls (termed the first phase by Kunze). The motor patterns which contribute to medial movements of the dorsal, lateral and posterior lateral gastric walls (the second phase) are illustrated in Figure 2A, B and C. Medial movements of the anterior dorsal gastric wall (the third phase) appear to be commanded by the motor pattern similar to that shown in Figure 2D, because the constrictor neurons in the son do not fire in this pattern. Maceration of ingested food may occur during these three phases while the pcp lateral channels are closed. Kunze [15] has seen the fourth phase which seems to occur sporadically. During this phase, digested food are transferred by simul- taneous muscular contractions from the cardiac to the pyloric stomach through the pcp lateral and cardio-pyloric channels. Such food movements have not been examined in this study (see next section). Functional relations between cardiac and pcp- pyloric cycles It has been shown that once the cardiac cycle is activated by stimulation of the son input fibers, the pep-pyloric cycle is suppressed [16]. On the other hand, it was found that the pcp-pyloric cycle units suppressed the cardiac cycle units: the PY neurons inhibited the CA1 and CA4 neurons (Fig. 5A, B); the PCP1, PCP2 and PD neurons also inhibited the CA1 and CA2 neurons (Fig. SC). In both cycles 308 the units fired alternately. The CAl and CA4 neurons control contractions of the lem] and pem, respectively, to bring the dorsolateral gastric wall medially in front of the pcp. Thus, the gastric wall can be tightly pressed to the dorsal floor of the pcp, resulting in blockade of the cardio-pyloric channel. The PY neurons command constrictions of the ampulla in the pyloric stomach, which cause a backflow of digestive juices to the cardiac stomach through the cardio-pyloric channel from the pylor- ic ampullary channels [15]. The inhibition of the cardiac constrictor neurons by the pyloric neurons probably makes the backflow possible. Figure 5C shows the pattern of motor outputs likely to represent the commencement of the transfer of digested food from the cardiac to the pyloric stomach: the CA neurons command constrictions of the gastric wall to force food suspensions backward, and the PCP and PD neurons command the opening of the pcp lateral and cardio-pyloric channels. In Panulirus, the CD neurons accelerate or inhibit the PD neurons [11]. Such modulation of the pcp-pyloric cycle by the cardiac cycle units appeared to be present in the Squilla STG (Fig. 4). However, the functional significance of such mod- ulation is still unknown. The present study has shown that the cardiac cycle of the STG in Squilla, as well as the pep-pyloric cycle, is of special interest for studying the characteristics of central pattern generators. Differences between the two cycles have been described. The cardiac and pcp-pyloric systems also provide useful models for understanding the mechanism underlying the generation of rhythmic motor patterns. The cellular properties and neural networks of the two subsystems of the STG in stomatopods will be studied for comparison with those in decapods which have been well analyzed. ACKNOWLEDGMENTS The author is grateful to I. M. Cooke, J. A. Benson and M. W. Miller for critical review and suggestions on this manuscript; H. Fukuyama and M. Funahashi for technical assistance; to R. Miyamoto for illustrations. This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan (No. 59540460). K. TAZAKI ie) 10 11 13 14 15 16 REFERENCES Selverston, A. I., Russell, D. F., Miller, J. P. and King, D.G. (1976) The stomatogastric nervous system: structure and function of a small neural system. Progr. Neurobiol., 7: 215-290. Wales, W. (1982) Control of mouth parts and gut. In “The Biology of Crustacea. Vol. 4. Neural Integration and Behavior”. Ed. by D. Sandeman, and H. Atwood, Academic Press, New York, pp. 165-191. Orlov, J. (1929) Uber den histologischen Bau der Ganglien des Mundmagennervensystems der Crustaceen. Z. Mikrosk.-Anat. Forsch., 8: 493-541. Maynard, D. M. (1966) Integration in crustacean ganglia. Symp. Soc. Exp. Biol., 20: 119-149. Maynard, D.M. and Dando,M.R. (1974) The structure of the stomatogastric neuromuscular sys- tem in Callinectes sapidus, Homarus americanus and Panulirus argus (Decapoda, Crustacea). Philos. Trans. R. Soc. Lond., B268: 161-220. Maynard, D. M. (1972) Simpler networks. Ann. NY. Acad. Sci., 193: 59-72. Hartline, D. K. and Maynard. D. M. (1975) Motor patterns in the stomatogastric ganglion of the lobster Panulirus argus. J. Exp. Biol., 62: 405-420. Dando, M.R. and Selverston, A. I. (1972) Com- mand fibres from the supraoesophageal to the stomatogastric ganglion in Panulirus. J. Comp. Physiol., 78: 138-175. Russell, D. F. (1976) Rhythmic excitatory inputs to the lobster stomatogastric ganglion. Brain Res., 101: 582-588. Russell, D. F. (1979) CNS control of pattern gener- ators in the stomatogastric ganglion. Brain Res., 177: 598-602. Moulins, M. and Vedel, J. P.(1977) Programmation centrale de l’activité motorice rhythmique du tube digestif antérieur chez les Crustacés décapods. J. Physiol. (Paris), 73: 471-510. Vedel, J.P. and Moulins, M. (1977) Functional properties of interganglionic motoneurons in the stomatogastric nervous system of the rock lobster. J. Comp. Physiol., 118: 307-325. Police, G. (1909) Sul sistema nervoso viscerale della Squilla| mantis. Mitt. Zool. Sta. Neapel, 19: 144-148. Reddy, A. R. (1935) The structure, mechanism and development of the gastric armature in Stomato- poda with a discussion as to its evolution in Decapoda. Proc. Indian Acad. Sci., B1: 650-675. Kunze, J. C. (1981) The functional morphology of stomatopod Crustacea. Philos. Trans. R. Soc. Lond., B292: 255-328. Tazaki, K., Miyatani, M. and Ando, F. (1986) The Stomatogastric Nervous System. II 309 anatomy and physiology of the stomatogastric ner- 17 Watanabe, A., Obara, S. and Akiyama, T. (1967) vous system of Squilla. I. The posterior cardiac plate Pacemaker potentials for the periodic burst dis- and the pyloric systems. J. Comp. Physiol., 159A: charge in the heart ganglion of a stomatopod, 521-533. Squilla oratoria. J. Gen. Physiol., 50: 839-862. so : i a rei onal : Ne 7 oe > ‘ ead ZOOLOGICAL SCIENCE 5: 311-321 (1988) Ultrastructure and Physiological Response of Leucophores of the Medaka Oryzias latipes MASATAKA OBIKA Department of Biology, Keio University, Yokohama 223, Japan ABSTRACT— Ultrastructure and physiological responses of leucophores in isolated scales of the medaka Oryzias latipes were studied. Many of the leucophores are in close association with overlying melanophores, and nerve fibers that run between the two cells frequently form synapses on both sides. This situation provides a very efficient way to conduct body lightening response, since the stimulation of single adrenergic fiber produces pigment aggregation in the melanophore and dispersion in the leucophore almost simultaneously. Although it appears to be rather infrequent, some nerve fibers enter into the cell body of leucophores. Spherical and tubular synaptic vesicles, and larger vesicles with an electron-dense central core are observed in single nerve fibers. Responses of leucophores are produced by selective migration of the pigment granules toward or away from the center of the cell. Numerous microtubules and 10 nm filaments run parallel to the long axis of the dendrites, though direct connection between these cytoskeletal elements and pigment granules has not been ascertained by electronmicroscopy. Pigment dispersion and aggregation proceed normally in the presence of cytochalasin B while colchicine and EHNA (erythro-9-3-(2-hydroxynonyl)adenine) potently inhibit © 1988 Zoological Society of Japan pigment aggregation. epinephrine. leucophores are microtubule-dependent. INTRODUCTION Leucophores in the integument of the medaka are generally found in close association with overlying melanophores to form the melanophore- leucophore combination. Pigment cells in this teleost are under the control of adrenergic nerves, and the prevailing alpha adrenergic receptors on melanophores make their pigment aggregate in response to nervous stimulation producing body lightening [1] while leucophores, which are pre- dominantly controlled by beta adrenergic recep- tors respond to the same signal with pigment dispersal that also enhances body blanching by increasing light reflectance [2, 3]. Thus, the combination of these two types of pigment cells represents an exquisite example of dermal chroma- tophore unit [4] in fish integument. Studies on the physiological responses of leucophores have so far been carried out at light microscopic level, but the Accepted September 30, 1987 Received September 2, 1987 NEM (N-ethylmaleimide) interferes with the pigment dispersion elicited by These results suggest that the intracellular movements of pigment granules in ultrastructural basis of their responsiveness has not yet been clarified. The present study deals with the ultrastructure of leucophore and melanophore-leucophore com- bination. Innervation of nerve fibers into leucophores and melanophore-leucophore com- bination is clearly shown for the first time. Leucophores contain numerous spherical pigmen- tary organelles that selectively migrate either centripetally or centrifugally upon aggregative and dispersive stimuli. Their cytoplasm possesses a moderately developed microtubule system in addi- tion to 10-nm filaments. Pharmacological study indicates that the microtubule system, rather than actin-myosin system, is involved in the motile mechanism. MATERIALS AND METHODS Leucophores in the dermis of adult Oryzias latipes were used in this study. Scales were plucked from the anterior-dorsal region. 312 M. OBIKA Electron microscopy Scales were fixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.2) for 30 to 60 min at room temperature, post-fixed in 1% OsQ, in the same buffer for 30min at room temperature, dehydrated through a graded series of alcohol and embedded in epoxy resin. Ultrathin sections were stained with uranyl acetate and lead citrate and observed in a JEOL 1008S electron microscope at the accerelation voltage of 80 kV. Physiological responses and chemicals Physiological and pharmacological studies were made on the materials from which overlying epidermis had been removed by a 30 min treat- ment in 0.25% collagenase (Worthington, type II) in a Ca-free teleost saline solution with gentle agitation. Response of leucophores was observed under a Nikon inverted microscope (Diaphot). Teleost saline solution contained 128mM NaCl, 2.6 mM KCl and 1.8 mM CaCl.and buffered with 5 mM Tris-HCl at pH 7.2 [3]. Epinephrine (Sigma), theophylline (Tokyo Kasei), colchicine (Sigma), N-ethylmaleimide (NEM, Kokusan Kagaku, Fic. 1. complex. Tokyo), 2, 4-dinitrophenol (Tokyo Kasei), potas- sium cyanide (Kokusan Kagaku), erythro-9-’3-(2- hydroxynonyl)adenine (EHNA, Burroughs Well- come) and synthetic teleost melanophore concen- trating hormone (MCH), which was generously supplied by Dr. M. E. Hadley of University of Arizona, were dissolved directly in the saline solution at appropriate concentrations. Stock solution of cytochalasin B (Aldrich) was made in dimethylsulfoxide (DMSO) and diluted with saline immediately before use. RESULTS Morphology of leucophore and its association with melanophore Although a few leucophores in the dorsal skin are found without having any obvious contact with other pigment cells, a large number of leucophores are observed immediately below overlying mela- nophores. The dendritic processes of mela- nophores_ occasionally extend downward and embrace the upper portion of leucophores. Figure 1 depicts an example where the structures like Overlying melanophore (M) is in close contact with a leucophore (L) to form a melanophore-leucophore Electron lucent vesicles in the leucophore (P) are pigmentary organelles characteristic of this cell type. Arrows indicate the sites where the pigment cells are closely attached. Bar represents 0.5 um. Ultrastructure of Leucophore 313 intercellular bridge and the accumulation of dense materials on the membranes of both chroma- tophores are seen (arrows). Though the cells are very closely attached, there is no evidence that indicates the actual cytoplasmic connection be- tween the two different types of chromatophores. Innervation to melanophore-leucophore combina- tion Axons with putative synaptic vesicles of various morphology are frequent near the chromatophore combination. Many of them run in the intercellu- lar space between melanophore and leucophore (Figs. 2 and 3) while some enter deeper into leucophores (Figs. 4 and 5). In any case, morpho- logical specialization of pre- and postsynaptic membranes is not distinct. Synaptic vesicles are either spherical or tubular and contain moderately electron dense material. Some of the larger vesicles contain electron dense core. The diameter of smaller vesicles and tubules is about 40 nm while the larger ones with central electron dense core mesure about 80 nm. These three types of synaptic vesicles occur in a single axonal fiber. Although melanophores have a relatively large number of synapses both on the outer (epidermal side) and lower (leucophore side) surfaces, axons on leucophores are rather infrequent compared with those locating on melanophores. Morphology of leucophores with aggregated and dispersed pigment granules Figure 6 depicts a typical profile of a leucophore with aggregated pigments. Central portion of the cell is occupied by an aggregate of spherical pigmentary organelles that contain some fuzzy, amorphous intravesicular substance. Pigmentary organelles are rather uniform in size (about 500 nm in diameter). The central cytoplasm contains ER, ribosomes, microtubules, 10-nm filaments and some other membrane-bound vesicles but mitochondria are always found in the dendritic processes. The dendrites are more or less flattened but retain their width after the withdrawal of the pigment. Smaller vesicles, mitochondria, a large number of ribosomes, 10 nm filaments and numer- ous microtubules aligned parallel to the long axis of the processes are prominent cytoplasmic organelles in dendrites. Golgi complexes are frequent in their proximal portion. In cells with dispersed pigment, dendrites contain pigmentary organelles in addition to the larger vesicles in various size and morphology (Fig. 7). The origin and function of these larger vesicles are unknown, but they sometimes contain intravesicular material similar to that found in pigment granules. A few mitochondria are seen in the central portion of the cell, though the majority of them are densely populated in the proximal portion of the dendrites. Microtubules and 10-nm filaments are also abun- dant in this region. These findings indicate that chromatophore responses of the leucophores are produced by the change in the intracellular dis- tribution of the pigmentary organelles. The effects of chemicals on leucophore responses EHNA Leucophores with their pigment aggregated in the saline solution responded to EHNA with pigment dispersion in a_ dose- dependent manner. This response was completely reversed by washing the specimens with physiolog- ical saline solution. At concentration of 30 uM, EHNA had no appreciable effect on aggregated leucophores, at 60 uM, however, it produced a slight pigment dispersal within 10 min. A 10 min incubation in 125 ~M EHNA produced an almost full dispersion of leucophores in 10 min. In 500 uM or 1 mM, dispersal was induced within 5 min. Figure 8 shows an example where aggregated leucophores in saline solution (Fig. 8a) became dispersed after 20 min incubation in 1 mM EHNA (Fig. 8b). Subsequent perfusion with 10 ~M epinephrine for 15 min in the presence of EHNA did not change their morphology in appreciable degrees (Fig. 8c). Further treatment of the cells with an aggregating agent, melatonin (1 g/ml) for 25 min produced only a very minute response (Fig. 8d). The effect of EHNA was reversed by washing the scale in saline solution for 20min and leucophores became punctate (Fig. 8e). These cells responded to epinephrine in 10 min as shown in Figure 8f. Colchicine Most of the leucophores in isolated scales remained aggregated in physiologi- cal saline. Transfer of these scales into 1 to 5 u«M 314 M. OBIKA PMS Fic. 2. A single nerve fiber (N) makes synaptic contacts with melanophore (M) and leucophore (L). Bar represents 0.5 yam. Fic. 3. A longitudinal section of a nerve fiber (N) between the dendrites of melanophore (M) and leucophore (L). Melanosomes are withdrawn from the dendrite. Synaptic vesicles of various size and morphology are seen. Ten nm filaments are found in leucophore. Scale bar represents 0.5 sm. colchicine produced a rapid dispersion of incubation in 1 mM colchicine (Fig. 9b). Melato- leucophores within a few minutes. Figure 9 shows nin, epinephrine and theophylline all failed to leucophores in saline (Fig.9a) and after 1hr_ elicit further response of leucophores in the Ultrastructure of Leucophore 315 Fic. 4. Cross-sectional profile of a nerve fiber (N). The fiber is encircled by leucophore (L) membrane. Bar: 0.5 yam. Fic. 5. Longitudinally sectioned nerve (N) found near the central portion of a leucophore. Bar: 0.5 um. presence of colchicine. responded normally to epinephrine with pigment Cytochalasin B Leucophore responses are _ dispersion. Removal of epinephrine produced found to be totally insensitive to cytochalasin B. _leucophore reaggregation. Figure 10 shows a Specimens treated in 10 g/ml cytochalasin B (final typical response of leucophores to cytochalasin B. concentration of DMSO was 0.25%) for up to3 hr Most of the leucophores in an isolated scale Fic. 6. Proximal portion of a dendrite of a leucophore with aggregated pigment. Pigment granules have migrated toward the cell center (far left), and the dendrite contains Golgi apparatus (arrowheads), mitochondria, microtubules (arrows) and other cytoplasmic organelles except pigment granules. Scale bar represents 1 «am. Fic. 7. Proximal portion of a dendrite of a leucophore with dispersed pigment. Pigment granules and some larger vesicles are now present in the dendrite. Bar: 1 pm. remained punctate in saline solution (Fig. 10a). perfusion with the medium containing both Incubation in cytochalasin B for 2hr did not epinephrine and cytochalasin B produced a change their shape (Fig. 10b) and the following prompt dispersal of the cells (Fig. 10c) within 7 Ultrastructure of Leucophore 317 Fic. 8. Effect of EHNA on the physiological response of leucophores. Aggregated leucophores in the saline solution (a) became dispersed by a 20 min incubation in 1 mM EHNA (b). No further dispersion was induced by a 15 min treatment with 10 ~M epinephrine (c). These cells did not respond to melatonin in appreciable degrees (d). The dispersive effect of the drug was completely reversed by washing in the saline (e), and the cells rapidly redispersed when treated with epinephrine (f). 113. Fic. 9. Effect of colchicine. Aggregated leucophores in the saline solution (a) became dispersed by a 1 hr incubation in 1 mM colchicine (b). 113. 318 M. OBIKA Fic. 10. incubation in cytochalasin B (b). cytochalasin B (c). 113. min. Close observation of the dendritic processes of dispersed leucophores indicated that their tips were more inflated compared to those dispersed in the absence of cytochalasin B. NEM NEM at concentrations between 0.1 to 2.5mM potently inhibited epinephrine- or theophylline-induced pigment dispersion — in leucophores. Application of the drug on dispersed leucophores (produced by _ theophylline-pre- treatment) did not induce pigment aggregation either. This drug has also a very potent inhibitory effect on melanophore aggregation that is normal- ly produced by epinephrine and MCH. Potassium cyanide and dinitrophenol (DNP) When the isolated scales were incu- bated in KCN at 5X10~* M for 40 to 60 min, dispersion of leucophores by epinephrine or theophylline was only partially inhibited. DNP at 1 mM potently inhibited leucophore dispersion by epinephrine and theophylline, but produced pig- ment aggregation when applied on dispersed leucophores. The effects of these metabolic inhibi- tors on leucophore responses were, however, not as prominent as those found on melanophores and xanthophores where the responses appear to be more susceptible to these chemicals. Leucophore response to MCH Leucophores with aggregated pigments pro- duced by placing scales in physiological saline, or Effect of cytochalasin B. Aggregated leucophores in the saline (a) remain aggregated during a 2 hr Rapid pigment dispersion was induced by epinephrine in the presence of those with dispersed state induced by theophylline treatment, were perfused with synthetic MCH (1 nM-1 4M). Neither pigment dispersion nor aggregation was induced although the hormone was potent enough to produce full pigment aggregation in neighboring melanophores within a few minutes at the lowest concentration employed. DISCUSSION Membrane specialization of melanophore-leuco- phore junction Although leucophores and melanophores of Fundulus are in close association as in the present species, no specialized junctional structure has been demonstrated in the earlier work [5]. In Oryzias, leucophores and melanophores appear to be more closely associated. Sometimes mem- branes of the two cells appeared to be tightly attached. Whether the association is simply hold- ing the two cells together (tight junction) or it actually functions as electrical or metabolic cou- pling (gap junction) remains to be investigated. Innervation into melanophore-leucophore com- bination Pharmacological evidence indicates _ that leucophores of the medaka are under the control of beta adrenergic receptors [2, 3]. Since rhythmic Ultrastructure of Leucophore 319 pigment granule aggregation and dispersion (pulsation) induced by Ba ions occur simul- taneously but in opposite directions in a mela- nophore-leucophore combination, it has been sug- gested that the two chromatophores are innervated by the same nerve. Furthermore, the evidence is presented that xanthophores are also controlled by the same nerve, thus providing an efficient way to adapt the fish to its environmental background [6]. Since the early work of Ballowitz [7], innervation into fish melanophores has been studied repeated- ly at light microscopic level (see [8, 9] for review). Adrenergic innervation to melanophores and erythrophores has also been demonstrated recent- ly by light microscopic autoradiography [10-12]. At electron microscopic level, chromatophore- neural junctions have been described in the mela- nophores of Fundulus [13], Chasmichthys [14] and the angelfish Pterophyllum [15]. However, in- nervation to bright-colored chroamtophores, 1.e. xanthophores and leucophores, has not been studied at ultrastructural level. The present study clearly shows that the melanophore-leucophore combination is, in fact, innervated by a single nerve fiber as Iwata and his collaborators have concluded from their physiological studies [6]. The synaptic structure is rather indistinct, membrane specialization with pre- and postsynaptic densities being only occasionally observed. Synapses are frequently found on the epidermal side of mela- nophore membrane in addition to those found in melanophore-leucophore junctions. Sometimes fibers run deeper into melanophores or leucophores. Observation on serial sections indi- cates that nerve fibers form en passant synapses as has been shown in the angelfish melanophores [15]. Morphology of leucophore with aggregated and dispersed pigments Ultrastructural observation revealed that, in contrast to the response of melanophores where the cytoplasm other than pigment granules translo- cates simultaneously [16, 17], aggregation or dis- persion of leucophores is produced by a selective translocation of pigmentary organelles in the cytoplasm toward the centripetal or centrifugal direction. Very selective movement of pigment granules conducted by a cytoskeletal meshwork has been reported in erythrophores [18]. In leucophores, however, direct association of the pigment granules with cytoplasmic microtubules or other cytoskeletal elements has not been demon- strated. Dendritic processes of a leucophore with aggregated pigments contain numerous microtu- bules, 10 nm filaments, free and membrane bound ribosomes, mitochondria and other membrane bound organelles but are entirely free of pigment granules. The central portion of the cell is, on the other hand, largely occupied by pigment granules but contains other cytoplasmic organelles in much less number compared to the peripheral region. Microtubules running parallel to the long axis of the dendrite are abundant while 10 nm filaments are more frequent near the central portion of the cell. Sometimes small bundles of 10 nm filaments were found at the boundary between the mass of pigment granules and the base of the dendrite. Mitochondria of leucophores, unlike those of melanophores that aggregate toward the perinu- clear region during pigment aggregation, did not change their distribution pattern drastically during pigment migration. This suggests that the pigment migration in leucophores proceeds more gently and selectively, provided that mitochondria trans- locate passively in both cases. Mechanism of pigment migration EHNA, a dynein-ATPase inhibitor, made leucophore disperse at concentrations as low as 60 uM. The effect was dose-dependent and was readily reversed by washing. Both in intact and detergent-treated, permeabilized preparations of Fundulus melanophores, EHNA at 2 mM blocked epinephrine-elicited pigment aggregation [19] and in melanophores of Oryzias, it inhibited epinephrine or MCH-induced melanosome aggregation at concentrations between 0.25 to 2 mM [20]. In Holocentrus erythrophores, the drug at 1 to 4 mM prevented the saltatory movement of pigment granules but epinephrine did induce pigment aggregation at slower rate than in un- treated cells [21]. Thus, the response pattern of the leucophores to EHNA is similar to that observed in melanophores, though leucophores appear to be more sensitive to the inhibitor. 320 Kinesin, a microtubule-dependent motor molecule that supports anterograde axoplasmic transport of cytoplasmic particles [22, 23], is reported to be relatively resistant to the effect of EHNA [24] or NEM [23]. The movement of pigment granules within leucophores in centripetal and centrifugal directions was totally arrested by the presence of NEM, and centrifugal displacement of the gra- nules was produced by EHNA. Although the effects of these drugs and those of uncouplers of oxidative phosphorylation may be partially due to the depletion of ATP, relatively high sensitivity of the leucophores to NEM and EHNA suggests the involvement of dynein-tubulin interaction in their responses, especially in pigment aggregation, as has been suggested in fish melanophores [19, 20] and melanoma cells [25]. The dispersive effect of colchicine is. common to all types of Oryzias chromatophores but hardly reversible. Pigment dispersion in leucophores was induced rapidly in the presence of relatively low concentrations of the alkaloid. Cytoplasmic microtubules appear to be resistant to this drug, and a derivative of colchi- cine, lumicolchicine, which does not bind to tubulin also blocks pigmentary responses in mela- nophores in a similar manner as colchicine does [26]. These observations suggest that the site of action of colchicine is not restricted on cytoplasmic microtubules. The involvement of actin-myosin system in the pigment movements is not likely since the responses were insensitive to cytochalasin B. It has recently been suggested that pigment dispersal and aggregation in fish xanthophores and melanophores are brought about by protein phos- phorylation and dephosphorylation, respectively [27-30]. Though the mechanism of pigment translocation in leucophores remains to be ascer- tained from this point of view, induction of pigment dispersal by theophylline or dibutylyl cyclic AMP [2] allows an assumption that the elevation of intracellular cyclic AMP level pro- duces pigment dispersal in leucophores as well as in the other types of chromatophores. REFERENCES 1 Fujii, R. (1961) Demonstration of the adrenergic M. OBIKA ie) 10 11 13 14 15 16 nature of transmission at the junction between melanophore-concentrating nerve and melanophore in bony fish. J. Fac. Sci. Univ. Tokyo, Sec. IV, 9: 170-196. Obika, M. (1976) An analysis of the mechanism of pigment migration in fish chromatophores. Pigment Cell, 3: 254-264. Iga,T., Yamada,K. and Iwakiri,M. (1977) Adrenergic receptors mediating pigment dispersion in leucophores of a teleost, Oryzias latipes. Mem. Fac. Lit. Sci., Shimane Univ. Nat. Sci., 11: 63-72. Bagnara,J.T., Taylor,J.D. and Hadley, M. E. (1968) The dermal chromatophore unit. J. Cell Biol., 38: 67-79. Menter,D.G., Obika,M., Tchen,T.T. and Taylor, J. D. (1979) Leucophores and iridophores of Fundulus heteroclitus: Biophysical and _ ultras- tructural properties. J. Morphol., 160: 103-120. Iwata, K.S., Takahashi, T. and Okada, Y. (1981) Nervous control in chromatophores of the medaka. In “Phenotypic Expression in Pigment Cells”. Ed. by M. Seiji, Univ. Tokyo Press, Tokyo, pp. 433- 438. Ballowitz, E. (1893) Die Nervenendigungen der Pigmentzellen. Z. wiss. Zool., 56: 673-706. Parker, G. H. (1948) Animal Colour Changes and their Neurohumours. Cambridge Univ. Press, Cam- bridge. Bagnare, J. T. and Hadley, M. E. (1973) Chroma- tophore and Color Change. Prentice-Hall, N.J. Yamada, K., Miyata, S. and Katayama, H. (1984) Autoradiographic demonstration of adrenergic in- nervation to scale melanophores of a teleost fish, Oryzias latipes. J. Exp. Zool., 229: 73-80. Miyata,S. and Yamada, K. (1985) Pattern of adrenergic innervation to scale erythrophores of the swordtail, Xiphophorus helleri. Zool. Sci., 2: 49-57. Miyata,S. and Yamada, K. (1987) Innervation pattern and responsiveness of melanophores in tail fins of teleost. J. Exp. Zool., 241: 31-39. Bikle, D., Tilney, L.G. and Porter, K. R. (1966) Microtubules and pigment migration in the mela- nophores of Fundulus heteroclitus L. Protoplasma, 61: 322-345, Fujii, R. (1966) A functional interpretation of the fine structure in the melanophore of the guppy, Lebistes reticulatus. Annot. Zool. Japon., 39: 185- 192. Schliwa,M. (1976) Fine structure of nerve- melanophore contacts in the angelfish Pterophyllum scalare. Cell Tissue Res., 171: 381-387. Obika, M. (1976) Pigment migration in isolated fish melanophores. Annot. Zool. Japon., 49: 157-163. Schliwa, M. and Euteneuer, U. (1978) Quantitative analysis of the microtubule system in isolated fish melanophores. J. Supramol. Struct., 8: 177-190. 18 19 20 21 22 23 24 Ultrastructure of Leucophore 321 Byers, H. R. and Porter, K. R. (1977) Transforma- tions in the structure of the cytoplasmic ground substance in erythrophores during pigment aggrega- tion and dispersion. J. Cell Biol., 75: 541-558. Clark, T. G. and Rosenbaum, J. L. (1982) Pigment particle translocation in detergent-permeabilized melanophores of Fundulus heteroclitus. Proc. Natl. Acad. Sci. USA, 79: 4655-4659. Negishi, S., Fernandez, H.R.C. and Obika, M. (1985) The effects of dynein ATPase inhibitors on melanosome translocation within melanophores of the medaka, Oryzias latipes. Zool. Sci., 2: 469-475. Beckerle, M. C. and Porter, K. R. (1982) Inhibitors of dynein activity block intracellular transport in erythrophores. Nature, 295: 701-703. Vale, R. D., Reese, T. S. and Sheetz, M. P. (1985) Identification of a novel force-generating protein, kinesin, involved in microtubule-based motility. Cell, 42: 39-50. Vale, R. D., Schnapp, B. J., Mitchison, T., Steuer, E., Reese, T. S. and Sheetz, M. P. (1985) Different axoplasmic proteins generate movement in opposite directions along microtubules in vitro. Cell, 43: 623- 632. Brady, S. T., Lasek, R. J. and Allen, R. D. (1985) Video microscopy of fast axonal transport in ex- truded axoplasm: A new model for the study of molecular mechanisms. Cell Motility, 5: 81-101. 25 30 Ogawa, K., Hosoya, H., Yokota, E., Kobayashi, T., Wakamatsu, Y., Ozato, K., Negishi, S. and Obi- ka, M. (1987) Melanoma dynein: evidence that dynein is a general “motor” for microtubule- associated cell motilities. Eur. J. Cell Biol., 43: 3-9. Obika, M., Turner, W. A., Negishi, S., Menter, D. G. Tchen,T.T. and Taylor, J.D. (1978) The effects of lumicolchicine, colchicine and vinblastine on pigment migration in fish chromatophores. J. Exp. Zool., 205: 95-110. Lynch, T. J., Taylor, J. D. and Tchen, T. T. (1986) Regulation of pigment organelle translocation. I. Phosphorylation of the organelle-associated protein p)/. J. Biol. Chem., 261: 4204-4211. Lynch, T. J., Wu, B., Taylor, J. D. and Tchen, T. T. (1986) Regulation of pigment organelle trans- location. II. Participation of a cAMP-dependent protein kinase. J. Biol. Chem., 261: 4212-4216. Rozdzial, M. M. and Haimo, L. T. (1986) Reacti- vated melanophore motility: Differential regulation and nucleotide requirements of bidirectional pig- ment granule transport. J. Cell Biol., 103: 2755— 2764. Rozdzial, M. M. and Haimo, L. T. (1986) Bidirectional pigment granule movements of mela- nophores are regulated by protein phosphorylation and dephosphorylation. Cell, 47: 1061-1070. ! ' ; = t ; . a x i , oS : at ; ; 4 7 ; ' ry ; as = ad : a : t ; Fe ’ aye na meh wg Oe M, Wee “1 asad "yin ZOOLOGICAL SCIENCE 5: 323-330 (1988) Conjugation in Tetrahymena: Its Relation to Concanavalin A Receptor Distribution on the Cell Surface YOSHIKO SUGANUMA and HirosH! YAMAMOTO! Biological Laboratory, Nara Sahojogakuin College, Rokuyaon-Cho, Nara 630, and ‘Department of Anatomy, Nara Medical University, Kashihara, Nara 634, Japan ABSTRACT—Cell surface events during conjugation of Tetrahymena thermophila were studied by electron microscopic examination of ferritin conjugated concanavalin A (F-Con A). Small amounts of ferritin particles (F-particles) were bound to the surface of cells in the nutritional and starved states, but there were no large clusters of F-particles. In regions where F-particles were scantly, the ectoplasmic layer (epiplasm) was directly under the plasma membrane and no alveoli were observed. In contrast, in cells during co-stimulation, large clusters of F-particles were seen on the presumptive junctional area (PJA) formed after the onset of co-stimulation between the complementary mating types, and on the side walls of ectoplasmic ridges of the oral apparatus. In regions of large clusters of F-particles, there was a thick, dense ectoplasmic layer under the plasma membrane and no alveoli were seen. In conjugants, F-particles were seen not only on the smooth PJA but also in zones of gaps in the junctional area between the two conjugants. These findings suggest that the Con A binding glycocalyx is anchored to the ectoplasmic layer, a kind of cytoskeleton, under the plasma membrane. © 1988 Zoological Society of Japan INTRODUCTION The protozoan ciliate Tetrahymena usually mul- tiplies asexually by binary fission, but it can also reproduce sexually through conjugation. Conjuga- tion can be induced artificially and distinguished into two successive processes: (i) an initiation process induced by removing nutrients from the culture medium, and (ii) a co-stimulation process which begins on mixing the two complementary mating types [1, 2]. Recent studies on the ultrastructure of cells during the conjugation pro- cess have shown that complementary mating type cells recognize each other during the co- stimulation period and then interact to form a presumptive junctional area (PJA) on the front side of the anterior region. This PJA is a special area that has no particular subpellicular organelles. The co-stimulation period is followed by the conjugation period in which the cells pair at the sites of their PJAs. Accepted September 18, 1987 Received August 5, 1987 Finally, the PJAs of the conjugated paris partially fuse to form intercellular bridges [3, 4]. On the basis of these findings, the co-stimulation process in this protozoan ciliate can be regarded as the period for formation of a specialized mem- brane area, which may be important in cell recognition, cell adhesion, and final cell fusion. Glycocalyx on the surface of the plasma membrane is thought to be involved in cellular events such as differentiation, malignant transformation, and cell-cell adhesion [5]. Addition of concanavalin A (Con A) to cultured mammalian cells induces gathering of ConA receptors to form a cap-like structure [6]. Indeed, the conjugation process of Tetrahymena, which includes cell transformation, recognition and adhe- sion, is inhibited by Con A treatment [7, 8]. A previous report from this laboratory described ultrastructural changes in Tetrahymena during the process of PJA (SSA) formation [3]. In the present study, ferritin conjugated concanavalin A (F-Con A) was used to examine the distribution of Con A receptors on the cell surface of Tetrahy- mena. Ww in) = MATERIALS AND METHODS The two complementary mating types used were II and IV (strain B) of Tetrahymena thermophila, which were kindly provided by Dr. T. Sugai, Ibaraki University, Mito, Japan. The ciliates were cultured axenically in growth medium (2% proteose peptone, 1% yeast extract, and 0.6% glucose) at 26°C. Late log phase ciliates (10° cells/ml) were harvested and washed three times with medium consisting of KCI (0.008%), NaCl (0.2%) and CaCl (0.012%) in redistilled Fic. 1. Y.SUGANUMA AND H. YAMAMOTO water with low speed centrifugation. Equal volumes of competent cells were mixed to induce subsequent conjugation. For determination of the distribution of Con A receptors on the cell surface, a suspension of the cells was incubated with 25 «g/ml of F-Con A (E. Y. Laboratories) for 20min at 26°C. Then the cells were collected by centrifugation and washed twice with 0.2 M phosphate buffer, pH 7.4. Specific binding of Con A to its receptors was confirmed in a control experiment using a-methyl- mannoside (20 mM). Section of a starved cell after incubation for 20 min with F-Con A. A few ferritin particles (F-particles) can be seen at the bottom of the ectoplasmic grooves at the ciliary base (arrows). The plasma membrane in this area is directly above a thin ectoplasmic layer (EL). Scale bar: 0.5 um. Fics. 1-7 are transmission electron micrographs of Tetrahymena thermophila. AL: alveoli, AEL: inner alveolus ectoplasmic layer, AZM: adoral zone of membranelle, C: cilium, CI: cisterna, KS: kinetosome, LT: longitudinal tubule, MI: mitochondria, MU: mucocyst, RI: ectoplasmic ridge. Fic. 2. Cross section through the upper part of the oral apparatus. The cells were mixed for 40 min, then incubated with F-Con A for 20 min. Numerous large clusters of F-particles can be seen on the surface of the plasma membrane of the smooth presumptive junctional area (PJA), but elsewhere there are few particles on the surface. Scale bar: 0.5 um. Fic. 3. Section of a closely adjacent part of the same cell as shown in Fig. 2. The EL in the PJA is connected with the inner alveolus ectoplasmic layer (AEL). Clusters of F-particles are found only on the plasma membrane of the PJA that is directly above the EL. Scale bar: 0.5 um. 325 Tetrahymena jugation in Con 326 Y.SUGANUMA AND H. YAMAMOTO Samples for electron microscopy were collected by low-speed centrifugation and fixed for 30 min at 0°C in freshly prepared fixative consisting of a mixture of solutions of 0.6% glutaraldehyde, 2% osmium tetroxide, and 1.2% potassium bichro- mate (2:1:1, v/v) adjusted to pH 7.4 with 0.2M phosphate buffer. Then they were dehydrated by rapid passage through a graded ethanol series and embedded in Epoxy resin containing Quetol 812 (11 g), DDSA (6g) and MNA (5.8 g). Ultrathin sections obtained with an LKB ultratome were stained with 1% aqueous uranyl acetate and lead citrate and examined with a JEOL 100-C electron microscope. OBSERVATIONS AND RESULTS Tetrahymena cells are pear-shaped, and are covered with numerous ridges stretching along the apse line. Cilia are arranged in a line along the ridges at the bottom. Cross sections, between the tip and upper edge of the oral apparatus of starved cells, have a highly undulating, ellipse-shaped profile. The region below the cytoplasmic mem- brane is mainly occupied by alveoli, kinetosomes, mucocysts, longitudinal tubules and other subpel- licular organelles. In the starved cell after incubation for 20 min with F-Con A (Fig. 1), a few F-particles are seen located exclusively around the ciliary base (arrows). In the areas where the F-particles are attached, there are no alveoli and an extremely thin ectoplasmic layer (EL) is seen directly below the cytoplasmic membrane. After co-stimulation for 40 min and additional incubation for 20 min with F-Con A, the unpaired cell has the profile of a highly undulating ellipse in cross section between the tip of the cell and the upper edge of the oral Fic. 4. Cross section through the adoral zone. Cells were mixed for 40 min and then incubated with F-Con A for 20 min. Large clusters of F-particles can be seen on the surface of the plasma membrane of the PJA (arrows) and on the side wall (F) of the ectoplasmic ridge (RI). Scale bar: 0.5 «am. Conjugation in Tetrahymena 327 apparatus (Figs. 2 and 3), like that of cells in the starved state (Fig. 1). But unlike during starvation the PJA appears during co-stimulation. A thick, dense EL appears under the cytoplasmic mem- brane of the PJA, and F-particles are only found on the surface that is directly above the EL, in areas of alveoli (AL), longitudinal tubules (LT) or cilia (C). The EL is seen as a single dense layer just beneath the cytoplasmic membrane of the PJA, whereas in areas around the PJA, it is displaced deep under the inner alveolus membrane (Figs. 2 and 3, AEL). The ectoplasmic layer is much thicker in cells in the co-stimulation period than in starved cells. Figure 4 shows a cross section through the adoral zone of an unpaired cell. There are numerus F-particles on the PJA (arrows) and on the side wall (F) of the adjacent ridges (RI). The cytoplasmic membrane associated with F- particles is directly above the EL and there are no other subpellicular organelles in the cortical zone. After co-stimulation for 60 min and incubation for 20 min with F-Con A (Figs. 5 and 6), cross sections through the junction area (JA) of con- jugants shows that numerous F-particles are spe- cifically distributed on regions of the PJA, that are not in contact, on the side walls of ridges (arrows) and in gaps between conjugant cells (Fig. 6, F). Figure 7 shows the PJA of a conjugant, without F-Con A incubation. The clear gap zone between the conjugant is filled with fine fibrous structures, possibly glycocalyx. Fic. 5. Cross section through the junction area (JA) of a conjugant. incubated with F-Con A for 20 min. Clusters of F-particles (arrows) can be seen on the surface of the PJA (arrows). Scale bar: 0.5 um. The cells were mixed for 60 min, and then 328 Y.SUGANUMA AND H. YAMAMOTO : ee : 6S OS Fic. 6. Section of the junction area (JA) of conjugants. The cells were mixed for 60 min, and then incubated with F-Con A for 20 min. F-particles can be seen singly or in clusters on the surface of the PJA (arrows) and in the gap zone between the conjugants (F). Scale bar: 0.5 um. Fic. 7. Section of the junction area of conjugants. The cells were mixed for 60 min. The gap zone between the conjugants (JA) contains many fine fibrous structures. Scale bar: 0.5 sam. Conjugation in Tetrahymena 329 DISCUSSION The present study on the surface of the cytoplas- mic membrane of Tetrahymena revealed the dis- tributions of Con A receptors in starved and conjugation-induced cells. The relationship of the distribution of Con A receptors with that of substructures in the cortical zone is noteworthy. The cortical zone of Tetrahymena contains various subpellicular organelles. The cytoplasmic membrane can be classified into the following three membrane areas according to differences in substructures in the cortical zone; (i) a ciliary area, (ii) a cortical area, and (iii) an area directly above the EL. The third type exhibits Con A binding activity specifically. During periods of nutrition and starvation, the third type is found only in restricted areas around ciliary bases and cell membranes that are directly above a thin EL. There are only a few sparsely distributed F- particles bound to the surface of such areas, so their Con A binding activity may be extremely weak. In studies by fluorescence microscopy with FITC-Con A, no Con A binding activity was detected on the cell surface during starvation before mixing complementary mating types [4, 9]. Since there were so few Con A receptors at the ciliary bases, no FITC could be detected in these regions by fluorescence microscopy. The most striking ultrastructural changes of the ectoplasm and cortex that occur after mixing starved complementary mating types are thicken- ing of the EL under the inner-membrane of the alveoli and formation of PJAs [3]. Allewell and others [10] proposed that the co-stimulation period should be distinguished into an activation period and maturation period. Morphologically, the former corresponds to the period of thickening of the EL and the latter to the period of PJA formation. The EL under the inner-alveolus membrane, which has thickened and increased in electron density, is morphologically similar to the EL in the PJA, and may be of similar composition to the latter. Surface areas displaying structural similarities to PJA’s are also found on the side walls of some cytoplasmic ridges in the cell tip near the adoral zone. Numerous ferritin particles are attached as large clusters to the outer surface of these special areas as well as to PJAs. After the co-stimulation period, cells can make contact with each other by forming PJAs. When the cells are in partial contact stage of conjugation, however, broad areas of PJAs around the junctional region of the one partner remain free from contact with PJAs of the other cell. Changes in FITC-Con A binding patterns during the conjugation process were reported by Wata- nabe et al. [11, 12]. The changes in the fluores- cence pattern they observed are similar to those in the F-Con A distribution pattern observed in the present study. The ring pattern of FITC-Con A described by Watanabe ef al. may correspond to the present F-Con A distribution pattern in regions where PJA’s of adjacent cells are not in contact. From previous studies with inhibitors of protein synthesis, a special kind of protein was concluded to be synthesized during the co-stimulation period [13]. This protein was proposed to be a glycopro- tein [14]. Watanabe er a/. [11] found that changes in the Con A binding pattern are stopped or eliminated by cycloheximide. The striking thick- ening of the EL during the activation period may thus reflect an increase in structural protein in the EL during this period. The close relationship between the distribution of Con A receptors on the cell surface and morphological alterations of the EL under the cytoplasmic membrane strongly suggests that the structural protein(s) is bound to the Con A binding glycocalyx on the cytoplasmic membrane. Since ConA receptors are rarely found in starved cells, the interaction between cells during the activation period is unlikely to be mediated by Con A receptors; rather, ConA receptors an- chored to the EL are likely to play some role in adhesion during the maturation period. REFERENCES 1 Bruns, P. J. and Brussard, T. B. (1974) Pair forma- tion in Tetrahymena pyriformis, an inducible de- velopmental system. J. Exp. Zool., 188: 337-344. Bruns, P.J. and Palestine, R.F. (1975) Co- stimulation in Tetrahymena pyriformis. A develop- mental interaction between specially prepared cells. Dev. Biol., 42: 75-83. in) 330 Suganuma, Y., Simode,C. and Yamamoto, H. (1984) Conjugation in Tetrahymena: Formation of a special junction area for conjugation during the co-stimulation period. J. Electron Microsc., 33: 10- 18. Wolfe, J. and Grimes, G. W. (1979) Tip trans- formation in Tetrahymena: A morphogenetic re- sponse to interactions between mating types. J. Protozool., 26, 82-89. Frazier, W. and Glaster, L. (1979) Surface compo- nents and cell recognition. Ann. Rev. Biochem., 48: 491-523. Irimura, T., Nakajima, M., Hirano, H. and Osawa, T. (1975) Distribution of ferritin-conjugated lectins on sialidase-treated membranes of human erythro- cytes. Biochim. Biophys. Acta., 413: 192-201. Ofer, L., Levkovitz, H. and Loyter, A. (1976) Con- jugation in Tetrahymena pyriformis. The effect of polylysine, concanavalin A and divalent metals on the conjugation process. J. Cell Biol., 70: 287-293. Frisch, A., Levkovitz, H. and Loyter, A. (1977) Inhibition of conjugation in Tetrahymena pyriformis by concanavalin A. Binding of concanavalin A to material secreted during starvation and to washed cells. Exp. Cell Res., 106: 293-301. Y.SUGANUMA AND H. 9 10 11 12 13 14 YAMAMOTO Frisch, A. and Loyter, A. (1977) Inhibition of conjugation in Tetrahymena pyriformis by Con A. Localization of Con A-binding sites. Exp. Cell Res., 110: 337-346. Allewell, N. M. and Wolfe, J. (1977) A _ kinetic analysis of the memory of a developmental interac- tion. Mating interactions in Tetrahymena pyriformis. Exp. Cell Res., 109: 15-24. Watanabe,S., Toyohara,A., Suzaki,T. and Shigenaka, Y. (1981) The relation of concanavalin A receptor distribution to the conjugation process in Tetrahymena thermophila. J. Protozool., 28: 171- 17S: Wolfe, J., Pagliaro,L. and Fortune,H. (1986) Coordination of concanavalin-A-receptor distribu- tion and surface differentiation in Tetrahymena. Differentiation, 31: 1-9. Allewell, N. M., Oles, J. and Wolfe, J. (1976) A physicochemical analysis of conjugation in Tetrahy- mena pyriformis. Exp. Cell Res., 97: 394-405. Van Bell, C. T. (1983) An analysis of protein syn- thesis, membrane proteins, and concanavalin A- binding proteins during conjugation in Tetrahymena thermophila. Dev. Biol., 98: 173-181. ZOOLOGICAL SCIENCE 5: 331-336 (1988) Fine Structure of the Dorsal Tongue Surface in the Japanese Toad, Bufo japonicus (Anura, Bufonidae) SHIN-ICHI IWASAKI and KAN KOBAYASHI Department of Anatomy, School of Dentistry at Niigata, The Nippon Dental University, Niigata 951, Japan ABSTRACT—The ultrastructure of the epithelial cells and sensory organs of the dorsal surface of the tongue of Bufo japonicus were investigated by scanning electron microscopy. The specimens were prepared using a method designed to remove the extracellular material which normally adheres to the tongue’s surface. Irregular undulant structures, or ridge-like papillae, which correspond to the filiform papillae of Rana, were compactly distributed over almost all of the dorsal surface of the tongue, while fungiform papillae were scattered amongst these ridge-like papillae. A round sensory disc was located on the top of each fungiform papilla. Latticework, which represented the outline of the boundary of each cell, was visible on the surface of each sensory disc. At higher magnification, we observed that the surface of almost every sensory disc was covered with a honeycomb structure, while a small number of cells with microvilli on their surfaces were scattered amongst them. Each sensory disc was encircled by a thin band of non-ciliated cells. Microridges were widely distributed on the epithelial cell surface of the ridge-like papillae. The observed micro-ornamentation of the lingual structure with its microridges and honeycomb structures may be related to the retention of mucus on the surface of © 1988 Zoological Society of Japan the anuran tongue. INTRODUCTION Filiform and fungiform papillae are distributed on the dorsal surface of the anuran tongue. A round sensory disc is located on the top of each fungiform papilla. Electron microscopic studies of the structure of the sensory discs include those of Graziadei [1], Graziadei and DeHan [2], During and Andres [3] and Gubo et al. [4], all of whom reported that the entire surface of each sensory disc of the tongue is covered with microvilli. However, Jaeger and Hillman [5] described the cytoplasmic ridges of the associated cells of the sensory disc, as well as interspersed cells with microvilli. More recent studies [6, 7] have re- vealed that when mucus is almost completely removed, most of the surface of the sensory disc is covered with a honeycomb structure. It is possible that these conflicting observations reflect inter- specific variations among the anuran species ex- amined. We have attempted to ascertain whether Accepted September 12, 1987 Received June 23, 1987 the honeycomb structure of the sensory disc occurs in the genus Bufo (Bufonidae) as well as in the frogs of the genus Rana (Ranidae). In all but the most recent reports on studies of the anuran tongue [4, 6, 7], the fine structure of the surface of the filiform papillar cells has been neglected. The present study examines the struc- ture of the surface of the ridge-like papillae in Bufo japonicus, since these structures may be analogous to the filiform papillae of Rana. MATERIALS AND METHODS Tongues from four male and three female adult japanese toads, Bufo japonicus, were used in the present study. The toads were perfused from the heart with Karnovsky fixative [8] under anesthesia with MS-222. The tongues were then removed and fixation was continued by immersion in the same solution. After rinsing in 0.1 M cacodylate buffer, several specimens were postfixed in phos- phate-buffered 1% osmium tetroxide solution [9] at 37°C for 2hr and then treated with 8N hydrochloric acid at 60°C for 30 min to remove 332 S. IWASAKI AND K. KOBAYASHI extracellular substances by acid hydrolysis. For scanning electron microscope (Hitachi S-500, S- use as controls, a few specimens were not sub- 800). jected to the postfixation and treatment with acid. All of the specimens were then dehydrated, critical-point dried and coated by gold-ion sputter- ing. Finally, the specimens were examied under a When specimens were postfixed in 1% osmium RESULTS Fic. 1. Scanning electron micrograph of the central dorsal surface of the tongue of Bufo japonicus. Rp: ridge-like papillae, Fu: fungiform papillae. Fic. 2. Ridge-like papillae from Bufo japonicus. Arrows show elevated intercellular borders. Asterisks indicate structures related to mucous secretion. Fic. 3. Higher magnification of polygonal, non-ciliated cells of the ridge-like papilla in Bufo japonicus. Mr: microridges. Arrow shows elevated intercellular borders. Fic. 4. Fungiform papillae from Bufo japonicus. Sd: sensory disc, Rp: ridge-like papillae. Dorsal Tongue Surface of Toad tetroxide and treated with 8 N hydrochloric acid, extracellular substances were almost completely removed. Irregular, undulant structures or ridge-like papillae are distributed in a compact arrangement over the entire dorsum of the tongue, except for its oS ian ye, ea sed ~ TUS AL winners % f kaon 4 ee Be Ks) #, 4 aie ue : ox Soy EGS I a SIRI Fic. 5. The surface of a sensory disc from Bufo japonicus. sae 55 . rf 1 fa peat hs th hn Secs So Zi Hcp ce ce aretronCaNEs af my 4 “S 3 833 anterior margin. Fungiform papillae, 100-150 ~m in diameter, are scattered among the ridge-like papillae (Fig. 1). The ridge-like papillae are 20-50 ~m in width. They are covered with polygonal, non-ciliated cells, the borders of which are elevated (Fig. 2, Hc: honeycomb structure. Arrows indicate microvilli. Fic. 6. Higher magnification of a sensory disc of the tongue of Bufo japonicus. He: honeycomb structure, Mv: microvilli. Fic. 7. Higher magnification of a sensory disc of the tongue of Bufo japonicus, without postfixation with osmium tetroxide and acid treatment. Mu: piled mucus. Fic. 8. Boundary region of a sensory disc and non-ciliated cell. Sd: sensory disc, Nc: non-ciliated cells, Rp: ridge-like papillae. Asterisks indicate structures related to secretion of mucus. 334 S. IWASAKI AND K. KOBAYASHI arrow). Mucus-secreting cells are scattered among these epithelial cells (Fig. 2, asterisks). At higher magnification (Fig. 3), fine plications, or microridges, are densely distributed on the sur- faces of non-ciliated cells. The elevated intercellu- lar borders are composed of bundles of such plications. In specimens which were not postfixed with osmium tetroxide, the surfaces of the ridge- like papillae are obscured by mucus. A sensory disc is located in the central area of the top of each fungiform papilla (Fig. 4). The surface of the disc has a latticework pattern which reflects the boundaries of the cells on the surface of the papilla. At higher magnifications of the Fic. 9. Long ridges in the area near the apex. Ci: ciliated cells. Fic. 10. Higher magnification of the long ridges in the area near the apex. Ci: ciliated cells. Arrow indicates the area without cilia. Fic. 11. Fic. 12. honeycomb structure. Sensory disc (Sd) in the area near the apex. Ci: ciliated cells. Higher magnification of a sensory disc in the area near the apex. Arrows indicate micorovilli. Hc: Dorsal Tongue Surface of Toad 335 sensory disc (Fig. 5), the cell surfaces resemble a honeycomb. Many processes, which are 0.1—0.3 ym long, are recognizable on the honeycomb-like structure. A few cells with microvilli are present among the honeycomb-like cells (Figs. 5 and 6). In specimens not subjected to postfixation and acid treatment, mucus forms a thin covering over the honeycomb framework of the cells (Fig. 7). Each sensory disc is surrounded by non-ciliated cells, which appear to be the same as those that form the ridge-like papillae (Fig. 8). A series of long ridges about 1 mm in width are present on the anterior margin of the dorsal surface of the tongue in parallel with its anterior edge (Fig. 9). The surfaces of the long ridges are covered almost entirely by ciliated cells. Ridge- like papillae composed of non-ciliated cells are not found in this area. Non-ciliated areas are scattered on the surface of these ciliated cells (Fig. 10, arrow). Structures similar to the sensory discs on the top of fungiform papillae occur among the ciliated cells on the surfaces of these ridges (Fig. 11). At higher magnification, these discs appear to be identical to the discs described above (Fig. 12, compare with Fig. 6). DISCUSSION In several earlier reports [1-3], the surface of the sensory discs of the frogs, Rana pipiens, Rana esculenta and Rana temporaria, was described as being extensively covered with microvilli. In contrast, Jaeger and Hillman [5] described the cytoplasmic ridges of the associated cells of the sensory disc, as well as interspersed cells with microvilli in Rana catesbeiana and Hyla arborea. In our specimens of Bufo japonicus from which the mucus was removed, the greater part of the surface of each sensory disc was found to be covered with a latticework pattern similar to that which was demonstrated by us in two species of Rana namely Rana catesbeiana [6] and Rana nigromaculata [7]. As in Rana, most of the surface of each lingual sensory disc of Bufo japonicus is covered with a honeycomb-like texture of cell surfaces, which originate from “associated cells” designated by Graziadei and DeHan [2]. The honeycomb-like texture is completely coincident with the “cyto- plasmic ridges” described by Jaeger and Hillman [5]. In the present study, thin processes were also recognized on the surface of the honeycomb-like textures, just as in the observation by Jaeger and Hillman [5]. They identified these structures as microvilli. However, we feel that these protru- sions are too small to be considered microvilli. Microvilli, located between these “associated cells”, may derive from the “sensory cell” described by Key [10]. The honeycomb-like structures may play a role in the retention of water and other mucous fluid on the surface of the sensory disc. In addition, the possibility that the honeycomb-like structure has the same function as the taste hair is undeniable. On the other hand, it has been shown in a previous study [7] and in the present study that, when the mucus which covers the lingual surface is not completely removed, the remaining mucus may be transformed into various crystal structures during the drying of the speci- mens. which are Among anurans, we were able to recognize some morphological differences in the tongues of two species of Rana [6, 7] and now can compare them to a species of Bufo, Bufo japonicus. In Bufo japonicus, the epithelium formed many ridge-like papillae, and the pores related to the secretion of mucus were not obvious on its surface, while in the two species of Rana, the epithelial surface formed many filiform papillae and there were many pores on its surface [6, 7]. In addition in Bufo japonicus, no ciliated cells were observed on the surface of the ridge-like papillae and on the surrounding areas of the sensory disc, while, in Rana, many ciliated cells were seen on the surface of the filiform papillae and the surrounding areas of the sensory discs. Structures and surface features which were somewhat similar to the ridge-like papillae in Bufo, were shown by Gubo et al. [4] in Bombina variegata. However, they did not de- scribe these papillae in detail. The differences between anurans belonging to the varied genuses may be ascribed to the local differentiation of function of the lingual mucosa. In Rana, microridges were reported to be widely distributed on the non-ciliated cells of the filiform papillae [6, 7]. The present study of Bufo japoni- cus revealed that microridges are also present on 336 S. IWASAKI AND K. KOBAYASHI the non-ciliated cells of the ridge-like papillae. In our study of Rana nigromaculata using transmis- sion electron microscopy [11], it appeared that a large fraction of the filiform papillar epithelial cells had both microridges on the free surface of cells and cellular processes on the surfaces which faced adjacent cells. Thus, the microridges may be the result of an altered pattern of arrangement of these cellular processes, which function as connecting structures between adjacent cells [12]. As sug- gested by Sperry and Wassersug [13], microridges may be important for holding mucus on the surface of the cells. Thus, the lingual microridges in Bufo may function to retain mucus on the dorsal surface of the tongue. REFERENCES 1 Graziadei, P.P.C. (1969) The ultrastructure of vertebrate taste buds. In “Olfaction and Taste”. Ed. by C. Pfaffmann, Rockefeller Univ. Press, New York, pp. 315-330. Graziadei, P. P.C. and DeHan, R. S. (1971) The ultrastructure of frogs’ taste organs. Acta Anat., 80: 563-603. 3 Diring, M. v. and Andres, K. H. (1976) The ultra- structure of taste and touch receptors of the frog’s taste organ. Cell Tissue Res., 165: 185-198. 4 Gubo, G., Lametschwandtner, A., Simonsberger, P. and Adam,H. (1978) Licht- und _raster- Nw nN 10 13 elektronenmikroskopische Untersuchungen an Gau- men und Zunge der Gelbbauchunke, Bombina variegata L. Anat. Anz., 144: 169-178. Jaeger, C. B. and Hillman, D. E. (1976) Morpholo- gy of gustatory organs. In “Frog Neurobiology”. Ed. by R. Linal and W. Precht, Springer-Verlag, Berlin, pp. 588-606. Iwasaki, S. and Sakata, K. (1985) Fine structure of the lingual dorsal surface of the bullfrog. Okajimas Folia Anat. Jpn., 61: 437-450. Iwasaki, S., Miyata, K. and Kobayashi, K. (1986) Studies on the fine structure of the lingual dorsal surface in the frog, Rana nigromaculata. Zool. Sci., 3: 265-272. Karnovsky, M.J. (1965) A _ formaldehyde- glutaraldehyde fixative of high osmolality for use in electron microscopy. J. Cell Biol., 27: 137A-138A. Millonig, G. (1961) Advantages of a phosphate buffer for OsO, solutions in fixation. J. Appl. Physics, 32: 1637. Key, E. A. (1961) Uber die Endigungsweise der Geschmacksnerven in der Zunge des Frosches. Arch. Anat. Physiol. wiss. Med., 28: 329-349. Iwasaki, S., Miyata, K. and Kobayashi, K. (1988) Fine structure of filiform papillar epithelium from the tongue of the frog, Rana nigromaculata. Zool Sci., 5: 61-68. Krsti¢, R. V. (1979) Ultrastructure of the Mamma- lian Cell. Springer-Verlag, Berlin, pp. 238-239. Sperry, D. G. and Wassersug, R. J. (1976) A pro- posed function for microridges on epithelial cells. Anat. Rec., 185: 253-258. ZOOLOGICAL SCIENCE 5: 337-346 (1988) Fine Structure of the Iris Muscle in the Japanese Common Newt, Cynops pyrrhogaster, with Special Reference to Innervation MITSUMASA OKAMOTO Department of Molecular Biology, Faculty of Science, Nagoya University, Nagoya 464, Japan ABSTRACT—The localization and structure of the iridial muscles and associated nerves of the newt (Cynops pyrrhogaster) were examined by electron microscopy with some histochemical studies of catecholamine and acetylcholinesterase. The sphincter muscle, which was composed of pigmented smooth muscle cells, was located circumferentially in the pupillary margin of the iris. Sphincter muscle cells formed occasional close contacts with each other by protruding cellular processes. Numerous gap junctions and desmosomes were observed between anterior and posterior pigment epithelium of the iris, but not between muscle cells. Nerve endings formed varicosities and were composed of agranular vesicles and/or granular vesicles. There was no dilator muscle in this species. Prominent circumferential catecholamine fluorescence was observed in the pupillary margin. Acetyl- cholinesterase positive and negative muscle cells and nerve fibers were found within the sphincter © 1988 Zoological Society of Japan muscle region. INTRODUCTION The iris sphincter and dilator muscles, known to be the iris muscle in the mammalian eye, change the size of the pupil and regulate the quantity of light. Studies on the iris muscles have been a fascinating subject from the standpoint of develop- ment and differentiation of the muscle, because these muscles are unique in that the sphincter muscle is the only vertebrate smooth muscle known to be derived from neuroectoderm [1-4] and the dilator muscle has myoepithelial charac- ters [1, 5-6]. They are innervated by an autonomic nervous system [7-10], which is advantageous, in that we can easily detect the action of nerves through miosis and mydriasis of the eye. In addition to morpholgical studies on the iris muscle [1-6, 11-13], numerous pharmacological and elec- trophysiological reports have been made on this nervous system using various kinds of animals [14— 20]. Studies on the fine structure of the iris muscle besides mammals have also been reported [21-24], Accepted September 24, 1987 Received August 7, 1987 but only a few works have been undertaken on amphibian species [25-27]. On the other hand, it has been well known that the newt has a capacity for regeneration from the mid-dorsal margin of the iris after having extir- pated an intrinsic lens. Numerous investigations at the light and electron microscopic level have been made on the process of lens regeneration (for review, see [28]). However, the iris muscle in the newt has so far received almost no attention in terms of the study of lens regeneration, although the sphincter muscle at the pupillary margin would appear to be closely relevant to lens regeneration. There have been no reports on the fine structure of the iris muscle of the Japanese common newt, Cynops pyrrhogaster and no histochemical works on amphibian iridial muscles and nerves. Thus, in the present study, detailed investigations on the fine structure of the iris muscle and some histo- chemical studies on the iridial muscles and associ- ated nerves were made in the adult Cynops pyrrhogaster. In addition, the present study also seeks to obtain basic information on the iris muscle, which presumably has some relevance to lens regeneration. 338 M. OKAMOTO MATERIALS AND METHODS Electron microscopy Adult Japanese newts, Cynops pyrrhogaster, were kept in an aquarium at 21+2°C. Newts for experiments were decapitated and the isolated dorsal heads were prefixed at 4°C overnight in 3- 6% glutaraldehyde in Hanks’ solution diluted to 80% of the original concentration for newts. After rinsing with the buffer, iridocorneal complexes were isolated and postfixed in cold 1% osmium tetroxide in the buffer for lhr. The tissue fragments were then block stained with 0.5-1% aqueous uranyl acetate solution for 1 hr, washed, dehydrated in graded series of ethanol, and embedded in Epon. Sections were cut with a glass knife or a diamond knife with a Reichert ultrami- crotome, collected on carbon-coated grids, stained with uranyl acetate and lead citrate, then ex- amined by a JEOL 100C electron microscope at 80 KV. Histochemical localization of catecholamines The glyoxylic acid fluorescence technique [29] was employed with a slight modification for the newt iris. The isolated iris rings of the newts were immersed in 2% glyoxylic acid in 0.1-0.2M Sérensen’s phosphate buffer (pH 7.0) for 30 min. Samples were mounted on slides and air dried, then heated for 4min on a hot plate at 100°C. Control samples were treated only with the buffer. The specimens were sealed in liquid paraffin and observed with fluorescence microscope. Demonstration of acetylcholinesterase at the elec- tron microscope level For the demonstration of sites of acetylcho- linesterase activity, the method of Karnovsky and Roots [30] modified by Tsuji [31] was employed with a slight modification for the newt iris. The specimens were fixed in 2% paraformaldehyde and 1.25% glutaraldehyde in 0.05-0.1.M Sé6rensen’s phosphate buffer (pH 7.4) at 4°C for 2-3 hr and washed overnight in the buffer. After rinsing in 0.1M acetate buffer (pH 6.2), the samples were placed in the incubation medium (1.7-8.5 mM acetylthicholine iodide, 0.1 M sodium acetate, pH 6.2, 0.01 M copper sulfate, 0.04M glycine, and 0.04M magnesium chloride) for 3hr at room temperature. Thereafter, the tissues were rinsed twice in the acetate buffer and placed in 3% potasssium ferricyanide solution for 30 min. All incubations were performed in the dark with gyration. After a rinse in the phosphate buffer, the tissues were processed for electron microscopy as previously described. Eserine (physostigmine), an inhibitor of cholinesterases, was used at the concentration of 10~*M to detect the presence of nonspecific esterases. Tissue samples were preincubated for 15 min in an eserine solution, and placed in the incubation mixture containing an inhibitor at the same concentration. RESULTS General features and localization of the iris muscle The iris can be easily recognized as a three- Fic. 1. Fic. 2. present side by side. Fic. 3. basal lamina. Fic. 4. scarcity of pigment granules within the cell. Fic. 5. x 23,000. x 23,000. parallel to each other and to the pupillary margin. A higher magnification of the sphincter muscle region shown in Fig. 1. 2,400. A horizontal section cut slightly obliquely to the iris diaphragm. The sphincter muscle cells are located Meridional section of the iris. The stroma of the iris faces the cornea. Montage. Bar represents 20 ym. Junctions between anterior and posterior pigment epithelium. Both a gap junction and desmosomes are A longitudinal section of the pigment epithelium. The surface of the pigment epithelium is surrounded by a The region is prominent with a x 1,300. Fics. 1-16 are all electron micrographs. Abbreviations: st, stroma; ap, anterior pigment epithelium of the iris; pp, posterior pigment epithelium of the iris; sm, sphincter muscle; g, gap junction; d, desmosome; bl, basal lamina; Nu, nucleus; mf, myofilament; db, dense body; pr, polyribosome; fr, free ribosome; m, mitochondria; pg, pigment granule; ne, nerve ending; J, cell junction; col, collagen fiber; agv, agranular synaptic vesicle; gv, granular synaptic vesicle. Ultrastructure of Iris Muscle in Newt 339 340 M. OKAMOTO layered structure (Fig. 1). The most anterior layer is the stroma of an iris, which contains several cell types of mesenchymal cells and blood vessels. The iridial stroma is rather scanty compared with the other two layers. The next layer is the anterior pigment epithelium. It is laterally continuous with the pigment epithelium of the retina. It comprises a single layer of cells filled with melanin granules which are round to oval in profile, and about 0.60.8 ~m in diameter. The third layer is the posterior pigment epithelium. It is laterally con- tinuous with the neural retina, also comprising a single layer of cells filled with round, relatively large melanin granules, about | ~m in diameter. Numerous gap junctions and desmosomes can be found between the anterior and posterior pigment epithelium (Fig. 2). The anterior and posterior surfaces of the pigment epithelium are surrounded completely by a basal lamina (Fig. 3). The sphinc- ter muscle cells are between the stroma and the anterior pigment epithelium (Fig. 4). They enclose a pupillary margin like a ring and are located parallel to each other and to the pupillary edge (Fig. 5). As the sphincter muscle cells have more scanty pigment granules than the surrounding pigment epithelial cells, their region can be easily recognized in the low magnification electronmicro- graph (Figs. 1, 4 and 5). Fine structure of iris muscle cells and associated nerve fibers The individual sphincter muscle cells contain bundles of myofilaments along the long axis of the cell (Figs. 6 and 7). In cross section of the filament bundles, thin (about 7 nm in diameter) and thick (about 20 nm in diameter) myofilaments can readi- ly be identified (Fig. 8). Dense bodies or dense plaques were scattered throughout the bundles of myofilaments and on the cytoplasmic side of the plasmalemma (Fig.6). Nuclei were slender in outline with many interdigitations along the nu- clear envelope. Mitochondria, free ribosomes and polyribosomes were seen in the vicinity of the conglomerizations of pigment granules and be- neath the cell periphery. Average diameter of the pigment granules in the sphincter muscle cells was 0.5 0.7 um; their size was consistent with those of the anterior pigment epithelium. Many smooth-surfaced vesicles were inter- spersed along the plasmalemma (Figs. 6, 8 and 9). These vesicles have been called caveolae, plas- malemmal vesicles, or micropinocytotic vesicles in the cells of the smooth muscle. In Figure 9, we can easily identify the two different figures of caveolae. In sections tangential to the cell surface, the caveolae were seen in circles, but in vertical sections, they were seen in a flask-shaped invagina- tion attached by a narrow neck region. Each sphincter muscle cell was surrounded by a basal lamina except the spots at which the muscle cells were closely adjacent (Figs.6 and 10). The sphincter muscle cells usually formed close con- tacts with neighboring cells by protruding cytoplas- mic processes. But no gap junctions and desmo- somes were seen between muscle cells, in contrast to the junctions found between the anterior and posterior pigment epithelium. Collagen fiber or bundles of fibers were occasionally present in the A longitudinal section of the sphincter muscle cell. Myofilaments, dense body, pigment granules, caveolae, free and poly ribosomes are found within the cytoplasm. Each muscle cell is in close contacts with cellular Fic. 6. processes. 13,500. Fic. 7. ment is clearly discernible. 37,000. Fic. 8. thin (small arrow) myofilaments are prominent. periphery. 37,000. Fic. 9. sectioned caveolae, respectively. 19,800. Fic. 10. processes. 18,150. Fic. 11. Collagen fiber or bundles of fibers muscles. 66,300. Fic. 12. are no dilator muscles on the anterior pigment epithelium facing the iridial stroma. scattering in the A higher magnification of the myofilament bundle sectioned longitudinally to the bundle. Each myofila- A higher magnification of the myofilament bundle sectioned crossly to the bundle. Thick (large arrow) and Numerous caveolae are also found at the cell Section cut partly tangential and partly longitudinal to the cell surface, showing crossly and longitudinally- The junction between two sphincter muscle cells. They closely contact each other with protruded cellular intercellular spaces of the sphincter Meridional section of the iris situated between the sphincter muscle region and the root of the iris. There 3,960. Ultrastructure of Iris Muscle in Newt 342 M. OKAMOTO intercellular spaces (Fig. 11). Mammalian dilator muscle is knowm to be situated peripheral to the sphincter muscle and to be continuous with the cell bodies of the anterior pigment epithelium [32]. The dilator muscle is therefore a partial specialization of cytoplasmic processes of the anterior pigment epithelium into the myoepithelium. The dilator processes in mammals are arranged in an overlapping manner somewhat like tiles on a roof. In the iris of Cynops pyrrhogaster, the dilator muscle could not be found even in detailed observations in the corre- sponding region of the mammalian dilator muscle (Fig. 12). Large bundles of nerve fibers locating near the muscle cells were found at various places in the sphincter muscle region (Fig. 13). In cross and Fic. 13. agranular synaptic vesicles. 29,500. Fic. 15. A nerve varicosity of agranular type. vesicles. 25,000. Fic. 16 A nerve varicosity of granular type. Agranular and granular vesicles are mixed in a varicosity. Section showing dense innervation in the sphincter muscle region. longitudinal section of nerve fibers, microtubules and neurofilaments were seen along the long axis of the fibers. Nerve endings were found close to the muscle cells (Fig. 14). A cleft of about 5-10 nm between nerve membrane and muscle mem- brane was usually found, but in most examples, the area of contact between nerve endings and muscle cells did not show junctional specializations. Nerve endings which usually form varicosities contain agranular (40-60 nm in diameter) and/or granular (90-130 nm in diameter) synaptic vesicles (Figs. 15 and 16). Varicosities were roughly clas- sified into two types. One is composed mostly of agranular vesicles of rather uniform diameter except the occasional existence of only few granu- Another is composed of granular, dense-cored vesicles and agranular, clear vesicles. lar vesicles. x 6,435. Fic. 14. Adjacent contact of a nerve ending and a sphincter muscle cell. The nerve varicosity contains numerous Almost all of the synaptic vesicles are composed of agranular x 25,000. Ultrastructure of Iris Muscle in Newt 343 Fic. 17. Histochemistry of catecholamine in the pupil- lary margin. a, Fibrous catecholamine fluorescence is found circumferentially at the pupillary margin. b, No fluorescence at the pupillary margin is found in the buffer-treated sample. 178. Localization of catecholamine and demonstration of acetylcholinesterase activities in iris muscle The fluorescence micrographs of pupillary mar- gin of the iris treated with glyoxylic acid are shown in Figure 17a. Fluorescence was prominent in fibrous structure circumferentially located at the pupillary margin. The fibers showed a knot-like structure in some places. No fluorescence was detectable in the buffer-treated samples (Fig. 17b). In samples treated with acetylthiocholine as the substrate of acetylcholinesterase, the enzymat- ic reaction product was observed randomly in the surface of the sphincter muscle cells (Fig. 18a). The precipitate also appeared randomly at nerve fibers and vesicle-filled varicosities. In addition to the presence of the acetylcholinesterase-positive cells and nerve fibers, there were some negative cells and nerve fibers associated with the surround- ing muscle cells (Fig. 18b). They appeared to have the same features as eserine-treated samples (Fig. 18c). DISCUSSION Present study firstly demonstrated the localiza- tion and the detailed structure of the iris muscle in the Japanese common newt, Cynops pyrrhogaster. Fic. 18. sphincter muscle region. positive muscles and nerve fibers are present in the samples treated with acetylthiocholine as the sub- Histochemistry of acetylcholinesterase in the a, Acetylcholinesterase- strate. b, Acetylcholinesterase-negative muscles and nerve fibers are also found in the sample tre- ated with the substrate. c, No reaction product is found in the eserine-treated samples. 7,750. In addition, some histochemical studies on the localization of catecholamines and_acetylcho- linesterases were made in the iris muscle of the newt. The sphincter muscle is usually classified as the smooth muscle, even if its developmental 344 M. OKAMOTO origin is very different from ordinary smooth muscle of mesodermal origin [32]. In contrast to the mammalian sphincter muscle, the iris muscle in the newt appears to maintain some characters of the pigment epithelium of the iris by including a considerable number of cytoplasmic pigment gran- ules within the muscle cells even in adult age. The persistence of a considerable number of pigment granules within the muscle cells and the lack of the dilator muscle in Cynops pyrrhogaster are consis- tent with results in other amphibian species [25- 27]. These results suggest an incomplete dif- ferentiation of the iris muscle in the amphibian eye. The present results showed that the sizes of the pigment granules in the sphincter muscle cells were similar to those of the anterior pigment epithe- lium. But, in the American newt, Taricha torosa, the sizes of the pigment granules were similar to those of the posterior pigment epithelium [26]. Based on this observation, Tonosaki and Kelly [26] proposed the notion that the sphincter muscle was derived from the posterior pigment epithelium. On the other hand, in the grass frog, Rana pipiens, the pigment granules of the muscle cells were of an intermediate size of the anterior and the posterior pigment epithelium [27]. Thus, it seems not to be adequate to amplify the notion about the origin of the sphincter muscle obtained in Taricha torosa to all amphibian species. It has been generally accepted in the mammalian eye that the iris sphincter or dilator muscles are innervated by excitatory cholinergic or adrenergic nerve fibers, respectively, and miosis or mydriasis is the result of contraction of these nerve fibers [33]. Then the question arises as to how miosis or mydriasis is caused in the newt eye which has no dilator muscle. Some suggestions will be provided in the present results and in those previously reported by others [12, 25]. Two types of varicosi- ties were observed in nerve endings in the present study. Richardson [12] has described two types of nerve endings in the iris muscle of the rabbit, the first containing numerous small, agranular vesicles with an occasional large dense-cored vesicle, the second also with agranular vesicles, but mixed with a large number of dense-cored vesicles of two different types. He suggests that the first, associ- ated with the sphincter, may be typical of cho- linergic innervation; the second, associated with the dilator, typical of adrenergic innervation. The features of nerve endings found in the iris muscle of the rabbit were very similar to those in Cynops pyrrhogaster. However, in the Cynops, there is no dilator muscle. Armstrong and Bell [25] found that the toad possesses no dilator muscle and the application of noradrenaline or sympathetic nerve stimulation causes pupillary dilation and acetylcholine or parasympathetic nerve stimulation produced pupillary constriction. From these results, they concluded that the sphincter muscle in the toad has a dual innervation. Based on the pharmacological and electrophysiological works, a dual innervation of the mammalian sphincter and dilator muscle has been postulated in various species including cat, rat, bovine, dog and human [14-20]. The present results may be explained by dual innervation of the sphincter muscle as in the toad. But it is also probable to postulate that both the sphincter and the dilator muscle are situated so to be mixed in the so-called mammalian “sphincter region.” The existence of acetylcholinesterase-positive and -negative muscle cells and nerve fibers in the “sphincter region” may support this idea. Because a dual innervation in only one type of the muscle makes it difficult to explain the mixed existence of acetylcholinesterase-positive and -negative mus- cles. Further examinations of the newt iris muscle will provide additional data to the above two possibilities. However, it is said that acetylcho- linesterase is not always an appropriate marker for the cholinergic nerve [34] and it has been ques- tioned that cholinergic and adrenergic nerve pro- files can be identified by the morphological types of synaptic vesicles [35]. Thus, it seems necessary to further examine the autonomic nerves by immunoelectron microscopic studies using anti- bodies against the neurotransmitters or related enzymes to the metabolism of nervous system such as tyrosine hydroxylase and cholineacetyltrans- ferase [34, 36]. Finally, as for relevance to lens regeneration, the present results disclosed the detailed structure of the iris muscle in the non-operated eyes before lentectomy. Thus, the behavior of the iris muscle Ultrastructure of Iris Muscle in Newt during lens regeneration is now under investiga- tion, because it would be interesting to know whether the differentiated iris muscle cells situated at the dorsal marginal iris can be transformed into lens cells. ACKNOWLEDGMENT The author wishes to express his gratitude to Dr. Terumasa Komuro of Ehime University for his warm encouragement and helpful discussion. 10 11 REFERENCES Tamura, T. and Smelser, G. K. (1973) Develop- ment of the sphincter and dilator muscles of the iris. Arch. Ophthalmol., 89: 332-339. Ruprecht, K. W. and Wulle, K. G. (1973) Licht- und elektronenmikroskopische Untersuchungen zur Entwicklung des menschlichen Musculus sphincter pupillae. Arbrecht v. Graefes Arch. klin. exp. Ophthalmol., 186: 117-130. Lai, Y.-L. (1972) The development of the sphincter muscle in the iris of the albino rat. Exp. Eye Res., 14: 196-202. Imaizumi, M. and Kuwabara, T. (1971) Develop- ment of the rat iris. Invest. Ophthalmol., 10: 733- 744. Lai, Y.-L. (1972) The development of the dilator muscle in the iris of the albino rat. Exp. Eye Res., 14: 203-207. Kelly, R. E. and Arnold,J.W. (1972) Myofila- ments of the pupillary muscles of the iris fixed in situ. J. Ultrastruct. Res., 40: 532-545. Laties, A. and Jacobowitz, D. (1964) A histochem- ical study of the adrenergic and cholinergic innerva- tion of the anterior segment of the rabbit eye. Invest. Ophthalmol., 3: 592-600. Csillik, B. and Koelle, G. B. (1965) Histochemistry of the adrenergic and cholinergic autonomic in- nervation apparatus as represented by the rat iris. Acta Histochem., 22: 350-363. Malmfors, T. (1965) The adrenergic innervation of the eye as demonstrated by fluorescence micro- scopy. Acta Physiol. Scand., 65: 259-267. Ivens, C., Mottram, D.R., Lever, J.D., Presley, R. and Howells, G. (1973) Studies on the acetyl- cholinesterase(Ache)-positive and -negative auto- nomic axons supplying smooth muscle in the normal and 6-hydroxydopamine(6-OHDA) treated rat iris. Z. Zellforsch., 138: 211-222. Tousimis, A. J. and Fine, B. S. (1959) Ultrastruc- ture of the iris: an electron microscopic study. Am. J. Ophthalmol., 48: 397-417. 12 14 19 345 Richardson, K. C. (1964) The fine structure of the albino rabbit iris with special reference to the identification of adrenergic and cholinergic nerves and nerve endings in its intrinsic muscles. Am. J. Anat., 114: 173-205. Gabella, G. (1974) The sphincter pupillae of the gunea-pig: structure of muscle cells, intercellular relations and density of innervation. Proc. R. Soc. Lond., B, 186: 369-386. Narita, S. and Watanabe, M. (1981) Response of the isolated rat iris sphincter to cholinergic and adrenergic agents and electrical stimulation. Life Sci., 29: 285-292. Narita, S. and Watanabe, M. (1982) Response of isolated rat iris dilator to adrenergic and cholinergic agents and electrical stimulation. Life Sci., 30: 1211-1218. Schaeppi, U. and Koella, W. P. (1964) Innervation of cat iris dilator. Am. J. Physiol., 207: 1411-1416. Schaeppi, U. and Koella, W. P. (1964) Adrenergic innervation of cat iris sphincter. Am. J. Physiol., 207: 273-278. Suzuki, R., Oso, T. and Kobayashi, S. (1983) Cho- linergic inhibitory response in the bovine iris dilator muscle. Invest. Ophthalmol. Vision Sci., 24: 760- 76S. Yoshitomi, T., Ito, Y. and Inomata,H. (1985) Adrenergic excitatory and cholinergic inhibitory innervations in human iris dilator. Exp. Eye Res., 40: 453-459. Yoshitomi, T. and Ito, Y. (1986) Double reciprocal innervations in dog iris sphincter and dilator mus- cles. Invest. Ophthalmol. Vision Sci., 27: 83-91. Gabella, G, and Clarke, E. (1983) Embryonic de- velopment of the smooth and striated musculatures of the chicken iris. Cell Tissue Res., 229: 37-59. Oliphant, L. W., Johnson, M. R., Murphy, C. and Howland, H. (1983) The musculature and pupillary response of the great horned owl iris. Exp. Eye Res., 37: 583-595. Kuchnow, K. P. and Martin, R. (1970) Fine struc- ture of elasmobranch iris muscle and associated nervous structures. Exp. Eye Res., 10: 345-351. Reger, J. F. (1966) The fine structure of iridial constrictor pupillae muscle of Alligator mississip- piensis. Anat. Rec., 155: 197-216. Armstrong, P. B. and Bell, A. I. (1968) Pupillary responses in the toad as related to innervation of the iris. Am. J. Physiol., 214: 566-573. Tonosaki, A. and Kelly, D. E. (1971) Fine struc- tural study on the origin and development of the sphincter pupillae muscle in the west coast newt (Taricha torosa). Anat. Rec., 170: 57-74. Nolte, J. and Pointner, F. (1975) The fine structure of the iris of grass frog, Rana pipiens. Cell Tissue Res., 158: 111-120. 28 29 30 31 32 346 Reyer, R. W. (1977) The amphibian eye, develop- ment and regeneration. In “Handbook of Sensory Physiology. VII/S5: The Visual System in Verte- brates”. Ed. by F. Crescitelli, Springer-Verlag, Ber- lin, pp. 309-390. Lindvall, O. and Bjorklund, A. (1974) The glyox- ylic acid fluorescence histochemical method: a de- tailed account of the methodology for the visualiza- tion of central catecholamine neurons. Histochemis- try, 39: 97-127. Karnovsky, M. J. and Roots, L. (1964) A direct- coloring thiocholine method for cholinesterases. J. Histochem. Cytochem., 12: 219-221. Tsuji, S. (1974) On the chemical basis of thiocholine methods for demonstration of acetylcholinesterase activities. Histochemistry, 42: 99-110. Fine, B. S. and Yanoff, M. (1979) Ocular Histolo- gy, Harper & Row Publishers, New York, 2nd ed., M. OKAMOTO 33 34 pp. 203-210. Newell, F. W. (1986) Ophthalmology: Principles and Concepts, C. V. Mosby Company, St. Louis, 6th ed., pp. 274-282. Pourcho, R. G. and Osman, K. (1986) Acetylcho- linesterase localization in cat retina: a comparison with choline acetyltransferase. Exp. Eye Res., 43: 585-594. Komuro, T., Baluk, P. and Burnstock, G. (1982) An ultrastructural study of nerve profiles in the myenteric plexus of the rabbit colon. Neuroscience, 7: 295-305. Thind, K. K. and Goldsmith, P.C. (1986) Ultra- structural analysis of synapses involving tyrosine hydroxylase-containing neurons in the ventral periventricular hypothalmus of the macaque. Brain Res., 366: 37-52. ZOOLOGICAL SCIENCE 5: 347-351 (1988) © 1988 Zoological Society of Japan In vitro Dimerization of I-protein, an A-I Junctional Component of Skeletal Muscle Myofibrils' Makoto TSUNEOKA*, KoscAK MARUYAMA and Kazuyo OHASHE Department of Biology, Faculty of Science, Chiba University, Chiba 260, Japan ABSTRACT—Chicken myofibrillar I-protein, which was purified using ammonium sulfate precipita- tion and DEAE-cellulose column chromatography, was separated into two fractions by gel filtration, disc alkaline electrophoresis, or SDS polyacrylamide gel electrophoresis without SH reagents. These fractions consisted of 100,000 dalton and 50,000 dalton components. The amount of the high molecular weight component increased under the oxidizing conditions, while the amount of the low one increased when SH reagents were added. On the other hand, antiserum raised against 50,000 dalton component reacted with both of them, as revealed by immunoelectrophoresis. Therefore, it is concluded that I-protein dimerizes under oxidizing solutions. However, dimeric I-protein did not inhibit the ATPase activity of actomyosin in vitro, whereas monomeric I-protein did. INTRODUCTION I-protein is a myofibrillar protein which was isolated from chicken and rabbit striated muscles in 1977 [1-3]. The apparent molecular weight of I-protein, which was estimated from the migration rate of SDS polyacrylamide gel electrophoresis, is approximately 50,000 [1]. This protein is localized at the A-I junctional region of myofibrils in fresh myofibrils [4] and inhibits the ATPase activity of actomyosin in vitro [2]. In the process of chicken I-protein purification, we often found the coexistence of 100,000 dalton protein with I-protein in the I-protein fractions. We examined the form of I-protein in various solutions and found that the 100,000 dalton protein is a dimeric form of I-protein. It was observed that the inhibitory effect on actomyosin ATPase activ- Accepted October 20, 1987 Received August 11, 1987 This work was supported by Grant-in-Aid for Scien- tific Research from the Ministry of Education, Science and Culture of Japan. Present address: Department of Physiology, Kansai Medical University, Fumizono-cho, Moriguchi, Osaka 570, Japan. To whom reprint requests should be addressed. a N w ity of these two forms of chicken I-protein were different. MATERIALS AND METHODS Preparation of I-protein and antiserum against I-protein I-protein was prepared from chicken breast muscle according to the method described in a previous paper [1], using a DEAE-cellulose col- umn. Antiserum against chicken I-protein was raised in a rabbit as mentioned before [4]. Electrophoresis Disc alkaline electrophoreses of Tris-glycine buffer (pH 8.8) system were performed according to the method of Davis [5], using 10% acrylamide gel as separating gels (375 mM Tris-HCl, pH 8.8) and 4% acrylamide gel as stacking gels (125 mM Tris-HCl, pH 6.8). SDS polyacrylamide gel elec- trophoreses were carried out essentially according to Weber and Osborn [6], using 10% acrylamide gels with or without 2-mercaptoethanol (2-ME). Determination of protein concentrations Protein concentrations were determined by 348 M. TSUNEOKA, K. MARUYAMA AND K. OHASHI means of biuret reaction or estimated from ultra- violet absorption at 280nm and 260nm in a Shimadzu spectrophotometer. Immunoelectrophoresis An immunodiffusion test of anti-I-protein anti- serum against disc electrophoresed I-protein was carried out according to the method of Matsuda ert al. [7]. A disc gel, on which I-protein was elec- trophoresed, was put on a slide glass and then 3 ml of 1% melted agarose containing 0.15 M NaCl and 20mM_ sodium phosphate buffer, pH7.2 was poured onto the slide glass. It was left for 30 min at room temperature. The agarose gel was grooved along the electrophoresed gel. Antiserum against I-protein was poured into the groove. This slide glass was placed in a moisture box overnight at room temperature. After dipped into PBS so as to remove soluble proteins, the slide glass was stained by 1% Amido Black in 7% acetic acid for 1 hr and subsequently destained by 4% acetic acid. Oxidization and reduction of I-protein Oxidized and reduced I-proteins were prepared according to the method described previously by Stewart [8]. I-protein was left for 2 hr at room temperature under an oxidizing condition: 1M o A280 0.1 O 50 NaCl, 25 mM sodium borate buffer (pH 9.3), and 25 mM CuCh, or a reducing condition: 1 M NaCl, 25mM sodium borate buffer (pH9.3), 5mM dithiothreitol (DTT). I-protein was also incubated in the same solution without CuCl, and DTT as control. Protein concentrations were finally ad- justed to 0.2 mg/ml. ATPase measurements The ATPase activity of actomyosin was deter- mined by measuring the amount of inorganic phosphate (Pi) liberated by the method of Taussky and Shorr [9]. The standard reaction mixture consisted of 43mM KCl, 1mM MgCl, 1mM ATP, and 10mM Tris-HCl, pH 7.5. Incubation was carried out for 5-20 min at room temperature. The specific activity was given as moles of Pi split per mg of myosin per minute. RESULTS Monomeric and dimeric I-protein molecules coex- isted in a solution When a DEAE-cellulose column purified I- protein fraction [1] shown in Figure 1,a was applied onto the Sephadex G-—200 column equi- a ib 'cayG se 100 150 200 Elution volume (ml) Fic. 1. Gel filtration chromatography of I-protein. Ion-exchange column purified I-protein was applied onto Sephadex G-200 column (1.890 cm). Fractionated proteins were elec- trophoresed by disc alkaline electrophoresis (a, b, c), or SDS-polyacrylamide gel elec- trophoresis with 2-ME (d, e). a; applied sample. b, d; protein eluted at 80 ml. c, e; protein eluted at 105 ml. Arrows indicate the electrophoresed fractions of b, d and ¢, e. Dimerization of I-protein 349 librated with 20 mM Tris-HCl, pH 7.5 and eluted with the same solution, the elution profile showed two peaks (Fig. 1). The first peak contained a protein of relatively low mobility by disc alkaline electrophoresis (Fig. 1, b) and a high mobility protein was eluted in the second peak (Fig. 1, c). These proteins in both peaks were electrophoresed with the same mobility on SDS polyacrylamide gels with 2-ME (Fig. 1, d, e). The ratio of the first peak to the second one calculated from the peak size was 5 : 8. Two protein bands on SDS polyacrylamide gels were electrophoresed at the positions of 100,000 dalton and 50,000 dalton proteins without any SH reagent (data are not shown). A small amount of two unidentified proteins, which molecular sizes were similar to that of 100,000 dalton protein, were contained in the first peak (Fig. 1, d). Anti-I-protein antiserum reacted with both of the high and low mobility proteins DEAE-cellulose column purified I-protein was electrophoresed on an alkaline disc gel. The gel was embedded in 1% agarose gel and reacted with anti-I-protein antiserum as described in Materials and Methods. A confluent immunoprecipitin line formed along the gel shows that the antiserum strongly reacted with both the high and the low mobility proteins of the same antigenicity (Fig. 2). The ratio of the large molecular sized protein to the small one under a reducing condition differed from that under an oxidizing condition Ion exchange column purified I-protein was incubated under oxidizing, reducing, or control conditions. The samples were electrophoresed in the system of Weber and Osborn without SH reagent. Each protein band stained with Coomas- sie Brilliant Blue R 250 was cut out and incubated overnight in 3 ml of 0.2% SDS and 0.2 M sodium phosphate buffer, pH 7.4, so as to extract the dye from acrylamide gel. The ratios of the large molecular sized protein were roughly estimated by measuring the absorbance of the dye solutions at 600 nm in a spectrophotometer. The sample left under an oxidizing condition contained more large molecular sized protein than the control sample. In the case of Figure 3, the ratio of the large Fic. 2. Immunoelectrophoresis. | Immunoelectropho- resis was carried out using 100 ug of I-protein and anti-I-protein antiserum. a, immunoelectrophoretic profile. The arrow head shows an immunoprecipi- tin line. b, disc alkaline electrophoresis of I- protein. molecular sized protein to the small one under an oxidizing condition was approximately 3: 2 (Fig. 3, b), while the ratio of the control sample was 1:4 (Fig. 3, a). On the other hand, the sample under the reducing condition mainly contained the small molecular sized protein and did little large molecu- lar sized one (Fig. 3, c). These two protein bands were electrophoresed at the position of 100,000 dalton and 50,000 dalton proteins. Therefore, it was concluded that these proteins were dimeric and monomeric I-proteins. Dimeric I-protein possessed little inhibitory action on actomyosin ATPase activity The inhibitory effects of I-protein in both 350 M. TSUNEOKA, K. MARUYAMA AND K. OHASHI Fic. 3. SDS-polyacrylamide gel electrophoresis of oxi- dized and reduced I-protein. a, control I-protein. b, I-protein under an oxidizing condition. c, I- protein under a reducing condition. les Pi i Hinieles i/mg/min nN 0.1 0 5 10 15 20 |-protein/Myosin(%) Fic. 4. The effect of dimeric or monomeric I-protein on the actomyosin ATPase. Ion-exchange column purified I-protein (©), dimeric I-protein (a), or monomeric I-protein (@) was added to the actomy- osin solution. The actomyosin ATPase was meas- ured described in Materials and Methods. dimeric and monomeric forms on actomyosin ATPase activity were examined (Fig. 4). Sephadex G-—200 column fractionated dimeric and monomeric I-proteins (Fig. 1) were used. Recon- stituted actomyosin was used. Final concentra- tions of myosin and actin were 0.8 mg/ml and 0.4 mg/ml, respectively. I-protein was added in va- rious ratios to myosin. Dimeric I-protein hardly affected the actomyosin ATPase activity. Although the inhibitory action was lower than that of an original I-protein sample without gel- filtration, monomeric I-protein inhibited the actomyosin ATPase activity. Some denaturation of the protein during purification may account for the low inhibitory action of monomeric I-protein. DISCUSSION Ion exchange column purified I-protein usually containd both dimeric and monomeric molecules and the ratio of the two varied in each preparation. Dimeric I-protein usually increased with the lapse of time after preparation, even when stored at 0°C. In the present study, we showed that oxidization can account for the dimerization of I-protein. Dimerized I-protein was stable in a solution. Even an SDS solution containing 2—ME could not always dissolve the dimers into monomers in a short period incubation (data are not shown). However, the condition under which all monomeric I-protein molecules dimerize could not be specified. Tropomyosin is a myofibrillar protein which also forms a dimer in a solution through disulfide bond [8]. Figure 3 shows that an oxidized I-protein solution contains more dimeric form and a reduced I-protein solution does more monomeric form. These results suggest that I-protein also forms a dimer through disulfide bond. It is assumed that I-protein may exist in a monomeric form under cellular circumstances. As is shown in Figure 4, dimeric I-protein lost the effect on the actomyosin ATPase activity while monomeric I-protein pre- served. I-protein is localized at the A-I junctional region of myofibrils [4] and bind to myosin filaments in vitro [3]. These suggest that monomeric I-protein molecules bind to the both ends of thick filaments of myofibrils. Dimerization of I-protein 3 REFERENCES Ohashi, K., Kimura, S., Deguchi, K. and Maruyama, K. (1977) I-protein, a new regulatory protein from vertebrate skeletal muscle I. Purification and charac- terization. J. Biochem., 81: 233-236. Ohashi, K., Masaki, T. and Maruyama, K. (1977) I-protein, a new regulatory protein from vertebrate skeletal muscle II. Localization. J. Biochem., 81: 237-242. Maruyama, K., Kunitomo,S., Kimura,S. and Ohashi, K. (1977) I-protein, a new regulatory protein from vertebrate skeletal muscle III. Function. J. Biochem., 81: 243-247. Ohashi, K., Tsuneoka, M. and Maruyama, K. (1985) I-protein is localized at the junctional region of A-bands and I-bands of chicken fresh myofibrils. Nn — J. Biochem., 81: 1323-1328. Davis, B. J. (1964) Disc electrophoresis — II. Method and application to human serum proteins. Ann. N.Y. Acad. Sci., 121: 404-427. Weber, K. and Osborn, M. (1969) The reliability of molecular weight determination by dodecyl sulfate- polyacrylamide gel electrophoresis. J. Biol. Chem.., 244: 4406-4412. Matsuda, R., Obinata, T. and Shimada, Y. (1982) Types of troponin components during development of chicken skeletal muscle. Dev. Biol., 82: 11-19. Stewart, M. (1975) Tropomyosin: Evidence for no stagger between chains. FEBS Letters, 53: 5-7. Taussky, H. H. and Shorr, E. (1953) A microcolor- imetric method for the determination of inorganic phosphorus. J. Biol. Chem., 203: 675-681. i Ae MaTBRE Ye ih a sadn t 2 a, ‘ats i. - light ie = 7 . 7 oy ¥% sviet, LAW GEON, os ian Tey, | bie emepaagyE ee ‘ ‘ F y Tne Aye my fw at iii) pie "he ey zi a eae a” a al | ne er a ’ hehe iAnev 442 A . er sited wath Ay ceepngeye aye ‘i K vherveta, = 3 SO | ere evi wid _ iy ZOOLOGICAL SCIENCE 5: 353-373 (1988) Oogenesis in the Medaka Oryzias latipes —Stages of Oocyte Development TAKASHI IWAMATSU, TADAYUKI OHTA, Emiko OsHma! and NorivosH! SAKAI Department of Biology, Aichi University of Education, Kariya 448, ‘Department of Breeding Research, Faculty of Agriculture, Nagoya University, Nagoya 464, and *Laboratory of Reproductive Biology, National Institute for Basic Biology, Okazaki 444, Japan ABSTRACT—An ovary of Oryzias latipes contains developing oocytes of various sizes and mor- phology during the breeding season. Formation of the egg membrane (chorion), changes in cell organelles and the volume of the developing oocyte as well as changes in follicle cells were investigated by ordinary light and electron microscopy. Basic morphological distinctions were primarily used to prepare a table of the developmental stages of Oryzias latipes oocytes. In addition to these distinctions, differences in proteins analyzed by SDS-polyacrylamide gel electrophoresis and the capacity for steroid production assayed by radioimmunoassay were examined in follicles at different stages. On the basis of these results, oogenesis is grossly classified into five phases (early and late previtellogenic phases, early and late vitellogenic phases and postvitellogenic phase) which are © 1988 Zoological Society of Japan divided in detail into ten stages (Stage I-X) of oocyte development. The present classification of developing oocytes was compared with early ones by other investigators. INTRODUCTION In teleost fishes, many investigations on fine structures of the oocyte during oogenesis have been reviewed [1-3]. In the medaka, Oryzias latipes, developing oocytes have already been investigated [4-17]. Since these investigators frag- mentally described developing oocytes with dif- ferent viewpoints toward oogenesis, they have not always provided sufficient details on the defined developmental stages of oocytes that are required for various physiological and biochemical studies [18, 19] of medaka oocytes. In order to establish a standard table of the developmental stages of medaka oocytes that can be used for cytological, physiological and _ bio- chemical studies, we investigated the characteris- tics of live and fixed follicles by light and electron microscopy and by a few biochemical methods, we also reviewed briefly the cytological and histologi- Accepted August 20, 1987 Received July 8, 1987 cal observations by early investigators. MATERIALS AND METHODS Adult medakas, Oryzias latipes (orange-red type), used in the present investigation were obtained from a fish farm in Yatomi (Aichi Pref. Japan). They were kept in glass aquaria under artificial reproductive conditions (light 14 hr, 26- 28°C) and fed mixed powders of shrimps and roasted wheat-grains at least three times a day. To test the ability of large oocytes to mature in response to hormone stimulation, intrafollicular oocytes isolated as described elsewhere [20] were incubated for 10 hr (26°C) in culture medium (90% Earle’s medium 199, Dainippon-seiyaku, Osaka, Japan) containing 17a, 208-dihydroxy-4-pregnen- 3-one (17a, 208-diOHprog, 100 ng/ml). Ten to fifteen oocytes were incubated in 5 ml of culture medium in a sterilized glass Petri-dish. At the end of incubation, maturation of oocytes was deter- mined by observing the breakdown of the germinal vesicle (GVBD) and the responsiveness to insemi- 354 T. Iwamatsu, T. Onta et al. nation. In the present study, the diluted Earle’s medium 199 was supplemented with 30 mg/l peni- cillin G potassium (Meiji-seika, Tokyo, Japan) and 60 mg/ml streptomycin sulfate (Meiji-seika) and adjusted to pH 7.4 with 1 M NaHCO. For light microscopic observations, some ovaries were fixed for 3-12 hr with Bouin’s fixative or glutaraldehyde dissolved in regular saline buffered to pH7.3 (4°C) with 0.1M phosphate buffer. Fixed samples were dehydrated in a graded etha- Paraffin sections 7 um in thickness were stained with Delafield’s haematoxylin. For transmission elec- tron microscopy, oocytes were fixed with modified Karnovsky’s fixative (pH 7.3) for 12 hr and post- fixed in 1% solution of buffered (0.1 M phosphate buffer, pH 7.3) osmium tetroxide for 2 hr (4°C). These samples were rinsed in 0.1 M_ phosphate buffer and then embedded in epoxy resin. Exami- nation of ultrathin sections stained with uranyl nol series and embedded in paraffin. acetate and lead citrate was conducted with a JEM-100 B and a JEM-100 CX electron micro- scopes. Fixed and dehydrated samples shrank as in Table 1. Follicles (80-90 per sample) of various sizes were homogenized in 10 mM Tris-HCl buffer (pH 6.8) at a ratio of 35 volumes of buffer to one volume of follicle or egg with 10 strokes of a teflon microhomogenizer. The homogenate was incu- bated for 3 min at 60°C, and was centrifuged at 1,000 g for 10 min (room temp.). The supernatant was mixed with an equal volume of the sample buffer containing 2% SDS, #£-mercaptoethanol, 20 mM Tris-HCl (pH 6.8) and 40% glycerin and incubated for 10 min at 60°C. It was then analyzed by SDS-—polyacrylamide gel electrophoresis (SDS-PAGE). The protein sam- ples (about 25 ug) prepared above were loaded on polyacrylamide gels (12%) containing 0.1% SDS, 0.0025% fresh ammonium persulfate, 0.00025% riboflavin, and 0.05% Temed. Gels were run at a constant 10 A until the tracking dye (bromophenol blue) had reached the bottom of the gel, which took approximately 6.5 hr (20°C). A common electrophoresis buffer (0.1% SDS-0.05 M _ Tris— 0.38 M glycine) was used. The gels were fixed for 10 hr in 45% methanol-10% acetic acid, stained for 6 hr with 0.05% Coomassie brilliant blue R- 250 dissolved in the above solution, and destained by diffusion in 5% methanol-10% acetic acid for 6-12 hr. Molecular weights were determined after SDS-PAGE comparing log-relative electropho- retic mobilities (Log-Rm) of standard proteins as references: carbonic anhydrase, ovalbumin, albu- min, phosphorylase, $-galactosidase and myosin (a MW-SDS-200 Kit from Sigma Chemicals, St. Louis, MO., USA). 17a, 208-DiOHprog and estradiol-17 (E>) were measured in conditioned culture medium (15 follicles/ml) with or without PMS (pregnant mare serum gonadotropin: Serotropin, Teikoku-zoki, Tokyo) by radioimmunoassay (RIA), as described in detail by Kagawa et al. [21] and Young et al. [22]. The samples of culture medium were assayed after extraction with five volumes of diethyl ether for 30 sec. The anti-E, and anti-17a, 208-diOHprog sera were provided by Dr. Y. Nagahama. The level of cross reactivity of anti-E, serum, with E5, estradiol-17a, estrone, estriol and testosterone was 100.0, 0.80, 3.20, 1.77 and 0.29%, respectively. The anti-17a, 20-diOHprog serum was highly specific, cross-reacting less than 0.01% with most of a wide range of ovarian steroids such as progesterone (4*-pregnen-3, 20-dione), —17a- hydroxy-4-pregnen-3, 20-dione and 20/-hydroxy- 4-pregnen-3-one. Only 17a, 20(-dihydroxy-5/- pregnan-3-one (2.4%) showed a cross reactivity above 1%. (2, 4, 6, 7-7H)E, and (1, 2, 6, 7-7H)17a, 208-diOHprog were used as antigens. In the present radioimmuno-assay system, both E, and 17a, 20-diOHprog standards of 20 pg/ml could be believably distinguished from the buffer blank with TaBLeE 1. Change in size of follicles due to dehydration after fixation Diameter (um) of live follicles <100 200 300 400 500-700 800-900 1100-1200 Bouin’s fixative < 66 152 267 356 = 450-630 ~=— 720-810 880-960 Modified Karnovsky’s fixative <5) 164 264 364 460-644 728-818 979-1068 Numbers in the table indicate diameter (um) of dehydrated follicles after fixation. Oogenesis in the Medaka 355 a 98% confidence limit, but for practical purposes, portions of media reading less than 30 pg/ml were considered to have nondetectable levels. RESULTS In a medaka during the reproductive season, an ovary contains oocytes in all phases of oogenesis (Fig. 1). Oocytes can be grossly divided into five phases and further into ten stages, based on major morphological characteristics of developing oocytes and follicles (Fig. 2). Figure 3 diagramati- cally illustrates these stages, based on the present observations obtained using light microscope. Ex- tracts of various sized follicles and eggs form numerous protein bands with SDS-PAGE (Fig. 4), which reveals some characteristic features. The capacity of follicles to produce E, and 17a, 20f8-diOHprog in respone to gonadotropin stim- ulation is also different in follicles of different sizes (Fig. 5). The developmental pattern of E, secre- tion by follicles is different from that of 17a, 206-diOHprog. Early previtellogenic phase Oocytes (Stages I and II of Fig. 3) in this phase (follicles approx. 20-90 ym) are the smallest grow- ing oocytes in the ovary of spawning females. These oocytes usually form clusters in the peripheral region of the ovary. Stage I (chromatin-nucleolar stage): This group of very small and transparent oocytes (follicles approx. 20-60 um, Figs. 1A, 2 and 3), possesses poorly developed tubular endoplasmic reticulum (ER) and mitochondria. The roughly spherical nucleus with chromatin-threads with a network appearance has many spherical nucleoli, unlike that of the very small (about 10 ~m) oogonia [6, 23] which have a large nucleolus in the nucleo- plasm and aggregations of germinal dense bodies Fic. 1. A: Smallest oocytes at Stage I. 600. Oocytes at the various stages of oogenesis in the Oryzias ovary. B: A section of ovary during the spawning season. Note large yolky oocytes, postovulatory follicles (POF), and previtellogenic oocytes with a yolk nucleus (arrows). Y, yolk mass. C: Follicle layers and chorion at the animal pole side of an oocyte (micropyle, arrow). x 150. x 60. 356 T. Iwamatsu, T. Outa et al. Fic. 2. Living follicles at various stages of oocyte development. Small- and medium-sized follicles (Stage I- VII). 150. Large-sized follicles (Stage VIII). X80. SV, short villi on the chorion; AF, attaching filaments on the chorion. Oogenesis in the Medaka 357 oS Wark tg 5 —800p) oO St Mil Fic. 3. Diagramatic illustration of growing oocytes at various stages. of oogenesis. OW, ovarian wall; PO,, postovulatory follicle just after ovulation; PO;, postovulatory follicle 24 hr after ovulation; A, animal pole; V, vegetal pole; Y, yolk mass; YV, yolk vesicle. and mitochondria in the cytoplasm. Oocytes are covered with an extremely thin single layer of il - 205K flattened follicle cells. mm . GK Stage II (perinucleolar stage): The small and i Se. oy transparent oocytes in the follicles vary from 61 * - 66K ym to 90 um in diameter (Fig. 2) and have a q so relately large nucleus with many spherical nucleoli arranged in the peripheral region (Fig. 3) and = haematoxylin-stainable cytoplasm containing a a = 8 3 Bak yolk nucleus (Balbiani’s body, Fig.1B). An A electron-dense matrix free from cytoplasmic “3 sa @ organelles is situated in the cortical region of the ee ooplasm (Fig. 6A). Mitochondria are arranged in = 3 the marginal region of the electron-opaque area of a the cytoplasm that contains poorly developed SS Se Fic. 4. SDS-polyacrylamide gel electrophoresis of ex- 9 2. 819 os os +. tracts of various sized follicles and eggs. The size asa 6 6 2 3 of the follicles and eggs (dot) is indicated at the 9 8 3 8 8 3 + bottom of each lane. Molecular weight values on the right indicate standard protein bands. 358 T. Iwamatsu, T. Onta et al. E. (ng /m) 17 4.20p-Di'OH prog ("3/m) Fic. 5. Production of E; and 17a, 20/- diOHprog by follicles of various sizes. Follicles were incubated with (shaded columns) or without PMS. Twenty folli- cles were incubated in 1 ml of culture medium. The medium was collected for determination of E, and 17a, 20f- diOHprog by radioimmunoassay. Bars represent means+SE of four incuba- tions. DIAMETER (pm) OF FOLLICLE tubular and smooth ER. The oocytes are covered with extremely flattened granulosa and _ thecal cells, which are separated by a basement mem- brane. Clustered microvillus-like projections in- crease on the restricted area of the oocyte surface facing the junctions of granulosa cells, which usually attach to each other by a number of desmosomes. Granulosa cells possess few micro- projections, a small number of mitochondria, very poorly developed ER and flattened nuclei with a nucleolus. Late previtellogenic phase (chorion-formation) In this phase, follicles containing transparent oocytes (Fig. 2) vary in diameter from 91 «m to 150 um (Stages III and IV of Fig. 3). The protein band pattern formed by SDS-PAGE is charac- terized by 53k M, as a major band and 42k, 63k and 67 k M, as minor bands (Fig. 4). Stage III (chorion-rudiment stage): The oocytes in these follicles (91-120 ym, Figs. 2 and 3) exhibit rudiments of short villi prior to chorion formation. In the cortical cytoplasm, the electron-dense ma- trix (Fig. 6B) has already disappeared from the cortical region, in which mitochondria and tubular ER are located. The monolayer of granulosa cells is interconnected tightly and circumferentially with many well developed desmosomes (Fig. 6B). Thecal cells are separated from the granulosa cells 1-1.5 ym in width by a basement membrane. The cytoplasm of the oocytes has a large yolk nucleus (Fig. 1B) and is stained less intensely with haema- toxylin. Stage IV (attaching filament and oil-droplet Fic. 6. Sections of ovarian follicles containing Stage II and Stage III oocytes. A: Oocyte (in a follicle about 65 ~m in diameter) surrounded by a very thin follicular epithelium consisting of flattened granulosa (G) and thecal (T) cell layers which are separated by a basement membrane (B). Note that there are only a few microprojections from the surface of the oocyte (O). The cortical cytoplasm of the oocyte is electron dense (DL). x 6,600. B: Oocyte (in a follicle about 100 4m in diameter) is surrounded by thick granulosa (G) and thecal (T) cell layers. desmosomes (arrows) meet. Clusters of microprojections from the oocyte are found where adjacent granulosa cells joined by Masses of long mitochondria are frequently observed. x 10,000. Oogenesis in the Medaka 359 a ke 360 T. Iwamatsu, T. Ounta et al. formation stage): The oocytes in these follicles (121-150 um, Figs. 2 and 3) possess minute bumps as rudiments of short villi (Fig.7A) on thin rudiments of chorion (about 0.1 ~m in thickness) among ooplasmic projections (Fig.7B). The oocyte surface possesses long microvillus-like pro- jections among which very thin and electron-dense rudiments of the chorion have begun to appear. In the cortical region of the oocyte, scattered tubular ER, mitochondria (approx. 0.3-0.5 wm in dia- meter) and Golgi complexes have increased markedly in number (Fig. 7B). The oocytes be- longing to this stage exhibit the differentiation of long chorion-villi (attaching filaments) at the vegetal pole region of the chorion (0.2—0.3 um in thickness) and still possess the yolk nucleus (Fig. 1) in the cytoplasm. The most developed yolk nucleus consists of thread-like bodies [24], mitochondria, ER, Golgi lamellae, vesicles and multivesicular bodies (internal small vesicles, a- bout 50 nm) (Fig. 8). Granulosa cells with desmo- somes at intercellular junctions have become thicker than in the previous stage, and their cytoplasmic organelles such as mitochondria, ER and Golgi complexes have increased. Early vitellogenic phase (pre-yolk formation stage) The size of the oocytes markedly increases in this phase. The oocytes in those follicles (151-400 ym, Stages V and VI of Figs. 2 and 3) possess yolk vesicles forming from one to several layers and oocupying the greater part of the cytoplasm. Minute fatty or oil droplets appear mainly in the deeper region of the cytoplasm at the end of this phase. Stage V (early yolk vesicle stage): The oocytes in these follicles (151-250 ym) are characterized by the appearance of 1-2 layers of small vesicles which contain granular materials with a lower electron density than the surrounding cytoplasm. The yolk nucleus is obscure or not observed in the cytoplasm. Oil droplets appear adjacent to the indented nuclear envelope. The nucleoplasm containing shortened lamp-brush chromosomes and ring-shaped nucleoli is deeply stained with haematoxylin. A new innermost layer has been found on the inside of the outermost layer of the chorion by deposition of electron-dense and amor- phous material (0.2-0.6 um in thickness) that is detectable as small vesicles in the vicinity of the oocyte surface (Fig. 9A). Numerous vesicular ER, lamella bodies [25] and mitochondria with minute electron-dense granules are distributed throughout the cortical cytoplasm of the oocytes (Fig. 9A). A special cell, presumably the micropylar cell, is observed among the granulosa cells with their desmosome junctions. Stage VI (late yolk vesicle stage): The oocytes in these follicles (251-400 4m) before yolk formation have 2-6 layers of large yolk vesicles which occupy the greater part of the cytoplasm. Lipid granules are found in the perinuclear region and small yolk granules appear in the cytoplasm at the end of this stage. The protein band pattern is characterized by several bands (less than 20k, 29k, 37k, 42k, 53-80 k M,; Fig. 4). The nucleoplasm has a high affinity for haematoxylin and contains shortened chromosomes and ring-shaped nucleoli. As the oocyte enlarges, the innermost layer of the chorion thickens from 1.5 um to 3 um (Fig. 9B). Electron- lucent vesicles are seen in the cortical cytoplasm (Fig. 9B), probably blebbing inward as pinocytotic vesicles (see [10, 26]). Granulosa cells have developed rough ER, dilated smooth lamellae and lysosome-like bodies, and are in contact with ooplasmic projections via gap junctions. The ability to produce E,j is first recognized in follicles of this stage, but these follicles do not increase their secretion of E, and 17a, 208-diOHprog in response to gonadotropin. Late vitellogenic phase In the oocytes within follicles (401-800 um, Stages VII and VIII of Fig. 3) of this phase, yolk Fic. 7. D, desmosome; G, granulosa cell; T, thecal cell. Follicles containing Stage IV and V oocytes. A: Follicle (about 125 4m in diameter, Stage IV) containing an oocyte that possesses bumps of short villi on the developing rudiment of the chorion. microprojections derived from the oocyte at the oocyte-follicle cell interface. 16,600. B: Follicle (about 155 zm in diameter containing Note numerous B, basement membrane; Stage V oocyte) showing formation of a very thin outermost chorion layer (COL) as an electron-dense layer. Mitochondria and Golgi complex (GO) are located at the peripheral region of oocyte (O). x 26,000. Oogenesis in the Medaka 361 362 T. Iwamatsu, T. OntTa et al. a : ; #9 A. a Nas is, * Fic. 8. Part of the yolk nucleus of an oocyte (in a follicle about 100 »m in diameter) showing numerous thread-like structures (TLB), tubular endoplasmic reticulum, Golgi lamellae, small Golgi vesicles and primitive mul- tivesicular bodies (MV). 31,000. platelets first appear and fuse with each other to form yolk masses in the cytoplasm among yolk vesicles. Cytoplasmic inclusions are pushed to- ward the peripheral region of the oocyte as the mass of yolk enlarges in the central part of the oocyte. The protein band pattern is characterized by the new appearance of 26k, 27k, 30k, 32k, 53 k, 78 k, 98-116 k (yolk proteins), 175 k and 205 Oogenesis in the Medaka 363 k M, bands and the disappearance of bands less than 20 k M,. Stage VII (early yolk formation stage): The oocytes in these follicles (401-500 um, Fig. 2) enlarge rapidly as fusion of small yolk platelets to form yolk masses among yolk vesicles advances. Protein bands of 98 k-116 k and 175 k M, first appear in this stage. The cortical cytoplasm contains a number of mitochondria, ER, yolk granules (YG), cortical alveoli (CA) and oil droplets (L) (Fig. 10A). A nearly formed micropyle (Fig. 1C) is easily recognized at the animal pole of the chorion. Cytoplasmic projections of oocytes are 0.2-0.3 ~m in width, and their cytoplasmic matrix is more electron-dense than that of the follicle cell projec- tions. At this stage, the egg nucleus (germinal vesicle) begins to migrate toward the animal pole where the chorion (about 5 wm in thickness) is thicker than at the vegetal pole region. Gap junctions between granulosa cells (Fig. 10B) are observed. Follicles of this stage are most active in secreting E, and secrete slightly detect- able (P<0.05) 17a, 208-diOHprog. Gonadotropin stimulates an apparent increase in E, secretion by follicles 450 ~m in diameter, while it does not increase their 17a, 208-diOHprog secretion (Fig. 5). Stage VIII (late yolk formation stage): The oocytes which have formed a large yolk sphere in these follicles (501-800 um, Figs. 1B and 2) show morphological differences between the animal and the vegetal poles. The chorion is thicker in the animal hemisphere than in the vegetal hemisphere, the micropyle is located at the animal pole, attaching filaments exist at vegetal pole (Fig. 1B), and the nucleus is located at the animal pole of the oocyte. The cortical ooplasm with the most elongated microprojections is filled with Golgi complexes, aligned ER, CA, yolk granules (YG) and ribosomes (Fig. 10D). Isolated these oocytes still fail to resume meiosis in response to matura- tion-inducing steroids (MIS; progesterone, or 17a, 206-diOHprog) during incubation. Granulosa cells, which display pronounced mitochondria with well-developed tubular cristae and a dense matrix, actively produce E; and 17a, 208-diOHprog. Pro- duction of these steroids is stimulated by gonado- tropin (P<0.05) (Fig. 5). Postvitellogenic phase The fully grown oocytes within preovulatory follicles (801-1200 zm) have the potential to initi- ate their maturation events in response to MIS. Follicles are capable of producing E, and 17a, 206-diOHprog (Fig. 5). Stage LX (maturation stage): The oocytes within the largest follicles (801-1200 um) 0-24 hr before ovulation are capable of undergoing maturation when incubated with MIS. The oocytes approaching the resumption of meiosis clearly exhibit a large germinal vesicle in the vicinity of the micropyle. Before initiation of maturation, the cortical cytoplasm exhibits many mitochondria, annulate lamellae, rough ER and Golgi complexes (Fig. 11). In maturing oocytes the nucleus changes with GVBD leading to formation of the metaphase II spindle of meiosis. The first polar body is extruded at the animal pole. In the granulosa cells of this stage, the nucleus is located in the vicinity to the basement membrane (Fig. 12A). Marked features of the granulosa cells before initiation of oocyte maturation are mitochondria with highly developed tubular cristae and dense-matrix, many vacuoles among the Golgi complexes, transport vesicles and dilated ER (Fig. 12B). All cytoplas- mic projections of both the oocyte and granulosa cells have withdrawn completely from the chorion by the time of ovulation. At the end of this stage, ovulation takes place: oocytes are squeezed from the vegetal pole side out of the follicular layers. Stage X (postovulatory stage): Ovulated oocytes (eggs, approx. 1200 um) exposing short villi and long attaching filaments on the chorion are in the ovarian lumen. Under the light micro- scope, the whole egg is semitransparent, and numerous cortical alveoli (CA, 0.4—40 am) and oil droplets (1-50 4m) are visible throughout the entire egg surface, except for a restricted area at the animal pole. A few Golgi complexes and no annulate lamellae are observed in the cortical cytoplasm, as reported in previous notes [9, 27]. The protein band pattern is characterized by faint protein bands of 175k and 205kM,, and reap- pearance of 20 k, 21k and 45 k bands (Fig. 4). 364 T. Iwamatsu, T. Onta et al. DISCUSSION The present study has summarized cytological and histological characteristics of developing oocytes of Oryzias latipes which were assigned to five phases and ten stages (Table 2). Phases were assigned from the viewpoint of vitellogenesis. The stages of oocyte development, which were clas- sified on the basis of observations of changes in the nucleus and the cytoplasm, and of the formation of the egg membrane (chorion), were compared with those reported in early investigations (Table 3). Oocytes less than 20 4m in diameter, which were designated as the chromatin-nucleus stage by Yamamoto and Yoshioka [17], seem to correspond to oogonia [23] with a large nucleolus and chroma- tin-threads in the nucleus. The perinucleolar stage (Pn) oocytes in the classification of Yamamoto and Yoshioka [17] are divided into Stage I (small oocytes less than about 20 um with the nucleus displaying several nucleoli and chromatin-threads) and Stage II (slightly larger oocytes, which have a large nucleus with many perinucleoli, as in Fundu- lus heteroclitus [28]). Yamamoto [15] and the present study, which employed electron micro- scopy, also divided the stage A oocytes of an early classification [16] into two stages based on fine structural differences in the ooplasm. Our clas- sification of the oocytes, based on changes in oocyte volume as well as cytoplasmic and nuclear morphology of oocytes surrounded by extremely flattened granulosa cells with well developed desmosomes, is as a whole consistent with the observations of Yamamoto [16]. Thus, the early previtellogenic phase before formation of the chorion is conveniently divided into Stage I (chro- matin-thread stage) and Stage II (perinucleolus stage). Concomitant with development of follicles, the growing oocytes in which the cytoplasm faintly stains with haematoxylin can be distinguished from those of the previous stage. The late previtel- logenic phase was established for two developmen- tal stages of oocytes (Stages HI and IV) with both the developing chorion and the most developed yolk nucleus present before formation of yolk vesicles. The present criterion for classification of these stages was the bipolar differentiation of the follicle, which was not studied in detail by any previous investigators. With advancing dif- ferentiation of the chorion, the egg polarity is recognized by appearance of attaching filaments on the chorion at the vegetal pole side in the vicinity of the yolk nucleus. Fine structural studies have revealed the yolk nucleus as an extensive aggregate of cell organelles such as thread-like bodies, mitochondria, Golgi complexes, smooth ER, vesicular bodies and plate structures. The thread-like bodies especially characterize the yolk nucleus [14]. Quite similar bodies in Xenopus oocytes (100 «m) have been described by Takamo- to [24]. During the late phase of previtellogenesis, another typical feature is the most highly- developed lamp-brush chromosomes that possibly correlate with the most highly-developed yolk nucleus. In addition at this stage the maximal development of pinocytosis is seen at the oocyte surface, and the yolk precursors are actively incorporated. Yamamoto [16] and Yamakawa [13] assigned the yolk formation stage to oocytes that enlarged as yolk vesicles and mass increased. The central zone of the cytoplasm has small yolk vesicles in the oocytes designated as being at the stage characterized by the dispersing yolk nucleus. The oocytes of Stage V are characterized by disappearance of the yolk nucleus and appearance of both a layer of yolk vesicles in the central part of the cytoplasm and two layers of the chorion. In addition to these features, other characteristics are appearance of oil droplets adjacent to the nucleus Fic. 9. Follicles containing Stage V and VI oocytes. A: Follicle (about 200 ~m in diameter) containing oocyte (Stage V) with thick outermost layer of the chorion. Microprojections from granulosa cells are observed at a low frequency. ooplasm. 32,000. Arrows indicate electron-dense material. B: A part of the chorion of an oocyte at Stage VI. It consists of outermost (COL) and 16,600. Inset: Lamellar body in the innermost (CIL) layers, into which microprojections (GP, from the granulosa cell; OP, from the oocyte) of both the oocyte (O) and the granulosa cell (G) insert. The granulosa cell surface is in contact with an ooplasmic projection (arrow heads). Arrows indicate surface. 20,000. small electron-lucent vesicles intact with the oocyte Oogenesis in the Medaka T. Iwamatsu, T. OuTA et al. 366 Fic. 10. Oogenesis in the Medaka 367 Fic. 11. Annulate lamellae in an oocyte at the maturation stage (Stage IX). Note the lamellae (AL) that are continuous at the end with lamellae of rough endoplasmic reticulum (ER). 55,200. Fic. 10. Portions of follicles containing Stage VII and VIII oocytes. A: Cortical cytoplasm of Stage VII oocyte. Note ooplasmic inclusions such as cortical alveoli (CA), numerous mitochondria, yolk granules (YG) and oil droplet (L). 7,000. B: Gap junctions among granulosa cells surrounding Stage VII oocyte. 44,800. C: Gap junctions (arrow heads) between Stage VII oocyte microprojections (electron dense) and granulosa cell. 25,700. D: Cortical cytoplasm of Stage VIII oocyte. Note piled endoplasmic reticulum (AER) and many Golgi complexes (GO). 23,100. T. Iwamatsu, T. Onta et al. 368 12: Fic Oogenesis in the Medaka 369 and ring-shaped nucleoli in the nucleoplasm. Our observations of Stage V oocytes are almost con- sistent with those reported by Yamamoto [14] in which irregularly shaped nucleoli become volumi- nous. At Stage V the two or more innermost layers are piled inside the outermost layer of the chorion. This mode of chorion formation has been proposed by Tesoriero [29]. The innermost layers are produced by the oocyte. Another characteris- tic of this stage is that yolk vesicles (or goutte claires, [30]), which contain PAS positive material [4, 13, 16, 31], are observed in the ooplasm. These vesicles correspond to the cortical alveoli (vesicles) in the mature egg. A change in the ER from a tubular to a vesicular form with ribosomes seems to reveal the beginning of active protein synthesis. In hypophysectomized females [32], which are responsive to E,, the ovary contains only young oocytes earlier than those of this stage. Oocyte development beyond Stage V depends on stimula- tion by gonadotropin. At the yolk vesicle stage of Yamamoto and Yoshioka [17], the oocytes that possess several layers of yolk vesicles in the cytoplasm are designated as Stage VI. In this stage, oocyte volume rapidly increases, but yolk globules (small yolk masses) are hardly observed. The oocytes in which small yolk globules appear between large yolk vesicles are designated as Stage VII, belong- ing to the late vitellogenic phase. The appearance of yolk globules is quite consistent with that of yolk proteins of 98-116kM,, as shown in Fundulus oocytes [33]. In this stage, small yolk masses are found in the peripheral region of the cytoplasm among yolk vesicles and numerous oil droplets. Beyond this stage, oocytes in which the yolk mass has rapidly enlarged in the central region are designed as Stage VIII, as described by Yamamoto [16], Yamakawa [13] and Yamamoto and Yoshioka [17]. In these growing oocytes, there are markedly increased microprojections of the oocyte surface toward the granulosa cells through the pore canals of the chorion, as well as projections of granulosa cells toward the oocyte surface. Gona- dotropin stimulates an apparent increase in E, secretion by follicles of both Stages VII and VIII, but follicles of Stage VII do not secrete increased 17a, 208-diOHprog in response to gonadotropic stimulation while those of Stage VIII do. Thus, follicles containing oocytes of Stage VII are different from those of Stage VII in their steroid synthesis capacities. As the yolk mass enlarges, the yolk vesicles and the nucleus (germinal vesicle) shift toward the peripheral region of the oocyte and the nucleus are finally located at the cortical cytoplasm (the animal pole) near the micropyle. Most investigators have reported the late stage of yolk formation as the maturation stage. However, oocytes within folli- cles about 800 zm in diameter are the first that are able to respond to the MIS reinitiating meiosis [20]. Therefore, we assigned Stage VIII to large oocytes that have not yet obtained the capacity for resuming meiosis. During the maturation stage, the nuclei of the granulosa cells move from the side facing the chorion to the opposite side of the cell. Additionally, cytoplasmic protrusions of both the follicle cells and the oocyte into the chorion have shrunk and withdrawn. The oocytes rapidly en- large up to about 1200 um in diameter, primarily due to hydration [5, 34], and ovulate from the vegetal pole region [8]. ACKNOWLEDGMENTS We are very thankful to Dr. Y. Nagahama, National Institute for Basic Biology, Okazaki, for supplying the antisera against steroids, and to Mr. T. Yokochi for his excellent technical assistance. We also wish to express our cordial thanks to Dr. C. A. Brown, Department of Pathology, School of Medicine, University of New Mexico, for her help in preparing the manuscript. Fic. 12. Preovulatory follicles (about 900 ~m in diameter). A: Special granulosa cells as progesterone-producing cells contain characteristic mitochondria with body. tubular cristae and dense matrix. Ly; 6,600. B: A portion of typical granulosa cell exhibits dilated Golgi lamellae (GO), endoplasmic lysosome-like reticulum and round mitochondria with many tubular cristae. Note gap junctions (arrow heads) between oocyte projections and granulosa cell. TV, transport vesicle. x 21,000. 370 T. Iwamatsu, T. OnTA et al. TABLE 2. Characteristics of the Phase Stage Folele General appearance Nucleus 8 size (yam) Pp 20 Centrally located and roughly I to spherical. Nucleoli and chromatin- 60 threads with network appearance = a m 61 Transparent cytoplasm and nucle- Centrally located with peripheral II to us with round nucleoli clearly nucleoli and thin lamp-brush 2 90 visible chromosomes 5 ob a ie} 2 Centrally located. Folded en- os 91 ; ; o Wl ie velope. Voluminous _ nucleoli om irregular in shape. Developed 120 lamp-brush chromosomes vo 3 4 1 Transparent cytoplasm and nucle- Centrally located. Folded en- Iv me us with irregular shaped nucleoli velope. Voluminous nucleoli ring- 150 clearly visible. Rudiments (bright shaped. Developed lamp-brush spots) of short villi on chorion chromosomes Transparent cytoplasm and _ nu- 151 . : . cleus with ring-shaped nucleoli. y ‘© Elongated short villi and attachi 250 ongated short villi and attaching Folded envelope. Nucleoplasm a filaments on chorion strongly stainable with haema- 5 toxylin. Lamp-brush chromo- w somes thicker and shorter. Ring- 251 Transparent cytoplasm. Attaching shaped nucleoli VI to filaments opaque in vegetal hemi- 400 sphere 1S) = v Cy = Attaching filaments opaque in Very irregular in shape. Displaced = 401 vegetal hemisphere. Dark lipid toward animal pole. Shortened > : : : : VII to granules in perinuclear region of chromosomes. Reduced in _vol- 500 cytoplasm filled with vacuoles ume. Nucleoplasm strongly staina- o (yolk vesicles) and oil droplets ble with haematoxylin. e Cortical alveoli (yolk vesicles) and Located at animal pole. Con- 501 : : ene ; VIII ' oil droplets dimly visible in trans- densed chromosomes massed in ns lucent cytoplasm surrounding yolk center of nucleoplasm sphere = Cortical alveoli and oil droplets Apparent large vacuole within 2 801 clearly seen in light cortical cyto- outer nuclear envelope at ovarian 2 S IX to plasm. A large germinal vesicle surface side. Nucleoli and nuclear | 8 1200 visible near micropyle until re- envelope disappear with resump- = 2 sumption of meiosis tion of meiosis bi as) Projecting short villi and attaching Sg = Pa filaments on chorion. Transparent Second metaphase figure at animal us x 1200 : : : 5 and light cytoplasm clearly show- pole in cortical cytoplasm je) ing cortical alveoli and oil droplets *Egg. Oogenesis in the Medaka various stages of oocyte development Cytoplasm Chorion Follicle cells A few microvilli and small number of poorly developed ER and Absent Extremely flat mitochondria Clusters of long microvilli (cyto- plasmic projections). Yolk nu- cleus, a small number of mito- Absent chondria and ER. Electron-dense cortex Cytoplasmic projections evenly distributed on whole surface. Yolk nucleus, increased mitochondria and ER. Faintly stainable with haematoxylin Outermost layer rudiments a- mong cytoplasmic projections of oocyte. Short villi bumps appear A single layer each of thecal and granulosa cells. Desmosomes de- veloped between adjacent granu- losa cells Most developed yolk nucleus. In- creased mitochondria in cortex. Faintly stainable with haema- toxylin Outermost layer (0.2-0.3 um) rudiments among long cytoplas- mic projections. Elongated short villi. Differentiation of attaching filaments Granulosa cells thick at attaching filament side Elongated cytoplasmic — projec- tions. Broken yolk nucleus and a single layer of small yolk vesicles Outermost layer 0.2-0.6 ~m in thickness. Two layers at end of this stage Micropylar cell detectable Very elongated cytoplasmic pro- jections. Two to several layers of large yolk vesicles occupy most of cytoplasm. Lipid granules at peri- nulear region and dense yolk granules Outermost layer 0.5-0.6 ~m and innermost layer 1.5-3 ~m. Mi- cropyle incomplete Thick granulosa cells in vegetal hemisphere. E, production Cytoplasmic projections irregular- ly contact with granulosa cells by gap junctions. Yolk vesicles fully occupy cytoplasm. Small yolk masses among yolk vesicles. In- creased oil droplets Thick in animal hemisphere. Outermost layer 0.6 ~m and in- nermost layer about 4 ~m E, production responsive to gonadotropic stimulation Most elongated cytoplasmic pro- jections. Yolk vesicles and oil drop- lets aligned in a single layer in cortex. Central yolk mass occupies most of cytoplasm. Very long attaching filaments at vegetal pole region, coiled around vegetal hemisphere. Ten ym in total thickness 17a, 208-DiOHprog and E, pro- duction responsive to gonado- tropic stimulation Cortical alveoli (yolk vesicles), oil droplets, mitochondria, vesicular ER, Golgi complexes, annulate lameilae and multivesicular bodies in thin cortex Cytoplasmic projections with- drawn at end of this stage Nucleus located at side of base- ment membrane. Sifted off at end of this stage (micropyle open). E, and 17a, 208-diOHprog produc- tion Transparent and light cytoplasm showing clearly cortical alveoli and oil droplets Projecting short villi and attach- ing filaments on chorion. Inner- most layers showing a stratiform state (12-14 layers) Absent 371 372 TABLE 3. T. Iwamatsu, T. Onta et al. Comparison of the present classification of the stages of oocyte development with those of early investigations of Oryzias latipes oogenesis (A) The present classification (B) Yamamoto, T. S. (1955) (C) Yamakawa, Y. (1959) (D) Yamamoto, M. (1964) (E) Yamamoto, K. and Yoshioka, H. (1964) (F) Iwamatsu, T. (1973) REFERENCES Guraya, S. S. (1986) The Cell and Molecular Biolo- gy of Fish Oogenesis. Karger, Basel & New York. Nagahama, Y. (1985) The functional morphology of teleost gonads. In “Fish Physiology, Vol. 9A”. Ed. by W. S. Hoar, D. J. Randall and E. M. Donaldson, Academic Press, New York, pp. 223-275. Wallace, R. A. (1985) Vitellogenesis and oocyte growth in nonmammalian vertebrates. In “Develop- mental Biology, Vol. 1”. Ed. by L. W. Browder, Plenum Press, New York, pp. 127-177. Aketa, K. (1954) The chemical nature and the origin of the cortical alveoli in the egg of the medaka, Oryzias latipes. Embryologia, 2: 63-66. Hirose, K. (1972) The ultrastructure of the ovarian follicle of medaka, Oryzias latipes. Z. Zellforsch., 123: 316-329. Hogan, J.C. (1978) An ultrastructural analysis of “cytoplasmic markers” in germ cells of Oryzias latipes. J. Ultrastruct. Res., 62: 237-250. Iwamatsu, T. (1973) On changes of ovary, liver and pituitary gland of the sexually inactive medaka (Oryzias latipes) under the reproductive condition. Bull. Aichi Univ. Educ. (Nat. Sci.), 22: 73-88. Iwamatsu, T. (1974) The medaka as a teaching 12 13 14 material. II. Oocyte maturation and fertilization. Bull. Aichi Univ. Educ. (Nat. Sci.), 23: 113-144. (In Japanese) Iwamatsu, T., Ohta, T., Nakayama, N. and Shoji, H. (1976) Studies of oocyte maturation of the medaka, Oryzias latipes. II]. Cytoplasmic and nu- clear changes of oocyte during in vitro maturation. Annot. Zool. Japon., 49: 28-37. Tesoriero, J. V. (1977) Formation of the chorion (zona pellucida) in the teleost, Oryzias latipes. 1. Morphology of early oogenesis. J. Ultrastruct. Res., 59: 282-291. Tsukahara, J. (1971) Ultrastructural study on the attaching filaments and villi of the oocyte of Oryzias latipes during oogenesis. Dev. Growth Differ., 13: 173-180. Yamakawa, Y. (1959) Cytological studies on the oogenesis of the medaka, Oryzias latipes. Report I. Cytological observation of the oogenesis of Oryzias latipes. Bull. Exp. Biol., 9: 46-56. (In Japanese) Yamakawa, Y. (1959) Cytological studies on the oogenesis of the medaka, Oryzias latipes. Report II. Cytochemical studies on the oogenesis of Oryzias latipes. Bull. Exp. Biol., 9: 57-66. (In Japanese) Yamamoto, M. (1963) Electron microscopic studies on the oogenesis and early development of the teleost, Oryzias latipes. Bull. Marine Biol. Station 15 16 17 18 19 20 21 De 23 24 Oogenesis in the Medaka Asamushi, Tohoku Univ., 9: 211-216. Yamamoto, M. (1964) Electron microscopy of fish development. III. Changes in the ultrastructure of the nucleus and cytoplasm of the oocyte during its development in Oryzias latipes. J. Fac. Sci. Univ. Tokyo, Sec. IV, 10: 335-346. Yamamoto, T.S. (1955) Morphological and cytochemical studies on oogenesis of fresh water fish, medaka Oryzias latipes. Jpn. J. Ichthyol., 4: 170-181. (In Japanese) Yamamoto, K. and Yoshioka, H. (1964) Rhythm of development in the oocyte of the medaka, Oryzias latipes. Bull. Fac. Fish. Hokkaido Univ., 15: 5-19. Nakano, E. and Ishida-Yamamoto, M. (1968) Up- take and incorporation of labeled amino acids in fish oocytes. Acta Embryol. Morphol. Exp., 10: 109- 116. Tsusaka, A. and Nakano, E. (1965) The metabolic pattern during oogenesis in the fish, Oryzias latipes. Acta Embryol. Morphol. Exp., 8: 1-11. Iwamatsu, T. (1978) Studies on oocyte maturation of the medaka, Oryzias latipes. VI. Relationship between the circadian cycle of oocyte maturation and activity of the pituitary gland. J. Exp. Zool., 206: 355-364. Kagawa, H., Takano, K. and Nagahama, Y. (1981) Correlation of plasma estradiol-172 and progester- one levels with ultrastructure and histochemistry of ovarian follicles in the white-spotted char, Salvelinus leucomaenis. Cell Tissue Res., 218: 315-329. Young, G., Crim, L. W., Kagawa, H., Kambegawa, A. and Nagahama, Y. (1983) Plasma 17a, 20/- dihydroxy-4-pregnen-3-one level during sexual maturation of amago salmon (Oncorhynchus rho- durus): Correlation with plasma gonadotropin and in vitro production by ovarian follicles. Gen. Comp. Endocrinol., 51: 96-105. Satoh, N. (1974) An ultrastructural study of sex differentiation in the teleost Oryzias latipes. J. Embryol. Exp. Morphol., 32: 195-215. Takamoto, K. (1979) Electron-microscopic studies on the origin of the germinal cytoplasm. Stud. Hum. 25 26 30 31 34 373 Nat., 13: 29-59. Riehl, R. (1977) Konzentrische “Lamellen” in jungen Oocyten von Noemacheilus barfatulus (L.) (Teleostei, Cobitidae). Biol. Zbl., 86: 523-528. Tesoriero, J. V. (1977) Formation of the chorion (zona pellucida) in the teleost, Oryzias latipes. I. Polysaccharide cytochemistry of early oogenesis. J. Histochem. Cytochem., 25: 1376-1380. Iwamatsu, T. and Ohta, T. (1977) Fine structure of loach oocytes during maturation in vitro. Dev. Growth Differ., 19: 213-226. Tokarz,R.E. (1978) Oogenical proliferation, oogenesis, and folliculogenesis in nonmammalian vertebrates. In “The Vertebrate Ovary”. Ed. by R. E. Johnes, Plenum Press, New York, pp. 145-179. Tesoriero, J. V. (1978) Formation of the chorion (zona pellucida) in the teleost, Oryzias latipes. III. Autoradiography of (+H) proline incorporation. J. Ultrastruct. Res., 64: 315-326. Konopacka, B. (1935) Recherches histochimiques sur le dévelopment des poissons. I. La vitellogenese chez le goujon (Gobio fluviatillis) et la carpe (Cyprinus carpio). Bull. int. Acad. pol. Sci. Lett. Ser. B. 1935: 163-182. Masuda, K., Iuchi, I., Iwamori, M., Nagai, Y. and Yamagami, K. (1986) Presence of a substance cross-reacting with cortical alveolar material in “yolk vesicles” of growing oocytes of Oryzias latipes. J. Exp. Zool., 238: 261-265. Iwamatsu, T. and Akazawa, Y. (1987) Effects of hypophysectomy and sex steroid administration on development of the ovary in Oryzias latipes. Bull. Aichi Univ. Educ. (Nat. Sci.), 25: 63-71. (In Japanese) Wallace, R. A. and Selman, K. (1985) Major pro- tein changes during vitellogenesis and maturation of Fundulus oocytes. Dev. Biol., 110: 492-498. Wallace, R. A. and Selman, K. (1978) Oogenesis in Fundulus heteroclitus. 1. Preliminary observations on oocyte maturation in vivo and in vitro. Dev. Biol., 62: 354-369. i vue gyrate diab sail ine Wf SRLS vin juitio ; Aik ee hae | “Tebaitl ally 4 Abi s i ) ie 1 Hi eet W 4 tr it ea’ 4 ZOOLOGICAL SCIENCE 5: 375-383 (1988) © 1988 Zoological Society of Japan Morphological Features of Embryogenesis in Drosophila melanogaster Infected with a Male-killing Spiroplasma Sum TSUCHIYAMA-OMmuURA, BUNGO SAKAGUCHI!, KATSUMI KOGA and DoNALD F. PouLsON” Laboratory of Sericultural Science, Kyushu University, Fukuoka 812, Japan, and *Department of Biology, Yale University, New Haven, Connecticut 06520, U.S.A. ABSTRACT—Morphological studies were conducted with Drosophila melanogaster embryos mater- nally infected with the ‘sex-ratio’ organism, a species of Spiroplasmas which specifically kill male zygotes. Abnormalities occurred at the stages from as early as cleavage or blastoderm to morpho- genetic movements. The most remarkable feature was defective organization of the ventral nervous tissues, a result which fits well with that of previous mosaic analysis showing that the focal region of infection localizes at the ventral midline of the blastoderm. INTRODUCTION It has been rather long since the finding of an abnormal ‘sex-ratio’ (SR) condition in Drosophila, where females exceed males in number disturbing normal sex ratio of 1:1 (see ref. [1] for review). This trait is cytoplasmically inherited from genera- tion to generation through ovaries and eggs [1-3]. Some of SR conditions have been proposed to be brought about by infectious, parasitic microorgan- isms [1] and finally it was demonstrated that the causal agent is a species of mycoplasma having a spiral form with a dimension of 4 to 8 um in length and 0.1 to 0.2 ~m in diameter [2]. This type of organism, although belongs to Spiroplasmas (cf. [3]), has often been called the sex-ratio organism and will be abbreviated as SRO. The SROs are found in the hemolymph of females of infected Drosophila strains and can be transferred to females of the same or other species by intra- abdominal injection, making non-SR strains into SR-ones [4]. Males are absent in the transferred as well as in the original SR strains, and _ this characteristic is accountable almost wholly through Accepted September 7, 1987 Received July 23, 1987 ' To whom reprints should be requested. mortality during embryogenesis [1]. We previously analyzed gynandromorph surviv- als in D. melanogaster having a maternally infected SRO line and suggested that the focus of action of the SRO locates in the close vicinity to the ventral midline [5]. In the present study, we observe by light and electron microscopes the SRO-infected zygotes of D. melanogaster at their embryonic stages and describe that the most markedly affected organ is in fact the ventral nervous system. MATERIALS AND METHODS Collection of eggs The Oregon-R strain of D. melanogaster in- fected with the SRO of the D. nebulosa origin was used, since this heterologous combination has offered a stable and convenient investigation system [5]. Female adults of the infected Oregon- R strain aged 7 to 10 days after emergence were crossed to young males of Oregon-R strain and fed on yeast-enriched food for 3 days at 25°C. Eggs were deposited at 25°C onto filter paper, which had been dipped in a suspension of yeast and placed on an agar plate. The eggs were collected at intervals and washed with distilled water for 376 S. TSUCHTYAMA-OmuRA, B. SAKAGUCHI et al. several times to remove the debris of yeast and then treated outright or incubated until desired age. As a control, normal Oregon-R eggs were processed equally as described above. Light microscopic observation of intact embryos Eggs collected were allowed to develop for 24 hr at 25°C. Then they were counted for hatchability and, after deprived of the chorions by brief treatment with 3% sodium hypochlorite, were observed for the terminal abnormality under a phase-contrast microscope. Another series of eggs were observed for the embryogenesis continuously. The chorions were removed as above. Then, young embryos under- going cleavage mitosis were selected by the method accordind to Bownes [6], submerged in the Drosophila Ringer solution on a glass depression slide to allow development and subjected to inspection individually under a_ phase-contrast microscope. Light microscopic observation of sectioned embryos Dechorionated embryos at desired age were fixed in a mixture of 4 parts 95% ethanol, 1 part 50% acetic acid and 1 part formalin or 25% glutaraldehyde in 0.1_M phosphate buffer at pH 6.8 shaken with octane for 2-Smin. Vitelline membranes were removed with fine tungsten needles in the aqueous phase of the fixing mixture [7]. The fixed embryos were washed with 70% alcohol, dehydrated with a series of ethanol and finally with n-butyl alcohol. To facilitate observa- tion, the eggs were stained in toto with 1% Eosine dissolved in 95% ethanol in the course of dehydra- tion. The stained and dehydrated embryos were embedded in a solution of butoxyethanol and glycol methacrylate (Polysciences Inc., a JB-4 embedding kit) [8], sectioned with a Porter-Blum Sorvall MT-1 ultramicrotome at 2 um thickness, and stained with Giemsa solution (diluted with 0.12 M phosphate buffer, pH 7.4) for 20 min. The stained materials were mounted in Permount and observed under a compound microscope. Electron microscopic observation Eggs were dechorionated, fixed and removed from vitelline membranes as described above. The fixation was continued after removing vitelline membranes to make the total time of fixation 4 hr. The specimens were washed with sucrose- phosphate buffer of the same osmolarity as that of the fixative and allowed to stand overnight at 0°C, rapidly dehydrated in ethanol series and embed- ded in Epon 812. Embryos were sectioned at 60- 150 nm thickness with a Porter-Blum Sorvall MT-1 ultramicrotome and stained with uranyl acetate and lead mixture. A JEM 7A type electron microscope was used for observation. Ovaries taken from female adults of the infected Oregon- R strain were fixed, embedded and sectioned followed by the observation for ultrastructure as above. RESULTS Hatchability and terminal abnormalities The eggs of D. melanogaster maternally infected with the SROs from D. nebulosa had the hatch- ability of about 50% (Table 1). The resulting adults were found to be all females, and thus the unhatched embryos should mainly be males which were killed during embryogenesis. About one fifth of the unhatched eggs showed no sign of develop- ment as seen in Figure 1 a and were white in color, indicating early lethality or block of fertilization (Table 1). The majority of the unhatched eggs exhibited somewhat developed features (Fig. 1b and c) and were brown in color, indicating late lethality (Table 1). Almost all the embryos of the late lethal group showed a common characteristic feature of morphology, with about one third of yolk mass at the posterior part and two thirds of cellular mass at the anterior part (see Fig. 1 b and c). The late lethal embryos had larval tissues in disordered arrangements at the cellular region (see below). Arrows show the yolk mass. The results from continuous observation of embryos indicated that cessation of development occurred at the stages from cleavage mitosis to the early morphogenetic movements. The embryos which were abnormal at the cleavage stage never developed beyond the blastokinesis stage and later broke down as non-brown lethal embryos. These Embryogenesis of SRO-infected Drosophila 377 TaBLeE 1. Hatchability and terminal phase of eggs of D. melanogaster infected with sex-ratio organisms® The number and percentage of eggs Strain hatched unhatched Total Early lethals : b or unfertilized” Late lethals SRO-infected 431° (48%) 80° (9%) 383" (43%) 894 Control 581° (93%) 26" (4%) 18° ( 3%) 625 a: Females of Oregon-R strain of D. melanogaster infected with the D. nebulosa SRO (or of the uninfected control) were aged 7 to 10 days after eclosion and crossed to males from Oregon-R. Eggs were collected at 1 or 2 hr intervals, incubated for 24 hr at 25°C, then the hatched individuals were counted and types of terminal abnormal embryos were estimated. b: See text and Fig. 1. c: The ratios of individuals developed into adults were about 94% (all females) for the SR-strain and about 45% females and 45% males for the control. d: One might argue that the embryos of the SR-strain contain spontaneously occurring lethals as deduced from these data. Although this possibility could not be ruled out, the overall observation presented in this study may be reasonable because the SR-condition is clear-cut as shown in c and because we found the most common features out of several hundreds of specimens. a C Fic. 1. Typical abnormal embryos infected with sex-ratio organisms. Dechorionated eggs were observed under a phase-contrast microscope as described in Materials and Methods. a) An early lethal embryo which stopped development at cleavage mitosis (aged 5 to 6 hr after oviposition). b) A late lethal embryo showing cellular mass at the anterior part and unused yolk mass at the posterior end (aged 20 to 22 hr after oviposition). c) A late lethal embryo with distinct larval tissues (aged 24 hr after oviposition). Scale, 100 sam. 378 S. TSUCHTYAMA-OMURA, B. SAKAGUCHI et al. were typical early lethal embryos like that shown in Figure 1 a. Some embryos indicated their ab- normality as the retardation of blastoderm forma- tion. These embryos in time become anomalous in appearance and finally indistinguishable from those which showed their first abnormality during early morphogenesis. These were typical late lethal embryos like those shown in Figure 1 b and c. Features of dying embryos Samples taken at various times of em- bryogenesis were sectioned and inspected after staining for histology in relation to the stage or state the respective embryos were in (Table 2). During the first 2 hr of development after oviposi- tion, difference between the SRO-infected embryos and the normai ones was seldom de- tected, and at 2 hr about 70% of the individuals were at the syncytial blastoderm stage. At the 3rd hr, infected samples showed significant delay of development, about 50% of which were still at the syncytial blastoderm stage and about 10% at cleavage mitosis (most of the normal embryos were already at or beyond the cellular blastoderm stage). At 3 to Shr after oviposition, the delay became greater (data not shown), and some of abnormal embryos began to break down exhibiting energid- like cytoplasmic islands which seemed to begin to fall apart. These energid-like structures consisted of disturbed fibers (remnants of spindle fibers and/or chromosomes?) as also seen in the SRO- infected D. willistoni embryos [9]. Some of other infected embryos were again at the cleavage stage or at the syncytial to cellular blastoderm stages. As to the embryos which showed the first sign of abnormality at later stages beyond gastrulation, the deviation from the normal course of develop- ment was expressed as the appearance of unusual, necrotic type of cells. These cells were character- ized by their strong stainability of the cytoplasm (dark blue with Giemsa), clear and large nucleus with a ring-shaped nucleolus-like structure and roundness of the cell surface under a light micro- scope. An electron microscopical appearance of this type of cells is seen in Figure 2. Autoradio- graphical study has suggested that these cells are inactive in RNA synthesis and protein synthesis (Tsuchiyama-Omura, unpublished data). They may therefore be a sort of retrograding cells. These occurred mainly in the neuroblast region 5- 6 hr of the development (Fig. 3). As the SRO- infected embryos develop, such cells gradually increase in number also in the hypodermis, midgut or other tissues as well and coming to make a cluster. More and more infected embryos were catego- rized as abnormal, even when they still did not reach the terminal features. At 9 hr after oviposi- tion the cells constructing the hypodermis, neuro- blast, ventral nervous system and mesodermal structures seemed to be somewhat disorderly intermingled with each other. During the succeed- TaBLE 2. Comparison of development between infected and control embryos* The number and percentage of eggs at the stage or state of Time of Strain Total observation 1: cleavage syncytial cellular early anteralzes mite blastoderm blastoderm _ gastrulation 1405h SR 1 ( 1%) 68 (94%) 3 ( 4%) 0 0 72 Shr a Control 1 ( 3%) 26 (72%) 3 ( 8%) 4 (11%) 2 ( 6%) 36 240.5h SR 5 (10%) 7 (14%) 33 (66%) 5 (10%) 0 50 Shr _ Control 1 ( 4%) 5 (18%) 19 (68%) 3 (11%) 0 28 340.5h SR 4 ( 4%) 7 ( 8%) 46 (52%) 23 (26%) 9 (10%) 89 Shr ~ Control 4 ( 7%) 0 8 (13%) 31 (51%) 18 (30%) 61 a: Females of SRO-infected or control Oregon-R strains were crossed as described in Table 1 and eggs were collected at indicated times, sectioned and inspected light microscopically. There were significant differences between SRO-infected and control embryos by means of %*-test in the 3rd-hr samples. Embryogenesis of SRO-infected Drosophila 379 Fic. 2. Electron micrograph of a part of an embryo infected with sex-ratio organisms. An late lethal embryo at 10 hr old was taken to show a necrotic, retrograding cell (a dense and round cell in the center). Arrows point the SRO-like structures seen in the intercellular areas with low electron density. Scale, 1 ~m. ing periods the hypodermic and mesodermal tis- sues followed the normal course of development, that is, the hypodermis thinned out dorsally at about 10 hr (Fig. 4), and the somatic visceral and pharyneal muscles appeared from this time on. The salivary gland with or without contents (mucoprotein?) could also be found in most of the abnormal embryos. On the other hand, the region of ventral nervous system in SRO-infected embryos was shorter than normal even before the onset of shortening and little nerve fibers were found. Nevertheless, the brain was clearly iden- tified at the ordinary place with normal configura- tion in most individuals. The fore- and hindgut looked quite normal, but the mid-ventral part of the midgut was disorganized having retrograding cells. Some of the latter tend to fall into the yolk mass which was abnormally concentrated in the center of an embryo (Fig. 4). The hypodermis began to show the sign of cuticle secretion after the 12th hour which is the normal time course of uninfected embryos. The gonad-like structures were occasionally observed. During the 11-14thhr after oviposition, the yolk mass began to extrude through the feeble and disorganized ventral nervous system leaving other tissues behind. After this incident, abnormal embryos could be easily recognized by the typical terminal phenotype of SRO-infected embryos. Following the extrusion of yolk, tissues seemed to develop autonomously; i.e. embryos from later samples showed more developed structures of tissues, although their interorganic arrangement was completely disturbed. This disturbance occur- red increasingly as the time went, changing most of the abnormal embryos into the typical late lethal ones by the 20-22nd hr after oviposition. Figure 5 illustrates such features. The unabsorbed yolk Fic. 5. 380 Fic. 3. Features of a 5 to 6 hr old embryo infected with sex-ratio organisms. A specimen was cut, stained and observed under a compound microscope as described in Materials and Methods. Arrows show necrotic cells. AM, amnion; EC, ectoderm; MS, mesoderm; NBL, neuroblast; PR, proctodaeum; ST, stomodaeum; YK, yolk mass. Scale, 100 pm. S. TSUCHTYAMA-OmuRA, B. SAKAGUCHI et al. oe — AMG Sa % Fic. 4. Features of a 10 hr old embryo infected with sex-ratio organisms. A specimen was cut, stained and observed as described in Fig. 3. Arrows indi- cate retrograding cells in the midgut, hypodermis etc. AMG, anterior midgut; AMSE, amnioserosa; HY, hypodermis; MMG, middle midgut; PMG, posterior midgut; VNS, ventral nervous system; YK, yolk mass. Scale, 50 yam. Features of 22 to 25 hr old embryos infected with sex-ratio organisms. Specimens were cut, stained and observed as described in Fig. 3. a, b and c are from different individuals to show variety in arrangement of larval tissues from embryo to embryo. Arrows indicate necrotic cells. BR, brain; CPA, cephalopharyngeal apparatus; HY, hypodermis; MG, midgut; MUS, muscles; SG, salivary gland; TR, trachea; YK, yolk mass. Scale, 50 um. Embryogenesis of SRO-infected Drosophila 381 resided at a posterior part (a, b and c) and sometimes at a lateral part (c) of the embryo. The tissue region gave an inside-out impression, with the hypodermis accompanying muscles almost at the central part. The midgut, trachea, cepha- lopharyngeal apparatus, brain, salivary gland, hypodermis and muscles were recognized, but most of them were located at unordered places (the number of cells which consist of each tissue seemed to be smaller than normal). The ventral nervous system was never found in a complete form; instead, the necrotic cells (arrows) were dispersed at the surface of the embryos. Another remarkable status found in the present Sask = Fic. 6. Electron micrograph of parts of ov study is the intercellular areas with low electron density as seen in Figure 2; these areas contained a number of peculiar structures pointed by arrows. We concluded the latter to be SROs from their form and their mode of existence. The intercellu- lar spacing and the SRO-like structures were not detected in the specimens from normal embryos. Moreover, these curious features were very similar to those observable in the ovaries of SRO-infected D. melanogaster (Fig. 6 and Niki, personal com- munication). These may be the first electron microscopic observation of the SRO-like struc- tures in the eggs and ovaries. structures are seen in the area with low electron density between the follicle cells (arrows). Scale, | ym. 382 S. TSUCHIYAMA-OmuRA, B. SAKAGUCHI et al. DISCUSSION In the SRO-infected strain, abnormalities may occur at different stages of embryogenesis. Never- theless, in the large majority of the lethal embryos, the ventral nervous system was found to be the most severely affected organ. The occurrence of necrotic, seemingly degenerating cells in cluster must be in close association with the SRO infec- tion, since in normally developing embryos of the control this type of cells were only occasionally found and never in cluster. These results in D. melanogaster were in agreement with those of the SRO-infected D. willistoni embryos [9]. Most tissues other than the ventral nervous system rather seemed to follow normal course of development in the SRO-infected strain although their arrangement in an embryo was highly abnor- mal. The ventral nervous system was never seen in a normal assembly by itself and the constituent cells became necrotic earlier than other cells. This apparent preference of the ventral nervous tissue in sensitivity to the SRO is compatible with the results of mosaic analysis of SRO-infected D. melanogaster [5] which indicated that the focus of SRO-lethal action included the ventral nervous system. Not only the SRO-lethal action but also many other embryonic lethalities in Drosophila were shown to have their focal region at the ventral side of an embryo [10]. It should be mentioned here that our previous gynandromorph analysis has suggested the mesoderm also to be included in the target of SRO action [5] but in the present study this seemed not the case. The discrepancy may be owing to the difference in developmental stage of observation: The present study was done with embryos whereas the mosaic analysis with adults. In addition to the major late lethal embryos, there also exist a minor group of SRO-lethal embryos, i.e. several percent of the total abnormal embryos stopped their development at the stages as early as cleavage or blastoderm, thus before the establishment of the ventral nervous system. We interpret the early mechanism of androcidal SRO action in terms of some perturbation due to the presence of the SRO. Such effects should be incurable by gene activation of the part of the hosts, since female embryos with double dosage of X-chromosome-linked genes can escape from this lethal action [4]. The plausibility of the inference may be resolved by microinjection of the SRO into normal embryos. The late effects of SRO-embryos might be explained by assuming that the ventral nervous system of the Drosophila embryo requires the expression of a large number of genes for its development and organization so that it is highly sensitive to disturbance owing to the hypothetical early SRO action. In support of this assumption is a preliminary observation by Nickla et al. [11], who indicated that a lethal mutation brings about the abnormalities at the ventral nervous system as well as the brain, although the genes concerned are known to be active in non-nervous tissues. ACKNOWLEDGMENTS This study was supported in part by a Grant-in-Aid for Scientific Research (No. 61480053) from the Ministry of Education, Science and Culture of Japan. REFERENCES 1 Poulson, D. F. (1963) Cytoplasmic inheritance and hereditary infections in Drosophila. In “Methodolo- gy in Basic Genetics”. Ed. by W. J. Burdette, Holden-Day, San Francisco, pp. 404-424. 2 Poulson, D. F. and Sakaguchi, B. (1961) Nature of “sex-ratio” agent in Drosophila. Science, 133: 1489- 1490. 3 Williamson, D. L. and Poulson, D. F. (1979) Sex ratio organisms (spiroplasmas) of Drosophila. In “The Mycoplasmas IIT”. Ed. by R. F. Whitcomb and J.G.Tully, Academic Press, New York, pp. 175-208. 4 Sakaguchi, B. and Poulson, D.F. (1963) Inter- specific transfer of the “sex-ratio” condition from Drosophila willistoni to D. melanogaster. Genetics, 48: 841-861. 5 Tsuchiyama, S., Sakaguchi, B. and Oishi, K. (1978) Analysis of gynandromorph survivals in Drosophila melanogaster infected with the male-killing SR organisms. Genetics, 89: 711-721. 6 Bownes, M. (1975) A photographic study of de- velopment in the living embryos of Drosophila melanogaster. J. Embryol. Exp. Morphol., 33: 789- 801. 7 Zalokar, M. (1971) Fixation of Drosophila eggs without pricking. Dros. Inf. Serv., 47: 128-129. Embryogenesis of SRO-infected Drosophila 383 8 Rice, T. B. and Garen, A. (1975) Localized defects of blastoderm formation in maternal effect mutants of Drosophila. Dev. Biol., 43: 277-286. 9 Counce,S.J. and Poulson, D. F. (1962) Develop- mental effects of the sex-ratio agent in embryos of Drosophila willistoni. J. Exp. Zool., 151: 17-31. 10 Hotta, Y. and Ishikawa, E. (1975) Mosaic analysis 11 of lethal genes of Drosophila. Protein, Nucleic Acid and Enzyme (Kyoritsu Shuppan, Tokyo), 20: 1234- 1256. (In Japanese). Nickla, H., Lilly, T. and McCarthy, A. (1980) Gene activity in the carnation-light synthetic lethal in Drosophila melanogaster. Experientia, 36: 402-403. a petal eo ga : Rien a oy hh st. ans ie (ea “i zB a Pt tf nee nf pom. pe Jett vt hy << ( vine YR TAY» é Par a . { = Sian age tay, CRW @ ee : Pen Ki ' 2 7 a - i . ‘hula mS a ae 7 at Nag 3 7 \ * ra PAL oh a! = ra | 7 3, 1 = _ 4 ZOOLOGICAL SCIENCE 5: 385-395 (1988) Glandular Epithelium Induced from Urinary Bladder Epithelium of the Adult Rat Does Not Show Full Prostatic Cytodifferentiation NaAoyaA SUEMATSU, HIROYUKI TAKEDA and TAKEO MIZUNO Zoological Institute, Faculty of Science, University of Tokyo, Hongo, Tokyo 113, Japan ABSTRACT— Urinary bladder epithelium of the adult rat formed prostate-like glands, when com- bined with fetal urogenital sinus mesenchyme in culture and grafted beneath the renal capsule of male rat hosts. With the progression of the gland formation, the bladder epithelium lost its alkaline phosphatase activity and antigenicity against anti-functional bladder epithelium-antiserum. Like the normal prostate, the induced epithelium expressed acid phosphatase, nonspecific esterase and anti- genicity against anti-human prostatic acid phosphatase-antiserum, but hardly expressed androgen receptors nor an antigen against anti-4-week prostate epithelium-antiserum. The SDS-PAGE patterns of total proteins in the induced glandular epithelium and in the normal epithelia of the urinary bladder and prostate revealed that the induced glandular epithelium loses some bands identified in the bladder epithelium and comes to express other bands similar to but not identical with those of the normal prostatic epithelium, suggesting that the induced gland does not differentiate as completely as the normal prostate, and this result is linked with extremely reduced androgen receptors in the induced © 1988 Zoological Society of Japan epithelium. INTRODUCTION The prostate glands are one of the mammalian male accessory sex organs and develop from their rudiments in the urogenital sinus, which are recognized as epithelial buds projecting into the surrounding mesenchyme. Lasnitzki and Mizuno [1] have shown that the sinus epithelium requires both androgens and the sinus mesenchyme to form prostatic buds in vitro. The androgen receptors observed by steroid autoradiography [2] concen- trate mainly in the sinus mesenchyme but are absent in the sinus epithelium until the lumen starts to be formed. It is likely, therefore, that the androgens stimulate first the sinus mesenchyme, and that the activated mesenchyme produces substances, which stimulate the epithelium to form buds, though their nature is yet unknown. The urinary bladder epithelium could form prostatic glands when combined with the sinus mesenchyme [3]. Furthermore, the bladder epi- Accepted October 22, 1987 Received August 28, 1987 thelium not only formed morphologically recogniz- able prostatic glands but also differentiated biochemically into prostatic glandular epithelium as assessed by protein profiles in two-dimentional gel electrophoresis [4]. In this paper we examined the nature of the prostate-like glands induced in the adult bladder epithelium and whether the induced glands accom- plish full prostatic differentiation. The degree of cytodifferentiation was judged by the types and amount of proteins counted in slab gel elec- trophoresis and by immunochemistry of the tissue. Further an attempt was made to relate the type of cytodifferentiation to the presence or absence of androgen receptors. MATERIALS AND METHODS Animals and tissues Inbred rats (Fischer 344, Charles River Japan Inc., Kanagawa) were mated during the night and copulation was confirmed by the presence of 386 N. SUEMATSU, H. TAKEDA AND T. MIZUNO spermatozoa in the vaginal smears on the following morning. Noon of the day was recorded as 0.5 days of pregnancy. Urinary bladder and uterus were obtained from adult pregnant rats. Urogenital sinuses, rectums and stomachs were excised from 16.5-day male and female fetuses. Rudiments of the seminal vesicle were dissected out from 19.5-day male fetuses. The excised urogenital sinus was then separated into ventral and dorsal halves that develop ventral and dorsal lobes of the prostate respectively. Collection of epithelium and mesenchyme The adult bladder and uterus were treated with collagenase (Worthington Biochemical Corp., Code CLS) 0.03% in Tyrode’s solution for 2 hr at 37°C and the fetal urogenital sinus, rectum, stom- ach and rudiments of seminal vesicle were treated with 0.06% collagenase for 30 min at 37°C. The collagenase-treated tissues were washed in 50% fetal bovine serum (FBS) in Tyrode’s solution for 2 hr with three changes, and the epithelium was separated from the mesenchyme with two watch- maker’s forceps under a stereomicroscope (x 20). Culture method A small fragment of the epithelium was placed upon a piece of the mesenchyme put on a membrane filter (Millipore Corp., pore size 0.8 ym), which was put on a Stainless grid placed in a glass dish filled with medium 199 with Earle’s salts (GIBCO) containing 20% FBS to the level of the membrane filter, cultured in a CO, incubator (5% CO>, 95% air, at 37°C) overnight and then the recombinate was grafted beneath the kidney cap- sule of syngeneic adult male hosts (8-20 weeks old) (designated as “in vivo cultivation”). After 3- 5 weeks, the grafts were harvested for examina- tion. Enzyme histochemistry Alkaline phosphatase The grafts were fixed in ice-cold 80% ethanol for 2 hr and embed- ded in paraffin (m.p. 46-48°C). The alkaline phosphatase was detected by the tetrazolium reaction [5] at pH 9.2-9.4. Acid phosphatase and nonspecific esterase The grafts were fixed in ice-cold 4% paraformaldehyde in 0.1M phosphate buffer (pH 7.2-7.4) for 2 hr, washed overnight in several changes of 5% sucrose at 4°C and frozen in isopentane (—190°C) chilled with liquid nitrogen. The acid phosphatase was assessed by azo-coupling method [6] at pH 5, and the nonspecific esterase, by the azo-coupling method [7] at pH 5. Immunohistochemistry As the first ligands for the indirect im- munofluorescence methods, the following anti- bodies were used: a rabbit polyclonal anti-human prostatic acid phosphatase-antibody (Miles-Yeda Ltd.), which stained patches of epithelial cells more intensely in the ventral prostate than in the dorsal one; mouse polyclonal anti-rat ventral or dorsal prostatic epithelium-antibodies of which antigen was homogenate of glandular cells of 4-week rat ventral or dorsal prostate respectively and the antibody was prepared in our laboratory according to the method of McKeehan et al. [8]; and a mouse polyclonal anti-rat urinary bladder epithelium-antibody prepared in our laboratory. Each antibody was applied to the sections of various tissues including urinary bladder, ventral and dorsal prostate, seminal vesicle, liver and kidney and thus confirmed to be specific to the tissues concerned. Tissues were fixed in methanol (—20°C) for 30 min, replaced in ethanol (—20°C) for 30 min, embedded in Polyester wax (melting point 37°C, BDH Chemicals Ltd.), and sectioned at 6 «m. They were treated with the primary antibody for 1 hr at room temperature and exposed to second antibodies: Goat anti-rabbit IgG antibody (Miles) or goat anti-mouse IgG antibody (Cappel), both conjugated with fluorescein isothiocyanate. The sections were examined with an Olympus epifluorescence microscope (BH2-RFK). After observation, the sections were washed in phos- phate-buffered saline and stained with haematoxy- lin-eosin (HX-E). Steroid autoradiography Tissues and grafts were labelled with [1, 2, 4, 5, 6, 7-H] dihydrotestosterone and then subjected to the thaw-mount autoradiographic technique. The details of steroid autoradiographic techniques Prostate Induction in Bladder Epithelium 387 have been described previously [2]. Polyacrylamide gel electrophoresis (PAGE) and immunoblotting Tissues were treated with collagenase in Ca’‘, Mg’*-free Tyrode’s solution for 1.5 hr at 37°C, washed with gentle stirring in Ca’*, Mg**-free Tyrode’s solution for 30 min at room temperature. Then isolated glandular epithelial cells were pul- verized in a stainless steel mill chilled with liquid nitrogen, dissolved in a SDS sample buffer includ- ing 5% (v/v) 2-mercaptoethanol and 2.3% (w/v) sodium dodecyl sulphate (SDS), and subjected to electrophoresis on a polyacrylamide 4/20 gradient gel with SDS (Daiichi Pure Chemicals Co.) accord- ing to Laemmli [9]. As molecular-weight markers, LMW calibration kit (Pharmacia) was used. After electrophoresis, proteins on the gels were either stained with silver or transferred electrophoretical- ly to Durapore filters (Millipore) according to Towbin et al. [10]. The filters were then treated first with the antibody to be examined, and next with goat anti-mouse IgG antibody conjugated with horseradish peroxidase (Miles), washed, and treated with 4-chloro-1-naphthol in the presence of hydrogen peroxide. RESULTS Morphological differentiation There was no contamination with epithelial cells in the isolated urogenital sinus mesenchyme throughout the ex- periments examined in serial sections of the separated components (14 cases in total). All of the 17 homotypic recombinates composed of fetal ventral sinus epithelium and its mesenchyme de- veloped fully differentiated ventral prostate glands with high secretory activity, when grown for 3.5 weeks beneath the kidney capsule of normal male hosts (Fig. 1), but all 5 recombinates failed to develop prostatic glands when grown in castrated male rats. Adult bladder epithelium recombined with fetal ventral sinus mesenchyme underwent morphological changes after in vivo cultivation in normal male hosts: after 1 week of cultivation, epithelial buds projected into the surrounding mesenchyme in 5 recombinates out of 5 (Figs. 2a and 2a’); after 2 weeks, lumina were formed in the & Fic. 1. A homotypic recombinate of fetal ventral sinus epithelium and its mesenchyme grown for 3.5 weeks in an adult male host. 280. extended buds (Fig. 2b); after 3 weeks, prostate- like glands were formed, but they were lined with epithelium generally shorter than that found in the homotypic recombinates and were rich in stroma and poor in secretion (Fig. 2c) in all 37 recombi- nates. After 8 weeks, there was no further dif- ferentiation in all 8 recombinates (Fig, 2d). In the castrated male hosts, the bladder epithelium was maintained in the same state as at the onset of culture and no glands were formed in all 12 recombinates even in the presence of the sinus mesenchyme. Mesenchymes of 16.5-day ventral and dorsal sinuses and of 19.5-day seminal vesicle induced prostate-like glands from the adult bladder epithe- lium, though the glands were morphologically different to some extent according to the kind of the recombined mesenchymes (Fig. 3a, b, and c), but fetal rectal mesenchyme did not. The other examined epithelia combined with fetal ventral sinus mesenchyme did not form any glandular structure even after 4 weeks’ in vivo cultivation: In all 3 recombinates, adult uterine epithelium made only a tube; Fetal rectal epithelium formed a vesicle, with epithelium that developed many goblet cells in all 4 recombinates; Fetal stomach epithelium formed a vesicle, lined with mucous epithelium characteristic of the adult stomach epithelium in all 3 recombinates. Functional differentiation The enzyme activi- ties were seen to change during the prostate-like gland formation. The epithelium of the adult 388 N. SuEMATSU, H. TAKEDA AND T. MIZUNO Fic. 2. weeks, X270; (c) 3.5 weeks, «270; bladder was positive for alkaline phosphatase activity (Fig. 4a) and negative for both acid phos- phatase (Fig. 4b) and nonspecific esterase (Fig. 4c) activities. In contrast, the glandular epithelium induced in the bladder epithelium by the fetal ventral sinus mesenchyme lost its alkaline phos- phatase activity (Fig. 4d) but showed acid phos- phatase (Fig. 4e) and nonspecific esterase (Fig. 4f) activities just like the acinar epithelium of the normal ventral prostate, which was negative for alkaline phosphatase (Fig. 4g) and positive for both acid phosphatase (Fig. 4h) and nonspecific Recombinates of adult bladder epithelium and fetal ventral sinus mesenchyme. (a) 1 week, «215; O 4 Cc or Y : e d (b) 2 (d) 8 weeks, x 135, cultured in adult male hosts. (a’) A cross section of the epithelial bud as seen in (a), 340. esterase (Fig. 41). The stromal cells immediately adjacent to the acinar epithelium were positive for alkaline phosphatase both in the induced and normal prostatic glands. The glandular structure induced by the mesenchymes of fetal dorsal sinus and seminal vesicle showed similar histochemical features as those induced by the mesenchyme of the fetal ventral sinus. There was thus no differ- ence between the heterotypic and homotypic recombinates as far as histochemical activities of these enzymes were concerned. The immunohistochemical study also suggested Prostate Induction in Bladder Epithelium 389 Fic. 3. sinus mesenchyme (b), or 19.5-day seminal vesicle mesenchyme (c) cultured in adult male hosts. that heterotypic functional differentiation occur- red in the induced glands. The bladder epithelium reacted with our anti-bladder epithelium- antiserum (Fig. 5) even after in vivo culture of 4 weeks as far as it maintained its original form seen at the time of recombination with sinus mesen- chyme. The antigenicity against the antiserum was lost in the epithelium, when induced to form glandular epithelium (Fig. 6). This result suggests that the bladder epithelium dedifferentiated to express specific enzyme activities following the morphological gland formation. Most of the induced glandular epithelium was found to be positive against anti-human prostatic acid phosphatase-antiserum (Fig. 7), although some glandular cells as well as the normal counterparts were negative. Oddly enough it did not express antigenicity against the anti-ventral prostatic epithelium-antiserum (Fig. 8a), which intensely stained the normal ventral prostatic epithelium in the control explant (Fig. 8c) but only faintly the ventral prostate 5 days after castration (Fig. 8d). In the immunoblotting, the anti-ventral prostatic epithelium-antiserum recognized many types of proteins specific to the ventral prostatic epithelium (Fig. 9). The induced glandular epithelium did not Heterotypic recombinates of adult bladder epithelium and 16.5-day ventral sinus mesenchyme (a), dorsal «215. respond to anti-dorsal prostatic epithelium- antiserum, which reacted with the normal dorsal prostatic epithelium (Fig. 8e). The steroid autoradiography showed heavy nuclear labelling with [*H] dihydrotestosterone (DHT) in homotypic explants of sinus epithelium with its mesenchyme (Fig. 10a). In contrast, the glandular epithelium induced from bladder epithe- lium showed no preferential nuclear uptake of the steroid, but the surrounding mesenchymal cells showed heavily labelled nuclei in all 5 recombi- nates (Fig. 10b). In the competition experiments in which the tissues were incubated both with the radioactive steroid and with a 400-fold excess of unlabelled steroid, the uptake of radioactive ster- oid by nuclei was completely abolished. We also compared the total proteins of the isolated glandular epithelial tissues with each other, which were separated by SDS-PAGE. The induced glandular epithelium differed qualitatively and quantitatively from the bladder epithelium but not from the ventral prostatic epithelium. The induced glandular epithelium lost the bladder epithelium-specific proteins (B, B’, Fig. 11), and showed prostatic epithelial proteins (P, P’). Moreover the induced glandular epithelium ex- N. Suematsu, H. TAKEDA AND T. MIZUNO 390 Prostate Induction in Bladder Epithelium 391 Fics. 5. and 6. Immunofluorescence with anti-bladder epithelium-antiserum (a) and HX-E (b) preparations of normal adult bladder (Fig. 5) and induced glands in the recombinates of adult bladder epithelium and fetal sinus mesenchyme (Fig. 6). 270. Fic. 7. acid phosphatase-antiserum. The bladder epithe- Immunofluorescence with anti-human prostatic lium-derived acinar epithelium showed intense reaction in the apical cytoplasm. 480. pressed the ventral prostatic epithelial protein (V) but not the dorsal prostatic epithelial proteins (D, D’), even when the bladder epithelium was recom- bined with dorsal sinus mesenchyme. DISCUSSION In the present study, we found that in recombi- nates of fully differentiated adult rat bladder epi- thelium with sinus mesenchyme, cultured in the presence of androgens, the bladder epithelium formed prostate-like glandular epithelium. The stratified transitional bladder epithelium which is normally alkaline phosphatase positive changed to a simple columnar glandular epithelium, became alkaline phosphatase negative and acquired acid Fic. 4. Histochemical assay of alkaline phosphatase (a, «280; d, x 140; g, 270), acid phosphatase (b, x 140; e, x 340; h, x 270) and nonspecific esterase (c, * 140; f, x 250; 1, 380). Dark areas indicate the enzyme positive sites. The features of these histochemical markers expressed in the glandular epithelium induced in the recombinates composed of adult bladder epithelium and fetal ventral sinus mesenchyme (d, e, f) were similar to those of acinar epithelium of normal ventral prostate (g, h, i) but different from those of adult bladder epithelium (a, b, c). 392 N. SUEMATSU, H. TAKEDA AND T. MIZUNO Fic. 8. phosphatase, nonspecific esterase and prostatic acid phosphatase antigen. Heterotypic mor- phogenesis has been reported in the studies of epithelio-mesenchymal interactions, but some- times it was not accompanied by heterotypic cytodifferentiation [11, 12]. However, the present study revealed that cytochemical differentiation took place concomitant with heterotypic mor- phogenesis, and the mesenchyme played the deci- sive role both in the morphogenesis and in the cytodifferentiation of the epithelium. The ventral-, dorsal- and coagulating-lobes of prostates develop from the urogenital sinus and the seminal vesicle, from the basal region of the Immunofluorescence with anti-ventral prostatic epithelium-antiserum of the induced glandular epithelium of bladder origin (a), ventral prostate gland in the control explant composed of sinus mesenchyme and its epithelium (c) and ventral prostate excised 5 days after castration (d). HX-E preparation (b) is the same with (a). Immunofluorescence with anti-dorsal prostatic epithelium-antiserum of normal dorsal prostate (e). x 270. Wolffian ducts. Although they are similar in histological appearance, closer examination re- vealed differences among them. For instance, the acini of the ventral prostate are tightly packed in very small amounts of stroma and are lined with columnar epithelial cells that have basally located nuclei and a prominent supranuclear clear area that corresponds to the location of the Golgi apparatus, while the acini of the dorsal prostate are loosely distributed within large amounts of stromal tissue and are lined mainly with cuboidal epithelial cells with centrally placed nuclei and a supranuclear clear area as described by Jesik et al. [13]. Prostate Induction in Bladder Epithelium 393 a bcd — origin ‘— 95K oe —-77K ~= —60K pce Sa: svete ’ Ss : ‘ ; Ue APS: Piinsghe is - * a ede ° 3 ss ‘ ¢ or? ee eb Fic. 9. Immunoblotting with anti-ventral prostatic epithelium-antiserum after electrophoresis on the gel containing SDS of crude extract of adult blad- der epithelium (a), glandular epithelium in the re- combinates composed of adult bladder epithelium and fetal ventral sinus mesenchyme (b), glandular epithelium in the control explants of fetal ventral sinus epithelium and its mesenchyme (c) and nor- mal ventral prostatic epithelium (d). Ten xg of pro- tein were applied to each lane. Oe of Fic. 10. Autoradiographs of explants incubated with [7H]DHT. Sections were stained with HX. The exposure period was 4 weeks. E, glandular epithelium. (a) Prostate gland formed in a control explant of fetal sinus. Epithelial cells exhibited intense nuclear labelling, while the intensity of labelling in the mesenchyme surrounding the epithelium is weaker. 1350. (b) Prostate-like gland induced in a recombinate composed of adult bladder epithelium and fetal sinus mesenchyme. Epithelial cells showed no strong nuclear concentration of grains, while the mesenchymal cells immediately beneath the epithelium (arrows) showed heavy accumula- tion on nuclei. 1350. Cunha et al. [3] showed that prostatic mor- phogenesis occurred in heterotypic recombinates of adult bladder epithelium and fetal sinus mesen- chyme, but they did not mention which region of the sinus mesenchyme was used. In the present study, we used three precisely defined mesenchy- mal regions isolated from 16.5-day ventral and dorsal sinus and from 19.5-day rudiments of seminal vesicle and found that our results agreed mostly with those of Cunha ef al. except in a few details: In all grafts, glands rich in stroma and poor in secretion were formed, irrespective of the region of the mesenchyme used; When the sinus mesenchyme of the ventral or dorsal region was 394 N. Suematsu, H. TAKEDA AND T. MIZUNO a bcecede | | |. D'(69K) B(63K)>—3 +p’(64K) B’(49K)> Fic. 11. SDS-electrophoretic patterns of proteins in adult bladder epithelium (a), acinar epithelium of recombinates composed of adult bladder epithelium and fetal ventral sinus mesenchyme (b) or fetal dorsal sinus mesenchyme (c), acinar epithelium of control grafts of ventral sinus (d) or dorsal sinus (e). Ten zg of protein were applied to each lane. Both (b) and (c) lost proteins B, B’ and expressed P, P’, V but not D, D’. used as an inductor, the acini formed in the recombinates resembled developing prostate in the histological character: the acini were loosely arranged within the supporting stroma and lined with cuboidal epithelial cells containing a centrally located round nucleus. When the mesenchyme of the seminal vesicle was used instead of the sinus mesenchyme, the acinus epithelium possessing long nuclei seen in the seminal vesicle was also induced (Fig. 3c). Neubauer et al. [4] reported that the induced acinar epithelium in the recombinates composed of adult bladder epithelium and fetal sinus mesen- chyme acquired androgen receptors but that the androgen binding activities were significantly low- er than those of the adult ventral prostate. We assume that the androgen binding activities they detected in the experimental grafts lies principally in the stroma rather than the epithelium. Because our autoradiographic studies have shown that androgens scarcely bind to the induced glandular epithelium in the heterotypic recombinates but mainly to the stromal cells, particularly, those in the immediate vicinity of the epithelium (Fig. 10b). Similar results have been obtained with the epithelia of androgen-receptor defective Tfm mice: the 7Tfm bladder epithelium as well as the sinus epithelium [14] formed prostate-like glands under the influence of normal sinus mesenchyme and expressed acid phosphatase and nonspecific esterase activities. It may, therefore, be assumed that the induced glandular epithelium is devoid of androgen receptors. The function of the androgen receptors in the acinar epithelium is still uncertain. In adult ventral prostate glands excised 5 days after castration the reaction against anti-ventral prostatic epithelium- antiserum was significantly lowered (Fig. 8d), sug- gesting that the antibody recognized mainly the androgen-dependent proteins. Our results showed that the induced glandular epithelium possessed neither normal androgen receptors nor antigenic- ity against the anti-ventral prostatic epithelium- antibody. It seems, therefore, that the epithelial androgen receptors are necessary for the synthesis of androgen-dependent proteins responsible for epithelial function. When adult bladder epithe- lium was combined with dorsal sinus mesenchyme, the induced glandular epithelium did not possess D’ (69 K)-protein specific to dorsal prostatic epithelium (Fig. 11). Though the molecular weight of D’ is similar to serum albumin, serum con- tamination is never expected, because we did not use serum in the procedure of sample preparation, and D’ is specific to the lane (e). This suggests that the androgen receptors also play an important role in the synthesis of lobe specific proteins. It can be concluded that adult bladder epithe- lium is able to respond to the stimulation of androgen-activated sinus mesenchyme and loses the characteristic proteins for the bladder epithe- lium and forms glandular structures with newly induced enzyme activities similar to those of the normal prostate. The induced epithelium, howev- er, is unable to attain full prostatic cytodifferentia- tion nor express lobe specificity due to a lack of normal androgen receptors. ACKNOWLEDGMENTS We thank Dr. Ilse Lasnitzki of the Strangeways Research Laboratory for constructive criticism and help in the preparation of the manuscript. This work was supported by a Grant-in-Aid from the Ministry of Education, Science and Culture of Japan. Prostate Induction in Bladder Epithelium 395 REFERENCES Lasnitzki, I. and Mizuno, T. (1977) Induction of the rat prostate gland by androgen in organ culture. J. Endocrinol., 74: 47-55. Takeda, H., Mizuno, T. and Lasnitzki, I. (1985) Autoradiographic studies of androgen-binding sites in the rat urogenital sinus and postnatal prostate. J. Endocrinol., 104: 87-92. Cunha, G.R., Fujii,H., Neubauer, B. L., Shan- non, J.M., Sawyer,L. and Reese, B. A. (1983) Epithelial-mesenchymal interactions in prostatic de- velopment. 1. Morphological observations of prosta- tic induction by urogenital sinus mesenchyme in epithelium of the adult rodent urinary bladder. J. Cell Biol., 96: 1662-1670. Neubauer, B. L., Chung, L. W. K., McCormick, K. A., Taguchi, O., Thompson, T. C. and Cunha, G. R. (1983) Epithelial-mesenchymal interactions in prostatic development. 2. Biochemical observations of prostatic induction by urogenital sinus mesen- chyme in epithelium of the adult rodent urinary bladder. J. Cell Biol., 96: 1671-1676. McGadey, J. (1970) Detection methods II. Hydro- lases. Tetrazolium reaction. In “Enzyme Histochem- istry (1979)”. Ed. by Z. Lojda, R. Gossraw and T. H. Schiebler, Springer-Verlag, Berlin, pp. 61-63. Lojda, Z. (1979) Detection methods II. Hydrolases. Simultaneous azo-coupling with esters of the naph- thol AS series. In “Enzyme Histochemistry (1979)”. Ed. by Z. Lojda, R. Gossraw and T. H. Schiebler, Springer-Verlag, Berlin, pp. 72-75. Davis, B. J. and Ornstein, L. (1959) Detection 10 11 13 14 methods IT. Hydrolases. Simultaneous azo-coupling with 1-naphthyl acetate. In “Enzyme Histochemistry (1979)”. Ed. by Z. Lojda, R. Gossraw and T. H. Schiebler, Springer-Verlag, Berlin, pp. 109-110. McKeehan, W. L., Rosser, M. P., Glass, H. A. and Fast,D. (1980) Prostatic binding protein: An androgen-dependent marker for prostate epithelial cells. Biochem. Biophys. Res. Commun., 95: 674— 681. Laemmli, U. K. (1970) Cleavage of structural pro- teins during the assembly of the head of Bacte- riophage T4. Nature, 227: 680-685. Towbin, H., Staehelin, T. and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacryl- amide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci. USA, 76: 4350-4354. Sakakura, T., Sakagami, Y. and_ Nishizuka, Y. (1979) Persistence of responsiveness of adult mouse mammary gland to induction by embryonic mesen- chyme. Dev. Biol., 72: 201-210. Yasugi, S. (1984) Differentiation of allantoic en- doderm implanted into the presumptive digestive area in avian embryos. A study with organ-specific antigens. J. Embryol. Exp. Morphol., 80: 137-153. Jesik, C. J., Holland, J. M. and Lee, C. (1982) An anatomic and histologic study of the rat prostate. The Prostate, 3: 81-97. Mizuno, T., Takeda, H. and Suematsu, N. (1986) Récepteurs d’androgénes de l’épithélium au cours de l’induction des glandes prostatiques a partir de Pépithélium du sinus urogénital de la Souris Tfm- mutant. C. R. Soc. Biol., 180: 593-595. i] Hah deo’ cast, 7 7 rit ' a ; a ia . eh = «hte te ; eee, ni foliwagerl ar Fy - had Hite ie es : ss Poy . : _ : CRE mM = 4 Dee oy te NEge 7 “ fish eee - stead ‘ — 7 = : i 4 ; = Nal t > u F 1 a ia io apy ee th wile : -_ a a * Hi Hye, dill | E uty Sa. : 7 = Se ZOOLOGICAL SCIENCE 5: 397-406 (1988) Development of an in situ Hybridization Method for Neurohypophysial Hormone mRNAs Using Synthetic Oligonucleotide Probes Susumu Hyopo, MAMORU FUJIWARA, SHIGERU KOZONO, Moriyuki Sato! and AKIHISA URANO Department of Regulation Biology, Faculty of Science, Saitama University, Urawa, Saitama 338, and 'Tokyo Research Laboratories, Kyowa Hakko Kogyo Co., Machida, Tokyo 194, Japan ABSTRACT— Vasopressin (AVP) and oxytocin (OXT) mRNAs are highly homologous. We de- veloped an in situ hybridization method to discriminate the AVP and the OXT mRNAs using synthetic 22mer deoxyoligonucleotides as probes which have several advantages over the use of cDNAs, e.g., highly specific, easy to obtain a designed probe, and easily accessible to cellular mRNAs. The probes were radiolabeled at the 5’ ends with *’P, applied to rehydrated paraffin sections of rat and/or toad hypothalami, and were visualized by autoradiography. RNase treatment before incubation with the probes and measurement of melting temperature showed that the probes actually paired with tissue RNAs. The specificity of hybridization signals was checked by the following tests: absorption test, competition test, a use of alternate probes complementary to the different regions of the same MRNA, cross species hybridization, and comparisons with the immunohistochemical localization of AVP and OXT in adjacent or the same tissue sections. These tests showed that the oligonucleotide probes specifically discriminate the AVP mRNA from the highly homologous OXT mRNA. Furthermore, cross species hybridization clarified that an oligonucleotide probe can discriminate nucleotide se- quences which include 2 mismatching bases. The use of multiple probes complementary to different loci in the same MRNA showed not only the specificities of the hybridization signals, but also its © 1988 Zoological Society of Japan usefulness to enhance hybridization signals. INTRODUCTION Arginine vasopressin (AVP) and oxytocin (OXT) are mammalian neurohypophysial hor- mones produced mainly in magnocellular neurons in the supraoptic (SON) and the paraventricular nuclei (PVN). They are released from neurosecre- tory terminals into blood capillaries in the neurohypophysis, and play important physiologi- cal roles, e.g., regulation of plasma osmolarity and blood pressure by AVP, and oxytocic action and stimulation of milk ejection by OXT. It is therefore important to examine expressions of AVP and OXT genes in magnocellular neurons in various physiological statuses. A recently de- veloped in situ hybridization (ISH) method is the Accepted October 13, 1987 Received August 12, 1987 most plausible candidate for this examination. The structures of rat AVP and OXT genes recently clarified [1] show that the AVP mRNA and the OXT mRNA share an extremely homolo- gous region, the exon B, the homology of which is about 95 %. Since cDNAs can hybridize with mRNAs the homology of which is approximately 65 % [2], the presence of the AVP mRNA has been detected with a spliced cDNA probe com- plementary to the glycoprotein encoding region which is not present in the OXT mRNA [3-7], while the OXT mRNA has been localized with a probe complementary to the 3’-end of neurophysin (NP) and the 3’-untranslated region [3]. A prob- lem arising here is the occurrence of vasotocin (AVT) especially in fetal brains of mammals [8, 9]. A possibility that the cDNA probes hybridize with the AVT mRNA makes it difficult to apply the ISH method in a study of ontogeny of the 398 S. Hyopo, M. Fustwara et al. neurosecretory system. Moreover, the use of cDNA probes entirely depends on their availabil- ity that requires facilities for recombinant DNA techniques. One of possible ways to overcome these problems is the use of synthetic deoxyoligo- nucleotides as probes for ISH, the technique developed in our laboratory [10-12]. The use of oligonucleotides as probes for ISH further can have several advantages over the use of cDNAs, that is, highly specific [2, 13, 14], easy to obtain a designed probe, easy to prepare and to label in an ordinary laboratory [15] and easily accessible to cellular mRNAs [16-18]. We designed 22mer oligonucleotide probes to discriminate localization of AVP and OXT mRNAs in paraffin sections. We further revised our previous ISH protocol [10, 11] by checking each staining step. A fixative solution was also carefully screened to stain the same or adjacent tissue sections by both ISH and immunohisto- chemical methods, because demonstration of AVP and OXT is crucial for better understanding of their gene expressions. Specificity of the present method was confirmed by various tests including cross species hybridization with the mRNAs of toad neurohypophysial hormones, nucleotide sequences of which were recently determined by Nojiri et al. [19]. Through the specificity tests, we tried to elucidate technical limitations of the present method and to confirm its general applica- bility in gene expression studies of many other peptides and proteinaceous hormones. MATERIALS AND METHODS Preparation of tissue sections Male Wistar-Imamichi rats (6-8 weeks old) and adult Japanese toads of both sexes captured in the autumn were obtained from commercial sources. They were killed by decapitation, and the hypothal- ami and the pituitaries were rapidly taken out and immersed in fixative solutions at 4°C for 2 days. Since a preliminary experiment showed that fixa- tion by perfusion markedly decreased hybridiza- tion signals, we preferred fixation of tissues by immersion. Fixatives tested were: Bouin’s solu- tion, modified Bouin’s solution which does not include acetic acid, 4% paraformaldehyde (PFA) in 0.05 M phosphate buffer (pH 7.3), a buffered solution containing 2% PFA and 1% glutaralde- hyde (GLA), and that including 2% PFA, 1% GLA and 1% picric acid (PA). As is described in the Results section, the mixture of PFA, GLA and PA (PGP solution) yielded satisfactory results in both ISH and immunohistochemical staining among these fixatives. Therefore, the PGP solu- tion was routinely used in the present study. After fixation, the hypothalami were washed in 70% ethanol at 4°C for 24 hr twice. They were then dehydrated through graded ethanols, and were embedded in paraplast. Serial transverse sections were cut at 8 or 10 «4m, separated into several groups, and were mounted on gelatinized slides. Some hypothalamic tissues were washed in cold 0.05 M phosphate buffer (pH 7.3) after fixa- tion, rapidly frozen in butanol cooled in dry ice-acetone, and were cut at 20 wm on a frozen microtome. They were also mounted on gelatin- ized slides. Preparation of synthetic oligonucleotide probes Four 22 mer oligonucleotide probes (Fig. 1) were synthesized by the phosphoramidite method [20] and were purified by polyacrylamide gel electrophoresis. They are complementary to the regions in the rat AVP mRNA encoding AVP (2- 9) and AVP-NP (1-8), to that in the rat OXT mRNA encoding OXT-NP (1-8), and to that in the toad AVT mRNA encoding AVT (—1 to 7). They are thus referred to as AVP, AVP-NP, OXT-NP and AVT-OXT probes, respectively. The nucleotide sequence of AVT/OXT probe is exactly complementary to the corresponding re- gion of OXT mRNA, while the nucleotide at position 9 of this probe is mismatched with the counterpart of AVP mRNA (i.e., 95% homology). The AVP probe has 2 mismatching positions with the AVT mRNA (91% homology), and 6 mis- matching positions with the OXT mRNA (73% homology). The AVP-NP and OXT-NP probes differ at 10 positions (55% homology). The homology of the corresponding region of toad mesotocin mRNA with the AVT-OXT probe is 77%, and that with the AVP probe is 73%. The probes were labeled at the 5’ ends with T4 Detection of Neuropeptide mRNAs 399 AVP mRNA Sig | AVP AVP-NP | GP | AVP(2-9) AVP-NP(1-8) AVP Cys Tyr Phe Gln Asn Cys Pro Arg Gly AVP mRNA 5’ UGC UAC UUC CAG AAC UGC CCA AGA GGA 3? Probe 3’ G AAG GTC TTG ACG GGT TCT CCT 5” Template AC TTC CAG AAC TGC AVP-NP Ala Tyr Ser Asp Met Glu Leu Arg AVP-NP_ mRNA GCC ACA UCC GAC AUG GAG CUG AGA Probe GG TGT AGG CTG TAC CTC GAC TC OXT mRNA Sig | OXT OXT-NP OXT-NP(1-8) OXT-NP Ala Ala Leu Asp Leu Asp Met Arg OXT-NP mRNA GCU GCG CUA GAC CUG GAU AUG CGC Probe GA CGC GAT CTG GAC CTA TAC GC AVT mRNA sig | avt | AVT-NP GP AVT ((-1)-7) AVT Ala Cys Tyr Ile Gln Asn Cys Pro AVT mRNA GCC UGC UAC AUC CAG AAC UGC CCC Probe GG ACG ATG TAG GTC TTG ACG GG Fic. 1. Design of synthetic oligonucleotide probes in the present experiments. Nucleotide sequences of the probes are shown with those of the com- plementary mRNAs. Amino acid sequences en- coded by the mRNA sequences are also shown. Numbers in parentheses indicate amino acid posi- tions. Sig, signal peptide; NP, neurophysin; GP, glycoprotein. polynucleotide kinase using [y—*°P] ATP to a final specific activity of 4-610’ cpm/zg by the proce- dure of Maxam and Gilbert [15]. Radioactivity of the labeled probe was measured by liquid scintilla- tion counting of Cerenkov radiation. Procedure for in situ hybridization After rehydration, tissue sections were treated with proteinase K (1 «g/ml; Sigma, type XI) in 0.1 M Tris buffer (pH 8.0) containing 50 mM EDTA at 37°C for 30min, and were briefly washed in doubly deionized water at room temperature. They were then rinsed in 2 SSC (1 x SSC contains 0.15M NaCl and 0.015 M_ sodium citrate), preincubated in a hybridization buffer (0.9M NaCl, 6mM EDTA, 0.2% bovine serum albumin, 0.2% Ficoll, 0.2% polyvinylpyrrolidone and 100 vg/ml denatured salmon sperm DNA in 90 mM Tris buffer, pH 7.5) at room temperature for 1 hr, and were placed in a moist chamber. The radiolabeled oligonucleotide probe was diluted to 1x 10* cpm/sl in hybridization buffer, and 80 A of the probe solution was applied to each slide glass. Sections were coverslipped, and were incubated at 30°C overnight. After removing coverslips in cold 6X SSC, the sections were washed in 6 x SSC firstly at 4°C for 10 min, then at about 20°C for 20 min twice, and again at 4°C for 10 min. The sections were then dehydrated through graded ethanols (70, 90 and 100%) containing 0.3M ammonium acetate and were air-dried. Thereafter, the sec- tions were dipped in Sakura NR-M2 emulsion diluted 3:2 with 0.3M ammonium acetate, air- dried for 30min, and were exposed for 1 to 3 weeks. After development in Kodak D-19 and fixation, they were dehydrated and were coverslip- ped with Permount (Fisher). Methodological checks Proteinase treatment The proteinase treat- ment after rehydration has been considered to increase accessibility of the probes to tissue mRNAs. We examined whether the proteinase treatment actually increase hybridization signals in the rat hypothalamic sections fixed by PFA only and those fixed by the PGP solution. Acetylation Acetylation of tissue sections was reported to decrease non-specific binding of probes, so that background could be reduced [21]. Therefore, proteinase treated rat tissue sections were immersed in freshly prepared 0.25 % acetic anhydride in 0.1M triethanolamine buffer (pH 8.0) for 10 min prior to preincubation. Effect of long-term storage of tissue sections Tissue sections from the same rats were separated into several groups, and were left unhydrated. They were kept in a desiccated box at a cool place. A group of tissue sections were periodically taken out, and the AVP mRNA was stained by ISH method during a period of more than 18 months. Concentration of labeled probes Hybridiz- ation mediums containing different levels of probe concentrations between 5X10* cpm/;d to 2x 10* cpm/yl were prepared, and were applied to tissue sections to determine an appropriate probe con- centration. Specificity tests for ISH Sections treated with proteinase were incubated with ribonuclease (100 vg/ml; BDH Chemicals) in 0.1 M Tris-HCl (pH 7.5) at room temperature for l hr, and were washed in doubly deionized water. The sections RNase_ pretreatment 400 S. Hyopo, M. Fustwara et al. were then hybridized with labeled AWVP-NP probe. Estimation of melting temperature (Tm) When a probe molecule is paired with the com- plementary mRNA region by hydrogen bonds, the Tm value experimentally determined was similar to those empirically determined and theoretically calculated [22]. As for oligonucleotide probes of around 20 bases, empirical Tm values for filter hybridization were between 50-60°C. An ex- perimental Tm value was determined by modifying the washing procedure after incubation with the probes, that is, tissue sections were washed 6xSSC at a series of graded temperature (18- 70°C) for 20 min after rinse in cold 6XSSC. The AVT/OXT probe was used in this experiment. In the hybridized rat sections, a 100 “mx 100 ~m square was settled in the OXT region of the PVN. The specific numbers of silver grains within the squares were determined, and were plotted to estimate graphically the Tm value after the logit transformation. Absorption test A 14mer template oligonu- cleotide complementary to the AVP probe (Fig. 1) was synthesized, and a 20-fold amount was added to a hybridization medium so as to absorb the probe. Rat hypothalamic sections were incubated in this absorbed hybridization medium. Competition test Rat hypothalamic sections were incubated in a hybridization medium contain- ing the labeled AVP-NP probe and a 10-fold amount of unlabeled AVP-NP probe. A similar experiment was performed also for the OXT-NP probe. As the control of these competition tests, the unlabeled mismatching probes were added to the hybridization mediums, e.g., the unlabeled OXT-NP probe to the labeled AVP-NP probe and vice versa. Use of different probes to the same mRNA The AVP and AVP-NP probes were com- plementary to different regions in the same mRNA. The localization of the AVP probe was thus compared with that of the AVP-NP probe. In addition, the same amounts of labeled AVP and AVP-NP probes were mixed so as to keep the radioactivity of incubation medium at 1x10* cpm/sd, and were applied to rat hypothalamic sections. Cross species hybridization The limitation of the oligonucleotide probes to discriminate mis- matching sequences was examined by using natu- rally occurring homologues of mRNAs of neurohy- pophysial hormones in the rat and the toad hypothalami. The AVP probe was applied to sections of the toad hypothalamus, and the result- ing hybridization signals were compared to those obtained by use of the AVT/OXT probe. Mean- while, the AVT/OXT probe was applied to sec- tions of the rat hypothalamus. Correspondence to immunohistochemical local- ization of neurohypophysial hormones The dis- tributions of hybridization signals were compared with immunohistochemical localization of neurohypophysial hormones in the same or adja- cent sections of the rat and toad hypothalami. For precise comparison, pairs of mirror image sections of the rat hypothalamus were utilized. Immunohistochemistry Tissue sections for immunohistochemistry were stained by the avidin-biotin-peroxidase complex (ABC) method using Vectastain ABC kit (Vec- tor), the procedure of which was described else- where [11]. In the present study, rabbit anti-A VP (Bioproducts, batch #001) was diluted 1:32,000 with phosphate-buffered saline (PBS) containing 0.5% bovine serum albumin; and rabbit anti-OXT (a gift from Professor S. Kawashima, Hiroshima University) was diluted 1:20,000 with PBS. Since the anti-AVP antiserum cross-reacts completely with AVT, toad hypothalamic sections were stained with this antiserum as was described previously [23]. ISH and immunohistochemistry double staining Tissue sections were first stained immunofluores- cently with fluorescein-labeled avidin D (Vector; diluted 1:250 with bicarbonate-buffered saline, pH 8.2) that was replaced with ABC, photographed with a fluorescence microscope, and were proces- sed for ISH. The use of an IgG-fractionated antiserum was required for this procedure. Other- wise, intensity of hybridization signals was markedly reduced probably by degradation of mRNAs by RNase in the serum. Specificity tests of immunohistochemistry In addition to the specificity tests previously de- Detection of Neuropeptide mRNAs 401 scribed [11], tissue sections were stained with AVP and OXT antisera preabsorbed with antigen conju- gated CNBr-Sepharose 4B (Pharmacia) columns. These tests confirmed the specificity of immunohis- tochemical stainings in the present study. RESULTS Autoradiographic silver grains that represent hybridization signals of the AVP, AVP-NP and OXT-NP probes were localized densely over the magnocellular neurons of the SON, the PVN, the circular nucleus, the anterior commissural nucleus (ACN) and other accessory magnocellular nuclei in the rat hypothalamus (Figs.2 and 3). The localization of AVP probe coincided with that of AVP-NP probe, while that of OXT-NP probe showed an independent pattern. The localization of hybridization signals was consistent with the immunohistochemical distribution of correspond- ing neurohypophysial hormones (Figs. 2 and 3). It was also true in the toad hypothalamus in which the AVT/OXT probe showed similar distribution to immunoreactive (ir) AVT in the magnocellular part of the preoptic nucleus (Fig. 5). In the rat hypothalamus, magnocellular neurons in the ven- Fic. 2. Immunoreactive (ir) AVP (a) and OXT (c) neurons and hybridization signals of the AVP mRNA (b) and the OXT mRNA (qd) in the paraventricular nucleus. Note parallel distribution of autoradiographic signals for the AVP mRNA (b) to AVP-ir neurons in mirror image section (a, counterstained with cresyl violet). Distribution of signals for the OXT mRNA (d) is also parallel to OXT-ir neurons in the adjacent section (c). Scale bar, 50 «zm. 402 S. Hyopo, M. Fustwara et al. Fic. 3. 4 4 .@ Immunoreactive (ir) OXT (a) and AVP (b, c) neurons and hybridization signals of the OXT mRNA (d) Ge ee and the AVP mRNA (e, f) in the anterior commissure nucleus (a, d), the circular nucleus (b, e), and the fornical nucleus (c, f). Autoradiographic signals for the OXT mRNA (d) are distributed parallel to OXT-ir neurons in the adjacent section (a). Signals for the AVP mRNA (e, f) are also distributed parallel to AVP-ir neurons in mirror image sections (b, c, counterstained with cresyl violet). Scale bar, 50 um. tral region of the SON and the dorsolateral region of the PVN were mainly AVP-ir, coinciding well with the localization of the AVP and AVP—NP probes. While the dorsal region of the SON, the ventromedial region of the PVN and the ACN were composed of OXT-ir neurons, and the OXT- NP probe was localized in these regions (Figs. 2 and 3). However, noticeable hybridization signals were not detected in the suprachiasmatic nucleus and the parvocellular part of the PVN which TABLE 1. include AVP-ir parvocellular Hy- bridization signals were also not found in the median eminence and the pars nervosa, the ter- minal regions of neurosecretory fibers. neurons. Preparation of tissue sections and methodological checks Among the fixatives tested, 4% PFA gave the most intense hybridization signals in the rat hypothalamus (Table 1). The use of a PGP Comparison of fixatives for in situ hybridization (ISH) of the AVP mRNA and immunohistochemistry (IHC) of AVP in the rat hypothalamus Fixatives Proteinase ISH me treatment 001 1285) Bouin’s solution Yes + Stet ++ Bouin’s without acetic acid Yes ++ NT oP or 4% PFA No +444 — ++ 2% PFA+1% GLA Yes ++ NT NT 2% PFA+1% GLA+1% PA Yes +++ = + 2% PFA+1% GLA+1% PA No + +++ +++ PFA, paraformaldehyde; GLA, glutaraldehyde; PA, picric acid. Note: —, not (or scarcely) stained; +, weakly stained; ++, moderately stained; +++, strongly stained; ++-+-+, very strongly stained; NT, not tested. ‘9 Anti-vasopressin antiserum (Bioproducts, batch 4001). >) Anti-vasopressin antiserum (Bioproducts, batch #1285). Detection of Neuropeptide mRNAs 403 solution also yielded satisfactorily intense staining results, when tissue sections were treated with proteinase prior to incubation with labeled probes. Results of hybridization in frozen sections were similar to those in paraffin sections. These observations were consistent among the four oligo- nucleotide probes, while effects of fixation on immunohistochemical staining were rather com- plex, that is, stainabilities differed between the antisera utilized (Table 1). We are currently using the PGP solution with a proteinase treatment in the hybridization procedure. The proteinase treatment of tissue sections fixed with the PGP solution markedly increased specific hybridization signals, although intensity of signals in 4 % PFA fixed sections was not increased by this treatment. Acetylation seemed to prevent not only non- specific background binding of labeled probes, but also their specific base-pairing with the com- plementary nucleotide sequences. Thus, we did not adopt this treatment in our method. On the other hand, the increase in probe concentration above 1X10* cpm/zl markedly augmented unde- sirable background. The probe concentration of 5x10° cpm/p gave clearly identifiable specific hybridization signals, although the signals were weak. These results indicate that the probe concentration around 1x10* cpm/zl may be appropriate for the oligonucleotide-mRNA ISH method for neurohypophysial hormones. The distributional pattern and intensity of hy- bridization signals in paraffin sections stored for up to 18 months were similar to those in the initial sections which were hybridized immediately after being cut. Specificity tests RNase pretreatment Hybridization signals in the magnocellular nuclei were almost completely diminished to the background level by the RNase pretreatment. Tm When the temperature of washing after hybridization with the AVT/OXT probe was raised to 50°C, hybridization signals were apparently reduced. Signals were further decreased along with elevation of washing temperature, and at about 65°C, almost all signals were removed. The value of Tm estimated from the plot (Fig. 4) was about 51°C for pairing between the AVT/OXT probe and the OXT mRNA. Absorption and competition tests Addition of excess amounts of the synthetic template and the unlabeled probe to the hybridization mediums markedly reduced specific localization of silver grains. On the other hand, an excess amount of unlabeled mismatching probe did not change the localization and intensity of hybridization signals. The use of alternate probes to the same mRNA Hybridization signals of the AVP and AVP-NP probes were localized in the same areas in the SON and the PVN. When the unlabeled AVP probe was added to the labeled AVP-NP probe and vice versa, hybridization signals were not altered. Furthermore, the application of mixed AVP and AVP-NP probes conspicuously increased hybri- dization signals (Fig. 6), showing that the AVP and the AVP-NP probes may not interact each other. Cross species hybridization In the magno- 100 Percent decrease of specific grains a oO fe) 20 40 60 80 Washing temperature (°C) Fic. 4. Tm analysis for pairing of the AVT/OXT probe and the OXT mRNA by in situ hybridization. (a) Thermal denaturation of probe-mRNA duplexes. (b) Logit transformation and estimation of Tm value, showing that the value is about 51°C for pairing of the AVT/OXT probe and the OXT mRNA. 404 S. Hyopo, M. Fustwara et al. Fic. 5. Immunoreactive (ir) AVT neurons (a) and hybridization signals of the AVT mRNA in the magnocellular part of the toad preoptic nucleus (b, c). The AVT/OXT probe yielded intense hybridization signals (b), which are distributed parallel to AVT-ir neurons (a). signals (c). Scale bar, 50 «m. reed obs . : irs Fic. 6. In situ hybridization of the AVP mRNA with the AVP probe only (a) and the mixture of AVP and AVP-NP probes (b) in the paraventricular nu- cleus. Note that the density of silver grains by the mixed probe is higher than by the AVP probe only. Scale bar, 50 um. In contrast, the AVP probe yielded only weak hybridization cellular part of the toad preoptic nucleus, the AVT/OXT probe yielded intense hybridization signals, while those given by the AVP probe were faint (Fig. 5). In contrast, in the rat hypothalamic sections, intensity of hybridization signals induced by the AVT/OXT probe was similar to that given by the AVP probe. The signals by the AVT/OXT probe were localized not only in the SON and the PVN regions where ir-OXT neurons are predomi- nant, but also in the regions occupied by ir-AVP neurons with similar intensity to that seen in the OXT regions. The same result was obtained in another independent study on the hypothalamus of the ICR strain mouse (unpublished). Distribution of ISH signals vs. that of AVP- and OXT-immunoreactivity As is described above, the distribution of hybridization signals was consist- ent with the immunohistochemical localization of related peptides. However, the intensity of hybri- dization signals did not necessarily correlate with that of immunoreactivity, as was reported pre- viously [11]. Immunoreactive neurons were some- times not labeled with the probe, and vice versa. DISCUSSION The present study showed that 22mer synthetic oligonucleotides as probes for mRNAs of neurohy- Detection of Neuropeptide mRNAs 405 pophysial hormones were localized in the hypo- thalamic magnocellular neurosecretory nuclei in the toad and the rat after ISH stainings. The distributions of the probes were consistent with those of immunoreactivities to related peptides, e.g., the AVP probe was localized in the dorso- lateral region of the PVN and the ventral region of the SON where ir-AVP neurons are predominant. However, suprachiasmatic neurons in which Uhl and Reppert [6] demonstrated intense hybridiza- tion signals for the AVP mRNA did not show noticeable hybridization signals in our study. Since the intense signals in the suprachiasmatic nucleus have been reported only by Uhl and Reppert, we consider that the above discrepancy is due to longer autoradiographic exposure time by them, judging from their published photographs. The disappearance of hybridization signals after the RNase pretreatment and the estimated Tm value indicate that the oligonucleotide probes were actually paired with tissue RNAs by hydrogen bonds. Other specificity tests showed that the present probes specifically recognize the com- plementary nucleotide sequences in particular mRNAs. The consistency of the distribution of hybridization signals for AVP and OXT mRNAs with those of AVP and OXT immunoreactivities further supports the occurrence of specific base pairings between the probes and the related tissue mRNAs. The discrepancy in the distribution of hybridization signals and immunoreactivities at the cellular level must be considered with information concerning secretory activity of neurosecretory neurons [11]. We thus convince that the AVP and AVP-NP probes were hybridized with the rat AVP mRNA, the OXT-NP probe paired with the rat OXT mRNA, and the AVT/OXT probe recognized the toad AVT mRNA and the rat OXT and AVP mRNAs. The cross species hybridization study clarified that the 22mer oligonucleotide probes discrimi- nated nucleotide sequences which include mis- matching bases at more than 2 positions, although one-point mismatching was not recognized. This result strongly supports a reliability of the present ISH method in the study of mRNAs for neurohy- pophysial hormones. Further, it suggests that the oligonucleotide-mRNA ISH technique is widely applicable to studies of gene expressions for various peptides and proteinaceous hormones with high fidelity. The method may also be employable in detection of expressed genes concerning he- reditary diseases. Our present study showed that, as to the hypothalamic magnocellular neurons, the distribu- tion of hybridization signals is coincide with that of immunohistochemical staining, indicating that the ISH method is sufficiently sensitive to study gene expression of neurohormones. Nonetheless, one of disadvantages in the ISH method using oligo- nucleotide probes is that labeling of multiple sites in a single probe molecule is rather difficult. The present result that an application of a mixture of the AVP and AVP_-NP probes yielded a marked increase in specific signals suggests a solution for the above problem, since an interaction between the AVP and AVP-NP probes seems to be negligible. Thus, a use of mixed probes each of which recognized a different region in the same mRNA probably enhances hybridization signals, when an increase in the sensitivity of the oligo- nucleotide-mRNA ISH method is required. ACKNOWLEDGMENT The authors would like to thank Professor S. Kawashima, Hiroshima University, for providing the antiserum to oxytocin. REFERENCES 1 Ivell,R. and Richter,D. (1984) Structure and comparison of the oxytocin and vasopressin genes from rat. Proc. Natl. Acad. Sci. USA, 81: 2006- 2010. Lathe, R. (1985) Synthetic oligonucleotide probes deduced from amino acid sequence data. Theoreti- cal and practical considerations. J. Mol. Biol., 183: 1-12. 3. McCabe, J.T., Morrell, J.1., Ivell, R., Schmale, H., Richter, D. and Pfaff, D.W. (1986) Jn situ hybridization technique to localize rRNA and mRNA in mammalian neurons. J. Histochem. Cytochem., 34: 45-50. 4 Sherman, T.G., McKelvy, J. F. and Watson, S. J. (1986) Vasopressin mRNA regulation in individual hypothalamic nuclei: a northern and in situ hybrid- ization analysis. J. Neurosci., 6: 1685-1694. 5 Uhl, G.R., Zingg, H. H. and Habener, J. F. (1985) i) 10 13 14 406 Vasopressin MRNA in situ hybridization: localiza- tion and regulation studied with oligonucleotide cDNA probes in normal and Brattleboro rat hypothalamus. Proc. Natl. Acad. Sci. USA, 82: $555-5559. Uhl, G. R. and Reppert,S.M. (1986) Supra- chiasmatic nucleus vasopressin messenger RNA: circadian variation in normal and Brattlebore rats. Science, 232: 390-393. Wolfson, B., Manning, R. W., Davis, L. G., Arent- zen, R. and Baldino, F., Jr. (1985) Co-localization of corticotropin releasing factor and vasopressin mRNA in neurons after adrenalectomy. Nature, 315: 59-61. Pavel, S. (1975) Vasotocin biosynthesis by neurohy- pophysial cells from human fetuses. Evidence for its ependymal origin. Neuroendocrinology, 19: 150- 159. Pavel, S. (1980) Presence of relatively high concen- trations of arginine vasotocin in the cerebrospinal fluid of newborns and infants. J. Clin. Endocrinol. Metab., 50: 271-273. Nojiri, H., Sato, M. and Urano, A. (1985) In situ hybridization of the vasopressin mRNA in the rat hypothalamus by use of a synthetic oligonucleotide probe. Neurosci. Lett., 58: 101-105. Nojiri, H., Sato, M. and Urano, A. (1986) Increase in the vasopressin MRNA level in the magnocellular neurosecretory neurons of water-deprived rats: in situ hybridization study with the use of synthetic oligonucleotide probe. Zool. Sci., 3: 345-350. Fujiwara, M., Hyodo, S., Sato, M. and Urano, A. (1985) Changes in vasopressin and oxytocin mRNA levels in the rat hypothalamus by oral hypertonic saline. Zool. Sci., 2: 990 (Abstract). Majzoub,J.A., Rich,A., vanBoom,J. and Habener, J. F. (1983) Vasopressin and oxytocin mRNA regulation in the rat assessed by hybridiza- tion with synthetic oligonucleotides. J. Biol. Chem., 258: 14061-14064. Wallace, R. B., Shaffer, J., Murphy, R. F., Bonner, J., Hirose, T. and Itakura, K. (1979) Hybridization of synthetic oligodeoxyribonucleotides to © X¥ 174 20 21 22 23 S. Hyopo, M. Fustwara et al. DNA: the effect of single base pair mismatch. Nucleic Acids Res., 6: 3543-3557. Maxam, A. and Gilbert, W. (1980) Sequencing end-labeled DNA with base-specific chemical cleav- ages. Methods in Enzymology, 65: 499-560. Brahic, M. and Haase, A. T. (1978) Detection of viral sequences of low reiteration frequency by in situ hybridization. Proc. Natl. Acad. Sci. USA, 75: 6125-6129. Lawrence, J. B. and Singer, R. H. (1985) Quantita- tive analysis of in situ hybridization methods for the detection of actin gene expression. Nucleic Acids Res., 13: 1777-1799. Moench, T. R., Gendelman,H.E., Clements, J. E., Narayan, O. and Griffin, D. E. (1985) Efficien- cy of in situ hybridization as a function of probe size and fixation technique. J. Virol. Meth., 11: 119-130. Nojiri, H., Ishida, I., Miyashita, E., Sato, M., Ura- no, A. and Deguchi, T. (1987) Cloning and se- quence analysis of cDNAs for neurohypophysial hormones vasotocin and mesotocin for the hypotha- lamus of toad, Bufo japonicus. Proc. Natl. Acad. Sci. USA, 84: 3043-3046. McBride, L. J. and Caruthers,M.H. (1983) An investigation of several deoxynucleoside phosphor- amidites useful for synthesizing deoxyoligonu- cleotides. Tetrahedron Lett., 24: 245-248. Hayashi, S., Gillam,I.C., Delaney, A.D. and Tener, G. M. (1978) Acetylation of chromosome squashes of Drosophila melanogaster decreases the background in autoradiographs from hybridization with [!*°I]-labeled RNA. J. Histochem. Cytochem., 26: 677-679. Kelsey, J. E., Watson, S.J., Burke,S., Akil, H. and Roberts, J.L. (1986) Characterization of proopiomelanocortin mRNA detected by in situ hybridization. J. Neurosci., 6: 38-42. Jokura, Y. and Urano, A. (1985) Projections of luteinizing hormone-releasing hormone and vasoto- cin fibers to the anterior part of the preoptic nucleus in the toad, Bufo japonicus. Gen. Comp. Endocri- nol., 60: 390-397. ZOOLOGICAL SCIENCE 5: 407-413 (1988) Effect of Hypothalamic Extract on the Prolactin Release from the Bullfrog Pituitary Gland with Special Reference to Thyrotropin-Releasing Hormone (TRH) TATSUNORI SEKI, SAKAE KikuyAMa! and Mitsuo Suzukr’ Department of Anatomy, Juntendo University School of Medicine, Bunkyo-ku, Tokyo 113, and ‘Department of Biology, School of Education, Waseda University, Shinjuku-ku, Tokyo 160, and *Department of Physiology, Institute of Endocrinology, Gunma University, Maebashi 371, Japan ABSTRACT— Acid extract of bullfrog hypothalami but not of rat hypothalami stimulated the release of prolactin (PRL) from the bullfrog pituitary gland in vitro. Since frog hypothalamus is known to contain thyrotropin-releasing hormone (TRH), a stimulator of PRL release, much more than rat hypothalamus, experiments were performed to determine whether PRL-releasing activity of the frog hypothalamic extract is derived from TRH it contains. PRL-releasing activity in the hypothalamic extract was nullified by incubation with bullfrog plasma. When the extract was fractionated on Sephadex G-25 chromatography, significant PRL-releasing activity was found in the fraction which is presumed to contain TRH. The activity of the hypothalamic extract was markedly reduced by co-incubation with IgG fraction separated from antiserum to TRH. These results indicate that TRH © 1988 Zoological Society of Japan is one of the major PRL-releasing factors in the hypothalamic extract. INTRODUCTION Acid extract of the amphibian hypothalamus is known to stimulate the release of prolactin (PRL) from amphibian pituitary glands [1-3]. However, little is known about the identity of PRL-releasing factor in the hypothalamic extract. On the other hand, synthetic thyrotropin-releasing hormone (TRH) shows a potent PRL-releasing activity in amphibians [2, 4-7] as well as in mammals [8]. In amphibians, it is not clear whether TRH regulates TSH secretion. According to the results obtained by most of the investigators, there is no indication that synthetic TRH stimulates TSH release from amphibian pituitary gland [9-12], while some recent data pointed out the possibility that TRH induces TSH release in the amphibian [13]. It has been reported that hypothalamic TRH concentra- tion is much higher in amphibians than in mam- mals [11, 14-16]. Immunohistochemical study of the bullfrog hypothalamus revealed that TRH Accepted September 29, 1987 Received August 26, 1987 neurons exist in the hypothalamus and their axons terminate in the median eminence, suggesting that TRH neurons have a physiological role in the regulation of adenohypophyseal hormone release [16]. The present investigation was undertaken to ascertain whether the PRL-releasing activity in the hypothalamic extract is due to TRH. A prelimi- nary report has been published elsewhere [3]. MATERIALS AND METHODS Incubation of pituitary gland Adult bullfrog (Rana catesbeiana) weighing 250-400 g were sacrificed by decapitation. The anterior pituitary gland was removed, weighed and hemisected. Two hemipituitaries were placed in a glass vial containing 200 wl of 67% Eagle MEM (Nissui Seiyaku Co., Ltd., pH 7.4). The vials were incubated in a Dubnoff metabolic incubator in an atmosphere of 95% O,-5% CO. After 1 hr of preincubaion, the medium was replaced with fresh One containing a test substance. Incubation was 408 T. Sexi, S. KIKUYAMA AND M. SuZUKI carried out for 8 hr at 25°C, since it was previously verified that bullfrog pituitaries incubated in the same condition as described above continue to release PRL until at least 28 hr at a linear rate [7]. Radioimmunoassay PRL in the medium was measured by a homolo- gous radioimmunoassay for bullfrog PRL de- veloped in Yamamoto and Kikuyama [17]. Preparation of tissue extract Fresh hypothalami were homogenized in cold 0.1 N HCl (50 -d/hypothalamus) with a teflon homogenizer and centrifuged at 10,000 xg for 30 min. The supernatant was neutralized with 1 N NaOH and recentrifuged. Acid extracts of frog cerebrum, frog neuro-intermediate lobe and rat hypothalamus were prepared in the same manner. Inactivation of PRL-releasing activity of hypotha- lamic extract by frog plasma Fifty microliters of the extract (one hypotha- lamic equivalent) was mixed with 50 vl of bullfrog plasma or distilled water and incubated for 3 hr at 37°C. After incubation, PRL releasing activity of this mixture was tested. Chromatography Hypothalamic extract, derived from 200 hypo- thalamic flagments was filtered through Millipore filter (HAWP) and applied to a 1.5x110cm Sephadex G—25 column. Each fraction consisting of 3 ml was collected by eluting with 0.1 N acetic acid. *H-TRH (1-proline-2, 3-H(N)-TRH, New England Nuclear) was also chromatographed to determine the position in which TRH is eluted. Fractions were assembled into three pools; frac- tion I (FI) eluted faster than TRH, fraction II (FII) eluted with TRH and fraction III (FIII) eluted later than TRH. The pooled fractions were lyophilized, dissolved in the incubation medium and tested for PRL-releasing activity. Inactivation of PRL-releasing substance by IgG from anti-TRH serum Immunoneutralization of hypothalamic TRH was performed by the use of IgG fraction from anti-TRH serum which had been prepared by immunizing rabbit with TRH conjugated to bovine serum albumin according to the procedures de- scribed previously [18]. Protein A (Pharmacia Fine Chemicals) was combined with CNBr-acti- vated Sepharose 4B (Pharmacia Fine Chemicals) by the method of Miller and Stone [19]. Anti-TRH serum or normal rabbit serum was applied to 0.84.5 cm protein A-Sepharose column. The IgG fraction was dialized, lyophilized and dis- solved in 50mM Hepes buffer containing 0.7% NaCl (HBS). Hypothalamic extract was incubated with the solution of IgG from anti-TRH serum, from normal rabbit serum (NRS) or HBS at 4°C for 48hr. After incubation, the medium was centrifuged. To test the PRL-releasing activity, aliquot of the supernatant was added to the medium in which pituitaries were placed. One milliliter of medium used for the test contained the extract from 0.2 hypothalamus which had been incubated with IgG obtained from 60 «1 of anti- serum or NRS. It has been ascertained that 1 pl of the antiserum can inactivate nanogram quantities of TRH. RESULTS The response of bullfrog pituitaries to various doses of frog hypothalamic extract is shown in Figure 1. Hypothalamic extract, equivalent to 0.01-1 hypothalamic fragment stimulated the re- lease of PRL from bullfrog pituitary gland in a dose-dependent manner. The dose of 0.1 and 1 hypothalamic equivalent per 1 ml medium caused a Significant increase in the PRL release. Effect of rat hypothalamic extract on the PRL release from the bullfrog pituitary gland was examined. No significant difference in the amount of PRL released into the medium was observed between the control medium (300+39 ng/mg pituitary) and the medium containing extract from one rat hypothalamus per milliliter (349+ 53 ng/ mg pituitary). PRL-releasing activities of extracts from the hypothalamus, cerebrum and neuro-intermediate lobe of the pituitary gland were shown in Figure 2. Addition of the extract from the hypothalamus or neuro-intermediate lobe to the medium significant- ly enhanced the release of PRL. The extract from PRL Releasing Activity in Frog Brain 409 PRL in medium (yg/mg pituitary) (e 0.01 0.1 1 HE HE HE /mi /mi /mi Fic. 1. Effect of various doses of hypothalamic extract (HE) on PRL release. One milliliter of medium contains acid extract from 0.01-1 hypothalamus. C, control medium. Each value represents mean + SEM for 7 determinations. Significance of differ- ence (analysis of variance): a vs b, a vs c, P<0.001. (pg/mg pituitary) PRL in medium Cc DW fPlasma + + 0.1 0.1 HE HE /mi /mi PRL in medium (pg/mg pituitary) NILE Cc CE HE Fic. 2. Effect of extract of cerebrum (CE), hypothal- amus (HE) and neuro-intermediate lobe (NILE) on PRL release. One milliliter of medium contains extract from 3 mg of each tissue. Each value repre- sents mean+SEM for 7 determinations. Signif- icance of difference (analysis of variance): a vs b, a vs c, P<0.001. the cerebrum caused a slight increase, but this increase was statistically not significant. When the hypothalamic extract was incubated with bullfrog plasma, the PRL-releasing activity was completely lost (Fig. 3). Figure 4 shows the profiles of Sephadex G—25 chromatography of the hypothalamic extract and the distribution of radioactivity of 7#H-TRH when chromatographed. Among the three pooled sam- ples, FI exhibited a marked stimulatory effect on the PRL release, while FHI had no releasing activity and FI had a slight but not significant Fic. 3. Effect of bullfrog plasma on PRL-releasing activity of bullfrog hypothalamic extract. One mil- liliter of medium contains extract from 0.1 hypotha- lamus subsequently incubated with distilled water (DW) or bullfrog plasma as described in Materials and Methods. Each value represents mean+SEM for 7 determinations. C, control medium. Signif- icance of difference (analysis of variance): a vs b, b vs c, P<0.001. 410 T. SEKI, S. KIKUYAMA AND M. SUZUKI 0.3 a 0.2 i ro) co “ = O o 3 0.1 © 2 7 (e) O 20 40 7) Q =} = 15 'o << 10 & Oo x oc Bedi = Oo =! | 60 80 Tube number (tube volume, 3 ml) Fic. 4. Sephadex G-25 column (1.5110 cm) chromatography of bullfrog hypothalamic extract and of *#H-TRH. Absorbance at 280 nm is denoted by the continuous line and radioactivity of 7H-TRH by the broken line. releasing activity (Fig. 5). Figure 6 shows the effect of immunoneutraliza- tion of TRH in the hypothalamic extract on the PRL release. Hypothalamic extract incubated with either IgG from NRS or HBS stimulated the PRL release. Incubation of the hypothalamic extract with IgG from anti-TRH serum consider- ably diminished the PRL-releasing activity of the hypothalamic extract. DISCUSSION It was demonstrated by the present experiment that the frog hypothalamic extract is effective in promoting PRL release from the bullfrog pituitary gland in vitro, while the rat hypothalamic extract is ineffective. It has also been reported that the frog hypothalamic extract stimulates the release of PRL from the rat pituitary gland [20]. These results indicate that the effect of frog hypothalamic extract on the PRL release is predominantly stimulatory. TRH is known to have a potent PRL-releasing activity in amphibians [2-6] as well as in mammals [8]. Among the three fractions separated from the bullfrog hypothalamic extract by Sephadex G-25 chromatography, the fraction (FII) which is pre- sumed to contain TRH had the strongest PRL- releasing activity. In the present experiment, no attempt was made to eliminate PRL-inhibiting factors presumed to be present in the hypothala- mic extract. Dopamine existing in the frog hypothalamus has been postulated to be a PRL- inhibiting factor [5, 21-25]. According to our test-run on Sephadex G-25 column, the monoamine was eluted later than TRH, indicating that dopamine will be included in FIII. Im- munoneutralization of the hypothalamic extract PRL Releasing Activity in Frog Brain 41] 1.0 0.5 PRL in medium (yg/mg pituitary) O Fill Fic. 5. Distribution of PRL-releasing activity among fractions obtained by gel-filtration of bullfrog hypothalamic extract on Sephadex G-25. One milliliter of medium contains each fraction derived from 0.1 hypothalamus. Each value represents mean+SEM for 7 determinations. Significance of difference (analysis of variance): a vs b, P<0.001. Cc Fl Fil with IgG separated from anti-TRH serum resulted in a considerable degradation of PRL-releasing activity. These results strongly suggest that PRL- releasing activity in the hypothalamic extract is largely derived from TRH it contains. In perinatal rats, it has also been demonstrated that the PRL- releasing activity in the hypothalamic extract is attributable to the existing TRH [26]. The PRL-releasing activity of the hypothalamic extract was completely abolished by incubation with frog plasma. Preliminary study also revealed that NRS as well as anti-TRH serum inactivated the PRL-releasing activity of the hypothalamic extract. This may be the result of enzymatic degradation of substances bearing PRL-releasing activity. It is well known that TRH is degraded rapidly in the blood [27]. Accordingly, we used IgG fraction instead of antiserum for immunoneu- tralization. The present experiment revealed that extract of neuro-intermediate lobe tissue possesses a marked PRL-releasing activity. Neuro-intermediate lobe tissue as well as the hypothalamus of bullfrogs (ug/mg pituitary) PRL in medium c HE HE HE + + + HBS NRS Anti-TRH Fic. 6. Effect of immunoneutralization of hypotha- lamic TRH on PRL release. One milliliter of medium contains acid extract from 0.2 hypothal- amus which had been subsequently incubated with 60 yl of SO mM Hepes buffer containing 0.7% NaCl (HBS), IgG from normal rabbit serum (NRS) or IgG from anti-TRH serum. Each value represents mean+SEM for 7 determinations. C, control medium. Significance of difference (analysis of variance): a vs b, a vs c, b vs d, c vs d, P<0.001; a vs d, P<0.01. contains a considerable amount of TRH [15, 16]. Accordingly, the stimulatory effect of the extract from neuro-intermediate lobe may partly be due to TRH. In mammals, it has been reported that extract of posterior pituitary or hypophyseal stalk contains a PRL-releasing activity and that the substance in the extract is distinct from TRH [28- 1]. In the present experiment, an excess amount of IgG from anti-TRH serum was applied for the immunoneutralization of TRH in the hypotha- lamic extract, taking into consideration of the TRH content in the bullfrog hypothalamus [16]. However, complete depression of PRL-releasing activity was not observed. This suggests the existence of PRL-releasing factors other than TRH in the frog hypothalamic extract. Several sub- stances such as vasoactive intestinal peptide (VIP) and peptide histidine isoleucine (PHI) are known to stimulate PRL release from the pituitary gland 412 T. Seki, S. KIKUYAMA AND M. SuzukKI in amphibians [32] as well as in mammals [33-36]. Identification of other PRL-releasing substances in the bullfrog hypothalamus is under way. ACKNOWLEDGMENTS This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture, Japan and by a grant from Waseda University. REFERENCES 1 Hall, T. R. and Chadwick, A. (1979) Hypothalamic control of prolactin and growth hormone secretion in different vertebrate species. Gen. Comp. Endo- crinol., 37: 333-342. 2 Hall, T.R. and Chadwick, A. (1984) Effect of synthetic mammalian thyrotropin releasing hor- mone, somatostatin and dopamine on the secretion of prolactin and growth hormone from amphibian and reptilian pituitary glands incubated in vitro. J. Endocrinol., 102: 175-180. 3 Seki, T. and Kikuyama, S. (1986) Possible involve- ment of thyrotropin-releasing hormone in the re- lease of prolactin from the hypophysis of the bullfrog. In “Pars Distalis of the Pituitary Gland —Structure, Function and Regulation”. Ed. by F. Yoshimura and A. Gorbman, Elsevier Science Pub- lisher B. V., Amsterdam, pp. 247-249. 4 Clemons, G.K., Russell,S.M. and Nicoll, C.N. (1979) Effect of mammalian thyrotropin-releasing hormone on prolactin secretion by bullfrog adeno- hypophyses in vitro. Gen. Comp. Endocrinol., 38: 62-67. 5 Kikuyama, S. and Seki, T. (1983) Dopamine and serotonine control of prolactin release in bullfrog. In “Neuropeptides, Neurotransmitters and Regulation of Endocrine Processes”. Ed. by E. Endroczi, Academiai Kiado, Budapest, pp. 317-324. 6 Kuhn,E.R., Kikuyama,S., Yamamoto, K. and Darras, V. M. (1985) Jn vivo release of prolactin in Rana ridibunda following an intravenous injection of thyrotropin-releasing hormone. Gen. Comp. En- docrinol., 60: 86-89. 7 Seki, T. and Kikuyama, S. (1986) Effect of thyro- tropin-releasing hormone and dopamine on the in vitro secretion of prolactin by the bullfrog pituitary gland. Gen. Comp. Endocrinol., 61: 197-202. 8 Vale, W., Rivier, C. and Brown, M. (1977) Regula- tory peptides of the peptides of the hypothalamus. Ann. Rev. Physiol., 39: 473-527. 9 Etkin,W. and Gona,A.G. (1986) Failure of mammalian thyrotropin releasing factor preparation 22 to elicit metamorphic response in tadpoles. Endocri- nology, 82: 1067-1068. Gona, A. G. and Gona, O. (1974) Failure of syn- thetic TRF to elicit metamorphosis in frog tadpoles or red-spotted newts. Gen. Comp. Endocrinol., 24: 223-225. Taurog, A., Oliver, C., Eskay, R. L., Porter, J. C. and McKenzie, J. M. (1974) The role of TRH in the neoteny of the Mexican axolotl (Ambystoma mexi- canum). Gen. Comp. Endocrinol., 24: 267-279. Vandesande, F. and Aspeslagh, M. R. (1974) Fail- ure of thyrotropin releasing hormone to increase '*°I uptake by the thyroid in Rana temporaria. Gen. Comp. Endocrinol., 23: 355-356. Darras, V. M. and Kiihn, E. R. (1985) Increased plasma levels of thyroid hormones in a frog Rana ridibunda following intravenous administration of TRH. Gen. Comp. Endocrinol., 48: 469-475. Giraud, P., Gillioz, P., Conte-Devolx, B. and Oliver, C. (1979) Distribution de thyroliberine (TRH), a@melanocyte-stimulating hormone (a- MSH) et somatostatine dans les tissue de la Greno- ville verte (Rana esculenta). C. R. Acad. Sci. Paris, Ser D., 228: 127-129. Jackson, I. M. D. and Reichlin, S. (1974) Thyrotro- pin-releasing hormone (TRH): Distribution in hypothalamic and extra-hypothalamic brain tissues of mammalian and submammalian chordates. En- docrinology, 95: 854-862. Seki, T., Nakai, T., Shioda,S., Mitsuma,T. and Kikuyama, S. (1983) Distribution of immunoreac- tive thyrotropin-releasing hormone in the forebrain and hypophysis of the bullfrog, Rana catesbeiana. Cell Tissue Res., 233: 507-516. Yamamoto, K. and Kikuyama, S. (1982) Radioim- munoassay of prolactin in piasma of bullfrog tad- poles. Endocrinol. Japon., 29: 159-167. Nishiyama, T., Heike, Y., Matsuzaki, T., Kawano, H., Daikoku,S. and Suzuki,M. (1983) Im- munoreactive TRH-containing neurones in the rat hypothalamus. Biomed, Res., Suppl. 4: 65-74. Miller, T. J. and Stone, H. O. (1978) The rapid isolation of ribonuclease-free immunoglobulin G by protein A-Sepharose affinity chromatography. J. Immunol. Methods, 24: 111-125. Kiihn, E. R. and Engelen, H. (1976) Seasonal varia- tion in prolactin and TSH releasing activity in the hypothalamus of Rana temporaria. Gen. Comp. Endocrinol., 28: 277-282. Alonso-Bedate,M. and Delgado,M.J. (1983) Effects of prolactin and bromocriptine in Discoglos- sus pictus Otth (anuran amphibian) tadpoles. Comp. Biochem. Physiol., 74A: 763-772. Kikuyama, S. and Seki, T. (1980) Possible involve- ment of dopamine in the release of prolactin-like hormone from bullfrog pituitary gland. Gen. Comp. 23 24 25 26 27 28 29 PRL Releasing Activity in Frog Brain Endocrinol., 41: 173-179. Platt, J. E. (1976) The effects of ergocornine on the tail height, spontaneous and T,-induced meta- morphosis and thyroidal uptake of radioiodide in neotenic Ambystoma tigrinum. Gen. Comp. Endo- crinol., 28: 71-81. Seki, T. and Kikuyama, S. (1979) Effect of ergocor- nine and reserpine on metamorphosis in Bufo bufo japonicus tadpoles. Endocrinol. Japon., 26: 675- 678. Seki, T. and Kikuyama, S. (1982) Jn vitro studies on the regulation of prolactin secretion in the bullfrog pituitary gland. Gen. Comp. Endocrinol., 46: 473- 479. Khorram, O., Depalatis, L. R. and McCann, S. M. (1984) Hypothalamic control of prolactin secretion during the perinatal period in the rat. Endocrinolo- gy, 115: 1698-1704. McKelvy, J. F. (1983) Enzymatic degradation of brain peptides. In “Brain Peptides”. Ed. by D. T. Krieger, M. J. Brownstein and J. B. Martin, John Wiley & Sons, New York, pp. 117-134. Boyd, II, A.E., Spencer, E., Jackson, I. M. D. and Reichlin, S. (1976) Prolactin-releasing factor (PRF) in porcine hypothalamic extract distinct from TRH. Endocrinology, 99: 861-871. Kokubu, T., Sawano,S., Shiraki, M., Yamasaki, M. and Ishizuka, Y. (1975) Extraction and partial purification of prolactin-release stimulating factor in bovine hypothalami. Endocrinol. Japon., 22: 213- 217. 30 32 34 36 413 Szabo, M. and Frohman, L. A. (1976) Dissociation of prolactin-releasing activity from thyrotropin re- leasing hormone in porcine stalk median eminence. Endocrinology, 98: 1451-1459. Yasuda, N., Yasuda, Y. and Greer,S.E. (1984) Heterogeneity of activity of the prolactin-releasing factor in the bovine hypothalamo-neurohypophysial complex. J. Endocrinol., 103: 243-249. Koiwai, K., Kikuyama, S., Seki, T. and Yanaihara, N (1986) Jn vitro effect of vasoactive intestinal polypeptide and peptide histidine isoleucine of prolactin secretion by the bullfrog pituitary gland. Gen. Comp. Endocrinol., 64: 254-259. Kato, Y., Iwasaki, Y.,_ Iwasaki, J., Abe, H., Yanaihara, N. and Imura, H. (1978) Prolactin re- lease by vasoactive intestinal polypeptide in rats. Endocrinology, 103: 554-558. Ruberg, N., Rotsztejn, W. E., Aranchibia, S., Bes- son, J. and Enjalbert, A. (1978) Stimulation of prolactin release by vasoactive intestinal peptide (VIP). Eur. J. Pharmacol., 51: 319-320. Samson, W. K., Lumpkin, M. D., McDonald, J. K. and McCann, S. M. (1985) Prolactin-releasing activ- ity of porcine intestinal peptide (PHI-27). Peptides, 4: 817-819. Werner, S., Hulting, A. L., Hékfelt, T., Eneroth, P., Tatemoto,K., Mutt, V., Moroder, L. and Wunsch, E. (1983) Effect of the peptide PHI-27 on prolactin release in vitro. Neuroendocrinol., 37: 476-478. #1 oui n "fay HVFicge~ i Pati A >: ene, yn: \ Pivar*s Ab \ CGM Fe =~ j & i, SPUERE ae aly wilt oy oo 7 - } - ( oT. — " _ | 4 jak Nar» oe eT ie — 7 : - i= To EO - , f ¥ : Pua yee é ies eos _ — 4 te * ZOOLOGICAL SCIENCE 5: 415-430 (1988) Morphology and Distribution of the Skin Glands in Xenopus laevis and Their Response to Experimental Stimulations Kerko FuskurA, SHINGO KuraBucHt!, MASAHIRO TABUCHI and SAKAE INOUE Department of Comparative Endocrinology, Institute of Endocrinology, Gunma University, Maebashi 371, and ‘Department of Anatomy, Nippon Dental University, Tokyo 102, Japan ABSTRACT—Four types of skin glands were observed in the skin of adult Xenopus laevis. Among them, the mucous and the granulated glands showed similar morphological characteristics with those previously reported. In this report, we made detailed description on the saccular gland (referred to as NP gland in the present study) of the male nuptial pad. The small granulated gland which has not been reported so far is also described. The granulated gland and the NP gland were observed to receive innervation inside the secretory compartments. The NP glands were only distributed in the nuptial pad of the male forelimb while the other three kinds of glands were widely distributed throughout the body surface. When the frogs were kept under waterless conditions, the mucous glands secreted a watery substance, the cross sectional area of the skin occupied by the mucous glands being reduced to 46-71% of the control. While, the granulated gland and the small granulated gland were less affected by the water deprivation. Evacuation of contents of the skin glands was recognized only in the mucous glands. After hypophysectomy, a significant decrease in the area of the NP glands occurred accompanied by the disappearance of the nuptial pad, while the mucous and the granulated glands were unchanged. A possible enhanced secretory activity of the small granulated glands was encountered after hypophysectomy. Isoproterenol induced the discharge of viscid white material from the contracted granulated gland both in vivo and in vitro, while other types of skin of glands seemed to © 1988 Zoological Society of Japan be unchanged by this chemical. INTRODUCTION The skin glands of amphibians have been clas- sified into major types, viz., the granulated glands and the mucous glands, based mainly on light microscopic observations [1, 2]. The granulated glands are involved in protecting the body through their poisonous or irritating secretions in response to violent stimulation [3], while the mucous glands produce a watery secretion containing Na* and Cl~ that keeps the skin moist and controls its permeability [4]. Among the skin glands of Xenopus laevis, the granulated gland has been intensely investigated mainly due to interest in the biochemistry of its secretion. The granulated gland of X. laevis has been reported to contain Accepted September 12, 1987 Received May 1, 1987 thyrotrophin releasing hormone, 5-hydroxy- tryptamine [5] and caerulein [6, 7]. Dockray and Hopkins [6] also reported that the dermal layer of the dorsal skin of X. /aevis contains several kinds of glandular structures, but there was no descrip- tion of gland other than the granulated and mucous glands. During the course of the investigation of the nuptial pad of male X. /aevis [8], two other types of skin glands could be differentiated. Berk [9] briefly described one of them as a simple saccular gland in the nuptial pad, and the other has not been possibly identified in any previous report cited. In the present paper we provide a classifica- tion of different types of skin glands of adult X. laevis after studying of fourteen regions all over the body surface of both sexes accompanied with a study of changes in these glands after exposing frogs to a waterless condition, hypophysectomy 416 K. FuyikURA, S. KURABUCHI ef al. and adrenergic stimulation. The possible rela- tionship of these glands to the internal and environmental modifications is also discussed. MATERIALS AND METHODS Animals The African clawed frogs X. laevis which were successively kept in our laboratory were used. Light and electron microscopy Excised skin pieces from frogs, after decapita- tion and pithing or under 0.5% ethyl m-amino- benzoate methanesulfonate anesthesia, were fixed with Bouin’s fluid, processed routinely for light microscopy and stained with Mallory’s trichrome. To study on the distribution of skin glands through the body surface, pieces of skin of about 2 cm square were excised from fourteen selected areas from both sexes as shown in Figure 3. To study the ultrastructure, skin pieces were fixed in 3% glutar- aldehyde followed by postfixation with 1% osmium tetroxide and embedded in an Epon- Araldite mixture. The prepared ultra-thin sections were observed with Hiracu! HS-8 and LEM 2000 electron microscopes. Response of the skin glands to various experimental conditions Dry milieu Adult frogs of various body weights were housed in separate containers with- out water (75-80% humidity and 21+0.5°C). Body weight was recorded every 24 hr exactly. At later stages of the experiment, pieces of skin in the central regions of ventral and dorsal sides were dissected out and processed for histological observation. Hypophysectomy Adult males with well- developed nuptial pads were hypophysectomized. Frogs then were observed for macroscopic changes in their nuptial pads. Representative frogs were sacrificed on 12 and 40 days and at 15 months after the operation for histological observation. All quantitative measurements of the area, length and number of glands and of skin were carried out with an image analyzing apparatus (VIP, 121CH, Olympus Co., Japan). Sympathomimetic stimulation The response of skin glands to isoproterenol was tested in vivo and in vitro as follows. Experiment 1; For the in vivo study, 2 ml of isoproterenol (ISP) (Sigma) at a concentration of 2 10~* M/100 g body weight was subcutaneously injected into the adult frogs. Experiment 2; An in vivo study concerning the interruption of vascularization and innervation of skin pieces was done as follows. Pieces of dorsal skin about 1.5xX1.0cm in size were first cut completely free from their surroundings, and were then stitched into the same position as before. Seven days after the operations, the frogs were subcutaneously injected with ISP and their re- sponses were observed. Experiment 3; For in vitro study explants about 2cm square were taken from the middle of the dorsal or ventral sides and were immersed in Ringer solution or TCM 199 solution for up to 14 days. Response of the freshly excised skin to ISP was first tested employing the solutions containing ISP in concentrations of 10~' to 10-°M. The pieces of skin incubated were transferred every other day to the solution containing 10~'M ISP to examine their reactivity. Noradrenaline (Sigma) was also used in vivo and in vitro study to test the reaction of the skin glands. RESULTS Morphological characteristics of the classified skin glands Through the microanatomical observations of the skin from several regions of the body surface, the following four types of skin glands were identified. The granulated gland The granulated gland was composed of a secretory compartment, a duct and an intermediated region connecting them. This type of gland was largest of the four types of glands observed, ranging 150 4m up to 200 um in height and 200 to 400 «m in diameter. Above all the dorsal side of a large-sized female observed had a group of the largest granulated glands, measuring from 350 to 450 ym in height and from 300 to 550 am in diameter. The secretory compart- Xenopus Skin Glands 417 ment consists of a syncytial sac surrounded by a myoepithelial cell layer (Fig. la). In a histological view of a gland which might be a median profile of a secretory compartment, there were observed twelve peripherally arranged nuclei profiles with their long axes parallel to the myoepithelial layer. The electron microscopic (EM) observations re- vealed that the secretory granules which occupy the greater part of the syncytial cell were elipsoidal in shape, measuring 2.7X0.8m in average. These secretory granules were membrane- bounded and varied in electron density. Their contents were heterogeneous showing electron dense stripes randomly distributed and running parallel to the long axes of the granules (Fig. 7a, b, c). In the cytoplasm, well-developed rough ER and Golgi apparatus were observed surrounding the nucleus. In the area of the Golgi apparatus small dense granules were prominent, some of them appeared to be larger and membrane- bounded containing homogeneous substance, which might be a sign of the formation of secretory granules. In the intermediate region between the secre- tory compartment and the excretory duct, two different types of small cells were observed. The cells located at the inner side, i.e. near the compartment, have a central large nucleus and distinct microvilli projecting into the ductal lumen with no secretory granules. On the other hand, the outer side cells were larger than the inner ones, interdigitated with each other and continued to the ductal cells. The excretory ducts of the glands were opened throughout its length onto the surface of the skin, and some secretory materials were occasionally found to remain in the ductal lumen. The duct was a bilayered structure whose cells are columnar, the outer cells being a little slenderer than the inner ones. The myoepithelial cells which completely enwrap the secretory compartment are cuboidal in cross section and their width and height were about 7-9 um and 4-6 yum, respectively. The individual myoepithelial cells in longitudinal sec- tions were found to face each other with indented cellular surface on which desmosomes were fre- quently seen. Thick myofilaments were distributed equally through the cytoplasm and the smooth ER were well developed in the peripheral cytoplasm facing the secretory cells (Fig. 7a) or adjacent to the myoepithelial cells. Numerous pinocytotic vesicles were found along the plasma membrane of the myoepithelial cells, especially facing the secre- tory cell cytoplasm. Irregular finger-like projec- tions from the secretory cells came close to the myoepithelial cells and desmosomes were found in these areas, especially in contracted cells (Fig. 7b). Extreme contractions of the myoepithelial layers of the granulated gland were always present in the biopsied specimens whose secretory products were expelled via the excretory ducts. Both myelinated and non-myelinated nerve bundles, capillaries and melanocytes were found in the subcutaneous connective tissue around the gland. In the vicinity of the granulated gland, nerve bundles were divided into small nerve branches and the non- myelinated nerve bundles were seen close to the glands. It was observed that these non-myelinated fibers run throught the intercellular spaces be- tween the adjacent myoepithelial cells and are distributed in the intercellular spaces between the myoepithelial cells and the secretory cells with a number of nerve ending structures on the surfaces of the myoepithelial cells or of the secretory cells (Fig. 2a). Occasionally these nerve endings were found to rest on the deeply concave surfaces of the secretory cells. No partial thickening like pre- or postsynaptic membrane could be found in these areas. Individual nerve endings were seen to contain a large number of small vesicles of the size of synaptic vesicles (about 50-60 nm), a few cored vesicles (about 100-150 nm) and mitochondria. The serially arranged pinocytotic vesicles along the surfaces of myoepithelial cells were frequently opened toward the intercellular spaces at places close to nerve ending structures. The mucous gland The mucous gland was composed of a secretory compartment, a duct and an intermediate region between them (Fig. 1b). The glands of this type in the male under normal conditions measured approximately 125 ~m in height and 184 um in width. In cross sectional profile the secretory cells were found to be arranged in a monolayered or a multilayered pattern along the periphery of the secretory compartment, and in horizontal section they appeared somewhat pentagonal in shape. Most of 418 K. Fuyikura, S. KURABUCHI et al. Xenopus Skin Glands 419 them were stained light blue with Mallory’s triple stain and some were stained dense blue. With the EM, the structural cells of a compartment could be classified into two types, one appearing as a low density cell possessing granules of low electron density and the other a high density cell with electron-dense granules (Fig. 1b). The latter type had a strongly positive periodic acid-Schiff (PAS) reaction. The first type of secretory cells was found to be the prevailing type while the second type was less represented, forming only about one quarter of the total number of secretory cells in a compartment. With the EM, the clear granules were occasionally seen to be vacuolated square, pentagonal or round in shape measuring 2.4 ~m along the longer axis and 1.7 4m along the shorter axis in average (Fig. 2b). Dark granules were not uniform, were of varied deformed eggplant shapes and contained irregular filamentous deposits that run roughly parallel to the long axis of the granules. They occasionally fused with each other. Numerous dark spots were also present in the granules. The nuclei of cells were situated basally and surrounded by a moderate number of rough ER. Many interdigitations and occasional desmo- somes were found in the intercellular space be- tween the neighbouring secretory cells. Toward the basal region of the excretory duct, the cell type in the secretory compartment changed abruptly. In this area the cells adjacent to the secretory cells were slender and elongated containing a moderate number of mitochondria with no secretory gran- ules. This cell type was adjoined to a cell aggregation forming the base of the excretory duct (Fig. 1b). In this cell aggregation, the inner side cells which were adjoined to the ductal cells were small and columnar in shape. They possessed large nuclei and were provided with numerous microvil- li. The outer side cells were elongated and almost the same size as the inner side cells. Well- developed interdigitations were observed between the inner and the outer side cells, occasionally with demosomes. These cells were adjoined to the bilayered ductal cells of which the inner side cells were of small and columnar shape. The outer side cells were slightly larger than the inner side cells. Cornified ductal cells were continuation of the epidermal cornified layer. The myoepithelial cells surrounding the mucous gland were flat and elongated and not so thick as those of the granulated gland (Figs. 1b and 2a), measuring 0.6 to 0.8 um in thickness which is about one tenth of that of granulated gland. The intercellular space between the secretory cells and the myoepithelial cells was of a simple appearance as compared with that of the granulated gland. Both myelinated and the non-myelinated nerve bundles were seen in the dermal connective tissue around the mucous glands (Fig. 2b). However, they did not pass through the intercellular space between the adja- cent myoepithelial cells. The small granulated gland The small granulated gland was also composed of a secretory compartment, a duct and an intermediate region between them (Fig. Ic). This gland has not been described in the literature cited so far. This type of gland was located close to the epidermal layer of the skin and was the smallest among the four types of skin glands observed in the present study. It averages 50 um in height and 60 «m in width. The nuclei, rough ER and Golgi apparatus were confined to the base with a large amount of secretory granules occupying most of the cytoplas- mic area. The secretory granules were of a uniform appearance and they became yellow with Mallory’s stain. With the EM, the granules were Fic. 1. Electron micrographs of four types of skin glands. a: A granulated gland showing the syncytial appearance of a secretory compartment and thick myoepithelial cell layer. On the right hand side of the figure, a thin myoepithelial cell layer of a mucous gland can be seen. SCG, secretory compartment of granulated gland; ML, muscle cell layer of granulated gland; MLM, muscle cell layer of a mucous gland. b: Mucous glands showing low density secretory cell (LSC) and high density secretory cell (HSC) intermediate region (IR) and excretory duct (D). E, epidermis; ML, myoepithelial cell layer. c: A small granulated gland showing secretory cell (SC), intermediate region with mitochondria-rich cells (MRC) and transition region (TR). D, secretory duct; ML, myoepithelial cell layer; E. epidermis. d: A NP gland showing secretory cells (SC), secretory duct (D) and intermediate region (IR). ML, myoepithelial cell layer; MNB, myelinated nerve bundle. 420 K. FusikurA, S. KURABUCHI et al. found to be electron-dense and spheroidal in shape with an average size of 1.93.2 um, including the largest one that measured around 6 yam in diameter (Fig. 1c). This is the largest among the secretory granules of the four kinds of skin glands observed. In the intermediate region adjacent to the secre- tory cells, mitochondria-rich cells were observed adjoining the transition region of the excretory duct (Fig. 2c). Other than numerous mitochon- dria, these cells were characterized by their micro- villi, large round nuclei and the cytoplasm free of secretory granules. The arrangement of cells at the transition region toward the excretory duct and that at the excretory duct were the same as those of the mucous gland. The myoepithelial cells discon- tinuously surrounded the secretory compartment. The myoepithelial cells of the small granulated gland were wider than those of the mucous gland, but narrower than those of the granulated gland. The myelinated and non-myelinated nerve bundles were seen to have access to the gland, and the nervous structures which seemed endings or near the endings containing number of small vesicles and some cored vesicles, were occasionally observed in the area close to the basement lamina of the myoepithelial cells (Fig. 2c). However, no innervation was found in the intercellular space between the myoepithelial cells and the secretory cells. NP gland NP glands were found only in the nuptial pad. The average size of the gland was 160 ~m in width and 99 wm in height in a frog weighing about 43g. This gland consists of a secretory compartment, a duct and an intermedi- a: A nerve ending structure in the intercellular space between myoepithelial cell and secretory cell of a granulated gland. N, nerve ending. b: A profile of a mucous gland showing a secretory cell with clear vesicles, and a non-myelinated nerve bundle and a myelinated nerve with access to the secretory compartment. MN, myelinated nerve, NN, non- myelinated nerve. c: A small granulated gland showing a mitochondria-rich cell in a secretory compartment and non-myelinated nerves which face a thin myoepithelial cell layer (ML). MRC, mitochondria-rich cell. d: A NP gland showing secretory cells with large secretory granules (SG), a myoepithelial cell in longitudinal section (ML), and a nerve ending (N) surrounded by secretory cells. Xenopus Skin Glands 421 ate region between them (Fig. 1d). The secretory compartment was composed of monolayered co- lumnar secretory cells which were arranged in the bottom of the compartment. The average size of secretory cells was 38 ~m in height and 7 «m in width. The secretory cells lodged with a large quantity of secretory granules and their nuclei were situated at the base. These granules were stained deep blue by Mallory’s triple stain, and yielded a strong positive reaction to PAS. With the EM, profiles of well developed layered rough ER and Golgi apparatus were seen near the nucleus, and the apical part of the cells was provided with numbers of microvilli protruding into the lumen of the compartment. The adjacent Male NP gland 73 60 30 cells were interdigitated with each other with occasional desmosomes. In a profile of a secretory cell, there were observed 50 to 60 membrane- bounded granules which were homogeneous and mostly spheroidal in shape. In the areas of Golgi apparatus there were occasionally observed small granules of varied electron density, measuring 1.81.2 wm in average, which seemed to be in a process of maturation. Some of these small granules were in shapes suggesting that secretory granules might also become larger by the fusion of small granules together. The glands were sur- rounded discontinuously with myoepithelial cells which were moderately developed as in the case of the small granulated gland. In the regions where 5 oa Small B 20 A 2 Tw = 40) ® 20 e oO oO 60 30 IV vVI VII Z 2 3 2 9 Tw c o o a Ventral side Fic. 3. Percent distribution of skin glands in selected regions of body surfaces. Dorsal side I-VII indicate areas selected for observations as shown in the right hand corner. 422 K. Fustkura, S. KURABUCHI et al. the secretory cells were not covered with myoepithelial cells, the basement lamina of the secretory cells directly faced the connective tissue surrounding the glands. In the connective tissue surrounding the NP glands, there were observed intense vascularizations. The terminal structures of the non-myelinated nerves were observed to rest on the surfaces of the secretory cells or the myoepithelial cells after passing through the inter- cellular space between the adjacent myoepithelial cells (Fig. 2d). Also the terminal structures were observed to make direct contact with the secretory cells at the place where the secretory cells had no myoepithelial cell covering. Distribution of skin glands through the body surfaces The distribution of four types of skin glands in selected areas of the body surface of both male and female frogs is shown in Figure 3. Distribution is shown as the number of each skin gland as a percentage of the total number of skin glands counted in the area selected for study. In the male, the percentages of mucous glands, granulated glands, small granulated glands and NP glands were 43, 22, 19 and 16% on the ventral side and 56, 32 and 12% on the dorsal side, respective- ly, while in the female they were 46, 35 and 19% on the ventral side and 45, 42 and 13% on the dorsal side, respectively. The percentage distribu- tion of NP glands was 73% at V and 38% at IV in the male. In the nuptial pad region occurrence of Large-sized Median-sized mucous glands and the granulated glands was infrequent. No NP glands were found in other regions. Changes in skin glands under the experimental conditions 1) Dry milieu Figure 4 shows the decrease in body weight after keeping in a waterless condition. The percent reduction in body weight varied among the ex- perimental frogs and the small-sized frogs were apt to show a larger reduction (32-34%) than the larger ones (21-34%). The skin pieces excised from the experimented frogs as well as the controls were subjected to histological studies to measure the heights of the skin and epidermis, sectional areas of the mucous glands, granulated glands and small granulated glands. In the control frogs, all these measurements in the females were larger than those in the males (Table 1). It is worth noting here that the area of the granulated glands on the dorsal side was about twice as great as that of those on the ventral side. When the frogs were kept in a waterless milieu for 3 days, all the measurements in the skin and the skin gland became smaller than those of the controls. The thickness of the skin became 73-83% of that of the controls, the epidermis 63-82%, and the areas of the mucous gland 29-54%, the granulated gland 63-80% and the small granulated gland 74-98%. Small-sized 10 9 (60-110g) (30-50g) (4-7q) a= D xR ~ 80 40) 2 S 7 oD c o & 60 i 30 2 ® @ = = fos x) = 40 Fis 20 8 ° g 7 3 ea = Oo 20 2 o 110 = o } o (0) : 0) {oe 12 OA ES 6 7 8 9 10 11 12 13 14 Frog number Fic. 4. Daily changes in body weight after keeping the animals under waterless conditions. Dark bars indicate reduced body weight as a percentage reduction as determined by final weighing. ( ) indicates body weight of frogs. "(9891-1 YUM ‘TQ'O>d) [OMUOS dy} YWM poseduOd se jURDYyIUZIS , ‘poarasqgo Joquinyy :(_ ) “AS Furs 4 (b2) (LS) (LS) (12) (€2) (61) (79) (fz) (€Z) (Sz) QTL FSIST «S88 PFIISLL «LEOFSIP'S «7FSS 49FOTS BLIFPBPT +6797 FSES' PH «S8SFSOE6 «I FOE 4S FLEE SSa[1oqe AA (02) (8S) (9s) (€2) (rZ) (€2) (09) (9S) (bZ) (v2) 677 F608'T = EOB'9FETH'ETL COVIFLEC OI 1F99 L¥SP9 66FOL'T CZTEFSES'SS OL8FHOELI ISS PPPs [BWION $ apis [esiog Se (61) (6t) (6S) (iz) (bz) (€Z) (IS) (ZL) (pz) (S2) Z 6EEF PILE «PEO'EFTIOOD 6LEFLEPO 47 F06 §=49FELD OZTFOOTT PLE TFZS9'OI §=«7Z7FOLO'V «=x TFSE 49 F687 SSoPIOIB AA c 2 (bZ) (6€) (19) (7) (v2) (ZZ) (Zs) (ci) (b2) (SZ) = 6LSFICTE ISTP FIROL = OOS‘ TF IPL'7Z = TFOIL «LF 89S «(907 F EPL'T €S6FS7I'ST CEH FOHGH'OI §=LF9S yh FULE [BWION = opis yeUudA > 3 Ege pojepnueiy snoonjy ane ULyS Eee a povejnuein snoony] Madar ul{s zu ul w/ ul zur Ul uw ul spurs ulys Jo seaiy urlys JO 1Y4sIOH spurs ulys Jo seoly uTys JO 14319 ayewlay ae SUONIPUOD ssafio}eM JopuN pue ([eWIOU) 19}eM Ul Jday sidayj sndouay Jo spue[s ulys pue UTYs Jo JUSWIOINsSeo| “| ATAV], 424 K. Fusikura, S. KURABUCHI et al. The present study revealed that the area of the mucous gland showed the most remarkable reduc- tion, especially in those on the ventral side (down to 37% in the male and 29% in the female). The sectional areas of their glands were reduced, and the evacuation of contents of the glands was also recognized histologically. In addition, low density cells in the compartment of the atrophied mucous gland seemed to be reduced in size more than high density cells. The granulated glands seemed to be unchanged. 2) Hypophysectomy A significant decrease in the area of the NP gland occurred after hypophysectomy and the size of gland became about half that of the unoperated control frog, while the mucous, the granulated and the small granulated glands were unchanged (Table 2, Fig. Sa, b). On day 12 after hypophysectomy, the NP gland cell already showed a decrease in diameter, area and height (Table 3). Such regres- sed features were persisted in the specimens kept for 15 months after hypophysectomy (Fig. 5c). The area of the secretory cell nuclei was un- changed. The amount of secretory granules was also decreased on day 12 and became much less on day 40 without recovery later. In association with regression of the NP gland, the numerous small epidermal spikes which specifically cover the nup- tial pad area were rapidly decreased in number followed by complete disappearance in the 40 day specimens. There were often seen profiles of the small granulated gland which seemed to show active secretion through their distended excretory ducts which contained a considerable amount of secretory material. Such a histological view was unusual in other experimental series as well as in the controls. 3) Isoproterenol (ISP) application Experiment 1: Within a minute after the injec- tion of ISP, the recipient frogs started to react to ISP and in about 5 min the entire body surface was covered with milky and viscid secretory material expelled from the granulated glands. Histological- ly, evacuation of the contents of the glands occurred only in the granulated glands in which the myoepithelial layer showed remarkable undulation accompanied by hypertrophy in some places. The mucous gland and the small granulated gland were TABLE 2. Effect of hypophysectomy on the areas of skin glands 40 days after the operation Area of gland in um? (Mean+S.E.) Skin glands Control Hypophysectomy Mucous 11,002 + 1,211 (71) 10,833 + 1,195 (61) Granulated 15 ,686 + 2,089 (25) 13,352+1,559 (26) Small granulated ( ), number of glands observed. * P<0.05 with t-test. 2,794 +441 (22) NP 9,779 + 1,766 (9) 2,975 +233 (21) 4,867 +412 (6)* TABLE 3. Morphological alterations in NP glands after hypophysectomy P 8g pop Nimnbes Diameter Cross sectional areas Height Areat Days f (4m, Mean+S.E.) (um2, Mean+S.E.) of of after land gland _ nucleus operation en qd long short Gland Lumen in compartment cell (a) (a)-(b) (vm) — (wm?) 0 23 99+2 160+6 12,549+881 1,877+216 10,671+718 30-55 25-34 12 days 28 8443** 13845 9,194 + 567* 2,715 +359 6,478 +409** 20-35 23-29 40 23 56+2** 130+7** 4,847+338* 1,478 +284 3,369+290** 10-20 23-27 15S months 23 69+2** 116+5** 5,783+362** 3,4074+355** 2,376+117** 5-20 26-32 * P<0.05, ** P<0.01, with t-test. + shows only ranges of measurements. Xenopus Skin Glands 425 Fic. 5. control (a), and operated animals in 40 days (b) Histological views of nuptial pads showing a and 15 months (c) after hypophysectomy. 53 (a), X145 (b), 470 (c). ES, epidermal spike; NP, NP gland; M, mucous gland; G, granulated gland; SG, small granulated gland. unchanged (Fig. 6b). Experiment 2: Upon injection of ISP, the operated skin areas on the backs of the frogs did not respond to a drug injected although the surrounding intact skin responded well in a way similar to that of normal frogs (Fig. 8). The operated skin area showed gradual response to injected ISP with time, following operation days. The central area of the operated skin pieces started to react weakly to the injected ISP by post- operative day 16. Approximately a half area of the operated skin piece was observed to respond to ISP on day 28. Fic. 6. Histological views of skin glands of a control skin (a) and of a skin affected 30 min after ISP injection (b). 110 (a), 110 (b). G, granu- lated gland; M, mucous gland; SG, small granu- lated gland. Experiment 3: The skin pieces taken from the body surfaces showed results similar to that in the Experiment 2 when they were immersed in the physiological saline or medium containing ISP in concentrations ranging from 10~' to 10~° M. With the EM, 30 min after immersion in ISP solution (10-'M), the intercellular space between the secretory cells and the myoepithelial cells was seen to be wider, and a large number of vascular structures appeared at the periphery of the secre- tory cells (Fig. 7d). Secretory granules were sparsely distributed (Fig. 7c). The fibrillar sarco- plasm of the myoepithelial cells had a condensed appearance. However, in the sarcoplasm facing the basal lamina of the secretory compartment, a number of small protrusions appeared in the specimens fixed 5 sec after ISP immersion (Fig. 7b). The higher the concentration of ISP applied the more rigorous was the response of the glands to it. The lower the concentration of ISP, the more time needed for the glands to respond and the weaker was their response. At a concentration of 10-°M the reaction was obscure. These skin 426 K. FustkuraA, S. KURABUCHI et al. ted aa sar Fic. 7. Electron micrographs of granulated glands of a control (a), and 5 sec (b) and 30 min (c and d) after immersion in ISP solution. Arrow shows small sarcoplasmic protrusions. vascular formations. pieces were able to respond to ISP even after they were kept for as long as 14 days (Fig. 9). However, in this case the response occurred sporadically and not through the entire skin surface. In vivo and in vitro application of noradrenaline at a concentration of 10~° M was able to induce a discharge of secretion. DISCUSSION Classification of skin glands The granulated and mucous glands have been reported to be the major types of the amphibian skin glands [1]. Mills and Prum [10] classified the mucous glands of Rana pipiens, R. temporaria and R. catesbeiana as mucous and seromucous glands on the basis of electron microscopic observations. According to this report the mucous gland has mitochondria-rich cells in the junctional area between the duct and the acinus while the seromu- cous gland has cells without granule in the same ML, myoepithelial cell layer; VC, region. Fox [11] introduced kinds of amphibian skin glands as poisonous, lumpy, callous, mucous, and so on. In Salamandrina terdigitata, three types of skin glands, that is the mucous, serous and the mixed type, have been described [12]. These indicate that the skin glands of amphibians vary in form, and the terminology for these glands has not been well established. In the case of the skin glands of X. laevis, most of the studies have been concentrated on the granulated gland due to the interesting biochemical nature of its secretion and a little has been elucidated about other type of skin glands, although it was reported that the dorsal skin of X. laevis contained several kinds of grandular structures opening via the epidermal ducts to the external surface [6]. In the present study, we were able to demonstrate four different types of skin glands in the body surface of adult X. laevis. These were the mucous, the granulated, the small granulated and the NP glands. It appeared that the small granulated gland and the NP gland as termed in the present report were different Xenopus Skin Glands 427 Response of skin granulated glands about 5 min after the injection of ISP. No response is seen in the operated skin (Experiment 2) at the center of Fic. 8. the back, 7 days post-operation. 0.55. Fic. 9. Response of the in vitro granulated gland kept in medium 199 (a) and in physiological saline (b) for 12 days followed by immersing them in the media containing 10-'M ISP for Smin. 2. types from the seromucous or mixed type which was described previously in other species, as well as from the mucous and the granulated glands of X. laevis in their morphology, and responses of glands to various types of stimulation will be mentioned later. Distribution of skin glands Except NP gland the other three types of skin glands are distributed throughout the body sur- face. The mucous gland and the granulated gland were found to be greater in number and larger in size. This is in agreement with a report by Sjoberg and Flock [13] on R. temporaria and R. esculenta. However, the NP gland was found to be distrib- uted only in the male nuptial pads. According to Berk [9], in the nuptial pad area there were only glands of a simple saccular type (NP gland) and no other glands were found in this area. However, in the present study it was found that the granulated and the mucous glands existed in the nuptial pad area although a few in number. Also quite a number of small granulated glands were observed in this area. Stimulation of glands The response of the mucous and the granulated glands to adrenergic agonists observed might be similar to that of the previous reports [4, 14, 15]. However, to what kind of stimulation the small granulated gland responds could not be clarified by the present experiment. We only found that some of the small granulated glands expelled much of their contents into the excretory ducts when the frogs were hypophysectomized. Many investiga- tors have reported that some kinds of glands were present underneath the papillae in thumb pads of male frogs and newts [16-18]. However it has been still not certain whether skin glands in the area of thumb pads are the same or different from those of the common skin glands. As has already been reported, the development of the pads was dependent on the androgenic hormonal cycle [16- 18], and the development of the nuptial pad was accompanied by the specially differentiated NP glands which occurred only in this area. It is still not certain, however, what kinds of substances are produced in NP gland and how this gland dis- charges its secretion upon what kind of stimula- tion. In our preliminary experiment using a pair of frogs in courtship, the nuptial pad of the male and the pieces of skin of the female biopsied during amplexus showed any noticeable histological changes. After hypophysectomy only this gland showed remarkable morphological regression of the secretory cells and this continued many days. Therefore the functional significance of NP gland is different from the mucous gland. The skin glands underneath the thumb pads in certain species have been classified as mucous glands [16- 18]. Since the structural integrity of the NP glands is absolutely associated with the development of nuptial pad, this gland must have something to do with sexual behavior of the male frog, but this 428 K. Fuyikura, S. KuraBucui et al. remains unresolved in the present experiment. The mucous glands lost their contents enor- mously when the frogs were kept in waterless conditions. The contents of other kinds of glands were unchanged, although their size became small due to possible loss of water in such abnormal conditions. The skin textures gradually evolved to adapt exterior circumstances [11] and in amphib- ians, ions and organic substances in water pas- sively pass through the skin [19]. The watery substance of the mucous glands may be secreted first to keep well being of an individual under the present severe experimental circumstances. In normal physiological condition it was reported that an active ion-transport rather than secretion of cellular product occurred in the secretory cells of the mucous gland, and this mechanism was not elucidated [20, 21]. Numbers of biologically active peptides and biogenic amines, similar to those of mammalians, and alkaloids have been found in amphibian skins [3, 25]. High concentrations of thyrotrophic releasing hormone (TRH) and _ 5-hydroxy- tryptamine (S-HT) in the granulated glands of X. laevis and R. pipiens have been demonstrated [22]. Dockray and Hopkins [6] and others [23, 24] demonstrated a large amount of caerulein in the granulated gland of X. laevis. These substances (TRH, 5-HT and caerulein) were proved to be discharged following adrenergic stimulation [6, 25]. In the present experiment we employed isoproterenol (ISP), one of the sympathomimetic substances, to observe the response of the granu- lated gland in normal and experimentally treated skin. Benson and Hadley [14] reported that adrenaline and noradrenaline stimulated secretion of the contents of the granulated gland but ISP was ineffective when R. pipiens and X. laevis were subjected to in vitro treatment with various sym- pathomimetic agents in concentrations of 107‘, 10~° and 10~° M. Holmes et al. [15] also observed the effectiveness of adrenaline and related sub- stances in stimulating glandular secretion of X. laevis skin and they ranked their stimulative effectiveness as follows; adrenaline was most effective followed by noradrenaline and phen- ylephrine, while isoprenaline and salbutamol of adrenergic agonist were ineffective. However, we were able to demonstrate that applications of ISP to X. laevis skin in vivo or in vitro caused positive responses. The secretion of a milky substance on their surfaces and changes in structures of the granulated glands occurred. Such an apparent difference between ISP effectiveness in the present series of experiments and others may not be due to the nature of the drug nor to a species specificity of the frogs employed but largely due to the concen- trations of ISP applied. In the present series of experiments we applied ISP in concentrations 107! to 10-°M, compared to 10-4M, 10-°M and 5xX10~°M in the experiments by Benson and Hadley [14] and 10°-4M and 10-°M in the experiment by Holmes et al. [15]. Our experiment 2, in which ISP was applied to the frogs which had the detached and sewed skin pieces in their backs, the skin pieces operated on showed exceedingly retarded response to ISP in the early post- operative period followed by gradual recovery. This may be largely due to the degree of recovery of vascularization after the operation in which the vascularization of the operated skin pieces in the early phases of healing might not be sufficient to carry the amount of ISP needed for stimulation of gland secretion. Innervations In the present series of experiments, we also surveyed innervations of the glands studied. All the skin glands have been reported to be sur- rounded by myoepithelial cells either completely [26] or incompletely [27]. The innervations in the small granulated gland and the NP gland in the male nuptial pad have not been studied before. In the granulated gland and the NP gland we were able to demonstrate nerve fibers in the secretory compartments, while in the mucous gland and the small granulated gland the nerve fibers only came to have access to the glands outside the secretory compartments. Dockray and Hopkins [6] showed that nerve endings were present in the secretory compartment of the granulated gland of X. laevis, and Sjéberg and Flock [13] also reported a similar finding in the granulated gland of R. temporaria and R. esculenta. In mammalian species, it was also reported that similar types of innervations existed in the nasal gland in Guinea pig [28] and Xenopus Skin Glands the lacrimal gland of sheep [29]. The present results also suggest that neural influence may be exerted in the skin glands differently, according to the individual gland, since some glands received innervation on the inside of the secretory compart- ments while other glands were not innervated inside the secretory compartments but were only accessed by the nerve ending structures outside the secretory compartment. The excretion of the contents of the glands may occur mainly due to contraction of the myoepithelial cells. Histological signs of contraction of the myoepithelial cells were evident through the reports by Holmes and Balls [26] and the results of the present series of experiments. However, the present study demon- strates that there were also nerve endings resting on the surfaces of the secretory cells and occa- sionally in a deep place far from the surface of the myoepithelial cells. Therefore it is important to study the receptor sites for the sympathomimetic substances not only through the surface of the myoepithelial cells but also of the secretory cells with thorough studies on alternations of nerve endings and the structures of the secretory cell cytoplasm, to elucidate the cellular mechanism by which the contents of the granulated gland are expelled upon nervous stimulation. REFERENCES 1 Noble,G. A. and Noble, E.R. (1944) On the histology of frog skin glands. Trans. Am. Microsc. Soc., 63: 254-263. 2 Joseph, W. and Vanable,J.R. (1964) Granular gland development during Xenopus laevis meta- morphosis. Dev. Biol., 10: 331-357. 3 Myers, C. W. and Daly, J. W. (1983) Dart-poison frogs. Sci. Am., 248: 96-105. 4 Campantico, E., Guardabassi, A. and Torasso, L. (1978) Histological changes in Xenopus laevis Daudin adult specimens kept under dry conditions, then moved back to their natural aquatic environ- ment. I. Skin, kidney and interrenal tissue. Arch. Sci. Biol., 62: 63-76. 5 Bennett, G. W., Marsden, C. A., Clothier, R. M., Waters, A. D. and Balls, M. (1982) Co-existence of thyrotrophin releasing hormone and_ 5-hydro- xytryptamine in the skin of Xenopus laevis. Comp. Biochem. Physiol., 72C: 257-261. 6 Dockray, G. J. and Hopkins, C. R. (1975) Caeru- lein secretion by dermal glands in Xenopus laevis. J. 10 11 12 14 15 16 17 18 19 20 429 Cell Biol., 64: 724-733. Wakabayashi, T., Kato,H. and _ Tachibana, S. (1985) Complete nucleotide sequence of mRNA for caerulein precursor from Xenopus skin: the MRNA contains an unusual repetitive structure. Nucleic Acids Res., 13: 1817-1828. Kurabuchi, S. and Inoue, S. (1981) Small spiny projections in the epidermis of the mature Xenopus laevis. Annot. Zool. Japon., 54: 182-190. Berk, L. (1939) Studies in the reproduction of Xenopus laevis. Il. The secondary sex characters of the male Xenopus: The pads. S. Afr. J. Med. Sci., 4: 47-60. Mills, J. W. and Prum, B. E. (1984) Morphology of the exocrine glands of the frog skin. Am. J. Anat., 171: 91-106. Fox, H. (1986) The skin of Amphibia. In “Biology of the Integument. Vol. 2. Vertebrates”. Ed. by J. Bereiter-Hahn, A.G.Matoltsy and K. Sylvia Richards, Springer-Verlag, Berlin, pp. 116-135. Delfino, G., Brizzi, R. and Calloni, C. (1982) De- velopment of cutaneous glands in Salamandrina terdigitata (Lacépéde, 1788) (Amphibia: Urodela); findings by light and electron microscopy. Z. Mi- krosk. Anat. Forsch., Leipzig, 96: 948-971. Sjoberg, E. and Flock, A. (1976) Innervation of skin glands in the frog. Cell Tissue Res., 172: 81-91. Benson, B. J. and Hadley, M. E. (1969) Jn vitro characterization of adrenergic receptors controlling skin gland secretion in two anurans Rana pipiens and Xenopus laevis. Comp. Biochem. Physiol., 30: 857— 864. Holmes, C.H., Moondi,P.S., Rao,R.R. and Balls, M. (1977) Jn vitro studies on the effects on granular gland secretion in Xenopus laevis skin of stimulation and blockade of a and f£ adrenoreceptor of myoepithelial cells. Cell Biol. Int. Rep., 1: 263- 270. Yoneyama, H. and Iwasawa, H. (1985) Annual changes in the testis and accessory sex organs of the bullfrog Rana catesbeiana. Zool. Sci., 2: 229-231. Iwasawa,H. and Asai,O. (1959) Histological observations on the seasonal change of testis and the thumb pad in the frog, Rana nigromaculata. J. Fac. Sci., Niigata Univ., Ser. H., 2: 213-219. Iwasawa, H. and Takasu, T. (1985) Study of thumb pad regions developed by the administration of testosterone in a young female of Rana nigromacula- ta with a supernumerary forelimb. Jpn. J. Herpetol., 11: 5-10. Lillywhite, H. B. (1971) Thermal modulation of cutaneous mucous discharge as a determinant of evaporative water loss in the frog, Rana catesbeiana. Z. Vergl. Physiologie, 73: 84-104. Thompson, I.G. and Mills, J.W. (1981) — Iso- proterenol-induced current changes in glands of frog 21 22 23 24 25 430 skin. Am. J. Physiol., 241: C250-C257. Thompson, I. G. and Mills, J. W. (1983) Chloride transport in glands of frog skin. Am. J. Physiol., 244: C221-C226. Bennett, G. W. Balls,M., Clothier, R. H., Mars- den, C. A., Robinson, G. and Wemyss-Holden, G. D. (1981) Location and release of TRH and 5-HT from amphibian skin. Cell Biol. Int. Rep., 5: 151- 158. Inselvini, M. (1975) First appearance of caerulein during the development of Xenopus laevis. Gen. Pharmacol., 6: 215-217. Seki, T., Kikuyama, S. and Yanaihara, N. (1985) Development of caerulein producing cells in Xeno- pus skin gland during metamorphosis. Zool. Sci., 2: 980. Mueller, G. P., Alpert, L., Reichlin,S. and Jack- son, I.M.D. (1980) Thyrotrophin-releasing hor- mone and serotonin secretion from frog skin are 26 27 28 29 K. Fusikura, S. KURABUCHI et al. stimulated by norepinephrine. Endocrinology, 106: 1-4. Holmes, C. and Balls, M. (1978) In vitro studies on the control of myoepithelial cell contraction in the granular glands of Xenopus laevis skin. Gen. Comp. Endocrinol., 36: 255-263. Neuwirth, M., Daly, J. W., Myers, C. W. and Tice, L. W. (1979) Morphology of the granular secretory glands in skin of poison-dart frogs (Dendrobatidae). Tissue and Cell, 11: 755-771. Yamamoto, T. (1968) On the fine structure of the terminal portion of nasal gland in guinea pig, with special references to the interrelationship between glandular cells and nerve endings. Arch. Histol. Jpn., 27: 311-325. Yamauchi, A. and Burnstock, G. (1967) Nerve- myoepithelium and nerve-glandular epithelium con- tacts in the lacrimal gland of the sheep. J. Cell Biol., 31: 917-919. ZOOLOGICAL SCIENCE 5: 431-435 (1988) © 1988 Zoological Society of Japan Circadian Rhythms in Locomotor Activity of the Hagfish, Eptatretus burgeri II. The Effect of Brain Ablation SADAKO OoKa-Soupa, HirosHt KABasAwa! and SEncHIRO KINOSHITA”’> Atomi Gakuen Junior College, 1-5-2 Otsuka, Bunkyo-ku, Tokyo 112, 'Keikyu Aburatsubo Marine Park Aquarium, 1082 Koajiro, Misaki, Miura-shi, Kanagawa 238-02, and *Misaki Marine Biological Station, University of Tokyo, 1024 Koajiro, Misaki, Miura-shi, Kanagawa 238-02, Japan ABSTRACT—The hagfish, Eptatretus burgeri, displays locomotor activity only during the first two thirds of the dark period under 12L: 12D (7: 00-19: 00 light, 19: 00-7: 00 dark), and shows a clear free-running rhythm under constant darkness. The altered activity in the animal, whose brain was surgically removed except for the medulla oblongata, assumed a peculiar pattern which can be described as follows: (1) The free-running rhythm in constant darkness disappeared. (2) Under 12L: 12D, motor activity in the dark period disappeared, and continuous activity was observed throughout the light period. (3) This continuous activity always appeared and remained throughout the light period in various light regimens and it seems to be a direct reaction to light. INTRODUCTION There are many reports concerning the localiza- tion of the circadian pacemaker. It has been suggested that it is in the optic lobes of the cockroach [1] and of the cricket [2], in the prothoracic gland of the moth [3] and in the eyes of a molluscan species [4]. In vertebrates, the suprachiasmatic nucleus of the rat [5], the pineal gland of the chick [6, 7], both the suprachiasmatic nuclesus and the pineal gland of the house sparrow [8] and the pineal gland of the lamprey [9] are candidate tissues in which circadian pacemakers may be located. It is remarkable that the pineal gland is supposed to play an important role in circadian control in lower vertebrates. The hagfish, one of the most primitive vertebrates, belongs to the same vertebrate group in which the lamprey is also included. However, the hagfish is supposed not to have a pineal gland [10]. In the present study, as the first step in deter- Accepted August 31, 1987 Received October 31, 1986 3 Present address: Saitama Medical School, 981 Kawakado, Moroyama-machi, Iruma-gun, Saitama 350-04, Japan. mining the localization of the circadian pacemaker in the animal, the effect of the brain ablation on motor activity patterns was investigated. Normally the animal shows a clear nocturnal rhythm under light-dark cycles and displays a free-running rhythm when placed in continuous darkness [11]. MATERIALS AND METHODS The hagfish, Eptatretus burgeri, were collected by use of a trap containing sardines as bait. For experiments both males and females were used, because it is not possible to distinguish males from females on the basis of outer appearance. Further- more, males and females are similar in their circadian rhythms. The methods and procedures for recording the motor activity of the animals were described in a previous paper [11]. The water temperature was kept at 15°C, and no food was given throughout the experiment. All surgical operations were done while the animals were lightly anesthetized with MS 222. The animal was fixed on a plastic stage, and the skin and the fibrous connective tissue covering the brain were cut longitudinally along the median axis. The brain was removed with a 432 S. Ooka-Soupa, H. KABASAWA AND S. KINOSHITA pair of scissors. The connective tissue was re- placed as it was originally, and the skin was sewn. Ten animals were subjected to a sham-operation. In them the skin and the fibrous connective tissue were cut but the brain was not disturbed. The animals were kept in a large aquarium under 12L:12D for two weeks prior to the recording of the behaviour in the experimental aquaria. RESULTS The intact animal clearly shows nocturnal swim- ming activity, which occurs only in the first two thirds of the dark period under 12L:12D (7:00- 19:00 light, 19:00-7:00 dark), and it displays a distinct free-running rhythm under constant dark- ness [11]. In the sham-operated animal activity rhythms were the same as in the intact animals both in the light-dark cycle and in constant darkness (Fig. 1). When the entire brain was removed, the animal immediately died. However, animals, in which the medulla oblongata was left intact, survived for at least two months. They were very active, swim- ming in a manner similar to the intact hagfish: swimming near the surface of the water and moving along the edges of the aquarium. The activity pattern in the brain-ablated animal was very different from that in the sham-operated one; under 12L:12D, the activity did not occur in the dark period, but appeared continuously throughout the light period, and under constant darkness, intermittent activity with no circadian rhythm was recorded. Figure 2 showed one of the nine records from brain-ablated animals. In Figure 3, various light-dark schedules were programmed for one brain-ablated animal. Simi- larly, activity was confined almost completely within light period under all the lighting conditions including 6L:6D and 77L. No transient activity was observed when the light-dark cycle was re- versed. DISCUSSION Ueck and Kobayashi [12] searched unsuccessful- ly for a pineal gland in the hagfish, Eptatretus burgeri. It is of interest to find the location of the 15 SUT ST ST TT eas SSS Oe 20 — MW Time in days DD 25 Te ee 30 ee 85 Sen es | 12 19 7 12 Time of day Fic. 1. Locomotor activity recorded for the sham- operated hagfish kept in 12L:12D and in constant darkness. The activity is indicated by the vertical marks on the time lines. The activity occurred only in the first two thirds of the dark period in 12L: 12D and displays a distinct free-running rhythm in con- stant darkness. These activity patterns are fairly the same ones as those in an intact hagfish. in hours circadian pacemaker in an animal which has no pineal. In the present study, we found that the circadian rhythm disappeared when the brain was removed, except for the medulla oblongata. This fact suggests that the circadian pacemaker may be in the brain. In these experiments locomotor activity was observed as the measurable result of the operations. Since the center controlling locomotor activity itself should exist in the brain and might be disturbed by our procedures then we cannot conclude definitively whether the circadian pace- Brain Ablation Effects on Hagfish Circadian Activity 43 Ww pce ree eee eadbelldadatlels ee Tee ee = 1 L_ 2 { ! 1s) SE eS 1 ue J oe ee ee ee ee ee ee ou uu y | (as OS ee Ce L ] ed OS ae LE aT Dee ee 11 = lL Bel nedl, a Te a = Tes rele! See lll. =e ee ol ” a SS ee ee ee > 1° Me a bg —i ___ os < ae ee —a_____ iS ees | i of ag tti“( CO © 15 =n ul tmiti ob ee | i of ao eet gee eRe Be i: aie ff - @ ee ee ee | = i _ it a eee! 2 eee = _ ff a | = | =— , | i@ _— 20 a on 8 ne | rt ee on a aut i | = ase | Ema —_f th i aL aia a tt = CE ee EE 4 a = a gan ii = a 25 = 4 Lif a i a a so & _ Ui i i BESS i mol __if 2) ey ES a ie 12 19 7 12 Time of day in hours Fic. 2. darkness. The underlines show the light period. Locomotor activity recorded for the operated hagfish kept in 12L: 12D and in constant In constant darkness, the free-running rhythm disappeared and intermittent activity were recorded throughout the period. In 12L: 12D, the animal showed activity throughout the light period but none in the dark. maker is located in the brain. Continuous swim- ming by animals lacking a fore- and mid-brain, however, argues that no essential motor control was impaired by the operation. The characteristic response to light changed after removal of the fore- and mid-brain in this animal even though optic function was lost. There- fore, the locomotor activity stimulated by light stimuli probably was in response to a photorecep- tor in the skin. Previous reports have established a photoreceptor in the skin of hagfish [13]. Light perceived through eyes has been supposed to control the nocturnal rhythm in intact hagfish because the nocturnal rhythm can not stay in dark period and free-runs after eye removal (unpub- lished data, Kabasawa and Ooka-Souda). Accordingly, the following innervation scheme can be postulated in control of the locomotor activity system in the hagfish. pc Geen — sctvity ae pacemanet center ight (in the brain) \ photorecepter (in the skin). >. Pinal cord ) muscle mass ACKNOWLEDGMENTS The authors wish to express their gratitude to Prof. A. Gorbman, University of Washington, for his help in preparing the manuscript. 434 S. Ooka-Soupa, H. KABASAWA AND S. KINOSHITA ERA MI VO Reha ee TOC Tee ee, ee Ce ee @ ~~ 15 $$ adhe = : ~~ II lds @ ae ae ee oo | ee | mo] TS a aN eg ii ae aoe a ae = = , es a es ||) ¢ ST 0 ee | ee Oe ooo —————— tS eee © a ee eee a ees a a oT ee eS —_ ——E—————————— i ee fom a 4 Nill eel Fic. 3. 30 mm: | Time of day in hours Locomotor activity recorded for the operated hagfish kept under the various light-dark programs indicated. The brain-ablated animal which was under such time schedules as 12L:12D, 77L, 21D, 12L: 12D, reversal of the 12L: 12D, 12L: 12D, 6L: 6D, 12L: 12D and reversal of the 12L: 12D successively alternatively behaved itself with light-active/dark-resting. REFERENCES Page, J. L. (1982) Transplantation of the cockroach circadian pacemaker. Science, 216: 73-75. Tomioka, K. and Chiba, Y. (1985) Circadian rhythm in the neurally isolated lamina-medulla- complex of the cricket, Gryllus bimaculatus. J. Insect Physiol., 32: 747-755. Mizoguchi, A. and Ishizaki, H. (1982) Prothoracic glands of the saturniid moth, Samia cynthiaracini, possess a circadian clock controlling gut purge timing. Proc. Natl. Acad. Sci. USA, 79: 2726-2730. Gene, B. D. and MecMahon, D. G. (1984) Cellular analysis of the Bulla ocular circadian pacemaker system III. Localization of the circadian pacemaker. J. Comp. Physiol. A, 155: 387-395. Inouye, S. T. and Kawamura, H. (1979) Persistence of circadian rhythmicity in a mammalian hypothala- mic “island” containing the suprachiasmatic nucleus. Proc. Natl. Acad. Sci. USA, 76: 5962-5966. Deguchi, T. (1979) A circadian oscillator in cultured cells of chicken pineal gland. Nature, 282: 94-96. Takahashi, J. S., Hamm, H. and Menaker, M. (1980) Circadian rhythms of melatonin release from individual superfused chicken pineal glands in vitro. Proc. Natl. Acad. Sci. USA, 77: 2319-2322. Takahashi, J.S. and Menaker, M. (1982) Role of the suprachiasmatic nuclei in the circadian system of the house sparrow, Passer domesticus. J. Neurosci., Brain Ablation Effects on Hagfish Circadian Activity 435 2: 815-828. Morita, Y. and Samejima, M. (1984) Control of diurnal circadian locomotor rhythm by direct photo- sensory pineal organ. In “Animal Behavior: Neurophysiological and Ethological Approaches”. Ed. by K. Aoki, S. Ishii and H. Morita, Japan Sci. Soc. Press, Tokyo/Springer-Verlag, Berlin, pp. 232- 241. Quentin, B. (1963) The central nervous system. In “The Biology of Myxine”. Ed. by A. Brodal and R. Fange, Scand. Univ. Books, Oslo, pp. 50-91. Ooka-Souda, S., Kabasawa, H. and Kinoshita, S. (1985) Circadian ryhthms in locomotor activity in the hagfish, Eptatretus burgeri, and the effect of reversal of light-dark cycle. Zool. Sci., 2: 749-754. Ueck, M. and Kobayashi, H. (1979) Neue Ergebnis- se zu Fragen der vergleichenden Epiphysenfor- schung. Verh. Anat. Ges, 73: 961-963. Newth, D.R. and Rose,M.D. (1955) On the reaction to light of Myxine glutinosa L. J. Exp. Biol., 32: 4-21. ae rin Lona aid, : i. ay perk ctf ie 7 wee ZOOLOGICAL SCIENCE 5: 437-442 (1988) © 1988 Zoological Society of Japan Circadian Rhythms in Locomotor Activity of the Hagfish, Eptatretus burgeri III. Hypothalamus: a Locus of the Circadian Pacemaker? SADAKO OoKA-SouDA and Hirosui KaBAsawa! Atomi Gakuen Junior College, 1-5-2 Otsuka, Bunkyo-ku, Tokyo 112, and 'Keikyu Aburatsubo Marine Park Aquarium, 1082 Koajiro, Miura-shi, Kanagawa 238-02, Japan ABSTRACT—By recording the locomotor activity rhythms of hagfish in which partial ablations in brain were made, this work was designed to determine the location of the circadian pacemaker. The characteristic rhythms were maintained in the absence of the optic tectum, but were lost in animal lacking the telencephalon or diencephalon. The telencephalon-diencephalon unit was more finely dissected for more precise localization. The rhythm still occurred without the upper or the lateral parts of this anatomical unit. Transecting the middle of the telencephalon had no effect on the rhythms displayed. These findings suggest that a crucial part (circadian pacemaker?) for the fish to be rhythmic may be in the ventromedial part of telencephalon-diencephalon, the hypothalamus, the frontal part of which extends under the telencephalon. INTRODUCTION The authors have demonstrated a clear circadian rhythm of the locomotor activity of the hagfish, Eptatretus burgeri, using an_ infra-red _ light- photocell system [1]. Attempts to determine the localization of the circadian pacemaker by surgical ablation have shown that the pacemaker may be in the brain anterior to the medulla oblongata [2]. In the present study, more precise examination was undertaken to identify more precisely the part of brain where the circadian pacemaker may be situated. MATERIALS AND METHODS All surgery was performed while the hagfish were lightly anesthetized by MS-222. Fixing the animal on a plastic stage, the skin and the fibrous connective tissue covering the brain were cut longitudinally along the median axis. Each part of the brain was removed with a pair of scissors. The connective tissue was replaced as it was originally Accepted October 1, 1987 Received February 14, 1987 and the skin was sewn together. The operated animals were kept in a large aquarium under 12L:12D for one week prior to recording the activity in the experimental aquaria. The method and procedure for recording the activity of the animal have been described in the previous papers [1, 2]. The existence of locomotive rhythm was detected by observation of the distribution of activity records. RESULTS Figure 1 shows diagrammatically the locality of the brain where the surgical cuts were made. The effect of each type of operation on the locomotor activity is summarized in Table 1. Animals without the optic tectum displayed a nocturnal rhythm under 12L:12D, and a free- running rhythm in constant darkness (Fig. 2). Without either the telencephalon and anterior part of hypothalamus, or the diencephalon, activ- ity was not rhythmic, but was irregularly intermit- tent under 12L:12D or under contant darkness (Figs. 3 and 4). When the dorsal part of the telencephalon- diencephalon was cut out horizontally to the S. OoKA-SOUDA AND H. Fic. TABLE 1. Surgical operations in brain and their effects KABASAWA 1. Diagrammatic representation of the dorsal (A), lateral (B) and ventral (C) structures of the brain in the hagfish. The hatched area on B corresponds to the removed region conducted in Fig.2. The hatched area on C corresponds to that in Fig. 6. The lines conducted in Figs. 7, 8, 9 and 10 are shown by a, b, c and d on A respectively. (Based on Kusunoki et al., 1981 [3]) on locomotor activity rhythm in the hagfish Specimens Free-running Nocturnal eeu any Aue tested, rhythm in rhythm in number constant darkness 12L:12D telencephalon + frontal 8 = = part of hypothalamus diencephalon 9 = = optic tectum 5} + tr dorsal part of (telencephalon + diencephalon) 1/4 2 + =f 1/2 3 + Ar 3/5 2 + + lateral halves of 5 E 44 (telencephalon + diencephalon) line-cut at a 1 A; nL b 2 ++ a c 2 = = d 1 = — See Fig. 1 for surgical operations. +; rhythm positive. —; rhythm negative. *+ the activity is not confined to the dark period. Hypothalamus and Circadian Activity in Hagfish 439 n ee ee) 1) VOUT TCC Cem) Ue zo} oo ll 5 — 11 sume, i itt 1 @ c Um Lm a — —i | 1) 1 os |S es | Lt tt a feces ih TT TT oT on | E 4Q aL [mime 1 a as = ee POS | os ee an | 12 19 7 12 Time of day in hours Fic. 2. Locomotor activity of the hagfish without the optic tectum kept under 12L:12D (7:00-19:00 light, 19:00-7:00 dark) and in constant darkness. The activity is indicated by the vertical marks on the time line. The operated hagfish shows activity in the dark period under 12L:12D, and exhibits, in constant darkness, the free running rhythm whose length is about 24 hr. 25. = 12 19 7 12 Ti me of day Fic. 3. Locomotor activity of the hagfish without the telencephalon (including the frontal part of hypothalamus) (see Fig. 1B). The operated hagfish shows intermittent activity both in 12L:12D and in constant darkness. This is suggestive of losing the rhythm without the tissues. The peripheral lighting conditions are the same as that explained in Fig. 2, and the following figures (Figs. 4-10) are also in the same. in hours pe ey 1 ft iii pee LI L 1 1 L py ft L 5 1 ll L = 1 ea 1 1 i o 1 1 > oe [eaiieaaa iid C) u uid 1 tit 4 Bio iit 1 4 J fit) 1 yt Lu ifbit 1 iu @ £ a eral eee) Qa 1 ut 1 1 Lit C) L peas a estes 1 E15 toe : rea 1 ! a 1 1 pe a Ll 14 it Hl et oe oe a wh) 20 Sasi ee eco el eee ee its nT | oe 12 19 7 12 Time of day in hours Fic. 4. Rhythm-negative activity without the diencephalon. depths of 1/4, 1/2 or 3/5, in all cases, the nocturnal rhythms appeared under 12L:12D and free run- ning rhythms appeared under constant darkness (Fig. 5). In absence of the lateral halves of the telen- cephalon-diencephalon, the rhythm appeared both under 12L:12D and under constant darkness. However, in the 12L:12D, the activity extended beyond the dark period and occurred even in the light period (Fig. 6). This phenomenon is similar to that in the eye ablation experiment (Kabasawa and Ooka-Souda, unpublished). After lineal cuts across the full depth of the brain both at the border between the bulbus olfactorius and the telencephalon, and at the middle of the telencephalon, there was motor activity in the dark period under 12L:12D, and the activity pattern showed a free-running rhythm under constant darkness (Figs. 7 and 8). The cuts both anterior to the posterior large habenula and at the border between the telencephalon and the diencephalon caused intermittent activity both under 12L: 12D and under constant darkness (Figs. 9 and 10). From the results described above, it appears that the hypothalamus plays an important part in 440 S. OoKA-SOUDA AND H. KABASAWA Time in days E 12 19 7 Time of day in hours Fic. 5. The original record is plotted twice. Rhythm-positive activity without the upper part (3/5) of telencephalon-diencephalon. 12 19 if 12 19 7 12 Time of day in hours Fic. 6. Rhythm-positive activity without the lateral halves of the telencephalon-diencephalon (see Fig. 1C). In this case, the activity shows a free-running rhythm not only in constant darkness but also in 12L:12D. circadian locomotor activity system, and it is possibly the locus for the circadian pacemaker function. As shown in Figure 1, the hypothalamus is located in the ventromedial part of the telen- cephalon-diencephalon and anteriorly it extended as far as the midline of the telencephalon. The results of the transectional experiments can be interpreted a3 meaning that the hypothalamus may include a circadian pacemaker. DISCUSSION There are many published reports concerning the localization of the circadian pacemaker in vertebrates. The suprachiasmatic nuclei of the rat [4], the pineal body of the chick [5], both the suprachiasmatic nuclei and the pineal body of the house sparrow [6, 7] and the pineal body of the lamprey [8] all have been proposed to be the sites Hypothalamus and Circadian Activity in Hagfish 44] eT VOT TTT tee Mm imine ee Te) TT ee) wee Terie Olver a > 10 —— miami 1 MO A o I it Ait me at | mM ao] ae Tenens V One OCT Te 1H Tite ele SIO Teeeneriee Velaro) i ALi Mad HAM IBUIEL I & STW Ce a ae an aa 15 tut WU MU tI © tun Tn Veeeel eee Te eet ee ee ae e Tei! TOT EO ETO TOE! VO TeT eee 0 a - uA TT SUT arerere wT) UB MASE Hi Mm) 20 uum oe MUN i es ee ce Tene enrT Meee 6). ee OTe ne ever ol eTOCs me tt Sele ele wereieeT | 1 eT Sete SORT viv) VOCan oT eT Um 1 | aS EEE 12 19 7 12 19 7 12 Time of day in hours Fic. 7. Rhythm-positive activity with cutting across the border between the bulbus olfactorius and the telencephalon (see Fig. 1A, line a). ” a Ty BS eS We ae ~ © me) c 5 —l Www ep [=] —_ NNN PR i a | MM iMti al Ba i "| 6 Lott BL cy) ! 1) ititis it at i i= —1 in tt Yt | 4 12 19 td 12 Time of day in hours Fic. 8. Rhythm-positive activity with cutting across the middle of telencephalon (see Fig. 1A, line b). of pacemaker function. It has been reported that the hagfish has no pineal body [9]. In hagfish in which the dorsal part (3/5) of diencephalon was removed there still was a circadian rhythm in locomotor activity. Thus even if there was a pineal-equivalent tissue in the dorsal thalamus it would not seem to have a pacemaker function. It is rare in lower vertebrates that the pacemaker is proposed to be situated in the hypothalamus. Continued efforts toward more precise localization of the hagfish circadian pace- maker in the hypothalamus using electrode lesion methods are in progress by the authors. 442 S. OoKA-SOUDA AND H. KABASAWA Time in days [ BS 12 19 7 12 Time of day in hours Fic.9. Rhythm-negative activity with cutting across the front of the posterior habenula (see Fig. 1A, line c). ® lit 4 4 1 11 ray > LL Ae I See | Se | ‘eo i) 1 i one © 1 : 1 4 SG L4 a4 re eee ee 5 tt if i ell al ie UG { £ TP wt it ta ra a ey Re OH 1s a oi LS! (a) oO £ — eel = 12 19 7 12 Time of day in hours Fic. 10. Rhythm-negative activity with cutting across the border between the telencephalon and the diencephalon (see Fig. 1A, line d). ACKNOWLEDGMENTS All these experiments were performed at Misaki Marine Biological Station, University of Tokyo. The authors wish to express their gratitude to Prof. A. Gorbman, University of Washington, for his help in preparing the manuscript. REFERENCES Ooka-Souda, S., Kabasawa,H. and Kinoshita, S. (1985) Circadian rhythms in locomotor activity in the hagfish, Eptatretus burgeri, and the effect of reversal of light-dark cycle. Zool. Sci., 2: 749-754. Ooka-Souda, S., Kabasawa,H. and Kinoshita, S. (1988) Circadian rhythms in locomotor activity in the hagfish, Eptatretus burgeri. Il. The effect of brain ablation. Zool. Sci., 5: 431-435. Kusunoki, T., Kadota, T. and Kishida, R. (1981) Chemoarchitectonics of the forebrain of the hagfish, Eptatretus burgeri. J. Hirnforsch., 22: 285-298. Kawamura, H. and Inouye,S. (1979) Circadian rhythm in a hypothalamic island containing the suprachiasmatic nuclei. In “Biological Rhythms and their Central Mechanism”. Ed. by M. Suda, O. Hayashi and H. Nakagawa, Elsevier/North-Holland Biomedical Press, Amsterdam, pp. 335-341. Deguchi, T. (1979) Circadian oscillator in cultured cells of chicken pineal gland. Nature, 282: 94-96. Zimmerman, N.H. and Menaker, M. (1979) The pineal gland; A pacemaker within the circadian system of house sparrow. Proc. Natl. Acad. Sci. U.S.A., 76: 999-1008. Takahashi, J. S. and Menaker, M. (1982) Role of the suprachiasmatic nuclei in the circadian system of the house sparrow, Passer domesticus. J. Neurosci., 2: 815-828. Morita, Y. and Samejima, M. (1984) Control of diurnal and circadian locomotor rhythm by direct photosensory pineal organ. In “Animal Behavior: Neurophysiological and Ethological Approaches”. Ed. by K. Aoki, S. Ishii and H. Morita, Japan Sci. Soc. Press, Tokyo/Springer-Verlag, Berlin, pp. 237- 241. Bone, Q. (1963) The central nervous system. In “The Biology of Myxine”. Ed. by A. Brodal and R. Fange, Scand. Univ. Books, Oslo, pp. 50-91. ZOOLOGICAL SCIENCE 5: 443-448 (1988) Feeding Responses of Pacific Snappers (genus Lutjanus) to the Yellow-bellied Sea Snake (Pelamis platurus) PAUL J. WELDON Department of Biology, Texas A & M University, College Station, TX 77843, U.S.A. ABSTRACT— Previous studies indicate that Pacific fishes refuse to attack the yellow-bellied sea snake (Pelamis platurus). Visual and non-visual cues are thought to be used to identify this snake. Three species of Pacific snappers — Lutjanus aratus, L. argentiventris, and L. guttatus —- were tested for reactions to pieces or extracts of P. platurus. Snappers presented with carcass pieces of a variety of stimulus animals regurgitated P. platurus more than others; one experiment was inconclusive. Snappers presented in darkness with pieces of stimulus animals attached to clips removed fewer P. platurus pieces than those of other species. Pieces of P. platurus skin were regurgitated more frequently, and removed from clips less frequently, than were pieces of skinned carcass or control animals. Large pieces (5.0 g) of P. platurus were regurgitated and rejected more often than were small pieces (1.3g). Snappers regurgitated and rejected fish pieces treated with a chloroform: methanol extract of P. platurus more than pieces treated with solvent alone. This study indicates that © 1988 Zoological Society of Japan snappers detect chemicals from this snake. INTRODUCTION The yellow-bellied sea snake (Pelamis platurus) is found from the west coast of Central America to the east coast of Africa [1]. Although predation pressure in tropical Pacific waters is notoriously intense [2], no species has been observed to eat this highly venomous, pelagic serpent. A couple of presumed predators (rather than scavengers) have regurgitated P. platurus [3, 4], but stomach con- tent analyses of a variety of predatory fishes — 457 fishes representing 25 species from Panama [1], 186 dolphins (Stenella spp.) and 79 tuna (Thunnus albacores) from the eastern Pacific [5], approx- imately 1000 sharks (mostly Carcharhinus falcifor- mis) from Baja California to Costa Rica (S. Kato, pers. comm.), and thousands of sharks from around Australia [6, 7] — failed to indicate this snake’s remains. Heatwole [6] inspected 19 species of marine snakes in Australian waters for pre- dator-induced injuries and found that P. platurus was the only amply represented species that lacked any sign of attacks by fishes. Wounds in P. Accepted October 20, 1987 Received February 26, 1987 platurus from the eastern Pacific have been documented, but at least some of these appear to be man-made rather than caused by would-be predators [8]. Observations of the reactions of birds and predatory fishes to P. platurus indicate that this snake is avoided. Naive herons and egrets in Panama presented with several eels, a terrestrial snake, and P. platurus fled only from the latter [9]. A variety of teleost and elasmobranch fishes refused to attack P. platurus or regurgitated pieces of this snake if they were ingested [10]. The cues by which potential predators recognize sea snakes have not been studied systematically, but several authors suggest that the conspicuous yellow (ventral) and black (dorsal) coloration of P. platurus acts as an aposematic signal [1, 6, 10, 11]. Rubinoff and Kropach [10] state that other cues also may be involved since snake pieces were rejected by Pacific fishes even after the snakes’ color patterns had been modified with marking pens, when snakes were skinned, and when snake pieces were wrapped in squid flesh. Rubinoff and Kropach [10] suggest that fishes identify snakes by chemical cues. This study examines the feeding reactions of 444 P. J. WELDON Pacific snappers (Lutjanus spp.) to pieces of P. platurus and other animals, and tests the accepta- bility to the fishes of different sized P. platurus pieces and those from different body parts. In addition, the reactions of snappers to chemicals extracted from P. platurus are examined. The Lutjanus species used in this study are found primarily on the continental shelf over rocky or sandy substrates [12], where they could encoun- ter sea snakes. Adult Lutjanus spp. generally are non-specialized feeders [13, 14]; they therefore were deemed appropriate to test as potential predators of sea snakes. Since some of the snappers tested in this study fed regularly only in darkness (at night), the descriptions and results of diurnal and nocturnal feeding tests are given separately. DIURNAL TESTS Reactions to different species Methods: Six mullet snappers, Lutjanus aratus (total lengths = 28-42cm), were maintained together in an 18.0X6.6 m tank filled one meter deep with recirculating water. The mean water temperature in this and other tanks ranged from 27 to 29°C. These fish had been captive for at least one year, during which they were fed shrimp on an irregular schedule. They were fed shrimp each of two days before testing began. Pieces of P. platurus; two fishes, Tylosurus fodiator and Pomadasys panamensis; and a squid, Lolliguncula panamensis, were presented to the snappers. The snakes used in this and in other experiments ranged from 34 to 62cm (total lengths). Stimulus animals were sacrificed and kept frozen no more than two days before material from them was used, unless stated otherwise. The stimulus animals were thawed and pieces, including integument and muscle, were cut from them (1.2-2.3 g). Pieces of squid were taken from the mantle. All pieces were wrapped in aluminum foil, kept frozen, and thawed at least 20 min before being presented to the fish. Pieces of the stimulus animals were dropped one at a time to the snappers. If a piece was ingested and not regurgitated within one minute, the next stimulus animal was presented. Observations on fish that regurgitated and reingested a piece were continued for an additional three minutes, begin- ning the moment the food was reingested. Fish that accepted a piece of stimulus animal after another fish had regurgitated it also were observed for an additional three minutes. Rejection was scored if a piece was ingested, regurgitated, and not consumed by any fish within three minutes of the last regurgitation. A session refers to the period during which the reactions of one or more fish to a piece of stimulus animal were observed. The snappers were tested until a total of 30 pieces of each stimulus animal were consumed or regurgitated during tests over two consecutive days. Tests were conducted between 13:00 and 16:00 hr. Results: P. platurus pieces were regurgitated 23 times during eight sessions; no other pieces were regurgitated and none were rejected. The Fisher exact-probability test indicates that regur- gitations occurred during significantly more ses- sions with P. platurus pieces than with the other species (P=0.002). Methods: Five yellow tail snappers, L. argen- tiventris (total lengths = 22-38 cm), and four spot- ted rose snappers, L. guttatus (19-36 cm), were maintained in a 4.44.0 m tank filled one meter deep with recirculating sea water. These fishes had been captive for 20 days before testing and were fed pieces of squid on each of three days before testing began. The stimulus animals were P. platurus; two fishes, Oligoplites mundus and Tylosurus fodiator, and an octopus, Octopus vulgaris. Pieces of stimulus animals were prepared as described in the previous experiment. Fishes were tested for four consecutive days between 12:00 and 14:00 hr, with responses to ten pieces of each stimulus animal scored per day. Results: P. platurus pieces were regurgitated four times during three sessions; no other pieces were regurgitated and none were rejected. The Fisher exact-probability test fails to indicate that this trend is significant (P>0.10). Feeding Responses of Pacific Snappers 445 Responses to different sized Pelamis pieces Methods: previous experiment were kept together in a 4.4x4.0m tank. They were fed shrimp each of two days before testing. Small (1.3 g) and large (5.0 g) pieces of P. platurus were freshly cut and presented to the snappers. Ten pieces each of small and large snakes were presented to the fish each day for three consecutive days between 10:00 and 11:00 hr. During these tests, fish sometimes approached the snake pieces, touched them with their snout, and moved away without ingesting them. Five L. argentiventris tested in the Results: The small pieces of P. platurus were regurgitated during one session; none were rejected. The large pieces were regurgitated or refused after snout contact during ten sessions; five were re- jected. The Fisher exact-probability test indicates that large pieces were refused and rejected during significantly more sessions than were small pieces (P<0.005 for both measures). Acceptability of different Pelamis body parts Methods: Five L. argentiventris were pre- sented with pieces of the following parts of P. platurus: muscle (1.0-1.5 g), dorsal whole skin, ventral whole skin (0.9-1.6 g), and shed skin (0.7- 1.2g). The snakes from which the muscle and whole skin samples were obtained were decapi- tated, eviscerated, and their whole skin was cut away from the carcass. The dorsal (black) and ventral (yellow) skin were separated. Shed skins, pooled from several collected off the surface of the snakes’ aquarium, were wrapped in aluminum foil and kept frozen. Each piece of shed and whole skin was folded and tied to present a compact morsel to the fish. Pieces of squid mantle (2.5 g) served as controls. A total of 12 pieces of each item was presented to the fish, six in each of two consecutive days. Tests were run between 12:00 and 13:00 hr. Results: P. platurus shed skins were regurgi- tated 11 times during eight sessions; they were rejected in seven sessions. No other pieces were regurgitated. The Fisher exact-probability test indicates that shed skins were regurgitated and rejected during significantly more — sessions (P<0.002 for both measures). Methods: Four L. guttatus, maintained in a 4.44.0 m tank filled 2.2m, were tested as de- scribed above, except that the fish Vomer declivif- rons was used as the control animal. Six pieces of each item were presented, three in each of two consecutive days. Results: The total number of sessions during which P. platurus dorsal, ventral, and shed skins were regurgitated (and total regurgitations) were 6 (7), 4 (10), and 4 (6), respectively. Rejection of these materials occurred during 6, 4, and 4 sessions respectively. No regurgitation was observed with either fish or P. platurus skinned carcasses. The Fisher exact-probability test indicates that, despite the small sample size, snake skins were regurgi- tated and rejected during more sessions than were pieces of fish or skinned snakes (P<0.05 in all cases). Responses to Pelamis extracts Methods: Four L. guttatus (total lengths = 37- 40 cm) were captured and maintained in a 9.0 x 6.6 m tank for one month before regularly accepting food (pieces of various fishes). They were fed fish three days before testing. Fifteen snakes (total lengths = 34-36 cm) were cut into 5 cm pieces and placed into one liter of a chloroform: methanol (2:1 v/v) solution. The solution was heated (40°C) for several hours and filtered. The filtrate was placed over a water bath (60°C) on a rotary evaporator for 2 hr, after which an aqueous solution was poured off and saved. The remaining residue was dissolved in 150 ml of chloroform and kept 13 hr in an open glass beaker under ventilation to remove the solvent. A total of 0.31 g of residue remained after the chloroform had evaporated. This residue was redissolved in 25 ml of chloroform. Strips of the integument and muscle of a fish (Scomberomorus sierra) were cut into 4.5 g pieces. The experimental pieces were soaked for 10 hr in 446 P. J. WELDON the aqueous snake extract solution and were injected with 0.5 ml of the chloroform-soluble carcass extract (some of which flowed out on to the surface). The control pieces were soaked in a distilled water:methanol solution (1:1 v/v) for 10 hr and injected with 0.5 ml of chloroform. Both experimental and control fish pieces were wrapped in aluminum foil and allowed to air-dry for 4 hr. They were kept refrigerated (4°C) for several days during tests. Fish were presented with 5-7 items a day between 09:00 and 13:00 hr and were observed for one minute. If regurgitation and reingestion occurred, the session was extended for three minutes. A total of 17 pieces each of control and Pelamis-treated pieces were presented. Results: The control pieces were regurgitated twice during one session; none were rejected. The Pelamis-treated pieces were regurgitated a total of 31 times during eight sessions and rejected in six sessions. The Fisher exact-probability test detects significantly higher rates of regurgitation (P<0.01) and rejection (P<0.001) with Pelamis-treated pieces. NOCTURNAL TESTS Reactions to different species Methods: Seven L. guttatus (total lengths = 23-28 cm) were captured and maintained for one month in a 9.06.6 m tank filled to 1.1 m deep. These fish refused to eat pieces of fish offered to them during the day, but at night they accepted food attached to wooden clips suspended from a string immersed into their tank. The string (11 m) was tied at opposite ends along one side of the tank. A metal weight was attached to each of nine wooden clips, placed 60cm apart on the string. The food items attached to the clips were situated from 20 to 100 cm from the floor of the tank during the test. Pieces from three stimulus species were pre- sented: P. platurus, Octopus vulgaris, and the fish, Scomberomorus sierra. Each piece (3.5 g) in- cluded the integument and underlying muscle tissue. Three pieces of each stimulus animal were presented during each one hour session. The pieces were attached in a random order on the clips, the lights were extinguished, and the string was lowered into the water; fish, therefore, were not permitted visual access to the items. After one hour, the string was raised out of the water, the lights were turned on, and the pieces remaining on the string were noted. Four consecu- tive sessions were run each of two nights, giving 24 presentations of each stimulus animal. All noctur- nal tests were conducted between 20:00 and 02:00 hr. Results: A total of 20, 6, and 1 pieces of P. platurus, octopus, and fish, respectively, remained attached to the clips. The %? test indicates that significantly more snake pieces were left attached than were the other items (P<0.001). Reactions to marine and terrestrial snakes The same fish, facilities, and protocol described in the previous test were used. Fish were pre- sented with pieces (3.5 g) of P. platurus, western diamondback rattlesnakes (Crotalus atrox), and the fish, Scomberomorus sierra. The rattlesnakes were obtained in Nolan County, Texas. The carcasses of both Pelamis and Crotalus had been kept frozen (—70°C) for 5 months and shipped on dry ice. A total of 36 of each item was presented to snappers over three nights. Results: A total of 31, 9, and 0 peices of P. platurus, rattlesnake, and fish, respectively, re- mained attached to the clips. The X? test indicates that significantly more P. platurus pieces were left attached than were the other items (P<0.001). Acceptability of different Pelamis body parts The same fish, facilities, and protocol described in the previous two tests were used in this experiment. Fish were presented with pieces of P. platurus whole skin (1.4 g), skinned carcass (3.2 g), and a fish (Epinephalus sp.) (3.5 g). A total of 36 of each item was presented to snappers over three nights. Results: A total of 24, 15, and 0 pieces of Pelamis skin, Pelamis carcass, and fish, respective- Feeding Responses of Pacific Snappers 447 ly, remained attached to the clips. The X? test indicates significantly more pieces of skin were left attached than were pieces of snake carcass or fish (P<0.05). DISCUSSION The reactions of Lutjanus spp. to the stimulus animals presented in this study indicate that P. platurus is least acceptable as food. Bullseye puffers (Sphoeroides annulatus) also rejected P. platurus more frequently than pieces of fishes or squid [8], as did spotted cabrillas (Epinephelus analogus) (Weldon, preliminary observation). These results agree with initial observations by Rubinoff and Kropach [10] on the rejection of this snake by numerous Pacific fishes. Kropach [11] suggested that Lutjanus spp. attend to visual cues in avoiding P. platurus., L. aratus and L. argentiventris immediately and rapid- ly swim away from P. platurus introduced into their tanks, even when dead snakes coated with varnish were used (Weldon, preliminary observa- tion). It seems likely that visual stimuli do elicit avoidance by snappers, but other cues may also be involved. The reactions of fishes during the diurnal tests were scored if the items presented were contacted by the snout or taken into the oral cavity. Visual cues may have been discerned, but the decision to swallow or regurgitate these materials likely was based upon post-ingestive information of a chemi- cal or mechanical nature. The results of the nocturnal experiments indicate that snappers dis- criminated between pieces of P. platurus and other animals under conditions where no visual informa- tion was available. The increased frequencies with which larger P. platurus pieces were regurgitated and rejected in tests with different sized pieces probably were due to mechanical difficulties en- countered in swallowing. MacLeish [15] describes an “underwater taste test” where Australian fishes regurgitated the carcass of a sea snake (not P. platurus). It has been unclear, however, whether the refusal of snakes by fishes is elicited by chemicals. The greater frequencies of regurgitation and rejection of fish pieces treated with P. platurus extract, compared to those observed with solvent-treated controls, support the hypothesis that fish attend to chemicals to avoid ingesting this snake. The solubility of these substances suggests that they are lipoidal. The results of the nocturnal test comparing fishes’ reactions to P. platurus and the rattlesnake, Crotalus atrox, suggest that snappers discriminate between these species on the basis of non-visual cues. Tests of the reactions of fishes to extracts of various snakes are needed. While P. platurus appears to be relatively predator-free, other marine snakes occasionally are eaten by sharks [7, 16]. Whether snakes differ in palatability is unknown. Pacific fishes, including Lutjanus spp., have been observed to reject P. platurus carcasses bereft of skin [1, 10]. No pieces of skinned P. platurus were regurgitated by the snappers in the diurnal tests here. L. guttatus rejected ventral, dorsal, and shed skins of P. platurus, and L. argentiventris rejected pieces of shed skin, which probably were recognized as inedible. The results of the nocturnal test with L. guttatus also indicate that whole snake skin is less acceptable than the skinless snake carcass. It is unclear whether the decision to regurgitate or eat different parts of P. platurus is based upon chemical or mechanical cues since the materials used differ in texture and consistency. Tests of fishes’ reactions to extracts from different body parts are needed to determine whether unpalatable chemicals are localized in the body of P. platurus. ACKNOWLEDGMENT I. Rubinoff generously offered facilities and logistical support for this research. K. Haworth, O. Vallarino, and A. Velarde provided expert field and laboratory assist- ance. S. Kato provided unpublished information on shark stomach contents. H. Drummond, D. R. Robert- son, and I. Rubinoff made helpful comments on an earlier version of this manuscript, and R. Eaton typed it. This study was supported by grants from the Smithsonian Institution and the Whitehall Foundation. REFERENCES 1 Kropach, C. N. (1973) A field study of the sea snake 448 P. J. WELDON Pelamis platurus (Linnaeus) in the Gulf of Panama. Ph. D. Diss., The City University of New York. Vermeij, G. J. (1978) Biogeography and Adapta- tion: Patterns of Marine Life. Harvard University Press, Cambridge. Heatwole, H. and Finnie, P. E. (1980) Seal preda- tion on a sea snake. Herpetofauna, 11: 24. Pickwell, G. V., Bezy, R. L. and Fitch, J. E. (1983) Northern occurrences of the sea snake, Pelamis platurus, in the eastern Pacific, with a record of predation on the species. California Fish Game, 69: 172-177. Perrin, W.F., Warner, R.R., Fiscus,C.H. and Holts, D. B. (1973) Stomach contents of porpoise, Stenella spp., and yellowfin tuna, Thunnus alba- cores, in mixed-species aggregations. Fish. Bull., 71: 1077-1092. Heatwole, H. (1975) Predation on sea snakes. In “The Biology of Sea Snakes”. Ed. by W.A. Dunson, University Park Press, Baltimore, pp. 233- 249. Lyle, J. M. and Timms, G. J. (1987) Predation on aquatic snakes by sharks from northern Australia. Copeia, 1987: 802-803. Weldon, P. J. and Vallarino, O. (1988) Wounds on the yellow-bellied sea snake (Pelamis platurus) from 10 16 Panama : Evidence of would-be predators? Biotro- pica, 20: 298-301. Caldwell, G. S. and Rubinoff, R. W. (1983) Avoid- ance of venomous sea snakes by naive herons and egrets. Auk, 100: 195-198. Rubinoff, I. and Kropach, C. (1970) Differential reactions of Atlantic and Pacific predators to sea snakes. Nature, 228: 1288-1290. Kropach, C. (1975) The yellow-bellied sea snake, Pelamis, in the eastern Pacific. In “The Biology of Sea Snakes”. Ed. by W. A. Dunson, University Park Press, Baltimore, pp. 185-213. Thomson, D. A., Findley, L. T. and Kerstitch, A. M. (1979) Reef Fishes of the Sea of Cortez. John Wiley & Sons, New York. Hobson, E. S. (1968) Predatory behavior of some shore fishes in the Gulf of California. Bull. Sports Fish. Wildlife, Res. Rep., 73: 1-92. Walford, L. A. (1937) Marine Game Fishes of the Pacific Coast from Alaska to the Equator. Universi- ty of California Press, Berkeley. MacLeish, K. (1972) Diving with sea snakes. Natl. Geogr., 141: 565-578. Heatwole, H., Heatwole, E. and Johnson, C. R. (1974) Shark predation on sea snakes. Copeia, 1974: 780-781. ZOOLOGICAL SCIENCE 5: 449-461 (1988) Geographical Differentiation in Populations of Japanese Dace Tribolodon hakonensis Deduced from Allozymic Variation Naoto Hanzawa!, NoBuHIKO TANIGUCHI” and KeEn-IcHI NUMACHE Ocean Research Institute, University of Tokyo, Minamidai, Nakano-ku, Tokyo 164, *Laboratory of Aquatic Ecology, Faculty of Agriculuture, Kochi University, Nankoku, Kochi 783, and *Otsuchi Marine Research Center, Ocean Research Institute, University of Tokyo, Akahama, Otsuchi, Iwate 028-11, Japan ABSTRACT— Allozymic variation of Japanese dace Tribolodon hakonensis was examined at 20 loci in 17 populations derived from Hokkaido to Kyushu. Intrapopulational variability indicated by the proportion of loci polymorphic and heterozygosity was close to the standard levels in vertebrates. On the other hand, allelic frequencies at five loci of the twenty were remarkably different among populations, and even replacement of predominant alleles was observed. The extent of interpopula- tional differentiation was estimated by coefficient of gene differentiation (G,,) and genetic distance (D). Levels of these indices were remarkably high compared with the levels known among popula- tions in many vertebrates. Based on a dendrogram, we classified the populations into three local groups; 1) Northern Group, 2) Southern Group, and 3) Lake Biwa Group. These results suggest that the populations of different local groups have been highly differentiated from each other and the © 1988 Zoological Society of Japan differentiation is due to geographical isolation. INTRODUCTION The genus Tribolodon which belongs to the family Cyprinidae has a unique characteristic on distribution: it is widely distributed from the upper reaches of rivers to the coastal regions of the sea in Japan and adjacent countries. The other genera of Cyprinidae do not show such a wide distribution. Therefore fishes of Tribolodon have been exten- sively studied from the viewpoints of physiology [1] and ecology [2]. However, their taxonomy had been confused and greatly different among tax- onomists [3-5], because the species of this genus have not been morphologically differentiated enough from each other. Our previous study using allozyme markers clearly showed that the four Accepted September 30, 1987 Received March 16, 1987 ' Present address: Department of Cytogenetics, National Institute of Genetics, Mishima, Shizuoka 411, Japan. species of Tribolodon classified by Nakamura [4] had been genetically differentiated enough as highly as the interspecific levels in vertebrates [6, 7]. Thus, biochemical analysis provided good information as for taxonomy. In addition, we examined TJ. hakonensis from the waters in Fukushima Pref. and Kochi Pref. in the previous work, and found the possibility that this species contained the highly differentiated populations [7]. In this work, we further analysed T. hakonensis collected at more locations from Hokkaido to Kyushu, evaluated the of intraspecific variability and the extent of interpopulational differentiation, and discussed the evolutionary aspects in populations of T. hakonensis. level MATERIALS AND METHODS Specimens We collected a total of 883 specimens of T. 450 N. Hanzawa, N. TANIGUCHI AND K. NUMACHI hakonensis in 17 waters from Hokkaido to Kyushu. Sample locations are shown in Figure 1, and sample sizes are given in Table 3. Fish were caught at a single location within a few days by a gill net, a casting net, and angling, and were frozen by dry ice at once and stored under —20°C. In the Otsuchi, the Monobe, and the Shimanto Rivers, specimens were collected at a few locations, the upper reaches, and the lower reaches or the coastal regions of the sea, which were isolated by dums at present. Electrophoretic analyses Horizontal starch gel electrophoresis was per- formed upon drip from tissue samples as described by Numachi [8]. The following four buffer systems (Table 1) were used: CAPM (citric acid, 4-(3- aminopropyl)-morpholine), pH 6.0 described by Clayton and Tretiak [9]; CAEA (citric acid, ie oe OTSUCHI R' r40°N N-(3-aminopropyl)-diethanol-amine), pH 7.0 and CT (citric acid, tris), pH 8.0 described by Numachi et al. [10]; and TBE (tris, boric acid, EDTA), pH 8.7 described by Numachi [11]. The staining methods were according to Shaw and Prasad [12], and Taniguchi and Numachi [13]. The gels stained were washed and dried by the method of Numachi [14]. A list of proteins analysed, their abbrevia- tions, E. C. numbers, locus designations, tissues, and buffers employed are shown in Table 1. Tissues of skeletal muscle, liver, heart, and blood were used. Mitochondrial and supernatant iso- zymes of AAT, IDH, MDH, and ME were discriminated. Locus and allele designation followed Hanzawa et al. [15]. The most common alleles in the populations of the Northern Group were desig- nated as 100 or —100, and the other alleles were designated by the relative differences in elec- Pacific Fic. 1. The sites where specimens of T. hakonensis were sampled. Genic Differentiation in Japanese Dace 451 TABLE 1. electrophoretic buffers employed Enzymatic and non-enzymatic proteins examined, locus designations, tissues, and Protein (Abbreviation; E.C. Number) Locus Tissue Buffer Aspartate aminotransferase (AAT; 2.6.1.1) m-Aat muscle CT Alcohol dehydrogenase (ADH; 1.1.1.1) Adh liver CAEA a-Glycerophosphate dehydrogenase (a-GDH; 1.1.1.8) a-Gdh-2 liver CAEA Glucosephosphate isomerase (GPI; 5.3.1.9) Gpi-l heart TBE Gpi-2 heart TBE Hemoglobin (HB) Hb-1 blood CT Hb-2 blood CT Isocitrate dehydrogenase (IDH; 1.1.1.42) s-Icd-1 liver CAEA s-Icd-2 liver CAEA Lactate dehydrogenase (LDH; 1.1.1.27) Ldh-l heart CT Ldh-2 muscle CT Ldh-3 liver CT Malate dehydrogenase (MDH; 1.1.1:37) s-Mdh-1 muscle CAPM s-Mdh-2 liver CAPM m-Mdh muscle CAEA Malic enzyme (ME; 1.1.1.40) s-Me muscle CAEA Phosphoglucomutase (PGM; 2.7.5.1) Pgm muscle CAEA 6-Phosphogluconate dehydrogenase (6-PGD; 1.1.1.44) 6-Pgd liver CAEA Sarcoplasmic protein (SP) Sp-2 muscle CAEA Sp-3 muscle CAEA trophoretic mobility of respective gene products with the reference allele, 700 or —100. Statistical analyses Amount of intrapopulational variability was evaluated by the proportion of loci polymorphic and heterozygosity. Degree of interpopulational differentiation was evaluated by coefficient of gene differentiation (Gs) [16] and genetic distance (D) [17]. Gsy value, which is calculated based on average heterozygosity, indicates the degree of allelic differentiation over all populations. D value, which is calculated based on allelic frequen- cies, indicates the degree of differentiation be- tween pairs of populations. Divergence time between the populations was estimated based on D value and gene substitution rate of 107 [16]. RESULTS Intrapopulational variation Twenty loci controlling 12 kinds of enzymatic and nonenzymatic proteins were examined. The genetic control by 20 loci was mainly deduced from the phenotypes of different species and their hybrids [6, 7, 15]. Allozymic variation was found at 14 of the 20 loci in sample populations of 7. hakonensis. Particularly, various polymorphisms were Observed at 5 loci, Adh, a-Gdh-2, Gpi-2, s-Mdh-2, 6-Pgd, and each allelic control was interpreted (Fig. 2). Deviations of observed num- ber of phenotypes from their expectations under Hardy-Weinberg equilibrium were not significant in all sample populations except the upper reach of the Shimanto River (at 6-Pgd, P<0.05). Intrapopulational variability of 7. hakonensis indicated by proportion of loci polymorphic and heterozygosity was given in Table 2. The propor- tion of loci polymorphic ranged from 0.05 to 0.50, and the mean value indicated 0.22+0.10. Heter- ozygosity observed ranged from 0.019 to 0.085, and the mean value indicated 0.041+0.019. Levels of these two indices were similar among sample populations. Heterozygosity observed agrees with its expectation (H°/H' +1). N. Hanzawa, N. TANIGUCHI AND K. NUMACHI 452 2@L/<2t <@l/o0r O0L/00t 28I/2@L £8L/00L 0or/oor ‘surayjed Suipueg aatjsodso1 ay} MOjaq UMOYs aie sadkjousH ‘sisuauoyoy “1 JO GOd-9 pue HAWS ‘IdD ‘Hdd ‘HAV 10} sussned onasoydooaja Jo soindy sneUayos pue sydeisojoyg = *Z “Ol 6II/6II 6L2I/00r O0L/00L oot/ta 6II/00L OOL/Tg 00t/o0t =a 691/00 O00T/00I o00L/0 0/0 0/0 = 00L/00L — = | == | =z — c-HOW-s a — ae — O00T-/EST- ads — 00L-/00L- 00I-/00I- 2%-/00T-U0I-/¢SI- 001-/00I- 26-/00r- DOr Dore hae et 8II/00L O0L/t6 O0T/00L OOT/00t oot/6e2 00L/00T rao Ha9-0 — as a) ay a — iz OLT/9LL 9éL/6LE G6IT/6II 6IL/00L oot/oor9Zt/9el 9ct/ert SIT/6rr 00t/00r a= = = < aaa) ae = a = — fer] — : — fee == =a P=} am _—— fees 1d9 ay Genic Differentiation in Japanese Dace 453 TABLE 2. Estimates of genetic variability at 20 loci within populations of 7. hakonensis en Proportion of Heterozygosity O/ {JE ne a polymorphic (Observed; H°) (Expected; H*) se Abira R. 0.30 0.039 0.040 0.976 Otsuchi upper reach 0.20 0.019 0.020 0.950 Otsuchi Bay 0.30 0.028 0.028 1.000 Hirose R. 0.25 0.046 0.047 0.979 Kinu R. 0.50 0.069 0.075 0.922 Kano R. 0.30 0.072 0.081 0.889 Aga R. 0.25 0.035 0.037 0.952 Iburibashi R. 0.15 0.030 0.029 1.039 Yoshii R. 0.10 0.019 0.019 0.947 Sufu R. 0.20 0.023 0.021 1.098 Chikugo R. 0.15 0.033 0.033 1.014 L. Biwa 0.40 0.085 0.089 0.954 Kumano R. 0.25 0.072 0.074 0.970 None R. 0.15 0.034 0.030 1.125 Monobe R. upper reach 0.05 0.026 0.024 1.111 lower reach 0.15 0.029 0.024 1.212 Kure R. 0.10 0.028 0.031 0.917 Shimanto R. upper reach 0.25 0.050 0.059 0.848 middle reach 0.20 0.048 0.049 0.974 lower reach 0.20 0.042 0.042 1.000 Hitotsuse R. 0.20 0.034 0.034 1.003 (0.22 +0.10) Interpopulational differentiation Gene constitution of sample populations col- lected at 21 locations in 17 waters from Hokkaido to Kyushu were compared (Table 3). Allelic frequencies among subpopulations within the same waters were very similar, and a significant devia- tion among allelic frequencies was not observed at all loci except Adh and a-Gdh-2 in the Shimanto River population. These subpopulations were isolated by dums at present, and fish of the Otsuchi Bay was particularly the amphidromous type. However, gene constitution of these subpopula- tions within the same waters was thus very similar. On the other hand, remarkable differences among populations derived from the different waters were found at five loci, Adh,a-Gdh-2, (0.041 +0.019) Gpi-2, s-Mdh-2, 6-Pgd. At Adh and a-Gdh-2 of the five, predominant alleles of /00 in the popula- tions of the Pacific coast of the northern Japan and of the coast of the Sea of Japan were replaced with the other alleles (J/9 or 0) in those of the Pacific coast of the southern Japan. Similarly, predomi- nant allele of 700 at 6-Pgd was replaced with 127 only in the Lake Biwa and the Kumano River populations. Population specific alleles with high frequencies were found, such as 176 at Adh in the Lake Biwa population and —/53 at Gpi-2 in the Kumano River population. The extent of interpopulational differentiation was evaluated by two kinds of indices, coefficient of gene differentiation (Gy) and genetic distance (D). As for the Otsuchi, the Monobe, and the Shimanto River systems, specimens collected at N. Hanzawa, N. TANIGUCHI AND K. NUMACHI 454 0000 000T 0000 O000T 0000 0000 O000T 0000 0000 0001 0000 0000 0000 7870 8120 7100 $960 £200 0000 000T “WU asns}owH 0000 O000T 0000 O00T 0000 £100 8860 0000 £100 8860 0000 0000 0000 €170 8820 0000 000T 0000 0000 O000T yoeol TOMO] 000°0 000'T 0000 O000T 0000 Sz00 $260 0000 0000 0001 0000 0000 0000 Ose0 05990 0000 0001 0000 0000 000T yoeol o[ppru 0000 000'T 0000 O000T 0000 0000 000T 0000 £100 8860 0000 0000 0000 8trO £990 e910 880 0000 0000 000'T yoear roddn “Yo oyRUWIYS 000°0 =000'T 000°0 O000°T 000°0 0000 000'T 000°0 000°0 000°T 000°0 0000 000°0 9800 +160 0000 000'T 000°0 000°0 000'T “YW any 000°0 =000'T 000°0 O000°T 000°0 0000 000'T 000°0 000°0 000°T 000°0 000°0 0000 7100 8860 0000 0001 000°0 0000 000'T yoeol TOMO] 000°0 000°T 000°0 000°T 000°0 000°0 000°T 000°0 000°0 000°T 00090 000°0 000°0 000° O000°T 0000 O000T 0000 0000 000'T yoear saddn “Y aqouojpy 000°0 + 000'T 000°0 000°T 000°0 0000 000T 0000 0000 O000T 0000 0000 0000 SszI0 S280 0000 O160 0600 0000 000'T “YU suON 000°0 =000'T 000°0 O000'T 0000 0000 FrSS5°0 9bb0 0000 O000T 0000 0000 0000 Tez0 6920 0000 O00T 0000 0000 000T “Y oueurny rc0'0 §=9260 0000 O000T 0000 LITO STI80 6100 0000 O000T 0000 0000 0000 I710 6280 Flr0 9750 00070 0000 0001 eMlg “7 000°0 = =000'T 0000 O000'T 000°0 000°0 000°T 000°0 000°0 O000°T 000°0 000°0 0000 192'0 6€70 0000 60F0 1680 0000 O000T “Yy osnyIyD 000°0 =000'T 000°0 O000'T 000°O cIOO 8860 0000 000°0 000'T 000°0 0000 0000 L060 £600 0000 910 +980 000°0 000'T “a WYNS 000°0 =000'T = 000°0 000'T 000°0 000°0 000°T 000°0 000°0 O000°T 000°0 0000 0000 £62'0 L070 0000 800 7960 000'°0 000T “a HYysoX 6200 1960 0000 O000'T 0000 S50 9FF'0 0000 000°0 000'T 000°0 0000 0000 ¢€IZ'0 8870 000°0 0000 000'T 000°0 000°T “UW rysequngy 0000 =000'T 000°0 O00'T 000° 8870 TCIZ0 0000 000°0 $860 000°0 SI00 0000 $860 STt0'0 000°0 0000 0001 000°0 000T “y esy Sc0'0 $460 0000 000'T 000°0 $Z0;O0 $7260 000°0 0000 000'T 0000 000°0 000°0 000°'T 0000 0000 0060 009°0 ILT0O 67280 “a ouey 000°0 =000'T §=000°0 §=000'T =000'0) = Fr7:0 =FFL'0) «C100 =—000'0)—:0L6'0—:000'0--: OL0'0-—-:000'0--:0Z6'0.-—s-: 0800-0000. FLT'0—-978'0—s«OFE'0—:0L90 “YU nury 000°0 =000'T 000°0 000'T 000°0 0870 07220 0000 0000 8860 0000 7100 0000 O000T 000°0 0000 £900 860 tzO 8910 “YU SSO 000°0 =886°0 =e10'0 O000'T 000°0 SzI'0 $280 0000 000°0 860 £90°0 0000 ¢€10°0 8860 0000 000°0 000°0 000'T SL0;0 $7260 Aey 1yonsiO 000°0 = 000°T §=000°0 «000° =000°0)—-880°0— £160 =—000'0)=—-000'0—:0S6'0—-0S0'0-—-000'0--:000'0—-: 000 T=—-:000'0-—-:000'0-—-:000'0--: OOD T= £90°0-—- 8£6°0 yoear raddn TyoNsIO 000°0 = 000'T_ =000'0 =LL6'0 —€70'0—680'0—-«1F6'0—-:000'0--:000'0-—-: 000 T_—-:000'0-—-:000'0-—s EZN'0—- LL60 0000 =0000 rL70 97Z0 £100 8860 “UW BQy Sol oor S8 Oo! cy Le- OOI-_—s EST--_—s BIT 001 16 6L 691 Oo! 0 9LT oll oor c9- = 00I- uones0T c-YPI-s T-Ypy-s c-1dy Tidy cYPO” 4YPV wy-u sisuauoyny “J yo suoyeindod ut 10] py IO} satouanbaay oyaPy “¢ ATAVL 455 Genic Differentiation in Japanese Dace vs 0000 0000 O00T ZITO 8080 000T 0000 6000 1660 0000 0000 000T 0000 0000 000'T = 000°0 = O00'T — 000°0 UY 9snsjouH OP 0000 0000 O000T SzeO0 SZ90 000T 0000 0000 000'T 000°O — 000°O —-000'T —-000°0—-000°0, 0007 T0000, ODOT 0000 Yoeot JOMO| OP 0000 = Sz00 S60 000 O00LO 00O0T 0000 0000 000'T 0000 0000 000'T 0000 0000 = 000'T 0000 = 000°T —- 000°0 yoeol o[ppru Ov e100 ¢cl00 Sl60 6770 IIZ0 O000T 0000 0000 000T 0000 000°O 000T 000° = 0000 —000'T =—000°0 000" T—-000°0 yorai soddn “ye oyuRWYs Ov 000°0 =000°0-000'T = 000°0 000 T 000 T — 000O 8880) EIPO =000°O) —-000°0 000 T0000: 000°0, 0007 T0000, 000° T —000°0 “YU Any Ip 0000 0000 O000T ZIOO 8860 O000T 0000 LI0.05, d.f.=1. ** 7=3.08, P>0.05, d.f.=1. *** 471.98, P>0.05, d.f.=1. 492 ACKNOWLEDGMENT This study was supported by the subsidy from the Ministry of Education, Science and Culture, Japan as a part of the Project for Preservation of Medaka as a Gene Resource. Nw REFERENCES Shimada, A., Shima, A. and Egami,N. (1988) Zool. Sci., 5: (in press). Shima, A. and Shimada, A. (1988) Mutation Res., (in press). Tomita, H.(1982) Medaka, 1: 7-9. Aida, T. (1921) Genetics, 6: 554-573. Naruse, K., Ijiri, K., Shima, A. and Egami,N. 12 13 K. Naruse, A. SHIMADA AND A. SHIMA (1985) J. Exp. Zool., 236: 335-341. Hyodo-Taguchi, Y. (1980) Zool. Mag., 89: 283- 301. (In Japanese) Shima, A., Shimada, A., Komura,J., Isa, K., Naruse, K., Sakaizumi, M. and Egami, N. (1985) Medaka, 3: 1-4. Sakaizumi, M., Moriwaki, K. and Egami, N. (1983) Copeia, 1983: 311-318. Haldane, J. B.S. (1919) J. Genet., 8: 299-309. Mather, K. (1935) J. Genet., 30: 53-78. Thorgaard, G. H., Allendorf, F. W. and Knudsen, K. L. (1983) Genetics, 103: 771-783. Streisinger, G., Singer, F., Walker, C., Knauber, D. and Dower, N. (1986) Genetics, 112: 311-319. Cherfass, N. B. (1977) Genetika, 14: 599-604. ZOOLOGICAL SCIENCE 5: 493-495 (1988) [COMMUNICATION] © 1988 Zoological Society of Japan T Cell-specific XTLA-1 Antigens from Xenopus laevis Tadpole and Froglet are not Identical SABURO NAGATA Tokyo Metropolitan Institute of Gerontology, Sakaecho 35-2, Itabashiku, Tokyo 173, Japan ABSTRACT—T cell-specific XTLA-1 antigen mole- cules were immunoprecipitated from tadpoles and frog- lets of J strain Xenopus laevis, and analysed by gel electrophoresis. Both the tadpole and froglet XTLA-1 molecules have an apparent molecular size of 120 KD as analysed by SDS-PAGE. But, after deglycosilation with endo F glycosidase, the froglet XTLA-1 molecules show more extensive charge heterogeneity than the tadpole ones do on two-dimensional gels. The results suggest that the XTLA-1 molecules partially changes their structure during metamorphosis. INTRODUCTION During amphibian metamorphosis, profound biochemical transitions occur in association with the morphological and physiological changes. In erythrocytes, for example, there is a complete switch in hemoglobin from larval to adult type molecules. It has been demonstrated that such a switch in hemoglobin types results from a replace- ment in the blood of larval hemoglobin-producing erythrocytes by newly differentiating adult hemo- globin-producing cells [1]. Recently, evidences have been accumulated in Xenopus supporting that similar biochemical and functional transitions may occur in association with the shift of lymphoid cell types during metamorphosis (see review [2]). The XTLA-1 antigen recognized by a mouse monoclonal antibody is the only thymus- dependent (T) cell-specific surface antigen that has so far been identified in Xenopus, and provides a useful marker for studying the development, dif- Accepted October 13, 1987 Received August 20, 1987 ferentiation and function of the T cell system of this animal [3-5]. The XTLA-1 molecules im- munoprecipitated from froglet thymocytes and splenocytes are glycoproteins of an apparent molecular size of 120k dalton (KD) [5]. In the present study, the biochemical characterization was carried out on the XTLA-1 molecules im- munoprecipitated from tadpoles as well as froglets. The results indicate that the tadpole XTLA-1 molecules have the same apparent molecular size as the froglet ones but differ in a charge heteroge- neity, suggesting that the XTLA-1 molecules may partially change their structure during metamor- phosis. MATERIALS AND METHODS Major histocompatibility complex (MHC) homozygous, partially inbred J strain Xenopus laevis were used. Mature females were injected with 300U of human chorionic gonadotropin (Gonatropin 1000; Teikoku Zoki Co.) to induce ovulation, and the eggs obtained were artificially fertilized in Steinberg’s solution according to the method described previously [6]. Embryos and larvae were kept in aquaria and their developmen- tal stages were determined by the Normal Table of Nieuwkoop and Faber [7]. The mouse monoclonal antibody, XT-1, was produced by the previously described method [5] and the IgG fraction of the hybridoma ascites, obtained by the affinity chro- matography on a protein A-Sepharose CL-4B (Pharmacia) column, was used for immunopre- cipitation. 494 S. NAGATA Thymuses were dissected from tadpoles between stages 55-56 (35 days after the fertilization) and from froglets of 8 months in age, and the thymo- cyte suspensions were prepared in amphibian phosphate buffered saline. Cells from 50 tadpoles or 10 froglets were pooled and labeled with '251. Cell labeling, immunoprecipitation, digestion with endo F glycosidase (glycopeptidase F-free; Boehringer Mannheim Biochemica) and gel elec- trophoresis were carried out exactly as described previously [5]. RESULTS AND DISCUSSION The lysates of radioiodinated thymocytes from tadpoles and froglets were immunoprecipitated by sequential incubations with the specific mono- clonal XT-1 antibody and protein A-Sepharose CL-4B beads, and analysed by SDS-PAGE. The results showed that both tadpole and froglet XTLA-1 molecules run as a single band around an apparent molecular size of 120 KD (Fig. A), con- firming the previous results on the adult molecules [5]. Since the endo F glycosidase-treated XTLA-1 molecules from froglet thymocytes were known to segregate into the characteristic pattern of spots on O’Farrell’s two-dimensional gel electrophoresis, immunoprecipitates from tadpoles and froglets were digested with endo F glycosidase and then analysed by two-dimensional gel electrophoresis. As shown in the previous study [5], the deglycosi- lated froglet XTLA-—1 molecules formed a number of spots (some are fused) with a slightly reduced molecular size near the acidic end of the gel (Fig. B). In contrast, the tadpole XTLA-1 molecules migrated into a more restricted charge heter- ogeneity under the same conditions, 1.e., several relatively basic spots seen on the gel of the froglet antigen were not detected on that of the tadpole antigen. Endo F glycosidase was reported to re- move both high-mannose and complex type gly- cans linked through asparagine to the peptide backbone [8]. The difference in charge distribu- tion patterns between the tadpole and froglet XTLA-1 molecules observed here is, therefore, assumed to represent a difference in other linked glycans and/or amino acid compositions of the 120 KD glycopeptides. During the metamorphosis of anuran amphib- ians, a profound reorganization occurs in the larval immune system as the tadpole changes its physiol- ogy, morphology and behavior. This reorganiza- B — NEPHGE ACIDIC > BASIC TADPOLE ', > 31OdavL 1319044 8 ADVd-Sds <— Vv FROGLET Fic. 1. Analysis of tadpole and froglet XTLA-1 mole- cules by SDS-PAGE (A) and O’Farrell’s two- dimensional gel electrophoresis (B). Thymocytes from J strain tadpoles and froglets were radioiodi- nated, solubilized with lysis buffer containing 1% Nonidet P-40, and immunoprecipitated with XT-1 monoclonal antibody followed by protein A- Sepharose CL-4B beads. For SDS-PAGE, im- munoprecipitates were boiled in SDS sample buffer containing 5% 2-mercaptoethanol and_ elec- trophoresed on 10% polyacrylamide gels. For two- dimensional gel electrophoresis, immunoprecipi- tates were deglycosilated with endo F glycosidase, and separated by nonequilibrium pH gel elec- trophoresis (NEPHGE) on first dimension. Elec- trophoresed tube gels were immersed in SDS sam- ple buffer and then subjected to second dimension SDS-PAGE. Electrophoresed gels were visualized by autoradiography on X-ray films. Molecular size standards (150 KD, 94 KD, 67 KD, 43 KD, 30 KD and 20.1 KD) were run on the same gels and indi- cated with arrowheads in (A). Xenopus T Cell Antigen 495 tion is suggested to involve a nearly complete replacement of lymphopoietic cells and a reorgan- ization of the microenvironment where lympho- cytes differentiate [2]. Thus, lymphopoietic cells in the early-larval thymus are replaced by precursor cells derived from a distinct compartment of the embryonic dorsal lateral plate mesoderm [9], so that the reconstituted lymphopoietic system may supply the froglet with adult type lymphocytes having “mature” immunological functions. Re- cently, it was demonstrated that the MHC class I molecules are not expressed on the lymphocyte surface until the metamorphic climax [10]. Although their function is remained to be clarified, the difference in two-dimensional gel patterns of the XTLA-1 molecules found in the present study might provide, as the MHC class I antigens, a marker to distinguish adult type T cells from larval type ones in Xenopus. Such a marker should prove useful in studying the cellular basis of the develop- ment of immune reactivity and tolerance during amphibian metamorphosis. ACKNOWLEDGMENTS This work was in part supported by the Grant-in-Aid (No. 60440100) from the Japanese Ministry of Educa- tion, Science and Culture. REFERENCES 1 Dorn, A. R. and Broyles, R. H. (1982) Proc. Natl. Acad. Sci. USA., 79: 5592-5596. Flajnik, M. F., Hsu, H., Kaufman, J.F. and Du Pasquier, L. (1987) Immunol. Today, 8: 58-64. 3 Nagata, S. (1985) Eur. J. Immunol., 15: 837-841. Nagata, S. (1986) Dev. Biol., 114: 389-394. Nagata, S. (1988) Zool. Sci., 5: in press. Moriya, M. (1976) J. Fac. Sci. Hokkaido Univ. Ser. VI, 20: 272-276. 7 Nieuwkoop, P. D. and Faber, J. (1976) The Normal Table of Xenopus laevis (Daudin). North Holland Publ. Co., Amsterdam. 8 Elder, J.H. and Alexander, S. (1982) Proc. Natl. Acad. Sci. USA, 79: 4540-4544. 9 Maeno,M., Tochinai,S. and Katagiri, C. (1985) Dev. Biol., 110: 503-508. 10 Flajnik, M. F., Kaufman, J.F., Hsu, E., Manes, M., Parisot,R. and Du Pasquier, L. (1986) J. Immunol., 137: 3891-3899. bo nn ZOOLOGICAL SCIENCE 5: 497-499 (1988) [COMMUNICATION] © 1988 Zoological Society of Japan Interspecific Transplantation of Developing Tissues and Their Subsequent Differentiation in Flies PAKKIRISAMY SIVASUBRAMANIAN Department of Biology, University of New Brunswick, Fredericton, New Brunswick, E3B 6E1, Canada ABSTRACT—Imaginal leg discs from the larvae of housefly, Musca domestica were cultured in the body cavity of the pupae of fleshfly, Sarcophaga bullata. The implanted discs were not rejected by the foreign species. On the contrary, they differentiated fully and completed metamorphosis in time according to their own develop- mental program. However, the tanning and hardening of the cuticle occurred along with that of the host after eclosion of the host fly. These developmental events are discussed in relation to the hormonal milieu of the host species. INTRODUCTION During the course of their development flies go through distinct stages such as larva and pupa before metamorphosing into adults. Several adult structures like the legs, wings, eyes, etc., differ- entiate in the pupal stage from groups of embryonic cells called imaginal discs. Simul- taneously, the central nervous system (CNS) too undergoes considerable reorganization and estab- lishes specific neural connections with the newly formed peripheral target tissues. A great deal of information is available with regard to the forma- tion of specific nerve connections between the CNS and target tissues of the same species such as crickets [1], fleshflies [2] and fruitflies [3]. Howev- er, very little is known whether this specificity is restricted within a species or it extends beyond species boundaries. To explore this aspect of neuron-target interaction, developing tissues of Accepted September 16, 1987 Received August 4, 1987 the housefly, Musca domestica, were cultured into the pupal body cavity of the fleshfly, Sarcophaga bullata, and this report describes the metamor- phosis of such transplanted imaginal discs. The differentiation of transplanted CNS of the housefly has been reported elsewhere [4]. MATERIALS AND METHODS The housefly, Musca domestica, and the fleshfly, Sarcophaga bullata, were cultured in the labora- tory under constant conditions of temperature (25°C) and photoperiod (16L:8D). The housefly larvae were raised in an artificial diet containing milkpowder, wheat bran and sawdust, while the fleshfly maggots were fed with fresh beef liver. Since the housefly is smaller in size with a shorter pupal life they were used as donors, while the larger fleshfly with longer pupal period served as hosts. The imaginal leg discs of mature 3rd instar larvae of Musca domestica were dissected in insect saline and implanted into freshly formed fleshfly prepupae. The transplantation method was based on the technique of Bhaskaran and Sivasubramanian [5] but slightly modified as de- scribed in Sivasubramanian and Nassel [2]. Of the total of 76 successful implants, 36 were recovered 6 days after the operation from the host pupa (total pupal period of donors) and 40 were recovered 12 days post operation after eclosion of the meta- morphosed host flies. The tissues were examined as whole mounts. 498 In vivo Culture of Imaginal Leg Discs RESULTS AND DISCUSSION Housefly, the donor, is about half the size of the fleshfly, the host. The imaginal leg discs of the housefly being very small (0.55 mm long; 0.38 mm wide) compared to the volume of the host pupa (170 mm*) several discs could be cultured in the same host. Accordingly, 2—4 discs from housefly larvae were transplanted into the fleshfiy pupae. The time taken to complete metamorphosis is also correspondingly shorter for housefly, i.e., 6 days as compared to 12 days for fleshfly. Therefore, the implanted leg discs were recovered from hosts 6 days after operation. As seen in Figure 1 the discs had fully metamorphosed in 6 days with well tanned bristles albeit in an uneverted condition because of their development inside the body cavity. However, the cuticle was still untanned. Figure 2 shows the metamorphosed leg discs recovered from an eclosed host fly (12 days post-operation). The cuticle of these legs were fully tanned. Although insect imaginal discs are routinely cultured in vitro for understanding of hormonal control of growth [6], eversion [7], biochemistry of developmental events [8], etc., in vivo culture is the preferred method of developmental biologists looking into the aspects of determination [9], pattern formation [10], axonal projection [2, 11] etc. In the latter method, the discs from larval stages are transplanted into the metamorphosing stages (pupa) of the same species and examined at the completion of metamorphosis of the host. In this procedure, one of the limiting factors is the volume of the host which restricts the size and number of discs that can be cultured. This problem was solved as reported in this communica- tion by using a larger species as host and a smaller one as a donor. The housefly discs not only survive but also complete their differentiation within the body cavity of fleshfly pupa. Molting hormone ecdysone is an essential re- quirement for differentiation of imaginal discs, and, according to Wentworth et al. [12], there is an ecdysone peak in the host Sarcophaga bullata at the time of transplantation. Therefore, it is not surprising that the housefly discs complete meta- morphosis within fleshfly pupae. Nevertheless, it is 2 Fic. 1. Metamorphosed leg discs. Two prothoracic leg discs recovered 6 days after transplantation showing fully tanned bristles. 35. Fic. 2. Metamorphosed leg discs. Two prothoracic leg discs recovered from eclosed host flies 12 days after transplantation showing fully tanned cuticle. 35. interesting to find that the transplanted discs follow their own inherent developmental time table to complete differentiation. That is, within a period of 6 days they were fully differentiated with well tanned bristles whereas bristle tanning begins much later, 10 days after pupariation in the host species [13]. However, the cuticular tanning of the implants occurs simultaneously with that of the host cuticle (Fig. 2). Hardening and tanning of the cuticle is a critically timed event that is controlled by the neurohormone bursicon secreted soon after eclosion of the host fly [14] and therefore the housefly legs too undergo tanning after the eclo- sion of the host fly. Thus, the Sarcophaga pupal body cavity acts as a suitable environment for the differentiation of housefly imaginal discs. The central nervous P. SIVASUBRAMANIAN 499 system of larval housefly also undergoes complete metamorphic reorganization within the fleshfly pupa [4]. Such a system has the potential to be exploited for studies of neuronal specificity by means of in vivo culture of the CNS with the discs of the same or different species. ACKNOWLEDGMENT This study was supported by a grant from the Natural Sciences and Engineering Research Council of Canada. REFERENCES 1 Murphey, R. K., Bacon, J. P., Sakaguchi, D. S. and Johnson, S. E. (1983) J. Neurosci., 3: 659-672. 2 Sivasubramanian, P. and Nassel,D.R. (1985) J. Comp. Neurol., 239: 247-253. 3 Schmid, H., Gendre, N. and Stocker, R. F. (1986) Dev. Biol., 113: 160-173. 4 Sivasubramanian, P. (1987) J. Neurochem., 48 (suppl): S 59. Bhaskaran, G. and Sivasubramanian, P. (1969) J. Exp. Zool., 171: 385-396. Davis, K. T. and Shearn, A. (1977) Science, 196: 1438-1440. Millner, M. J. (1977) J. Embryol. Exp. Morphol., 37: 105-117. Nishirua, J. T. and Fristrom, J. W. (1975) Proc. Natl. Acad. Sci. USA, 72: 2984-2988. Hadorn, E. (1965) Brookhaven Symp. Biol., 18: 148-161. Bryant, P. J. (1975) J. Exp. Zool., 193: 49-78. Stocker, R. F. and Schmid, H. (1985) Experientia, 41: 1607-1609. Wentworth, S. L., Roberts, B. and O’Conner, D. (1981) J. Insect Physiol., 27: 435-440. Sivasubramanian, P. and Biagi, M. (1983) Int. J. Insect Morphol. Embryol., 12: 355-359. Fraenkel, G. and Hsiao, C. (1965) J. Insect Phys- iol., 11: 513-556. é * ™ o ZOOLOGICAL SCIENCE 5: 501-503 (1988) [COMMUNICATION] © 1988 Zoological Society of Japan Preference for Striped Backgrounds by Striped Fishes YASUTOSHI KOHDA and MUNETAKA WATANABE Department of Biology, College of Liberal Arts and Sciences, Okayama University, Tsushima, Okayama 700, Japan ABSTRACT — Six available species of freshwater fishes, two cross-striped, two lengthwise-striped and two non- striped ones, were exposed to vertically and horizontally striped backgrounds. The two cross-striped fishes pre- ferred to rest at vertically striped sites over horizontally striped ones. One of the two lengthwise-striped fishes tended to rest at horizontally striped sites. These results imply that many striped fishes prefer resting at sites with stripes similar to their own. The behavior of one non-striped fish suggests that there may be a factor other than a fish’s own stripes that causes preference of vertically striped sites. INTRODUCTION Cryptic animals must merge with their back- grounds [1, 2]. It is known that some cryptic animals choose resting places appropriate to their body colorations [3-5]. Kohda and Watanabe [6] showed that the freshwater serranid fish oyani- rami, Coreoperca kawamebari, which has cross stripes on its body, chooses to rest at vertically rather than horizontally striped sites. Do cross-striped fishes other than the oyanirami have the same preference? Do lengthwise-striped fishes prefer horizontally striped sites? Do non- striped fishes have any preference between verti- cally and horizontally striped sites? In the present study, we tested the preferences of six available species of freshwater fishes including the oyani- rami for striped sites. MATERIALS AND METHOD Ten individuals from each of six species were Accepted August 31, 1987 Received July 13, 1987 used; two cross-striped fishes, C. kawamebari (7.8-8.8cm in total length) and Macropodus chinensis (3.2-4.0 cm), two non-striped, Carassius auratus (6.3-8.0cm) and Acheilognathus limbata (5.0-7.4 cm), and two lengthwise-striped, Barbus titteya (3.0-3.4cm) and Melanochromis auratus (3.5-5.1 cm). The four former species were col- lected in Okayama Prefecture, Japan, while the latter two were obtained from a _ tropical-fish dealer. As all the specimens were small, we used an experimental apparatus different from that used in our previous study [6]. Instead, we utilized a gray plastic tank 15010050 cm high (20 cm deep), which had two vertically striped and two horizon- tally striped shelters (Fig. 1B). Shelters were transparent plastic boxes (15 x 1515 cm), which had three striped side walls and one open side, and were put on squares (25 20cm) bordered by a light green line (Fig. LA). The stripes were 2 mm black bands with 2 mm transparent intervals. A fish was placed in the center of the tank, and the time the fish spent in each square was recorded for 30 min. The fish was left in the tank and the next day its position was recorded again for 30 min. Five of the ten fishes of each species were tested under one arrangement of shelters (Fig. 1B), and the other five were tested under the reverse arrangement. Preference for stripes by each species was tested by the two-tailed matched pairs signed test [7]. RESULTS AND DISCUSSION In the first-day test, the two cross-striped spe- cies, C. kawamebari and M. chinensis, preterred 502 Y. KOHDA AND M. WATANABE oO foe) I———— 100 — > Fic. 1. Schematic representation of striped shelters and experimental apparatus. (A) Striped shelters (15 x 1515 cm) having three vertically or horizontally striped side walls and one open side were put on squares (25 x 20 cm) bordered by a light green line. The stripes were 2 mm black bands with 2 mm transparent intervals. (B) One arrangement of striped shelters in the experimental tank (150 10050cm). V: vertically striped shelters, H: horizontally striped shelters. Half of the specimens were tested under this arrangement of shelters and the other half were tested under the reverse arrangement. Ist day 2nd day vertically striped zones to horizontally striped ones (both r/n=0/9, P<0.01) (Fig. 2). The two length- wise-striped and the two non-striped species showed no preference (r/n=2/10, 1/8, 2/10, 2/6, all P>0.05). In the second-day test, one of the two length- wise-striped fishes, M. auratus, preferred horizon- tally striped sites (r/n=0/9, P<0.01) (Fig. 2). One of the two non-striped fishes, A. limbata, showed a slight preference for vertically striped zones (r/ n=1/9, P<0.05). The other four species showed no preference (1/n=4/9, 3/8, 2/9, 3/10, all P>0.05). On the first day, the specimens were unfamiliar C1 C2 N1 Fic. 2. The time that the fishes spent in the striped sites during the 30 min observation time. One dot represents the data of one specimen. The foot of the perpendicular from a dot on the V-axis is the time (min) spent in the vertically striped sites and the foot on the H-axis is the time spent in the horizontally striped sites. 1st day: the results of the tests immediately after placing the specimens into the experimental tank, 2nd day: the results after one-day of adaptation. Cl: Coreoperca kawamebari, C2: Macropodus chinensis, N1: Carassius auratus, N2: Acheilognathus limbata, L1: Barbus titteya, L2: Melanochromis auratus. *: P<0.05, **: P<0.01. N2 LI L2 Preference of Stripes by Fishes 503 with the experimental tank, and might have been frightened and thereby motivated to seek refuge. On the second day, they probably knew the geography in the tank, and their motivation to seek refuge was probably smaller than on the first day. No fish preferred the same striped zone on both days, but we can infer that the species which chose either zone on either day have a preference for that stripe. Kohda and Watanabe [6] showed that the oyanirami, a cross-striped fish, prefers vertically striped sites and this was reconfirmed in the present study. The other cross-striped fish, M. chinensis, showed the same preference. A length- wise-striped fish M. auratus, preferred horizontally striped sites. These three fishes tended to rest at shelters with stripes similar to their own. One lengthwise-striped fish, B. titteya, showed no pref- erence. None of the four striped fishes showed a preference for stripes unlike their own. These results imply that many striped fishes prefer resting at sites similar to their own body stripes, thereby camouflaging themselves. Further studies with more striped species are needed to prove this hypothesis. One non-striped species, C. auratus, showed no preference, and this result conforms to our hypoth- esis. However, the other non-striped one, A. limbata, showed a slight preference for vertical stripes over horizontal stripes on the second day. This observation means that there may be a factor other than a fish’s own stripes causing a preference for vertically striped sites. In order to determine what this factor is, we need to observe the behavior of many non-striped fishes that show a preference for vertical stripes and to find which character of these fishes correlates with their preference for vertical stripes. ACKNOWLEDGMENTS We wish to express our sincere thanks to Dr. Hiromi Iwagaki and Ms. Yurie Hiraoka Nakatsuka for their help in carrying out pilot experiments, and to Professor Susumu Ishii of Waseda University for advice concerning the statistical analyses. REFERENCES 1 Wickler, W. (1986) Mimicry in Plants and Animals. McGraw-Hill, New York. Edmunds, M. (1974) Defence in Animals. Longman, Harlow. Ergene, S. (1950) Z. vergl. Physiol., 32: 530-551. Kettlewell, H. B. D. (1955) Nature, 175: 943-944. Sargent, T. D. (1968) Science, 159: 100-101. Kohda, Y. and Watanabe, M. (1986) Ethology, 72: 185-190. 7 Siegel, S. (1959) Nonparametric Statistics for Be- havioral Sciences. McGraw-Hill, New York, pp. 68- TS: te Nn WW 1988 MEETING OF THE AMERICAN SOCIETY OF ZOOLOGISTS and AMERICAN MICROSCOPICAL SOCIETY, ANIMAL BEHAVIOR SOCIETY, THE CRUSTACEAN SOCIETY, INTERNATIONAL ASSOCIATION OF ASTACOLOGY, SOCIETY OF SYSTEMATIC ZOOLOGY, AND WESTERN SOCIETY OF NATURALISTS SAN FRANCISCO HILTON & TOWERS SAN FRANCISCO, CALIFORNIA DECEMBER 27 - 30 Housing Rates: $66 Single, Double, Triple & Quad Call for Papers: April 1988 Abstract Deadline: August 8, 1988 For Oral and Poster Presentations SYMPOSIA/WORKSHOPS: Recent Developments in the Study of Animal Migration Parasites and Sexual Selection Evolving Concepts of Chemical Mediation: A Symposium In Honor of Howard A. Bern Marine Invertebrate Allorecognition and the Evolution of Immunity Concepts of Efficiency in Biological Systems Concepts of Adaptation in Aquatic Animals: Deviations from the Terrestrial Paradigm Cellular and Molecular Biology of Pattern Formation Developmental Neurobiology of the Cnidaria Antarctic Marine Biology Chemical Factors that Influence the Settlement and Metamorphosis of Marine Invertebrate Larvae Cracking a Black Box: Field Inferences in the Ecology of Marine Invertebrate Larvae Species and Evolution in Clonal Organisms Biology of Nonmammalian Chordate Testis A History of Regeneration Research Science As A Way of Knowing — Cell and Molecular Biology Sex Attraction, Mating Behavior and Insemination in the Crustacea The Complete Biology of Giant Kelp Workshop on Science Comes to California Workshop on Research-Education at Small College and Universities: Quality Science on a Frayed Shoestring KKK KKK KKK KK KK KK Hosted by the California Academy of Sciences -— OO Daphne G. Fautin and Ralph |. Smith Co-Chairpersons of the Local Arrangements Committee S ALN FRA NCISCO NUMEROUS SOCIALS, SPECIAL PROGRAMS COMMERCIAL EXHIBITS..... JOB PLACEMENT SERVICE..... For more information, contact: Mary Adams-Wiley, Executive Officer American Society of Zoologists 104 Sirius Circle Thousand Oaks, California 91360 Telephone: (805) 492-3585 ES £ s Wa Published by (4 reniia tion the Japanese Society of Developmental Biologists The journal is devoted to the publication of original papers dealing with any aspects of developmental phenomena in all kinds of organisms, including plants and micro-organisms. Papers in any of the following fields will be considered: developmental genetics, growth, morphogenesis, cellular kinetics, fertilization, cell division, dormancy, germination, metamorphosis, regeneration and pathogenesis, at the biochemical, biophysical and analytically morphological levels ; reports on techniques applicable to the above fields. At times reviews on subjects selected by the editors will be published. Brief complete papers will be accepted, but not preliminary reports. Members of the Society receive the Journal free of charge. Subscription by institutions is also welcome. Papers in Vol. 30, No. 2. (April 1988) 11. REVIEW: T. KisHimoto: Regulation of metaphase by a maturation-promoting factor. 12. S. Tanimoto and M. Morisawa: Roles for potassium and calcium cahnnels in the initiation of sperm motility in rainbow trout. 13. K.MIKAMI-TAKEI, A. FUJIWARA, and I. YASUMARU: The acrosome reaction induced by dimethylsulfoxide in sea urchin sperm. 14. T. KAWASHIMA and T. NAKAZAWA : Stimulation of protein synthesis in the mitochondria of sea urchin embryos before gastrulation. 15. G. Casazza, R. DE Samtis, and M.R. Pinto: Plasma membrane glycoproteins of Ciona intestinalis spermatozoa that interact with the egg. 16. M. YAmacucui, T. Niwa, M. Kurita, and N. Suzuki: The participation of speract in the acrosome rection of Hemicentrotus pulcherrimus. 17. R.MortyaMa and K. YANAGISAWA: Protein synthesis initiated by cell fusion in Dictyostelium discoideum. 18. A.M.Montes and W.K. MorisHiGce: Lung-derived growth factors: Ontogenic shift in parahormone secretion in the perinatal rat lung. 19. J.C. Lasse, A. Picarp, and M. Doree: Does the M-phase promoting factor (MPF) activate a major Ca’*-and cyclic nucleotide-independent protein kinase in starfish oocytes? Development, Growth and Differentiation (ISSN 0012-1592) is published bimonthly by The Japanese Society of Developmental Biologists, Department of Biology, School of Education, Waseda University, Tokyo 160, Japan. 1987: Volume 29. Annual subscription U. S. $ 110.00 including air speed delivery except Japan. Application to mail at second class postage rate is pending at Jamaica, NY 11431, U.S.A. Outside Japan: Send subscription orders and notices of change of address to Academic Press, Inc., Journal Subscription Fulfillment Department, 6277 Sea Harbor Drive, Orlando, FL 32887, U.S. A. Send notices of change of address at least 6-8 weeks in advance. Please include both old and new addresses. U.S.A. POSTMASTER: Send changes of address to Development, Growth and Differentiation, Academic Press. Inc., Journal Subscription Fulfillment Department, 6277 Sea Harbor Drive, Orlando, FL 32887, U.S. A. In Japan: Send nonmember subscription orders and notices of change of address to Business Center for Academic Societies Japan, 16-3, Hongo 6-chome, Bunkyo-ku, Tokyo 113, Japan. Send inquiries about membership to Business Center for Academic Societies Japan,4—16, Yayoi 2-chome, Bunkyo-ku, Tokyo 113, Japan. Air freight and mailing in the U.S. A. by Publications Expediting, Inc., 200 Meacham Avenue, Elmont, NY 11003, U.S.A. NARISHIGE THE ULTIMATE NAME IN MICROMANIPULATION OUR NEW MODELS MO-102 and MO-103 MAKE PRECISION MICROMANIPULATION SO EASY! (Photo: by courtesy of Olympus Optical CO., LTD,) SOME FEATURES of MO-102 and MO-103: * The manipulator head is so small that it can be mounted directly on the microscope stage. There is no need for a bulky stand. * Hydraulic remote control ensures totally vibration-free operation. * 3-D movements achieved with a single joystick. Micromanipulators Microelectrode pullers Stereotaxic instruments NARISHIGE SCIENTIFIC INSTRUMENT \2=&7 LABORATORY CO.,LTD. 4-9-28, Kasuya, Setagaya-ku, Tokyo 157 JAPAN Telephone: 03-308-8233 Telex: NARISHG J27781 Sophisticated Balance between Safety and Centrifugation Capability without Compromise. Centrifuge in Integrated with A Refrigerator Extra-Quiet Operation Ease of Loading/ Unloading The Rotors Quick Start/ Quick Stop High Quality Triple Safety Design Corrosion Resistance TOMY CORPORATION 1002 SOLEIL NARIMASU BLDG., 31-8, NARIMASU 1-CHOME, ITABASHI-KU, TOKYO 175 JAPAN TLX:02723111 TOMYCO J CABLE: TOMYSHO TOKYO FAX:(GIIl GII)(03)930-7010 TOMY SEIKO CO.,LTD. 2-2-12, ASAHICHO NERIMA-KU, TOKYO 176 JAPAN TEL:(03)976-3411 TEL:(03)976-3111 HIGH SPEED REFRIGERATED MICRO CENTRIFUGE move. MR-150 MANUFACTURER (Contents continued from back cover) hybridization method for neurohypophysial hormone mRNAs using synthetic oligonu- cleotide probes Seki, T., S. Kikuyama and M. Suzuki: Effect of hypothalamic extract on the prolactin re- lease from the bullfrog pituitary gland with special reference to thyrotropin-releasing hormone (TRH) Morphology Fujikura, K., S. Kurabuchi, M. Tabuchi and S. Inoue: Morphology and distribution of the skin glands in Xenopus laevis and their re- sponse to experimental stimulations Behavior Biology Ooka-Souda, S., H. Kabasawa and S. Kinoshi- ta: Circadian rhythms in locomotor activity of the hagfish, Eptatretus burgeri. 11. The effect of brain ablation Ooka-Souda, S. and H. Kabasawa: rhythms in locomotor activity of the hagfish, Circadian Eptatretus burgeri. Ill. Hypothalamus: a locus of the circadian pacemaker? ....... 437 Weldon, P.J.: Feeding responses of Pacific snappers (genus Lutjanus) to the yellow- bellied sea snake (Pelamis platurus) ...... 443 Kohda, Y. and M. Watanabe: Preference of striped backgrounds by striped fishes (COM- MUNICATION) Ecology and Taxonomy Hanzawa, N., N. Taniguchi and K. Numachi: Geographical differentiation in populations of Japanese dace Tribolodon hakonensis de- duced from allozymic variation Konishi, K. and R. Quintana: The stages of three pagurid crabs (Crustacea: from Hokkaido, larval Anomura: Paguridae) Japan Sawada,I. and A.L.Molan: Two new- hymenolepidid cestodes, Vampirolepis mola- ni sp. n. and V. iragensis sp. n., from Iraqi Dats: saeparcn a nesise asetnancce nines seeine cee 483 ZOOLOGICAL SCIENCE VOLUME 5 NUMBER 2 APRIL 1988 CONTENTS Obituary ...4seced cseeoes adeno eeee 213 Genetics and Immunology REVIEWS Naruse, K., A.Shimada and_ A. Shima: Meusy, J.-J. and G. G. Payen: Female repro- duction in malacostracan Crustacea Nishioka, R. S., K. M. Kelley and H. A. Bern: Control of prolactin and growth hormone secretion in teleost fishes ORIGINAL PAPERS Physiology Ozaki, M.: A possible sugar receptor protein found in the labellum of the blowfly, Phor- mia regina Okano, Y., E. David, K. Honda and S. Inoué: Auditory evoked potentials dynamically re- lated to sleep-waking states in unrestrained Tats, wer whieg ancaiWekestnawian ae anerestamnets aie: 291 Tazaki, K.: The anatomy and physiology of the stomatogastric nervous system of Squilla. II. The cardiac system Obika, M.: Ultrastructure and physiological response of leucophores of the medaka Oryzias latipes Cell Biology Suganuma, Y. and H. Yamamoto: Its relation to con- Conjuga- tion in Tetrahymena: canavalin A receptor distribution on the cell surface Iwasaki, S. and K. Kobayashi: of the dorsal tongue surface in the Japanese toad, Bufo japonicus (Anura, Bufonidae) 331 Okamoto, M.: cle in the Japanese common newt, Cynops Fine structure Fine structure of the iris mus- pyrrhogaster, with special reference to in- nervation Gene-centromere mapping for 5 visible mutant loci in multiple recessive tester stock of the medaka (Oryzias latipes) (COM- MUNICATION) Nagata, S.: T cell-specific XTLA-1 antigens from Xenopus laevis tadpole and froglet are not identical (COMMUNICATION) Biochemistry Tsuneoka, M., K. Maruyama and K. Ohashi: In vitro dimerization of I-protein, an A-I junctional component of skeletal muscle myofibrils Developmental Biology Iwamatsu, T., T. Ohta, E. Oshima and N. Sakai: Oogenesis in the medaka Oryzias latipes—stages of oocyte development ...353 Tsuchiyama-Omura, S.,_ B.Sakaguchi, K. Koga and D.F. Poulson: Morphological features of embryogenesis in Drosophila melanogaster infected with a male-killing spiroplasma Suematsu, N., H.Takeda and_ T. Mizuno: Glandular epithelium induced from urinary bladder epithelium of the adult rat does not show full prostatic cytodifferentiation ....385 Sivasubramanian, P.: Interspecific transplan- tation of developing tissues and their subse- quent differentiation in flies (COM- MUNICATION), 2. ..65.52.02 ge eee eee 497 Endocrinology Hyodo, S., M. Fujiwara, S. Kozono, M. Sato and A. Urano: Development of an in situ (Contents continued on inside back cover) INDEXED IN: Current Contents/LS and AB & ES, Science Citation Index, ISI Online Database, CABS Database Issued on April 15 Printed by Daigaku Printing Co., Ltd., Hiroshima, Japan ZOOLOGICAL SC] ae An Internatio nal Jou ZOOLOGICAL SCIENCE The Official Journal of the Zoological Society of Japan Editor-in-Chief: The Zoological Society of Japan: Hideshi Kobayashi (Tokyo) Toshin-building, Hongo 2-27-2, Bunkyo-ku, Ma vaeine ECuOL) ee Tokyo 113, Japan. Tel. (03) 814-5675 Seiichiro Kawashima _ (Hiroshima) Oficers: Assistant Editors: Takeo Machida (Hiroshima) Sumio Takahashi (Hiroshima) President: Nobuo Egami (Tsukuba) Secretary: Hideo Namiki (Tokyo) Treasurer: Tadakazu Ohoka (Tokyo) Kazuyoshi Tsutsui (Hiroshima) Librarian: Shun-Ichi Uéno (Tokyo) Editorial Board: Howard A. Bern (Berkeley) Walter Bock (New York) Aubrey Gorbman (Seattle) Horst Grunz (Essen) Robert B. Hill (Kingston) Yukio Hiramoto (Chiba) Susumu Ishii (Tokyo) Yukiaki Kuroda (Mishima) Koscak Maruyama (Chiba) Roger Milkman (Iowa City) Hiromichi Morita (Fukuoka) Kazuo Moriwaki (Mishima) Tokindo S. Okada (Okazaki) | Andreas Oksche (Giessen) Hidemi Sato (Nagoya) Hiroshi Watanabe (Shimoda) | Mayumi Yamada (Sapporo) Ryuzo Yanagimachi (Honolulu) ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and communications in the broad field of Zoology. The journal was founded in 1984 as a result of unification of Zoological Magazine (1888-1983) and Annotationes Zoologicae Japonenses (1897-1983), the former official journals of the Zoological Society of Japan. ZOOLOGICAL SCIENCE appears bimonthly. An annual volume consists of six numbers of more than 1000 pages including an issue containing abstracts of papers presented at the annual meeting of the Zoological Society of Japan. MANUSCRIPTS OFFERED FOR CONSIDERATION AND CORRESPONDENCE CONCERN- ING EDITORIAL MATTERS should be sent to: Dr. Seiichiro KAWASHIMA, Managing Editor, Zoological Science, Zoological Institute, Faculty of Science, Hiroshima University, 1-1-89 Higashisenda-machi, Naka-ku, Hiroshima 730, Japan, in accordance with the instructions to authors which appear in the first issue of each volume. Copies of instructions to authors will be sent upon request. SUBSCRIPTIONS. ZOOLOGICAL SCIENCE is distributed free of charge to the members, both domestic and foreign, of the Zoological Society of Japan. To non-member subscribers within Japan, it is distributed by Business Center for Academic Societies Japan, 6-16-3 Hongo, Bunkyo-ku, Tokyo 113. Subscriptions outside Japan should be ordered from the sole agent, VSP, Utrechtseweg 62, 3704 HE Zeist (postal address: P.O. Box 346, 3700 AH Zeist), The Netherlands. Subscription rates will be provided on request to these agents. New subscriptions and renewals begin with the first issue of the current volume. All rights reserved. No part of this publication may be reproduced or stored in a retrieval system in any form or by any means, without permission in writing from the copyright holder. © Copyright 1988, The Zoological Society of Japan Publication of Zoological Science has been supported in part by a Grant-in-Aid for Scientific Publication from the Ministry of Education, Science and Culture, Japan. ZOOLOGICAL SCIENCE 5: 505-506 (1988) © 1988 Zoological Society of Japan General Introduction to the Special~Issue be on Advances in Cell Division Research _ The publication of special issues of Zoological Science was proposed by the Editorial Board and approved by the council of the Zoological Society of Japan in 1986. This special issue on “Advances in Cell Division Research”, dedicated to Emeritus Professor Katsuma Dan, was suggested by Professor N. Egami, the former Editor-in-Chief and the present President of the Zoological Society and is issued with the full support of the Editorial Board of Zoological Science. Emeritus Professor Dan is now 83 years old, yet he remains active and is working by himself on cell division of embryos of marine invertebrates at the Misaki Marine Biological Station. Since he began work on the surface potential of Arbacia eggs in the laboratory of the late Professor L. V. Heilbrunn, he has dedicated himself for almost 60 years to the study of cell division, which has profoundly influenced various areas of biological science. Professor Dan graduated from the Department of Zoology of the Tokyo Imperial University (now the University of Tokyo) in 1929. He joined the late Professor Heilbrunn’s laboratory at the University of Pennsylvania in December of 1930, and in 1934 he received the Ph. D. degree there. After returning to Japan he was appointed instructor at the Misaki Marine Biological Station of the Tokyo Imperial University. Two years later, he married Jean Clark of New Jersey. His well-known studies on the behavior of the cell surface during cleavage began after he returned to Japan in 1937 and continued until 1947. The series of his experiments became the basis for a proposed model for the mechanism of cytokinesis, the astral model that elucidates the role of the spindle elongation in the initial shrinkage during constriction of the cell (1943) (see the chapter by Professor Dan herein). Furthermore, his perforation experiments during this period (1943) beautifully demonstrated the fascinating rules of interactions between the mitotic apparatus and the egg cortex, which later contributed to the well-known concept of the “cleavage signal” proposed by Professor Daniel Mazia in 1961 in The Cell, III. In 1943, he was appointed lecturer at the Tokyo Imperial University. In 1945 he was the last to leave the Misaki Marine Biological Station at the end of the war, when the Station was occupied by the American forces. He made great efforts to take the Station back from the Occupation forces, and on December 31, 1945, it was returned to the University of Tokyo. Professor Dan returned to the Station in the summer of 1946, starting a series of studies with Jean on the mechanism of unequal division (1947). Soon he was appointed full professor at the Tokyo Metropolitan University, with the strong support of Emeritus Professor Daigoro Moriwaki, the former president of the National Institute of Genetics at Mishima. In 1950, he was awarded the Zoological Society of Japan Prize for his studies on the mechanism of cell division in marine eggs. With one of his students, Professor Shinya Inoue, he had the great success of observing birefringence of dividing cells for the first time (see the chapter by Professor Inoue). Soon he was invited to work on the mitotic apparatus with his longtime friend, Professor Daniel Mazia. They succeeded for the first time in isolating the mitotic apparatus and characterizing its protein components. This pioneering work has had a major influence on later investigations of the structure and function of the mitotic apparatus. Many of his colleagues can recall Professor Dan’s memorable lecture on the isolated mitotic apparatus at the 1952 annual meeting of the Zoological Society. 506 H. SAKAI Although he was reluctant to be an educational administrator, he was elected President of Tokyo Metropolitan University in 1965, and served until 1973, for an eight-year period that included the years of serious student unrest in Japan. During this period, he also served as President of the Zoological Society of Japan (1967-1972) and the Japanese Society of Developmental Biologists (1968-1972). After retiring from Tokyo Metropolitan University in 1973, he returned to his laboratory in the Misaki Marine Biological Station to work on the mechanism of unequal cleavage. He was invested with the Second Order of Merit in 1976. Last year, he was named a Person of Cultural Merit in the annual Honors list for his long dedication to biological science. We are pleased to dedicate the chapters of this special issue to Emeritus Professor Dan Festschrift in honor of his long devotion to cell division research. Finally, I would like to express many thanks to Drs. D. Mazia, Y. Hiramoto, M. Yoneda, H. Sato and I. Mabuchi for their valuable collaboration in the organization and editorial works of this issue. March 1988 HIKOICHI SAKAI ZOOLOGICAL SCIENCE 5: 507-517 (1988) Mechanism of Equal Cleavage of Sea Urchin Egg: Transposition from Astral Mechanism to Constricting Mechanism KatsumA Dan! Misaki Marine Biological Station, Koajiro, Miura-shi, Kanagawa 238-02, Japan ABSTRACT — In past 40 years, there have been two opposing theories on the mechanism of cleavage for sea urchin eggs. One was the author’s theory in which elongation of the spindle is a main source of force which is assisted by the presence of two asters. This will be called “astral mechanism”. The other was Schroeder’s constricting ring theory. He argues that the ring consists of bundles of actin fibers and by the contraction of these fibers, they pinch a cell into two equal halves. The strong point of the astral mechanism is that it has quantitative verifications for the degree of equatorial shrinkage and also for the depth of forming furrow. Its shortcoming is that it works for only a few minutes at the beginning of cleavage. After the equatorial surface goes into a stretching phase, the theory no longer works. The strong point of the constricting mechanism is that it provides an ideal way to pinch a cell into two parts but its shortcoming is that the ring cannot be found as long as the egg remains spherical which means that a spherical cell can never start to cleave. Emphasis of the present paper is to point out that both mechanisms are at work in a process of cleavage but they are activated in succession. Furthermore, by the author’s analysis, it is possible to © 1988 Zoological Society of Japan determine the time point of switching from the astral mechanism to the constricting mechanism. HISTORICAL Course of advancement of our knowledge on the mechanism of cell division of sea urchin egg can be divided into three steps. Setp 1 In 1943, the present author proposed a scheme for division of the sea urchin egg [1]. There, the author contended that the two asters have their astral rays crossing on the equatorial plane (see Fig. 4 of this paper). If these asters are pushed away from each other by an intervening and elon- gating spindle, the equatorial surface shrinks as indicated by coming closer of two marker particles. The degree of shrinkage of the equatorial surface and the depth of the deepening furrow can be accounted for semi-quantitatively by the scheme. Accepted March 17, 1988 Correspondence should be addressed to the following home address: 3-19-8 Kami-Yoga, Setagaya-ku, Tokyo 158, Japan. For convenience, this scheme will be called the astral mechanism. However, after the beginning of furrowing, in a few minutes, the equatorial surface begins to stretch. Once this stage is reached, the proposed scheme no longer holds. In other words, the stage in which the astral mechanism is valid is restricted to the initial phase of the furrow formation, in spite of the fact that it is provided with semi- quantitative proofs. Step 2 This step spans between 1953 to 1956. In 1953, Swann and Mitchison in England [2] and in 1956, Hiramoto in Japan [3] conducted two different experiments and arrived at closely similar conclu- sions. The former workers tried to destroy the spindle rapidly by concentrated colchicine solutions and the latter worker tried to remove the spindle by sucking it in a capillary. Both groups reached a unanimous conclusion that if the spindle is de- 508 K. DAN stroyed or removed during metaphase, the egg remained spherical forever, but if operated at the ana- or telophase, the egg could cleave completely in spite of the absence of the spindle. These experiments show that either the reaction of a dividing cell changes in the course of cell division or a new division mechanism comes into play in the latter half of the division process, although the above workers failed to pinpoint what was a real nature of the change. Step 3 In 1972, by using the eggs of Arbacia punctulata, Schroeder discovered a special structure under- lying the surface of cleavage furrow in thin sections by transmission electron microscopy [4]. Later he succeeded in identifying the structure as bundles of actin fibers running along the bottom of the furrow EF If the bundles of actin fibers encircle the bottom of the furrow, it is not hard to imagine that the bundle would contract and pinch the cell to com- plete separation. This offers a concrete way of pinching which Swann and Mitchison as well as Hiramoto alluded to before. In passing, it must be mentioned that in 1965, Hiramoto injected a large amount of paraffin oil or sea water into the sea urchin egg which was about to divide. He reported that in spite of an enor- mous inflation of volume and shifting of the spin- dle to the cell periphery, the cleavage furrow appeared exactly at the place anticipated from the position of the spindle before the injection, the furrow dividing the inflated cell body exactly into two equal halves [6]. This experiment, however, is directed to a slightly different aspect of division process from those quoted above. It means that once a constricting ring is formed, its position cannot be shifted by inflating the cell volume. ANALYSES Concerning step 1 When the author’s original work was done, it was during the World War II. As the result, various quantities (surface shrinkage, surface stretching, spindle length etc.) were measured on camera lucida drawings. This must have sliced off reliability of the data in the eyes of readers. Besides, the distribution of the Journal in which the work was published was very limited at that time. For these reasons, new sets of data were collected. A brief explanation of the author’s scheme may be necessary here. When two kaolin particles are attached to the equatorial surface of the egg, the distance between them shrinks for first few mi- nutes and then it begins to stretch. This behavior is specific to the equatorial surface and not shared by other regions of the egg surface (Figs. 1 and 2) [7]. This shrinkage begins with the onset of spindle elongation together with an elongation of the cell OND Gea OG Fic. 1. A record of surface behavior of the egg of Mespilia globulus during cleavage as revealed by attaching marker particles. (Reproduced by permis- sion of Protoplasma). -20 Fic. 2. Graphic presentation of the data of Fig. 1. A sharp contrast in behavior between the surface of the furrow and that of adjacent regions. (Repro- duced by permission of Protoplasma). Cleavage Mechanism of Sea Urchin Egg 509 body in the direction of the spindle axis (Fig. 3) (Ay. Now the astral rays of the two asters are crossing on the equatorial plane (Fig. 4). Since the tips of the rays are anchored to the cortical gel (Fig. 1 in [3]), they are not free to move separately but they must move as a mass. When the astral centers are pushed away from each other by the elongation spindle, the loci of the tips of the rays will be as shown in Figure 5. It is obvious that while the equatorial surface is being pulled in toward the spindle, it shrinks at the same time. Here, I would like to call attention of readers to the fact that the author applies the term “furrow” to the stage when a circular contour of the cell becomes flat as shown in Figure 5. In the litera- ture, some people use the term only after an indentation appears on this flat surface. Fic. 3. Simultaneity among spindle (@) and cell body (©) elongation and shrinkage of furrow surface (x). Fic. 4. Two examples of Clypeaster eggs just before cleavage. (Please ignore attaching particles). One can see crossing of the rays on the equatorial regions. The widest crossing ranges are indicated by a pair of lines outside the pictures. Since rays are radiating out from a center, it is extremely difficult to get clear images of crossing rays along their lengths. This drawback is compensated by flattening the cell or by moving focus up and down. Fic. 5. Loci of the tips of crossing rays as astral centers are pushed away by elongating spindle. The equato- rial surface shrinks as it is drawn in toward the spindle. 510 K. DAN Going over to Figures 6 and 7, these are mi- crophotographs of the eggs of Clypeaster japonicus with two particles of activated clay attached on the equatorial surface. In taking these photographs, special care was taken for the following conditions. (a) Eggs should be selected in which the spindle is strictly parallel to the substratum so that it won’t tilt during measurement. (b) Two attached parti- cles must be situated roughly on the widest cros- sing range of the crossing rays (Fig. 4) If the Fic. 6. An egg of Clypeaster japonicus in cleavage. (Times of photographing: 0”, 1°45”, 2°50’, 3’20’, 4°45’, 600’). Two dots on upper furrow surface are attached clay particles. The distance between them reaches a minimum in the 3rd picture. Astral centers are separating owing to a push by the elongating spindle. (Smallest division of the scale is 10 um). In explanatory drawing, clay particles are represented by arrow heads and positions of the particles anticipated by the author’s theory (obtained by the methods in Fig. 5) are shown by X. Cleavage Mechanism of Sea Urchin Egg 511 Fic. 7. Another example of the same experiment as Fig. 6. (Times of photographing: 0”, 3'00", 4°30”, 5°00”, 5°30”, 630°’). The astral rays are sharper here than Fig. 6. By the time constricting ring begins to work (4th—Sth photos), crossing of the rays is no longer discernible and the rest of the rays are bending into a fountain figure as the result of push against the polar surface. (Smallest division of the scale is 10 «m). positions of the particles do not coincide with actual crossing range, good agreement between the degree of observed approach of the particles and the degree of calculated shrinkage by drawings of the intersection points cannot be expected [1]. The present author’s crossing range could be a concrete basis for Rappaport’s ‘joint astral in- fluence’ [8]. (c) Camera must be sharply focussed on the two astral centers. Owing to the fact that particles are on the cell periphery, one tends to get an impression that sharp focussing on the cell contour is important, which, however, is not the case. Since the contour is being looked down upon tangentially from above, a slight up and down deviation of the image of the cell contour does not affect the calculation while slight errors in posi- tioning of astral centers upsets the accuracy of the calculation to a large extent. In the accompanying explanatory drawings of Figures 6 and 7, attaching particles are indicated by arrow heads and separating centers of the asters or the tips of the elongation spindle are shown by black dots. The particles, astral centers and cell contour are traced from the photographs. In the first explanatory figure,4 segments of line are obtained by connecting the particles and the astral centers. These represent a pair of crossing rays and two non-crossing rays. From the second drawing on, keeping the lengths of the segments unchanged and using them as radii, 4 circles are drawn placing the centers of the circles at moving positions of the astral centers. Intersections of the circles are indicated by X. These X’s show the positions where the particles would be by the astral mechanism. Degrees of shrinkage of particle distance and $12 K. DAN TABLE 1. Changes of the particle distance (P) and intersectional distance (D) in the egg shown in Fig. 6 P D Time distance % distance % 0” *12.3mm 100.0 *12.3mm 100.0 100” 11.8 95.9 10.8 87.8 1°45” *10.3 83.7 * 9.9 80.5 2ai5:° 10.2 82.9 10.2 82.9 2’50° * 9.8 79.7 * 9.4 76.4 3°30" “11-1 90.2 * 8.0 65.0 410” 12.9 104.9 TD 61.0 4°45” *13.0 105.7 * 6.9 55.1 520” 16.2 151.7 6.1 49.6 600" *19:2 156.3 * 6:0 48.2 Series VII: 24.0°C, crossing angle 62°. Underlined numerals are minimum values. Asterisks are cases in Fig. 6. those of the intersectional distance are shown in Tables 1 and 2 and plotted in Figures 8 and 9. In the Tables and Figures, all the data are given, among which cases illustrated in the photographs are indicated with asterisks. Two features are obvious. (1) The distance between actual paricles and that of the intersec- tions decreases pari pasu until the end of the shrinkage phase of the equatorial surface. During the following stretching stage, although the in- tersections continue to come closer, particles move away from each other. (2) Roughly until this turning point (third drawing of Fig. 6, 3rd—4th drawings of Fig. 7) the intersection points fall on the contour line of the deepening furrow. Coinci- dence between the transition from shrinkage to stretching and the point of disparity between the positions of particles and intersections is almost perfect in Figure 6 but it is slightly off in Figure 7. However, on examining the data of Figure 7 close- ly, it is revealed that although the particle distance of the 8th point or the 4th drawing is certainly smaller than the preceding step, the difference between them is less than in previous steps (see Table 2 and Fig. 9). This may mean that an actual minimum must have happened between them, at some time during 30 second interval and the seeming minimum point must be representing very beginning of the stretching phase. Repeating the conclusion once more, during the T T T 1 = lope fe Ronin 1 2 3 4 5 6 Fic. 8. A graph showing changes in distance of particles (open circles) and of intersections (filled circle) of the data of Fig.6. Asterisks are cases shown in photographs. Cleavage Mechanism of Sea Urchin Egg TaBLE 2. Changes in the particle distance (P) and intersectional distance (D) in the case shown in Fig. 7 P D Time distance % distance % 0” *14.7mm 100.0 14.7 mm 100.0 2°00” 14.1 95.9 14.1 95.9 2°30" 12.0 81.6 12.0 81.6 300” *11.9 81.0 10.5 71.4 3°30" 11.0 74.8 10.5 71.4 400” 10.0 68.0 10.5 71.4 4°30” - 91 61.2 9.0 612 5°00” 78.9 60.5, 8.0 54.4 5°30” * 9.9 67.3 6.8 46.3 6°00 12.6 85.7 7.0 47.6 6°30” 712.9 87.8 6.9 46.9 700 17.0 1S a7 6.9 46.9 Series I: 24.5°C, crossing angle 58°. Underlined numerals are minimum values. Asterisks are cases shown in Fig. 7. % words, from this stage on, the actual furrow is 120-4 always deeper than expected by the astral mechan- ®: ism which, in turn, suggests that the constricting mechanism is being activated. Concerning step 2 It was an important discovery that both English and Japanese investigators became aware of the fact that the cell behavior changes between metaphase and anaphase. Since Swann and Mitchison were using a polar- ization microscope, they could discern the chromo- some stage more or less [2]. On the other hand, since Hiramoto was using phase contrast optics, he could not see the condition of chromosomes [3]. The present author presumes that Hiramoto’s criteria of meta- and anaphase were probably whether the cell was spherical or with a furrow. However, this bears a relation to a delicate point of discussion in the next section. ae 1 9 94 4 5 6 72min Fic. 9. A similar graph as Fig. 8 for the data of Fig. 7. shrinkage phase of the equatorial surface, the astral mechanism can account for the degree of — Concerning step 3 shrinkage quantitatively and it can also define the depth of the forming furrow. However, once the stretching phase of the equatorial surface begins, the behavior of the particles and the intersections diverge and the positions of the intersections begin to go outside of the furrow contour. In other By the first impression, Schroeder’s thesis sounds almost perfect, since he succeeded to iden- tify the special structure underlying the cleavage furrow as bundles of actin fibers. To scrutinize his idea we must keep in mind a very crucial statement of his that the ring does not exist as long as the cell 514 K. DAN remains spherical while it apears only after a furrow is formed (pp.422—425 in [4]). On the other hand, he also tries hard to show that the ring is present from the very beginning of furrow forma- tion: 20 sec after the end of anaphase (pp. 422, 425, and Fig. 6 in [4]). Of course, we can under- stand his strong wish to make the ring appear as early as possible and to make it do the entire work of furrow formation, since the constricting mechanism is the only scheme he has in mind. At any rate, Schroeder’s two statements even sound somewhat contradictory. There are two papers reporting the absence of the constricting ring in early phase of nuclear division. Usui and Yoneda [9] showed a series of meridional thin sections of the eggs of Hemicentro- tus pulcherrimus in early phase of mitosis. They concluded that until late telophase when a slight decrease in equatorial diameter is percieved, the ring is absent ([9], Fig. 3a). But in eggs 1 min after furrow formation a ring is there (Fig. 3b in [9]). Yonemura and Kinoshita [10] obtained peeled cortices isolated by Vaquier’s cortical lawn method [11] of Clypeaster eggs. They stained the cortices by phallacidin conjugated with nitrobenzox- adiazole to examine the distribution of F-actin. When cortices are prepared from telophase cells, they could recognize constricting rings as bands of F-actin, while they failed to find them when taken from anaphase cells. At any rate, the ring does not seem to exist through anaphase to the first half of telophase when the spindle elongation is taking place. Actually the situation is more complicated than Schroeder’s notion. By Rappaport’s experiments, we now know that some time before actual forma- tion of a furrow, a precursor or some basis of a future furrow is laid by induction by the asters [12]. That is to say that once the induction becomes completely effective, the presence of the asters is no longer required for the formation of the furrow. The time when this induction takes effect is 4 min before the appearance of the furrow in Echinar- achnius parma. Comparable figures for Japanese sea urchins by Hamaguchi are of the same order of magnitude [13]. Taking advantage of this situa- tion, Rappaport prepared a cylindrical egg in a glass capillary of 80 um in diameter and after leaving the mitotic apparatus at one place during effective time for the induction, he moved it to a new location in the egg. Thereby he succeeded in making furrows appear and regress more than 10 times in a single cell [14]. It was mentioned earlier that Schroeder con- fronted a very queer fact that he could not find the constricting ring unless the egg has already a furrow. Giving a direct expression, it means that spherical eggs can never start to cleave. This is a serious matter. If the induction story is superim- posed on this fact, the situation becomes still more complicated. Facing such a complicated circum- stance, the author feels that there may be a missing link in the story which has been overlooked so far. This link could be called an assembly-initiating factor and the factor could be the spindle elonga- tion itself, as the result of which dormant induction effect is made apparent as a constricting ring. From the author’s standpoint, the astral mecha- nism invariably precedes the constricting mecha- nism. In the past, people, not being aware of this fact, thought the asters offer sufficient and ade- When people use mitotic poisons to suppress cleavage, the author believes that the effects of the poisons are primarily on the spindle elongation and inhibi- tion of furrow formation is a secondary result. Then what is the merit of spindle elongation? (1) In our daily experience, under some adverse con- ditions, a shallow furrow which once appeared often regresses. Rappaport would say that, in this case, induction is inadequate or it is during latent period. But the peiod under consideration is the intervening time (4 min) between the induction and realization of the furrow. As far as this period is concerned, the autnor feels that it would be as likely as Rappaport’s interpretation as to think that shallow furrow is formed by the elongation of the spindle and the next step of assembly of actin fibers is failing. Quoting Rappaport: the result indiates (Fig. 7) that in these cylindrical cells, a considerable portion of the increase in interastral distance takes place after the position of the fur- row is fixed (p. 570 in [12]). (2) From Rappaport’s measurements [12], the strength of cleavage stimulus of the asters seems to decrease more or less proportionally to the dis- quate explanation for furrow formation. Cleavage Mechanism of Sea Urchin Egg 515 tance between the aster and the cell surface. In general, in intact spherical eggs, presence of only one aster is considered to be inadequate to form a furrow [7, 15, 16]. However, even under this condition, by reducing the distance between the asters and cell surface, say, by making the cell cylindrical, the cell can divide. Under normal condition, while the cell remains spherical, asters lie closer to the polar surface than to the equatorial surface. But when the spindle elongates, the cell changes its shape to oblate or cylindrical, bringing the equatorial surface closer to the asters. This is exactly the same condition as Rappaport’s experi- ment making cell cylindrical to make the asters and cell surface come to proximity in order to enhance interaction between them [15]. (3) The force of cleavage of sea urchin egg has been measured [18] and the figures are around 1.5 39-5 x10~? dyne. As long as the furrow keeps on constricting with this force, the efficiency of pin- ching will become greater as the diameter of the ring decreases. Stating it in the reverse way, at the beginning of constriction when the diameter of the ring is maximal, the efficiency of pinching is minimal. The astral mechanism comes into play at this moment as if it helps the constricting mecha- nism to start more easily. As a matter of fact, according to analysis of cleavage force in our other papers [19, 20], the time when the largest force is required is the initial stage of division when the egg goes off the spherical shape to an oblate form and the rest of the division process goes downhill forcewise. (4) Since an illustration in the paper of Yoneda and Dan [20] is very elucidating for the situation under discussion, it is partly reproduced here (Fig. Fic. 10. An egg of Temnopleurus hardwicki compressed between two parallel glass plates with the spindle oriented vertically. Force of 2.7 10~* dynes is applied continuously from above. The egg elongates against the force till the 3rd frame and then slackens. (reproduced by courtesy of J. exp. Biol.). 516 K. DAN 140 - Control eggs é ® 120+ 2 ee a 3 100 ~ b f Stalk width (°%) 80 - 90 70 SO 30 ' Lo) es Vipe Tae a t Compressed eggs 10 ' Corrected time (min) Fic. 11. Similar measurements as Fig. 10 for eggs of Pseudocentrotus with and without the weight. (by permission of J. exp. Biol.). 10). In this experiment, eggs of Temnopleurus hardwicki at about the time to divide are placed between two horizontal glass plates with the spin- dle oriented vertically. The lower glass plate is fixed unmovable while the upper plate is movable and can press on the egg. Force of 2.7x10~3 dynes is applied continuously throughout the observation. It is clear that the egg elongates against the weight of 2.7 10~* dynes until the third picture and then it slackens. The same situation is found in the eggs of Pseudocentrotus depressus (Fig. 11). In the light of the astral mechanism, this can be interpreted as showing that until the third photo- graph, cleavage furrow is being formed by the spindle elongation, namely astral mechanism and from the 4th on, the constricting ring begins to work. As a matter of fact, by comparing 3rd and 4th pictures of Figure 10, in the former, the blasto- mere surface is smooth and the tip of the furrow is roundish, while in the latter, the cell surface is beginning to wrinkle and the furrow tip is becom- ing pointed, showing it is being dragged by the constricting ring. Comparing Figures 6 and 7 on one hand and Figure 10 on the other, a point of departure from the astral to constricting mechanism looks to have occurred somewhat later in the latter case than in the former examples. This may be because the imposed pressure may have delayed the time of shifting from the astral to constricting mechanism in the latter case. DISCUSSION In the foregoing description, a point of interest is that the author’s astral mechanism just fits into the complicated period. As the result, Schroeder does not have to worry about early appearance of the ring. Furthermore, although the ring is an ideal tool for pinching, the ring itself has no information as to in which direction it should be formed. The mitotic apparatus not only gives directional information by induction but also gives an impetus to the assembly of actin fibers. From this stage on, the ring carries out the rest of the cleavage process. In other words, the mitotic apparatus is a planner and the ring is an executer, so to speak. At any rate, the author thinks that both mechanisms are involved in cleavage process and a baton is handed over from astral to constrict- ing mechanisms in early stage of the process. REFERENCES 1 Dan, K. (1943) Behavior of the cell surface during cleavage. VI. On the mechanism of cell division. J. Fac. Sci. Tokyo Imp. Univ., Sec. 4, 6: 323-368. 10 Cleavage Mechanism of Sea Urchin Egg Swann, M. M. and Mitchison J. M. (1953) Cleavage in sea-urchin eggs in colchicine. J. exp. Viol., 30: 506-514. Hiramoto, Y. (1956) Cell division without mitotic apparatus in sea urchin eggs. Exp. Cell Res., 11: 650-656. Schroeder, T. E. (1972) The constricting ring II. Determining its brief existence, volumetric changes and vital role in cleaving Arbacia eggs. J. Cell Biol., 53: 419-454. Schroeder, T. E. (1975) Actin in dividing cells; contractile ring filaments bind heavy meromyosin. Proc. Natl. Acad. Sci. USA, 70: 1688-1692. Hiramoto, Y. (1965) Further studies on cell divi- stion without mitotic apparatus in sea urchin eggs. J. Cell Biol., 25: 161-168. Dan, K., Yanagita,T. and Sugiyama, M. (1937) Behavior of the cell surface during cleavage. I. Protoplasma, 28: 66-81. Rappaport, R. (1975) Establishment and organiza- tion of the cleavage mechanism. In “Molecule and Cell Movement”. Ed. by S. Inoué and R. E. Stephan, Raven Press, N. Y., pp. 287-304. Usui, N. and Yoneda, M. (1982) Ultrastructural basis of the tension increase in sea-urchin eggs prior to cytokinesis. Dev. Growth Differ., 24: 453-465. Yonemura, S. and Kinoshita, S. (1986) Actin fila- ment organization in the sand dollar egg cortex. Dev. Biol., 115: 171-183. 11 14 16 17 18 19 20 517 Vacquier, V. D. (1975) The isolation of intact cor- tical granules from sea urchin eggs: calcium ions trigger granule discharge. Dev. Biol., 43: 62-74. Rappaport, R. (1981) Cytokinesis: cleavage furrow establishment in cylindrical sand dollar eggs. J. exp. Zool., 217: 365-375. Hamaguchi, Y. (1975) Microinjection of colchicine into sea urchin eggs. Dev. Growth Differ., 17: 111- ibe Rappaport, R. (1985) Repeated furrow formation from a single mitotic apparatus in cylindrical sand dollar eggs. J. exp. Zool., 234: 167-171. Rappaport, R. (1986) Mitotic apparatus-surface in- teraction and cell division. J. Inverteb. Reprod. Dev., 9: 263-277. Rappaport, R. (1965) Geometrical relation of cleavage stimulus in invertebrate egg. J. Theoret. Biol., 9: 51-66. Hiramoto, Y. (1971) Analysis of cleavage stimulus by means of a micromanipulation of sea urchin eggs. Exp. Cell Res., 68: 291-298. Rappaport, R. (1977) Tensiometric studies of cyto- kinesis in cleaving sand dollar eggs. J. exp. Zool., 201, Suppl. 3: 375-378. Dan, K. (1963) Force of cleavage of the dividing sea urchin egg. Symp. Int. Soc. Cell Biol., 2: 261-276. Yoneda, M. and Dan, K. (1972) Tension at the surface of the dividing sea-urchin egg. J. exp. Biol., 57: 575-587. ZOOLOGICAL SCIENCE 5: 519-527 (1988) Mitotic Poles in Artificial Parthenogenesis: A Letter to Katsuma Dan DANIEL MAZIA Hopkins Marine Station of Stanford University, Pacific Grove, CA 93950, U.S.A. ABSTRACT—The two steps in the classical procedure for artificial parthenogenesis in the sea urchin egg are analyzed. Step I (typically a treatment with butyric acid) can be considered to be the turning-on of the cell cycle, as in normal fertilization. Early events such as Ca*t release and elevation of intracellular pH lead into chromosome replication and a mitotic cycle. However, NonpoLAR MA are produced. The basic structure of the NONPOLAR MA is like that of a normal MA, but it can not make poles; chromosomes can not orient and can not separate and the egg can not divide. Step II (typified by a treatment with hypertonic sea water) results in the production of a great variety of MA, mostly abnormal. MonopoLar MA, which orient and engage chromosomes, are common, but the half-spindles often fail to engage all the chromosomes. There are various MA which seem to be partially POLAR and partially NONPOLAR. BIPOLAR MA may be formed, but often these do not engage all the chromosomes. There are great variations of the poles: some make asters, some do not; some are compact, some are so flat that the chromosomes diverge in anaphase. If the eggs divide at all, the blastomeres are likely to be aneuploid. The fact that truly normal parthenogenetic development is rare is explained by the likelihood of aneuploidy. The results are best interpreted in the light of Boveri’s hypothesis that the maternal centrosome becomes disordered (but is not destroyed) in the maturation of the egg; it is restored to its active form by the parthenogenetic procedure. An explanation can be found in a linear model of the structure of the centrosome (Mazia, D. (1987) Int. Rev. Cytol., 100: 49-91). The linear model would allow for the denaturation of the centrosome in the unfertilized egg and its renaturation by the parthenogenetic procedures. The high incidence of abnormal poles would follow from the probability of errors in the refolding of the very long linear structure. © 1988 Zoological Society of Japan Dear Kady: Our tribute is more than praise; it is thanks. My thanks go back very far — some 55 years. You were a graduate student, I was an ignorant under- graduate who had been accepted by our teacher, Lewis Victor Heilbrunn, into his laboratory fami- ly. I learned so much from you then and have learned much from you since; I learned a lot from your most recent paper. I want to tell you about some work that I have been doing over the last 10 years or so. Let me claim the right of old friendship to say what I have to say in the form of a letter to you. Ina letter between friends, one may look more deeply into the past and imagine further Accepted March 17, 1988 into the future than is welcome in a normal scien- tific publication. The topic is artificial parthenogenesis, especially in the sea urchin egg. The letter has some of the elements of a review, but a personal letter does not require the citation of the larger literature. (The field began to languish so long ago that Albert Tyler’s publication in 1941 [1] covers a major part of the literature). In this letter I also report original but hitherto-unpublished results of experi- ments that I have been doing over more than a decade. Mostly I will be trying to account for artificial parthenogenesis in the light of present views about the nature of the centrosome. Others are free to read and cite this letter, although artificial parthenogenesis does not now interest many biologists. It was exciting and even 520 D. MaZzia sensational before our time because the genesis of fatherless sea urchins by chemical means (“Che- mische Entwicklungserregung”, as Loeb [2] called it) defied the essential idea of fertilization. After that excitement was over, the intrinsic problem —the formation of a bipolar mitotic apparatus in an unfertilized egg—remained unsolved. The problem was hard to think about at a time when there was so much scepticism about the very “reality” of the mitotic apparatus and when cell biologists doubted the universal ocurrence of cen- trosome. An historian can trace the main ideas about artificial parthenogenesis back to two great cell biologists, Jacques Loeb and Theodor Boveri. (The two have been the subjects of interesting biographies [3, 4]). To say that they disagreed would be an understatement. Each simply dismis- sed the other’s idea of the very nature of the problem. The outcome, viewed some 80 years later, was that each identified one part of the problem correctly but saw that part as the whole problem. Loeb’s central thesis [2] was that the egg could be excited by changes of ions; thereby he minimized the role of the spermatozoon in ferti- lization. Boveri [5], who had discovered that the spermatozoon contributes the centrosomes which then provide the poles of the mitotic apparatus, responded to the challenge of parthenogenesis by hypothesis that: the parthenogenetic procedures resurrect the maternal centrosome, which had become disordered during the maturation of the egg. Considering artificial parthenogenesis in the sea urchin egg as a two-step procedure I will consider how Loeb’s point of view explains step I (acti- vation by butyric acid or other means) and how Boveri’s thoughts define Step I (restoration of mitotic poles by a treatment with hypertonic sea water). When Jacques Loeb came here to Pacific Grove, to work in his little laboratory on Monterey Bay, not far from where I am now writing (in the Loeb Building of the Hopkins Marine Station), he was prepared by previous experience to study artificial parthenogenesis as a one step process. He ex- pected to arouse development in unfertilized eggs of Strongylocentrotus purpuratus by a one-step treatment with hypertonic sea water. The results were unsatisfactory, stimulating him to the chain of ingenious experiments that led to the two-step method. The two steps of the procedure that was desig- nated as the “improved method” are: Step I, a brief exposure to a dilute solution of butyric acid in sea water, after which the eggs were returned to sea water, and Step II, an interval of exposure to hypertonic sea water, at about 1.5 the osmolar- ity of normal sea water. Many other agents can be substituted for butyric acid in Step I. My present favorite is CO>-sea water, artificial sea water made up in ordinary carbonated water. (Loeb had used CO; I can only guess why he preferred butyric acid). Different solutes could be used to raise the osmotic pressure in Step II. The method could not be completely standardized with respect to the timing of Step I and Step II; it called for trial-error sampling of the time of starting and the duration of exposure. Nowadays we can understand Step I reasonably well. The broadest statement is that Step I turns on the Cell Cycle, which is repressed in unfertilized sea urchin eggs. Observations of the outcome of Step I assure us the chain of events is about the same as is found at fertilization: the membrane events that precede Ca**-release; the pH changes that follow the Ca**-release; the phosphorylations of nucleosides and the initiation of DNA replica- tion that follows the pH changes, and the further cell cycle events that lead to the production of an MA. The turning-on of these Cell Cycle events by agents other than sperm is Step I of the parthe- nogenetic procedure. Nowadays, as we know more about the turning-on the Cell Cycle by weak acids, weak bases, IP3, Ca-ionophores, non- electrolytes, proteolytic enzymes, etc., etc., we have to be reminded that we do not yet know how the events are started by a spermatozoon. It is confusing to speak of “parthenogenetic activation”, as some writers do, because Step I does not bring about division and development. (It may do so in kinds of eggs in which both steps can be carried out in a single treatment). It was fortunate that Loeb moved to Pacific Grove, where he was forced to sort out the two quite different components of artificial parthenogenesis. Artificial Parthenogenesis 521 Step I, applied to unfertilized sea urchin eggs, leads to the formation of a NoNpoLAR MA. The NonPOLAR MA is very different from a MONOPOLAR MA. We will define the NonpoLar MA after we have described the MonopoLak MA. That may sound perverse, but the MoNopoLarR MA has taught us what a single true mitotic pole can do. The NonpoLar MA is seen in the living cells as a symmetrical aster; it was seen long ago by earlier workers on parthenogenesis. It can be isolated by methods of isolating MA. The fine structure resembles that of a normal MA, but the microtu- bules radiating from the center do not seem to engage the kinetochore in a normal way [6]. In the Nonpoar MA, the chromosomes go through their cycle of condensation, splitting and recondensa- tion. The split chromosome does not move apart. The chromosomes show no orientation (Fig. 1A). There is no polarity; no spindle or half-spindle can be distinguished from the monaster. Chromo- somes are not separated and the completed cycle produces a single diploid nucleus. The cycle may be repeated, again forming a NonpoLar MA with doubled numbers of chromosomes [7]. In modern language the NonpoLar MA appears to be gener- ated by an MTOC, but that MTOC can not function as a pole. We will try the hypothesis that the NonpoLtaR MA expresses the action of the de- generated (denatured) but still-existing maternal centrosome, the survival of which was postulated by Boveri. Running ahead of my story, I can say that the center of the Nonpotar MA contains centrosomal material, as judged by osmiophilic foci [6] and by staining with anticentrosomal anti- body. (Heide Schatten and Gerald Schatten allow me to cite their unpublished finding). The evidence from all the biochemical and cyto- logical studies tells us that Step I starts up the Cell Cycle in unfertilized eggs. What we have seen of the NonpoLtar MA allows us to pursue the idea that the maternal centrosome persists in the ma- ture unfertilized egg but is incapable of functioning as a pole. In my experience, the definition of a pole be- came clear only from a closer study of the Monopotar MA [8] and from consideration of other modern work on Monopotar MA [9]. Working with sea urchin eggs, we can make large numbers of monopolar MA by the “Mercaptoetha- nol Experiment” [10-13]. That useful stratagem traces back to the days when you and I were so fascinated by the role of thiol groups in the MA. With correct timing of the period during which fertilized eggs are kept in beta-mercaptoethanol, we can persude the eggs to divide into four blasto- meres, each of which receives one centrosomal unit (monovalent) instead of the normal pair (biva- lent) of centrosomal units. In the next cycle, each blastomere makes a MonopoLar MA; literally, there is one pole and there is clearly a half spindle [8]. In the following cycle, we see that the pole of the MonopoLar MA has doubled; now there is a bipolar MA and the blastomeres divide. (The cycles of doubling and division of poles in the normal Cell Cycle are so precise! — and necessarily so for the success of cell reproduction. How can we reconcile the normal reproductive behavior of poles with the opinions that they can arise de novo, opinions that have been so influen- tial in the literature of artificial parthenogenesis!) The close study of the MoNopoLarR MA yields, I believe, a very clear idea of what a true pole can do. Without any interaction with a second pole the MonopoLar MA can (1) form a distinct half spin- dle; (2) gather the chromosomes at the “equato- rial” edge of the half spindle; (3) orient and engage the chromosomes so that one kinetochore of a sister-pair points at the pole and makes connec- tions to the pole; (4) carry out anaphase (or an equivalent change that results in the restoration of an interphase nucleus near the pole). We will see that a MoNopoLaR MA is a likely outcome of Step II of the parthenogenetic proce- dure. In trying to understand Step IJ, it is helpful to realize that some of the literature exaggerates the success of the parthenogenetic methods, at least with sea urchin eggs. As students, we may have received from textbooks the impression that the technique of artificial parthenogenesis is the equivalent of pushing a few buttons: carry out Step I and Step II and watch the nice plutei appear! As teachers, we have had to face disappointed stu- dents who tried the procedures in the laboratory. Perhaps some investigators (and species of eggs) have produced large numbers of normal embryos, $22 D. Mazia but careful observers, in the past [14] and more recently [15, 16] comment on the rarity of normal larvae and on the progressive attrition of the developing population. That is unfortunate for those who want to study orphan sea urchins. The “bad” results turn out to be the most interesting to those of us who are trying to understand mitotic poles — if we make the effort to diagnose the varied pathologies of the imperfect mitotic appa- ratus that we find in many eggs. Boveri’s proposed in 1901 [5] that the maternal centrosome is inactivated during the maturation of the sea urchin egg but remains and is reconstructed during the parthenogenetic procedures. In 1988, I am adding the statement that the mature egg contains a denatured centrosome that can generate a Nonpoar MA after Step I of the parthenogenetic procure. After Steps I and II, the centrosome can be renatured and can carry out the functions of pole. (The terms “denaturation” and “renatura- tion” anticipate the hypothesis that the centrosome is a fundamentally linear structure whose activity depends on its 3-dimensional conformation.) Extensive cytological studies show that the re- naturation of the centrosomes is, most often, im- perfect. The MA display an extravagant variety of abnormalities in the forms of the poles and in the engagement of the chromosomes by the poles. It becomes clear why perfect parthenogenetic de- velopment is so rare. The variety of abnormalities is immense, but a few examples tell the essential story. (1) Many eggs do not make a pole at all; Step II has not worked. One sees only NoNpoLAR MA (Fig. 1A) as they have been described earlier [e.g. 6; 7]. (2) In some cases a seemingly perfect MONopo- LAR MA is formed (Fig. 1C). The mitotic appara- tus as a whole is seen as a single aster. The chromosomes are arranged on a metaphase plate on the boundary of a half-spindle. (3) One sometimes sees MA that are intermedi- ate (Fig. 1B) between the Nonporar (as in Fig. 1A) and the well-developed Monopotar MA (Fig. 1C). The chromosomes are arranged on one sector of the spherical monaster, as though engaged to a very wide half-spindle. (4) Studying many of the cases where Monopo- LAR MA are produced, one often sees MA in which some of the chromosomes are oriented on a plate (rather an arc, as is typical of the monopolar MA) on the virtual “equator” at the margin of a half spindle. Others can lie anywhere around the aster (Fig. 1D). Obviously the renaturation of the cen- trosome has been imperfect. To some degree, the such Monopotar MA retains properties of a NonpoLar MA as judged by the behavior of some of the chromosomes. Such defects are common and they predict the production of aneuploid blasto- meres. (5) The MA may be BIPoLar to some degree. The common defects are not familiar from experi- Fic. 1. The first mitotic phase after parthenogenetic treatment. The MA is a monaster; MTs arise from a single center. A. Nonpotar MA. Chromosomes arranged all around the monaster. B. Indication of formation of MONOPOLAR MA. Chromosomes arranged on side of the monaster, interpreted as a “metaphase” configuration on a wide half-spindle. Intermediate between A and C. C. Monopolar metaphase. D. MonopoLtar MA in which some chromosomes are not engaged on metaphase plate. These figures are fluorescence images of isolated MA stained for DNA with DAPI (4,6-diamidino-2- phenylindole). parthenogenetic treatments. Bar=10 micrometers. Haploid chromosome sets indicate that the eggs were in the first mitotic period after the Artificial Parthenogenesis 523 ence with sick MA in fertilized eggs. (My parthe- nogenetic MA are not sick; they are trying hard to make a good mitotic apparatus and are frustrated by the very improbability of the perfect renatura- tion of a centrosome.) Figure 2 tells just about the whole story. Some of the chromosomes are engaged in a spindle and are going through a plausible anaphase. Other chromosomes are not engaged; they lie outside the spindle, experiencing a NoNpOLAR mitosis. The photograph may or may not convince you of the observed fact that the chromosomes outside the spindle have split, as they should have done at the onset of anaphase. If you have noticed that one pole lacks an aster you are correct, although the evidence is clearer in Figure 3. (6) Even when chromosomes are in a bipolar spindle, sister kinetochores may fail to make the right connections and are not always separated. One finds cases of non disjunction; sister chromo- somes go to the same pole (Fig. 2). (7) In many of the cases where bipolar MA are formed, the shapes of the poles may be strange. One may find no asters, or unequal asters or one pole with no aster. More interesting, in the light of the linear model of the centrosome that I will be proposing, are anaphases (Fig. 3A) or telophases a Pa (Fig. 3B) which indicate that one group of chromo- somes has converged toward a more compact pole while the other has moved on a divergent course, as though moving to a very flat pole. Such an image is incomprehensible if we think of a centro- some as a particle, but is reasonable according to the idea that a centrosome may have many shapes, depending on the folding of a long and linear fundamental structure [17, 18]. If interest in artificial parthenogenesis ever re- vives, the new devotees will have to think about the problem of timing. The first cell divisions in parthenogenetic embryos appear much later than they do in fertilized eggs. As shown in Figures 1 and 2, the timing of the chromosome cycle and of the provision of poles may be out of phase. The monopolar and more-or-less bipolar MA shown in Figures 1 and 2 both belong to the first mitotic cycle; chromosome numbers are haploid. Thus, at the same time on the chromosome clock, the poles have either failed to form (Fig. 1A); barely formed (Fig. 1B); formed as good MoNopoLar (Fig. 1C); formed as a bad Monoporar (Fig. 1D) or it may have started to make two poles, separated but usually pretty poor compared to normal poles. Eggs with good MonopoLar MA may go through another cycle to produce bipolar MA that distrib- oe Fic. 2. Examples of defective bipolar MA in artificial parthenogenesis. Some chromosomes are not engaged in spindle; they have split but do not separate toward poles. Some of chromosomes on spindle appear double, indicating non-disjunction; sister chromosomes are going to same pole. Aceto-orcein squash preparations. Bar= 10 micrometers. 524 D. Mazia Fic. 3. Differences in shapes of poles. One group of chromosomes converges toward a compact pole while the other group diverges as though toward a very wide flat pole. The figures show rather extreme cases of this common occurrence. A. Anaphse. B. Telophase. Aceto-orcein squash preparations. Bar=10 micrometers. ute diploid complements of chromosomes. Then a gap in our information becomes evident. It arises from our habit of going home to dinner after a long day in the laboratory. A great many workers in this field must have had the same bad habit; they are content to report the numbers of “swimming embryos” seen on the next morning. The observations I present above do explain why most of the parthenogenetic embryos develop so poorly. Aneuploidy is very likely. The aneu- ploidy will take its developmental toll progressive- ly, as gene regulation becomes more important and the embryo can no longer depend on stored messages and modification of molecules in the cytoplasm. A few papers in the literature describe the abnormal MA in artificial parthenogenesis. The paper of Wilson [14] and an excellent study by Marianne von Ledebur-Villiger [15] contain draw- ings that illustrate a number of the mitotic abnor- malities I have presented as photographs. Von Ledebur-Villiger also made measurements of DNA content of parthenogenetic blastomeres which indicated the frequent occurrence of aneu- ploidy. My long study, from which I illustrate only a few examples, has been directed toward the varieties of defective poles as possible indications of the process of renaturing centrosomes. The original concept of the centrosome as a “polar corpuscle” would not imply such flexibility. We now see Step II as the restoration of Mitotic PoLes (mostly defective) in unfertilized eggs which would otherwise have produced a NONPOLAR MA. It would seem obvious that the effects of the hypertonic medium would arise from an increase in the internal concentration of ions or other solutes. That interpretation became doubtful with the discovery that Step II can be carried out by replacing normal cell water with D,O. In my experience, D2O does not work well with eggs of Strongylocentrotus purpuratus but sea water made up in DO is especially effective in experiments with Lytechinus pictus. (In a typical experiment, the eggs might be kept in sea water made up in concentrated D,O between the 20th and 40th minute after Step I.) The eggs do not shrink, although cytoplasmic particles may aggregate into coarser clumps. From what I have read, the behavior of electro- lytes does not change so very much when the solvent is D2O. So, one speculates that the effects of hypertonic sea water depend less on the concen- trations of intracellular solutes than they do on changes of the intracellular water. The suspicion Artificial Parthenogenesis would apply to the effects of hypertonic sea water as well as to those of D,O. Hypertonic sea water at 1.5 isosmolar would remove a great deal of the “free” water of the cell. At the very least, macromolecular —and even larger— units would become more crowded and interactive sites would come closer. Speculation about water recalls the fate of theo- rists who, over the years, have indulged in special “cell water” (bound water, structured water) that is different from ordinary water. Cell water is an intoxicating beverage. At the moment, it is re- spectable to think about a role of more-structured water in the cell, thanks to the introduction of modern methods such as NMR. The time has come when colleagues are beginning to study changes of properties of water during the mitotic cycle in sea urchin eggs [20]. For my part, I do predict that the configuration of mitotic poles will become more explicable when we know more about cell water. That is not a lot of progress from Boveri’s [5] comment (freely translated) that “...the transfer into the Loeb medium brings the egg protoplasm into a condition under which the egg centrosome can function. /f this explanation is correct then the effect of (hypertonic medium) has nothing in com- mon with that of the spermatozoon, despite the same end-result”. The controversies about artificial parthenogene- sis have played a large part in the evolution of our ideas about centrosomes. There can not be many questions more important than: is the centrosome a universal, permanent, reproducing organ of the eucaryotic cell; or is it an agent that can be generated de novo? A _ profound question, be- cause the permanent, reproducing organ accounts logically for the fact that cells divide into two, generation after generation. The centrosome was very much in decline at the time when we were students. The cytologists had failed to find it in many cells. The defect was not in their eyes, but in their erroneous preconception. They were looking for a compact particle. The centrosome is not a particle. The closer study of artificial parthenogenesis directs us to a model of the centrosome. Recipro- cally, the model helps us to understand the hap- Nn to nN penings of parthenogenesis. The model [17, 18] calls for a linear centrosome, very long in its extended form, more comparable in length to a chromosome than to a macro- molecule. The thread (or ribbon) bears micro- tubule-initiating units; each unit determines the origin and the direction of a microtubule. The very long centrosome can fold in definite conforma- tions; these conformations are expressed in the shapes of the microtubular structures they gener- ate. The shapes of the centrosome change during the normal mitotic cycle [5, 17, 18]; correspond- ingly the shapes of poles change. The compact centrosomes at metaphase in sea urchin eggs generate pointed spindles, while the flattened cen- trosomes at late anaphase explain the more barrel- shaped spindles. The unusual shapes of poles that have been described above (Fig. 3A, 3B) would also be examples. Here I am citing facts, based on the resolution of centrosomes with the aid of anticentrosome antibodies as well as other criteria such as osmiophilia [11, 23]. Nowadays, anticentrosome antibodies are reas- suring us that we can expect to find centrosome at mitotic poles in all kinds of cells and that the centrosomes are not necessarily compact particles and can change their shapes. The model helps to explain some facts—and deficiencies—in the interpretation of artificial parthenogenesis as the renaturation of centro- somes. (1) A deficiency—there is still no direct evi- dence for centrosomes in the mature unfertilized sea urchin egg. If a denatured centrosome is present, it may be present as a very much un- raveled thread, too fine to have been resolved by the immunofluorescent staining reported so far. (2) When a NonpoLar MA results from the Step I treatment, it does seem to be organized by a centrosome. The centrosome is compact at this stage. It was identified originally by osmiophilic foci [6]. I am allowed to say that it can be identified with the help of anticentrosome anti- bodies (personal communication, H. and J. Schat- ten). It can still be considered to be a denatured centrosome; the postulated extended thread has collapsed into a randomly folded compact body that can not form a pole. 526 D. MaZzia (3) Step II provides conditions for renaturation, but a perfect refolding of the extended thread will be rare. By the stage at which the cell cycle commands the formation of an MA, some regions of the centrosome may have the necessary con- figuration for MONOPOLAR behavior (generation of some sort of half-spindle) while other parts of the very large structure still can only function in the NONpPOLAR mode. The problem posed by the cytasters seem less formidable in the light of increasing experience. For the above description of the variants of the parthenogenetic MA, I used experiments in which there were no cytasters. In fact some of the MA observed had no asters at all—or asters could be unequal or poles could be very flat. The view that cytasters (generated de novo) are taking over the roles of mitotic poles does not describe what we actually see. In many cases where cytasters were present, they do not participate as poles. If mitotic poles came from cytasters, we would expect to find nice MONopOLAR, BIPOLAR or MULTIPOLAR MA rather than the zoo of abnormal MA that is actually encountered. There is no need to take your time with various explanations of cytasters. Cytasters may defy logic but they refuse to go away (I do not want to risk the seductions of Lawyer-Science in speaking to you who are a model of Scientist-Science.) The idea of the linear centrosome invites experi- ments that test the ideas of its linearity, its native and denatured states and its expression as a mitotic pole. Indeed, we now have evidence that the conformation is sensitive to low temperatures [24]. In the experiments, the forms of the centrosomes were observed by staining with anticentrosome antibody. Fertilized eggs at a prometaphase stage were stored at 0°C. The centrosomes collapsed in a single compact mass. After return to normal temperature, the centrosome material unwound quickly as an irregular thread then reassociated to form a bipolar spindle with flat poles. The observation gives comfort to the hypothesis of the linear centrosome and some assurance that ideas about the unfolding and refolding of the centro- some thread can be defended. As I reflect on centrosomes and mitotic poles and the fabrication of mitotic apparatus, I think of your wonderful new work on unequal division, and memory goes back to 1952 and the first isolation of MA with unequal asters from clam eggs [25]. Your opinion that unequal asters can explain unequal division was a revelation to me then. Surely an unequal division that invokes an MA with asters of different size implies an MA whose poles are different. Therefore the centrosomes are differ- ent. What would you expect from experiments that look into differences in centrosomes at two poles of an MA? Can you imagine that differentia- tion seen in unequal divisions in cell lineages might be an expression of differentiation of centrosomes, manifested as differences of asters at the two poles of a decisive division? If I were a better writer, I would know how to express every sentence of this letter as a question to you—?—., not as a statement... Then, I could be learning. We, your friends and your students and their students have been learning from you for half a century and it is we who are honored by this opportunity to thank you. Your friend, Daniel Mazia ACKNOWLEDGMENT I express my gratitude to Dr. Colin Pittendrigh, former Director of the Hopkins Marine Station, for calling me to Pacific Grove. It will be evident that much of the substance of this contribution comes out of collabora- tions with Dr. Neidhard Paweletz of the Deutsches Krebsforschungszentrum, Heidelberg and with Dr. Heide Schatten and Dr. Gerald Schatten, of the Uni- versity of Wisconsin. The National Science Foundation to support my work through Grant DCB-8401901. REFERENCES 1 Tyler, A. (1941) Biol. Rev. Cambridge Philos. Soc., 16: 291-335. 2 Loeb, J. (1909) Die chemische Entwicklungser- regung des tierischen Eies. (Kuenstliche Parthe- nogenese). Berlin, Julius Springer Verlag. 3 Baltzer, F. (1967) Theodor Boveri; Life and Work. (English translation by J. Oppenheimer). Univ. California Press, Berkeley. 4 Pauly, P. J. (1987) Controlling Life. Jacques Loeb 11 12 Artificial Parthenogenesis and the Engineering Ideal in Biology. Oxford Univ. Press, New York. Boveri, T. (1901) Zellen-studien IV. Ueber die Natur der Centrosomen. Fischer Verlag, Jena. Paweletz, N. and Mazia, D. (1979) Eur. J. Cell Biol., 20: 37-44. Mazia, D. (1974) Proc. Natl. Acad. Sci., 72: 690- 693. Mazia, D., Paweletz, N., Sluder, G. and Finze, E.- M. (1981) Proc. Natl. Acad. Sci., 78: 337-381. Bajer, A.S. and Mole-Bajer, J. (1979) In “Cell Motility: Macromolecules and Organelles”. Ed. by S. Hatano, H. Ishikawa and H. Sato, Univ. Tokyo Press, Tokyo, pp. 569-592. Mazia, D., Harris, P. J. and Bibring, T. (1960) J. Biophys. Biochem. Cytol., 7: 1-20. Paweletz, N., Mazia,D. and Finze, E.-M. (1984) Exp. Cell Res., 152: 47-65. Sluder, G. (1978) In “Cell Reproduction”. Ed. by E. R. Dirksen, D. M. Prescott and C. F. Fox, Academic Press, New York, pp. 563-570. Sluder, G. and Rieder, C. (1985). J. Cell Biol., 100: 887-896. 14 15 527 Wilson, E. B. (1901) Roux Arch. Entwicklungs- -mech., 12: 529-589. von Ledebur-Villiger, M. (1972) Exp. Cell Res., 72: 285-308. Brandriff, B., Hinegardner, R. T. and Steinhardt, R. (1975) J. Exp. Zool., 192: 13-24. Mazia, D. (1984) Exp. Cell Res., 153: 1-15. Mazia, D. (1987) Int. Rev. Cytol., 100: 49-92. Wilson, E. B. (1928) The Cell in Development and Heredity, 3rd. ed. Macmillan, New York. Schatten, H., Schatten, G, Mazia, D., Balczon, R. and Simerly, C. (1986) Proc. Natl. Acad. Sci., 83: 105-109. Mazia, D. and Dan, K. (1952) Proc. Natl. Acad. Sci., 38: 826-838. Mazia, D. (1975) Ann. N.Y. Acad. Sci., 253: 7-13. Endo, S. (1980) Dev. Growth Differ., 22: 509-516. Mazia, D., Schatten,H., Coffe,G., Szoeke, E., Howard, C. and Schatten, G. (1987) J. Cell. Biol., 105: no.4, part 2: 206a. Dan, K., Ito, S. and Mazia, D. (1952) Biol. Bull., 103: 292. ZOOLOGICAL SCIENCE 5: 529-538 (1988) © 1988 Zoological Society of Japan The Living Spindle SHINYA INOUE Marine Biological Laboratory, Woods Hole, Massachusetts 02543, USA ABSTRACT— Experience relating to earlier studies on the birefringence of the mitotic spindle and spindle fibers in living cells is first reviewed. This is followed by a description of the dynamic arrangement of micotubules in spindle fibers in living cells studied recently with high resolution video polarized light microscopy. KD’S LAB The urchin eggs had been collected, washed gently in a hand centrifuge, and mixed just minutes ago with a dilute sperm suspension. “What’s the rate of fertilization?” KD (Dan-san, Dan-sensei, Professor Katsuma Dan) would ask. “75%? Bet- ter try another female,” he’d say. “You'd just be kidding yourself if you worked with cells that aren’t completely healthy.” That was the way KD lived his science, and the way he trained the upstarts working in his laboratory. The second floor laboratory at the Misaki Marine Biological Station, overlooking the waters of Aburatsubo and Moroiso Bay, was where so many of us were exposed to KD’s challenges, to his way of biology, and to the enchantment of science. “The goddess of science is terribly jealous,” he’d utter when he’d feel that we were not fully engag- ing our wits in carrying out an experiment. Or, he might even shout, “Try putting yourself in the animals’ shoes!” when we’d missed accounting for a parameter that critically affected the organism or embryo. Most of the time these criticisms seemed born more out of exasperation rather than as tongue lashing from a hard task master. Even those who felt cowed or rebellious at these words were work- ing in KD’s lab because that was simply what they wanted to do. They were neither paid assistants nor students who were required to be there. They were a school teacher, a brash graduate student, a junior colleague, or one just fascinated by biology; Accepted April 4, 1988 anyone who cared enough to be there and who dared take on the challenge. That was the makeup of those who frequented KD’s and Jean’s labora- tory in the early post-war days (ca. 1946-1949). To Jean and KD who had survived the double hardships of raising a young family and keeping up research under the challenging conditions of the bombed-out nation, while coping with the complex adjustments of a couple whose countries were in mortal conflict, those post-war days brought forth incalculable strength, joy, and optimism. Their mood pervaded the Laboratory just returned to the Japanese biologists and students by the occupation forces (with KD’s intervention: “The last one to go,” now prominently featured at the MBL Library in Woods Hole). KD’s cramped lab, saturated with the scent of marine organisms, bustled with activity oblivious to, and actually challenged by, the absence of all but a few essential tools for research. Some mornings the marine scent would be taken over by the delectable aroma of buttered toast smeared with honey that Jean and KD would share with us. Jean would have rounded up the butter, and the honey had been spun from KD’s own hives in Nagai. In the lab, Kayo was raising miniature adult sand dollars from the four blastomeres she’d cut apart free hand, KD and Jean were busy writing up their papers on the role of spindle elongation in astral cleavage, Endo-san had given up his hard- ware business to explore fertilization, and I was tinkering with pieces of optics trying to visualize spindles in living sea urchin eggs following Runn- str6m’s (in fixed plant cells [1]) and Schmidt’s [2] 530 S. INOUE earlier observations. SPINDLE BIREFRINGENCE With a mercury arc lamp and a pair of calcite prisms loaned us by Professor Koana rigged to the microscope, and with the window shades drawn tught, we finally caught glimpses of the football- shaped birefringence just before each division of the transparent Clypeaster and Spirocodon eggs. Those were the spindles that we had attempted to see at KD and Jean’s home in Kudan during an airraid blackout (ca.1942). There was no question that we could detect the spindle birefringence now, that was, until I read- justed the microscope (while observing the back aperture of the objective lens to maximize the extinction). What I’d believed would give us a much better image had made the spindle birefrin- gence disappear altogether. “See, I told you to leave well enough alone,” was KD’s admonish- ment. Red-faced, I continued to tinker for another month before it dawned on me that the strain birefringence in the lens was not hindering us, but was in fact helping us visualize the weakly birefrin- gent spindle. The lens was acting as a compensator and raising the contrast and field brightness! Through such a hard lesson, how could one ever forget how helpful the bias retardation is [3]. As I learned then (and Michael Swann and Murdoch Mitchison were also finding out in Edinburgh half way around the globe [4]), a low retardation compensator not only allows us to measure the birefringence, but is indeed an_ indispensable friend for the biologist who is trying to visualize the orderly but dynamic alignment of just a few molecules in the living cell (Fig. 1). In retrospect, W.J. Schmidt in Giessen had figured this out a decade before us, but my poor performance in the three year German classes had ill-equipped me to grasp the important teachings in Schmidt’s two classic volumes [2, 5]. JEAN’S HOMECOMING From Jean who had just visited her homeland, there was doubly good news. Asa present for KD, she had secured (with support from the American if Fic. 1. Birefringence of dividing egg of a jellyfish Spi- rocodon saltratrix. The mitotic spindle and astral rays exhibit a longitudinally positive birefringence while the cell surface exhibits a tangentially negative birefringence. From Inoué and Dan [3]. Philosophical Society) one of the first phase con- trast microscopes produced by Bausch and Lomb in Rochester, New York. She had also arranged that I apply to graduate school at Princeton, with a loan from her sister Peggy Chittick of Bridgeport, Living Spindle Fic. 2. Fic. 3. Birefringence of spindle fibers and fibrils in Chaetopterus oocyte. From Inoué [8]. Detailed structure of Chaetopterus spindle (Fig. 2) as I interpreted in 1951 [8]. 532 S. INOUE Connecticut, to help finance the trip. The phase microscope, as documented in KD’s autobiography [6], led Jean back into the lime light of biology and to her discovery of the sperm acrosomal reaction [7]. I myself in the meantime had sped to Princeton, and to Woods Hole about which we had heard so much from KD and Jean. SPINDLE FIBERS With guidance and encouragement from my mentor Kenneth Cooper at Princeton and with help from the Osterhouts in Woods Hole, using a moderately improved polarizing microscope I was able to settle a 50-year controversy regarding the reality of spindle fibers and fibrils in intact living cells (Figs. 2-4, [8]). So entrenched was the idea that spindle fibers were invisible in living cells that E. B. Harvey, upon first seeing my time-lapse movies in the Lillie Auditorium asked, “Were Fic. 4. spindle fiber [8]. those cells alive?” She could not accept the fact that the dividing Chaetopterus and Lilium cells which distinctly showed the spindle fibers (owing to their weak but distinctly higher birefringence) were in fact alive! Albert Tyler from Cal. Tach. was also at the MBL in Woods Hole those summers (1949-1950). At his suggestion, I looked into the effects of colchicine on spindle birefringence. In dilute col- chicine protected from blue light, the birefringence of the metaphase-arrested Chaetopterus spindle faded away in just a few minutes. The chromo- somal spindle fibers were the last to go. As they were disappearing, the fibers would shorten (with- out fattening) and pull the chromosomes and inner centrosome to the animal pole surface where the outer spindle pole was anchored [8, 9]. When colchicine was washed out, the spindle and its birefringence grew back. This experience with colchicine, and later with | 10.um Spindle fiber birefringence. Early anaphase in pollen mother cell of Easter lily. Chr: chromosome, spf: Living Spindle 533 low temperature and high hydrostatic pressure, gave birth to the notion that spindle and astral fibers and their fibrils (later shown to be micro- tubules) could dynamically assemble from pre- formed subunits and grow, and that they could disassemble and shorten [10-12]. In other words, the spindle fibers and fibrils could polymerize and push, or depolymerize and pull [13]. ISOLATED SPINDLE A few years after my early MBL days, KD was visiting Dan Mazia who had recently been appointed to the faculty at the University of Cali- fornia at Berkeley. Dan was a long standing friend of both KD’s and Jean’s from their Pennsylvania and Woods Hole days in the early 1930s. In the basement lab of the Medical School in Seattle where I was an Instructor at the University of Washington, I got an excited call from Dan who told me that he and KD had just figured out to isolate the mitotic apparatus [14]. Could he fly up with some samples to see if the isolates were birefringent? Next day Dan was bubbling with excitement at the reassuring sight under my polar- izing scope, now significantly refined mechanically thanks to inputs from my colleague Wayne Thorn- burg and chairman Stan Bennett. After establishing the positive birefringence of the isolated mitotic apparatus, Dan and I decided that the isolate’s response to colchicine would reveal how native the isolates were. But, as the colchicine solution that we perfused had just reached the isolates under the slide, the micro- scope field and the room went completely dark. Fortunately we had already photographed several isolates before the power went out, but we had no clear answer regarding the sensitivity of the iso- lates to colchicine. The inability of colchicine to affect the isolates (event those isolated according to more recent, gentler methods that yield spindles which de- polymerize in the cold) still remains an enigma today. Once the conditions required to maintain colchicine sensitivity in the isloates are found, we will presumably also have found the conditions needed to reliably study anaphase movement in the isolates. Unlike striated muscle and cilia, the mitotic structures have tenaciously refused to re- veal their force-generating machinery. VIDEO ENHANCEMENT Many years later, KD and Kayo were visiting my laboratory at the MBL in Woods Hole. The scope was now equipped with a rectifier with a rectifier for DIC as well as for pol, and we could directly image the spindle in the somewhat opaque Spisula egg thanks to the contrast enhanced by video. On his previous visit to the States, KD had worked together with Sus Ito from Harvard [15]. The two had isolated Spisula spindle with asym- metric asters and related these to their role in asymmetric cleavage. video-enhanced polarized light microscopy, we Now with time-lapsed, could directly follow the growth, migration, and striking oscillation of the spindle seeking its polar anchorage point [9, 16] The dynamic behavior of the spindle, centro- somes, and asters, presumably reflecting the life- like extension and shortening of their microtu- bules, continues to amaze and challenge investiga- tors (e.g., see [17, 18]). And as we improved the imaging capability of the microscope, the dynamic behavior of each individual microtubule became ever more evident. In the next section of this paper, I shall describe some observations that Ted Salmon, Lynn Cas- simeris, and I made on dividing newt lung epithe- lial cells, using a recent version of the high extinc- tion polarizing microscope. Despite KD’s admoni- tion, I had not, since those days at the Misaki Labs, been able to leave the microscope alone. MICROTUBULE DYNAMICS IN SPINDLE FIBERS The following observations were made with the high extinction polarizing microscope recently equipped with a fiber optic light scrambler [19; 20, Figs. III-11,12].. The scrambler provides a high intensity, uniform illumination that homogeneous- ly fills the back aperture of the N.A. 1.35 rectified condenser (made by Nikon for Plan Apo objective lenses). Illuminated with this condenser, the video-enhanced image, formed with the new 534 S. INOUE (1987) series N.A. 1.4 Plan Apo objective lenses, gives a depth of field as shallow as 0.1 ~m in polarized light microscopy [21]. The newly available shallow depth of field (i.e., outstanding optical sectioning capability and very high axial resolution) was now coupled with the inherently high lateral resolving power and the further improved corrections just incorporated by Nikon and Zeiss in the new series 1.4 N.A. Plan Apochromatic objective lenses. When the contrast of the faint polarized light image produced by this high resolution system was enhanced with video (analog output of Dage/MTI model 65 Newvicon camera digitally processed, in continuous background subtraction and 4-frame jumping average mode, with Universal Imaging’s Image-1/AT [20, 22]), the behavior of individual microtubules and their bundles could be clearly captured in the live, newt (Taricha granulosa) lung epithelial cells (Figs. 5-8). In early prometaphase of the Taricha cells, we find a discrete, positively birefringent thread, a- bout 1 4m in diameter, attached to the kinetochore of each chromosome. According to Rieder’s high voltage EM studies, the thread is a bundle of a dozen or so, tightly-packed, parallel microtubules [25]. Polewards, beyond the birefringent “bundle” or “cable”, the microtubules splay apart from each other, then mix with polar microtubules as well as other kinetochore microtubules (Figs.5 and 6). Each cable alternately grows and shortens with the progression of prometaphase, but on the whole they gradually shorten so that by the onset of anaphase few cables are present. Throughout prometaphase to anaphase, short, very weakly birefringent “rods” appear and dis- appear stochastically throughout the spindle. The rods are regions of microtubules, including the kinetochore microtubules splayed poleward from the bundle, which transiently associate with other microtubules in parallel pairs over a few micron lengths. The transient lateral associations each last Fic. 5. Prometaphase in newt lung epithelial cell [24]. high-resolution polarized light image. See Fig. 6 for interpretation of this video-enhanced, Living Spindle 535 Fic. 6. Schematic of microtubule distribution and behavior as interpreted from dynamic playback of laser disk recordings of the video-enhanced, high-resolution polarized light images such as those in Figs. 5, 7, and 8 [23]. Note the remarkable similarity of this schematic to those by Nicklas er a/. derived by another approach, i.e., serial thin section electron microscopy [27]. Fic. 7. Early anaphase of cell shown in Fig. 5 [23]. The birefringent cables are mostly gone and replaced by many short-lived birefringent rods. (See Fig. 6 for definition of cable and rod.) 536 S. INOUE Fic. 8. Telophase of the same cell in Figs. 5 and 7. In such very thin optical section, the density of microtubules is low enough so that individual microtubules are clearly imaged (arrow). ({23]; see [22, 24] on microtubule distribution in Haemanthus endosperm, observed in fixed cells stained with immuno-gold antitubulin, also with digitally enhanced high resolution polarized light microscopy). for only a few seconds, but on the whole they form what could be considered a fringed micellar struc- ture (Figs. 6 and 7). According to these observations, what was called a chromosomal spindle fiber in the classical literature (e.g., see Schrader [26]), or that we observed with lower resolution in polarized light microscopy [8, 10, 13], turns out to be a mixture of kinetochore and non-kinetochore microtubules. The different microtubules in the chromosomal spindle fiber stochastically associate laterally with each other and presumably form an anastomosing, fringed micellar gel. We believe that the fringed micelle structure, formed by the short-lived, dynamic lateral associa- tion between microtubules, explains the modest mechanical integrity of the whole spindle and of the chromosomal and other spindle fibers. It also accounts for their lability when repetitively teased with a glass microneedle [27,28]. The dynamic fluctuation of microtubules due to transient lateral association (or zipping as the Bajers may prefer to call it), presumably together with the dynamic instability of the microtubules (see below), also accounts for the Northern lights flickering of spindle fiber birefringence that is seen at lower resolution [10]. AND NOW? Given the dynamic lateral association coupled with the dynamic instability of microtubules (wherein each microtubule extends steadily until it suddenly undergoes rapid shortening as postulated by Tim Mitchison and Mark Kirschner [17] and observed under the light microscope by Horio and Hotani [29] and ourselves—see companion paper by Inoué [30]), it is no wonder that microtubules Living Spindle exhibit complex life-like behavior. In addition to the dynamic properties of the fibrillar microtubules themselves, also calcium- regulating vesicles, microtubule-associated pro- teins, and other factors modulate the stability of the microtubules in local parts within a living cell. Together they account for the dynamic, enigmatic behavior of the centrosomes, spindle fibers, and asters, which govern much organizational activity so crucial to the cell’s survival and proper function. In exploring these mysteries, on the one hand, the diversity of pattern expressed by different life forms may provide us with clues, by exaggerating or omitting some elements involved in the complex set of interactions (e.g., see [26, 31-33]). On the other hand, we need to learn much more about the state of the microtubules and modifying factors, not only as isolated purified elements, but as they function dynamically and locally in intact living cells. To this end, direct studies of living cells with the light microscope as a non-invasive, analytical probe should play increasingly important roles, especially when combined with video. Images with high lateral and axial resolution can provide better 3-dimensional insight (Fig. 8), and the video- enhanced contrast can reveal subtler changes in fine structure and molecular organization of living cells with reduced exposure to light [20]. We should also use the microscope more effec- tively to modify minute targeted regions in living cells, not only with UV or other high energy irradiation, but with wavelengths and energy levels selected to induce subtler chemical modifications [34-36]. Combined with a good choice of reagents that are activated or inhibited by those wavelengths, we can better probe for local factors that govern the intricate, dynamic organization of the living cell and its spindle. ACKNOWLEDGMENT This paper is dedicated to Professor Katsuma Dan on his 83rd birthday. I recall with great pleasure the support and challenges provided by KD, which started as early as 1941 when I was still a student at the Musashi Higher School. The recent work described here and preparation of the manuscript were supported by NIH grant R37 GM31617-06 and NSF grant DCB 8518672. NO 10 11 13 14 537 REFERENCES Runnstrom, J. (1929) Uber die Veranderung der Plasmakolloide bei der Entwicklungserregung des Seeigeleies. II]. Protoplasma, 5: 201-310, Fig. & Table 4. Schmidt, W. J. (1937) Die Doppelbrechung von Karyoplasma, Zytoplasma und Metaplasma. Proto- plasma-Monographien, Vol.2. Gebruder Born- traeger, Berlin. Inoué, S. and Dan, K. (1951) Birefringence of the dividing cell. J. Morphol., 89: 423-455. Swann, M.M. (1951) Protoplasmic structure and mitosis. II. The nature and cause of birefringence changes in the sea-urchin egg at anaphase. J. Exp. Biol., 28: 434-444. Schmidt, W. J. (1934) Polarisationsoptische Analy- se des submikroskopischen Baues von Zellen und Geweben. In “Handbuch der biologischen Arbeits- methoden.” Ed. by E. Abderhalden, Urban und Suhwarzenberg, Berlin and Vienna, Sec. 5, Part 10, pp. 435-665. Dan, K. (1987) “Uni toh kataru.” (Dialogue with sea urchin.) Gakkai Shuppan Center, Tokyo. Dan, J.C. (1954) Studies on the acrosome. II. Acrosome reaction in starfish spermatozoa. Biol. Bull., 107: 203-218. Inoué, S. (1953) Polarization optical studies of the mitotic spindle. I. The demonstration of spindle fibers in living cells. Chromosoma,5: 487-500. Lutz, D. A., Hamaguchi, Y. and Inoué, S. (1988) Micromanipulation studies of the asymmetric posi- tioning of the maturation spindle in Chaetopterus sp. oocytes. I. Anchorage of the spindle to the cortex and migration of a displaced spindle. Cell Motility and Cytoskeleton, in press. Inoué, S. (1964) Organization and function of the mitotic spindle. In “Primitive Motile Systems in Cell Biology”. Ed. by R.D. Allen and N. Kamiya, Academic Press, New York, pp. 549-598. Inoué, S. (1981) Cell division and the mitotic spin- dle. J. Cell Biol., 91: 131s—147s. Inoué, S., Fuseler, J., Salmon, E. D. and Ellis, G. W. (1975) Functional organization of mitotic micro- tubules. Physical chemistry of the in vivo equilib- rium system. Biophys. J., 75: 725-744. Inoué, S. and Sato, H. (1967) Cell motility by labile association of molecules. J. Gen. Physiol., 50: 259- 292. Mazia, D. and Dan, K. (1952) The isolation and biochemical characterization of the mitotic appa- ratus of dividing cells. Proc. Natl. Acad. Sci., 38: 826-838. Dan, K. and Ito,S. (1984) Studies of unequal cleavage in molluscs. I. Nuclear behavior and anchorage of spindle pole to the cortex as revealed 16 18 23 24 538 by isolation technique. Dev. Growth Differ., 26: 249-262. Dan, K. and Inoué, S. (1987) Studies of unequal cleavage in molluscs. I]. Asymmetric nature of the two asters. Int. J. Invertebr. Reprod. Dev., 11: 335- 353. Mitchison, T. and Kirschner, M. (1984) Dynamic instability of microtubule growth. Nature, 312: 237- 242. Salmon, E. D., Leslie, R. J., Saxton, W. M., Karow, M. L. and McIntosh, J. R. (1984) Spindle microtubule dynamics in sea urchin embryos: Analy- sis using a fluorescein labelled tubulin and measure- ments of fluorescence redistribution after photo- bleaching. J. Cell Biol., 99: 2165-2174. Ellis, G. W. (1985) Microscope illuminator with fiber optic source integrator. J. Cell Biol., 101: 83a. Inoué, S. (1986) Video Microscopy. Plenum, New York. Inoué, S. (1988) Imaging of unresolved objects, superresolution, and precision of distance measure- ment, with video microscopy. In “Fluorescence Microscopy of Living Cells in Culture: Quantitative Fluorescence Microscopy: Imaging and Spectro- scopy”. Ed. by D.L. Taylor and Y.-L. Wang, Methods in Cell Biology, Vol. 30, Academic Press, New York. In press. Inoué, S. (1987) Video microscopy of living cells and dynamic molecular assemblies. Applied Optics, 26: 3219-3225. Cassimeris, L., Inoué, S. and Salmon, E. D. (1988) Microtubule dynamics in the chromosomal spindle fiber: Analysis by fluorescence and high resolution polarization microscopy. J. Cell Motil. Cyto- skeleton, in press. Inoué, S., Molé-Bajer, J. and Bajer, A. S. (1985) Three-dimensional distribution of microtubules in Haemanthus endosperm cell. In “Microtubules and Microtubule Inhibitors”. Ed. by M. De Brabander and J. De Mey, Elsevier, Amsterdam, pp. 15-30. S. INOUE 25 26 27 28 29 36 Rieder, C. and Bajer, A. S. (1977) Heat-induced reversible hexagonal packing of spindle microtu- bules. J. Cell Biol., 74: 717-725. Schrader, F. (1953) Mitosis. The Movements of Chromosomes in Cell Division. Columbia Univ. Press, New York, 2nd ed. Nicklas, R. B., Kubai, D. F. and Hays, T. S. (1982) Spindle microtubules and their mechanical associa- tions after micromanipulation in anaphase. J. Cell Biol., 95: 91-104. Begg, D. A. and Ellis,G. W. (1979) Micromani- pulation studies of chromosome movement. II. Birefringent chromosomal fibers and the mechanical attachment of chromosomes to the spindle. J. Cell Biol., 82: 542-554. Horio, T. and Hotani, H. (1986) Visualization of the dynamic instability of individual microtubules by dark-field microsocpy. Nature, 321: 605-607. Inoué, S. (1988) Manipulating single microtubules. Protoplasma: Noburo Kamiya Festschrift, in press. Bélar, K. (1926) Der Formwechsel der Protisten- kerne. Ergeb. Fortschr. Zool., 6: 1-420. Inoue, S. and Ritter, H., Jr. (1978) Mitosis in Barbulanympha. 1. Dynamics of a_ two-stage anaphase, nuclear morphogenesis, and cytokinesis. J. Cell Biol., 77: 655-684. Wilson, E. B. (1928) The Cell in Development and Heredity. Macmillan, New York, 3rd ed. Englemann, Th. W. (1882) Ueber Sauerstoffaus- scheidung von Pflanzenzellen im Mikrospektrum. Bot. Zeitung, 40: 419-426. Aronson, J. and Inoué, S. (1970) Reversal by light of the action of N-methyl N-desacethyl colchicine on mitosis. J. Cell Biol., 45: 470-477. Hiramoto, Y., Hamaguchi, M.S., Nakano, Y. and Shoji, Y. (1984) Colcemid UV-microirradiation method for analyzing the role of microtubules in pronuclear migration and chromosome movement in sand-dollar eggs. Zool. Sci., 1: 29-34. ZOOLOGICAL SCIENCE 5: 539-544 (1988) Measurement of Spindle Birefringence by the Optical Integration Method Yosuitaro NaKANo! and YuKkio HirAmoto? Biological Laboratory, Tokyo Institute of Technology, Tokyo 152, Japan ABSTRACT—A method was developed to determine the total amount of birefringence within a given area in the field of a polarization microscope using photometry (optical integration method). By setting a Brace-K6ohler compensator inserted in the optical path at 20° in reference to the crossed polarizer and analyzer axes, it was possible to integrate the phase retardation over an area in the specimen by measuring the light intensity passing through that area. Because the integral of the phase retardation along the cross line of the mitotic spindle is proportional to the number of microtubules in the cross-section that contains the cross-line in echinoderm eggs, the number of microtubules in the cross-section could be determined from the intensity of the light passing through a rectangular area crossing the spindle. Distribution of the number of microtubules in the cross-section of the spindle along its axis was determined at various stages of mitosis in eggs of the sand dollar, Clypeaster japonicus. Effects of Colcemid on the microtubule distribution in the mitotic spindle were determined by measuring the change in birefringence distribution in the spindle after application of seawater containing © 1988 Zoological Society of Japan Colcemid to the egg in mitosis. INTRODUCTION The structure of the mitotic spindle has been examined by many investigators using polarization microscopy since Inoué [1] showed the presence of birefringent fibers in the mitotic spindle in living cells. that many microtubules are present parallel to the axis of the spindle and that spindle fibers observed by light microscopy are the bundles of microtu- bules. Sato et al. [2] concluded that the birefrin- gence of a spindle isolated from a starfish oocyte is caused mainly by microtubules in the spindle, by making a quantitative comparison of the measured birefringence and form birefringence expected theoretically from the number of microtubules in the spindle determined by electron microscopy. Hiramoto et al. [3, 4] extended this conclusion to living sand dollar eggs by their quantitative analy- Electron microscopic studies have shown sis of the spindle birefringence, and concluded that Accepted April 21, 1988 ' Present address: Instruments Division, Nikon Cor- poration, Sakae-ku, Yokohama 244, Japan. * To whom reprints should be requested at the following address: The University of the Air, Wakaba, Chiba 260, Japan. the number of microtubules in the spindle can be estimated by measuring the birefringence of the spindle in living cells. However, it took a full minute or so to obtain a single birefringence distribution in a cross-section using their method. An increase in the speed of measurement is re- quired because the spindle structure changes con- siderably within this period. In the present study, we devised a method to determine the intergral of the birefringence phase retardation over an area in a microscopic field, where the retardation is not uniform, by measuring light intensity passing through that area by polar- ization microscopy (optical integration method). We succeeded in obtaining the distribution of microtubules in the spindle at various stages of first mitosis in sand dollar eggs whereby each measure- ment was possible within 10 sec. OPTICAL INTEGRATION METHOD FOR POLARIZATION MICROSCOPY We used a microscopic birefringence detection system described by Hiramoto ef al. [3] to make birefringence measurements. The system con- sisted of a polarization microscope with a rectified 540 Y. NAKANO AND Y. HIRAMOTO condenser and rectified objectives, a system to illuminate a limited area in the specimen on the microscope stage with 546nm monochromatic light, and a system to measure and display the intensity of light passing through a definite area at the center of the illuminated area. A Brace- Kohler compensator with 27.7 nm (0.319 radian) retardation for 546nm monochromatic light was inserted between the crossed polarizer and analyz- er in the optical path. Hiramoto et al. [3] showed that the intensity of the light passing through the same microscopic birefringence detection system as used in the pres- ent study changed as expected from Jerrard’s theoretical equation [5] when the compensator angle was changed. According to Jerrard [5], the intensity (I) of the light passing through the cros- sed polarizer and analyzer, between which a com- pensator plate and a birefringent specimen are inserted, is given by [=I sin* cos 2 sin?2y 4+ ie ; . - 202 5 (sin 6 ;sin 6 2sin 2 Y ;)+sin°—} (1) where I, is the intensity of the light emerging from the polarizer, ¥,; and 6, are the azimuth and the phase difference of the compensator plate, respec- tively, and ¥> and 6 are those of the specimen whose axis is set at 45° to the polarizer axis (cf. [3]). In the present measurements where 61 was 0.319 radian, I=I,[0.0252 cos 6 sin’2 ¥ +0.157 sin § 2 sin 2yq;+sin?(64/2)]. (2) As shown in this equation, the relation between I and 6 > (retardation of the specimen) is not linearly related when ¥ ; (azimuth of the compen- sator) is small. However, the relation approaches linear if the azimuth of the compensator is in- creased. The solid curve in Figure 1 indicates the theoretical relation between I and 6 > when the compensator plate is set at 20° ( ¥ ;=0.349 radian), in which I increases almost linearly as 6 2 increases at a rate of 0.126/nm. This relation between I and 6 > was confirmed by the following experiment. Chips of mica pre- sy 204 ES a c o a} c ’=) te o o) > . =) © a x ) — —— —— fe) 2 4 6 Phase retardation ( nm ) Fic. 1. Relation between the intensity of light passing through the specimen and its birefringence phase retardation in the optical system used. Light intensity is expressed by the ratio to the light intensity of the background without birefringence. Solid line indicates the relation calculated from Jerrard’s theoretical equation [5]. Circles indicate experimental results using mica chips. Broken line indicates the inclination of 0.13/nm retardation. pared by grinding mica plates in a Waring blender and suspending them in water were put on the stage of the birefringence detection system. The light passing through the chips was determined at various points of uniform phase retardation. The retardation was determined at the same points by the ordinary extinction method (by finding the compensator angle in which light intensity be- comes minimal). As shown by the open circles in Figure 1, the experimental results almost fit the theoretical curve. This implies that the phase retardation (62) of the specimen can be deter- mined from the ratio(x) of the intensity of the light passing through the specimen to the intensity of the light passing through the same area without birefringence by the equation 62=(x—1)/0.126=7.9(x—1) (nm). (3) Because the retardation in every part of the speci- men is proportional to x— 1 in it, the average phase retardation in an area where the retardation is uneven can be determined by the same method. The retardation can be integrated over a region from the total intensity of light passing through Measurement of Spindle Birefringence 541 that region using B=(7.9X 107 ’)a(x—1) (4) where B is the integral of the birefringence (in cm?) over the region (a cm” in area). THE NUMBER OF MICROTUBULES IN THE SPINDLE OF SAND DOLLAR EGGS DURING MITOSIS The number of microtubules at various points in the spindle at various stages of first mitosis was determined in eggs of the sand dollar, Clypeaster japonicus by the optical integration method men- tioned above. Principle of the determination of microtubule num- bers in cross-section of the spindle The birefringence of the spindle is mainly due to the form birefringence of microtubules aligned in parallel to the spindle axis in echinoderm eggs [2, 4]. According to Hiramoto et al. [4], the number of microtubules per unit cross-section of the spin- dle is proportional to the coefficient of birefrin- gence of the spindle, and the total number of microtubules in the cross-section of the spindle in the living cell (N;) is given by N, =(2.08 x 10'°)M,. (5) where M is the areal integral of the coefficient of birefringence over the cross-section of the spindle (in cm?). The areal integral of the coefficient of birefrin- gence over the cross-section of the spindle, viz. the integral of the retardation along the cross-line of the spindle (M;), can be determined from the intensity (Is) of light passing through a rectangular area whose long sides cross the spindle perpen- dicularly and cover the spindle width. Thus, M, =(7.9X 107 ’)L(Is/Ip—1) (6) in which L is the length (in cm) of the long side of the rectangular area crossing the spindle and Ig is the intensity of light passing through an area in the background with the same size and the same absorption as the rectangular area crossing the spindle. From Eqs. 5 and 6 the total number of microtu- bules (N,) in the cross-section of the spindle is given by N, =(1.65 x 10’)L(Is/Ip — 1) (7) Materials and procedure of measurement Eggs were deprived of fertilization membranes and hyaline layers by treating with a 1M urea solution for 1 min shortly after insemination and kept in Ca-free artificial sea water (Ca-free Jamar- in-U; Jamarin Laboratory, Osaka). Shortly before the onset of mitosis, they were put into normal artificial sea water (ASW, Jamarin-U), and then placed on a poly-L-lysine-coated glass slide. Pieces of polyester film 60 sm in thickness were used as spacers, and a cover-slip was put over them. Because the eggs were compressed between the glass slide and cover-slip, most spindles in mitosis were oriented in parallel to the plane of the glass slide. An egg at the onset of first mitosis was brought into the microscopic field of the birefringence detection system, and the spindle axis was oriented at 45° in reference to the polarizer axis. The azimuth of the compensator was set at 20°. A rectangular area (18 “mx 4 am) at the plane of the specimen illuminated with 546nm monochromatic light by inserting a rectangular window of an appropriate size at the plane conju- gate to the specimen plane, in reference to the condenser lens so that the long sides of the window image might cross the spindle axis at right angles. The intensity of light passing through the rectangu- lar window corresponding to 15 ~m x1 ym at the center of the illuminated area was cast on the photocathode of a photomultiplier tube to meas- ure the intensity of the light after passing through another window placed at the plane conjugate to the specimen plane in reference to the objective lens (cf. [3]). A point on the extension of the spindle axis was brought to the center of the illuminated area at the beginning of measurement and the stage of the was microscope was moved with a micromotor in the direction of the spindle axis at a speed of 5 m/sec to record the intensity of light passing through the rectangular area crossing the spindle. The length 542 Y. NAKANO AND Y. HIRAMOTO of the area (15 um) was large enought to cover the spindle width. A single measurement of the change in light intensity along the whole spindle axis could be made within 10sec or so. The intensity of light passing through the background (Ip in Eqs. 6, 7) was measured by setting the background at the center of the microscopic field or setting the spindle axis in parallel to the polariz- er axis. The numbers of microtubules in cross-section (N,) at various positions along the spindle axis were obtained by substituting the intensity of light thus measured and the light intensity of the back- ground for Is and Ig in Eq. 8. N, =(2.5 X 10*) (Is/In—1) (8) which was obtained by substituting 1.5 10~* cm (15 wm) for L in Eq. 7. At each interval between successive measure- ments of birefringence, the glass slide was transfer- red to the stage of a differential interference microscope (Optiphot, Nikon) to take a micro- graph of the same egg used in the birefringence measurement and then transferred back to the stage of the polarization microscope for the next measurement. In this way, the morphological processes of mitosis including the positions of chromosomes and centrosomes which were dif- ficult to determine accurately by polarization mi- croscopy could be recorded in the same egg. Changes in microtubule distribution in the spindle during mitosis Figure 2 shows a series of measurements of an egg in mitosis. In this egg, the distribution of the number of microtubules along the spindle axis was measured 7 times (a-g). The positions of the centrosomes at the center of the aster (circles) and those of the chromosomes (triangles) in this figure were obtained from differential interference micrographs taken during the intervals between successive birefringence measurements. The tim- ing of birefringence measurements and differential interference microphotography is shown by hori- zontal arrows pointing to the time scale on the right. These records are similar to those in a previous report [4], although the method of measurement was greatly improved in the present 8 2 = - > > o | = 44 3 Cc 6 > oO a O D 4 One rs 6 ' | \ s (3) | oO Cc Oo 0 sella a pls 5 | \ E I \ 3 | | \ otubu =) fF MUcr [o} Ss ee oO — a= = o fa) Db oe Number Oo E es in ¢ oO gS ——— Q E o Time TS mem eNO ee PM -30 fe) 30 Distance from the equatorial plane (jum) Fic. 2. Change in the distribution of microtubules in the spindle of a sand dollar egg at first mitotis. Hatched area indicates the amount of microtubules. Circles indicate the positions of the centrosomes and triangles, the positions of the chromosomes. Time and stage of mitosis are indicated on the right side. Horizonal arrows indicate time of measurement. PM, M, A and T are prometaphase, metaphase, anaphase and telophase, respectively. study. The area of the hatched region in Figure 2, which is the integral of the number of microtubules along the spindle length from one end to the other, indicates the sum of the lengths of all microtubules contained in the spindle. It can be seen in Figure 2 that the amount of microtubules expressed by the sum of the lengths of microtubules increases in prometaphase, metaphase and early anaphase and decreases in late anaphase and telophase. Measurement of Spindle Birefringence 543 M —-__—=—_ A —T— n o 2 104 s+ ee oS 384 Oo rao) = ec 6 o= a ce. Ad D0 love Er O ad Cc ie} o ol — = “© 4 2. 3 4. 5 6 7 & © 40 Time ¢ mit ) Fic. 3. Change in the total amount of microtubules in the spindle of a sand dollar egg at first mitosis. The total amount of microtubules is expressed by the sum of the lengths of all the microtubules. M, A, and T are metaphase, anaphase and telophase, respectively. Figure 3 shows the change in the total amount (length) of microtubules in the spindle during mitosis from another egg. The above results indicate that formation and destruction of microtu- bules occur during mitosis in parallel to their translocation in the spindle. Effect of Colcemid on the microtubule distribution in the spindle Colcemid (N-methyl-N-desacethylcolchicine) is known to be an agent which can destroy microtu- bule structures. The effects of Colcemid on micro- tubule distribution in the spindle were examined by the optical integration method. When the egg reached metaphase, ASW con- taining 10° °-10- © M Colcemid (Sigma Chemicals, St. Louis, MO) was substituted for the medium of the egg on the glass slide by adding it at one edge of the cover-slip and placing a piece of filter paper at the opposite edge. Figure 4 shows typical results. The hatched area in the figure a indicates the distribution of microtu- bules in the spindle in ASW and those in b~g are the microtubule distributions after ASW contain- ing Colcemid (Colcemid-SW) was substituted for the medium. It was noted that both the microtu- bule number and the length of the spindle gradual- ly decrease after the medium is replaced with Colcemid SW. It appears that the disassembly of vs) A ee | Cc [e) es oO Mo ie5) ' uv) % — oO = Oo - i= = E wo 4 = iS ® =] Ee wo 5 ee im = 2 5 jo) a — ie} E O = = 9 5 | ie ive) oO Q iS = Zz -30 fe) 30 Distance from the equatorial plane Cum) Fic. 4. Effect of Colcemid on the distribution of micro- tubules in the metaphase spindle of a sand dollar egg. Circles indicate the positions of the centrosomes. Triangles indicate the position of chromosomes re- maining in the equatorial plane of the spindle. Time of application of Colcemid-SW is indicated on the right side. Horizontal arrows indicate time of measurement. microtubules is more conspicuous at the polar regions than at the central regions of the spindle. Chromosomes stay in the equatorial plane and the distance between the centrosomes decreases. This evidence suggests that the centrosomes are still attached to the spindle even after the disassembly of microtubules at the polar ends of the spindle. DISCUSSION Optical integration method The optical integration method described in the present study is a simple method to determine the average or total birefringence phase retardation of a microscopic body when the axes of birefringence are parallel to one another and the intensity of birefringence is uneven. If the birefringence is regarded to be form birefringence due to the fibrous elements being aligned in the same direc- 544 Y. NAKANO AND Y. HiRAMOTO tion as the mitotic spindle in echinoderm eggs, the total amount of fibrous elements in the body can be estimated by this method. In using this method, the elimination of stray light coming from background is important be- cause the background is quite bright owing to the large azimuth of the compensator. Limitation of the illuminated area by using a window inserted in the optical path for illumination may be effective in minimizing measurement errors due to back- ground brightness (cf. [3]). The fact that the experimental results using mica chips fit so well with the theoretical curve expected from Jerrard’s equation [5] may indicate that this method can be of practical use. As shown by Eqs. 1| and 2, an increase in the azimuth of the compensator plate increases the linearity between the light intensity and the re- tardation of the specimen, while it decreases the ratio of the light intensity in the birefringent region to the light intensity in the background. We decided that the azimuth of the compensator plate should be set at 20° considering the fact that the linearity and the accuracy of the present ex- perimental conditions depends on the strength of the birefringence of the specimen and the phase retardation of the compensator plate. We demonstrated changes in the distribution and number of microtubules in the spindle under circumstances of normal division and immersion in seawater containing Colcemid, an agent inhibiting polymerization of tubulin into microtubules. We confirmed our previous results [4] on microtubule distribution in normal mitosis and presented more detailed data. By application of Colcemid-SW to the cell in mitosis, microtubules disappeared pre- dominantly at the polar regions as compared with the equatorial regions of the spindle. This observation appears to contradict the idea that the depolymerization of microtubules at a steady state predominantly occurs at the plus ends of microtu- bules [6], because most of the spindle microtubules are oriented so that the plus ends point toward the equatorial region (cf. [7]). However, further stu- dies on the mode of disassembly of microtubules by Colcemid in living cells and more quantitative information on the dislocation of microtubules in Colcemid-treated cells are required for a detailed discussion of the effect of Colcemid on the micro- tubule structure of the spindle. Differences in stability between kinetochore microtubules and nonkinetochore microtubules to colchicine re- ported by Salmon et al. [8] should also be consi- dered in this discussion. ACKNOWLEDGMENT This work was supported in part by a Grant-in-Aid for Scientific Research No. 61480019 from the Ministry of Education, Science and Culture, Japan awarded to Y. H. REFERENCES 1 Inoué,S. (1952) The effect of colchicine on the microscopic and submicroscopic structure of the mitotic spindle. Exp. Cell Res., 2 (Suppl.): 305-318. Sato, H., Ellis, G. W. and Inoué, S. (1975) Micro- tubular origin of mitotic spindle form birefringence. Demonstration of the applicability of Wiener’s equa- tion. J. Cell Biol., 67: 501-517. 3 Hiramoto, Y., Hamaguchi, Y., Shdji, Y. and Shimo- da, S. (1981) Quantitative Studies on the polarization optical properties of living cells. I. Microphotometric birefringence detection system. J. Cell Biol., 89: 115- 120. 4 Hiramoto, Y., Hamaguchi, Y., Shoji, Y., Schroeder, T. E., Shimoda, S. and Nakamura, S. (1981) Quan- titative Studies on the polarization optical properties of living cells. H. The role of microtubules in bire- fringence of the spindle of the sea urchin egg. J. Cell Biol., 89: 121-130. 5 Jerrard,H.G. (1948) Optical compensators for measurement of elliptical polarization. J. Opt. Soc. Am., 38: 35-59. 6 Horio, T. and Hotani, H. (1986) Visualization of the dynamic instability of individual microtubules by dark-field microscopy. Nature (Lond.), 321: 605-607. 7 Euteneuer, U. and McIntosh, R. (1981) Structural polarity of kinetochore microtubules in PtK;, cells. J. Cell Biol., 89: 338-345. 8 Salmon, E.D., McKeel,M. and Hays, T. (1984) Rapid rate of tubulin dissociation from microbutubles in the mitotic spindle in vitro measured by blocking polymerization with colchicine. J. Cell Biol., 99: 1066-1075. Nw ZOOLOGICAL SCIENCE 5: 545-552 (1988) © 1988 Zoological Society of Japan In vivo Cytochemistry in Cell Division YUKIHISA HAMAGUCHI Biological Laboratory, Tokyo Institute of Technology, O-okayama, Meguro-ku, Tokyo 152, Japan INTRODUCTION The movement of chromosomes during cell divi- sion and the dividing process of the cell have well been documented. The mitotic apparatus plays an important role in translocating chromosomes while the contractile ring plays an important role in dividing the cell. These two structures are fun- damentally composed of two different cytoskeletal elements, i.e., microtubules in the mitotic appa- ratus and microfilaments in the contractile ring. These elements are too thin to be observed by ordinary light microscopy, although they can be observed individually by electron microscopy after fixation and staining. It is well-known through biochemical analysis that tubulin and actin are the principal constituents of microtubules and mi- crofilaments, respectively, and that the dynamics of microtubules and microfilaments have been extensively investigated in vitro. However, the in vivo dynamics of these molecules in living cells have not yet been investigated to a satisfactory degree. It is not clear how the mitotic apparatus and the cleavage furrow are composed of micro- tubules and microfilaments, respectively, or how the motive force responsible for chromosome movement and dividing processes is generated in these structures during cell division (for a review, see [1]). In vivo cytochemistry was introduced to supple- ment cytological and biochemical analyses by Taylor and Wang [2]. They injected fluorescently labeled actin into living cells and reported the incorporation of the exogenous actin into the microfilamentous structures in living cells [2]. In this brief review, I summarize results obtained by in vivo cytochemistry in the last ten years, from the time studies were first begun. Accepted April 20, 1988 METHOD FOR IN VIVO CYTOCHEMISTRY Procedures of in vivo cytochemistry are as fol- lows (see [3, 4] for reviews). 1. Prepare a molecule having biological spe- cificity 2. Label with probes 3. Purify a functional and optimally labeled con- jugate 4. Characterize the labeled conjugate a. probe/molecule ratio b. site of labeling c. check the functional activity of the conju- gate in vitro 5. Microinject the labeled conjugate into living cells 6. Determine the functional activities in vivo The molecules having biological specificity used in in vivo cytochemistry are divided into three groups: 1) cytoskeletal proteins, 2) antibodies against cytoskeletal proteins, and 3) agents specific to cytoskeletal proteins. Fluorescent probes are the most useful among the various probes because they are detectable in living cells by fluorescence microscopy. Since fluorescence conjugates are called fluorescent analogs, in vivo cytochemistry where fluorescent analogs are used is called fluorescent analog cytochemistry. In the past ten years, in vivo cytochemistry has become a useful tool mainly due to the improvements of two tech- niques, micromanipulation and video microscopy. Micromanipulation or microinjection was greatly advanced by the introduction of useful micromani- pulators and the development of a method of microinjection [4-6]. The development of elec- tronics has been of great significance, especially in digital image using video which can raise the luminescence of the microscopic image and en- hance the contrast of the image [7, 8]. 546 Y. HAMAGUCHI Using these fluorescent analogs, we can investi- gate the following functions of the cytoskeleton and cytoskeletal proteins in vivo by means of determining the functional activities in vivo as well as in vitro (for details, see the following sections). 1. Dynamic distribution 2. Local physiological characteristics such as polymerization and pH (fluorescence of fluorescein is pH-sensitive) 3. Molecular dynamics (assembly-disassembly equilibrium, dynamic redistribution, and diffusion constant) 4. Inhibition CYTOSKELETAL PROTEINS Mitosis Tubulin is one of the proteins which has been investigated extensively by in vivo cytochemistry during mitosis because it is a conservative protein and can copolymerize with the endogenous tubulin in the injected cell. Rich sources of tubulin can be found in brains and sperm flagella. Less than 3% of the endogenous tubulin did not perturb mitosis or cleavage when injected into sea urchin eggs [9]. Travis et al. [10] first prepared native and fluores- cent analogs of brain tubulin and microtubule- associated proteins; fluorescent analog of tubulin showed polymerizability and these fluorescent ana- logs showed affinity for unlabeled proteins. Keith et al. [11] reported that the fluorescent analog of tubulin was incorporated into the microtubular network in the cell even though the analogs of microtubule-associated proteins were not incorpo- rated into the microtubules when injected into the living cell. In sea urchin eggs, the fluorescent analog of tubulin was incorporated into microtubu- lar structures such as the mitotic apparatus and the sperm aster, and the distribution in the mitotic spindle of the labeled tubulin was coincident with the distribution of birefringence [9, 12, 13]. Even after the formation of the mitotic apparatus, tubu- lin was found to be incorporated within 20-30 sec of injection, and fluorescence redistribution occur- red within this period after photobleaching in sea urchin eggs, indicating that tubulin in microtubules is quickly exchangeable with that in the cytoplasm [14, 15]. Exchangeability in the mitotic cell was 18-fold more than that in the interphase cell [16]. The effects of microtubule depolymerizing agents such as colchicine and nocodazole, and microtu- bule stabilizing agents such as taxol on the mitotic apparatus were examined in vivo [14, 15, 17]. Mitchison and Kirschner [18] prepared biotiny- lated tubulin, which is not fluorescent but is visual- ized by electron microscopy using an antibody against biotin followed by a secondary antibody coupled to collloidal gold. Using this analog, Mitchison et al. [19] reported that astral microtu- bules incorporated the analog very rapidly, but, by contrast, kinetochore microtubules incorporated it at a slower rate at metaphase and not at all during anaphase. Saxton and McIntosh [17], using the fluorescent analog of tubulin, showed that interdigitated mi- crotubules in the interzonal region can undergo antiparallel sliding and therefore, that the mitotic spindle can elongate during anaphase. Hamaguchi et al. [15] also reported that the photobleached region moved poleward at the same rate as spindle elongation in metaphase and anaphase during fluorescence recovery after photobleaching (Fig. 1). However, Wadsworth and Salmon [14] re- ported that the translocation of the bleached re- gion could not be detected at metaphase. The injection of calcium-saturated calmodulin caused a significant shortening of the kinetochore and interpolar microtubules into which the fluores- cent analog of tubulin had been incorporated but did not cause shortening of the astral microtu- bules. The spindle quickly recovered its normal form [20]. Calmodulin was extracted from brains and testes, purified, and labeled fluorescently. The fluorescent analog of calmodulin was also used to investigate the function of the protein in mitosis [21-24]. The fluorescent analog of calmodulin accumulated in the mitotic apparatus (Fig. 2), and bound to microtubules in a calcium-independent manner even though it usually binds to many enzymes in a calcium-dependent manner when it acts as a modulator protein [21]. The distribution of calmodulin is different from that of microtu- bules and dynein, although the role of calmodulin in mitosis might be imagined to regulate assembly In vivo Cytochemistry 547 Fic. 1. A series of fluorescence micrographs of anaphase mitotic apparatus in which fluorescence was redistributed when the photobleached area (the arrow) was perpendicular to the spindle axis. The numbers in these micrographs represent the time (sec) after the end of photobleaching. The fluorescent analog of tubulin was injected into sand dollar eggs at prophase. (From Hamaguchi et al. [15] with the permission of Cell Struct. Funct.) and disassembly of microtubules and/or dynein ATPase activity (21-23, 25, 26]. Scherson et al. [27] and Drubin and Kirschner [28] reported that tau protein and MAP2, microtu- bule associated proteins, were injected into living cells after labeling and were associated with micro- tubules, in contrast to the report by Keith ef al. [11]. These proteins did not perturb mitosis and tau protein was found to induce tubulin assembly and to stabilize microtubules in vivo although they were not always involved in all types of cells [27, 28]. Cytokinesis Rabbit skeletal muscle actin was first used in the study of fluorescent analog cytochemistry by Taylor and Wang [2, 29]. They reported that the fluorescent analog of actin was incorporated into microfilamentous structures in the cortex of sea urchin eggs for only a short period after fertiliza- tion when injected into unfertilized eggs [29]. This was, however, inconsistent with the biochemical and cytological analyses. Hamaguchi and Mabuchi [30] reinvestigated actin dynamics in the living cell during early development using sea urchin egg actin labeled with fluorescent probes which can modify two different groups of actin. Fluorescein- labeled actin was incorporated into microvilli and the cortex through cleavage. No difference was found in fluorescence between the region of the cleavage furrow and the rest of the cortex, which indicates that as shown in biochemical and electron microscopical analyses [31, 32], microfilaments of the contractile ring are not newly synthesized from the cytoplasmic actin pool, but are reconstructed from the cortical microfilamentous network pre- viously formed. Accumulated actin molecules in the cortex, which may be polymerized, may ex- change quickly with diffusible cytoplasmic actin molecules because the fluorescence in the cortex quickly redistributed after photobleaching [30]. Fluorescent analogs of light chains of muscle myosin were injected into dividing cells [33, 34]. Myosin light chains were concentrated in the cleav- age furrow, which suggests that fluorescent analogs of myosin light chains from muscle can be readily incorporated into non-muscle myosins and then used to follow the dynamics of myosin distribution in living cells. The fluorescent analog of alpha-actinin isolated from sea urchin eggs or muscle was readily in- corporated into the microfilamentous structures of the egg cortex although the localization of the fluorescent analog of egg alpha-actinin was distinct compared with that of the analog of muscle alpha- actinin [35]. The fluorescent layer at the cleavage furrow region seemed slightly thicker than that in the polar region. Just after the meiotic apparatus arrived at the cortical region of starfish oocytes, egg alpha-actinin accumulated in the region where Y. HAMAGUCHI 548 Fic. 2 In vivo Cytochemistry 549 Fic. 3. Distribution of the fluorescent analog of egg alpha-actinin in a starfish oocyte during the first polar body formation. The numbers represent time (min: sec) after treatment with 1-methyladenine. The upper and lower rows are fluorescence and differential interference micrographs, respectively. (From Hamaguchi and Mabuchi [36] with the permission of Cell Motil. Cytoskeleton) the polar body was expected to form, suggesting that the meiotic apparatus induces differentiation of the cortex so as to form a polar body (Fig. 3) [36]. Fluorescent analogs of tropomyosin and filamin, other microfilament-associated proteins, were in- corporated into stress fibers of interphase cells but were not reported in dividing cells [37, 38]. ANTIBODIES AGAINST CYTOSKELETAL PROTEINS Monoclonal antitubulin antibody (YL1/2) which reacts specifically with tyrosinated alpha-tubulin was fluorescently labeled and injected into living cells [39, 40]. At lower concentrations, the anti- body did not perturb cell division and was incorpo- rated into the mitotic apparatus. At present, only one antibody metioned above has been applied in vivo. However, since antibodies against different types of tubulin molecules have been obtained, the dynamic distribution corresponding to each type of tubulin molecule may be investigated [41-44]. Moreover, peptide antibodies are simple to pre- pare [45]. They are antibodies against short pep- tides whose sequence corresponds to the sequence Fic. 2. Localization of the fluorescent analog of calmodulin in a blastomere during the 8-cell stage. Differential interference micrographs show the mitotic stage (left column) and fluorescence micrographs show calmodulin localization (right column). The fluorescent analog of calmodulin was injected into a fertilized egg at prophase. (From Hamaguchi and Iwasa [21] with the permission of Biomed. Res.) 550 Y. HAMAGUCHI of a part of a cytoskeletal protein. They react specifically with different molecules of cytoskeletal proteins even with different molecular sites of cytoskeletal proteins. Dynamic distribution and specific inhibition of each cytoskeletal protein may be investigated using these antibodies. AGENTS SPECIFIC TO CYTOSKELETAL PROTEINS Phallotoxins, toxic hepta-peptides from a mushroom, are filamentous-actin specific drugs [46]. Fluorescent analogs of phallotoxins which show no significant alteration in the binding capac- ity to filamentous-actin can be used to visualize actin containing cytoskeleton [47-49]. The fluorescent analog of phalloidin was injected into living eggs in order to investigate F-actin localiza- tion [49]. Fluorescein-labeled phalloidin became localized in the cortical layer of both unfertilized and fertilized eggs soon after injection. No differ- ence was detected, however, in fluorescence be- tween the region of the cleavage furrow and the rest of the cell cortex during cytokinesis, which is consistent with the results obtained using the fluorescent analog of actin. Distinct fluorescence was not detected in the mitotic apparatus, which indicates that microfilaments are not concentrated in the mitotic apparatus. The fluorescent analog of colchicine, a tubulin binding agent, was used to visualize soluble tubu- lin, but not microtubules in fixed cells [50]. PERSPECTIVES Further advancements in equipment and tech- niques for in vivo cytochemistry are outlined below. The recent development of the scanning confo- cal light microscope [41] offers the advantages of greater resolution than that attained by conven- tional microscopy and improved rejection of out- of-focus noise which provides better optical sec- tioning of thick specimens such as living cells. Cytoskeletal elements can more easily and clearly be visualized by fluorescent analog using a scan- ning confocal light microscope as a fluorescence microscope. Owing to advances in video microscopy, small granules have now become visible’ with differential interference microscopy [7, 52]. Small particles such as colloidal gold particles and bacteria are valuable tools with wide applicability to the study of the functions of cytoskeletal pro- teins within living cells [52, 53]. Gold particles coupled to a monoclonal antibody reacting with the alpha-subunit of tubulin stayed in an entirely fixed position in the cell after injection, which indicates that microtubule treadmilling does not seem to be involved in microtubule dynamics in the cell [52]. Phycobiliproteins, constituents of the light- harvesting apparatus of blue-green bacteria, red algae, and cryptomonads, can be used as fluores- cent probes because they are highly fluorescent [54]. If fluorescent intensity greatly increases, detection of cytoskeletal elements may be im- proved. Oi ef al. [54] reported on the various phycobiliprotein conjugates to a molecule having biological specificities such as _ phycoerythrin- immunoglobulin, phycoerythrin-protein A, and phycoerythrin-avidin conjugates. Local structural changes in cytoskeletal proteins can be determined in vivo if the fluorescent probe responsible for the condition of the cytoskeleton such as pyrene is used. Fluorescence intensity of N-(1-pyrenyl)- iodoacetamide-labeled actin was enhanced by a factor of about 25 on polymerization [55]. In the near future, in vivo cytochemistry is expected to be applied to many different cyto- skeletal proteins in various dividing cells using the improvements mentioned above, and may become an even more powerful technique in the study of cell division. ACKNOWLEDGMENT This review is dedicated to Professor Katsuma Dan on the occasion of his 83rd birthday. I wish to thank Dr. M. S. Hamaguchi for her valuable criticism. This work was supported by Grants-in-Aid from the Ministry of Education, Science and Culture of Japan (61304008, 62300004, and 62540538). REFERENCES 1 Ishikawa, H., Hatano, S. and Sato, H. (1985) Cell 10 11 12 13 14 In vivo Cytochemistry Motility: Mechanism and Regulation, Univ. Tokyo Press, Tokyo. Taylor, D.L. and Wang, Y.-L. (1987) Molecular cytochemistry: Incorporation of fluorescently labeled actin into living cells. Proc. Natl. Acad. Sci., 75: 857-861. Taylor, D. L. and Wang, Y.-L. (1980) Fluorescent- ly labelled molecules as probes of the structure and function of living cells. Nature, 284: 405-410. Kreis, T. E. and Birchmeier, W. (1982) Microinjec- tion of fluorescently labeled proteins into living cells with emphasis on cytoskeletal proteins. Int. Rev. Cytol., 75: 209-227. Hamaguchi, Y. (1986) Introduction to microinjec- tion. Saiho (Cell), 18: 36-40. (in Japanese) Hiramoto, Y. (1984) Micromanipulation. In “Cell Biology”. Ed. by T. Niizu and T. Okigaki, Maruzen Co. Ltd., Tokyo, pp. 277-297. (in Japanese) Inoué, S. (1986) Video Microscopy, Plenum Press, New York. Arndt-Jovin, D. J., Robert-Nicoud, M., Kaufman, S.J. and Jovin, T. M. (1985) Fluorescence digital imaging microscopy in cell biology. Science, 230: 247-256. Hamaguchi, Y., Toriyama, M., Sakai, H. and Hira- moto, Y. (1985) Distribution of fluorescently labeled tubulin injected into sand dollar eggs from fertilization through cleavage. J. Cell Biol., 100: 1262-1272. Travis, J.L., Allen, R.D. and Sloboda, R. D. (1980) Preparation and characterization of native, fluorescently labelled brain tubulin and microtubule- associated proteins (MAPs). Exp. Cell Res., 125: 421-429. Keith, C. H., Feramisco, J. R. and Shelanski, M. (1981) Direct visualization of fluorescein-labeled microtubules in vitro and in microinjected fibro- blasts. J. Cell Biol., 88: 234-240. Salmon, E. D., Leslie, R. J., Saxton, W. M., Karow, M. L. and McIntosh, J. R. (1984) Spindle microtubule dynamics in sea urchin embryos: Analy- sis using a fluorescein-labeled tubulin and measure- ments of fluorescence redistribution after laser photobleaching. J. Cell Biol., 99: 2165-2174. Hiramoto, Y., Hamaguchi, Y., Hamaguchi, M. S. and Nakano, Y. (1985) Roles of microtubules in pronuclear migration and spindle elongation in sand dollar eggs. In “Cell Motility: Mechanism and Reg- ulation”. Ed. by H. Ishikawa, S. Hatano and H. Sato, Univ. Tokyo Press, Tokyo, pp. 349-356. Wadsworth, P. and Salmon, E. D. (1986) Analysis of the treadmilling model during metaphase of mito- sis using fluorescence redistribution after photo- bleaching. J. Cell Biol., 102: 1032-1038. Hamaguchi, Y., Toriyama, M., Sakai, H. and Hira- moto, Y. (1987) Redistribution of fluorescently 16 19 20 551 labeled tubulin in the mitotic apparatus of sand dollar eggs and the effects of taxol. Cell Struct. Funct., 12: 43-52. Saxton, W.M., Stemple, D.L., Leslie, R. J., Sal- mon, E.D., Zavortink,M. and McIntosh, J. R. (1984) Tubulin dynamics in cultured mammalian cells. J. Cell Biol., 99: 2175-2186. Saxton, W. M. and McIntosh, J. R. (1987) Inter- zone microtubule behavior in late anaphase and telophase spindles. J. Cell Biol., 105: 875-886. Mitchison, T. J. and Kirschner, M. W. (1985) Prop- erties of the kinetochore in vitro. I. Microtubule nucleation and tubulin binding. J. Cell Biol., 101: 755-765. Mitchison, T., Evans,L., Schulze, E. and Kir- schner, M. (1986) Sites of microtubule assembly and disassembly in the mitotic spindle. Cell, 45: 515- 527. Keith, C. H. (1987) cium-calmodulin on mitosis in PtK2 cells. Motil. Cytoskeleton, 7: 1-9. Hamaguchi, Y. and Iwasa, F. (1980) Localization of fluorescently labeled calmodulin in living sea urchin eggs during early development. Biomed. Res., 1: 502-509. Hamaguchi, Y. and Iwasa, F. (1982) Localization of fluorescently labeled calmodulin in living sand dollar eggs during early development. In “Biological Func- tions of Microtubules”. Ed. by H. Sakai, H. Mohri and G. G. Borisy, Academic Press, Tokyo, pp. 199- 210. Zavortink, M., Welsh,M.J. and McIntosh, J. R. (1983) The distribution of calmodulin in living mitotic cells. Exp. Cell Res., 149: 375-385. Luby-Phelps, K., Lanni,F. and Taylor, D. L. (1985) Behavior of a fluorescent analogue of calmo- dulin in living 3T3 cells. J. Cell Biol., 101: 1245- 1256. Hisanaga, S., Tanaka, T., Masaki, T., Sakai, H., Mabuchi, I. and Hiramoto, Y. (1987) Localization of sea urchin egg cytoplasmic dynein in mitotic apparatus studied by using a monoclonal antibody against sea urchin sperm flagellar 21S dynein. Cell Motil. Cytoskeleton, 7: 97-109. Hisanaga, S. and Pratt, M. M. (1984) Calmodulin interaction with cytoplasmic and flagellar dynein: Calcium-dependent binding and stimulation of ade- nosinetriphosphatase activity. Biochemistry, 23: 3032-3037. Scherson, T., Kreis, T. E., Schlessinger, J., Lit- tauer, U. Z., Borisy,G. G. and Geiger, B. (1984) Dynamic interactions of fluorescently labeled micro- tubule-associated proteins in living cells. J. Cell Biol., 99: 425-434. Drubin, D. G. and Kirschner, M. W. (1986) Tau protein function in living cells. J. Cell Biol., 103: Effect of microinjected cal- Cell 29 30 Ww i) 33 34 35 36 37 40 41 552 2739-2746. Wang, Y.-L. and Taylor, D. L. (1979) Distribution of fluorescently labeled actin in living sea urchin eggs during early development. J. Cell Biol., 82: 672-679. Hamaguchi, Y. and Mabuchi, I. (1988) Accumula- tion of fluorescently labeled actin in the cortical layer in sea urchin eggs after fertilization. Cell Motil. Cytoskeleton, 9: 153-163. Mabuchi,I., Hosoya,H. and Sakai, H. (1980) Actin in the cortical layer of the sea urchin eggs. Biomed. Res., 1: 417-426. Usui, N. and Yoneda, M. (1982) Ultrastructural basis of the tension increase in sea-urchin eggs prior to cytokinesis. Dev. Growth Differ., 24: 453-465. Mittal, B., Sanger, J. M. and Sanger, J. W. (1987) Visualization of myosin in living cells. J. Cell Biol., 105: 1753-1760. McKenna, N. M. and Wang, Y.-L. (1987) Localiza- tion and exchangeability of microinjected smooth muscle myosin LCyo in living cells. J. Cell Biol., 105: 127a. Mabuchi, I., Hamaguchi, Y., Kobayashi, T., Hosoya, H., Tsukita,S. and Tsukita,S. (1985) Alpha-actinin from sea urchin eggs: Biochemical properties, interaction with actin, and distribution in the cell during fertilization and cleavage. J. Cell Biol., 100: 375-383. Hamaguchi, Y. and Mabuchi,I. (1986) Alpha- actinin accumulation in the cortex of echinoderm eggs during fertilization. Cell Motil. Cytoskeleton, 6: 549-559. Wehland, J. and Weber, K. (1980) Distribution of fluorescently labeled actin and tropomyosin after microinjection in living tissue culture cells as observed with TV image intensification. Exp. Cell Res., 127: 397-408. Mittal, B., Sanger, J. M. and Sanger, J. W. (1987) Binding and distribution of fluorescently labeled filamin in permeabilized and living cells. Cell Motil. Cytoskeleton, 8: 345-359. Wehland, J. and Willingham, M. C. (1983) A rat monoclonal antibody reacting specifically with the tyrosylated from of a-tubulin. II. Effects on cell movement, organization of microtubules, and in- termediate filaments, and arrangement of Golgi elements. J. Cell Biol., 97: 1476-1490. Warn, R. M., Flegg, L. and Warn, A. (1987) An investigation of microtubule organization and func- tions in living Drosophila embryos by injection of a fluorescently labeled antibody against tyrosinated a-tubulin. J. Cell Biol., 105: 1721-1730. Thompson, W. C., Asai, D.J. and Carney, D. H. (1984) Heterogeneity among microtubules of the cytoplasmic microtubule complex detected by a 42 44 48 49 50 51 52, 53 54 55 Y. HAMAGUCHI monoclonal antibody to alpha tubulin. J. Cell Biol., 98: 1017-1025. Gundersen, G. G. and Bulinski, J.C. (1986) Dis- tribution of tyrosinated and nontyrosinated a- tubulin during mitosis. J. Cell Biol., 102: 1118-1126. Piperno, G., LeDizet, M. and Chang, X.-J. (1987) Microtubules containing acetylated a-tubulin in mammalian cells in culture. J. Cell Biol., 104: 289- 302. Lopata, M. A. and Cleveland, D. W. (1987) In vivo microtubules are copolymers of available #-tubulin isotypes: Localization of each of six vertebrate /- tubulin isotypes using polyclonal antibodies elicited by synthetic peptide antigens. J. Cell Biol., 105: 1707-1720. Bulinski, J. C. (1986) Peptide antibodies: New tools for cell biology. Int. Rev. Cytol., 103: 281-302. Wieland, T. (1977) Modification of actins by phallo- toxins. Naturwissenschaften, 64: 303-309. Wulf, E., Deboben, A., Bautz, F. A., Faulstich, H. and Wieland, T. (1979) Fluorescent phallotoxin, a tool for the visualization of cellular actin. Proc. Natl. Acad. Sci. USA, 76: 4498-4502. Barek, L.S., Yocum, R. R., Nothnagal, E. A. and Webb, W. W. (1980) Fluorescence staining of the actin cytoskeleton in living cells with 7-nitrobenz-2- oxa-1, 3-diazole-phallacidin. Proc. Natl. Acad. Sci. USA, 77: 980-984. Hamaguchi, Y. and Mabuchi, Y. (1982) Effects of phalloidin microinjection and localization of fluores- cein-labeled phalloidin in living sand dollar eggs. Cell Motil., 2: 103-113. Albertini, D. F. and Clark, J. I. (1981) Visualiza- tion of assembled and disassembled microtubule protein by double label fluorescence microscopy. Cell Biol. Int. Rep., 5: 387-397. White, J. G., Amos, W. B. and Fordham, M. (1987) An evaluation of confocal versus conventional imag- ing of biological structures by fluorescence light microscopy. J. Cell Biol., 105: 41-48. De Brabander, M., Geuens,G., Nuydens,R., Moeremans, M. and De May, J. (1985) Probing microtubule-dependent intracellular motility with nanometre particle video ultramicroscopy (nanovid ultramicroscopy). Cytobios, 43: 273-283. Hynes, T. R., Block, S. M., White, B. T. and Spu- dich, J. A. (1987) Movement of myosin fragments in vitro: Domains involved in force production. Cell, 48: 953-963. Oi, V.T., Glazer, A.N. and Stryer,L. (1982) Fluorescent phycobiliprotein conjugates for analyses of cells and molecules. J. Cell Biol., 93: 981-986. Kouyama, T. and Mihashi, K. (1981) Fluorimetry study of N-(1-pyrenyl)iodoacetamide-labelled F- actin. Eur. J. Biochem., 114: 33-38. ZOOLOGICAL SCIENCE 5: 553-562 (1988) © 1988 Zoological Society of Japan Computed Profiles of Compressed Sea-Urchin Eggs with Elastic Membranes Mitsuki YONEDA Department of Zoology, Faculty of Science, Kyoto University, Kyoto 606, Japan ABSTRACT— A computer simulation was made for profiles of compressed sea-urchin eggs with various amounts of elasticity at their surfaces. Fluidity of cytoplasm and negligible bending stress of the surface were assumed a priori. Based on the profiles thus calculated, the relation of the force of compression and the degree of flattening was predicted and matched with the observed force and flattening of unfertilized eggs of Hemicentrotus pulcherrimus. The experimental data were shown to fit well with the calculated force-deformation curve on the assumption of non-elasticity at the surface. The possibility that the egg surface had an elasticity modulus of 10° dyne/cm? was precluded. Whether the surface of sea-urchin eggs is elastic or not has been the subject of a debate between Hiramoto [1-5] and myself [6-8]. Both claims have emerged from experimental data on the force required to compress the egg by the “compression method” originally developed by Cole [9]. When an unfertilized sea-urchin egg with the diameter Zo is compressed between a pair of parallel plates with a known force (F), the thick- ness (Z) of the egg reaches an equilibrium within a few minutes. By measuring the thickness of a single egg under varing amounts of force, we can determine the relation of the: force (F) and the thickness (z= Z/Zo) as a “F-z curve” (Fig. 1). Hiramoto adopted a mathematical procedure different from mine to calculate the tension (T) working at the egg surface from measured values of force (F). This gave rise to the differing views as to the physical nature of the egg surface; Hiramoto [1] converted the measured force into excess inter- nal pressure P=F/ 2 D* by measuring the radius (D) of the circle of cell surface in contact with the plates of compression (Fig. 1). By also measuring the radii of curvatures (R, and R;>) of the egg surface, he calculated the meridional and equato- rial tensions (T;, and Ty) with two formulae, mRYP=227R5T,+F and P=T,/R,}+T7/R>. Accepted April 20, 1988 E Z coe) Paes Thickness Fic. 1. Sea-urchin egg compressed to thickness (Z) between a pair of parallel plates under force (F) of compression. Ry, and R; are radii of the principal curvature of the egg surface at the equator. D is the radius of the circular area in contact with the plate. Typical curve of the force versus flattening of the egg is shown schematically to the right. Both tensions were found to increase as the egg surface is stretched upon compression, which lead him to claim the presence of elasticity at the egg surface. The method I adopted equates the work of compression, —FdZ, with the work of stretching the surface, TdS. Hence F=T(—dS/dZ), or — m TZo (—ds/dz) (1) where s=S/So is the total surface area (S) of the compressed egg divided by the initial surface area, So= 2 Zo. I calculated the tension (T) by eq. (1) employing an empirical expression of the surface area derived from measured Z, R, and R>, and found that the calculated tension now remains constant in spite of the change in surface area. 554 S O — Wn 8 E Surface Area Fic. 2. Conceptual illustration of the relation between the tension and the degree of stretching of the egg surface. A: Non-elastic surface. The tension is independent of the surface area. B: Elastic surface. The tension increases upon surface stretching. To is the initial tension when the egg is a sphere. Thus I concluded that the egg surface is not elastic, a view diametrically opposite to Hiramoto’s. Figure 2 is a conceptual illustration of the contrast between the calculated results. Thus the debate has arisen as to the validity of both procedures. I feel that the term D used by Hiramoto will not be accurately measured because the “angle of contact” is close to 180°. Conversely, Hiramoto (1976) argues that “it is questionable whether —dS/dZ could be determined with a reasonable accuracy” because “increase in surface area dS by compression dZ is very small”. Both Hiramoto and I have reinforced our own theories by supplementing with circumstantial evidence, which however, does not logically prove or dis- prove the validity of either calculation. I [6, 7] earlier found an analytical solution for the profile of the compressed egg, assuming non- elasticity of the egg surface. The calculated con- tour coincided with the actual contour of the compressed egg. Calculated values of —ds/dz were also consistent with those derived from meas- ured geometrical parameters. The agreement be- tween the calculation and observation, while sup- porting the view of non-elasticity of the egg sur- face, does not prove it however, since an elastic membrane may also show similar contours and similar values of —ds/dz. I recently found a mathematical procedure to derive the profile of elastic shells by numerical calculation. The theoretical contour thus obtained provides a means of directly evaluating the unpro- cessed raw data of the force versus compression without relying upon the geometrical parameters of the egg, as presented here. M. YONEDA CALCULATION Mathematical formulation Since the profile of the compressed egg is axially symmetric, the problem is finding the meri- dian of the egg by numerical calculation. Four assumptions can be made: that 1) the volume of the egg remains unchanged upon compression, 2) the surface does not resist bending, 3) the inner cytoplasm is fluid, and 4) hydrostatic pressure inside the egg cytoplasm does not affect the egg shape. Based on these assumptions, the egg is looked upon as a thin elastic shell encircling in- compressible fluid, to which the classical mem- brane theory of shells can be applied. Two kinds of mathematical expressions for bal- ance of forces are known to apply to the thin elastic shell loaded with external forces along the axis, as given by Timoshenko and Woinowsky- Krieger [10]. One is the familiar Laplace formula which relates the internal pressure P with the tensions (T,. and Ty) as P=T/RL+T7/Rr (2) where T, is the tension in the direction of the meridian (meridional tension) and Ty (transverse tension) is the tension in the direction perpendicu- lar to T, (Figs.3 and 4). R, is the radius of curvature of the meridian and Ry is the radius of curvature in the direction perpendicular to R; and, by the theory of differential geometry, equal to the line segment normal to the meridian cut by the symmetry axis. Another expression for axially loaded shell is did ¢ (T_X)—TrRicos # +YR_X=0 [eq. (f) on page 434 in Timoshenko and Woinow- sky-Krieger [10]]. In the case of compressed shells, the term Y, expressing the force acting tangentially to the shell, is null. Defining the orthogonal coordinates (x—y) as shown in Figure 3 in which the line x=0 denotes the symmetry axis, we obtain dX/d¢ =R,cos ¢. Hence, didX(XT,)=Tr (3) The calculations to follow are to find the profile of the egg which satisfies both eqs. (2) and (3). To Form of Compressed Sea-Urchin Eggs 55) Fic. 3. Geometric parameters describing the compress- ed shell. Parameters Ry and Ry in eq. 2 correspond to r_(i) and r+(i) in this figure. Fic. 4. Its meridional Rectangular surface element. (S_o) and transverse (S;o) dimensions in the spheri- cal shell change to S; XS; upon compression. The initial tension working at element Ty changes to T; and Ty in the meridional and transverse directions, respectively. begin with, let us suppose a rectangular surface element of the shell with the meridional and tranverse dimensions S;9 and Sto (Fig. 4A). When the shell is a sphere, the meridional and tranverse tensions coincide at Ty. Compression will change the dimension of the element to S; and S,> (Fig. 4B). “Relative elongation” E,; and Ey of the element, are defined here as E_=(Si—Sy0)/Sio, Er=(Sr—Srto)/Sto (4) Denoting the thickness and Young’s modulus of the shell by h and E respectively, T, is given as E,.+sEy Eh (+E, )(i+8,) t—s? TL=Tot+ where s is the Poisson ratio of the material of the shell and usually assigned with the value of 0.5 for rubbery materials. Now the term “elasticity in- dex”, Q=Eh/(1—s’)/Tp, which represents the magnitude of elasticity (E) relative to the initial tension To, is defined. Then, letting th=T,/To, E, +0.5Ey CE Ceee, ftp oI. Q (5) Similarly, t¢(=Ty/To) is calculated by Er+0.5E, C+EnU+E1) tr=Ty/To=1+ Q (6) All parameters used in calculations are express- ed in their relative values, z=Z/Zo, th=T /To, tr =Ty7/To, r= R{/Ro, tr=Ry/Ro, d=D/Ro and p= P R)/To, where Ro, Zo, To are the initial radius, diameter and tension, respectively, when the shell is a sphere. Expressing eqs. (2) and (3) by these parameters, we have p=ty/r_t tyr (7) and d/dx(x t_)=dt,/dx+t_=tr (8) Starting profile Calculation of the profile of the compressed shell was started by assuming a semicircular con- tour at the free surface, and by assigning a proper value of d(=D/Ro) so that the volume encompas- sed by the shell retains the volume of the sphere with a unit radius. The boundary between the surface in contact with the plate and the free surface will be called “the border”, hereafter. Fine detail of the shell was described by posi- tions in orthogonal co-ordinates of “nodes” which divide the meridian into 48 “segments” (Fig. 5). The nodes are serially numbered from 0 (pole) to 0 eM Seg as =— d= =| a Seem eerie - < 0 Se £, Sc i — Uo - +. - U — —. A 43. |B 48 Fic. 5. Form of meridians of the initial (A) and com- pressed (B) shells defined by node nos. 00 to 48. The node at the border is “node M”. Initial lengths of “segments” delimited by adjacent nodes are Sco (at the contacting surface) and S;o (at the free surface). On compression, they change to Sc and S,, respectively. Up and U are the position on the X-coordinates of the center of the segment. 556 M. YONEDA 48 (equator). Initially, they were properly spaced along the surface of the compressed shell so as to allocate either one of the nodes on the border. This is denoted by node M. Initial spacing of the nodes is Sc on the surface contacting the plate and Si on the free surface (Fig. 5B). Such a distribu- tion of the nodes on the compressed shell was transcribed back to the initial sphere, accommo- dating the spacing to different meridional lengths between the compressed and spherical shells, so that the spacing among nodes 0 to M (Sco) and the spacing among nodes M to 48 (S;0) retain the ratio Sc/S_. The original positions of the center of each segment on the X-coordinate (Up in Fig. 5) in the sphere varies among segments and is expressed as Uo(i) ((=0 to 48). The corresponding value for the position in the compressed profile is U(i). Due to the axial symmetry of the shell, the ratio U(i)/Uo(i) is numerically equal to S7/Sro (cf. eq. (4)) which is the rate of stretching of the rectangular element shown in Figure 4 along the transverse direction. In routine calculations, Ey was given as U(1i)/Uo(i). The nodes 0 to M at the contacting surface were defined by their positions in orthogonal co- ordinates as x(i) and y(i), but nodes M+ 1 to 48 at the free surface were primarily defined by r, (i) and 6 (i) for convenience in calculation, and x(i) and y(i) for i=M-+1 to 48 were given successively by r.(i) and @ (1) (see Fig. 3), starting from node M (border) as % (i)= (i— 1) + A (i) x(i)=x(i—1)+r (i) [sin ¢ (i)—sin ¢ (i—1)] y(i)=y(i—1)+r_(1) [cos ¥ (i)—cos ¥ (i—1)] S_(i) 1s directly obtained as @ (i)r_(1). Iteration for fine adjustment of the profile The compressed shell was defined in the forego- ing section by rather arbitrary parameters under given flattening (=z) and elasticity index (Q). This is the starting model to be examined for local balance of forces. First, the rates of stretching (E, and Ey) of each segment obtained by eq. (4) were used to calculate the local tensions t,(i) and t;(i) by eqs. (5) and (6). The pair of t.(i) and ty(i) determined for each segment was then examined as to whether they satisfy both eqs. (7) and (8), which are two fun- damental expressions of the balance of forces in elastic shells under compression, by calculating imbalance between the lefthand and righthand sides of the equations under given t,(i) and ty(i). According to the amount of the imbalance, neces- sary adjustments were made for parameters r (i), @(i), or d to improve the profile of the shell. Iteration of the adjustment is explained in the Appendix. The whole procedure was repeated using the renewed parameters ry(i), 6 (i) and d until the calculated profile substantially satisfied the conditions predicted by eqs. (7) and (8). RESULTS The profile of the shells under a given degree (z) of compression is determined by the elasticity index Q=Eh/(1—s’)/To as the only parameter. Hiramoto (1963) calculates Young’s modulus (E) for the surface membrane of unfertilized eggs of Hemicentrotus pulcherrimus as 1.2 x 10° dyne/cm? when the membrane thickness (h) is 3.1 sm, and the initial tension Tp is 0.032 dyne/cm. These values give the value of Q as 16. The calculations were, therefore, made of the profile of a shell compressed from z=0.90 down to 0.40, and assigned with varying elasticity indices (Q) ranging from 0 (no elasticity) to 16. The results are summarized in Table 1 and show the geometrical parameters of the compressed shell. In real eggs, the force (F) of compression is counterbalanced by the force F= 2 PD* due to internal pressure (P), and since PD? can be written as (pd?/2) mT Zo, the value of pd*/2 in Table 1 is identical to (—ds/ dz) in eq. (1). The present values of pd*/2 for Q=0 were found to coincide with —ds/dz (not shown here) derived in a previous paper [11] by an analytical solution of the profile. Typical examples of the contour of the shell under compression (z) in 4 steps for Q=0 and 16 are drawn in Figure 6. The contours in the pres- ence of elasticity (shown by circles) tend to be- come flatter near the equatorial surface than the contour in the absence of elasticity (lines), result- ing in a diminished largest diameter of the shell. Conversely, the contours of the elastic shell exhibit steeper curvatures at the region near the boundary nN nN ~ Form of Compressed Sea-Urchin Eggs TaBLE 1. Geometric parameters of elastic shells under compression z u d Pp pd? /2 z u d p pd? /2 Q= 0 Q=5 90 1.022 0.215 2.046 0.047 90 1.018 0.247 2.080 0.063 85 1.038 0.284 2.082 0.083 85 1.031 0.325 253) 0.113 .80 1.057 0.350 Dldd 0.130 .80 1.048 0.397 2.254 0.177 aS) 1.079 0.416 2.178 0.188 mS) 1.069 0.466 2.387 0.259 70 1.104 0.484 2.243 0.262 70 1.093 0.534 2.562 0.365 65 1132 0.554 2.321 0.356 65 1122 0.602 2.788 0.505 .60 1.166 0.628 2.417 0.476 60 1.156 0.672 3.080 0.696 BO) 1.205 0.708 2.536 0.635 95 1.196 0.746 3.458 0.963 50 1.250 0.794 2.683 0.846 50 1.243 0.826 3.950 1.348 oe) 1.304 0.890 2.869 1.136 45 1.299 0.915 4.596 1.924 40 1.370 0.997 3.110 1.546 40 1.365 1.016 5.453 2.815 QO 1 Q= 10 .90 1.021 0.222 2.053 0.050 90 1.015 0.270 Dll 2: 0.077 .85 1.037 0.293 2.096 0.090 85 1.027 0.354 2225 0.139 .80 1.055 0.361 2151 0.140 80 1.043 0.430 2.385 0.221 75 1.076 0.428 2.220 0.203 we) 1.063 0.501 2.602 0.326 70 1.101 0.495 2.306 0.283 70 1.087 0.568 2.889 0.465 65 1.130 0.566 2.414 0.386 65 1.116 0.633 3.264 0.654 .60 1.163 0.639 2.549 0.520 60 1.151 0.699 3a751 0.916 Sp) 1.202 0.718 2.720 0.700 AO) 1.191 0.768 4.388 1.293 50 1.248 0.803 2.936 0.946 50 1.239 0.843 9.222 1.853 45 1.303 0.897 3.214 1.293 45 1.296 0.927 6.325 pia hes) 40 1.368 1.003 3.578 1.800 40 1.364 1.024 7.798 4.085 O23 Q = 16 .90 1.020 0.229 2.060 0.054 90 1.013 0.291 2.152 0.091 .85 1.035 0.302 2.110 0.096 85 1.024 0.381 2.315 0.167 .80 1.053 0.371 2177 0.149 80 1.039 0.460 2551 0.269 wd 1.074 0.438 2.262 0.271 a fs) 1.058 0.531 2.873 0.404 70 1.099 0.506 2.370 0.303 70 1.082 0.596 3.297 0.585 65 1.128 0.576 2.507 0.416 65 1.111 0.658 3.850 0.832 .60 1.161 0.649 2.682 0.564 60 1.146 0.719 4.570 1.182 aa )s) 1.201 0.726 2.904 0.766 eB) 1.188 0.783 5.513 1.691 50 1.247 0.810 3.189 1.047 50 1.237 0.854 6.755 2.461 45 1.301 0.903 3.560 1.451 45 1.294 0.934 8.404 3.665 40 1.367 1.008 4.047 2.053 40 1.363 1.028 10.614 5.610 z: thickness, u: largest diameter, d: radius of contacting area, p: internal pressure, Q: elasticity index, Eh/(1—s*)/Tp than does the non-elastic shell. The calculated — tennis ball filled with water under compression [7], radius of the area of contact increased as Q_ when the value of Q was expected to be much increased. This character of the elastic shell was higher than 16. qualitatively identical to the behaviour of a real Such a definite and systematic dissimilarity be- 558 M. YONEDA -+- | | | -+- | | | Fic. 6. Calculated contours of elastic and non-elastic shells. Numerals indicate relative thickness (z). Short dot-dash lines indicate the axes of the shells. Solid lines: Non-elastic surface (Q=0). Circles: Elastic surface (Q=16). Arrows point to the border between the contacting and free surface. TABLE 2. (assembled by Prof. Yukio Hiramoto) Relative thickness (Z) Force of compression (F) (x 10 * dyne) 0.90 0.90 tween the two contours is, however, not very large. As is evident from Table 1, the largest diameters differ, at the most, by only 2% (at z= 0.70). Matching these contours with the actual contour of the real eggs would require high-quality pictures of several numbers of sea urchin eggs under compression before a fair judgement could be formed. Yet Table 1 provides another means of testing the presence or absence of elasticity by using the values of pd’/2 as a proportionality factor for predicting the force of compression (F). Prof. Yukio Hiramoto of the University of the Air kind- ly supplied me with his data on the force required to compress unfertilized eggs of Hemicentrotus pulcherrimus (Table 2). The calculated values of pd*/2 for each degree of compression were multi- Force of compression measured in unfertilized eggs of Hemicentrotus pulcherrimus 0.60 8.75 0.80 Zo20 0.70 4.60 plied by a certain factor (mathematically equal to ToZo) so as to get the best fit for the data (Fig. 7). As in the case of matching contours, here, too, the predicted forces of compression of elastic and non-elastic shells were not very different, both fitting substantially well with the observed data. This is not entirely unexpected, however. In such a range (z=0.6) of mild compression, the surface is only slightly stretched and the elasticity, if any, will not largely affect the issue. Data presented in my earlier paper [6] will now be utilized since it includes 50 readings on 11 unfertilized eggs of Hemicentrotus pulcherrimus (Table 3) covering the range of compression down to 0.485. For each egg, the force of compression at z=0.75 was estimated by proper intrapolation of measured forces, as indicated in Table 1 under the = iz) co Force (x 10 “dyne ) i) observed force (dots) at z=0.90 to 0.60. obtained from Prof. Yukio Hiramoto. 0.8 Relative Thickness Fic. 7. Calculated force-deformation curves fitted to Non-elastic (Q=0) surface. surface (Q= 16). TABLE 3. Form of Compressed Sea-Urchin Eggs 559 Data Solid line: Broken line: Elastic column “F7;”. To even out the variation among eggs, the measured forces were normalized so that the estimated values of F7; would come to unity. The normalized forces were plotted on a logarith- mic scale against the degree of flattening (Fig. 8) in order to critically examine the F—z curve rather than the absolute values of force. The calculated force for the shells with or without elasticity (Q=0 and 16) were similarly normalized to have a unit force at z=0.75 and also plotted on the logarithmic scale. A substantial divergence in the calculated forces between elastic and non-elastic shell was noted. Plots of measured forces are now scattered around the theoretical curve for Q=0 (solid line) and favors the claim of non-elasticity of the mem- brane. Owing to the nature of logarithmic plot- ting, the points representing small forces for mild compression (z=0.75 to 0.9) are widely scattered, yet they are located closer to the solid line (Q=0) than to the broken line (Q=16). Both calculated curves happen to be quite linear on a logarithmic Force of compression (F) and thickness (Z) measuerd in unfertilized eggs of Hemicentrotus pulcherrimus Egg F575 Zo 1 0.20 0.53 1.16 1.66 2.21 3.27 0.68 93.0 (.900) (.795) (.680) (.595) (.550) (.485) 2 0.26 0.64 0.90 1.58 2.50 0.53 94.0 (.860) (.725) (.665) (.570) (.495) 3 0.64 1.15 1.79 2°96 0.68 98.0 (.780) (.680) (.580) (.530) 4 0.38 0.92 1.56 ei: 0.75 97.5 (845) (715) (630) (575) 5 0.15 0.62 0.89 1235 1.81 0.49 97.0 (.855) (.735) (.645) (.590) (.535) 6 0.48 0.99 1.52 213 0.84 95.5 (.830) (.725) (.665) (.605) 7 0.33 0.56 0.85 1.61 0.51 93.5 (.825) (720) (.670) (570) 8 0.20 0.43 0.70 1.05 151 2.10 0.58 92.0 (.870) (.770) (.705) (.645) (.605) (.560) 9 0.31 0.67 1.00 1.63 23 0.55 93.0 (.840) (.735) (.650) (.570) (.510) 10 0.32 0.59 1.20 2.02 0.66 97.0 (.860) (.760) (.650) (.580) 11 0.29 0.67 0.95 0.37 95.0 (.780) (.665) (.600) Taken from the data presented in (6). parentheses: Relative thickness (z=Z/Zo ). proper intrapolation is shown under “F;;”. Upper rows: Force in 10~* dyne. Lower rows in Force to compress the egg to z=0.75 estimated by Zo: initial diameter(m). 560 M. YONEDA 05 log ( F/F75) fe) fe) On 05 0.6 0.8 0.9 Relative Thickness ‘(z) Fic. 8. Data measured at z=0.90 to 0.485 and the calculated forces for Q=0 (solid line) and Q= 16 (broken line) are plotted on a logarithmic scale against the relative thickness, and normalized so that the values at z=0.75 (“F;;”) come to unity. The scale bar at the upper-left corner indicates a 30% change of force. TABLE 4. Culculated slopes of the log F—Z/Zo curve Q 0 1 2 slope —2.60 —2.66 —2.72 5 8 12 16 =2.85 — 2195 —3.04 —3.12 Calculated force at z=0.75, 0.70, 0.65, 0.60, 0.55, 0.50 were linearized. scale in the low range of z (<0.75). The slopes of the (log F) —z curves are shown in Table 4. On the other hand, the slope of the linear regression curve for 36 points (in common logarithm) measured in the range of z from 0.75 to 0.48 was found to be —2.57. The error variance S*=(Syy—Syy7/Sxx)/(n —2) of the slope was 0.00065 where the “sums of products”, S,x, Sxyy and S,, were 0.178, —0.459 and 1.202, respectively, and n=36. Based on these parameters, the probable range of the slope is —2.57+0.17 (ty S*/S,, =2.7<0.061=0.17) at p <0.01. Comparison with the calculated slopes precludes the possibility that the surface of unfer- tilized sea-urchin eggs has elasticity index as high as 16, or the Young’s modulus of 10° dyne/cm? as claimed by Hiramoto [1]. A small amount of elasticity up to Q=2 is marginally probable, but making a distinction between non-elasticity and elasticity as low as Q=2 (Young’s modulus of 10° dyne/em*) appears very difficult using ex- perimental data from the compression method. Thus, the data presented do not support the pre- sumption of elasticity. REMARKS The present paper reports success in calculating the profile of spherical shells with given elasticity, which could serve to simulate deformed animal cells. The present calculation was based on sim- plifying assumptions. Among them, the assump- tions that there is no bending stress in the mem- brane and that the inner cytoplasm is fluid should be kept in mind when simulating the calculated profile with living cells. As for unfertilized eggs of the sea-urchin, I earlier estimated that the bending stress of the membrane would be negligible in the compression experiment [6]. Measurement of the Form of Compressed Sea-Urchin Eggs 561 viscoelasticity of the cytoplasm of unfertilized sea- urchin eggs by Hiramoto [12] indicates that de- formation of the cytoplasm will not create any permanent stress. Thus, I believe that the present calculation will apply at least to unfertilized sea urchin eggs. Although the present study precluded the pres- ence of elasticity as high as 10° dyne/cm* based on Hiramoto’s calculation, there are still two sets of data which seem to contradict my claim of non- elasticity. One is the observation of Hiramoto [2] on centrifuged sea-urchin eggs, and the other is my own experience with highly compressed sea-urchin eggs [13]. Critical examinations of these data have not yet been performed. ACKNOWLEDGMENT I wish to express my gratitude to Prof. Yukio Hiramo- to of the University of the Air who kindly provided me with his data on the compression experiment. I also thank Prof. Susumu Ishii of Waseda University for in- structing me on the statistical analysis of measured and calculated slopes of F—z curves. This work was sup- ported in part by a grant from the Ministry of Education, Science and Culture of Japan (61304008, 61490016). APPENDIX: DETAILS OF THE ADJUSTMENTS The parameters of the shell were modified by the following five kinds of adjustments. 1) Pressure adjustment (by eq. (7)) The internal pressure p(i) to be held by each segment was calculated by eq. (7) using t,(i) and tr(i). The calculated pressures for segments M+ 1 to 48 were averaged. Based on the premise of uniform pressure in egg cytoplasm (cf. assumption 4 given earlier). the value of r,(i) for each segment was modified so that the calculated pressures [p(i)] would come closer to their average. Throughout this adjustment, @ (i) was also modified to com- pensate for the change in r,(i) so that S; (i)[=r,(i) @ (i)] remains unchanged with the hope that the balance of tensions, which is largely susceptible to change in the length of the segment (S,), is not drastically disturbed. 2) Tension adjustment at the free surface (by eq. (8) In numerical calculation, the value dt, /dx in eq. (8) was replaced by t,(i+1)—t,(i) divided by (x(i+1)—x(i—1))/2. The balance of forces at each node (M+ 1 to 48) was examined by putting t,(i), tr(i), and x(i) into eq. (8). According to the imbalance thus detected, the position of each node (M+1 to 48) was shifted along the meridian by modifying @ (i) to improve the balance. 3) Tension adjustment at the contacting sur- face (by eq. (8)) The tension balance at each node on the contact- ing surface was adjusted in a procedure similar to adjustment 2 except that the modification was made to x(i) (i=1 to M—1), instead of @ (i). 4) Tension adjustment at the border (by eq. (8)) Separate adjustments of the tension balance for the contacting and free surfaces (adjustments 2 and 3) will concentrate the imbalance at the border (node M) where the contacting and free surfaces meet. Since the two segments neighboring the border are assigned different resting lengths, Sco and S;o, the adjustment at the border was achieved by shifting the position of node M along the surface of the initial sphere, which inovlves reciprocal changes in Sco and S;o with the accom- panying modification of E,, t, and ty. This results in reducing the imbalance between the two seg- ments at the border. This adjustment of the resting lengths induces a subtle disturbance of the balance in all but one (border) nodes which are carried over to subsequent iterations. 5) Volume adjustment Adjustments 1 and 2 usually change the calcu- lated volume of the shell. To return it to its initial volume, a certain factor was multiplied to all sets of r_(i) at the nodes M+1 to 48, and x(i) of nodes 1 to M. A modification of the radius d of the contacting surface is performed for this occasion. The iteration consisting of these 5 separate adjustments, each involving a negative feedback of imbalances to the original parameters ry(i), 6 (1) and d, was repeated several times until a) the pressures calculated for all segments became uni- form within 0.01%, b) tension imbalances between any adjacent segments became less than 0.1% and c) deviation of the volume from unity became less than 10°. The chief difficulty encountered in such a scheme of “multiple adjustments” was that any 562 M. YONEDA (1) Pressure { (5) volume | (2) tension (free surface) (503%) (5) mie (80%) (3) Tension (contacting surface) (70%) (4) eee (border ) (1003) (2) een (free surface) (50%) ' (5) Volume Fic.9. Iteration of adjustments. Percentages in pa- rentheses are the rate of negative feedback to para- meters. one of the adjustments (pressure, for example) often tended to amplify the imbalance of other aspects (tension for example) resulting in a never- ending oscillation of the calculated profile, even when only moderate changes were imposed on original parameters in a single step of adjustment by limiting the rates of feedback to low levels. A scheme for a successive sequence of adjustments which modified the profile steadily and quickly to its equilibrium was fortunately found after several trials. Figure9 is the final version of present iteration in which the adopted rates of negative feedback are specified in percentages. Numerical calculations were automatized by a computer program called “ELAS”, written in Basic and adapted to a personal computer (Hitachi S1-40). tO 10 11 13 REFERENCES Hiramoto, Y. (1963) Mechanical properties of sea urchin eggs. I. Surface force and elastic modulus of the cell membrane. Exp. Cell Res., 32: 59-75. Hiramoto, Y. (1976) Observations and measure- ments of sea urchin eggs with a centrifuge micro- scope. J. Cell. Physiol., 69: 219-230. Hiramoto, Y. (1970) Rheological properties of sea urchin eggs. Biorheology, 6: 201-234. Hiramoto, Y. (1976) Mechanical properties of sea urchin eggs. III. Visco-elasticity of the cell surface. Dev. Growth Differ., 18: 377-386. Hiramoto, Y. (1987) Evaluation of cytomechanical properties. In "Cytomechanics”. Ed. by J. Bereiter- Hahn, O.R. Anderson and W.-E. Reif, Springer, Berlin. pp. 31-46. Yoneda, M. (1964) Tension at the surface of sea- urchin eggs. A critical examination of Cole’s experi- ment. J. Exp. Biol., 41: 893-906. Yoneda, M. (1973) Tension at the surface of sea urchin eggs on the basis of ‘liquid drop’ concept. Ady. Biolphys., 4: 153-190. Yoneda, M. (1976) Temperature-dependence of the tension at the surface of sea-urchin eggs. Dev. Growth Differ., 18: 387-389. Cole, K. S. (1932) Surface force of the Arbacia egg. J. Cell. Comp. Physiol., 1: 1-9. Timoshenko, S. and Woinowsky-Krieger, S. (1970) Theory of Plates and Shells, McGraw-Hill, New York, 2nd ed., pp. 434. Yoneda, M. (1986) The compression method for determining the surface force. Methods Cell Biol., 27: 421-434, Hiramoto, Y. (1969) Mechanical properties of the protoplasm of the sea urchin egg. I. Unfertilized egg. Exp. Cell Res., 56: 201-208. Yoneda, M. (1980) Tension at the highly stretched surface of sea urchin eggs. Dev. Growth Differ., 22: 39-47. ZOOLOGICAL SCIENCE 5: 563-572 (1988) Effect of Hexyleneglycol on Meiotic Division of Starfish Oocytes WAKAKO YAMAO and TAIKO MIKI-NOUMURA Department of Biology, Ochanomizu University, Ohtsuka, Tokyo 112, Japan ABSTRACT—Following 1-methyladenine treatment, maturing oocytes of the starfish, Asterina pecti- nifera, were transferred to sea water containing 2.5% hexyleneglycol (HG). After 20-30 min, the meiotic apparatus (MA), at first clearly visible as a tiny point close to the cell membranes, began to increase in size. After a time, most of the larger MA became situated perpendicular or parallel to the cell membranes, but in the case of a few oocytes near the cell center. The diameter of the aster of the isolated MA at metaphase of HG-treated oocytes two or three times longer than that in the control experiment. Depending on its particular location in the oocytes during maturation, the larger MA induced three different events during meiotic division: giant polar body formation, unilateral furrowing in most oocytes, and equal cell division in a few oocytes. The frequency of these events depended upon when the oocytes were transferred to HG-SW during maturation. Based on these findings, we attempted to present an explanation for the events induced by HG. When maturing oocytes were transferred to HG-SW, HG diffused gradually into them, stabilizing the migrating MA in its particular position during maturation, and causing its size to increase through the reassembly of microtubules around it. The larger MA may cause the meiotic division of oocytes to change, leading to the subsequent inducement of the above three events, depending on their particular © 1988 Zoological Society of Japan location in the oocytes during maturation. INTRODUCTION Mazia and Dan in 1952 [1] were the first to succeed in isolating the mitotic apparatus (MA) from synchronous dividing sea urchin eggs, and confirmed the MA to be a rigid structure which appears and then disappears during mitosis. They accomplished this in two steps: light fixation of dividing cells by application of cold ethanol and dispersion of cytoplasm around the MA with de- tergents. After several attempts, Kane [2] finally devised a simple one-step isolation method, treating a synchronous culture of dividing sea urchin eggs with 1 M (12.8%) hexyleneglycol (HG), maintain- ing pH at 6.4. He reported that various non-sulfur containing six-carbon glycols could be used for the isolation. Rebhun and Sawada [3] found that the volume and birefringence of the MA of sea urchin eggs increases with the addition of HG at meta- Accepted May 14, 1988 phase. In their study of the effect of HG on MA, Endo et al. [4] recently found that the MA of sea urchin eggs in sea water containing 5% HG, possesses a surprisingly large number of microtu- bules conspicuously uniform in length. Studying the polymerization process of porcine brain tubulin in the presence of HG, we recently found that HG promotes microtubule polymeriza- tion, shifting the equilibrium between micro- tubules and tubulin to the polymer side, and accelerating the initial velocity of the polymeriza- tion (Yamao and Miki-Noumura, unpublished data). In consideration of the above, we attempted to bring about some change in the polar body forma- tion of starfish oocytes by HG, since HG stabilizes MA [2] and induces an increase in the volume and birefringence of MA [3]. Using starfish oocytes, HG was first noted to cause additional microtu- bules to reassemble around the meiotic apparatus (MA), causing it to become much larger than that in the control. Depending on the particular posi- 564 W. YAMAO AND T. MIKI-NOUMURA tion of MA in the oocytes during maturation, the larger MA induced the occurrence of three distinct events during meiotic division: formation of a giant polar body, unilateral furrowing or equal cell division. MATERIALS AND METHODS Materials A starfish, Asterina pectinifera, was used, and a sea urchin, Hemicentrotus pulcherrimus, was used for a preliminary experiment. Immature oocytes of the starfish were prepared by treating the ovaries with Ca-free artificial sea water (CaFSW), in order to remove follicular envelopes. The oocytes were pooled and sus- pended in sea water (SW) at 17°C, after washing several times with CaFSW. More than 95% of these isolated oocytes were intact and immature. Sea urchin eggs were obtained by injecting 0.5 M KCI into the body cavity. Shedded eggs were suspended in SW at about 18°C, after washing several times with SW. Hexyleneglycol (HG) treatment The isolated immature oocytes were suspended in SW containing 5x 10° M 1-methyladenine (1- MeAde) to induce meiosis. periods, from 20 to 50 min, equal volumes of SW containing 5% hexyleneglycol were added to the suspended SW of the oocytes. The oocytes were cultured continuously in it. To observe the polar body formation clearly, the fertilization mem- branes were sometimes elevated, by treating the oocytes with 10mM caffeine for 5 min, before suspending in 1-MeAde-SW. Such a 5 min treat- ment in caffeine was confirmed not to induce a parthenogenetic response, only the elevation of the membrane, as reported by Obata and Nemoto [5]. Observation took place under a_phase- contrast (XF-ph) and a high sensitivity polarizing differential (recti-Nomarski, HPO) microscope (Nikon Optical Co., Tokyo). The phase-confrast microscope was equipped with a Nikon PFMB After various set camera, and photographs were taken on Neopan F film (Fuji Co., Tokyo) with an exposure time of 1/4 -1/30 sec. Isolation of meiotic apparatus (MA) We have used the word “MA” here for the meiotic apparatus of starfish oocytes, because Dan and co-workers used “MA” for meiotic apparatus of Spisula oocytes in their papers [6]. Two methods, using hexyleneglycol (HG) or glycerol/DMSO, were used for isolation of the MA, as described previously [2, 4]. Compositions of the isolation medium were as follows: HG isolation medium (15% (V/V)HG, 1 mM EGTA, 10mM KH>PO,4, pH 6.2) and glycerol/DMSO- isolation medium (1M_ glycerol, 10% (V/V) DMSO, 1 mM EGTA, 0.5 mM MgCh, 1% (V/V) Nonidet P—40, 10 mM MES pH 6.2). Oocytes at the desired stage were collected by low-power centrifugation washed twice with ten volumes of 1M dextrose, to favor dispersion of cytoplasm. The pellet of oocytes was suspended in ten volumes of the isolation medium. The suspen- sion was shaken up and down several times in a centrifugal tube to disperse cytoplasm around the MA. The isolated MA was then collected by low-power centrifugation. RESULTS Observation under a phase-contrast microscope Confirmation was first made on the effects of hexyleneglycol (HG) by using fertilized sea urchin eggs, as previously reported by Endo et al. [4]. When fertilized eggs reached prometaphase or metaphase in the first cleavage, they were transfer- red to artificial sea water containing 5% HG (HG- SW). After 15 min in HG-SW, a clear zone appeared around the mitotic apparatus (MA) of prometaphase eggs, encircling the MA and causing its width to increase with time (Fig. la). The two asters of the metaphase MA became larger (Fig. 1b), and could be discerned more clearly. They appeared to become slightly separated from each other during the 15min in HG-SW. The two larger asters apparently consisted of two larger monasters side by side, resembling a pair of glasses (Fig. 1c). Based on these observations, we next attempted to bring about some changes in the polar body formation of starfish oocytes using HG. Effect of HG on Meiosis 565 Fic. octNpM Fe As is already known, immature oocytes of starfish start meiosis in sea water within 20 min at 20°C, following the addition of 1-methyladenine (1-MeAde). The germinal vesicles in the oocytes begin to break down in 1-MeAde-SW and then disappear. The first polar body is formed after 70- 80 min (Fig. 2a, b), and the second one, 110-120 min after 1-MeAde treatment. In this experiment, the concentration of HG in SW was determined to be 2.5%, after various concentrations from 1 to 10% HG had been added to the SW. The oocytes were transferred to 2.5% HG-SW after 40min of 1-MeAde treatment. Each oocyte possessed a tiny meiotic apparatus (MA) which could initially be clearly discerned as a small point under a phase-contrast microscope. In about 15min, the MA of HG-treated cells Fertilized sea urchin (Hemicentrotus) eggs. 15 min after suspension in 5% HG-SW at prometaphase. 15 min after suspension in 5% HG-SW at metaphase. 30 min after suspension in 5% HG-SW at metaphase. Phase-contrast micrographs. Bar represents 100 um. became distinct and larger, with its orientation and shape quite evident (Fig. 3a). The oocytes in the control had a smaller MA situated close to the cell membranes and visible as a clear point. After 30 min, the larger MA of HG-treated oocytes appeared to consist of two larger asters, side by side, resembling a pair of glasses, as was also noted in HG-treated eggs of the sea urchin. The larger MA was situated close to the cell membranes, and apparently attached to the cell cortex. The long axis of the MA was directed parallel or perpen- dicular to the cell membranes (Fig. 3b, c). When the long axis of the MA became perpen- dicular to the cell membranes (Fig. 3d), the larger MA then continued to grow and after about 70-80 min, extruded a giant polar body, which was about 10-18 ~m in diameter. Its length was about two or Fic. 2. Polar body formation of the oocytes in the control experiment. Extruded polar body in control experiment about 60 min after 1-MeAde treatment. a: lower magnification, b: higher magnification. Phase-contrast micrographs. Bar represents 100 “am. 566 W. YAMAO AND T. MIKI-NOUMURA Fic. 3. Polar body formation of the oocytes suspended in 2.5% HG-SW . Giant polar body extruded from the oocytes. a: 30min after HG-SW treatment. b: 30min. Larger MA oriented perpendicular to the cell membrane. c: 30 min. Larger MA oriented parallel to the cell membrane. d-f: Extrusion process, d: 40 min, e: 90 min, f: 140 min, g and h: 180 min, after HG—SW treatment. Bar represents 100 «m, which is shown for photos at higher magnification. Phase-contrast micrographs. i-k: Isolated meiotic apparatus (MA) from starfish oocytes. i: MA in the control experiment. j and k: Larger MA isolated from HG-treated oocytes. Bar represents 20 ~m. Phase-contrast micrographs. Effect of HG on Meiosis 567 three times that of the control (compare Fig. 3g, h and Fig. 2a, b). Immediately beneath the extruded polar body, one larger aster continued to remain and never disappeared (Fig. 3e, f). In no case did a HG-treated oocyte extrude a second polar body, in contrast to the control oocytes. For a comparison of the MA in HG-treated and control oocytes, MA were isolated from both sources as described in “Methods”. The MA in the control consisted of a spindle body and two smaller asters after 40 min in 1-MeAde. In some cases, a normal aster about 6 “m in diameter was present on one side. The other aster had its flattened end on astral rays attached by some vesicles or mem- brane (Fig. 3i). MA isolated from HG-treated oocytes exhibited well-developed asters on both sides of the spindle, which was about 9 to 10 ym in diameter, and increased to about 20 um after 30 min of HG-SW treatment. The MA appeared to be comprised of a spindle body and two larger asters of the same diameter (Fig. 3}, k). The aster with its flattened end on the astral rays could not be found in the MA of HG-treated oocytes. With its long axis parallel to the cell membranes, MA continued to grow, and the cell surface im- mediately above the middle point of the MA gradually became concave (Fig. 4A). The concav- ity progressed more on one side with time, in a manner similar to the cleavage furrowing of medu- sa (Spirocodon) eggs. The two well-developed asters were situated just under the concave cell surface on both sides of the furrow region (Fig. 4B, C). Unilateral furrowing did not divide the oocytes completely, and later, the furrow some- times ceased to be apparent. The MA was often situated near the cell center of HG-treated oocytes, but the reason is unknown at present. It continued to grow, becoming much larger than previously observed in giant polar body formation and unilateral furrowing. The MA consisted of two larger asters. The chromosomes were observed near the equator region, and some- times formed vesicles or lumps. The cell surface over the middle of the MA along the long axis became concave on both sides (Fig. 4D, E). The cleavage furrow became deeper on both sides of the cell surface, often constricting the oocytes equally (Fig. 4F). Each divided cell contained a Fic. 4. A-C, Unilateral furrowing of the oocytes suspended in HG-SW. A: 50 min, b: 100 min, c: 140 min after suspension in HG-SW. D-F, Equal cell division of oocytes suspension in HG-SW. D: 120 min, E: 200 min, F: 230 min, after suspended in HG-SW. Bar represents 100 ~m. Phase-contrast micrographs. 568 W. YAMAO AND T. MIKI-NOUMURA well-developed larger MA, with vesicles or lumps of chromosomes. The MA failed to disappear even upon completion of division. The two equally divided cells ceased to undergo further division. Time course of events in HG-treated starfish oocytes Following HG-SW treatment of oocytes during maturation, the following three events were observed to occur: formation of a giant polar body, unilateral furrowing and equal cell division. When the oocytes were transferred to HG-SW after 40 min of 1-MeAde treatment, HG at 2.5% in SW induced formation of a giant polar body after 80 min of 1-MeAde treatment. The frequency of occurrence of this event was measured every 10 min. Using 100 oocytes, we found that it was 85% at 100 min, decreasing to 20% at 130 min. It increased again, reaching 40% at 160 min. Uni- lateral furrowing first began after 80 min, and reached a maximal value of 30% at 200 min. It decreased again at 240 min. Equal cell division was first noted at 160 min and gradually occurred in 5% of the oocytes at 200 min. These decreases in frequency of occurrence with time may possibly have resulted from recovery to the beginning state, due to the incomplete division of the giant polar body or incomplete unilateral furrowing. When the oocytes started constricting again, the occur- rence of a giant polar body formation and unilater- al furrowing peaked for a second time. In order to further examine the above events, we studied the effects of HG-SW on oocytes, GVBD 0 10 20 30 40 a a a iali extending the time of HG treatment from 20 to 50 The procedure is shown in Figure 5. The relationship between the frequency of these events and time in HG—SW is illustrated in Figure 6. The frequency was maximal at about 100 min and 140- 160 or 200 min following 1-MeAde treatment. The increasing and decreasing trends were essentially the same as those of the control during meiotic division, corresponding to the formation of first and second polar bodies, in spite of some delay in HG-treated oocytes. It thus appeared that the cycle of maturation division may possibly exert some effect on the occurrence of these events which were induced by HG. The frequency of giant polar body formation depended on the par- ticular time at which the oocytes were subjected to HG-SW treatment. Later treatment resulted in a higher frequency. When oocytes were transferred to HG-SW after 50 min of 1-MeAde treatment, polar body formation occurred in about 90% of the oocytes, while transfer after 20 min reduced it to about 40% during a period of 100 min. Essentially the same results occurred at 140-150 min, when the second peaks or second maximal values were observed. As shown in Figure 6a, transfer at 50 min resulted in a frequency of about 90%, but only 5-8% at 20 min. Unilateral furrowing was maximal or peaked two times, at 100 and 200 min, the latter exceeding the former. A transfer time of 30 min following 1-MeAde-SW treatment resulted in a frequency of 40%, while at 50 min, the frequency was 20% at min. Ist pb. Homogeneous eoindle 50 min 225% HG in ASW 1-MeAde Fic. 5. The experimental procedure. Effect of HG on Meiosis 569 Bigger polar body a ) 100 Unilateral furrow b O 100 200 min. Equal division Cc lo 20 HG edd 0 100 200 min Fic. 6. Time course of frequency of three changes induced by HG. The oocytes were transferred to HG-SW at various time, from 20 to 50 min after 1-MeAde treatment. @ 20 min, © 30 min, #40 min, 450 min after 1-MeAde treatment. Arrows indicate the transferring time of the oocytes to HG-SW. Ordinate: Frequencies (%) of each event. Abscissa: Time (min) after 1-MeAde treatment. a: Giant polar body formation. b: Unilateral furrowing. c: Equal cell division. the second peak during 200 min (Fig. 6b). The equal division peaked only once at 200 min. It attained a maximum of 20% during 200 min, as a result of a transferring time of 20 min, but only 2- 3% at 50 min, as shown in Figure 6c. The frequen- cies of both unilateral furrowing and equal division were greater with early HG—SW treatment, but polar body formation was more frequent with later treatment. The frequencies of these events seem to be a function of the time of HG-SW treatment. In spite of the considerable deviation due to in- complete meiotic synchronization in_ starfish oocytes, the relationship is adequately apparent. That is, greater unilateral furrowing and equal division resulted from earlier HG—SW treatment, while later treatment induced more frequent giant polar body formation. Next, the occurrence of these events was observed in the HG-treated oocytes at 40 min following 1-MeAde treatment, to determine which events occurred when the MA was in a certain location or with a certain orientation in the oocytes. About sixty oocytes were continuously observed under a recti-Nomarski microscope (a high sensitivity polarizing differential interference contrast microscope). The position of the MA in oocytes was found by measuring its distance from the center of the oocytes after determining the focal plane. In order to represent where the MA situates in the oocyte, the sphere of oocyte was divided into two parts. The radius of the oocyte was expressed as 0% at the center and 100% at the cell surface. One area of the oocyte was occupied by radius from 0% to 50%, and another area, by radius from 50% to 100%. In about 90% of the oocytes, the MA was situated near the cell surface, i.e., in the area occupied by 50-100% radius of an oocyte to the cell surface. Unilateral furrowing and giant polar body formation were observed in 90% of the oocytes situated in that area. In a few cases, the MA was found near the center of the oocytes, that is, in the area occupied by 0-50% radius from the cell center, reaching S-7%. The location of the MA in the oocytes was substantially consistent with the frequencies of the three events, and thus may be closely related. DISCUSSION Effects of hexyleneglycol on the MA At meiotic division, the MA could be clearly seen as a tiny point close to cell membranes in 570 W. YAMAO AND T. MIkI-NOUMURA starfish oocytes under a phase-contrast micro- scope. It grew with time, with its long axis perpendicular to the cell membranes. The small area of the cell cortex to which the astral rays were attached, appeared as being pushed out; it was finally extruded as a polar body at the first meiotic division. Based on a study of Spisula oocytes, Dan [6] has proposed that “anchorage” of the spindle to the cortex at its one pole is a common phe- nomenon in the unequal division of polar body formation. As shown in “Results”, when the oocytes were transferred to HG—SW after 1-MeAde treatment, three distinct events occurred as a result: giant polar body formation, unilateral furrowing and equal cell division. Observation of HG-treated oocytes indicated that the MA at metaphase grew in size. The extruded polar body had a diameter of about 10-18 «am, which was two or three times that of a normal polar body. The MA at the time of unilateral furrowing and equal cell division was much larger than that in the control. Although not investigated in detail by electron microscopy, the astral rays in the MA appeared similar to those of the MA in sea urchin eggs. In Figure 4D and E, the MA has a remarkable number of microtubules very fine in appearance. The asters contained many microtubules of uniform length, thus consti- tuting a layer or membrane surrounding the MA in dividing cells. The two large asters in the MA had the appearance of a pair of glasses, surrounded by a clear distinct zone. This was also observed in sea urchin eggs by Endo et al. [4]. HG-—SW induced growth of the MA by causing the number of microtubules around it to increase. In fact, a preliminary study of HG-effects on tubulin polymerization of porcine brain, indicated that HG shifts the equilibrium between the micro- tubules and tubulin to the polymer side and further accelerates the initial velocity of the polymeriza- tion (Yamao and Miki-Noumura, unpublished data). This shows HG to be the cause for the greater number of microtubules of uniform length around the MA in dividing cells, possibly by rapid organization of many microtubule nucleating cen- ter around the MA, to induce additional microtu- bule assembly, and to shift the equilibrium be- tween tubulin and the microtubules to the poly- mer side. Time course of the three events in oocytes induced by HG-SW The following events were induced by 2.5% HG in SW: giant polar body formation or unilateral furrowing in most of the oocytes, and equal cell division in a few cases. The frequency initially increased, reached a plateau, decreased and then increased again, as shown in Figure 6. The time course of this increase and decrease was essentially in agreement with that of meiotic division, the first and the second polar body formations. The cell cycle of oocyte maturation may have some effect on the time course of these events. As described in “Results”, in most cases, the long axis of the MA was directed parallel or perpendicular, close to the cell membranes after a break-down of the germinal vesicles. In 90% of the oocytes, the MA was close to the cell mem- branes, in spite of the perpendicular or parallel long axis orientation. Giant polar body formation and unilateral furrowing were also induced in these oocytes. Unilateral furrowing was maximal at 30 min after 1-MeAde treatment. Equal division was so at 20 min, and giant polar body formation at 50 min. Although incompletely synchronous meiosis in the starfish oocytes may possibly bring about a time deviation in the occurrence of each event, the order of occurrence appeared to be, equal divi- sion, unilateral furrowing, and giant polar body formation. This order suggests the location and orientation of the MA after a break-down of the germinal vesicles. It is initially parallel to the cell membranes and then becomes perpendicular to them. In the former case, it would grow in size through the action of HG, followed by inducement of the unilateral furrow at the cell cortex in the middle of the MA. In the latter case, the increase in size of the MA would possibly cause extrusion of a giant polar body, attached with astral rays to the small area of cell cortex. In about 5% of the oocytes, the MA may be situated near the cell center and may not migrate to the cell surface following the break-down of germinal vesicles. The MA may become larger at the cell center, thus inducing a furrow in the middle of the oocytes, and causing them to divide in half. Effect of HG on Meiosis In this present experiment, HG diffused grad- ually into the oocytes following their transfer to HG-SW, and stabilized the migrating MA in situ. Furthermore, HG promoted reassembly of micro- tubules in oocytes and resulted in greater growth of the MA which may be the possible cause of the three distinct events. The results of the present research are presented schematically in Figure 7, although nothing definite can be said about equal division owing to the low frequency of its occur- rence. E. B. Wilson’s book, The Cell in Develop- ment and Heredity, presents figures describing MA orientation during meiosis [7], which are consistent with the present ideas concerning the MA during A “ ' © ie S OO Fic. 7. Proposed explanation of our results. ie 571 meiosis in starfish oocytes. As mentioned above, Dan has proposed that the migration of the meiotic spindle to the cell cortex and its becoming situated close to it are phenomena that occur in the case of polar body formation, based on observation of the meiotic division of Spisula oocytes [6]. However, we have so far been unable to directly observe the migration of the MA to the cell cortex after the break-down of germinal vesicles in starfish oocytes. Examination of the processes of meiotic division of starfish oocytes through electron mi- croscopy should provide further clarification of this phenomenon. 4 c - C Normal maturation process is shown by 1-4, enclosed in the box. A—C: Events induced by HG. A: Unilateral furrowing. B: Formation of a giant polar body. C: Equal cell division. 572. W. YAMAO AND T. MIkKI-NOUMURA ACKNOWLEDGMENTS We thank Miss Atsuko Hanayama for her kind help in preparing this manuscript. Thanks are also due to the staff of Tateyama Marine Laboratory, Ochanomizu Uni- versity, for providing facilities and materials for this study. This work was supported in part by a Grant-in- Aid for scientific research from the Ministry of Educa- tion, Science and Culture, Japan. 1 ine) REFERENCES Mazia, D. and Dan, K. (1952) The isolation and biochemical characterization of the mitotic apparatus of dividing cells. Proc. Natl. Acad. Sci., 38: 826-838. Kane, R. (1962) The mitotic apparatus: Isolation by controlled pH. J. Cell Biol., 12: 47-55. 3 Rebhun, L. I. and Sawada, N. (1969) Augmentation and dispersion of the in vitro mitotic apparatus of living marine eggs. Protoplasma, 68: 1-22. Endo, S., Toriyama, M. and Sakai, H. (1983) The mitotic apparatus with unusually many microtubules from sea urchin eggs treated by hexyleneglycol. Dev. Growth Differ., 25: 307-314. Obata, C. and Nemoto, S. (1984) Artificial parthe- nogenesis in starfish eggs: Production of parthe- nogenetic development through suppression of polar body formation by methylxanthines. Biol. Bull., 166: 525-536. Dan, K. and Ito, S. (1984) Studies of unequal cleav- age in Molluscs: I. Nuclear behavior and anchorage of a spindle pole to cortex as revealed by isolation technique. Dev. Growth Differ., 26: 249-262. Wilson, E. B. (1925) The Cell in Development and Heredity, Macmillan, New York, pp. 498-503. ZOOLOGICAL SCIENCE 5: 573-584 (1988) © 1988 Zoological Society of Japan Gamete Interactions and Sperm Incorporation in the Nemertean, Cerebratulus lacteus FRANK Lonco, Wa us H. Ciark, Jr.! and GERTRUDE W. HINSCH? Department of Anatomy, University of lowa, lowa City, IA 52252, ' Bodega Marine Laboratory, University of California, P.O. Box 247, Bodega Bay, CA 94923, and ?Department of Biology, University of South Florida, Tampa, FL 33620, U.S.A. ABSTRACT — Light and electron microscopic observations have been carried out with Cerebratulus gametes prior to and immediately following sperm-egg fusion. Cerebratulus eggs released into sea water underwent germinal vesicle breakdown and a massive exocytosis of cortical vesicles concomitant with the elevation of a chorion. Contact of the sperm with the egg surface induced the acrosome reaction. This was followed by gamete membrane fusion which occurred between the acrosomal process and an egg microvillus. Although the base of the microvillus involved in gamete fusion enlarged to permit entry of the spermatozoon, a fertilization cone failed to form and actin filaments did not accumulate at the site of sperm incorporation. Morphological changes in the egg cortex and extracellular coverings, that might be involved with a block to polyspermy, were not apparent. INTRODUCTION Sperm and eggs of the nemertean worm, Cere- bratulus lacteus, are released from male and female animals into their respective burrows. Cur- rents then carry the gametes into the open sea water where fertilization takes place [1]. Fertiliza- tion of Cerebratulus eggs can also be achieved in the laboratory as animals are easily maintained in sea water aquaria and gametes may be collected as described by Kume and Dan [2] and Costello et al. [3]. Eggs dissected from animals are in the germi- nal vesicle stage and if fertilized become polysper- mic. If left uninseminated in sea water, however, the germinal vesicle breaks down and a chorion lifts from the surface of the egg. Eggs fertilized at this stage are monospermic [2]. Wilson [4] noted that a “favorable character” of Cerebratulus eggs for cleavage studies is the absence of a fertilization membrane. The lack of a morphological correlate that might be involved with a block to polyspermy is in agreement with recent observations by Kline et al. [5] who demonstrated electrical changes in the plasma membrane of Cerebratulus eggs that Accepted March 17, 1988 are involved in polyspermy prevention. The structure and chemistry of Cerebratulus sperm have been enigmas. Afzelius [6] described a distinct acrosome in the sperm of the nemertine, Malacobdella grossa. The presence of an acro- some in Cerebratulus sperm, however, could not be morphologically substantiated by Olds and Au- stin [7] nor could Metz [8] demonstrate sperm lysins even though eggs possess a heavy investing layer (chorion). Considering the features of this evolutionary distinctive organism [9] and its gametes, it is surprising that C. /acteus has not been used more extensively for investigations of fertilization and embryogenesis. In light of this paucity of informa- tion, we have carried out observation on Cerebra- tulus gametes prior to and immediately following sperm-egg fusion. MATERIALS AND METHODS Collection of Gametes Cerebratulus lac- teus were generously provided by Drs. Douglas Kline and Laurinda Jaffe. Methods for handling adults and obtaining gametes were as described by Costello et al. [3]. Eggs were obtained from 0.5 to 574 F. Lonco, W. H. CLark and G. W. HINSCH 1.0cm segments cut from the posterior end of female worms. The segments were cut several times in sea water to release eggs which were then collected with a Pasteur pipette, filtered through cheesecloth and suspended in 50 ml sea water until fertilized. Sperm were obtained from a 0.5 to 1.0cm segment cut from the posterior end of a male worm just prior to insemination. The segment was suspended in 2 ml sea water and cut to release sperm. Sperm were diluted to a final concentration of about 1:10,000 for insemination. Electron Microscopy Segments from the posterior end of female worms, unfertilized eggs suspended in sea water for | to 60 min and insemi- nated ova were fixed in 3% glutaraldehyde in sea water at 4°C for lhr. Fixed specimens were washed in sea water overnight, post-fixed in 0.5% OsO, in sea water at 4°C for 1 hr, dehydrated with ethanol and embedded in Spurr’s embedding medium [10]. Thin sections stained with uranyl acetate and lead citrate were examined with a Philips EM 300 transmission electron microscope. Fluorescence Microscopy Fertilized eggs were fixed for 1 hr at 4°C with 3% paraformal- dehyde in sea water, washed for 2 hr in sea water containing 50 mM NH,Cl, and rinsed for 5 min in 0.1% Triton X-100 in sea water. To visualize the DNA of incorporated sperm nuclei, inseminated eggs were stained for 30min at 37°C in 10 um Hoechst 33342 (Sigma) in phosphate-buffered saline (2.9mM NaH>,PO,, 7mM Na,HPO, and 136.9mM NaCl, pH7: PBS) and washed three times in PBS. Filamentous actin was localized in Hoechst-stained zygotes with rhodamine- phalloidin [11]. Specimens were stained in 0.5 yvg/ml rhodamine-phalloidin for 30min, rinsed three times in PBS, and mounted in 90% glycerol. Preparations were observed with a Nikon inverted microscope fitted with an epifluorescence attach- ment. RESULTS Structure of the Egg Cortex and Extracellular Matrix The cortices of ovarian eggs, which were at the germinal vesicle stage of meiosis, were filled with yolk granules and numerous vesicles containing a filamentous material (Fig. 1). The plasma mem- branes of these eggs were reflected into numerous microvilli (Fig. 1) that projected into a narrow (0.3-0.4 um) perivitelline space; consequently, the microvilli were bent and usually only seen in cross and oblique section. The extracellular layers of eggs consisted of a thin envelope or chorion, composed of filamentous material, and immediate- ly superficial to this layer, a jelly layer. The innermost region of the jelly layer was composed of filamentous material that formed a coarse re- ticulum; this graded into a region containing a more dispersed material. Egg Changes on Exposure to Sea Water When Cerebratulus eggs were exposed to sea water, they initiated germinal veiscle breakdown; meiotic maturation progressed to metaphase I and then arrested. Concomitantly, changes in the egg cortex and extracellular layers took place. The jelly layer expanded and the chorion elevated enlarging the perivitelline space (Figs. 2 and 3). Fic. 1. Section of an ovarian Cerebratulus egg. The cortex is projected into microvilli (M) and filled with numerous vesicles (V) that are released upon the egg’s exposure to sea water. Extracellular structures associated with the egg’s surface include a chorion (C) and a jelly layer consisting of a dense internal reticulum (Jp) and a more flocculent outer region (Jp). x 15,000. Fic. 2. Fic. 3. Nomarski micrograph of an egg exposed to sea water depicting the elevation of its chorion (C) and numerous scalloped structures along its surface which represent regions of dehiscing vesicles (arrows). Section of an egg exposed to sea water in which the jelly layer has swollen (not depicted); the chorion (C) has x 310. elevated and vesicles within the cortex have begun to exocytose (arrow). In contrast to microvilli (M) of ovarian eggs, those of ova released into sea water project at more-or-less right angles from the egg surface. x 19,000. Fic. 4. Cortex of an egg 60 min after its exposure to sea water which lacks the numerous vesicles found in ovarian specimens. x 29,000. Typically the bases of microvilli that project from the egg surface are constricted (arrows). w ww Fertilization of Cerebratulus Eggs 576 F. Lonco, W. H. Ciark and G. W. HINscH This elevation accompanied the exocytosis of vesi- cles contained within the egg’s cortex. The release of these vesicles appeared to be random, in groups, and involved extensive portions of the egg surface (Fig. 2). By 15 min virtually all of the vesicles had exocytosed and the egg cortex was essentially devoid of such structures (Fig. 4). With the elevation of the chorion and expansion of the perivitelline space, mirovilli became erect and oriented perpendicular to the egg surface. Struc- turally, microvilli at this stage were approximately 1 ym in length by 0.2 «m in diameter. Characteris- tically, the bases of microvilli were constricted to a region 0.05 ~m in diameter. Internally, the micro- villi lacked a prominent microfilamentous core (Fig. 4). Sperm Structure Cerebratulus sperm consisted of a distinct head, mitochondrial-middle piece and flagellum. The head was a tapering crescent-shaped structure, approximately 16 «zm in length, containing an acro- some and nucleus (Fig.5). The nucleus, which made up most of the sperm head, consisted of uniformly electron dense chromatin. The acrosome was an elongated ellipsoid about 1 um in length by 0.3 «m in diameter (Figs. 6 and 7). Its posterior aspect was invaginated to form a fossa approximately 0.2 um deep. Within the fossa and the space separating the acrosome and nucleus was accumulation of electron dense, postacrosom- al material. The middle piece of the sperm possessed four mitochondria (2 4m in length by 0.5 ~m in diam- eter), which were situated in depressions along the base of the nucleus (Figs. 8 and 9). Approximately one third the length of the mitochondria extended posteriorly, surrounding a cytoplasmic compart- ment in which were found the proximal and distal centrioles (Fig. 8). The proximal centriole was situated partially in a fossa at the posterior end of the sperm nucleus (Fig. 8). The distal centriole was oriented perpen- dicular to the proximal; only its anterior region was situated within the space circumscribed by the mitochondria. The portion of the distal centriole which was not surrounded by mitochondria was associated with nine pericentriolar processes that projected from its triplet tubular wall and termi- nated in thickened tips (Figs. 8 and 10). In cross section the nine pericentriolar processes defined a wheel-like structure that marked the anterior mar- gin of the sperm tail (Fig. 10). Sperm-Egg Interaction and the Acrosome Reaction Examination of living specimens revealed that sperm were capable of swimming in and out of the perivitelline space, i.e., through the chorion that surrounded the egg. During its transit through the chorion the acrosome remained intact (Figs. 11 and 12). An acrosome reaction did not occur until the sperm apex came into contact with an egg microvillus (Fig. 13). Fusion of the sperm plas- malemma and acrosome membrane occurred at the apex of the sperm head, thereby releasing the acrosomal contents (Fig. 13). An acrosomal pro- cess formed via the inversion of the former acro- somal membrane and the modification of postacro- somal substance. The acrosomal processes in specimens that appeared to have just undergone the acrosome reaction were associated with some filamentous material (Figs. 13 and 14). This mate- rial joined the acrosomal process to the egg sur- face. Gamete Fusion and Sperm Incorporation Gamete fusion occurred via the tip of the acro- somal process and a microvillus (Fig. 15). Inter- estingly, the base of the microvillus involved in gamete membrane fusion enlarged before the con- tents of the spermatozoon moved through it, into the egg cortex (Fig. 15). Other than an enlarge- ment of the base of the microvillus involved in gamete fusion, no other specializations of the egg cortex were found in association with the incorpo- rating sperm, e.g., the appearance of microfila- ment bundles and a fertilization cone (Fig. 16; see also Figs. 20-22). In addition to an absence of fertilization cone formation in metaphase I eggs, immature, germinal vesicle oocytes also failed to develop such a cortical specialization (Figs. 17- 19). When inseminated eggs were treated with rho- damine-phalloidin there was a general diffuse staining throughout the cytoplasm indicating the presence of filamentous actin (Figs. 23 and 24). Fertilization of Cerebratulus Eggs 577 ‘ SA ey x Ex Fic. 5. Longitudinal section of a Cerebratulus sperm depicting its crescent-shaped nucleus (N), acrosome (A), mitochondria (M) and proximal elements of the flagellum (F). x 9,000. Fics. 6 and 7. Longitudinal sections through the acrosome (A) and anterior of the sperm nucleus (N). Between these two organelles is some electron dense material (arrows) that surrounds the anterior aspect of the nucleus and projects into a small fossa at the posterior of the acrosome. Fig. 6, x 37,000; Fig. 7, « 59,000. Fics. 8 and 9. Longitudinal- and cross-sections of Cerebratulus sperm along the region of the middle piece. The anterior aspect of the four mitochondria (M) fit into shallow invaginations along the posterior of the sperm nucleus (N). The posterior aspect of the mitochondria define a compartment which contains two centrioles. The proximal centriole (PC) is located in a shallow fossa at the posterior of the nucleus. Just posterior to the proximal centriole is the distal (DC) from which the axonemal complex (Ac) and pericentriolar processes project (P). Figs. 8 and 9, x 28,000. Fic. 10. Cross section of a spermotozoon at the level of the distal centriole. Nine pericentriolar processes, which project from the distal centriole, are depicted. x 40,000. 578 F. Lonco, W. H. CLark and G. W. HINScH Fic. 11A-C. Serial sections of a sperm (S) traversing the chorion (C) which has lifted from an egg surface. The intact acrosome (A) of this specimen is depicted in Figs. 11C and 12. An accumulation of rhodamine-phalloidin staining was not observed in association with the incorpo- rated spermatozoon. Once fusion had taken place the entire cytoplas- mic contents of the sperm moved into the cortical ooplasm (Figs. 20 and 21). The sperm flagellum extended from the surface of the egg (Fig. 22). In the present study we were unable to determine the eventual fate of this structure. It is interesting to note, however, that the tips of the sperm pericentriolar processes were closely associated with the zygote plasmalemma (Fig. 22). x 2,900. DISCUSSION Other than recent investigations of the electrical properties of eggs following insemination and aspects of protein synthesis during early develop- ment [12, 13], Cerebratulus have been little used for contemporary studies of fertilization processes. The present observations provide structural fea- tures of gamete interactions leading to the acro- some reaction, sperm-egg fusion and sperm incor- poration in the nemertean, Cerebratulus. These results amplify much earlier light microscopic studies in this phylogenetically unusual organism and have a direct bearing on fertilization mecha- Fic. 12. acrosome (A) is intact. Fic. 13. x 20,500. Anterior aspect of the sperm shown in Fig. 11C which has penetrated the chorion surrounding an egg. The Anterior aspect of a spermatozoon that has contacted the surface of an egg and has undergone the acrosome reaction. The membranous sleeves (MS), that consist of the outer acrosomal membrane and the plasmalemma, partially surround the forming acrosomal process (P). Some filamentous material joins the acrosomal process and the egg surface. x 41,000. Fic. 14. Fic. 15. Anterior aspect of a sperm in contact with an egg microvillus (M) via its acrosomal process (P). Sperm that is in the process of fusion with an egg. Gamete fusion oocurs at the tip of the acrosomal process x 52,000. (P) and an egg microvillus (M). Gamete fusion results in slight enlargement at the base of the microvillus (*). x 41,000. Fic. 16. x 33,500. Fused sperm and egg. The site of cytoplasmic continuity of the sperm and egg is depicted by the arrow. Fertilization of Cerebratulus Eggs 579 580 F. Lonco, W. H. CLark and G. W. HINscu Fics. 17 to 19. "ee Stages of sperm incorporation (S) into germinal vesicle (GV) containing eggs. An elevation of egg cytoplasm, a fertilization cone and characteristically observed at the site of gamete fusion in other organisms, e.g., sea urchins, fails to form in inseminated, immature Cerebratulus eggs. Eggs inseminated at metaphase I of meiosis in which sperm incorporation is not associated with the Fics. 20 and 21. formation of a projection of cytoplasm, the fertilization cone. x 750. MA, meiotic apparatus; M, microvilli; S, incorporated sperm. Fig. 20, « 750; Fig. 21, * 27,000. nisms in other organisms. Eggs Cerebratulus eggs are enveloped by distinct extracellular coats, a chorion and a jelly layer. Upon exposure to sea water the chorion lifts from the surface of the egg. The concomitant dehiscence of cortical vesicles with the elevation of the chorion suggests that these two processes may be linked. That is, the vesicle contents may become hydrated upon dehiscence thereby ex- panding the perivitelline space and lifting the chorion. A similar mechanism has been shown to be involved in the elevation of the vitelline layer to form the fertilization membrane in sea urchins Fertilization of Cerebratulus Eggs 581 Fic. 22. Cortex of an inseminated egg depicting the posterior aspect of an incorporated sperm nucleus (N), mitochondria (M), and axonemal complex (Ac). The surface of the egg immediately superficial to the incorporated sperm does not show signs of fertilization cone formation. Portions of the incorporated pericentriolar processes are shown at the arrows. 28,000. Fics. 23 and 24. Fluorescence micrographs of an inseminated egg stained with Hoechst for DNA (Fig. 23) and rhodamine-phalloidin for filamentous actin (Fig. 24). The inseminated egg shows a diffuse staining for filamentous actin; specific rhodamine-phalloidin-staining in the vicinity of the incoporated sperm nucleus (arrow) is not apparent. The maternal chromatin is out of the plane of focus in Fig. 23. x 450. [14]. changes, at the site of sperm incorporation or Unlike sea urchins, however, the elevation of along other regions of the egg cortex that might be the chorion does not occur as a result of sperm-egg — involved with a block to polyspermy, were not interaction nor does the elevated chorion act asa = observed. These data are in agreement with barrier to sperm. Sperm were seen to move morphological observations of Kline ef al. [5] who through this structure with apparently little effort | also demonstrated a rapid positive-going shift in [see also 5]. Sperm-egg fusion is not associated | membrane potential coincident with insemination. with a cortical reaction in this species. Structural They suggest that this change in potential acts as a 582 F. Lonco, W. H. CLark and G. W. HInscu block to polyspermy which lasts for approximately 60 min in Cerebratulus. Sperm The general structure of the sperm of Cerebratulus resembles that observed in the nemertean, Malacobdella grossa [6]. Elaborate pericentriolar processes radiate from the distal centrioles of Cerebratulus sperm; structures similar to those described in other species [15]. In the sperm of Malacobdella these structures are de- scribed as radiating “rods” that terminate in an electron dense annulus [6]. In Hydactinia sperm, the elements that project from the distal centriole contain actin [15]. The acrosome of Cerebratulus sperm consists of a vesicle and distinct subacrosomal fossa contain- ing electron dense postacrosomal material. The organization of this material is morphologically similar to that described for sperm of Malacobdella [6]. We have observed the acrosome reaction in Cerebratulus sperm and its is functionally similar to that observed in other invertebrate sperm [16, 17]. Interestingly, the acrosome reaction in Cerebratu- lus does not take place when sperm swim through the jelly coat and chorion. Instead, this reaction is induced by an association of the sperm with the egg surface. In every instance examined, acro- some reacted sperm were found in close associa- tion with egg microvilli, suggesting that a microvil- lus-sperm interaction leads to the release of acro- somal contents and the formation of an acrosomal process. It is interesting to note that Metz [8] was unable to demonstrate sperm-lysins in Cerebratu- lus. If in fact lysins are required for sperm penetration through the chorion one would expect their release prior to/during chorion passage by the sperm. In addition, sperm attachment and fusion with the egg occurs at a microvillus, the latter apparently induces the resumption of meiotic maturation [3]. A similar sperm-egg relationship has also been described for Chaetopterus [18]. Sperm Incorporation The present obser- vations substantiate and extend previous light mi- croscopic studies [19-21] demonstrating that sperm incorporation in Cerebratulus is not accom- panied by the formation of a fertilization cone in either germinal vesicle or metaphase I eggs. Sperm incorporation in sea urchin germinal vesicle (immature) eggs is distinguished by the formation of extremely large fertilization cones in contrast to those formed by mature (pronuclear) ova [22]. Prior to sperm-egg interaction, microvilli pos- sess constricted bases and do not appear to have a distinctive microfilamentous core. Although the base of the microvillus involved in gamete fusion enlarges slightly, the formation of a cytoplasmic projection, as observed in sea urchin and other invertebrates and vertebrates [22], does not occur in Cerebratulus. In other species that have been studied, namely sea urchins and mice [23-25], actin filaments accumulate in the cytoplasm sur- rounding the entering sperm. In sea urchins the accumulation of microfilaments is believed to be derived from the in situ polymerization of monomeric actin [26, 27], which in turn aggregate to form large bundles that are organized with the same polarity [28]. Such an accumulation of filamentous actin in Cerebratulus zygotes was not detected either by rhodamine-phalloidin staining or electron microscopy. These observations indicate that fertilization cone formation and the accumulation of filamen- tous actin are not obligatory for sperm incorpora- tion. They are similar to previous investigations of fertilized sea urchin eggs treated with cytochalasins [23, 25]. When sea urchin eggs are treated with cytochalasins immediately after gamete fusion, fer- tilization cone formation and actin polymerization fail to occur. Nevertheless, the sperm nucleus and mitochondria enter the egg cortex and male pro- nuclei form [23]. The inhibition of microfilament and fertilization cone formation in sea urchin zygotes treated with cytochalasins also suggests that these processes are coupled [see 28], i.e., fertilization cone formation is dependent upon actin polymerization [22, 27]. The absence of fertilization cone formation in Cerebratulus may result from a failure of actin to polymerize and accumulate in the region of sperm incorporation. The results presented here and those employing sea urchin zygotes treated with cytochalasins [22, 25] beg the question of possible mechanisms in- volving the movement of the sperm contents into the egg cortex. Fertilization of Cerebratulus Eggs ACKNOWLEDGMENTS This paper is dedicated to Professor Katsuma Dan, an international scientist whose many outstanding research contributions have had major influences in cellular and developmental biology. The authors wish to thank Drs. Laurinda Jaffe, Doug- las Kline and Frederick Griffin for the contribution of their time and valuable discussions during the course of this study. The assistance of Ms. Susan Cook and the typing of Ms. Tena Perry are gratefully acknowledged. This investigation was supported by funds from the NIH (F.J.L.) and National Sea Grant College Program, De- partment of Commerce, under grant number NA9SAA- D-SG140 project number R/A-61 (W.H.C.). 10 11 REFERENCES Wilson, C. B. (1900) The habits and early develop- ment of Cerebratulus lacteus (Verrill). A contribu- tion to physiological morphology. Q. J. Microsc. Sci., 43: 97-198. Kume,M. and Dan,K. (1968) Invertebrate Embryology. Nolit Publishing House, Belgrade, Yugoslavia. Costello, D. P., Davidson, M. E., Eggers, A., Fox, M. H. and Henly, C. (1957) Methods for Obtaining and Handling Marine Eggs and Embryos. Lancaster Press, Inc., Lancaster, PA. Wilson, E. B. (1903) Experiments on cleavage and localization in the nemertine egg. Arch. Entwick- lungsmech., 16: 411-460. Kline, D., Jaffe, L. and Tucker, R. P. (1985) Ferti- lization potential and polyspermy prevention in the egg of the nemertean, Cerebratulus lacteus. J. Exp. Zool., 236: 45-52. Afzelius, B. (1971) The spermatozoon of the nemertine, Malacobdella grossa. J. Submicrosc. Cytol., 3: 181-192. Olds, P. J. and Austin, C. R. (1967) Entry of Cere- bratulus spermatozoa into Echinarachnius eggs. Biol. Bull., 133: 477. Metz, C. B. (1957) Specific egg and sperm sub- stances and activation of the egg. In “Beginnings of Embryonic Development”. Ed. by A. Tyler, R. C. Von Borstel and C. B. Metz. Am. Assoc. Adv. Sci., Washington, D.C., pp. 23-69. Barnes, R. D. (1980) Invertebrate Zoology. Saun- ders, Philadelphia, PA. Spurr, A. K. (1969) A low viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res., 26: 31-43. Wieland, T. and Faulstish, H. (1978) Amatoxins, phallotoxins, phallolysin and antamanide: The biologically active components of poisonous Amani- 15 16 17 26 2] 583 ta mushrooms. CRC Crit. Rev. Biochem., 5: 185- 260. Candelas, G. and Monroy, A. (1968) Protein synthe- sis during maturation and early development of the egg of Cerebratulus lacteus. Exp. Cell Res. , 52: 664- 667. Kline, D., Jaffe, L. A. and Kado, R. T. (1986) A calcium-activated sodium conductance contributes to the fertilization potential in the egg of the nemer- tean worm Cerebratulus lacteus. Dev. Biol., 117: 184-193. Schuel, H. (1978) Secretory function of egg cortical granules in fertilization and development. A critical review. Gam. Res., 1: 299-382. Kleve, M. G. and Clark, W.H. (1980) Association of actin with sperm centrioles: Isolation of centriolar complexes and immunofluorescent localization of actin. J. Cell Biol., 86: 87-95. Dan, J. C. (1967) Acrosome reaction and lysins. In “Fertilization, vol. 1”. Ed. by C. B. Metz and A. Monroy. Academic Press, New York., pp. 237-293. Tilney, L. G. (1975) The role of actin in nonmuscle cell motility. In “Molecules and Cell Movement”. Ed. by S. Inoué and R. E. Stevens. Raven Press, New York., pp. 339-388. Anderson, W. A. and Eckberg, W.R. (1983) A cytological analysis of fertilization in Chaetopterus peramentaceus. Biol. Bull., 165: 110-118. Yatsu, N. (1904) Experiments on the development of egg fragments in Cerebratulus. Biol. Bull., 6: 123-136. Yatsu, N. (1909) Observations on ookinesis in Cere- bratulus lacteus Verrill. J. Morphol., 20: 353-401. Chambers, R. (1933) The manner of sperm entry in various marine ova. J. Exp. Biol., 10: 130-141. Longo, F. J. (1987) Fertilization. Chapman and Hall, New York. Longo, F. J. (1980) Organization of microfilaments in sea urchin (Arbacia punctulata) eggs at fertiliza- tion: Effects of cytochalasin B. Dev. Biol., 74: 422- 433. Maro, B., Johnson,M.H., Pickering,S.J. and Flach, G. (1984) Changes in actin distribution dur- ing fertilization of the mouse egg. J. Embryol. Exp. Morphol., 81: 211-257. Cline, C. A. and Schatten, G. (1986) Méicrofila- ments during sea urchin fertilization: fluorescence detection with rhodaminyl phalloidin. Gam. Res., 14: 277-291. Spudich, A. and Spudich, J. A. (1979) Actin in Triton-treated cortical preparations of unfertilized and fertilized sea urchin eggs. J. Cell Biol., 82: 212- 226. Cline, C. A., Schatten, H., Balezon, R. and Schat- ten, G. (1983) Actin-mediated surface motility dur- ing sea urchin fertilization. Cell Motility, 3: 513- 584 F. Lonco, W. H. CLark and G. W. HINscH villi, and the fertilization cone of sea urchin eggs. J. 524. Cell Biol., 87: 771-782. 28 Tilney, L. G. and Jaffe, L. A. (1980) Actin, micro- ZOOLOGICAL SCIENCE 5: 585-601 (1988) © 1988 Zoological Society of Japan Centrosomes, Centrioles and Post-translationally Modified Microtubules during Fertilization HEIDE SCHATTEN, CATHY HOWARD, GERARD COFFE, CALVIN SIMERLY and GERALD SCHATTEN Integrated Microscopy Resource for Biomedical Research, University of Wisconsin, 1117 West Johnson Street, Madison, WI 53706, U.S.A. I. INTRODUCTION Fertilization requires an elaborately choreo- graphed sequence of motions, all of which are effected by rearrangements of the cytoskeleton (review in [1]). The first example of an alteration of the cytoskeleton is found in the sperm as the microfilaments of the acrosomal process assemble ({2—5] for review). This occurs while the microtu- bules in the sperm axoneme are sliding and bend- ing to produce the waveform motion which propels the sperm through the insemination fluid ([6, 7] for review). Once the sperm contacts and binds to the egg surface (review in [8, 9]) microfilaments at the egg cortex (reviewed by [10]) as well as actin binding proteins including fodrin [11] serve a critical role in the physical incorporation of the sperm in some, but not all [12], fertilization systems. The fertilized egg surface erupts with elongated microvilli (re- viewed by [13]) and clathrin-coated endocytosis has been observed by Fisher and Rebhun [14]. As the sperm nucleus decondenses into the male pronucleus (reviewed by [15, 16]), typically a monastral array of microtubules, the sperm aster, assembles. The sperm astral microtubules perform the crucial role of uniting the male and female pronuclei. In many fertilization systems the two pronuclei fuse during interphase, though in some systems such as mammals ({17] for review) and ascidians [18], the pronuclei do not fuse during interphase. Instead the male and female chroma- tin condense into individual chromosome sets at prophase of first mitosis. The completion of ferti- Accepted March 17, 1988 lization, signalled by the intermingling of the parental genomes is formally achieved at meta- phase of first mitosis when the chromosomes are aligned on the metaphase plate. Since most eggs are fertilized as oocytes rather than as mature pronucleate eggs (reviewed by [19]), the functioning of microtubules during the completion of meiotic maturation and the motility of cortical microfilaments in the constriction of the polar bodies must also be included in a considera- tion of cytoskeletal reorganization during fertiliza- tion. The foundation for the elucidation of cytoskele- tal activity during fertilization has already been laid, and it is now possible to begin a molecular characterization of the responsible proteins and the post-translational modifications of these mole- cules throughout the cell cycle. The objectives of this chapter are to evaluate the composition, be- havior and mode of inheritance of the centrosome, to investigate the status of the centriole during fertilization and to consider post-translational modifications of a-tubulin during meiosis, fertiliza- tion and mitosis. Il. CENTROSOMES A. Paternal Inheritance of Centrosomes The centrosome is the structure which specifies the configurations of assembling microtubules, and in doing so, defines cell polarity either by shaping mitotic spindle axes or by controlling the organiza- tion of the interphase microtubule array. Boveri [20] recognized the central importance of the cen- trosome and proposed that the centrosome is one 586 H. SCHATTEN, C. Howarp et al. of the vital components contributed by the sperm to the egg at fertilization. In this view, the egg loses its centrosomes during oogenesis and there- fore cannot normally organize a bipolar mitotic spindle after artificial activation. By the paternal contribution of centrosomes, the fertilized egg is able to first organize the microtubules comprising the monastral sperm aster, and after duplication, the bipolar mitotic apparatus. While this classic view of centrosomal inheri- tance was well documented by Boveri’s hematoxy- lin images, the ability to directly trace centrosome configurations has only recently been rediscov- ered. Calarco-Gillam ef al. [21] found an autoim- mune serum which reacts specifically with pericentriolar material and this serum has been used to investigate centrosome shape in a wide variety of somatic and germ cells ranging from plants [22] to eggs and embryos of sea urchins [23] as well as mice [21, 23, 24]. Recently a mouse monoclonal antibody to a 68 kD protein has been shown by Schatten et al. [25] to react specifically with centrosomes from sea urchin eggs and embryos. This monoclonal antibody may well overcome many of the limitations of routine availability and sufficient quantity necessary for a more complete characterization of the centro- some. The organization of the centrosome during ferti- lization and first division in sea urchins is shown in Figure 1. The unfertilized egg does not bind the monoclonal antibody to the 68kD antigen. However one or two spots are usually found at the base of the sperm head, and this material is introduced to the egg at sperm incorporation. As the microtubules of the sperm aster develop, the centrosome is found as a compact sphere at the center of the aster and adjacent to the male pronucleus (Fig. 1A). It then spreads around the decondensing male pronucleus, forming into an arc (Fig. 1B) and it enlarges and spreads to oppos- ing poles of the zygote nucleus by the streak stage (Fig. 1C). The centrosome undergoes a_ characteristic change in shape during mitosis. At prophase and metaphase, the centrosomes are compact spheres (Fig. 1D). By anaphase the centrosomes flatten (Fig. 1E). At telophase the centrosomes enlarge into large ovoid structures; the direction of spread- ing is perpendicular to the axis of the first mitotic axis and parallel with the axes for the next mitosis. Evidence that the centrosome is contributed by the sperm at fertilization is rather strong in a wide variety of animals. In systems as diverse as cten- ophores [26] through amphibians [27] a monastral sperm aster is organized from a site adjacent to the sperm nucleus and few other microtubules are observed after the loss of the second meiotic spindle. Antitubulin immunofluorescence micros- copy in sea urchins [28-30], clams [31] and ascid- ians [18] provide more direct evidence that micro- tubules are predominantly organized around the sperm centrosome after fertilization. However mammalian fertilization may not follow this pat- tern. B. The Question of Maternal Inheritance of Cen- trosomes in Mammals The observation that microtubules in fertilized mouse eggs were organized by numerous cytoplas- mic sites first leads to the suggestion that the origins of the microtubule organizing centers Fic. 1. Centrosomes during sea urchin fertilization and division. At fertilization the centrosome is introduced into the egg by the sperm and it is detected here with a monoclonal antibody to Drosophila intermediate filament protein. Initially the centrosome is quite compact and in tight association with the male pronucleus (A). It organizes the radially symmetric sperm aster. Later (B), the centrosome disperses into an arc and the sperm aster enlarges. At the end of first interphase (C), at 60 min post-insemination, the centrosome separates in two; asters are associated with each centrosome. At metaphase (D), the centrosomes are quite compact, as the characteristic mitotic spindle with well elaborated asters and a fusiform spindle develops. The metaphase chromosomes are aligned at the spindle equator. During the remainder of the mitotic cycle, the centrosomes expand and enlarge in the direction perpendicular to the spindle axis (E). The asters enlarge as microtubules clear from the astral centers. The separated chromosomes decondense into karyomeres, which will fuse to reconstitute the blastomere nuclei. B-E are triple stained for microtubules (MTs: left), centrosomes (CENTROS: middle) and DNA (DNA: right). A: double stained for microtubules and DNA with a corresponding centrosome image at the same stage. Lytechinus pictus. Bars: 10 um. A, D and E: reprinted, with permission, from [55]. Centrosomes and Modified a-Tubulin H. ScHATTEN, C. Howarb et al. CENTROS Centrosomes and Modified a-Tubulin 589 CENTROS Fic. 3. Centrosomes in parthenogenetically activated mouse oocytes. The centrosomes in parthenogenetically activated mouse oocytes behave in the same fashion as those during normal fertilization. At the time for mitotic prometaphase (A) they aggregate to form a bipolar barrel-shaped spindle and by telophase they separate and move to the poleward faces of the blastomere nuclei (B). These observations support the hypothesis that the centrosome in this mammal is maternally inherited. Images are double stained for DNA (left) and centrosomes (right). Bars: 10 ~m. Fic. 2. Centrosome during mouse fertilization and division. The unfertilized mouse oocyte, which is ovulated arrested at second meiotic metaphase, displays numerous centrosomal foci in addition to those found at the poles of the meiotic spindle (A). Each centrosomal focus is associated with a microtubule-containing cytaster. The centrosome does not appear to be contributed by the sperm at fertilization and instead these maternal centrosomal foci associate with each pronucleus during interphase (B). They duplicate by the completion of first interphase and at prophase (C) aggregate and move towards the cell center in apposition with the condensing chromosomes. At metaphase (D), the centrosomes are positioned to form broad spindle poles and a barrel-shaped anastral spindle emerges. At telophase (E), the centrosomes are found on the poleward faces of the decondensing blastomere nuclei. All images are triple stained for microtubules (MTs: left), DNA (DNA: middle) and centrosomes (CENTROS: right). Bars: 10 wm. Reprinted, with permission, from [12]. 590 H. SCHATTEN, C. Howarp et al. (MTOCs) in this mammal did not follow the expected pattern [32]. By the use of the autoim- mune serum reactive with centrosomes [21], Schat- ten et al. [32] hypothesized that the mouse centro- some is maternally inherited. This hypothesis was based on the observations that mouse sperm did not react with this autoim- mune serum, while numerous cytoplasmic foci were detected in the unfertilized oocyte. Each centrosomal focus organized a cytaster, detectable with antitubulin labeling, and larger foci were After sperm incorporation, when a single prominent sperm aster is expected, the mouse egg displays numerous cytoplasmic asters, each with their own centrosomal particles (Fig. 2B). The incorporated sperm nucleus does not appear to be associated with any centrosomal particles initially. The male pronucleus associates with the cytoplasmic centro- somal particles at the same time as the female pronucleus associates with them. It is likely that there is an attraction of centrosomes for interphase nuclear surfaces, and that this interaction con- tributes to the motions which result in pronuclear apposition. At the end of first interphase (Fig. associated with larger asters (Fig. 2A). CENTROS Fic. 4. 2C), the centrosomal particles aggregate near the condensing chromatin at the cell center. First mitosis in the mouse is quite unusual. Instead of the typical fusiform mitotic spindle with associated asters, the mouse spindle at metaphase is often barrel-shaped and devoid of asters (Fig. 2D). It is more reminiscent of a plant cell spindle [33, 34] than that expected in animal cells. This observation is reinforced by the fact that the mouse spindle at first mitosis is organized in the apparent absence of functional centrioles [32]. This topic of centrioles is considered in the next section. The hypothesis that the centrosomes are mater- nally inherited in mice can be experimentally tested. One prediction is that parthenogenetically activated unfertilized oocytes should contain all the information necessary to form a normal bipolar mitotic apparatus: the centrosomes which serve as the mitotic poles should not require any paternal contribution. In Figure 3, mouse oocytes parthe- nogenetically activated by ethanol are found to organize normal bipolar mitotic spindle (Fig. 3A, B) which divide the eggs from one into two at a frequency of greater than 50%. This is quite Aggregation of the sea urchin centrosomes at prometaphase into a single particle. When sea urchin eggs are transferred to ice temperatures at prometaphase, the two centrosomes aggregate into a single compact particle. After recovery from the cold for only 30 sec, a monaster of microtubules is observed. The chromosomes remain condensed, but separate from the compact centrosome. Triple stained for centrosomes (CENTROS: left), microtubules (MTs: middle) and DNA (DNA: right). Bar: 10 um. Fic. 5. Centriole appearance during mouse embryogenesis. Centrioles are not found in unfertilized oocytes (A) or in blastocysts as studied with 0.5 ym sections using high voltage electron microscopy. The centrosomes appear as dense osmiophilic material at the spindle poles. By seven-days after mating, centrioles are found at spindle poles (B). The centrioles which appear at this stage have the typical orthogonal pattern of 9+0 triplet microtubules. Bars: A: 1 ym, B: 100 nm. Centrosomes and Modified a-Tubulin 591 592 H. ScHATTEN, C. Howarp et al. unlike the situation in lower animals in which parthenogenesis can occur but the spindles that emerge are typically irregular and only rarely bipolar. C. Physical Treatments Modify Centrosome Con- figurations Since the shape of the centrosome is critical in the establishment of a microtubule array, it is of great importance to understand the forces which direct centrosome shape and positioning. A recent collaborative effort has discovered the remarkable effect of physical treatments on centrosome shape [35]. In Figure 4, centrosomes are demonstrated to collapse on each other due to the effects of prolonged cold treatments. Sea urchin eggs at prometaphase, when the two mitotic centrosomes are rather condensed, are incubated for over eight- een hours at 0°C. The microtubules of the mitotic apparatus largely disassemble, the chromosomes remain in their condensed state, and the two centrosomes coalesce into one compact particle. This observation suggests that centrosome shape might well be altered by physical treatment, much in the same way that the state of chromosome condensation can be modified. Il. CENTRIOLES A. Centriole in Mouse Embry- ogenesis Appearance Centrioles are thought to follow a pattern simi- lar to centrosomes during gametogenesis: they are retained at spermatogenesis and lost during oogenesis. High voltage electron microscopy (HVEM) by Rieder et al. [36] demonstrates ele- gantly the progressive loss of centrioles during maturation in starfish oocytes. The absence of centrioles in mouse oocytes was first shown by Széllési et al. [37]. Wooley and Fawcett [38] noted the absence of centrioles in rat sperm. HVEM observations (Fig. 5) confirm the finding that centrioles are absent in unfertilized mouse oocytes and show the dense osmiophilic material at the poles of the meiotic spindle charac- teristic for centrosomes (Fig. 5A). In light of the unusual “plant-like” shape of the first mitotic apparatus, the ultrastructure of the spindle poles was examined [32]. Surprisingly a structure similar to the sperm centriole was found in association with the remnants of the sperm axoneme, but centrioles were not observed at the spindle poles. This leads to the suggestion that the first mitotic spindle in the mouse is organized in the absence of centrioles. This suggestion raises questions about the functions of centrioles in animal cells as well as their mode of inheritance and timing of appear- ance during embryogenesis. To address this latter question, mouse embryos and fetuses were fixed and processed for HVEM at progressively later stages. The advantages of mil- lion volt high-voltage electron microscopy for this investigation includes the ability to locate cen- trioles in thick sections more rapidly than would be possible with conventional thin sections studied with transmission electron microscopy. At all stages until seven days after mating, quite late in fetal development, centrioles have not been lo- cated. At this time, centrioles are found in a few cells and it appears that centrioles emerge slowly and asynchronously during mouse development Fic. 6. Acetylated a-tubulin antibody localization during completion of second meiosis. At ovulation, mouse oocytes are arrested at metaphase (A); tyrosinated a-tubulin antibody (YL1/2) staining reveals microtubules of the anastral barrel-shaped spindle and small cytoplasmic asters heavily decorated (B) while the acetylated a- tubulin antibody only weakly recognizes the polar ends of the spindle (C). At anaphase (D), the chromosomes begin migrating to their respective poles; tyrosinated a-tubulin antibody (YL1/2) and acetylated a-tubulin antibody are indistinguishable in their recognition of the spindle and forming interzonal microtubules (E and F). By late telophase (G), the midbody connecting the polar body with the egg proper has replaced the spindle apparatus and is equally recognized by both a-tubulin (H) and acetylated a-tubulin antibodies (1). Note, however, that acetylated a-tubulin antibody does not stain the cytoplasmic asters during meiosis (I). In addition to the meiotic midbody, the sperm axoneme remains acetylated during interphase (J and K). All images triple labeled for DNA, glutamate (H) or tyrosinated (B, E) a-tubulin antibody, and acetylated a-tubulin antibody. DNA (left panel), rat or rabbit antitubulin antibody (center panel), and acetylated a-tubulin antibody (right panel) on metaphase, anaphase, and late telophase of meiosis in the mouse. Bars: 10 «am. Centrosomes and Modified a-Tubulin 593 594 H. SCHATTEN, C. Howarp et al. (Fig. 5B). Magnuson and Epstein [39] have de- scribed the appearance of centrioles in cultured mouse blastocysts at approximately six days after mating. IV. POSTTRANSLATIONAL MODIFICA- TIONS OF a-TUBULINS A. Modifications and Their Implications for the Fertilization Process Microtubules are composed primarily of heter- odimer subunits of a-tubulin and f-tubulin, as well as numerous microtubule associated proteins (MAPS; reviewed by [40]). Recently it has been shown that tubulin can be post-translationally modified in a few ways including the acetylation of lysine in a-tubulin [41-44], and by the removal and later readdition of the carboxyterminal tyrosine of a-tubulin [45-48] referred to as detyrosination and tyrosination. The mouse oocyte at fertilization presents a useful model for exploring the mechanisms re- sponsible for the new appearance of some microtu- bules and the selective disappearance or stabiliza- tion of others. The oocytes are ovulated arrested at second meiotic metaphase and the resumption of meiosis Occurs upon activation. New cytoplas- mic microtubules proliferate after sperm entry during interphase. The microtubules comprising the incorporated sperm axoneme and the meiotic spindle disassemble with the exception of those destined to form the midbody between the oocyte and the second polar body [32]. B. Acetylation a-Tubulin in the microtubules of mouse oocytes and embryos is acetylated and detyrosinated in a specific spatial and temporal sequence. In unferti- lized oocytes, which are arrested at second meiotic metaphase (Fig.6A; DNA: Hoechst dye 33258 DNA fluorescence), comparisons of total microtu- bule image (Fig. 6B; MTs: microtubules) with the binding pattern of acetylated-a-tubulin antibody (Fig. 6C; Ac-a-T: acetylated-a-tubulin) show that the acetylated form is found primarily at the spindle poles though weaker staining is noted throughout the spindle; the cytoplasmic asters (Fig. 6B; arrows) are not acetylated. At meiotic anaphase (Fig. 6D), the spindle microtubules (Fig. 6E) are heavily decorated with the acetylated antibody (Fig. 6F). At the completion of second meiosis (Fig.6G), when cytasters (Fig. 6H: arrows) and the midbody between the oocyte and second polar body are apparent, only the midbody is acetylated (Fig. 61; MB: midbody). The mouse sperm axoneme contains acetylated- a-tubulin, expected from the studies of Piperno and Fuller [44] who investigated sperm from sever- al species, including mammals. After sperm incor- poration and during the phases of pronuclear development and apposition, the axoneme retains its ability to bind the antibody to acetylated-a- tubulin (Fig. 6J and 6K). The midbody which resulted from the completion of second meiosis retains its ability to bind the acetylated-a-tubulin antibody. However comparisons of the total tubu- lin image (Fig. 6H) with that for acetylated-a- Fic. 7. Acetylated a-tubulin localization during first mitosis. DNA (left panel), affinity purified porcine brain antitubulin antibody (middle panel), and acetylated a-tubulin (right panel) from prophase through telophase in mouse oocytes. At prophase, the male and female chromatin condense separately (A) as each pronucleus is enveloped by a sheath of microtubules (B); this dense array of microtubules is not acetylated (C). By prometaphase, the chromatin has fully condensed (D), the nuclear membranes have disappeared, and the mitotic spindle begins to take shape (E). Acetylated a-tubulin antibody is weakly localized around the developing spindle poles (F). By metaphase, the fertilization process has concluded with the intermingling of the fully condensed male and female chromatin (G) and the formation of a barrel-shaped spindle (H); however, the acetylation of the spindle remains weak at the poles (I). At early anaphase (J and K), there is a dramatic increase in acetylation of the spindle microtubules, particularly at the spindle poles (M). By late anaphase (N) the polar microtubules appear to lose their affinity for the acetylated-a-tubulin antibody and instead, interzonal microtubules become labeled (O and P). At telophase each daughter cell has a monaster of microtubules extending from the nuclear surface of each chromatin mass towards the cell surface as well as an association with the mitotic midbody (Q and R); only the midbody is acetylated (S). All images are triple-labeled for DNA, total tubulin and acetylated a- tubulin antibody except M, which is a single stained image of acetylated-a-tubulin at anaphase. Bars: 10 «am. Centrosomes and Modified a-Tubulin 596 H. SCHATTEN, C. Howarp et al. tubulin (Fig. 61) demonstrates that the microtu- bules comprising the numerous cytasters in the oocyte are not acetylated (arrows: Fig. 61). The meiotic midbody remains acetylated throughout the remainder of first interphase. The incorporated sperm axoneme is frequently found to splay into several fibers towards the completion of the first cell cycle. First mitosis follows a pattern similar to the second meiotic division. At prophase, only the meiotic midbody binds the acetylated antibody strongly. At prometaphase and metaphase (Fig. 7A, D and G), a barrel shaped mitotic spindle emerges (Fig. 7B, E and H) and the acetylated-a- tubulin stains the polar microtubules weakly (Fig. 7F and I). At anaphase (Fig. 7J) most of the fibers are acetylated with an abundance of the polar microtubules (Fig. 7M: arrowheads). At telophase (Fig. 7N), interzonal microtubules (Fig. 70) des- tined to form the mitotic midbody are found to be acetylated (Fig. 7P). After first division (Fig. 7Q), microtubules (Fig. 7R) are found in the monasters extending from the blastomere nuclei to the oppos- ing cell surfaces as well as the interzonal microtu- bules which will form the midbody. Only the forming midbody microtubules (Fig. 7S; arrow- heads) are acetylated. At the second meiotic and first mitotic divisions, the microtubules surrounding the centrosomes are acetylated at metaphase and there is an increase in acetylation of all spindle microtubules at ana- phase: this pattern is not observed precisely during later mitoses. The two-cell mouse embryo which has an interphase array of microtubules binds the acetylated a-tubulin antibody only at the mitotic midbody. Midbodies, however, remain acetylated at these later development stages. To study the pattern of acetylation of a-tubulin at higher resolution, high voltage electron micros- copy was performed on extracted oocytes with immunogold (Fig. 8). At meiotic metaphase, the majority of the acetylated microtubules are found at the spindle poles with only sparse detection of acetylated microtubules within the spindle proper. Figure 8A is a low magnification HVEM image of a meiotic spindle at metaphase. In Figure 8B, at higher magnification, acetylated microtubules are detected by immunogold labeling. In contrast only sparse acetylation is found on the microtubules at the spindle equator (Fig. 8C). C. Detyrosination Microtubules can also be post-translationally modified by the loss of the carboxyterminal tyro- sine amino acid residue from a-tubulin [45, 46, 48]. Detyrosinated microtubules appear to be older, more stable microtubules and in the mouse oocyte, the sperm axoneme is found to be the only class of microtubules which is uniquely detected with a rabbit affinity purified antibody to detyrosinated-a- tubulin (courtesy of Dr. Bulinski). Eichenlaub- Ritter et al. [49] have examined the pattern of detyrosinated microtubules in ageing mouse oocytes and showed the presence of both types of a-tubulin. All the other classes of microtubules, including the meiotic and mitotic spindles, the meiotic and mitotic midbodies, the cytasters in meiotic and mitotic cytoplasms and the interphase microtubule complex all appear to be composed of a mixture of tyrosinated and detyrosinated microtubules. The images obtained from a global analysis of fixed material do not yet address the question of instan- taneous turnover and dynamics of individual mi- crotubules, a problem which awaits improvements in imaging technologies for cells as large as oocytes and embryos. Vv. CONCLUSIONS AND FUTURE PROSPECTS These studies comparing the origins of centro- somes and centrioles during fertilization and the post-translational modifications of a-tubulin have Fic. 8. Immuno-gold labeling of acetylated-a-tubulin with high voltage electron microscopy. Meiotic spindle microtubules in unfertilized mouse oocytes at metaphase are primarily acetylated at the spindle poles (A). The microtubules in the middle of the spindle including those at the kinetochores are less heavily acetylated. B: High magnification image showing heavy labeling of immunogold against acetylated-a-tubulin along the microtubules at the centrosomes. C: The microtubules at the spindle equator are only weakly acetylated at metaphase. Bars: 100 nm. Centrosomes and Modified a-Tubulin 597 598 H. SCHATTEN, C. Howarp et al. several implications regarding the manner in which microtubule arrays are organized and stabilized. Microtubule configurations appear to be spec- ified by the combination of labile, actively growing microtubules emanating from the centrosome and stable, perhaps older microtubules which are post- translationally acetylated or detyrosinated. For example in the two-cell mouse egg (Fig. 7), the midbody microtubules are acetylated and stable to cold or drug treatments. In contrast the monastral microtubules extending from the centrosomes on the polar faces of the blastomere nuclei are deacetylated, sensitive to cold and drug treat- ments, and are likely the result of rapid turnover due to the dynamic instability of growing microtu- bules [50]. Perhaps the stabilization of certain microtubules by post-translational acetylation or detyrosination assists in the generation of persist- ent architecture without the energetic costs of maintaining an array of microtubules involving the constant equilibrium between assembly and dis- assembly. Centrosomes appear to be paternally inherited in most all animals: animals ranging from coelenterates to lower vertebrates including fish and amphibians. In the mouse evidence is accumu- lating supporting the theory that centrosomes are of maternal origin. If studies on other mammals support this conclusion, then the evolutionary question as to when and why centrosomes switched from a paternal pattern to a maternal one must be posed. Perhaps the typical pattern of paternal inheri- tance is designed to ensure biparental fertilization. The requirement for the sperm centrosome would decrease the likelihood of successful natural par- thenogenesis, if for example the second polar body is not properly extruded. This idea is supported by studies on artificial activation [51]; most metabolic processes are properly initiated, including DNA synthesis (reviewed by [52]), but without the sperm centriole the egg cannot form a mitotic apparatus and divide. If this is correct, why then might mammals violate this seemingly sensible scheme? Abnormal fertilization and embryogenesis in mammals repre- sent a significant risk to the mother, a risk which is not found in any other class of animals. Perhaps to reduce this jeopardy, mammalian reproduction switched from a reliance on biparental centrosom- al contributions to one needing biparental chromo- some contributions. This idea is well supported by the elegant studies of Surani and Barton [53] demonstrating that fertilized eggs with two mater- nal or two paternal pronuclei cannot develop to term and by Sapienza et al. [7] which extends these observations on the molecular differences between sperm and egg DNA. If a strict requirement for biparental genomic contributions became the norm in mammals, then the requirement to maintain fidelity to the sperm centrosome could be relaxed. Perhaps the sperm might contribute some centrosomal material, perhaps the egg could also contribute: oogenesis might not destroy centrosomes and spermatogene- sis might not necessarily retain them. In this manner, one prediction might be that in some mammals either or both parent could contribute centrosomes and there might be heterogeneity among various mammalian species. Evidence to support this is emerging from the studies of Kola and Trounson [54]; human oocytes polyspermical- ly fertilized in vitro divide in a pattern not pre- dicted from studies on invertebrates or mice. While it is tempting to indulge in these specula- tions, answers supporting or refuting some assumptions are possible to obtain. The events during parthenogenesis, when the paternal con- tribution is lacking, or polyspermy, when the paternal contribution is multiplied, will likely pro- vide important evidence. Future studies on other rodents and non-rodent mammals may confirm the homogeneity among mammalian species and in- vestigations on birds and reptiles will help fill some evolutionary gaps. In conclusion, the centrosomes and centrioles are paternally inherited in most animals. However the mouse and perhaps other mammals violate this scheme: centrosomes appear to be of maternal origin, and centrioles emerge only late in fetal development probably also from maternal sources. Microtubule patterns are generated not only by the discriminating positioning of centrosomes, but also by the selective post-translational acetylation and detyrosination of a-tubulins. The rapid discoveries on the molecular basis of microtubule dynamics Centrosomes and Modified a-Tubulin will undoubtedly result in an even greater under- standing on the role, regulation and mechanisms of force generation of the microtubule-mediated mo- tions at fertilization. ACKNOWLEDGMENTS It is indeed our pleasure to acknowledge all of our wonderful colleagues and collaborators who have con- tributed to various aspects of this research. These investigators include Drs. David Asai, Harald Biess- mann, Chloe Bulinski, Peter Cooke, Daniel Mazia, Es- ter Sz6ke and Marika Walter. We are also grateful for the technical assistance of Kathleen Johnson and Cody Martin. The research reviewed here was funded by research grants from the National Institutes of Health. The IMR in Madison is an NIH Biomedical Research Technology Resource. REFERENCES 1 Schatten, G. (1982) Motility during fertilization. Int. Rev. Cytol., 79: 59-163. 2 Garbers, D. L. (1988) The regulation of spermato- zoan function by the egg. In “The Molecular Biology of Fertilization”. Ed. by H. Schatten and G. Schat- ten, in press. Academic Press, New York. 3 Lardy, H. and San Agustin, J. (1988) Caltrin and Calcium regulation of sperm activity. In “The Cell Biology of Fertilization”. Ed. by H. Schatten and G. Schatten, in press. Academic Press, New York. 4 Schackman, R. (1988) Ionic regulation of the sperm acrosome reaction and stimulation by egg-derived peptides. In “The Cell Biology of Fertilization”. Ed. by H. Schatten and G. Schatten, in press. Academic Press, New York. 5 Tilney, L. (1976) The polymerization of actin. II. How non-filamentous actin becomes nonrandomly distributed in sperm: Evidence for the association of this actin with membranes. J. Cell Biol., 69: 51-72. 6 Gibbons, I. R. (1981) Cilia and flagella of eukary- otes. J. Cell Biol., 91: 107S-124S. 7 Sapienza, C., Peterson, A. C., Rossant, J. and Ball- ing, R. (1987) Degree of methylation of transgenes is dependent on gamete of origin. Nature, 328: 251- 254. 8 Ruiz-Bravo, N. and Lennarz, W. J. (1988) Recep- tors and membrane interaction during fertilization. In “The Molecular Biology of Fertilization”. Ed. by H. Schatten and G. Schatten, in press. Academic Press, New York. 9 Shur, B. D. (1988) Galactosyltransferase as a recog- nition molecule during fertilization and develop- ment. In “The Molecular Biology of Fertilization”. Ed by H. Schatten and G. Schatten, in press. 11 13 14 16 17 18 20 in) in) N Ww Academic Press, New York. Longo, F. J. (1988) Egg cortical architecture. In “The Cell Biology of Fertilization”. Ed. by H. Schatten and G. Schatten, in press. Academic Press, New York. Schatten, H., Cheney, R., Balczon, R., Cline, C., Simerly, C. and Schatten, G. (1986) Localization of fodrin during fertilization and early development in sea urchins and mice. Dev. Biol., 118: 457-466. Schatten, H., Schatten, G., Mazia, D., Balczon, R. and Simerly, C. (1986) Behavior of centrosomes during fertilization and cell division in mouse oocytes and in sea urchin eggs. Proc. Natl. Acad. Sci. USA., 83: 105-109. Schroeder, T. (1981) Interrelations between the cell surface and the cytoskeleton in cleaving sea urchin eggs. In “Cytoskeletal Elements and Plasma Mem- brane Organization”. Ed. by G. Poste and G. L. Nicolson, Elsevier/North Holland, Amsterdam, pp. 170-216. Fisher, G. W. and Rebhun, L. I. (1983) Sea urchin egg cortical granule exocytosis is followed by a burst of membrane retrieval via uptake into coated vesi- cles. Dev. Biol., 99: 456-472. Poccia, D. (1988) Reactivation and remodelling of the sperm nucleus following fertilization. In “The Molecular Biology of Fertilization”. Ed. by H. Schatten and G. Schatten, in press. Academic Press, New York. Zirkin, B.R., Perreault,S.D. and Naish, S. J. (1988) Formation and function of the male pronu- cleus during mammalian fertilization. In “The Molecular Biology of Fertilization”. Ed. by H. Schatten and G. Schatten, in press. Academic Press, New York. Austin, C. R. (1965) New Jersey. Sawada, T. and Schatten, G. (1988) Anti-tubulin indirect immunofluorescence microscopy of Ascid- ian eggs during fertilization and early development. Cell Motil. Cytoskeleton, in press. Rothschild, L. V. (1956) _ Fertilization. Publ., London. Boveri, Th. (1900) Zellen-Studien IV. Ueber die Natur der Centrosomen. Fischer, Jena, Germany. Calarco-Gillam, P. D., Siebert, M. C., Hubble, R., Mitchison, T. and Kirschner, M. (1983) Centro- some development in early mouse embryos as de- fined by an autoantibody against pericentriolar material. Cell, 35: 621-629. Wick, S.M. (1985) Immunofluorescence micros- copy of tubulin and microtubule arrays in plant cells. Cell Biol. Int. Rep., 9: 357-371. Schatten, G., Schatten, H., Spector, I., Cline, C., Paweletz,N., Simerly,C. and Petzelt,C. (1986) Latrunculin inhibits the microfilament-mediated Fertilization. Prentice Hall, Methuen 25 28 30 31 34 600 processes during fertilization, cleavage and early development in sea urchins and in mice. Exp. Cell Res., 166: 191-208. Maro, B., Howlett,S.K. and Webb,M. (1985) Non-spindle microtubule organizing centers in metaphase-II arrested mouse oocytes. J. Cell Biol., 101: 1665-1672. Schatten, G., Simerly,C., Asai,D., Szdke, E., Cooke, P. and Schatten, H. (1987) Acetylation and detyrosination of a-tubulin in microtubules during mouse fertilization, mitosis and early development. J. Cell Biol., 105: 258a. Carre, D. and Sardet, C. (1984) Fertilization and early development in Beroe ovata. Dev. Biol., 105: 188-195. Klymkowsky, M. W. (1988) Cytoskeletal reorgan- izations during amphibian fertilization and early development. In “The Cell Biology of Fertilization”. Ed. by H. Schatten and G. Schatten, in press. Academic Press, New York. Balczon, R. and Schatten, G. (1983) Microtubule- containing detergent extracted cytoskeletons in sea urchin eggs from fertilization through cell division: Antitubulin immunofluorescence microscopy. Cell Motility, 3: 213-226. Bestor, T. H. and Schatten, G. (1981) Antitubulin immunofluorescence microscopy of microtubules present during the pronuclear movements of sea urchin fertilization. Dev. Biol., 88: 80-91. Harris, P., Osborn, M. and Weber, K. (1980) Dis- tribution of tubulin-containing structures in the egg of the sea urchin, Strongylocentrotus purpuratus. J. Cell Biol., 84: 668-679. Kuriyama, R., Borisy, G. G. and Masui, Y. (1986) Microtubule cycles in oocytes of the surf clam, Spisula solidissima: An immunofluorescence study. Dev. Biol., 114: 151-160. Schatten, G., Simerly, C. and Schatten, H. (1985) Microtubule configurations during fertilization, mitosis, and early development in the mouse and the requirement for egg microtubule-mediated motility during mammalian fertilization. Proc. Natl. Acad. Sci. USA., 82: 4152-4156. Bajer, A. S. and Mole-Bajer, J. (1986) Reorganiza- tion of microtubules in endosperm cells and cell fragments of the higher plant Haemanthus in vivo. J. Cell Biol., 102: 263-281. Schmit, A.-C. and Lambert, A.-M. (1987) Charac- terization and dynamics of cytoplasmic F-actin in higher plant endosperm cells during interphase, mitosis and cytokinesis. J. Cell Biol., 105: 2157- 2166. Mazia, D., Schatten, H., Coffe, G., Szdke, E., Ho- ward, C. and Schatten, G. (1987) Aggregation of the mitotic centrosomes into a single spherical cen- trosome by cold treatment in sea urchin eggs. J. Cell 36 37 38 39 40 41 42 44 45 46 47 48 49 50 H. SCHATTEN, C. Howarb et al. Biol., 105: 206a. Rieder, C. L., Miller, F. J., Davison, E., Bowser, S.S., Lewis, K. and Sluder, G. (1987) Proc. Elec- tron. Microsc. Soc. Am., 45: 578-581. Sz6llési, D., Calarco, P. and Donahue, R. P. (1972) Absence of centrioles in the first and second meiotic spindles of mouse oocytes. J. Cell Sci., 11: 521-541. Wooley, D. M. and Fawcett, D. W. (1973) The de- generation and disappearance of the centrioles dur- ing the development of rat spermatozoon. Anat. Rec., 177: 289-302. Magnuson, T. and Epstein, C. I. (1984) Oligosyn- dactyly — a lethal mutation in the mouse that results in mitotic arrest very early in development. Cell, 38: 823-833. Olmsted, J. B. (1986) Microtubule-associated pro- teins. Ann. Rev. Cell Biol., 2: 421-457. L’Hernault,S. W. and Rosenbaum, J.L. (1985) Reversal of the posttranslational modification on Chlamydomonas flagellar a-tubulin occurs during flagellar resorption. J. Cell Biol., 100: 457-462. L’Hernault,S. W. and Rosenbaum,J.L. (1983) Chlamydomonas a-tubulin is posttranslationally modified in flagella during flagellar assembly. J. Cell Biol., 97: 258-263. Piperno, G., LeDizet,M. and Chang, X. (1987) Microtubules containing acetylated a-tubulin in mammalian cells in culture. J. Cell Biol., 104: 289- 302. Piperno, G. and Fuller, M.T. (1985) Monoclonal antibodies specific for an acetylated form of a- tubulin recognize the antigen in cilia and flagella from a variety of organisms. J. Cell Biol., 101: 2085- 2094. Gundersen, G. G., and Bulinski, J. C. (1986) Dis- tribution of tyrosinylated and nontyrosinylated a- tubulin during mitosis. J. Cell Biol., 102: 1118-1126. Gundersen, G. G., Kalnoski, M. H. and Bulinski, J. C. (1984) Distinct populations of microtubules: Tyrosinylated and nontyrosinylated a-tubulin are distributed differently in vivo. Cell, 38: 779-789. Vacquier, V. D. (1981) Dynamic changes of the egg cortex. Dev. Biol., 84: 1-26. Geuens, G., Gundersen, G. G., Nuydens, R., Cor- nelissen, F., Bulinski, J.C. and DeBrabander, M. (1986) Ultrastructural colocalization of tyrosiny- lated and detyrosinylated a-tubulin in interphase and mitotic cells. J. Cell Biol., 103: 1883-1893. Eichenlaub-Ritter, U., Chandley, A. C., and Gos- den, R. G. (1986) Alterations to the microtubular cytoskeleton and increased disorder of chromosome alignment in spontaneously ovulated mouse oocytes aged in vivo: An immunofluorescence study. Chro- mosoma, 94: 337-345. Kirschner, M. and Mitchison, T. (1986) Beyond self-assembly: From microtubules to morphogene- 54 Centrosomes and Modified a-Tubulin sis. Cell, 45: 329-342. Loeb, J. (1913) Artificial Parthenogenesis and Ferti- lization. Columbia Univ. Press, New York. Epel, D. (1980) Ionic triggers in the fertilization of sea urchin eggs. Ann. N.Y. Acad. Sci., 339: 74-85. Surani, M. A. H. and Barton, S.C. (1983) De- velopment of gynogenetic eggs in the mouse: Im- plications for parthenogenetic embryos. Science, 222: 1034-1036. Kola, I. and Trounson, A. (1987) Tripronuclear ap) 601 human oocytes: Altered cleavage patterns and sub- sequent karyotypic analysis of embryos. Biol Re- prod., 37: 395-401. Schatten, H., Walter, M., Mazia,D., Biessmann, H., Paweletz, N., Coffe, G. and Schatten, G. (1987) Centrosome detection in sea urchin eggs with a monoclonal antibody against Drosophila intermedi- ate filament proteins: Characterization of stages of the division cycle of centrosomes. Proc. Natl. Acad. Sci., USA., 84: 8488-8492. ZOOLOGICAL SCIENCE 5: 603-611 (1988) © 1988 Zoological Society of Japan Spontaneous Aster Formation in Cytoplasmic Extracts from Eggs of the Surf Clam RoBeERT E. PALAZZO, JENNIFER B. BRAWLEY and LIONEL I. REBHUN Department of Biology, University of Virginia, Charlottesville, VA 22901, U.S.A. ABSTRACT— Asters form spontaneously in cytoplasmic extracts prepared from eggs obtained from the surf clam Spisula solidissima. Astral birefringence is augmented and asters are stabilized by the addition of hexylene glycol. Asters form spindle-like structures, contain particles associated with their periphery, and aggregate to form multi-astral complexes. Studies with video microscopy and electron microscopy indicate that asters formed in vitro are composed of microtubules radiating from a central zone which contains a centriole. Asters were isolated and their associated proteins analyzed. In addition, astral microtubules and microtubule associated proteins were isolated by temperature dependent polymeriza- tion-depolymerization. Numerous proteins coassemble with astral tubulin through cycles of assembly- disassembly. We suggest that this will be a useful system for the isolation and characterization of centrosomal components which direct microtubule organization in cells. INTRODUCTION Microtubules are generally not randomly arranged in cells but are organized around discrete regions of the cytoplasm [1-3]. These foci of radiating microtubules, known as microtubule organizing centers (MTOCs) [4], are sites of initia- tion of microtubule assembly [1, 2] and may serve to capture preexisting microtubules [5]. Microtu- bule organizing centers play a significant role in cell division [5-7], organization of cellular mor- phology [2], and possibly in directing cellular migrations [8, 9]. In spite of the obvious signif- icance of MTOCs to cell functions, remarkably little is known of their chemical composition or their regulation [1]. This is largely due to the inability to isolate and purify MTOCs on a suf- ficient scale for preparative and analytical bio- chemical approaches. We describe here a system which allows the isolation of MTOCs from the surf clam Spisula solidissima. Spisula is available during the summer months off the north east coast of the United States at the Marine Biological Laboratory at Accepted April 21, 1988 Woods Hole, Massachusetts. During the summer when the animals are gravid as much as 30 ml of eggs can be obtained from the gonad of a single female. The eggs are obtained at diakinesis of the first meiotic division [10]. They contain a large germi- nal vesicle which houses a prominent nucleolus and condensed chromosomes. Eggs can be ferti- lized with sperm or activated parthenogenetically by the addition of KCI to an egg sea water suspen- sion [11]. Usually within 8 min after the addition of KCl at room temperature the nucleolus dis- appears and the germinal vesicle envelope breaks down (GVBD). By 12 min a prominent meiotic spindle is formed within the egg center. The spindle subsequently migrates to the egg cortex where it performs oscillatory movements, a pro- cess reported in meiotic stages by Rebhun [10] and studied extensively in embryos undergoing cleav- age by Dan and Inoue [12]. Finally, the egg undergoes two sequential meiotic divisions result- ing in the formation of two polar bodies. In recent years cytoplasmic extracts prepared by centrifugal crushing of cells in very low dilution conditions have been used to study nuclear en- velope breakdown and assembly [13-18], chroma- 604 R. E. PaLtazzo, J. B. BRAWLEY AND L. I. REBHUN tin condensation and decondensation [14, 15], and DNA synthesis [19]. Using similar cytosolic ex- tracts prepared from Spisula we found that asters formed spontaneously when the extract was warmed to room temperature. We report our initial studies using video and electron microscopy, which indicate that asters contain centrioles con- firming the observation of Weisenberg and Rosen- feld [20]. Further, we have isolated these asters in sufficient quantity for biochemical analysis and described both astral protein composition and the astral microtubule associated proteins. MATERIALS AND METHODS Egg preparation and activation Eggs were prepared by dissection of the gonads from Spisula solidissima females. The eggs were released into sea water by snipping the ovary with scissors and gently agitating the resulting pieces. The egg-sea water suspension was poured through cheese cloth to remove unwanted debris and gonadal tissue. The eggs were then washed at least three times by successive resuspension and settling with a minimum of 1,000 <’s volume of sea water per wash. Eggs were activated by the addition of 0.5 M KCI to a final ratio of 86: 14 sea water to 0.5 M KCl according to Allen [11]. Preparation of cytosolic extract Cytosolic extracts were prepared from Spisula eggs whose vitelline membranes were removed by washing twice using successive cycles of 1,000 x g x 30 second centrifugations and resuspension with 10mM NaPO,, 1M glycerol at pH 8.0-8.2, fol- lowed by one wash with 1M glycerol. Eggs were then resuspended in ice cold 20 mM PIPES buffer containing 100 mM KCl, and 5mM MgSOu,, pH 7.2 and quickly centrifuged at 1,000 xg for 30 sec. The supernatant was removed and the pellet was loosened and eggs broken by resuspension with a pipette. The suspension was centrifuged at ap- proximately 10,000xg for 10min at 4°C which resulted in separation into three layers; an upper lipid layer, a lower yolk layer and a middle cyto- plasmic layer similar to the preparations reported by others in amphibian eggs [14, 15]. The cyto- plasmic extract was removed and centrifuged again at 10,000xg for 10min at 4°C and the middle cytoplasmic layer collected and kept on ice until use. Light and video microscopy Twenty microliter drops of the extract were placed onto a glass slide and sealed with a coverslip lined with silicon grease. Aster formation was observed with a Zeiss photomicroscope-I equipped with strain free polarization lenses and micro- graphs were taken with Kodak Tri-X 400 film developed with Diafine. Video microscopy was carried out with a Zeiss Axiophot microscope equipped with a Hamamatsu C-2400 Newvicon camera linked to a Sony 5800H video recorder. Photographs of video images were taken from a Javelin video monitor using Kodak T-Max 400 film developed with T-Max developer. Electron microscopy After resuspension in MEMG (see below) and centrifugation, aster pellets were fixed with 1% glutaraldehyde in MEMG followed by postfixation with 1% osmium tetroxide, dehydrated through an ethanol series, and embedded in epon. Thin sections were prepared, stained with lead citrate and uranyl acetate and micrographs taken using a Hitachi HU-11E] at 75 kv. Aster isolation Cytosolic extract, with or without added hex- ylene glycol to 3%, was incubated at room temper- ature for 5 min. Aster formation was monitored as described above with polarization microscopy. The cytosolic extract was then diluted with 10 volumes of MEMG (20 mM MES, 1 mM EGTA, 1 mM MgsSOu,, 10% glycerol, pH 6.5) and centri- fuged at 1,000xg for 10min. The pellets from multiple samples were pooled and washed by resuspension in MEMG (minimum of 100 volumes) and centrifuged again at 1,000xg. The washed pellet was resuspended in a 100 volume of microtubule reassembly buffer (100 mM PIPES, 1mM EGTA, 1 mM MgSOug, pH 6.9) and centri- fuged at 1,000 g. The supernatant was aspirated and GTP added to the pellet from a 100 stock to a final concentration of 1mM. The aster suspen- Aster Formation in Cytoplasmic Extracts 605 sion was then incubated on ice for lhr to de- polymerize microtubules as monitored by polariza- tion microscopy. The solution was centrifuged at 39,000 x g for 30 min at 4°C and the supernatant and pellet collected. The supernatant was warmed to 30°C for 30 min to repolymerize microtubules which were then centrifuged out of solution at 39,000 g for 30min at 30°C. The microtubule pellet and supernatant were collected. Samples at various stages of the preparation were taken for electrophoretic analysis as described below. Electrophoresis Proteins from the various stages of the astral protein preparation steps described above were analyzed by SDS gel electrophoresis according to Laemmli [21] using a running gel composed of 7.5% acrylamide and 4 M urea. Gels were stained with either Coomassie blue or silver [22] to visual- ize proteins. RESULTS Aster formation and migration Eggs were activated with KCI (as described above) for 4 min, washed, and cytosolic extract prepared as described in Materials and Methods. Within 2 min after warming the extract to room temperature numerous asters were visible with polarization microscopy throughout the slide (Fig. la). The number of asters formed varied between preparations but was generally between 20-30 per 10 field as visualized with a Zeiss photomicro- scope-I with an optovar setting of 1.25. Aster formation occurred in almost all preparations with few exceptions. Addition of hexylene glycol to the cytosolic extract to a final concentration of 3% before warming to room temperature resulted in the Fic. 1. Aster formation in cytoplasmic extracts. With polarization microscopy asters were observed within two minutes of warming extracts to room tempera- ture (a). Addition of hexylene glycol to the extract resulted in augmentation of astral birefringence (b). In the presence of hexylene glycol birefringent fibers formed in the background approximately 10 min after warming to room temperature (c). (Magnifica- tion: 108x). stabilization and augmentation of the birefrin- gence of asters (HG-asters) (Figs. 1b, 3 and 4). In addition, diasters and multiple aster complexes were observed (Figs. 2, 3 and 4b). After 10 min of warming the extract (with hexylene glycol) numer- 606 R. E. PALAzzo, J. B. BRAWLEY AND L. I. REBHUN Fic. 2. Aggregation of asters. Aggregates of asters (a) and eventually astral complexes (b) formed in ex- tracts. By 60 min after warming the extract to room temperature in the presence of hexylene glycol these complexes formed tight masses of asters which could be removed from the extract with forceps (b). (Magnification: 108 x ). ous fibers were observed in the background (Fig. 1c). With time (10-20 min) asters congressed and ultimately formed multi-astral complexes (Fig. 2). Indeed, if the reactions were allowed to occur in a test tube, the asters and secondary fibers eventual- ly formed a complex which separated from the rest of the solution and could be removed with forceps (Fig. 2b). Preliminary evidence suggests that this movement and aggregation was independent of actin polymer formation since these migrations occur if cytochalasin-B is added to the extract before inducing aster formation (data not shown). Fic. 3. Nomarski video-microscopy of an aster and a spindle-like structure. Asters were composed of fibers radiating from a central zone (a). Some asters formed spindle-like structures (b) which contained particles (small arrowheads) within their mid-zone. Fibers were also found radiating from particles not associated with astral centers (large arrowhead in b). (Magnification: 1,242). Video-microscopy and electron microscopy HG-asters were studied using Nomarski optics and visualized using video imaging. The asters were composed of fibers radiating from a central zone (Figs. 3 and 4). Asters contained so many radial fibers that they appeared birefringent even with Nomarski optics. Within the central zone of the asters a small dense structure was observed (Fig. 4). All asters contained at least one such structure, although occasionally asters which con- tained two were found (data not shown). Analysis by electron microscopy confirmed the video- microscopy observation that asters were composed of radiating microtubules (Fig. 4c). Single cen- trioles were found at the center of the asters (Fig. Aster Formation in Cytoplasmic Extracts 607 Fic. 4. Asters contain centrioles. High magnification Nomarski-video microscopy (a and b) revealed that astral centers were composed of a zone (large arrow- heads in a and b) from which fibers radiated. These zones from two or more asters were occasionally found to overlap as if fused (b). Within these zones small refractile bodies were found (small arrow- heads in a and b). Electron micrographs indicated that centrioles were present at the center of asters (c). (Magnification: 3,105 x (a and b) and 13,230 (c)). 4c), although on one occasion two centrioles were observed. Asters tended to form multiple aster complexes composed of two or more asters (Fig. 4b). These multi-astral complexes contained multiple centers with overlapping central zones which tended to distort and fuse rather than maintain discrete independent astral centers (Fig. 4b). Thus, the pericentriolar zone of asters is capable of fusion and interaction with the same zone of other asters. Nevertheless, we were always able to find a density corresponding to a centriole at the center of each zone (Fig. 4). Biochemical analysis of astral components Asters formed in the presence or absence of added hexylene glycol were isolated and analyzed by SDS-electrophoresis. As expected, one of the major components was tubulin. Few qualitative differences were observed when aster and HG- aster proteins were compared (Fig. 6). In addi- tion, quantitative differences, particularly with re- spect to tubulin and high molecular weight pro- teins were found (Fig. 6). Thus hexylene glycol increased the tubulin content of asters but did not induce significant qualitative changes in astral composition. Astral rays were depolymerized by cold treat- ment in a microtubule reassembly buffer (Fig. 5). The cold-labile components were separated from cold-stable material by centrifugation. In addition to tubulin, a number of other proteins became soluble upon treatment with cold temperatures and GTP (Fig. 6). When the cold-labile fraction was separated from the cold-stable components and warmed to 30°C, tubulin polymerized (Figs. 5c and 6) and a number of other proteins (presumably MAPs) coassembled (Fig. 6). Interestingly, some astral proteins which were cold labile and presum- ably solubilized as astral microtubules depoly- merized, did not reassemble (coassemble) with the tubulin upon warming. Finally, while microtu- bules form, no asters were observed in the poly- merized supernatant when assayed by polarized light microscopy (Fig. 5c), suggesting that the MTOCs were stable to treatment of cold tempera- ture and GTP, and consequently pelleted during the 39,000 x g centrifugation. 608 R. E. PaLazzo, J. B. BRAWLEY AND L. I. REBHUN DISCUSSION Previous studies have suggested that cytoplasmic extracts can be useful in the study of cell cycle dependent events in vitro. For example, cytoplas- mic preparations from sea urchin eggs can induce sperm chromatin to decondense [23]. In addition, extracts from Xenopus laevis oocytes induce a wide variety of processes in vitro such as DNA synthesis [19], formation of pronuclei [15], nuclear envelope assembly [13, 15, 16], chromatin decondensation [15, 23], chromosome condensation [14, 17], nu- clear envelope breakdown [14, 17, 18] and spindle formation [14] in vitro. Finally, Weisenberg et al. reported that asters formed spontaneously in homogenates prepared from Spisula oocytes [20] and sea urchin eggs [24]. We have adapted the methods of Masui [25], Benbow and Ford [19] and Lohka and Maller [14, 15] to prepare similar cytoplasmic extracts from eggs obtained from Spisula. With these extracts we have consistently observed the spontaneous formation of asters in vitro. Birefringent asters formed within two minutes upon warming the extracts to room temperature and were augmented and stabilized by the addition of hexylene glycol, a known microtubule stabilizing agent [26]. Numer- ous astral complexes were observed including monasters, diasters or spindle-like structures, and multiastral complexes of three or more asters. In addition, aster formation was followed by the formation of non-astral background fibers which were also birefringent and which varied in amount from preparation to preparation. Preliminary evi- dence suggests that these background fibers con- tained tubulin and associated proteins in the form of free microtubules. Once formed, asters and background fibers aggregated together to form multi-astral aggregates. The significance of astral aggregation is not clear, but one possibility is that this in vitro process is related to spindle migrations in living Spisula eggs [10, 12]. We have studied asters with video-microscopy and electron microscopy and confirmed the pre- Fic. 5. Isolation of asters and cycling of astral tubulin. Asters formed in the presence of hexylene glycol were isolated and washed in microtubule stabilizing buffer (a). Astral rays were depolymerized by treatment with cold temperatures and GTP (b). Cold stable material was separated by centrifugation and microtubules were isolated by temperature de- pendent polymerization-depolymerization cycles (c). (Magnification: 108 x). Aster Formation in Cytoplasmic Extracts 609 Fic. 6. 12345 67 8 Protein composition of asters and microtubule associated proteins analyzed by SDS electrophoresis. Lanes: 1) spontaneous asters, 2) hexylene glycol asters, 3) cold insoluble and 4) cold soluble astral proteins, 5) and 7) supernatants and 6) and 8) pellets of first and second cycle polymerizations respectively. Quantitative but no qualitative differences in proteins were found when hexylene glycol was used to augment aster birefringence (compare lanes | and 2). Several proteins (small arrows) in addition to tubulin (large arrows) coassembled during temperature dependent polymerization-depolymerization of microtubules. Numbers on left indicate approximate molecular weight in kilodaltons. vious reports by Weisenberg [20, 24] that a) asters are composed of microtubules radiating from a central zone which contains a centriole, and b) asters form spindle-like structures and _ collect particles at their periphery. Using video- microscopy we have shown that each aster con- tained a single dense body which when investi- gated with electron microscopy was found to be a single centriole except in a one case where we found an aster containing two centrioles. We have presented evidence which suggests that the zone surrounding the centriole is dynamic and capable of interaction with the pericentriolar region of other asters so that when two or more asters were found in proximity, their central zones tended to fuse. We have extended these studies by developing methods for the isolation of asters formed in cytoplasmic extracts and have begun to character- ize their constitutive proteins. The protein com- position of asters formed spontaneously was com- pared to that of asters augmented with hexylene glycol. Although some minor differences in pro- tein composition were observed, the major effect of hexylene glycol was to increase the astral tubu- lin content. Electrophoretic analysis of cycled astral tubulin revealed that a number of proteins, presumably MAPs, coassembled with microtu- bules. In addition, investigation of the cycled tubulin fractions with the polarization microscope revealed that although birefringent fibers were observed, no asters were present. This result suggests that the component(s) responsible for MTOC activity was separated from the fraction during the 39,000g centrifugation step which followed the initial cold depolymerization of astral microtubules. With the use of video-microscopy we observed that particles present in the crude cytoplasmic extract much smaller than astral centers could also serve as sites from which fibers extended although such fibers were much fewer in number than those surrounding astral centers. These particles may be related to the granules from sea urchin spindles identified by Endo [27, 28] and isolated by Sakai and colleagues [29, 30] which are capable of or- ganizing microtubules (MTOGs). These investiga- tors have identified a granule associated 51 Kd protein which retains MTOC activity in vitro. In addition, antibodies raised against this protein localize to astral centers [30]. These results suggest that individual proteins may be responsible for 610 R. E. Patazzo, J. B. BRAWLEY AND L. I. REBHUN MTOC activity. Whether or not similar proteins exist in Spisula asters formed in vitro is not clear at this time, but since we can now isolate asters in sufficient quantities for preparative and analytical biochemistry we can begin to address such ques- tions. ACKNOWLEDGMENTS This work was done with funds supplied by National Institutes of Health Grant No. GM 36550 to L.I. Rebhun. Dr. Palazzo was supported by Fellowships From National Institute of Health Grant Nos. 5T32HD07192 and 5—532-GM11502-02. REFERENCES 1 Brinkley, B. R. (1985) Microtubule organizing cen- ters. Ann. Rev. Cell Biol., 1: 145-172. McIntosh, J. R. (1983) The centrosome as an orga- nizer of the cytoskeleton. Mod. Cell Biol., 2: 115- 142. 3. Tucker, J. B. (1979) Spatial organization of micro- tubule organizing centers and microtubules. J. Cell Biol., 99: 55s—62s. 4 Pickett-Heaps, J.D. (1969) The evolution of the mitotic apparatus: an attempt at comparative ultra- structural cytology in dividing plant cells. Cytobios, 1: 257-280. 5 Picket-Heaps,J.D., Tippit,D.H. and Porter, K. R. (1982) Rethinking mitosis. Cell, 29: 729-744. 6 Inoue, S. (1981) Cell division and the mitotic spin- dle. J. Cell Biol., 91: 131s—147s. 7 Mazia, D. (1984) Centrosomes and mitotic poles. Exp. Cell Res., 153: 1-15. 8 Gotlieb, A. I., May, L. M., Subrabmanyan, L. and Kalnins, V. I. (1981) Distribution of microtubule organizing centers in migrating sheets of endothelial cells. J. Cell Biol., 91: 589-594. 9 Koonce, M.P., Cloney,R. A. and Berns, M. W. (1984) Laser irradiation of centrosomes in newt eosinophils: Evidence for centriole role in motility. J. Cell Biol., 98: 2222-2229. 10 Rebhun, L. I. (1958) Studies of early cleavage in the surf clam Spisula solidissima, using methylene blue and toluidine blue as vital stains. Biol. Bull., 117: 518-545. 11 Allen, R.D. (1953) Fertilization and _ artificial activation in the egg of the surf clam Spisula solidis- sima. Biol. Bull., 105: 213-239. 12 Dan, K. and Inoue, S. (1987) Studies of unequal cleavage in Molluscs II. Asymmetric nature of the two asters. Int. J. Invertebr. Reprod. Dev., 11: 335- 354. to 13 14 15 16 17 18 19 20 21 pH 23 24 25 26 2) 28 Burke, B. and Gerace, L. (1986) A cell free system to study reassembly of the nuclear envelope at the end of mitosis. Cell, 44: 639-652. Lohka, M. J. and Maller, J. L. (1985) Induction of nuclear envelope breakdown, chromosome con- densation, and spindle formation in cell-free ex- tracts. J. Cell Biol., 101: 518-523. Lohka, M. J. and Maller, J. L. (1987) Regulation of nuclear formation and breakdown in cell-free ex- tracts of amphibian eggs. In “Molecular Regulation of Nuclear Events in Mitosis and Meiosis”. Ed. by R. A. Schlegel, M.S. Halleck and P.N. Rao, Academic Press, Inc., New York, pp. 67-109. Newport, J. (1987) Nuclear reconstitution in vitro: stages of assembly around protein-free DNA. Cell, 48: 205-217. Newport, J. and Spann, T. (1987) Disassembly of the nucleus in mitotic extracts: membrane vesicular- ization, lamin disassembly, and chromosome con- densation are independent processes. Cell, 48: 219- 230. Suprynowicz, F. A. and Gerace, L. (1986) A frac- tionated cell-free system for analysis of prophase nuclear disassembly. J. Cell Biol., 103: 2073-2081. Benbow, R. M. and Ford, C. C. (1975) Cytoplasmic control of nuclear DNA synthesis during early de- velopment of Xenopus laevis: a cell-free assay. Proc. Natl. Acad. Sci. U.S.A., 41: 639-652. Weisenberg, R. C. and Rosenfeld, A. C. (1975) In vitro polymerization of microtubules into asters and spindles in homogenates of surf clam eggs. J. Cell Biol, 64: 146-158. Laemmli, U. K. (1970) Cleavage of structural pro- teins during the assembly of the head of bacte- riophage T4. Nature, 227: 680-685. Wray, W., Boulikas, T., Wray, V. P. and Hancock, R. (1981) Silver staining of proteins in polyacryl- amide gels. Anal. Biochem., 118: 197-203. Kunkle, M., Magun, B. and Longo, F.J. (1978) Analysis of isolated sea urchin nuclei incubated in egg cytosol. J. Exp. Zool., 203: 381-390. Weisenberg, R. C. (1987) Assembly of sea urchin egg asters. In “Cell Reproduction”. Ed. by E. R. Dirksen, D. M. Prescott and C. F. Fox., Academic Press, New York, pp. 366. Masui, Y. (1982) Oscillating activity of maturation promoting factor (MPF) in extracts of Rana pipiens eggs. J. Exp. Zool., 224: 389-399. Rebhun, L. I., Jemiolo, D. K., Ivy, N., Mellon, M. and Nath, J. (1975) Regulation of the in vivo mitotic apparatus by glycols and metabolic inhibitors. Ann. N. Y. Acad. Sci., 253: 362-377. Endo, S. (1979) The clusters of granular material around the centriole during mitosis. Cell Struct. Funct., 4: 71-74. Endo, S. (1980) Further observations of the clusters 29 Aster Formation in Cytoplasmic Extracts of granular material around the centriole in the sea urchin egg: Changes in distribution during mitosis. Dev. Growth Differ., 25: 307-314. Toriyama, M., Endo, S. and Sakai, H. (1984) Aster formation in vitro is nucleated by granules isolated from the mitotic apparatus. Cell Struct. Funct., 9: 30 611 213-224. Toriyama, M., Ohta, K., Endo, S. and Sakai, H. (1988) 51 Kd protein, a component of microtubule- organizing granules in the mitotic apparatus in- volved in aster formation in vitro. Cell Motil. Cytosk., 9: 117-128. —_ o - on - < ™~ 7 1h pele a ay 3 i : =e ~ tae tee — Co ee We ‘ail ; =f m a ey! Me , OG | riot ey _— ’ au ©. ee ™~. : i / 7 re a C ‘ { es, | ZOOLOGICAL SCIENCE 5: 613-621 (1988) Mitotic Apparatus-Associated 51-kD Protein in Mitosis of Sea Urchin Eggs KUNIHIRO OHTA, MASARU TORIYAMA, SACHIKO ENDO and HIKOICHI SAKAI Department of Biophysics and Biochemistry, Faculty of Science, University of Tokyo, Tokyo 113, Japan ABSTRACT— The centrosome of the sea urchin egg at metaphase consists of the centrioles and clusters of granular material with an average diameter of 90 nm which we call microtubule-organizing granules (MTOGs). MTOGs isolated from the isolated mitotic apparatus initiated astral microtubules in the presence of exogenous tubulin with the plus end distal to the center. Phosphocellulose column chromatography enabled the fractionation of a protein fraction from solubilized MTOGs which was capable of self-assembling into granules and initiating astral microtubules with plus end also distal to the center. A 51-kD protein with an isoelectric point of 9.8 was a major component of the granules. Lysine was the most prominent amino acid residue of the 51-kD protein. Polyclonal antibodies against the 51-kD protein stained the center of asters reconstructed in vitro. The antibody almost totally inhibited the aster forming ability of MTOGs. The antibody stained centrosomal regions, the proximal end of astral microtubules and half spindles. Immunoflorescence patterns of the mitotic apparatus stained with monoclonal antibody against the 51-kD protein were almost the same as those stained with polyclonal antibody. Microinjection of the monoclonal antibody into sea urchin eggs revealed that the antibody totally inhibited the formation of the mitotic apparatus when injected before prometaphase. We suggest that the 51-kD protein is a major component of the centrosome and plays a role in the initiation of astral and spindle microtubules at mitosis of the sea urchin egg. © 1988 Zoological Society of Japan INTRODUCTION Tubulin self-assembles into microtubules with the aid of microtubule-associated proteins under physiological medium conditions in vitro. Nuclea- tion of assembly and elongation of polymers are well known processes in microtubule reconstitu- tion. However, the in vitro properties of microtu- bule assembly do not necessarily apply to microtu- bule dynamics in live cells, in which most of the microtubules are thought to be site-initiated from the so-called microtubule-organizing center (MTOC) [1], a ubiquitous structure that spatially organizes microbutules in cells [2, 3]. In mitotic cells, the centrosome that consists of pericentriolar materials and a pair of centrioles initiates astral and spindle microbutules. Forma- tion of aster- and spindle-like structures in a homogenate of surf clam egg was described by Accepted April 14, 1988 Weisenberg and Rosenfeld [4] who suggested that the granular material surrounding the centriole in the aster may be an MTOC. Furthermore, isolated centrosomes or mitotic centers radially nucleated microtubules [5-7]. The structures responsible for the initiation of astral microtubules are not the centrioles but the pericentriolar material, and the concept that pericentriolar material is of primarily importance to this structural organization was brought forth by Gould and Borisy [8] who demon- strated that microtubule assembly is nucleated from isolated pericentriolar material. This nucle- ating ability was shown to be cell cycle- or mitotic cycle-dependent [9, 10]. Actual involvement of the pericentriolar material, not the centrioles, for spindle organization in live cells was then shown by Berns et al. [11, 12] in irradiation experiments of the centrosome with a laser microbeam. The pericentriolar material in cultured cells was called the pericentriolar cloud as an amorphous assembly [13]. In contrast, the pericentriolar 614 K. Outa, M. Toriyama et al. material in sea urchin eggs was shown to consist of clusters of granular material [14] that surrounded the centrioles, which later came to be called micro- tubule-organizing granules (MTOGs) [15]. Exten- sive observations using electron microscopy showed that the clusters of granular material play a role in nucleating and organizing microtubules with marked changes in shape during mitosis [16- 18]. This brief review describes recent studies on the 51-kD protein that is responsible for the aster- forming activity of MTOGs in the sea urchin egg. ASSEMBLY OF MICROTUBULE-ORGANIZING GRANULES AT THE POLES Prophase of the first division cycle of the sea urchin egg is characterized by the beginning of assembly of electron dense granules on the both sides of the nucleus [16]. The average diameter of the granules is estimated to be 90 nm. The gran- ules gradually form clusters surrounding the cen- trioles. Microtubules seem to be nucleated by the granules to form asters, because they are focused on the clusters of such granules (Fig. la, arrow). At the very beginning of the aster formation at prophase, microtubules are observed to be con- nected to small clusters of granules. MTOGs accumulate around the poles from prometaphase through metaphase and accompany the growth of astral microtubules. Shortly after the breakdown of the nuclear envelope, no microtubules are observed in the nuclear region. However, the clusters of MTOGs which face the nucleus rapidly initiate spindle microtubules. The spindle microtu- bules are not curved but straight at the beginning of the formation of the spindle, therefore crossing each other around the equator (Fig. 1b). This is followed by a bending of the microtubules to the form of the spindle, probably due to the formation of crossbridges among the microtubules. Fic. 1. Electron microscopic image of mitotic Hemicentrotus eggs [16]. Eggs at mitosis were fixed with glutaraldehy- de and processed for sections of 0.3 um in thickness. a: Beginning of prometaphase, b: Prometaphase, c: Metaphase, d: Anaphase. Arrows show clusters of granular material (MTOGs). Bar: 5 ~m. (By permission of Dev. Growth Differ.) 51-kD Protein in the Mitotic Apparatus 615 Metaphase is characterized by a maximum accu- mulation of MTOGs at both poles in the form of a spherical mass surrounding the centrioles (Fig. 1c). When the astral microtubules grow further at anaphase, the shape of the clusters of MTOGs changes from a sphere to a disc perpendicular to the axis of the spindle, forming flat centrosomes (Fig. 1d). Furthermore, the cluster adheres to the daughter nuclei at the end of mitosis [17]. That the clusters of MTOG initiate microtubules was confirmed by experiments in which MTOGs are dispersed by the action of hexyleneglycol [19]. When sea urchin eggs at prometaphase are treated with sea water containing 5% hexyleneglycol, assembled clusters of MTOGs at the pole disperse to form a centrosphere-like region. MTOGs are redistributed at the periphery of the ‘pseudo- centrosphere’ and each of the dispersed granules initiates microtubules so that an unusually large number of microtubules are formed, each focusing on the dispersed granules. Similar dispersion of MTOGs and growth of unusually large numbers of microtubules are observed when hexyleneglycol treatment is carried out at metaphase [19]. INITIATION OF MICROTUBULES BY FRAG- MENTS OF THE CENTROSOME IN VITRO The first step in the investigation of the mecha- nism of microtubule initiation in vivo is the isola- tion of the centrosome and analysis of the ability to initiate microtubules. Isolated mitotic apparatuses favor a large supply of functional centrosomes. When the mitotic apparatuses isolated by the DMSO-glycerol method are chilled and gently homogenized, the isolated centrospheres initiate microtubules to form asters [20]. | Further homogenization of the isolated mitotic apparatuses causes fragmentation of the centrosomes, resulting in the formation of numerous small asters when combined with tubulin. An indirect im- munofluorescence image of the aster is shown in Fic. 2. ym. b: Dark-field microscopic image. Asters reconstructed from MTOGs and tubulin. MTOGs from isolated a: Immunofluoresence image. mitotic apparatuses were incubated with tubulin for 10 min and fixed with formaldehyde, followed by processing for immunofluorescence staining using anti-tubulin antibody and fluorescein-labeled goat anti-rabbit IgG. Bar: 5 Solubilized mitotic apparatuses were column chromatographed on phosphocellulose to obtain a 0.5M KCl eluate, which was dialyzed against a solution of low ionic strength. Granules formed were incubated with tubulin for 10 min. Bar: 10 um. 616 K. Outa, M. Toriyama et al. Fic. 3. Electron microscopic image of astral centers [21]. a: MTOGs from isolated mitotic apparatuses. b: MTOGs reconstructed from 0.5 M KCl eluate on a phosphocellulose column, to which a KCI extract of whole metaphase eggs were applied. Bar: 100 nm. (By permission of Alan R. Liss, Inc., New York) Figure 2a where the microtubules form an astral configuration. Measurement of the growth rate of the microtubules reveals that the plus end is distal to the center [21]. Electron microscopy discloses that the center of these asters consists of small clusters of 10 to 20 granules with diameters ranging from 40 to 140 nm (Fig. 3a), and usually initiating ~100 microtubules. Therefore, it is conceivable that these asters are formed from fragmented centrosomes. We also noted that these granular aggregates appear to be derived only from centro- somes as judged by electron microscopy of the isolated mitotic apparatuses. When a homogenized suspension of the isolated mitotic apparatuses is treated with trypsin, asters are no longer formed [20]. In contrast, treatment with RNase A does not destroy aster forming activity, but it is susceptible to heat treatment, i.e., incubation at 60°C for 2 min almost totally de- stroys it. SOLUBILIZATION OF PROTEINS AND RECONSTITUTION OF MTOGs Aster forming activity in the isolated mitotic apparatuses can be solubilized in a solution of higher ionic strength in the presence of glycerol. Dialysis of a 0.5 M KCI extract of mitotic appa- ratuses causes formation of granular aggregates from which small asters are formed when incu- bated with tubulin [20]. Phosphocellulose column chromatography of the solubilized MTOG fraction enabled separation of a protein component that is capable of being reconstituted into granules by dialysis against a solution of low ionic strength and capable of initiating astral microtubules when incu- bated with tubulin. Starting from a KCl extract of whole fertilized eggs, a protein fraction is obtained by phosphocellulose column chromatography, and forms granules by dialysis that are capable of nucleating astral microtubules. Polarity of the astral microtubules is always such that the plus end is distal to the astral center. The same holds true with astral microtubules initiated from MTOGs freshly prepared from isolated MAs [21]. Fur- thermore, when microtubules are first initiated from MTOGs in a short time and then biotin- labeled tubulin is added (Fig. 4), one can measure the growth rate of microtubules more easily with the plus end distal to the center. Electron microscopy shows that such granular aggregates resemble MTOGs with average di- ameters ranging from 100 to 300 nm (Fig. 3b). Small clusters that consist of 10 to 20 unit granules usually initiate more than 50 microtubules. Be- cause quick dilution of a 0.5M KCI extract of mitotic apparatuses leaves most of the proteins solubilized in solution, it is conceivable that the granules reconstitute in such a way that the constit- uent proteins self-assemble in an order dependent upon a gradual decrease in ionic strength. 51-kD Protein in the Mitotic Apparatus 617 Fic. 4. Growth of astral microtubule from MTOGs. Isolated mitotic apparatuses were dispersed in the cold to fragment the centrosome. The granular suspension was briefly incubated with porcine brain tubulin, followed by further incubation with biotin-labeled tubulin. Microtubules reconstituted from labeled tubulin were visualized by rhodamine-labeled avidin. Bar: 5 um. THE 51-kD PROTEIN, A MAJOR PROTEIN COMPONENT OF MTOGs Starting from solubilized mitotic apparatuses, a fraction eluated by 0.5M KCl from the phos- phocellulose column contained a 51-kD protein as a major component and several minor compo- nents. The 51-kD protein is a basic protein with a pl of 9.8. Amino acid analysis of the 51-kD protein showed that the most prominent amino acid residue is lysine [21]. Polyclonal antibodies raised in rabbits against the 51-kD protein and affinity purified, stained only the center of asters reconstructed from tubu- lin and MTOGs prepared from the isolated mitotic apparatuses [21]. They also stained MTOGs re- constructed from the 0.5M KCI eluate, which astral microtubules were initiated. These from results indicate that the 51-kD protein is a major component of MTOGs in the mitotic apparatus. Antibody blocking experiments are necessary in order to further assess the role of the 51-kD protein in the initiation of astral microtubules from MTOGs. The MTOG fraction freshly prepared from the isolated mitotic apparatuses was preincu- bated with polyclonal antibody in the presence of 50% glycerol which stabilizes aster forming activ- ity, followed by the addition of tubulin. Compared with the control, in which incubation was carried out in the absence of the antibody, the formation of asters was substantially suppressed (up to 60%) by the antibody [20]. However, inhibition did not reach 100% in the presence of glycerol. Glycerol at a concentration of 50% exhibited an inhibitory effect of antibody binding to antigens [22]. When MTOG fraction prepared from freshly isolated MAs was first incubated with the antibody in the absence of glycerol followed by tubulin addition, aster formation was almost totally suppressed (up to 95%). Therefore, it seems safe to conclude that the 51-kD protein is involved in the initiation of microtubules from the MTOGs in vitro. 618 K. Onta, M. TortyaMa et al. INHIBITION BY Ca?*+ OF ASTER FORMING ABILITY OF MTOGs The effect of Ca?* on aster forming ability was examined using MTOGs prepared from isolated mitotic apparatus. The MTOGs were preincu- bated in a solution containing various concentra- tions of Ca**, followed by the addition of a tubulin solution containing EGTA to measure microtu- bule initiation. Figure 5 shows that concentrations of Ca’** as low as 1 uM begin to block the aster forming ability of MTOGs. The time course of inhibition showed that a preincubation of ~5 min with 20 “M Ca** was critical to the MTOGs loss of ability. An incubation of the Ca’*-treated MTOGs with an excess amount of EGTA prior to the addition of tubulin caused the MTOGs to restore the ability to initiate microtubules. Ca** has been believed to play a role in mitosis since microtubules of sea urchin eggs have been shown to be highly susceptible to a micromolar level of free Ca ions [23]. Localization of calmodu- lin in the mitotic spindle [24, 25] and sensitivity of number of asters 50 YA 6 5) 4 pCa Fic. 5. Inhibition of the aster forming ability of MTOGs by Ca. MTOGs from isolated mitotic apparatuses were first incubated with various con- centrations of Ca ions using Ca buffer for 20 min at 0°C. A tubulin solution containing excess EGTA was then mixed with the MTOG suspension and incubated for another 30 min at 35°C, followed by a count of the number of asters. Two series of independent experiments are shown. tubulin to Ca**/calmodulin [26, 27] suggest that Ca?t/calmodulin may be involved in de- polymerization of the spindle microtubules after anaphase, provided that free Ca ions are released from the Ca?*-sequestering vesicles [28] in the spindle. Although Ca ions seem to be necessary after anaphase [29], they must be sequestered at the beginning of the formation of the mitotic apparatus to support the microtubule nucleating function of the centrosome. INTRACELLULAR LOCALIZATION OF THE 51-kD PROTEIN DURING MITOSIS DETECTED BY IMMUNOCYTOCHEMISTRY USING MONOCLONAL AND POLYCLONAL ANTI- BODIES AGAINST THE 51-kD PROTEIN Monoclonal antibodies against the 0.5M KCl eluate on a phosphocellulose column were catego- rized into 4 groups [30]. Of these, HP1 and HP2 clones secreted IgG which cross-reacts with a protein of M. W. 51,000 in the eggs of the sea urchins, | Hemicentrotus pulcherrimus —_ and Pseudocentrotus depressus. On the other hand, the eggs of Clypeaster japonicus, Anthocidaris crassis- pina and Temnoplurus hardwicki were shown to contain a protein of a slightly higher molecular weight (52,000) that specifically reacts with both of the antibodies. Indirect immunofluorescence stainings of sec- tioned sea urchin eggs using monoclonal and polyclonal antibodies produced the same results [22, 30]. The 51-kD protein is localized in the centrosomal regions of the MAs from prophase through anaphase. At prometaphase, the astral microtubules are mostly attached to the centro- some so that the centrosphere cannot be observed at this stage. When the egg reaches metaphase, the proximal ends of the astral microtubules appear to detach from the centrosome, thereby forming the centrosphere region. The detached microtubules appear to be capped by granules at their proximal ends. It is possible that these granules are stained by the antibody. Figure 6a shows the staining of a metaphase egg with anti- tubulin antibody and Figure 6b, that of an early anaphase egg with anti-51-kD protein antibody showing immunofluorescence in the centrosomal 51-kD Protein in the Mitotic Apparatus 619 Fic. 6. Localization of the 51-kD protein in the mitotic apparatus. Paraffin-embedded sections were incubated with anti-tubulin antibody (a) and anti-51-kD protein antibody (b), followed by staining with fluorescein-labeled secondary antibody. a: Metaphase, b: Early anaphase. Bar: 10 um. region, the proximal region of the astral microtu- bules and the half spindles. The spindle is always stained with antibody. Although the function of the 51-kD protein in the spindle is unknown at present, it may have a role in stabilizing spindle microtubules which lie through the spindle matrix. A preliminary experiment showed that the 51-kD protein causes microtubules to line up side by side (unpublished data). It is necessary to determine the sites where the 51-kD protein is situated in the spindle with a spatial reference to spindle microtu- bules. If the 51-kD protein is actually involved in the initiation of microtubules in vivo, we can test another experimental design in such a way that unfertilized sea urchin eggs are artificially acti- vated to determine whether the centers of the cytasters can be stained with antibody against the 51-kD protein. This was the case for Hemicentro- tus eggs parthenogenetically activated by taxol [30]. MICROINJECTION OF MONOCLONAL ANTIBODY TO THE 51-kD PROTEIN INTO MITOTIC EGGS The polyclonal antibody was highly monospe- cific to the 51-kD protein [21]. antibody concentration was not high enough for However, the some other purposes, i.e., experiments such as microinjection for analyses of the effect of anti- bodies need _ higher IgG. Although one of the monoclonal antibodies, HP1 antibody, does not inhibit the aster forming activ- ity of the MTOGs prepared from the isolated MAs, it suppressed the formation of the MA when injected before prophase [22]. Therefore, cleav- age did not occur (Fig. 7). concentrations of Differential interfer- ence microscopy enables us to observe the astral rays in the Clypeaster egg. When the antibody was injected into the egg at prometaphase, however, growth of the astral microtubules was not inhib- ited, but the formation of the spindle was suppress- ed. The spindle formed is extremely short and in the form of a sphere (Fig. 8). This inhibition of spindle formation brings about total suppression of the karyokinesis. Sometimes, an incomplete fur- 620 K. Onta, M. Tortyama et al. F Fic. 7. Phase-contrast micrographs of a Pseudocentrotus egg injected with anti-51-kD protein antibody into one of the blastomeres after the first cleavage. Monoclonal antibody (6.1 pl) was injected into the lower blastomere. Note that the upper blastomere continued to divide. a: 120 min after fertilization, b: 148 min, c: 195 min. Bar: 50 um. Leh ee 3 Fic. 8. Polarized-light micrographs of a Clypeaster egg injected with anti-51-kD protein antibody at prometaphase [22]. a: Control, b: Injected. After injection, the growth of asters did not appear disturbed, but the spindle formation did. Bar: 10 um. (By permission of Springer-Verlag New York Inc.) row will be formed, but the furrow does not proceed further, and eventually regresses. The apparent contradiction which arises because the antibody that does not inhibit aster formation in vitro totally suppresses the formation of the MA can be elucidated if we assume that the antibody blocks the assembly of MTOGs into the centro- some. It is important to make further observations of the changes in the distribution of the MTOGs by injection of the antibody using the techniques of immunocytochemistry. In conclusion, the 51-kD protein was shown to be a component of MTOGs in the centrosome. This protein seems to be responsible for the initia- tion of microtubules in the formation of asters and the spindle when it is incorporated into MTOGs possibly associated with other constituents as yet unidentified. This series of work using the isolated mitotic apparatuses [19-22] is possible today, thanks to Professors Daniel Mazia and Katsuma Dan and their pioneering work on mitotic apparatus 36 years ago [31]. This paper is dedicated to Profes- sor Emeritus Katsuma Dan. ACKNOWLEDGMENT This work is supported by a Grant-in-Aid for scientific research from the Ministry of Education, Science and Culture of Japan (60065005). 14 15 51-kD Protein in the Mitotic Apparatus REFERENCES Pickett-Heaps, J. D. (1969) Cytobios, 1: 257-280. Pickett-Heaps, J. D. and Tippit, D. H. (1978) Cell, 14: 455-467. Brinkley, B. R. (1985) 145-172. Weisenberg, R. C. and Rosenfeld, A. C. (1975) J. Cell Biol., 64: 146-158. McGill, M. and Brinkley, B. R. (1975) J. Cell Biol., 67: 189-199. Ann. Rev. Cell Biol., 1: Snyder, J. A. and McIntosh, J. R. (1975) J. Cell Biol., 67: 744-760. Pepper, D. A. and Brinkley, B. R. (1979) J. Cell Biol., 82: 585-591. Gould, R. R. and Borisy,G.G. (1977) J. Cell Biol., 73: 601-615. Telzer, B. R. and Rosenbaum, J. L. (1979) J. Cell Biol., 81: 484-497. Kuriyama, R. and Borisy,G.G. (1981) J. Cell Biol., 91: 822-826. Berns,M. W., Rattner,J.B., Brenner,S. and Meredith, S. (1977) J. Cell Biol., 72: 351-367. Berns, M. W. and Richardson, S. M. (1977) J. Cell Biol., 75: 977-982. Robbins, E., Jentzsch, G. and Micali, A. (1968) J. Cell Biol., 36: 329-339. Endo, S. (1979) Cell Struct. Funct., 4: 71-74. Endo, S., Toriyama, M. and Sakai, H. (1985) In “Cell Motility: Mechanism and Regulation”. Ed. by H. Ishikawa, S. Hatano and H. Sato, Univ. Tokyo Press, Tokyo, pp. 403-414. 16 17 621 Endo, S. (1980) Dev. Growth Differ., 22: 509-516. Paweletz, N., Mazia,D. and Finze,E.M. (1984) Exp. Cell Res., 152: 47-65. Mazia, D. (1984) Exp.Cell Res., 153: 1-15. Endo, S., Toriyama,M. and Sakai, H. Dev. Growth Differ., 25: 307-314. Toriyama, M., Endo, S. and Sakai, H. (1984) Cell Struct. Funct., 9: 213-224. Toriyama, M., Ohta, K., Endo, S. and Sakai, H. (1988) Cell Motil. Cytoskel., 9: 117-128. Ohta, K., Toriyama, M., Hamaguchi, Y., Maeka- wa, S., Endo, S. and Sakai, H. (1988) Protoplasma, in press. Nishida, E. and Kumagai, H. (1980) J. Biochem. (Tokyo), 87: 143-151. Hamaguchi, Y. and Iwasa, F.(1980) Biomed. Res., 1: 502-509. Zavortink, M., Welsh,M.J. and McIntosh, J. R. (1983) Exp. Cell Res., 149: 375-385. Nishida, E., Kumagai, H., Ohtsuki, I. and Sakai H. (1979) J. Biochem. (Tokyo), 85: 1257-1266. Kumagai, H. and Nishida, E. (1980) Biomed. Res., 1: 223-229. Silver, R. B., Cole, R. D. and Cande, W. Z. (1980) Cell, 19: 505-516. Hepler, P. K. and Callaham, D. A. (1987) J. Cell Biol., 105: 2137-2143. Ohta, K., Toriyama, M., Endo, S. and Sakai, H. (1988) Cell Motil. Cytoskel., in press. Mazia, D. and Dan, K. (1952) Proc. Natl. Acad. Sci. USA, 38: 826-838. (1983) ZOOLOGICAL SCIENCE 5: 623-638 (1988) The Thermodynamics of Molecular Association in the Mitotic Spindle with or without Heavy Water (D0)! Hipemi SAto and JosEPH BRYAN?” Sugashima Marine Biological Laboratory, Nagoya University, Sugashima, Toba, Mie 517, Japan, and *Department of Cell Bioligy, Baylor College of Medicine, One Baylor Plaza, Houston, Texas 77030, U.S.A. ABSTRACT— Addition of heavy water (DO) or temperature shift to optimum range enhances spindle birefringence (BR) and volume in living, dividing cells. Knowing the form BR of spindle reflects the amount of oriented microtubules, we analyzed the association-dissociation reaction of the spindle with thermodynamic approach. Meiotic metaphase spindle of the mature oocyte of Pisaster ochraceus was chosen as experimental material. From BR of equilibrated metaphase spindle of living oocyte, and of spindles isolated in 12% hexylene glycol, pH 6.3, at various temperatures, we calculated thermo- dynamic parameters for the association reaction: In HO (19 parts of 0.53 M NaCl plus 1 part of 0.53 M KCl), AH=58.9 kcal/mole, AS=205.9eu, and AF=—1.1 kcal/mole. In 45% D.O, AH=29.55 kcal/mole, AS=106.3 eu, and AF=—0.9 kcal/mole. This reaction is observed between 3 and 13°C, at which temperature the spindle BR is maximum in both H,O and D,O. Kinetic data are obtained by BR measurements of rapidly isolated spindles after a varying interval following a temperature shift or the addition of D,O. Both association and dissociation processes appear to follow first-order kinetics. Activation energy (E,.:) for association reaction in HO is 41 kcal/mole; for D2O, 39 kcal/mole; and for dissociation reaction on removing D3O, 15 kcal/mole. For each condition these data are consistent with the hypothesis that the spindle BR measures the reversible polymerization of tubulin molecules into linearly aggregated polymers (microtubules) in a first-order reaction, even though associated or regulatory proteins are required in vitro tubulin polymerization. However, the difference of both AH and AS in H,O and D,O, and high but similar E,,, in the association reaction suggest the spindle reaction in fact to be a two step reaction such as: Tubulin dimer ai activated form Bis microtubules where, K, is primarily governed by D,O and Kz by temperature. The low E,,; in the back reaction of the D2O case suggests that the microtubules in vivo may first break down into random oligomers and then dissociate into tubulin dimers. feulated INTRODUCTION pices © 1988 Zoological Society of Japan thermodynamic parameters of the Heavy water (D2O) and temperature could in- dependently enhance the form birefringence (BR) and the volume of the living mitotic spindle [1, 2]. Response of the mitotic spindle to these two physical parameters was completely reversible and could be repeated several times within certain limits [1-3]. Thermodynamics of the deuterated and non-deuterated spindle in vivo [4, 5] indicated that significant differences existed between the Accepted April 9, 1988 ' This paper is dedicated to Emeritus Professor Katsuma Dan. deuterated and non-deuterated system [2-9]. However, few attempts were made to analyze the kinetics of the increase in volume and BR effected by temperature and DO. It partly could be due to the technical difficulties associated with measuring small retardation changes in the rapidly evolving system, even though some successful reports final- ly appeared to be real [2, 10, 11]. Another reason was based on the doubt of form BR yielded by mitotic spindle and reservation was expressed as spindle BR was not due to the oriented molecules or microtubules but yielded by the complex of gelated micells. However, Sato et al. [12] clearly demonstrated the applicability of Wiener’s equa- 624 H. Sato AND J. BRYAN tion to the spindle form BR, and proved that the spindle BR reflected exactly oriented amount of microtubules within spindle using the isolated mitotic spindle from mature oocyte of starfish, Pisaster ochraceus. This could be an ideal material to circumvent the problem of temperature or DO effect on mitotic spindle and to obtain imformation of comparative value on another system [4, 5, 7]. The mature oocytes of Pisaster ochraceus, whose spindle persisted in meiosis I for one hour at 13°C, provided a plentiful source of material, thus were used as experimental material. These mitotic spindles in meiosis I have a strong BR, physically stable, and exhibited relatively slow response to a variety of physical parameters or chemical rea- gents [2, 12, 13]. Using this material, we attempted to further refine the comparison between the deuterated and the non-deuterated systems and to begin a kinetic analysis of the molecular association of the mitotic spindle. We shall report our results obtained as follow- ings. MATERIAL AND METHODS Mature oocytes of the starfish, Pisaster ochraceus which were naturally arrested in meiosis I, were used in the present study due to the physical stability of the mitotic spindle. The methods of obtaining and preparing oocytes for spindle isolation and the procedures used for isola- tion have been described previously [12, 13]. The BR measurements used for the thermo- dynamic analysis were made on spindles isolated from oocytes which were washed twice in isotonic 0.53 M NaCl-27 mM KCI (“19:1” solution) and allowed to equilibrate in this solution at the given temperature for at least 5-10 min. The kinetic studies indicated that this was sufficient time to reach a new equilibrium. The kinetic measure- ments of BR were done by rapidly transferring oocytes into the “19:1” solution to the desired experimental condition (transfer time shoud be less than 10sec), then subsequently transferring one volume of oocytes to 10 volumes of isolation medium at various time intervals. The isolation medium consisted of 12% hexylene glycol and 10 mM potassium phosphate, pH 6.3 at 20°C [14]. Oocytes were spun in a hand centrifuge, the pellet collected and resuspended in 5 volumes of ice cold 12% hexylene glycol, pH 6.3 solution. The oocyte suspension was vortexed to lyse the oocytes and to free the spindles from the oocytes. Spindles were concentrated by centrifugation at 1,000g for 5 min. These concentrated preparations were stored on ice and retardation (BR) measurements were carried out within 6 hr. In particular case, correc- tions were made following the decay curve of BR of stored spindles [13]. This procedure served well to quickly stabilize the spindle at any given point in the reaction and preserve it for subsequent re- tardation measurement. Optical methods A modified Leitz Ortholux-polarizing micro- scope was used as the basic optical stand. A Zeiss Brace-Koehler type, rotating mica compensator with a retardance of 22.3nm was used as the measuring device. Rectified strain free optics [15] were used to permit precise retardation measure- ments of spindles with great accuracy. A Leitz illuminator with HBO 200WL 2 (Osram, the Netherlands) combined with Baird Atomic in- terference filter and heat absorbing filters to pro- vide an intense mercury green light (546 nm) was used as routine illumination. Polaroid HN-22 non-laminated polarizing sheet film mounted on a light tight plastic support was used as the polarizer. A built-in Leitz polarizing film was used as the analyzer mounted behind the objectives. Extinc- tion factor (Iparatte/Icross) WaS Maintained in an order of 10*. All photographic records were made on either Agfa KB-17 or Kodak Plus-X films exposed with the Leitz Orthomat automatic camera. Koehler illumination was maintained throughout all retardation measurements. Calculation of BR Since the retardation to be measured was much less than the wave length of illuminating light used for the measurements, the following formula was used (see, [1, 12]). sin AX sm Ax 5 comP .sin2 @’ Molecular Association of the Mitotic Spindle 625 where, @ =azimuth angle between the vibrating plane of the incident polarized light, @ ‘=measured compensation angle of the speci- men by Brace-Koehler compensator, A Xspec=retardation of specimen (or BR), AXcomp=known retardance of compensator (22.3 nm in our case). The standard deviations were calculated from the average triplicate measurements of 10 to 15 randomly sampled spindles at each point using the formula, oe where x =one value in a series of measure- ments, x =arithmatic mean of the series, N=number of values in the series. Electron microscopy Metaphase spindles isolated were immediately cooled to 4°C, washed twice, centrifuged at 1,250 xg for Smin, then fixed for 30min in 3% glutaraldehyde-12% hexylene glycol at pH 6.3. The preparations were washed twice with cold isolation medium, and postfixed for 30 min in 1% OsO,-hexylene glycol solution at pH. 6.3, then again washed twice with isolation medium. De- hydration started with 30% ethanol-12% hexylene glycol, then 50% ethanol-12% HG. Hexylene glycol was ommitted from the graded concentra- tions of ethanol above 70%. After dehydration, spindles were placed in propylene oxide, followed with a mixture of equal proportion of propylene oxide and Epon 812 (Shell Chem. Corp., CA). Embedding was accomplished in hard Epon rather than Araldite to provide good preservation of fine structure without shrinkage or compression upon sectioning. Sections were made with a Porter- Blum Ultramicrotome MK-II and examined with Phillips model 200 electron microscope. RESULTS Effect of D2O The effect of D,O on the spindle BR and volume was shown in Figure 1. Spindle BR and volume apparently increased as seen in bl and b2 of Figure 1. As shown in Figure 2, isolated spindle from the mature oocyte was mainly occupied by oriented microtubules and little else. Thus, D,O effect was studied by transferring mature oocytes in “19:1”—H.O solution to various concentration of “19:1”-D>0O solution. All concentrations stud- ied were done at 13°C, and equilibrium times were from 5 to 10 min. Examples of these experiments are shown in Figure 3, and from the retardation measurements, we noticed the spindle retardation became maximum at 45% D>,O-“19:1” solution, and similar results were obtained in 45% D,O- artificial sea water. If the oocytes in DO were transferred to “19:1”—H,O, spindles returned to normal size and BR. This effect was reversible and can be repeated several times. The spindles isolated at various concentrations showed a pronounced volume increase, which can be roughly estimated by considering the spindle to be a rotating ellipsoid whose long axis was equal to the pole-to-pole distance and whose short axis was equal to the width at the mid point between the poles, the equator. Using this model, we com- pared spindles with and without D,O and con- firmed an average 8 fold increase in volume in 45% D0 spindle. This indicated definite increase of spindle microtubules and number altered from 4,200 microtubules in control spindle (Fig. 2) to 10,000 microtubules in 45% DO spindles within 5 min. Pole-to-pole distance also increased suggest- ing elongation of each microtubules also oocurred within DO, not disturbing the initial distribution pattern of microtubules within the spindle. Figure 4 showed a cross section of DO spindle with a same magnification with Figure 2b. Although not commonly noted, rapid alteration of pH from 6.3 to 6.8 during spindle isolation produced microtu- bule dissociation and lateral split of microtubules. Likewise, in the early stages of dehydration, change in pH or osmolarity could easily induce “C-tubules” within isolated spindles even during or after glutaraldehyde or OsO, fixation [2, 12]. Thermodynamic analysis The retardation of spindles isolated from oocytes which have been fully equilibrated in “19:1”-H,O or —D,O, at various temperarture 626 H. Sato AND J. BRYAN bETOECRICUTIS! Fic. 1. Meiosis I spindles in mature oocytes of Pisater ochraceus. al and a2; A control spindle with different contrast. bl and b2; A spindle treated with 45% D,O is shown in different contrast. Polarization microscpy. Scale; 1 div.=10 yam. has been studied. Concentration of D,O was 45% because it was the best concentration in term of BR, spindle volume and reversibility (Fig. 3). Plots of retardation as absolute temperature (“K) for both series were given in Figure5. As ex- pected, the “19:1”-45% DO values were always greater that those in “19: 1”-H,O. Both plots have a maximum at 13°C, a temperature very close to the normal water temperature in which the mature organism was found. Further analysis of the association reactions was carried out using the equilibrium model of Morales and Inoué [16]. This model assumed that the spindle could be described by an equilibrium between oriented, birefringent material (B) and a pool of unoriented material (A, —B), where A, was the value of retardation measured with the polarization microscope. The equilibrium was then written as: A,-B—==B and a generalized equilibrium constant calculated from: Molecular Association of the Mitotic Spindle 627 Fic. 2. a; Tubulin immunofluorescence of isolated spindles from mature oocytes of P. ochraceus. b; Cross section of an isolated meiosis I spindle is mainly occupied by microtubules and little else. Some of the vesicles in the photograph are degenerated mitochondria resulting from the isolation procedure (cited from Sato er al. [8], pp. 556 as a control electron micrograph). K=B/A,—B A, was taken as the maximum value of retardation which B approached to asymptote as the tempera- Using this approach, thermodynamic parameters _ ture increased [4, 5]. In the present work, A, has were calculated. Morales and Inoué [16] suggested been used as the selected parameter to give the 628 H. Sato AND J. BRYAN best fit to van’t Hoff’s plot, using the assumption that AH was independent of temperature in a very limited temperature range. Two methods of esti- mating A, gave similar results. With this approach, the Arrhenius plots [log B/(A,—B) versus 1/temperature, where B was spindle retardation and A,, the asymptote B approaches] were made and the results were given in Figure 6. Two different A, values, 4.4 for H,O-spindle and 6.3 for D,O-spindle, were given to establish the van’t Hoff’s relationship. The Fic. 3. al and a2; An equilibrated spindle at 18°C. b1 and b2; A spindle treated with 45% D,O and equilibrated at 13°C. Polarization microscopy. Scale; 1 div.=10um. results obtained by Carolan et al. [4, 5] for Pecti- naria and those of Inoué [17] for Chaetopterus are also shown for comparison. The calculated para- meters for both H,O- and D,O-spindles indicated that the association reactions of spindles were endothermal with a large positive AH and have a large positive entropy change. The calculated parameters were consistently lower for D2O equili- brated spindles than for the H,O equilibrated spindles. In addition, the parameter A, for the D,O samples was significantly larger, suggesting Molecular Association of the Mitotic Spindle 629 Fic. 4. Electron micrograph of a cross section of an isolated spindle treated with 45% DO. Spindle width is increased but the density of microtubules per ~m* remains constant. Note the significant increase of C-tubules which are caused by the rapid association reaction. an increase in pool size. A similar analysis for the — and we believe this discrepancy could be caused by spindle association reaction has hardly been ap- adding new parameters made spindle association plicable for temperature above 13°C (see, Fig. 3), | reaction more complicate way and disturb the 630 Retardation (nm) H. Sato AND J. BRYAN VSiei—a)? Retardation vs. Temperature Standard Deviation prot sq - vatlein ey" ' N 6.0 5 Equilibriation Time 5.0 45%D20-19 1(Na:K) ia “Y - - a - Be - - - ae (bs - Ze XN =” A a rss 7 - 4.0 2.0 ia 0%D,0-19 1(Na:K) 276.15 278.15 280.15 282.15 284.15 286.15 288.15 290.15 292.15 T(°K) 4°c 7°c 10°c 13°¢ 15°c 16°C 19° Fic. 5. Retardation measurements of isolated spindles equilibrated at various temperatures with or without DO. Log B/(Ac —B) Spindle Association Reaction H,0 D0 OH= 58.9 Kcal /mole AH=29.55Kcal/mole 4S =209.5 eu 4S=106.5eu AF=-I.1Kcal/mole AF= -09 Kcal/mole Pectinaria HO D,0 } AH = 82 Kcal/ mole AH= 59 Kcal/mole AS= 286eu AS= 208eu Chaetopterus 4OH=29Kcal /mole 4S =l00eu 0.1 3.48 3.50 3.52 3.54 356 3.58 360 3.62 7K x 1073 Fic. 6. van’t Hoff plot of spindle association reaction. Molecular Association of the Mitotic Spindle 631 thermodynamic analysis. However, same tendencies were confirmed on the D,O depending spindle association in dividing sea urchin eggs [7, 9, 18], and even in vitro tubulin polymerization with DO [6]. Kinetic measurements A. Association reactions The time course of the association of tubulin molecules into the spindle microtubules was car- ried out in order to investigate the kinetics of these reactions. Temperature shift was used as the parameter, where oocytes were equilibrated for 10 min in “19:1”—H,O at 4°C, then transferred to the desired higher temperature. For D,O measurements, oocytes were equili- brated for 10 min at a given temperature in “19: ”_H,O, then transferred to “19:1”-45% DO at the same temperature. Spindles were isolated at various times using the method outlined pre- viously. Figure 7 showed photographic illustrations of isolated spindles of both methods in one sequence. Oocytes were equilibrated at 4°C in “19: 1”-H,O and transferred to “19: 1”—H,O at 13°C, then after 5 min finally transferrd to “19:1-D 0 (45%)” at 13°C. The increase in BR and spindle volume was quite apparent. The retardation values ranged from 1.2nm at 4°C to 5.3 to 5.5nm at 13°C in “19:1"-45% DO. Using ellipsoidal assumption described before, the volume change for the same condition was calculated as 70 to 80 fold. The time course of the association reactions in “19:1”-H,O using the temperature shift method was illustrated in Figure 8. These data have been analyzed in terms of the equilibrium model in the following manner. Even for complicated reac- tions, the return to equilibrium after a small per- turbation usually followed first order kinetics, therefore it was assumed that the rate of growth of spindle (microtubule polymerization) was pro- portional to the amount of unpolymerized mole- cules, the tubulin molecules. d/d, (polymer) ~ (monomer) In the symbols previously defined, this can be written as: which could be integrated and rearranged to give: logio (Ag —B)=—k’t+ constant 4°C 13°C 1’ 13°C 3’ Fic. 7. 13°C 5’ 13°C-45%D.20 " uw 1 3" 5 Isolated spindles at various steps in temperature depending or D,O depending reactions. al and a2; An equilibrated spindle at 4°C. When oocytes are teansferred to 13°C, spindle volume and BR rapidly increased (b1, b2; cl, c2) and reached to maximum in 5 min (d1, d2). When equilibrated spindles at 13°C are transferred to 45% DO at 13°C, again spindle volume and BR go up and reach maximum values in 5 min. Polarization microscopy. Scale; 1 div.=10 ~m 632 H. Sato AND J. BRYAN Association Reaction of the Mitotic Spindle Parameter 8 Temperature Retardation (nm) 0 2 3 4 5 6 10 Minutes Fic. 8. Temperature depending association reaction of spindle in “19: 1”-H,O. H20 ASSOCIATION REACTION 70.0 109), (Ao- B) > > / N“N ? 5 10°c 10.0 d/dt (Polymer) ~ (Monomer) 2 dB/dt = K (Ao-B) faB/(Ao-B8) = JR-at —In(Ao- B)= Kt +Constant, 13°c 50 10,9 (Ao—B) = -K t + Constant, K'=K/2.303 = (min"') R=-K't+Constant, —log (Ac~B) 7% 10g, (Ae-B) = — (3.00 x 10~?)t + 15509 10% log, (Ae~B) = — (5.202 x10-*)t +15957 [3° log, (Ae-B) == (11.335 x107#)t+ 16824 pe Ee (9) 2 4 6 8 10 Minutes Fic. 9. Rate of reactions of temperature depending association of mitotic spindles. Molecular Association of the Mitotic Spindle 633 k’ =k/2.303 The data were fitted to this equation using a least squares procedure; measurements from 2 to 6 min were used for this procedure. The results of this procedure was shown in Figure 9. The rate con- stants (k) from these measurements were: TC-k=11.5X10~4-sec™! 10°C-k =20.0 10~*-sec™! 13°C-k=43.5x10~*-sec”! Using these constants, the apparent activation energy or temperature characteristic for the H2O association was determined from the Arrhenius plot, as shown in Figure 10. The calculated values from the relation: d log k Pier d(1/T) then, E,.,.=41.3 kcal/mole. A similar analysis has been carried out for the D.O transfer method. The time course of these reacrions is shown in Figure 11. The first order rate constants obtained from these data were: TC-k=11.3X10~*-sec”! 10°C-k=23.8107*-sec™! H,0 Association Reaction 50.05 i ° °4 K7°= 11 5124 x 10°4(sec™') x K10° = 79.9653 x 10-4(sec"') S | K13°= 43.5037 x 107*(sec-!) Lo) is 4 5.04 log 10*K = —9.0 x 103 (1/T) + 33.1213 Fo= -p atnk d(I/T) =-(1.99)(2303)(-9.0x103) Eact. = 41.25 Kcal/mole sal T T + T 7 =| 3.48 3.50 352 3.54 3.56 358 \/T (x1073 °K7!) Fic. 10. Arrhenius plot and the kinetic data obtained from the temperature depending association reac- tions. Association Reaction of the Mitotic Spindle by D2O 13°C = OW 45% D20 lO2 Cis eh ——— 7°C= " Retardation (nm) 0 | 2 3 4 6 10 Minutes Fic. 11. Association reactions depending on “19:1”-45% D.,O of mitotic spindles at given temperatures. 634 H. Sato AND J. BRYAN 13°C-k=48.3 x 107*- sec! B. Dissociation reaction The time course of the dissociation reaction of the spindle has been studied using the reverse of the procedures described for initiating association. Oocytes in “19:1”-45% DO at a given tempera- D2O Association Reaction O— 45% Dz20 ture were transferred to “19: 1—-H,O” at the same . temperature. The results for this type of experi- An Arrhenius plot from these values showed in Figure 12 gives an apparent activation energy of 39.3 kcal/mole. 500 ment were shown in Figure 13. These data were treated as follows; it was assumed that the initial rate of dissociation was proportional to the initial amount of oriented material (B), then: dB ———~B dt or, = 100 3 dB = K 7°c = 11.284 x 10°* ap kr BI me K 10% = 23834 x10 * 5 50 K13%¢ = 48.267 10-*(sec~!) which can be integrated to give: log 104K = -8.55 x 109(1/T) + 31.5680 3 d ink log;o9B= —k’ _,-t+constant Ea= -R — d(1/T) = —(1.99)(2.303)(-8.55 x 10°) and, = 39 3Kcal /mole , K_, k 1 2.303 The calculated first order rate constants from '- : ; : , . = this procedure at each temperature were: 3.48 3.50 352 354 356 3.58 WT Gsig-2°K) 13°C-k=24.2 10 *- secs" Fic. 12. Arrhenius plot and the kinetic data obtained 10°C-k=18.3* 1074: sec! from the D,O depending association reaction. PC-k=13.9X 107*-sec—! Dissociation Reaction at Various Temperatures 6.0 Parameter = 45% D»0 H90 5.0 \ > y a » be . | SS Se 240 ~~} ne c a rae eo > = — : ~n ~~_ J ~--------#--------- |3°C--~--—--- = a) eee iS) ae ee ee ee One ee ee eee = 3.0 ae eee 10°C =) => - - -- —-------7% --— - e ® (oa 2.0 1.0 I-71. or — fe) \ 2 3 4 5 6 10 Minutes Fic. 13. Dissociation reaction of mitotic spindle subtracting 45% D.O at given temperatures. Molecular Association of the Mitotic Spindle 635 D2O Dissociation Reaction 0% 0%D,0 10° : {igre 45% eo 100 ° eo ei . 2 oO Bs 50 Logiy 1O*K_,= -322x10°(I1/T) + 11.6314 rae _p din K-I d(I/T) = -2.303 R(-3.22 x10?) =14.8 Kcal/Mole 10 ST A [a 3.48 350 352 3.54 3.56 358 VT (x107? °k7!) Fic. 14. Arrhenius plot and kinetic data obtained from D.,O depending dissociation reactions. These values gave the Arrhenius plot illustrated in Figure 14. The apparent energy of activation calculated from this plot was E,.,= 14.8 kcal/mole. When temperature jump method was used from higher to lower temperatures, somewhat aberrant results were observed. The spindles lost oriented material much more rapidly at the lower tempera- The first order decay constants for this process were: tures. 4°C-k=68 x 10~4-sec! 7TC:-k=30.6X10~*-sec™! 10°C-k =23.110~*-sec™! The use of an Arrhenius plot and these data led to an apparent negative activation energy. DISCUSSION Isotope effect of heavy water, especially D,O, was one of the attractive subjects and many reports were published concerning the biochemistry and the physiology of living and dividing cells [19-23]. However, heavy water depending enhancement of spindle BR and volume was founded by Inoué and Sato [1] using the metaphase arrested meiosis I spindle of mature oocyte of Pectinaria gouldi and spindle in dividing egg of sea urchin, Lytechinus variegatus. The increase of volume and BR of spindle depended on the concentration of DO, but the maximum increases were obtained by applying 45% D,O during the metaphase or onset of anaphase. This maximum concentration to achieve the maximum effect of DoO was common for metaphase spindles in eukaryotes even in tissue culture cells [3, 8]. The DO depending associa- tion reaction was rapid, and a new state of equilib- rium being reached within 90sec in developing Japanese sea urchin eggs [9, 18], about 2 min in Pectinaria oocyte [4, 5], and required 5 min in mature oocytes of present material. The D,O effect is completely reversible and can be repeated many times on the same spindle. H,O'* has no effect whereas HDO and Hbo'® have a half effect of DxO. pD, which can be expressed as p-H and; p H=pH reading + } pH where, § pH=0.3314-n+0.0776-n* and n is the mole fraction of DO, has also no significant effect for the spindle association reaction. In many respects, D2O effect is quite similar to the elevat- ing temperature within the physiological range. However, the spindle became overstabilized or freezed when higher concentration of D2O was applied. From the electron micrographs shown in Figures 2 and 4, we confirmed the coefficient of BR, (n.— no), was 5x10 *, and the average measured densi- ty of microtubules in both types of spindle was constant and 108/um*. However, microtubule number increased from normal 4,200 to 10,000 in DO, and also microtubules were elongated [8]. When observed in the living and dividing sea urchin eggs under the polarized light, the spindle displayed its dynamic nature and appeared as a very labile structure. To explain this dynamic state, Inoué [15, 17, 24] and Inoue and Sato [1] postulated a “dynamic equilibrium model” for the molecular nature of spindle. In this model, the birefringent spindle fibers were composed of oriented microtubules which were in equilibrium with a pool of polymerizable tubulin dimer. To 636 H. Sato AND J. BRYAN translate this model into terms which could be quantitated, we presumed that there were only two states of the spindle molecules (tubulin), poly- merized and unpolymerized. Oriented materials, microtubules, could be measured directly as re- tardation (B). The maximum retardation was taken as a measure of the total molecules available for spindle assembly (A,). The amount of un- oriented material at any given temperature was the difference between the two. This type of model gives a linear relationship on a van’t Hoff plot of; log B/(A,—B) vs 1/T. From this, it was deduced that the orientation (tubulin polymerization) was an endothermal process with a large positive heat of activation (AH) and a very large increase in entropy (AS). It was subsequently proposed that the high heat and entropy increase could be ex- plained on the basis of hydrophobic or apolar interactions. Higher temperatures were believed to dissociate bound water from the protein subunit and therefore permitted them to interact more strongly. Several problems arose in the actual application of the two state model to interprete present re- These primarily concerned with the measurement of the total unpolymerized material [25-27], the interpretation of microtubules, and interpretation of the effect of D2O substitution on the calculated thermodynamic parameters. Pre- vious attempts appeared to underestimate the total amount of material which was potentially available for incorporation into the spindle [4, 5, 17]. First, the addition of D,O invariably produced retarda- tion which was larger than the original estimates of available material [7, 18, 22, 27, 28]. Observations on spontaneously occurring tri- or tetra-polar spin- dles or the tubulin paracrystals induced by vinblas- tine and Colcemid [2, 18] apparently indicated the pool size might be a great deal larger than any of the previous estimates. In fact, a ten fold increase sults. in the amount of organized material was not uncommon in the tetrapolar configuration in the presence of DO. It appeared that a given orient- ing centers could interact with or assemble only a limited amount of the total available material [26]. The addition of more orienting centers resulted in the assembly of more monomeric or polymerizable material. The substitution of DO, on the other hand, appeared to strengthen the interactions of a given orienting center [10,28]. Despite these obvious difficulties which arose in attempting a molecular interpretation, the two state model seems to be the best first approxima- tion available and its use was justified until suf- ficient contradictory evidences were available. As was pointed out previously, the technical difficulties involved in making very rapid retarda- tion measurements prohibited studies on the time course of spindle assembly. This problem has been overcome through the use of rapid stabilization of spindles and the choice of a more slowly respond- ing organism. The results demonstrated that when the spindle equilibria were shifted only a small fraction, the kinetics governing the response could be adequately described by simple first order dif- ferential equations. This did not imply that the reactions themselves were necessarily first order. Rather than to follow the usual procedure of calculating relaxation times, the data have been expressed in terms of Arrhenius plots and the corresponding apparent energies of activation (Eact). This allowed a comparison of the present data with a number of other processes. The calculated E,., for the assembly process were rather large, approximately 40 kcal/mole. These values are greater than those usually recorded for specific enzymatic reactions but approach the values measured for the denaturation of several proteins. This suggested that the polymerizable tubulin underwent a conformational change upon assembly. Similar energies of activation have been reported. The energies of activation for the D2O and H,O reactions were very similar; 39.3 vs 41.3 kcal/mole. This fact and the numerical similarity of the rate constants at each temperature stongly supported the idea that there was no difference between the mechanism of assembly in D,O and H,0. The partial disassembly of the spindle after transfer from DO to HO was well behaved in the sense that it followed an Arrhenius type rela- tionship with a positive Ear. The Ect, 14.5 kcal/ mole, was approximately one-third that for the assembly step, suggesting there was less of an energy barrier to dissociation. The dissociation reaction of the spindle studied Molecular Association of the Mitotic Spindle 637 using temperature shifts from higher to lower T produced somewhat anomalous results which sug- gested that the mechanism of temperature depend- ing dissociation might be more complex. Here, the rate of retardation decrease became greater at lower temperatures. This led to the calculation of negative E,. It seems more reasonble to believe that perhaps spindle microtubules which presum- ably assembled by end or ends [28], treadmilling [29], dynamic instability [10] or break and mend [10, 30, 31], we feel it dissociated in a more complex manner by breaking into smaller pieces or by shortening and splitting. In order to adequately describe the present results, we believe two state equilibrium model [17] must be modified to some extent. One mod- ification was to assume that it was not possible to adequately measure the pool size which should be very large compared to the amount of oriented material. This modification led to the conclusion that AH was no longer independent of tempera- ture. This type of modification should be con- sidered in the future. An alternative modification was the introduction of second reaction, primarily introduced by Taka- hashi and Sato [18]. We suggest Figure 15 as a possible model. Two reactions were involved: one of which was primarily influenced by temperature, the other being mainly influenced by D,0. The breaking to smaller fragments is shown to illustrate the possible complexity of the dissociation by temperature. D020 Temperature Ky K2 —_—_—_—_—_—_> _—__ A go AO 4 SS B Tubulin dimers Activated form Microtubules (Spindle Proteins) (Mitotic Spindle) \ Randomized / Oligomers Fic. 15. A model postulated to interprete the tempera- ture or D,O depending spindle association and dissociation reaction. The temperature sensitive reaction was the rate limiting step for which the energies of activation have been measured. This reaction was tacitly assumed to be the assembly reaction although other interpretations were possible. The effect of D,O on the temperature sensitive step was reflected in the difference in AH and AS for the normal and DO substituted systems. As de- scribed before, the decrease in AH and AS upon D0 substitution suggested strongly that hy- drophobic or apolar interactions were not in- volved. We believe we are still far from real mechanism of the assembly and disassembly of tubulin or microtubule dynamics in vivo. Further study must be carried on physico-chemical bases. ACKNOWLEDGMENTS This work was supported by Grant-in-Aid for Scientific Research provided from the Ministry of Education, Science and Culture of Japan, numbered as 60480020, and Grant-in-Aid for Co-operative Researches num- bered as 613040008, 62300004 and 62304062, and also Grant-in-Aid for Developmental Scientific Research numbered as 62890004. We wish to extend our sincere thanks to all faculty and staff members in the Sugashima Marine Biological Laboratory, Nagoya University, for their kind coopera- tion and assistance during the course of present work. We also extend our sincere appreciation to Dr. R. Fernald’, who had been devoted for the Friday Harbour Laboratories as the director, and the staff of that labo- ratories where we made most of spindle isolations and retardation measurements during the summers of 1967 to 1969. Assistance for electron microscopy appreciated to Mrs. Bush’. REFERENCES 1 Inoué, S. and Sato, H. (1967) Cell motility by labile association of molecules— The nature of mitotic spindle fibers and their role in chromosome move- ment. J. Gen. Physiol. , 50: 259-292. Sato, H. (1975) The mitotic spindle. In “Aging Gametes”. Ed. by R. Blandau, S. Kager A. G., Basel, pp. 19-49. 3 Sato, H., Kato, T., Takahashi, T. C. and Itoh, T. J. (1982) Analysis of D2O effect on in vivo and in vitro tubulin polymerization and depolymerization. In “Biological Functions of Microtubules and Related Structures”. Ed. by H. Sakai, H. Mohri and G. G. Borisy, Academic Press, Tokyo, pp. 211-226. 4 Carolan, R.M., Sato, H. and Inoué,S. (1965) A thermodynamic analysis of the effect of DO and H,0 on the mitotic spindle. Biol. Bull., 129: 402. (Abstract) 5 Carolan,R.M., Sato,H. and Inoué,S. (1966) Further observations on the thermodynamics of the bho 10 11 14 15 16 17 18 19 638 living mitotic spindles. Biol. Bull., 131: 385. (Ab- stract) Ito, J. T. and Sato, H. (1984) The effect of deute- rium oxide (7H,O) on the polymerization of tubulin in vitro. Biochim. Biophys. Acta, 800: 21-27. Salmon, E. D. (1975) Spindle microtubules: Ther- modynamics of in vivo assembly and role in chromo- some movement. Ann. New York Acad. Sci., 253: 383-406. Sato, H., Ohnuki, Y. and Sato, Y. (1979) Assembly and disassembly of the mitotic spindle. In “Cell Motility: Molecules and Organization”. Ed. by S. Hatano, H. Ishikawa and H. Sato, Univ. Tokyo Press, Tokyo, pp. 551-568. Takahashi, T. C. and Sato, H. (1982) Thermody- namic analysis of the effect of D,O on mitotic spindle in developing sea urchin eggs. Cell Struct. Funct., 7: 349-357. Mitchison, T. J. and Kirschner,M. W. (1984) Dynamic instability of microtubule growth. Nature (Lond.), 312: 237-242. Salmon, E. D., Leslie, R. J., Saxton, W. M., Karow, M. L. and McIntosh, J. R. (1984) Spindle microtu- bule dynamics in sea urchin embryos: Analysis using a fluorescein-labeled tubulin and measurements of fluorescence redistribution after laser photo- bleaching. J. Cell Biol., 99: 2165-2174. Sato, H., Ellis, G. W. and Inoué, S. (1975) Micro- tubular origin of mitotic spindle form birefringence: Demonstration of the applicability of Wiener’s equation. J. Cell Biol., 67: 501-517. Bryan, J. and Sato, H. (1970) The isolation of the meiosis I spindle from the mature oocyte of Pisaster ochraceus. Exp. Cell Res., 59: 371-378. Kane, R. E. (1965) The mitotic apparatus: Physical- chemical factors controlling stability. J. Cell Biol., 25: 137-144. Inoué, S. (1981) Cell division and the mitotic spin- dle. J. Cell Biol., 91: 131-147. Morales, M. and Inoué, S. (see, 17) Inoué, S. (1959) Motility of cilia and the mechanism of mitosis. Rev. Mod. Physics, 31: 402-408. Takahashi, T. C. and Sato, H. (1984) Yields of tubulin paracrystals, vinblastine-crystals, induced in unfertilized and fertilized sea urchin eggs in the presence of D,O. Cell Struct. Funct., 9: 45-52. Gross, P. R. and Spindel, W. (1960) The inhibition of mitosis by deuterium. Ann. New York Acad. 20 im) ie) 23 i) Nn 26 27 28 29 30 31 H. Sato AND J. BRYAN Sci., 84: 745-754. Gross, P. R. and Spindel, W. (1960) Heavy water inhibition of cell division: An approach to mecha- nism. Ann. New York Acad. Sci., 90: 500-522. Kritchevsky, D. (1960) Deutrium isotope effects in chimistry and biology. Ann. New York Acad. Sci., 84; 573-781. Marsland,D. and Zimmermann, A.M. (1965) Structural stabilization of the mitotic apparatus by heavy water in the cleaving eggs of Arbacia punc- tulata. Exp. Cell Res., 38: 306-313. Thomson, J. F. (1963) Biological Effects of Deuter- ium. A Pergamon Press Book, Macmillan Co., New York. Inoué, S., Fuseler, J., Salmon, E. D. and Ellis, G. (1975) Functional organization of mitotic microtu- bules. Physical chemistry of the in vivo equilibrium system. Biophys. J., 15: 725-744. De Brabander, M., Geuens, G., Nuydens, R., Wil- lebrods, R. and DeMey,J. (1981) Microtubule assembly in living cells after release from nocoda- zole block. Cell Biol. Int. Rep., 5: 913-920. Harris, P., Osborn, M. and Weber, K. (1980) Dis- tribution of tubulin containing structures of the sea urchin Strongylocentrotus purpuratus from fertiliza- tion through first cleavage. J. Cell Biol., 84: 668- 679, Sluder, G. (1976) Experimental manipulation of the amount on tubulin available for assembly into the spindle of dividng sea urchin eggs. J. Cell Biol., 70: 75-85. McIntosh, J. R. (1979) Cell division. In “Microtu- bules”. Ed. by K. Roberts, and J.S.Hymes, Academic Press, New York, pp. 428-441. Margolis, R. L. and Wilson, L. (1981) Microtubule treadmills—possible molecular machinery. Nature (Lond.), 293: 705-711. Salmon, E. D., Saxton, W. M., Leslie, R. J., Karow, M. L. and McIntosh, J. R. (1984) Diffusion coefficient of fluorescein-labeled tubulin in the cyto- plasm of embryonic cells of sea urchin Lytechinus variegatus: Measurement by video image analysis of fluorescence redistribution after photobleaching. J. Cell Biol., 99: 2157-2164. Saxton, W. D., Stemple, D. L., Leslie, R. J., Sal- mon,E.D, Zavortink,M. and McIntosh, J. R. (1984) Tubulin dynamics in cultured mammalian cells. J. Cell Biol., 99: 2175-2186. ZOOLOGICAL SCIENCE 5: 639-644 (1988) Metaphase to Anaphase Transition of Sea Urchin Eggs Examined in Caffeine-Induced Monasters! PATRICIA J. HARRIS Department of Biology, University of Oregon, Eugene, Oregon 97403, U.S.A. ABSTRACT—Monasters were produced by treating mitotic sea urchin eggs for 15 min with 10 mM caffeine. Microtubules of the mitotic apparatus disappeared and the centrosomes moved to the chromosomes. Eggs that had not yet passed the metaphase/anaphase transition regressed to early prophase, while the anaphase eggs continued on to form interphase nuclei. On recovery in normal sea water, all eggs formed a large monaster, but some were just entering mitosis, while the others were entering interphase, a half cycle out of phase. New microtubules growing from the centrosomes of the anaphase eggs were resistant to the caffeine treatment. The difference in stability between the old and the new microtubules may reflect some changes occurring in the centrosome itself at the transition to © 1988 Zoological Society of Japan anaphase. system for studying mitosis. INTRODUCTION The transition from metaphase to anaphase is marked by the abrupt onset of a number of proces- ses, the most conspicuous being the separation of the chromatids and the beginning of their move- ment toward opposite poles. There are other pro- cesses associated with mitosis, however, that begin long before this transition point and extent long after it. For example, a poleward force is exerted on the chromosomes probably as soon as they become attached to kinetochore microtubules (MTs) in prometaphase [1], and in the astral mitosis of sea urchin zygotes, asters and their centrospheres enlarge continuously from the time of their origin in early prophase until late anaphase and telophase [2]. A useful tool for studying these processes is the mitotic monaster, frequently found naturally in newt lung cell cultures, where they have been used for the analysis of prometaphase chromosome movements [3], and artificially produced in sea urchin eggs treated with mercaptoethanol, where they have been used for studies of centrosome Accepted March 17, 1988 ' This paper is dedicated to Professor Katsuma Dan. The large size of the caffeine-induced monasters makes them an excellent experimental duplication [4]. The structural characteristics of monasters produced by artificial activating agents in unfertilized eggs have been explored by Paweletz and Mazia [5]. Monasters can also be produced by treatment of prometaphase sea urchin eggs with 10 mM caffeine for 10-15 min [6]. This causes the breakdown of aster MTs and shrinkage of the spindle until the two centrosomes are brought together at the metaphase plate. On recovery a large monaster, twice the size of the normal mitotic aster, grows and proceeds through all the stages of mitosis. The size of this aster makes it especially useful for experimentation. Recently Mazia et al. [7] have reported that in- cubation of prometaphase eggs at 0°C for 18 hr shrinks the spindle and causes the two centrosomes to fuse at the metaphase plate in a manner very similar to that after 15 min in caffeine. Described in this paper are the results of one of a number of experiments using a 15 min pulse of 10 mM caffeine to produce monasters. The inherent asynchrony in different batches of eggs, and the difficulty in determining the exact stage at which caffeine is added often results in samples contain- ing intact prophase nuclei or eggs that have ad- vanced into anaphase, and confuses the interpreta- tion of results. In this case approximately half the 640 P. J. Harris eggs were either in prometaphase or early ana- phase with no polyspermy, making the stages easy to distinguish. The recovery patterns of each of these populations of eggs is compared and the significance of their differences with regard to the transition from one MT system to another is discussed. MATERIALS AND METHODS Gametes of the sea urchin Strongylocentrotus purpuratus were obtained by injection of 0.5M KCl. Eggs were fertilized and the fertilization membranes removed by addition of mercapto- ethylgluconamide (Vega Biochemicals, Tucson, Arizona) at a concentration of 0.1 g/100 ml of egg suspension. After 15 min the softened membranes were stripped by passing through Nitex mesh. The eggs were washed by settling and decanting, and then incubated in normal filtered sea water at 12°C with constant stirring. At prometaphase, eggs were collected and re- suspended in 10 mM caffeine in sea water for 15 min, then washed and allowed to recover in nor- mal sea water. Samples were taken at 10 min intervals through 60 min, at the time of the control second division. Eggs were fixed in 1% paraform- aldehyde in 0.4M Na acetate pH 6.5 for 20-30 min, dehydrated in ethanol series, and embedded in 20% methyl/80% butyl methacrylate in gelatine capsules, as described by Harris and Rubin [8]. For brightfield light microscopy, 1.0 ~m thick sec- tions were stained with 1% azure 2, 1% methylene blue in 1% Na borate and mounted under cover- slips with immersion oil. RESULTS Examination of sectioned control eggs fixed at the time of beginning caffeine treatment showed that there were approximately equal numbers of prometaphase (Fig. la) and anaphase (Fig. 1b) stages. Actually, Figure la is probably closer to late metaphase than prometaphase, but represents the latest stage before anaphase onset. After 15 min in 10 mM caffeine the asters of the promet- aphase eggs had disappeared and the spindle had shrunk, moving the centrosomes (i.e. centrioles and peri-centriolar material) to the metaphase plate, where they often fused to form a single spherical body surrounded by the chromosomes (Fig. 2a). In the anaphase eggs (Fig. 2b) the asters and kinetochore MTs were also depolymerized, and the centrosomes were apparently moved to the chromosomes, even though the chromosomes were no longer held together at the metaphase plate. Unlike in the prometaphase eggs, a number of MTs radiated from the centrosomes and also appeared to traverse the interzonal region between the separated sets of chromosomes. When caffeine treated eggs were returned to normal sea water to recover for 10min, pro- metaphase eggs formed a monaster, with MTs growing out from the centrosome and apparently pushing the condensed chromosomes outward as they grew (Fig. 3a). The resulting structure might be considered a monaster prometaphase. In the Fics. 1-5. monaster formation. Fic. la,b. Mitotic stage at time of caffeine treatment. Comparison of prometaphase (vertical column a) and anaphase (vertical column b) at different stages of The metaphase figure in la is the latest stage in this population. The mid-anaphase stage in 1b is midway between very early anaphase and early telophase found in this population. Fic. 2a,b. After 15 min in 10 mM caffeine-sea water. The centrosomes move to the chromosomes in both cases. Within this time period no MTs are present in metaphase eggs, but are seen originating from the centrosomes at the poleward sides of the two sets of anaphase chromosomes. Fic. 3a,b. After 10 min recovery in normal sea water. All eggs have monasters, but the prometaphase eggs form a mitotic figure, while the anaphase eggs produce an interphase MT system. The two populations are now a half-cycle out of phase. Fic. 4a,b. After 40 min recovery. Prometaphase eggs remain in the mitotic state, but increase the size of the centrosphere. The anaphase eggs are entering prophase of the next division. Fic. 5a,b. After 60min recovery. Prometaphase eggs begin to decondense their chromosomes and enter interphase. Anaphase eggs are now entering the next division, which is directly from one to four cells. Caffeine-Induced Monasters 641 x ater 2 642 P. J. Harris anaphase eggs (Fig. 3b), the chromosomes decon- densed and began to fuse to form a centered interphase nucleus, while aster MTs continued to grow to form a large monaster. Thus, while all eggs in the 10min recovery sample had large monasters, they actually represented two separate populations that were a half division cycle out of phase. After 40 min recovery, the most conspicuous change in the prometaphase eggs was the great enlargement of the aster center, or centrosphere (Fig. 4a). The chromosomes remained condensed at the periphery of this region and aligned radially with respect to its center. Aster MTs inserted at the periphery, but did not penetrate the centro- sphere. In some sections, aggregates of material in the center of this region contained one and some- times two densely staining dots, possibly the cen- trioles. The reconstitution of the nucleus in the anaphase eggs was accompanied by a flattening of the nucleus and formation of a clear disk similar to the classical streak of the normal division cycle. During this time the centers or centrosomes, re- taining their MTs, became separated and associ- ated with the surface of the nucleus. Figure 4b shows a nucleus following the streak formation, with MTs somewhat reduced in number just before the beginning formation of mitotic asters. The centrosphere region of the original pro- metaphase eggs continued to grow, until at 60 min recovery it reached a diameter at least five times that of the 10 min recovery sample. At this time the chromosomes began to decondense and coalesce to form interphase nuclei. There was some asynchrony, but most eggs were in the stage pictured in Figure5a. Enlarging chromosomal vesicles moved toward the center of the aster, and new MTs originating from the aggregated material in the aster center rapidly grew to form a new monaster similar to the 10 min recovery sample of the original anaphase population. Further de- velopment of the prometaphase population apparently followed the same course as the ana- phase population, but with approximately 60 min delay. While the nuclei were reforming in the prometaphase eggs, the anaphase eggs had entered mitosis, with most at metaphase or anaphase of a one-to-four division (Fig. 5b). A description of the later division cycles of these populations will be reported elsewhere. DISCUSSION The actual mechanism by which caffeine affects the mitotic apparatus is not known, but indirect evidence suggests that it may bring about a release of intracellular calcium by affecting the calcium sequestering system [9, 10]. Whatever the mechanism is, the rapid depolymerization of MTs and the equally rapid recovery of the ability to reassemble them makes caffeine a useful tool for studying mitotic events. In the studies reported here, two populations of eggs within the same experimental sample, those just before and those just after the metaphase to anaphase transition, and thus only a few seconds apart in developmental time, were separated by half a cell cycle by treatment with caffeine. Those eggs that had not passed the transition point regressed to the start of mitotic apparatus forma- tion, while those that had passed that point con- tinued their development. The role of calcium in the metaphase to ana- phase transition is suggested by calcium-dependent changes in chlorotetracycline fluorescence at anaphase onset [11], anaphase acceleration or de- lay with injection of EGTA-calcium buffers in mammalian cells [12], and a calcium transient at the onset of anaphase in sea urchin eggs [13] as well as in mammalian cells [14], detected with the fluorescent probe Fura-2. The exact function of the calcium release is not known, but it may be a signal for the dephosphorylation of the cyclically phosphorylated mitotic proteins, which occurs at the onset of anaphase [15]. At least some of these phosphorylated proteins are associated with the centrosome [16]. One of the consequences of the metaphase to anaphase signal appears to be some change in the ability of the centrosome to nucleate MTs. After 15 min in the 10 mM caffeine sea water, the pro- metaphase centrosome showed no evidence of MT nucleation, while in the anaphase eggs MT growth from the centrosomes indicated either new nuclea- tion or continued growth onto already nucleated MTs. These MTs may represent the new growth Caffeine-Induced Monasters 643, that begins at late anaphase and telophase in untreated eggs [8]. In both the prometaphase and the anaphase eggs, the old mitotic apparatus disappeared com- pletely, demonstrating a difference in stability between the old and the newly forming MTs. One explanation for this difference may be that the centrosome loses its association with the old MTs before it begins to support new MT growth in anaphase. Endo [17] has described changes in the clusters of granular material around the centrioles of sea urchin eggs during mitosis, and noted that at metaphase there is a sharp separation of the granu- lar clusters from the surrounding radial MT zone. Endo suggested that by metaphase the main func- tion of these clusters has been accomplished, although she notes their association with MTs again at later stages. These later MTs may be those we see at anaphase, which are stabilized by their attachment to the centrosome. The growth of the centrosphere in the recover- ing prometaphase eggs is quite spectacular. Un- fortunately we have not yet found a good antibody to centrosomes that will react with our sectioned material, and so we can say little about the expan- sion and other shape changes of the centrosome as described by Mazia [18] and Schatten et a/. [19]. In the paraformaldehyde fixed eggs described here, centrioles stain as very small dense dots, while the pericentriolar material that makes up the bulk of the centrosome is very lightly stained and difficult to identify. Therefore it is impossible to say from this work how much of the centrosomal material remains attached to the proximal ends of the aster MTs in these enlarged centrospheres, how much remains in the aster center, and how much may be distributed back into the cytoplasm. The brief paraformaldehyde fixation and methacrylate embedding permit immunofluores- cence localization of tubulin and fluorescent stain- ing of the chromosomes. We hope to develop further probes to study the distribution of mitotic proteins. Certainly the caffeine-induced monas- ters provide an excellent experimental system for these studies. ACKNOWLEDGMENTS This work was supported by grant PCM 8409573 from the National Science Foundation (U.S.A.). REFERENCES 1 Bajer, A.S. and Molé-Bajer, J. (1972) dynamics and chromosome movement. Int. Cytol. (suppl.), 3: 1-271. Wilson, E. B. (1928) The Cell in Development and Heredity, The Macmillan Co., New York, 3rd ed. 3 Bajer, A.S. (1982) Functional autonomy of monopolar spindle and evidence for oscillatory movement in mitosis. J. Cell Biol., 93: 33-48. 4 Sluder, G. (1978) The reproduction of mitotic cen- ters: new information on an old experiment. In “Cell Reproduction: in Honor of Daniel Mazia”. Ed. by E. R. Dirksen, C. M. Prescott and C. F. Fox, Academic Press, New York, pp. 563-569. 5S Paweletz, N. and Mazia, D. (1979) Fine structure of the mitotic cycle of sea urchin eggs activated by ammoniacal sea water. Eur. J. Cell Biol. , 20: 37-44. Spindle Rev. nN 6 Harris, P. J. (1983) Caffeine-induced monaster cycl- ing in fertilized eggs of the sea urchin Strongy- locentrotus purpuratus. Dev. Biol., 96: 277-284. 7 Mazia, D., Schatten, H., Coffe, G. Sz6ke, Howard, C. and Schatten, G. (1987) Aggregation of the mitotic centrosomes into a single spherical centro- some by cold treatment in sea urchin eggs. J. Cell Biol., 105 (4, pt. 2): 206a. 8 Harris, P. J. and Rubin, B. P. (1987) Transition from mitosis to interphase in sea urchin first divi- sion: immunofluorescence studies of tubulin dis- tribution in methacrylate sections. J. Histochem. Cytochem., 35: 343-349. 9 Weber, A. (1968) The mechanism of the action of caffeine on sarcoplasmic reticulum. J. Gen. Phys- iol., 52: 760-772. 10 Kiehart, D. P. (1981) Studies on the in vivo sensitiv- ity of spindle microtubules to calcium ions and evidence for a vesicular calcium sequestering sys- tem. J. Cell Biol., 88: 604-617. 11 Wolniak,S.M., Hepler, P. K. and Jackson, W. T. (1983) Ionic changes in the mitotic apparatus during the metaphase/anaphase transition. J. Cell Biol., 96: 598-605. 12 Izant, J. G. (1983) The role of calcium ions during mitosis. Calcium participates in the anaphase trig- ger. Chromosoma (Berl.), 88: 1-10. 13, Poenie,M., Alderton,J., Tsien, R.Y. and Steinhardt, R. (1985) Changes of free calcium levels with stages of the cell division cycle. Nature (Lond.), 315: 147-149. 14 Poenie, M., Alderton, J., Steinhardt, R. and Tsien, 16 17 644 R. (1986) Calcium rises abruptly and _ briefly throughout the cell at the onset of anaphase. Sci- ence, 233: 886-889. Karsenti, E., Bravo, R. and Kirschner, M. (1987) Phosphorylation changes associated with the early cell cycle in Xenopus eggs. Dev. Biol., 119: 442- 453. Kuriyama, R. (1987) Rapid turnover of phosphate in a centrosomal component of dividing sea urchin eggs. J. Cell Biol., 105 (4, pt. 2): 284a. Endo, S. (1980) Further observations on the clusters P. J. Harris 18 19 of granular material around the centriole on the sea urchin egg: changes in distribution during mitosis. Dev. Growth Differ., 22: 509-516. Mazia, D. (1984) Centrosomes and mitotic poles. Exp. Cell Res., 153: 1-15. Schatten, H., Schatten, G., Mazia, D., Balczon, R. and Simerly, C. (1986) Behavior of centrosomes during fertilization and cell division in mouse oocytes and in sea urchin eggs. Proc. Natl. Acad. Sci. U.S.A., 83: 105-109. ZOOLOGICAL SCIENCE 5: 645-651 (1988) © 1988 Zoological Society of Japan Marked Elongation of the Anaphase Spindle by Treatments with Local Anesthetics in Sea Urchin Eggs MANABU K. KoJIMA Department of Biology, Faculty of Science, Toyama University, 3190 Gofuku, Toyama-shi/ken 930, and Sugashima Marine Biological Laboratory, Nagoya University, Sugashima-cho, Toba-shi, Mie-ken 517, Japan ABSTRACT— When fertilized sea urchin eggs are immersed in 1/32-1/16M urethane (ethyl carba- mate) sea water, cytoplasmic division is suppressed but the nuclear division continues as normal. In these eggs, aster formation is considerably disturbed but spindle can be formed and, moreover, can elongate markedly during the late anaphase-telophase. Such a spindle elongation occurs when eggs are put into a urethane solution at any time between 5 min after fertilization and the first metaphase. This elongation is inhibited if eggs, pretreated with 1/32 M urethane, are exposed to 0.1 mM colcemid, or 1/ 8 M urethane sea water in which nuclear division does not occur. Elongation of the anaphase spindle is also induced by treatments with isopropyl N-phenyl carbamate (IPC), procaine and tetracaine. However, caffeine treatments are not effective. The mechanism of spindle elongation is discussed from the standpoint of microtubule dynamics. INTRODUCTION It has been stated before that when fertilized sea urchin eggs are treated with Monogen or urethane sea water in appropriate concentrations, the cyto- plasmic division is suppressed while the nuclear division goes on as normal and thus forming multi- nucleated eggs [1]. Recently, it has become clear that Monogen suppresses the cytoplasmic division by mostly imparing the egg cortex, without affect- ing the aster formation. Urethane, however, in- hibits the division by mainly disturbing the forma- tion of mitotic apparatus [2]. Furthermore, it was also found that urethane in appropriate concentra- tions can induce marked elongation of the ana- phase spindle [2, 3]. Therefore the purpose of this paper is to report on this induced elongation of the anaphase spindle. It seems very worthwhile study- ing the induction mechanism of spindle elongation using local anesthetics, not only for elucidation of the mechanism of mitosis itself, but also for the analysis of microtubule dynamics. This paper is the first in a series of such studies. Accepted March 17, 1988 MATERIALS AND METHODS Materials Eggs of the following four species of Japanese sea urchins were used as materials: Anthocidaris crassispina, Temnopleurus toreumaticus, Pseudo- centrotus depressus and Hemicentrotus pulcherri- mus. Since all of these species essentially gave the same results, experimental data of only the last species will be presented in this paper. Chemicals Stock solutions were as follows; 1 M solution of ethyl urethane (Katayama) in distilled water, 10 mM solution of procaine hydrochloride (Sigma) in sea water, 5 mM solution of tetracaine hydrochlo- ride (Sigma) in sea water, a saturated solution of isopropyl N-phenyl carbamate (Sigma) in sea wa- ter, 50 mM solution of caffeine (Katayama) in sea water, 2X10 *M solution of colcemid (Sigma) in Abbreviations used: PIPES, 1,4 piperazine diethanesul- fonic acid; DTT, dithiothreitol; TAME, p-tosyl-L- arginine methyl ester hydrochloride. 646 M. K. Kosima sea water and 1% solution of Monogen (Dai-ichi Kogyo Seiyaku) in sea water. These stock solu- tions were diluted to various concentrations by adding sea water before use. Measurements of the spindle length Eggs were photographed using an ordinary light microscope at regular intervals after fertilization. Measurements of the spindle length were made both from projected negatives and from printed photographic papers, which were enlarged to the definite magnification, using 20 eggs at different mitotic stages. In some experiments, photographs were taken with Nomarski (Olympus) and polariz- ing (Leitz) optics so as to be able to observe the features of spindle elongation in eggs treated with the above mentioned chemicals. Isolation of mitotic apparatus Thirty seconds after fertilization, the eggs were immersed in | M urea solution and gently blown and sucked with a pipette to remove the fertiliza- tion membrane. Two minutes later, they were returned to normal sea water and washed three { Fic. 1. times by exchanging the sea water. Such denuded eggs were put into sea water containing various concentrations of local anesthetics at definite inter- vals after fertilization and allowed to develop there. At the desired stage, the eggs were transfer- red to Isolation Medium A (10mM PIPES, pH 6.9;5 mM EGTA; 0.5 mM MgSOy,; 1 M glycerol; 2 mM DTT; 1mM TAME). They were then re- transferred to Isolation Medium B (Isolation Medium A+1% Emulgen 810 or Nonidet P—40). In this medium, all eggs swelled within a few minutes. One drop of such an egg suspension was placed on a slide glass and gently pushed under a cover slip. In this way, the mitotic apparatus were easily isolated. observed and photographed with phase contrast (Olympus) or Nomarski (Olympus) optics. These mitotic apparatus were RESULTS In the first series of experiments, eggs were immersed in urethane solution of various concen- trations 5 min after fertilization and the features of Spindle elongation in urethane-treated eggs (19°C). Hemicentrotus eggs were exposed to 1/32 M urethane sea water solution at 5 min after fertilization and their mitotic figures were photographed with ordinary light microscope (A and C) and polarizing microscope (B and D). A and B: Control eggs (75 min after fertilization). C and D: 1/32 M urethane-treated eggs (95 min after fertilization). Spindle Elongation by Local Anesthetics 647 the mitotic apparatus in the first cleavage cycle were determined. At the same time, another group of eggs were exposed to Monogen sea water in graded concentrations in order to compare the effects of this reagent with that of urethane. Some differences were noted in sensitivity to both chem- icals in the different batches of eggs. However, generally speaking, when eggs are treated with 1/32-1/16 M urethane solutions which correspond to the minimum concentrations for inhibition of cytoplasmic division, the spindle can be formed although formation of the asters is considerably suppressed. In this case, the length of the ana- phase spindle, i.e. the pole-to-pole distance, be- comes much longer than that in the control eggs. It is very interesting to note that this elongation occurs even when the eggs themselves do not elongate and continue to remain spherical (Figs. 1 and 2). On the contrary, in Monogen-treated eggs, their mitotic apparatus are very similar in dimen- sion and morphology to those of the control ones but they are unable to divide (Fig. 2). In eggs treated with urethane, a fibrous structure connect- ing the two mitotic centers or nuclei can be easily detected (Figs. 1 and 2). Therefore, isolation of such an elongated anaphase spindle from urethane-treated eggs was tried and it became clear, as shown in Figure 3, that two daughter Fic. 2. Comparison of mitotic figures between Monogen- and urethane-treated eggs (19°C). Hemicentrotus eggs were transferred to 1/32 M urethane and 1/64% Monogen solutions at 5 min after fertilization and their mitotic figures in slightly depressed eggs were photographed with Nomarski optics. A and B: Control eggs (75 and 85 min after fertilization). C and D: 1/32 M urethane-treated eggs (90 and 110 min after fertilization). E and F: 1/64% Monogen-treated eggs (75 and 110 min after fertilization). In Monogen-treated eggs, furrowing sometimes occurred, but such cleavage furrow, once formed, regressed afterwards (F). 648 M. K. Kouima mee Fic. 3. Isolation of elongated mitotic spindle from urethane-treated eggs (18.5°C). Hemicentrotus eggs were deprived of fertilization membranes immediately after fertilization and, 15 min later, immersed in 1/32 M urethane solution. These denuded eggs were then transferred to the isolation medium at 90 min after fertilization and their mitotic spindles were isolated according to the procedure described in “Materials and Methods”. A: A metaphase spindle isolated from control eggs (75 min after fertilization). B-D: Elongated mitotic spindle which isolated from 1/32 M urethane-treated eggs (90 min after fertilization). Arrow head in D indicates the daughter nucleus detached from the spindle fiber. One div.=10 um. nuclei are connected by a fibrous structure, the so-called “continuous fiber”. This fact suggests to us that two nuclei may be pushed away from each other by stretching this fiber. In the second series of experiments, maximum pole-to-pole distances were measured in the con- trol, Monogen- and urethane-treated eggs. One of the results of such measurements is given in Table 1. From this table it will be seen that the pole-to- pole distances in urethane-treated eggs reach near- ly twice those in both the control and _ the Monogen-treated ones even when eggs still remain spherical before cleavage, and these interpolar distances are greater than the maximum values obtained just after the cleavage in the control ones. In addition, it should be noted that if the egg diameter is assumed as being 100, the maximum spindle length in undivided, control and Monogen- treated eggs is estimated at approximately 35. This coincides with the value, 35.47, calculated from Figure 6 in the paper by Dan [4]. In the third series of experiments, effects of colcemid on spindle elongation induced by urethane-treatments were examined. Eggs were pretreated with 1/32 M urethane solution from 5 min after fertilization. Seventy minutes later, half of them were put into 1/32-1/8 M urethane solu- tions and the rest were transferred to 5-100 uM colcemid solutions. One example of such experi- ments is represented in Table 2. As will be seen from this table, urethane-pretreated eggs are put into 1/32-1/16 M urethane solutions at 75 min after fertilization, their anaphase spindles can elongate to the same degree as those in eggs continuously exposed to 1/32 M urethane from 5 min after ferti- lization. When the pretreated eggs are transferred to 1/8M urethane in which nuclear division is disturbed, no marked spindle elongation occurs afterwards. On the other hand, it was also re- vealed when eggs are placed in 100 uM colcemid solution after exposure to 1/32 M urethane for 70 min, their spindles do not elongate further, but if a lower concentration of this drug is applied, some elongation of the spindle takes place. These facts suggest that, as concerns spindle elongation, urethane may affect some cytoplasmic changes during the late anaphase-telophase but not the changes which have occurred before the anaphase. Spindle Elongation by Local Anesthetics 649 TaBLe 1. Pole-to-pole distances in urethane- and Monogen-treated eggs (19.5°C) Pole-to-pole distances (sm) Egg diameter = — Control (ym) 1/32M urethane 1/64 % Monogen just before furrowing just after furrowing 95.7+0.4 3317 52.8422.2 64.5+3.6 33.4+1.3 (100)* (159.8)* (195.0)* (101.1)* (100)? (34.6)? ( 55.2)” ( 65.3)? ( 34.9)? Hemicentrotus eggs were put into 1/32M urethane and 1/64 % Monogen solutions at 5 min after fertilization and were then photographed at 10 min intervals for 90 min. The pole-to-pole distances were measured from photomicrographic records using 20 eggs at each of the mitotic stages. Values in this table represent maximum interpolar distances in undivided, urethane- and Monogen-treated eggs. Each value is the mean (um) + S.E. In addition, values of the spindle length just before furrowing (80 min after fertilization) and just after its completion (87 min after fertilization) in untreated eggs were given as the control. a: The numbers in parentheses express the percentage of each pole-to-pole distance when the value of the distance in eggs just before furrowing is assumed to be 100. b: The numbers in parentheses indicate the percentage of each pole-to-pole distance when the value of the egg diameter before cleavage is postulated as being 100. TaBLe 2. Effects of colcemid on spindle elongation in urethane-treated eggs (18.5°C) com 125 min after fert. isethane urethane colcemid 1/32 M 1/32M 1/16M 1/8M 5 uM 10 uM 20 ¢M 100 uM 24.9+1.1 54.4+4.3 57.1+4.0 32°64:3:3 38.7+4.5 37.9+2.2 35.8+1.1 25.9+2.6 (100) (219.0) (229.7) (131.3) (155.8) (152.3) (144.3) (104.2) Hemicentrotus eggs were exposed to 1/32 M urethane sea water at 5 min after fertilization and 70 min later, they were transferred to solutions containing urethane or colcemid in various concentrations, respectively. The length of the mitotic spindle was measured from photomicrographs taken at 75 and 125 min after fertilization (for more details, see “Materials and Methods”). Each value is the mean (“m) + S.E. The numbers in parentheses express the percentage of each pole-to-pole distance when the value of the spindle length in 1/32 M urethane-treated eggs at 75 min after fertilization is assumed to be 100. In the last series of experiments, it was deter- mined whether marked elongation of the anaphase spindle is induced not only by the treatment with ethyl urethane but also by treatments with other agents known to inhibit cytokinesis by disruption of the mitotic apparatus [5-8]. In the present experiments, effects of procaine, tetracaine and caffeine, and isopropyl N-phenyl carbamate (IPC) were tested. Eggs were exposed to sea water solutions of the above-described chemicals in var- ious concentrations at 10 min intervals from 5 min after fertilization up to the beginning of the first cleavage and the features of spindle formation were examined. Generally speaking, ethyl urethane (1/32-1/16 M) and IPC, a derivative of phenyl urethane, (40-50% dilution of the stock solution) are most effective in inducing spindle elongation and two amine compounds, procaine (10-15 mM) and tetracaine (0.2-0.5 mM), follow the former two in effectiveness. Such spindle elongation occurs when eggs are put into sea water solutions of each of these four above-mentioned agents in appropriate concentrations at any time from Smin after fertilization up to the first metaphase (75 min after fertilization). Caffeine has no effect on spindle elongation under ex- perimental conditions as used in the present study. 650 M. K. KosIMa DISCUSSION In the present study, it becomes clear that 1) when sea urchin eggs are put into 1/32-1/16M urethane sea water at any time between 5 min after fertilization and the first metaphase, aster forma- tion is considerably suppressed so that cytokinesis does not take place, but nuclear division goes on and, moreover, spindles elongate markedly during the late anaphase-telophase; 2) this spindle elonga- tion is inhibited when 1/8 M urethane or 0.1 mM colcemid, in which nuclear division does not occur, is applied to eggs pretreated with 1/32 M urethane, 3) such an elongation of the anaphase spindle can be induced by treatments with procaine, tetracaine and IPC, but not by caffeine treatment. It has been known that ethyl urethane and other anesthetics block cleavage in sea urchin eggs [9- 12]. In 1984, for instance, Rappaport [12] reported that ethyl urethane (0.06 M) reduces the size of the mitotic apparatus and blocks cleavage in sand dollar eggs. However, he did not refer to a spindle elongation, though the concentration of urethane used in his experiments was almost the same as that employed in the present study. Such a discrep- ancy between the two papers may depend upon a difference in experimental methods and materials used. Now, some questions arise: firstly, why can the spindle elongate as it does in urethane-treated eggs which have very poor asters? No definite answer can be made as yet concerning this question. However, it can be explained as follows. In normal cleavage, from metaphase to mid- anaphase, the astral rays are straight and compara- tively short. At the late anaphase, when the asters develop fully, the tips of the rays extend and reach the egg cortex. Soon afterwards elongation of the spindle takes place and the rays are no longer straight but bent, exhibiting a “fountain figure” [4, 13, 14]. This means that as the spindle elongates the polar asters are pushed against the cell mem- brane so that the spindle is compelled to elongate against the resistant force of the asters in the opposite direction. As shown already in the pres- ent paper, the size of the asters in urethane- treated eggs is much smaller than that in the control eggs. Therefore, such resistant force of the asters is assumed to be considerably weaker in urethane-treated eggs. If so, it seems reasonable that spindle can elongate much easier so that their length becomes longer than that in the control ones. In fact, a fibrous structure connecting the two asters or nuclei is actually isolated from such urethane-treated eggs and, as indicated in Table 1, maximum values of the spindle length in undi- vided, urethane-treated eggs reach nearly twice as long as those in the control eggs just before furrowing and, furthermore, these values are still greater than those in the control eggs just after cleavage. Then the next question arises: How is the motive force for such a spindle elongation produced dur- ing the late anaphase-telophase? To this question, we again cannot completely answer yet, but at least the following may be said. In the present paper, it was revealed that the spindle in eggs treated with urethane in appropriate concentra- tions can elongate to nearly twice the length of those in the control eggs, while colcemid and urethane in higher concentrations inhibit spindle elongation. These facts strongly suggest to us the possibility that, in the mechanism of spindle elongation, not only the mutual sliding of the interpolar microtubules is involved but also elongation of those microtubules themselves is very important. At any rate, investigation of this possibility must be continued. In addition, Saxton and McIntosh [15] very recently proposed a model for anaphase B in Ptk, cells in which plus end elongation of interdigitated microtubules and antiparallel sliding both contrib- ute to chromosome separation. ACKNOWLEDGMENTS This work was done at the Sugashima Marine Biologi- cal Laboratory, Nagoya University. The author wishes to express his sincerest gratitude to Prof. Hidemi Sato, Director of the Laboratory. Thanks are due to Dr. Te Ohara for her helpful assistance in the course of this work and to Prof. W. P. Hardwick, Sugiyama-jogakuen University, for reading the manuscript. REFERENCES 1 Kuno,M. (1954) Comparative studies on ex- Spindle Elongation by Local Anesthetics 651 perimental formation of multinucleated eggs of sea urchins by means of various agents. Embryologia, 2: 43-49. Kojima, M. K. (1978) On the cleavage in Monogen- or urethane-treated eggs of the sea urchins. Zool. Mag. (Tokyo), 87: 314. (in Japanese) Kojima, M. K. and Ohara, T. (1982) Effects of local anesthetics and a herbicide on the mitosis. Dev. Growth Differ., 24: 410. Dan, K. (1943) Behavior of the cell surface during cleavage. VI. J. Fac. Sci., Tokyo Imp. Univ., Sec. IV, 6: 323-368. Cheney, R. H. (1948) Caffeine effects in fertiliza- tion and development in Arbacia punctulata. Biol. Bull., 94: 16-24. Kiehart, D. P. (1981) Studies on the in vitro sensi- tivity of spindle microtubules to calcium ions and evidence for a vesicular calcium-sequestering sys- tem. J. Cell Biol., 88: 604-617. Harris, P. (1983) Caffeine-induced monaster cycling in fertilized eggs of the sea urchin Strongylocentrotus purpuratus. Dev. Biol., 96: 277-284. Jackson, W. T. (1969) Regulation of mitosis. II. Interaction of isopropyl N-phenyl carbamate and melatonin. J. Cell Sci., 5: 745-755. 9 10 11 13 14 Wilson, E. B. (1901) Experimental studies in cytol- ogy. II. Some phenomena of fertilization and cell division in etherized eggs. Arch. Entwicklungs- mech., 13: 353-373. Kobayashi, N. (1962) Cleavage of the sea urchin egg recovering from the cleavage-blocking effect of de- mecolcine. Embryologia, 7: 68-80. Rappaport, R. (1971) Reversal of chemical cleavage inhibition in echinoderm eggs. J. Exp. Zool., 176: 249-255. Rappaport, R. and Rappaport, B. N. (1984) Divi- sion of constricted and urethane-treated sand dollar eggs: A test of the polar stimulation hypothesis. J. Exp. Zool., 231: 81-92. Yatsu, N. (1909) Observations on ookinesis in Cere- bratulus lateus Verrill. J. Morphol., 20: 353-402. Balcezon, R. and Schatten, G. (1983) Microtubule- containing detergent-extracted cytoskeletons in sea urchin eggs from fertilization through cell division: Antitubulin immuno-fluorescence microscopy. Cell Motility, 3: 213-226. Saxton, W. M. and McIntosh, J. R. (1987) Inter- zonal microtubule behavior in late anaphase and telophase spindle. J. Cell Biol., 105: 875-886. ZOOLOGICAL SCIENCE 5: 653-665 (1988) © 1988 Zoological Society of Japan Control Mechanisms of Mitosis: The Role of Spindle Microtubules in the Timing of Mitotic Events GREENFIELD SLUDER Worcester Foundation for Experimental Biology, 222 Maple Avenue, Shrewsbury, Massachusetts 01545, U.S.A. Cell division in higher eukaryotes consists of a series of nuclear and cytoplasmic events that have a highly conserved sequence and a predictable timing. It is of utmost importance for the cell to exert control at both the spatial and temporal levels. The chromosomes and the cytoplasmic components of the mitotic apparatus must be precisely arranged to insure equal partitioning of the daughter chromosomes into separate, equal nuclei. In addition, the cell has to tightly control when mitotic events occur. For example, anaphase chromosome movement must not start before the chromosomes are properly oriented and aligned on the metaphase plate. The cell must not cleave until the chromosomes have adequately separated. Also, the cell should not reform nuclear envelopes and start the next cell cycle before anaphase is complete. Mistakes in either spatial arrangements or in timing lead to an abnormal division and the loss of viability for the daughter cells. In this article we review our studies which have sought to develop a better understanding of how the cell controls the timing of events during the mitosis portion of the cell cycle. Since we do not have the space to fully develop the data, the reader is directed to the original works for a rigorous demonstration of the points we will make. For our work, we used fertilized eggs from the sea urchins Lytechinus variegatus and Lytechinus pictus. Their large size, optical clarity, and the rapidity of their cell cycle make these eggs an exceptionally favorable experimental system. Im- portantly, the lessons learned from the sea urchin egg are directly applicable to other types of cells. The sequence of mitotic events for these eggs is Accepted March 17, 1988 shown in Figure 1. Figure 2 shows the average normal timing of the events we used as temporal markers. Our studies were derived from work dating back to the 1930’s which showed that the cell cycle in a variety of cell types can be stopped or prolonged in ‘metaphase’ by blocking spindle assembly with colchicine [1-4]. In the 1960’s, other studies showed that colchicine, at moderate doses, pre- vents microtubule assembly by specifically binding to the tubulin dimer [5—7]. We reasoned that if the rate at which the cell traverses mitosis can be altered by preventing microtubule assembly, spin- dle microtubules might be an important part of the mechanisms that control the mitosis portion of the cell cycle. This possibility was especially attractive given our knowledge that microtubules are re- quired for establishing the spatial arrangement of organelles at mitosis and for the execution of most mitotic events [8, 9]. As we will discuss below, spindle microtubules, in fact, do have a dual role during mitosis; they are not only necessary for the accomplishment of mitotic events, but also in- fluence when the cell will ‘decide’ to execute these events. In starting our work [10], we wanted to compare the cell cycle timing of normal fertilized eggs to eggs in which microtubule assembly had been experimentally reduced or blocked. For this we used Colcemid, a derivative of colchicine, which binds the tubulin dimer more tightly and is effec- tive at lower external concentrations than colchi- cine [11]. We had to be certain that Colcemid, as we used it, blocked microtubule assembly without toxic side-effects that could non-specifically alter a cell cycle timing. Thus, we developed a new and simple way to apply Colcemid [12]. Since con- 654 G. SLUDER Fic. Fic. 1. Upper: First and second mitoses in a L. variegatus egg; Polarization microscopy. (a) before first nuclear envelope breakdown: (b) nuclear envelope breakdown: (c) metaphase: (d) early anaphase: (e) telophase and cleavage: (f) prophase of second mitosis: (g) second nuclear envelope breakdown: (h) prometaphase of second mitosis. Minutes before and after first nuclear envelope breakdown are shown in the lower corner of each frame. 50 ym per scale division. Lower: First mitosis in a L. variegatus egg; differential interference contrast microscopy. Only the nuclear region is shown. (a) before nuclear envelope breakdown: (b) late prometaphase: (c) mid-anaphase: (d) telophase reformation of karyomeres: (e) telophase fusion of karyomeres: (f) daughter nuclei separated by cleavage furrow. Minutes before and after nuclear envelope breakdown shown in lower corner of each frame. 10 4m per scale division. NEB, Ana NER NEBs Ana NER NEB; Ana ee ee es ee 10 8 2l 7 8 19 9 Time —> 2. Normal timing of mitotic events (at 22°C) used as temporal markers. Two cell cycles are displayed on a time axis. ‘NEB,, NEB>, NEB;’ refers to nuclear envelope breakdown. ‘Ana’ is anaphase onset as detected with the polarization microscope. ‘NER’ indicates nuclear envelope reformation. The numerical values are average times, in minutes, for the intervals between the events indicated. Spindle Microtubules in Mitosis 655 tinuous immersion of eggs in colchicine can lead to continued binding of the drug beyond that neces- sary to block microtubule assembly [5, 6], we treated the eggs in early prophase for a short time (4 min) with a moderate concentration (5x 10° M) of Colcemid. By varying the duration of the pulse, we could precisely control the portion of the tubulin pool that was inactivated before mitosis [12]. In sea urchin eggs, Colcemid binds the tubulin dimer tightly; a single pulse administered in early prophase can completely prevent microtu- bule assembly for at least several cell cycles. Since the eggs are washed free of residual Colcemid, there is no continued binding of the drug beyond that necessary to block microtubule assembly. By being careful to use only the minimally effective dose of drug and conducting a series of control experiments with photo-chemically inactivated Colcemid, we showed that the changes in cell cycle timing we observed in our studies were due solely to the loss of spindle microtubules, not toxic side-effects of the drug. Percent NEB 40 46 52 First, we sought to determine if the prophase assembly of astral microtubules effects the time of first nuclear envelope breakdown, as had been previously suggested [13]. We compared the time- course of first nuclear envelope breakdown in untreated eggs and eggs treated with Colcemid to prevent microtubule assembly (Fig. 3). The fact that we found no significant difference in the time course of nuclear envelope breakdown between the control and treated populations indicated that the assembly of astral microtubules does not in- fluence the timing of interphase events leading to mitosis. We next investigated how far the cell cycle of sea urchin eggs will progress when microtubule assem- bly is prevented. Do the eggs arrest in mitosis or do they continue cycling? We always found that the cell cycle continues when microtubule assem- bly is prevented; Colcemid treated eggs show a repeated cycle of nuclear envelope breakdown and nuclear envelope reformation (Fig. 4). Also, the chromosomes repeatedly double in a normal 58 64 Minutes After Fertilization Fic. 3. Percent nuclear envelope breakdown (NEB) as a function of time after fertilization. Filled circles: untreated control eggs. Open circles: eggs treated for 3.5 min with 5x 10° ° M Colcemid. Open squares: eggs treated for 7 min with 5x 10~°M Colcemid. 100-180 eggs were scored for each data point. Inset: before and after nuclear envelope breakdown (fixed eggs). Differential interference contrast micrographs. 10 um per scale division. 656 G. SLUDER Rarer 5 inert NON eo Sean Snore: Oe a: os eigenen : Fic. 4. Disappearance-reappearance cycling of nuclear fashion as seen by the incorporation of *H Thymi- dine into DNA [14] and increases in chromosome number (Fig. 5). We noted that Colcemid treated eggs always spend more time between nuclear envelope break- down and nuclear envelope reformation than the controls. Therefore, we quantitated the timing of these events for both control and Colcemid treated cultures. Instead of working with populations of cells, we followed a number of individual eggs in vivo. This allowed us to circumvent a significant portion of the asynchrony found in any population of eggs. Much of the population asynchrony comes from variability in the period between fertilization and first nuclear envelope breakdown; thereafter, the timing of cell cycle events is reasonably con- stant from egg to egg. Thus, we normalized the time of nuclear envelope breakdown to zero for each control or treated egg we followed, because microtubule assembly does not influence when the cell undergoes first nuclear envelope breakdown (Fig. 3). We found that Colcemid treated eggs, on an average, spend twice as much time between nuclear envelope breakdown and nuclear envelope reformation than control eggs for both first and envelopes in an egg treated for 4min with 5x10~°M Colcemid. (a) before first nuclear envelope breakdown; (b) after nuclear envelope breakdown; (c) karyomeres forming; (d) karyomeres swell; (e) karyomeres synchronously break down; (f) karyomeres form for the second time; (g) these swell; (h) they break down. Minutes before and after first nuclear envelope breakdown are show in the lower corner of each frame. Differential interference contrast micrographs of the same living egg. 10 4m per scale division. second mitoses (Fig. 6). Interphase, the period from nuclear envelope reformation to the follow- ing nuclear envelope breakdown, is the same in both populations. Parenthetically, this last observation provides additional evidence that the Colcemid treatment we used does not produce detectable non-specific side effects. We next wanted to know if the Colcemid treated eggs show ‘C-anaphase’ (splitting apart of the chromatids as in anaphase onset but without chromosome movement) and if so, determine when this event occurs relative to nuclear envelope breakdown. Does it happen at the normal time for anaphase onset or is it delayed? The chromosomes become hyper-condensed during the prolonged mitosis and all split synchronously in each cell before nuclear envelope reformation (Fig. 5). Without spindle microtubules the chromatids do not move apart. By fixing aliquots of control and Colcemid treated eggs at three minute intervals, we found that this ‘C-anaphase’ splitting of the chromosomes in Colcemid treated cultures occurs 25 min or more after nuclear envelope breakdown. Anaphase onset in normal eggs occurs 10 min after nuclear envelope breakdown. Thus, in Colcemid Spindle Microtubules in Mitosis 657 Do SOM Se rd male ie: ec) a ee Fic. 5. (a-j) Chromosome morphology in Colcemid treated eggs which do not assemble any spindle microtubules. (a) Before nuclear envelope breakdown. (b-f) Progressive condensation of chromosomes at increasing times after nuclear envelope breakdown. (g) ‘C-anaphase’ splitting of the chromosomes. (h) Telophase decondensa- tion of chromatin. (i) Karyomeres. (j) Increased number of chromosomes at second nuclear envelope breakdown. (k-m) Metaphase, anaphase onset, and late anaphase in untreated eggs. Eggs were fixed in acid-alcohol and stained with orcein. Phase contrast micrographs. 10 «m per scale division. a NER, NEB» NER> Cont. Ps VISOR Aue laminven ne a7 % 1 (21) N_ (23) SS (23) =< | ne ‘ a. | SS aX SS | << Se a Colc:! ~. eet ere 1 37.8min SL 22.3min \ 27.5min == 1 (26) \ (23) (19) ie + + + ' ' ' ' + + ie) 10 20 30 40 50 60 70 80 Minutes After First NEB Fic. 6. Timing of nuclear envelope breakdown (NEB) and nuclear envelope reformation (NER) in untreated (upper line) and treated eggs (lower line). Eggs were treated for 4 min with 5x 10~° M Colcemid ~20 min before first nuclear envelope breakdown. The horizontal lines represent time axes. First nuclear envelope breakdown is normalized to 0 min for all individual eggs. The mean times of nuclear envelope reformation and breakdown are shown by filled circles on the time axes. The heavy horizontal bars delimit the 95% confidence limits of the means. The larger numbers under each time axis show the mean duration of the various intervals. The small numbers in parentheses give the sample sizes. treated eggs, the period between nuclear envelope tion follows the ‘C-anaphase’ splitting of the breakdown and anaphase onset is prolonged. In chromosomes at approximately the normal inter- Colcemid treated eggs, nuclear envelope reforma- val of 8 min. Thus, microtubule assembly in- 658 G. SLUDER fluences only the time from nuclear envelope breakdown to anaphase onset: the prometaphase/ metaphase portion of mitosis. Recently, the general applicability of these re- sults to other species of echinoderms was brought into question by a study using high concentrations of colchicine on eggs of Strongylocentrotus purpur- atus and Dendraster excentricus [15]. This study showed that 1-2mM colchicine, applied con- tinuously, blocked spindle assembly and altered cell cycle timing in qualitatively different fashions that those we had earlier reported for Lytechinus. Eggs of S. purpuratus were completely arrested in mitosis. For D. excentricus mitosis was of normal duration while interphase was prolonged. We were concerned that the relatively high dosages of colchicine used had toxic side effects (reviewed in [14]) which could cause these surprising patterns of timing. Therefore, we reinvestigated the role of spindle microtubules in the timing of the cell cycle of these two species of echinoderms using short treatments of Colcemid to block microtubule assembly [14]. The timing patterns we observed were entirely consistent with those we obtained from our studies on Lytechinus. In addition, we performed a series of control experiments using lumi-colchicine, a photo-isomer of colchicine which does not bind to tubulin [16] but shows the same toxic side effects as the native compound [17, 18]. One mM lumi-colchicine alone partially in- hibits microtubule assembly and prolongs both mitosis and interphase. Thus, we feel that the timing patterns observed in the colchicine study are most easily explained by a combination of the specific and non-specific effects of continuously applied 1-2 mM native colchicine. To further explore the reiationship between spindle microtubule assembly and the timing of mitotic events, we delayed the onset of microtu- bule assembly for increasing periods of time after nuclear envelope breakdown and determined the extent to which this perturbation influenced the times of anaphase onset and entry into the next cell cycle. We blocked microtubule assembly and then followed individual eggs as they entered first mito- sis. At various times after nuclear envelope break- down, we irradiated individual eggs on the micro- scope with 366 nm light to photochemically inacti- vate the Colcemid thereby allowing the eggs to assemble spindle microtubules. We wanted to know if anaphase onset and telophase events occurred at the normal times after nuclear en- velope breakdown or alternatively, if the time of irradiation determined when the eggs would initi- ate anaphase and finish mitosis. This would tell us whether or not these events are controlled by a microtubule independent clock mechanism. Con- trol irradiations of untreated eggs showed that the short (15 sec) irradiations used in this experiment do not alter the timing of mitotic events. When Colcemid treated eggs are irradiated at nuclear envelope breakdown, they assemble a functional spindle and divide in a normal fashion. The time from nuclear envelope breakdown to anaphase onset is the normal 10 min. Telophase events and second mitosis are also normal. We then delayed the irradiation for increasing amounts of time after nuclear envelope break- down. The eggs always assemble functional spin- dles over the normal time course and initiate anaphase, on average, 10 min after the irradiation. This is even true for eggs irradiated as long as 14 min after nuclear envelope breakdown, a time when the egg would normally have been in late anaphase (Fig. 7). These results show that the time of anaphase onset is not solely controlled by a microtubule independent clock mechanism but rather, is determined by the assembly of spindle microtubules. After anaphase onset, the sequence of telophase events proceeds with normal timing regardless of the duration of prometaphase. Second nuclear envelope breakdown follows first anaphase onset by a constant interval which is the same as that in the irradiated control eggs. Similar experiments on second division eggs showed that our results were not peculiar to the first division cycle. These results also show that Colcemid (or col- chicine) does not arrest cells at ‘metaphase’ as is often said. From the standpoint of the cell cycle’s progress through mitosis, Colcemid arrests sea urchin eggs at the start of prometaphase for a significant amount of time. Had the eggs truly been arrested at metaphase, the time from irradia- tion to anaphase onset would not have been the 10 minutes we consistently observed; the irradiation Spindle Microtubules in Mitosis 659 MINUTES BETWEEN NEB, and IRRAD. NEB, Ana NEB, Control (n=32) / ee \ \ \ \ \ a 1-4.5 (n=12) = 7 * . | \ \ \ \ \ s 6-9.5 (n=10) 1 \ \ ENN i fe ‘ =e = a | ‘\ ‘ \ 10-14.0 (n=23) \ : LE de tad {fo é a oe} Cole \10min \ 36min 5 Irrad : \ —_—_;— { H+-———+ +} _ | = -20 0 10 20 30 40 50 60 Minutes After First NEB Fic. 7. Results of Colcemid reversal experiments. Fertilized eggs were treated for 3.5 min with 5 x 10° M Colcemid in early prophase of the first division (cross hatching). They were irradiated for 15 sec with 366 nm light at times ranging from 0.5 to 14 min after first nuclear envelope breakdown (NEB). The untreated control eggs (top line) were irradiated shortly after first nuclear envelope breakdown. Timing data from the Colcemid-treated eggs are collected into three classes based upon the number of minutes between nuclear envelope breakdown and irradiation. The light horizontal lines are the time axes. The time of first nuclear envelope breakdown is normalized to 0 min for all individual eggs. The mean times of irradiation, anaphase onset, and second nuclear envelope breakdown are shown as filled circles on the time axes. The heavy horizontal bars delimit the 95% confidence limits of the means. The parallel dotted lines are drawn through the irradiation, anaphase, and second nuclear envelope breakdown means to emphasize the constancy of the interval between irradiation and anaphase, as well as the interval between anaphase and second nuclear envelope breakdown. The numbers in parentheses give the sample sizes. to anaphase onset interval would have depended on how long we delayed the irradiation. When we irradiated Colcemid treated eggs more than 14 min after nuclear envelope breakdown, we observed incomplete and abortive spindle assem- bly which indicated that the eggs had spontaneous- ly begun to finish mitosis and were entering the next cell cycle. This was understandable given our demonstration that the cell cycle does eventually continue when no spindle microtubules are assem- bled. We next wanted to determine if the extent of spindle microtubule assembly influences the dura- tion of prometaphase. Do eggs with small spindles traverse mitosis at the same rate as eggs with normal sized spindles? We allowed eggs to divide once and treated them late in first telophase with a short (1-3 min) pulse of Colcemid. At second nuclear envelope breakdown, the blastomeres assembled barrel-shaped spindles of reduced length and birefringence. Just after second nuclear envelope breakdown we irradiated one daughter blastomere with 366nm light which led to the immediate growth of that cell’s spindle to a normal size. This provided us with a control cell against which to compare the timing of the blastomere with a diminished spindle. The cell with the normal sized spindle always enters anaphase and finishes mitosis earlier than the cell with the short spindle (Fig. 8). Here, prometaphase in the blasto- mere with the diminished spindle is approximately 50% longer than prometaphase in the irradiated control cell. By varying the duration of the Col- cemid treatment we could vary the size of the spindles assembled at second mitosis. We found that the smaller the spindle, the later the cell initiates anaphase. After anaphase onset, telo- phase events proceed with normal kinetics regard- less of the size of the spindle. Collectively, these experiments show that spin- dle microtubules are necessary not only for the accomplishment of mitotic events, but also play an important role in the mechanisms that determine when the cell will ‘decide’ to initiate anaphase and 660 G. SLUDER K i F Nar 268 Fic. 8. Mitosis of daughter cells with different-sized spindles. This zygote was treated for 2.5 min with 5x10 °M Colcemid at first cleavage. Second nuclear envelope breakdwon occurred synchronously in both daughter cells. Shortly thereafter, the upper blastomere was irradiated for 15 sec with 366 nm light. The lower blastomere was irradiated for 15 sec when it was in telophase to equalize the doses of 366 nm light. (a) prometaphase in irradiated and unirradiated blastomeres; (b) anaphase in the irradiated blastomere while lower blastomere is in metaphase; (c) telophase and cleavage in irradiated cell while the unirradiated cell is in anaphase. The large spindle initiated anaphase 3 min earlier than the smaller one; (d) telophase and cleavage in lower blastomere; (e) nuclear envelope fully reformed in upper two cells but not in lower two; (f) prometaphase in upper cells while lower two are still in interphase; (g) anaphase in upper cells and prometaphase in lower cells; (h) cleavage in upper cells and anaphase in lower cells. Minutes after second nuclear envelope breakdown are shown in the lower corner of each frame. Polarization micrographs; 50 4m per scale division. finish mitosis. The cell appears to have a fun- damental rhythm that allows more time for mitosis than is actually needed under normal circum- stances. This fundamental rhythm is revealed when microtubule assembly is completely blocked. Starting with nuclear envelope breakdown, there is a waiting period that provides the cell with wide temporal tolerances to assemble the labile spindle structure in preparation for division. Within this period, spindle microtubules are part of the mechanism that leads to a necessary physiological change that triggers the cell to initiate anaphase, finish mitosis, and start the next cell cycle. Once triggered, the events that finish one mitosis and lead to the next mitosis, proceed at a normal pace independent of microtubule assembly. At the next nuclear envelope breakdown, the timing of the cell cycle again becomes sensitive to the assembly of spindle microtubules. Thus, the cell cycle can be thought of as a stopwatch. Once the hand comes into the mitosis portion of the cycle, the watch can be reset to the start of the next cycle by mecha- nisms involving spindle microtubules. If these microtubules are not assembled, the watch will continue to cycle but does so at a fundamental rate. The observation that once the cell cycle is prolonged, subsequent cycle(s) are not shorter than normal, rules out the possibility that the cell cycle timing is governed solely by a continuous oscillator such as the ones proposed for Physarum and Xenopus eggs [19, 20]. Furthermore, our studies indicate the existence Spindle Microtubules in Mitosis 661 of a physiological change that is an important transition point in the cell cycle. This change commits the cell to initiate anaphase, finish mito- sis, and start the next cell cycle. In a normal cell, the splitting of the kinetochores at anaphase onset is the first visible manifestation that this change has occurred. However, from the standpoint of cell cycle timing, it does not matter if the chromosomes actually move apart or not. Once this transition point has passed, cleavage, reformation of nuclei, centrosome duplication, and the start of the next cell cycle follow as a temporal linkage group whose timing is insensitive to the extent of microtubule assembly. Such a transition point was first sug- gested by Mazia and termed ‘a point of no return’ [21]. Although we cannot yet identify this event in molecular or ionic terms, we can detect it by a change in the functional properties of the cell cycle. Before this point, the progress of the cell cycle through mitosis is influenced by microtubule assembly. After this point, the cell cycle proceeds at a normal rate, irrespective of the state of its microtubules. This is not surprising since late anaphase and telophase are the times when the cell normally disassembles spindle microtubules anyway. Although we do not think that the anaphase splitting of the chromosomes per se is the causal event that commits the cell to finish mitosis, we must consider the possibility that the time of anaphase onset is determined by the assembly of a threshold quantity of microtubules sufficient to physically pull the chromosomes apart. This is not likely given observations which indicate that force production by the spindle does not trigger ana- phase onset. First, chromatids will synchronously pop apart without microtubules pulling on them [3, 10, 22, 23]. Second, unattached chromosomes or acentric fragments that lie outside the spindle split and separate slightly when the chromosomes in the spindle start their normal anaphase movements [24, 25]. Thus, a change in cytoplasmic conditions, not force production by the spindle, determines when chromosomes can move apart in anaphase. Third, relatively few microtubules are required to produce or transmit the force necessary for chromosome movement [26]. Our work to this point left us with a seemingly simple question: How could spindle microtubules participate in the mechanisms that control the duration of the prometaphase/metaphase portion of the cell cycle? Possibly, the cell monitors what fraction of the tubulin pool is used in spindle assembly. Alternatively, the effect of microtu- bules on cell cycle timing may be indirect, as for example, through the activity of some microtu- bule-associated enzymatic process that is active only on the assembled microtubules. A _ third possibility is that the organization of microtubules in a mitotic apparatus determines the distribution of organelles or cytoplasmic factors, which in turn influence the timing mechanisms. In other words, microtubules may act as a structural scaffolding that has the proper geometry for some important process to proceed. To test between these alternatives, we investi- gated the relationship between the spatial arrange- ment of spindle microtubules and the duration of the mitosis [27]. To alter the organization of the spindle without significantly reducing the total tubulin polymer in the egg, we used a microneedle to cut spindle at the metaphase plate and separate the two half spindles. Would a cell with a cut spindle traverse mitosis at the same rate as a cell with an intact spindle? We performed the opera- tion on second division eggs so that we could use one daughter blastomere as a control cell. At the onset, we needed to know if the ma- nipulation produced any non-specific, irreversible damage that would slow the cell cycle. Obviously, unhealthy eggs could not be expected to show normal timing. Therefore, we conducted a series of control experiments in which we stirred the cytoplasm with a microneedle and even moved the spindle without cutting it. In all cases, the manipu- lated blastomeres divide in complete synchrony with their controls. In addition, we cut spindles and then pushed the two half-spindles back together again. The half-spindles always reassoci- ate to form a functional spindle of normal appear- ance. Even though mitosis is slightly prolonged in these cases, subsequent mitoses are normal. Thus, we were assured that the manipulations did not reduce the viability of the eggs. Blastomeres with cut spindles always spend sig- nificantly more time in mitosis than their same 662 G. SLUDER embryo controls (Fig. 9). Here, the experimental cell is just finishing second mitosis when the con- trol is well into third mitosis. Since we could not observe anaphase chromosome splitting in vivo, we used the rapid telophase fading of astral bire- fringence as a measure of when the cell finishes mitosis. We found that blastomeres with cut spindles take, on average, 48 min to proceed from nuclear envelope breakdown to telophase as opposed to 15 min for the control cells. Once the manipulated cell enters telophase, the time to the start of the next mitosis is approximately normal. Thus, rearranging the spindle prolongs only the prometaphase/metaphase portion of the cell cycle. Importantly, the cell cycle of the manipulated cell never resynchronizes with that of the control blas- tomere. These results show that the timing mechanisms for mitosis do not monitor the utiliza- tion of the tubulin pool and are not dependent on some enzymatic process that is based strictly on the total amount of tubulin polymer in the cell. As a second way to produce spindles of altered geometry, we indirectly induced the formation of monopolar spindles using methods described else- where [28, 29]. The timing of blastomeres with monopolar spindles was compared to same- embryo blastomeres in which a cleavage furrow had failed thereby allowing two monopoles to AU S34 P i Fic. 9. come together to form a functional bipolar spindle of normal appearance. By using such cells as controls we were certain that all cells had been exposed to the same experimental regime. Cells with monopolar spindles always spend significantly more time in mitosis (49 min on average) than the same-embryo cells with bipolar spindles (15 min on average). These times are remarkably similar to those observed in the micromanipulation ex- periments. Presently, we cannot completely ex- plain why eggs with spindles of altered geometry spend more time in mitosis than eggs that assemble no spindle microtubules (Fig. 6). In part, the answer lies in the fact that the eggs used in the cut spindle and monopolar spindle experiments (L. pictus) were run at a slightly lower temperature than those used in the Colcemid experiments (L. variegatus). However, other factors may be in- volved, and this issue deserves further study. At this point, we were aware that our microma- nipulation and monopolar spindle results might be explained by the fact that half-spindles or monopo- lar spindles have less total tubulin polymer than normal spindles. Cells with diminished spindles always spend more time in mitosis than cells with normal sized spindles (Fig. 8). For the micromanip- ulation experiments, the total number of microtu- bules is approximately normal on a total cell basis, Development of a cell with a cut spindle; (a) before manipulation, both daughter cells are in early prometaphase of second mitosis; (b) spindle in lower blastomere cut with microneedle; (c—e) manipulated cell stays in mitosis while control completes mitosis; (f-h) manipulated cell enters telophase and cleaves while control enters third mitosis; (i) manipulated cell reforms nuclei (arrows) while control cleaves; (j) manipulated cell enters third mitosis while control cells are preparing for fourth division. Minutes after second nuclear envelope breakdown are given in the lower corner of each frame. Polarization micrographs; 10 ~m per scale division. Spindle Microtubules in Mitosis 663 but the half-spindles are sufficiently separated that they might reside in physiologically distinct re- gions. Conceivably the timing of mitotic events is more sensitive to the quantity of microtubules in a given volume of cytoplasm than the total amount of tubulin polymer in the cell. To explore these issues, we compared the timing of second division eggs containing Colcemid-diminished bipolar spin- dles to those containing monopolar spindles or cut spindles. The following observations indicate that monopolar spindles should have at least as much total polymerized tubulin as diminished bipolar spindles. We compared typical monopolar, nor- mal bipolar, and Colcemid diminished bipolar spindles (Fig. 10). The half-spindle length and birefringence are approximately the same in monopolar and normal bipolar spindles. Also, the extent of astral development is the same. Dimi- nished spindles, on the other hand, are less than half as long as a normal spindle, have very small asters, and have half the normal half-spindle bire- fringence. Fic. 10. Size comparison for (a) monopolar spindle, (b) normal bipolar spindle, (c) Colcemid-diminished spindle. All photographs are printed at the same magnification. The birefringence of the region between the chromosomes and the pole is the same in the monopolar and bipolar spindles. The meas- ured birefringence of the diminished spindle is about half that of the other two, but is not evident in this photographic reproduction. Polarization micro- graphs; 10 «m per scale division. We found that cells with monopolar or cut bipolar spindles spend, on average, more than twice as much time between nuclear envelope breakdown and telophase as cells with diminished bipolar spindles (48 vs 22 min). A bipolar spindle, even though much smaller than normal, enables a cell to traverse mitosis faster than a spindle of altered geometry. These results clearly show that the timing mechanisms are more sensitive to the spatial arrangement of spindle microtubules than the total quantity of microtubules assembled. This observation is important because it rules out sim- ple models for the involvement of spindle microtu- bules in the timing mechanisms. For example, if astral microtubules are helping to move substances and organelles, such as Ca** sequestering vesi- cles, within the cell, a monopolar spindle with its significantly longer, more plentiful astral microtu- bules should be more effective than the two almost imperceptable asters of the diminished bipolar spindle. How then could spindle microtubules participate in the mechanisms that control the timing of mitotic events? Although we do not have an answer to this question, we have noted correla- tions that suggest directions for future research. We noted that the spindle perturbations which influence when the cell will ‘decide’ to initiate anaphase and finish mitosis, also influence the extent to which an egg forms a cleavage furrow: 1. When spindle microtubule assembly is pre- vented, mitosis is significantly prolonged and a cleavage furrow is not formed. 2. When micro- tubule assembly is diminished, anaphase onset is delayed and only weak cleavage furrows, if any, form. There is a correlation between the size of the spindle, the delay in anaphase onset, and the strength of the cleavage furrows formed. 3. Cells with monopolar spindles traverse mitosis slowly and usually do not show any surface activity that could be regarded as a cleavage equivalent. 4. If a spindle is cut and the half spindles moved apart, mitosis is significantly prolonged and only weak furrows, if any, are formed. In addition, studies by Rappaport have shown that astral spacing in- fluences furrow formation. If asters are too far apart, furrows do not form [30]. These correlations, in concert with studies show- ing that the cleavage furrow is triggered (but not yet visible) at the time of anaphase onset [30], suggest that we should look to aster-cortex interac- tions for more insight into how spindle microtu- bules participate in the timing of the mitosis por- tion of the cell cycle. Perhaps an aster-cortex 664 G. SLUDER interaction leads to a local physiological change which is the key transition point that commits the cell to initiate anaphase and to finish mitosis. This aster-cortex interaction is also important in or- ganizing the cleavage apparatus, as Rappaport has shown [30]. In saying this, we do not want to imply that the assembly of the contractile ring of actin filaments or the physical constriction of the cell plays a causal role in committing a cell to finish mitosis and start the next cell cycle. The first visible manifestation of the furrow occurs well after the cell has passed the transition point that commits it to finish mitosis. We simply want to draw attention to aster-cortex interactions as a possibly key factor in the triggering of the transi- tion point as well as the formation of the cleavage furrow. In conclusion, our work has shown that the assembly of spindle microtubules in a spatially correct fashion is an important facet of the mechanisms that control the timing of mitotic events. Of what relevance could this be to the dividing cell? Perhaps an answer to this question is that an important aspect of mitosis is the establish- ment and maintenance of a number of specific spatial relationships within the dividing cell. Be- fore the cell commits itself to anaphase, the poles must separate to establish a division axis; the chromosomes must be aligned on the metaphase plate; and daughter chromatids must be oriented to opposite poles. Failure to maintain these spatial relationships results in abnormal division and the eventual loss of viability for the daughter cells. To minimize the chances of mistakes, the control mechanism for the timing of mitotic events has two important features. First, it provides more than enough time for the assembly of the labile spindle structure. Second, both the quantity and the spatial arrangement of spindle microtubules help determine when the cell will execute the critical transition point that commits it to finish mitosis and start the next cell cycle. In this way the cell can insure that it is ready to divide correctly before it actually commits itself to completing the process. ACKNOWLEDGMENTS These studies were conducted in the laboratories of Drs. Shinya Inoue, Hidemi Sato, and Daniel Mazia. We thank them for their help and the use of their facilities. We also thank Dr. George Witman for a critical reading of this manuscript. Ms. Sandy Johnson and Ms. Carol Savage were of invaluable assistance in the typing of this manuscript. We also express our appreciation of Mr. Rick Miller’s important efforts in reprinting the figures. Lastly, we thank the Rockefeller University Press for copyright permission to reproduce figures from The Journal of Cell Biology, 80: 674-691, 1979 and 97: 877- 886, 1983. REFERENCES 1 Biesele, J. J. (1958) Mitotic Poisons and the Cancer Problem. Elsevier Scientific Publ. Co., Amsterdam. 2 Deysson, G. (1968) Antimitotic substances. Int. Rev. Cytol., 24: 99-148. 3 Eigsti, O. J. and Dustin, P. (1955) Colchicine- in Agriculture, Medicine, Biology and Chemistry. Iowa State College Press, Ames, Iowa. 4 Kihlman, B. A. (1966) Actions of Chemicals on Dividing Cells. Prentice-Hall, Inc., Englewood Cliffs, N. J. 5 Taylor, E. W. (1965) The mechanism of colchicine inhibition of mitosis. I. Kinetics of inhibition and the binding of *H-colchicine. J. Cell Biol., 25: 145-160. 6 Borisy,G.G. and Taylor, E. (1967) The mecha- nism of action of colchicine; binding of colchicine- 3H to cellular protein. J. Cell Biol., 34: 525-533. 7 Borisy, G. F., Olmsted, J. B., Marcum, J.M. and Allen, C. (1974) Microtubule assembly in vitro. Fed. Proc., 33: 167-174. 8 Nicklas, R. B. (1971) Mitosis. In “Advances in Cell Biology, Vol 2”. Ed. by M. Prescott, L. Goldstein and E. McConkey, Appleton-Century-Crofts, New York, pp. 225-297. 9 Tilney, L. G. (1975) Origin and continuity of micro- tubules. In “Origin and Continuity of Cell Organelles”. Ed. by J. Rinert and H. Ursprung, Springer-Verlag, Berlin, pp. 222-260. 10 Sluder, G. (1979) Role of spindle microtubules in the control of cell cycle timing. J. Cell Biol., 80: 674-691. 11 Ray,K., Bhattacharyya,B. and _ Biswas, B. B. (1980) Colcemid and colchicine binding to tubulin: Similarity and dissimilarity. Eur. J. Cell Biol., 2: 288. 12 Sluder, G. (1976) Experimental manipulation of the amount of tubulin available for assembly into the spindle of dividing sea urchin eggs. J. Cell Biol., 70: 75-85. 13. Bajer, A. and Mole-Bajer, J. (1969) Formation of spindle fibers, kinetochore orientation, and be- havior of the nuclear envelope during mitosis in 14 16 17 18 19 20 21 Spindle Microtubules in Mitosis endosperm, fine structural and in vitro studies. Chromosoma, Berl., 27: 448-484. Sluder, G., Miller, F. J. and Spanjian, K. (1986) The role of spindle microtubules in the timing of the cell cycle in echinoderm eggs. J. Exp. Zool., 238: 325-336. Yoneda, M. and Schroeder, T. E. (1984) Cell cycle timing in colchicine-treated sea urchin eggs: Persist- ent coordination between the nuclear cycles and the rhythm of cortical stiffness. J. Exp. Zool., 231: 367- 378. Borisy, G. G. (1971) Mechanism of U. V. reversal of colchicine inhibition of microtubule aggregation in vitro. Proc. 11th Ann. Meeting of Am. Soc. for Cell Biol., 27: 35a. Mizel,S.B. and Wilson, L. (1972) Nucleoside transport in mammalian cells. Inhibition by colchi- cine. Biochem., 11: 2573-2578. Stadler, J. and Franke, W. W. (1974) Characteriza- tion of the colchicine binding of membrane fractions from rat and mouse liver. J. Cell Biol., 60: 297-303. Tyson, J. and Kauffman, S. (1975) Control of mito- sis by a continuous biochemical oscillation: synchro- nization; spatially inhomogeneous oscillations. J. Math. Biol., 1: 289-310. Hara, K., Tydeman, P. and Kirschner, M. (1980) A cytoplasmic clock with the same period as the divi- sion cycle in Xenopus-laevis eggs. Proc. Natl. Acad. Sci. USA., 77: 462-466. Mazia, D. (1961) Mitosis and the physiology of cell division. In “The Cell, Biochemistry, Physiology, 665 Morphology, Vol. 3”. Ed. by J. Brachet and A. Mirsky, Academic Press, Inc., NY, pp. 77-412. Levan, A. (1954) Colchicine-induced c-mitosis in two mouse ascites tumors. Hereditas, 40: 1-64. Mole-Bajer, J. ( 1958) Cine-micrographic analysis of c-mitosis in endosperm. Chromosoma, Berl., 9: 332-358. Zirkle, R. E. (1956) Cellular Aspects of Basic Mechanisms in Radiobiology. Nuclear Science Series Report, No. 8. Research Council, Washing- ton, DC., Publication No. 450, p. 1. Bajer, A. (1958) Cine-micrographic studies on chromosome movements in /-irradiated cells. Chro- mosoma, Berl., 9: 319-331. Nicklas, R. B. (1985) Mitosis in Eukaryotic Cells: An Overview of Chromosome Distribution. Aneu- ploidy. Ed. by V.L. Dellarco, P.E. Voytek and A. Hollaender, Plenum Publishing Corp. Sluder, G. and Begg, D. (1983) Control mecha- nisms of the cell cycle: Role of the spatial arrange- ment of spindle components in the timing of mitotic events. J. Cell Biol., 97: 877-886. Mazia, D., Harris, P. J. and Bibring, T. (1960) The multiplicity of mitotic centers and the time-course of their duplication and separation. J. Biophys. Biochem. Cytol., 7: 1-10. Sluder, G. and Begg, D. A. (1985) Experimental analysis of the reproduction of spindle poles. J. Cell Sci., 76: 35-S1. Rappaport, R. (1971) Cytokinesis in animal cells. Int. Rev. Cytol., 31: 169-213. ZOOLOGICAL SCIENCE 5: 667-675 (1988) © 1988 Zoological Society of Japan The Mechanism of Ooplasmic Segregation in the Ascidian Egg TomMoo SAWADA Department of Anatomy, Yamaguchi University School of Medicine, Ube City, Yamaguchi 755, Japan INTRODUCTION Ooplasmic segregation, the accumulation, re- placement or separation of certain components in egg cytoplasm is seen in many kinds of animals [1]. The mechanism of ooplasmic segregation, as gathered from investigations performed on several types of animals, seems to be associated with the contractile activity of the egg cortex and micro- tubular structures such as the sperm aster. Cortical contraction probably depends on the actin fila- ments at the egg cortex, and microtubules also play an important role in cell division and pro-nuclear movement at the same time [2, 3]. In ascidian eggs, the actin filaments and microtubules actually provide a motile system for ooplasmic segregation at the same time they exhibit successive changes in pursuit of meiosis or mitosis. Thus, the mechanism of ooplasmic segregation can not be separated from the mechanism of the cell division, especially in very early stage of development, i.e., from fertilization through the first cleavage when several kinds of movements are occurring together in a short time. OOPLASMIC SEGREGATION IN ASCIDIANS Ooplasmic segregation in ascidian early de- velopment was observed in detail by Conklin in 1905 [4]. His report laid the foundation for all subsequent investigations and is still the basis of work performed today. He was the first to distin- guish three different cytoplasms in the egg of ascid- ian Styela partita according to pigmentation and was able to trace their movement and segregation. Accepted May 11, 1988 He designated three cytoplasms (endo-, meso- and ectoplasm) and then five cytoplasms (ecto-, endo-, myo-, chymo- and chorda-neuroplasm) according to the fates of the blastomeres which included each cytoplasm [4, 5]. Conklin and subsequent workers theorized that some component of those cytoplasms may control the differentiation of the cells which are each provided with different cytoplasms by cleavage. Ascidian early development is thus basically “mosaic” [5-8], i.e., an embryo can not form a certain organ if it does not have the corre- sponding cytoplasm or the certain component in the corresponding cytoplasm. Recent biochemical and molecular studies have established that certain differentiation, namely the determinants, are segregated into the specific cell lines in ascidian early development [9- 14]. Myoplasm, the cytoplasm which is segregated into the cell line to form embryonic muscle cells [4], is the most well-investigated among the cyto- plasms defined by Conklin. Myoplasm includes numerous mitochondria [7, 8, 15, 16] in every species as well as specifically pigmented granules components essential for cell or other characteristic granules in some species [4, 7, 16]. According to recent studies [10, 11], myoplasm probably includes the determinant for muscular development. Myoplatm is distributed just beneath the egg cortex in unfertilized eggs as a layer which is thick near the vegetal pole and absent around the animal pole (Figs. 2A and 4A). The inner part of the egg is occupied by many yolk granules [4, 7, 16]. Within 2-3 min after fertiliza- tion, the myoplasm moves down and gathers near the vegetal pole (Figs. 2B and 4B, C, D). At the same time, inner yolk granules automatically fill 668 T. SAWADA part of animal side of the myoplasm [4, 16, 17]. This movement is the first phase of ooplasmic segregation [7]. Myoplasm moves up again to the equatorial region after the second meiosis is com- pleted (Fig. 4G, H, I). This is the second phase [7] of ooplasmic segregation and the myoplasmic cres- cent, the so-called yellow crescent according to Conklin’s study in Styela partita, is formed by this movement. The side on which the myoplasmic crescent is formed corresponds to the posterior side of the embryo [4, 7, 8]. The movement of myoplasm described above is very clear and easy to trace even in species whose eggs do not have a special pigmentation in the myoplasm, by staining for mitochondria using Janus green [18], rhodamine 123 [19] or ethydium bromide [20]. Therefore, the movement of myo- plasm has been observed mainly in ooplasmic segregation and the mechanism of its movement has been investigated well. STUDIES ON THE MECHANISM OF OOPLASMIC SEGREGATION There have been only few investigations regard- ing the mechanism of ascidian ooplasmic segrega- tion in contrast to the many studies on the role of ooplasmic segregation in development. Costello [21] tried to explain ooplasmic movement of asci- dian eggs by the ‘diffusion effect’ [22], although he was not successful. As already mentioned, until first mitosis, ascid- ian ooplasmic segregation can be divided into two phases. They differ in the manner of cytoplasmic movement and in the motile mechanism for it. The first phase movement of ooplasmic segregation takes place within a short time (Fig. 4A, B, C, D). It starts within 30 sec after fertilization and occurs in 2-3 min in Ciona savignyi* [16]. It is known that the egg cortex contracts and the test cells and particles on the egg surface are carried to the vegetal pole at the same time the myoplasm moves down to the vegetal pole [16, 24]. Recent evidence strongly supports the speculation that this contrac- tion provides the motive force of the first phase * Previously known as Ciona intestinalis in Japan; re-defined as Ciona savignyi [23]. movement. The Ciona savigni [16] egg was used to make detailed observations of the cortical contrac- tion because shape modification upon cortical con- traction is dramatic and an exceptionally clear constriction, not apparent in other species, in the course of this contraction (Fig. 1). The results suggested that the constriction is formed by the contraction of the cortex as a whole of the vegetal side, as clearly indicated by the accumulation of microvilli on the vegetal side surface which are equally distributed before fertilization. The dis- appearance of microvilli on the animal side indi- cates that the animal surface is expanded. Upon further contraction of the vegetal side cortex, the constriction shifts (Fig. 1) toward the vegetal pole [16]. It is as if the original cortex of the unfertilized egg, together with the subcortical myoplasm underlying the cortex, peels off and accumulates at the vegetal pole. This contraction is independent of the inner cytoplasm and is probably caused by the contractile structure specified at the cortex [25, 26]. This contractile structure is believed to be actin filaments because cytochalasin B [16, 27, 28] ~~” C FE Fic. 1. Modification of egg shape in the first phase of ooplasmic segregation of Ciona savignyi. A. Unfer- tilized egg, the bar indicates 50 ~m. B-E. Succes- sive shape modification of fertilized eggs from 1 to 3 min after insemination. The ‘cap’ is formed at the vegetal pole in E. F. After the formation of the first polar body (pb). ap, animal pole. vp, vegetal pole. (From Sawada and Osanai 1981 with permission). Mechanism of Ascidian Ooplasmic Segregation 669 inhibits contraction and cytoplasmic movement. The existence of actin filaments in the contracting cortex was revealed [29, 30] and its distribution corresponds to the manner of cortical contraction [30]. This cortical contraction seems to be trig- gered by an increase in cytoplasmic Ca‘ *-ion. For calicium ionophore A23187 or direct injection of Cat *-ion [31] is able to induce the contraction. An increase in Ca* *-ion during normal fertiliza- tion has also been detected [32]. The control mechanism which induces polarized contraction of such a wide area of egg cortex is not clear. More on this matter is discussed later. In spite of the progress made in the investiga- tions of the first phase movement, there have been only a few studies on the second phase movement. The second phase has sometimes lacked in inves- tigations and arguments about the mechanism of ascidian ooplasmic segregation, it is an important step because this movement actually completes the myoplasmic crescent. As mentioned before, this phase is clearly independent and different from the first phase. Early investigators reported coinci- dences in the localization of sperm aster and myoplasm and their movement, expecting a role of the sperm aster as motile mechanism in this phase [4, 7, 8]. Inhibition of myoplasmic movement by colcemid and nocodazole in the second phase (Sawada and Schatten, unpublished data) supports the speculation that microtubular components, such as the sperm aster, are involved in the second phase movement of myoplasm. Detailed observa- tions of the microtubular distribution in opaque ascidian eggs became possible using im- munofluorescence [20]. By this method, serial changes in the microtubules of ascidian eggs during fertilization and ooplasmic segregation were re- vealed (Fig. 4) and experimental studies on the microtubular structures became possible. SCHEMATIC MODEL OF THE MECHANISM OF OOPLASMIC SEGREGATION First phase movement After recent studies on the mechanism of the first phase movement, Jef- fery [29, 33, 34] and I [16, 25, 35] proposed schematic models for the cortical contraction and cytoplasmic movement. Both of us mention that Fic. 2. Light micrographs of an unfertilized egg (A) and fertilized egg at the late stage of cytoplasmic movement (B) in Ciona savignyi. Subcortical gran- ules (scg), specifically distributed in myoplasm and used as marker for the distribution of myoplasm, exist in the subcortical layer except for around the animal pole in the unfertilized egg (A). Myoplasmic layer (relatively dark area indicated by thick arrow in B) makes a fold and a gap is formed between the myoplasmic layer and the densely-stained cortex (c in B) of the vegetal area. ap, animal pole. vp, vegetal pole. Bar is 10 ~m. (B, from Sawada 1983 with permission). 670 the bipolar cytoplasmic segregation in the first phase is caused by the polarized contraction of the network of actin filaments underlying the egg surface and that this contraction is probably trig- gered by an increase in Ca**-ion which widely occurs in animal eggs. However, there is a differ- T. SAWADA ence between the two models with regard to the distribution of the cortical actin network in the unfertilized egg. In Jeffery’s model (1983, Fig. 3), the network covers all of the egg. In my model (Fig. 4A, B, C, D), an unequal distribution of actin filaments exists prior to fertilization and the Fic. VP 3. The Jeffery and Meier model (1983) for the first phase of ooplasmic segregation. (A) An unfertilized egg. The vertical arrow indicates the future focal point for segregation. (B—D) Fertilized eggs in the process of ooplasmic segregation. The arrows show the directions of ooplasmic movement; solid line arrows represent myoplasm and broken line arrows represent the endoplasm. (E) Zygote which has completed the first phase of ooplasmic segregation. In each diagram, the thick egg boundaries represent the parts of the plasma membrane with PML (plasma membrane lamina, actin filament-network) underneath, and the thin egg boundaries represent the parts of the plasma membrane without PML. The structures attached to the inside of the thick egg boundary represent the deep filamentous lattice (DFL, probably consisting of intermediate filaments) and its associated components, pigment granules and localized mRNA molecules of myoplasmic cytoskeletal domain. Ap, animal pole; VP, vegetal pole. (From Jeffery and Meier 1983 with permission). Fic. 4. My model for the first movement (A-D) and schematic representation of the serial changes of myoplasmic localization and sperm or mitotic aster in subsequent stages (E-L). Thick lines (in A-D) represent the cortical actin filament-network. The distribution of the cortical actin filament is ignored in E-L because it is not clearly known at those stages yet. Shape changes in E-L are also ignored for the same reason. Approximate time after insemination is indicated at the lower right side of the figures. The cortical actin filaments are absent or very sparse near the meiotic spindle (ms) at the animal pole in the unfertilized egg (A). After the activation of the egg, the actin filament-network begins to contract so as to form a contracting ‘basket’ with its mouth near the animal pole. This contraction creates a constricition at the mouth of the ‘basket’ (B), causes the drifting of the constriction toward the vegetal pole (C) and finally results in the formation of a small cap at the vegetal pole (D). The myoplasm (dotted area) is dragged by this contraction, and infolding at a late stage of its movement (C), is gathered near the vegetal pole (D). The inner yolk is pushed out of the constricted part (the mouth of the ‘basket’) toward the animal pole (B-D, arrows indicate the direction of movement of the inner yolk). The meiotic spindle rotates during this movement and becomes vertical to the egg-surface (E). + BSA (+) v ae 44 v » vyvyvyv < € < < < i ¢ a | + DACM-labeled = | | actin ( @ ) we: y S , EGTA wash f 5 ¥ @/ eacOe + factor ([ |) Rs eee “vy vvvvyvyvvvvevy sO era ( vy v Fic. 1. Schematic diagram of a system for quantification of the dissociation of 45K-A. Dissociation Factor of 45K-A 693 DEAE -buffer and incubated the plate for | hr at 4°C. After removal of unadsorbed 45K protein, 150 wl of BSA (1 mg/ml) dissolved in DEAE- buffer was added to each well and the plate was incubated again for 1 hr at 4°C to eliminate non- specific adsorption of actin in the later step. Ex- cess BSA was removed and the wells were washed several times with G-buffer containing 2mM CaCl,. DACM-labeled G-actin in G-buffer plus 2 mM CaCl, was added to each well to form a complex with the adsorbed 45K protein. After 30 min incubation at 4°C, free DACM-labeled actin was removed and the wells were washed several times with DEAE-buffer. A 150 «1 aliquot of the 0D 280 0.4 0.3 0.2 0.1 sample fraction containing 45K-A dissociating fac- tor(s) was added to each well and incubated at 4°C. After certain periods, the increase in fluorescence of the sample fraction was monitored by a Shima- dzu RF-540 spectrofluorophotometer at the ex- citation wave length of 395 nm and emission wave length of 460 nm. Neither of the fluorescence spectrum nor in- tensity of DACM bound to actin was affected in the presence of either Ca** or EGTA (data not shown). The amount of adsorbed DACM-labeled 45K-A was estimated to be 0.5 yg/well by dissolv- ing adsorbed proteins with 1 N NaOH followed by fluorophotometry. DACM fluorescence was not 10 FLUORESCENCE Fraction No. Pro 10 15>:20-25: 30. -35-40.45 S) 02- = 018 0 0 90'°55:°60' 65°70 °M Fic. 2. DEAE-cellulose column chromatography. Sea urchin egg crude extract (240 mg protein) was dialyzed against DEAE-buffer and loaded onto a DEAE-cellulose column (3.2 15cm). The column was washed with DEAE-buffer and eluted with a 0-0.3 M linear NaCl gradient. Top: An elution pattern of the protein which was monitored by absorbance at 280 nm (solid line). 45K-A dissociating activity was monitored by fluorescence of DACM (broken line). Bottom: SDS-PAGE pattern of the eluted proteins. Numbers at the top indicate fraction numbers. Marker proteins (M) are BSA (MW, 68,000), ovalbumin (45,000), carbonic anhydrase (30,000), Soybean trypsin inhibitor (20,000), cytochrome C (13,000). 694 M. OHNUMA AND I. MABUCHI released at all by incubation up to 24hr with DEAE-buffer which contained EGTA. This sug- gested that DACM-labeled actin was not adsorbed non-specifically but formed a complex with 45K protein that was not dissociated in the presence of EGTA. Egg extract was dialyzed against DEAE-buffer overnight and applied to a DEAE-cellulose col- umn which was preequilibrated with DEAE- buffer. Proteins were eluted with a 0-0.3 M linear NaCl gradient. The dissociating activities of 45K- A were eluted as two peaks (Fig. 2). The fractions eluted between 0.09-0.11 M NaCl (DEAE frac- tion), that is the major peak, were pooled and used in the following experiments. The DEAE fraction was diluted to various pro- tein concentrations and assayed for 45K-A dis- sociating activity (Fig.3). The rate of 45K-A dissociation was dependent upon the protein con- centration of the DEAE fraction. The amount of FLUORESCENCE Fraction No. Fr. 15 18 21 24 27 30 33 36 39 42 Fic. 3. Sephacryl S—300 column chromatography. DEAE fractions 25-30 were pooled, concentrated by Centriflo C-25, applied to a Sephacryl S—300 column (1.2 x60 cm) which was preequilibrated and eluted with F-buffer. Top: Solid line represents an elution pattern of the protein which was monitored by absorbance at 280 nm. The 45K-A dissociating activity was monitored by fluorescence of DACM (broken line). Stokes radius of the 45K-A dissociation factor (s) was estimated from the elution volume of standard marker proteins (inset). Standard marker proteins used for estimation were y-globulin (y-Glb), BSA and ovalbumin (OA). Bottom: SDS-PAGE pattern of the active fractions. Numbers at the top indicate fraction numbers. Those at the right indicate MW x 1073. Dissociation Factor of 45K-A 695 fluorescence dissociated reached a plateau level within 18 hours. The final amount of the fluores- cence dissociated was proportional to the protein concentration of the DEAE fraction. This sug- gested that the dissociation was not due to pro- teolysis or other enzymatic reactions. In order to investigate the nature of the dis- sociating factor, the DEAE fraction was treated with heat, trypsin, or RNase (Table 1). Heat treatment at 100°C for 3 min reduced the dissociat- ing activity to 2% of the original activity, while trypsin digestion reduced it to 24%. In contrast, RNase treatment did not affect the dissociating activity, suggesting that the dissociating factor(s) in the DEAE fraction is protein. Since the factors could be other modulating proteins, the effects of depolymerizing proteins on the dissociation of 45K-A were investigated by incubating 45K-A with profilin or depactin. However, the dissocia- tion of 45K-A was not observed (Table 2). The DEAE fraction was concentrated with Cen- triflo C-25, applied to a Sephacryl S—300 (Pharma- cia) column, and eluted with F-buffer (10 mM 3-[N-Morpholino] propanesulfonic acid (MOPS), 0.1M KCl, 0.2 mM ATP, 0.2 mM DTT, 0.5 mM EGTA, 1 yzg/ml leupeptin, and 5mM NaNs, pH 7.2) (Fig. 4). The dissociating activity was eluted as a symmetric peak at fractions 25-30. The top actin- actin- TABLE 1. dissociating activity. fraction corresponded to the Stokes radius of 4.1 nm. The elution pattern of proteins was investi- gated by SDS-PAGE and was compared with that of the activity. It was found that three polypeptides, the molecular weight of which were 40,000, 60,000, and 70,000, coeluted with the activity. DISCUSSION Fluorescence energy transfer is often used to detect association or dissociation of proteins [17]. At first, we attempted to study the dissociation of 45K-A by this method. N-iodoacetyl-N’-(5-sulfo- l-naphthyl) ethylenediamine (IAEDANS) and fluorescein isothiocyanate (FITC) were chosen to be energy donor and acceptor, respectively. However, in this experiment, the addition of EGTA was required to remove Ca** and this step had an effect on the fluorescence of FITC. Although the presence of either EGTA or Ca?* alone did not interfere with FITC fluorescence, FITC fluorescence was quenched in the presence of a micromolar order of EGTA-Ca?* complex (data not shown). This made it impossible to quantify the amount of dissociated 45K-A by fluorescence energy transfer. Therefore, we had to devise a new system. To quantify the amount of dissociated 45K-A, 45K Effect of heat, trypsin or RNase treatment on the 45K-A Relative amount of Sens dissociated 45K-A Untreated DEAE fraction (122,g/ml) 100 Heat-treated DEAE fraction (100°C, 3 min) 2 Trypsin-digested DEAE fraction (1 «g/ml, 5 min) 24 RNase-treated DEAE fraction (10 g/ml, 5 min) 100 TABLE 2. dissociation of 45K-A. Effect of low molecular weight actin-modulating proteins on the Sample Relative amount of dissociated 45K-A DEAE fraction (122yg/ml) Sea urchin egg profilin (13 ~g/ml) Starfish oocyte depactin (20 ug/ml) 100 3 7 696 M. OHNUMA AND I. MABUCHI FLUORESCENCE O42 Bras 18 24 TIME (hr) FLUORESCENCE 0 0.05 0.10 0.15 Conc. of crude factor (mg/ml) Fic. 4. Time course of the dissociation of 45K-A in the presence of the DEAE fraction. Top: The DEAE fraction having dissociating activity was diluted to various protein concentrations and 45K-A was dissociated by the addition of the diluted fractions. The amount of the dissociated 45K-A was monitored by the fluorescence of DACM. @: 41 vg protein/ml. m: 61 ug/ml. a: 122 ug/ml. When DEAE-buffer was added instead of the DEAE fraction, no fluorescence was observed (©). Bottom: The amount of released fluorescence at 24 hr after addition of the DEAE fraction (oridinate) was plotted against the protein concentration of the added fraction (abscissa). protein was adsorbed to the microtiter plate and with sea urchin egg extract, and that 45K-A dis- DACM-labeled actin was added in the presence of — sociating activity was due to protein factors. The Ca*t to form 45K-A. By use of this system, we Stokes radius of the factors was estimated to be 4.1 found that 45K-A was dissociated by incubation nm. Dissociation Factor of 45K-A The time course study (Fig. 3) showed that the dissociation of 45K-A was proportional to the protein concentration of DEAE fraction added. This suggested that the dissociation was not due to digestion of proteins by proteases or to other enzymatic reaction. The dissociation took 6-18 hours to reach a maximum level. We know of no reason why 45K-A dissociated so slowly, but it is likely that the conditions we used for dissociation were not optimal. The dissociating activity decreased during the course of purification and we have not yet purified the factor(s). The reason for this decrease in the activity is not known. We know that 45K protein from sea urchin egg severs the actin filament in the presence of mi- cromolar Ca?* [5]. Gelsolin and fragmin possess a similar activity [18-20]. These proteins also form complexes with actin in the presence of Ca** [17, 21], and once complex is formed, it is not dissoci- ated readily by removal of Ca** with EGTA [22]. These EGTA-stable complexes do not have an actin-severing activity, but cap the barbed end of the actin filament and thus cause the de- polymerization of actin. Therefore, the complex that is formed upon the fragmentation of actin caps the newly appearing barbed end of the actin fila- ment. The 45K-A dissociating factor(s) may re- move the 45K protein from the capping end of the actin filament to allow further elongation of the filament from this end at the rearrangement of the actin cytoskeleton during the development of the egg. This might result in the recirculation of the 45K protein. It is necessary to purify the dissociat- ing factor and to clarify the mode of dissociation of 45K-A. ACKNOWLEDGMENTS This work was supported by Grants-in-aid for Scien- tific Research from the Ministry of Education, Science and Culture, Japan. REFERENCES 1 Mabuchi, I. (1986) Biochemical aspects of cyto- kinesis. Int. Rev. Cytol., 101: 175-213. 2 Otto,J.J., Kane,R.E. and Bryan,J. (1979) Formation of filopodia in coelomocytes: Localiza- 10 11 13 14 697 tion of fascin, a 58,000 dalton actin crosslinking protein. Cell, 17: 285-293. Mabuchi,I. and MHosoya,H. (1982) Actin- modulating proteins in the sea urchin egg. II. Sea urchin egg profilin. Biomed. Res., 3: 465-476. Hosoya, H., Mabuchi,I. and Sakai, H. (1982) Actin modulating proteins in the sea urchin egg. I. Analysis of G-actin-binding proteins by DNase I- affinity chromatography and purification of a 17,000 molecular weight component. J. Biochem., 92: 1853-1862. Wang, L.-L. and Spudich, J. A. (1984) A 45,000- mol-wt protein from unfertilized sea urchin egg severs actin filaments in a calclum-dependent man- ner and increases the steady-state concentration of nonfilamentous actin. J. Cell Biol., 99: 844-851. Hosoya, H. and Mabuchi, I. (1984) A 45,000-mol- wt protein-actin complex from unfertilized sea urchin egg affects assembly properties of actin. J. Cell Biol., 99: 994-1001. Mabuchi, I., Hamaguchi, Y., Kobayashi, T., Hosoya, H., Tsukita, Sa. and Tsukita, Sh. (1985) Alpha-actinin from sea urchin eggs: biochemical properties, interaction with actin, and distribution in the cell during fertilization and cleavage. J. Cell Biol., 100: 375-383. Hosoya, H., Mabuchi, I. and Sakai, H. (1986) An 100-kDa Ca**-sensitive actin-fragmenting protein from unfertilized sea urchin egg. Eur. J. Biochem., 154: 233-239. Mabuchi,I. and Kane, R. E.(1987) A _ 250K- molecular-weight actin-binding protein from actin- based gels formed in sea urchin egg cytoplasmic extract. J. Biochem., 102: 947-956. Ohnuma, M. and Mabuchi,I. (1986) The 45K molecular weight actin-modulating protein from sea urchin egg forms a complex with actin in the pre- sence of calcium ions. J. Biochem., 100: 817-820. Coluccio, L. M., Sedlar, P. A. and Bryan, J. (1986) The effects of a 45,000 molecular weight protein from unfertilized sea urchin eggs and its 1:1 actin complex on actin filaments. J. Muscle Res. Cell Motility, 7: 133-141. Kurth, M. C. and Bryan, J. (1984) Platelet activa- tion induces the formation of a stable gelsolin-actin complex from monomeric actin. J. Biol. Chem., 259: 7473-7479. Spudich, J. A. and Watt, S. (1971) The regulation of rabbit skeletal muscle contraction. I. Biochemical studies of the interaction of the tropomyosin- troponin complex with actin and the proteolytic fragments of myosin. J. Biol. Chem., 245: 4866- 4871. Laemmli, U. K. (1970) Cleavage of structural pro- teins during the assembly of the head of bacte- riophage T4. Nature (London), 277: 680-685. 15 16 17 18 19 698 Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) Protein measurement with the folin phenol reagent. J. Biol. Chem., 193: 265-275. Laurent, T. C. and Killander, J. (1964) A theory of gel fileration and its experimental verification. J. Chromatogr., 14: 317-330. Giffard, R.G., Weeds, A.G. and Spudich, J. A. (1984) Ca?+-dependent binding of severin to actin: A one-to-one complex is formed. J. Cell Biol., 98: 1796-1803. Yin, H. L. and Stossel, T. P. (1980) Purification and structural properties of gelsolin, a Ca?*-activated regulatory protein of macrophages. J. Biol. Chem., 255: 9490-9493. Hasegawa, T., and Takahashi, S., Hayashi, H. 20 21 22 M. OHNUMA AND I. MaABuCcHI Hatano, S. (1980) Fragmin: a calcium ion sensitive regulatory factor on the formation of actin filaments. Biochemistry, 19: 2677-2683. Yamamoto, K., Pardee, J.D., Reidler, J., Stryer, L. and Spudich, J. A. (1982) Mechanism of interac- tion of Dictyostelium severin with actin filaments. J. Cell Biol., 95: 711-719. Bryan, J. and Kurth, M.C. (1984) Actin-gelsolin interactions: evidence for two actin-binding sites. J. Biol. Chem., 259: 7480-7487. Kurth, M. C., Wang, L.-L., Dingus, J. and Bryan, J. (1983) Purification and characterization of a gelsolin-actin complex from human platelets. J. Biol. Chem., 258: 10895-10903. ZOOLOGICAL SCIENCE 5: 699-711 (1988) Actin in Cytokinesis: Formation of the Contractile Apparatus Epwarp M. Bonper!, Douctas J. FISHKIND, JOHN H. HENSON, NINA M. CotrRAN and Davip A. BEGG Department of Biological Sciences, Rutgers University, Newark, NJ 07102 and Department of Anatomy and Cellular Biology, Harvard Medical School, LHRRB, Boston, MA 02115, U.S.A. ABSTRACT—The phenomenon of cytokinesis in animal cells is driven by the actin and myosin based contractile apparatus located within the cell’s cleavage furrow. In this report, we have examined the organizational dynamics of the sea urchin egg’s cortical actin cytoskeleton during the process of cytokinesis. Whole mounts, frozen sections and isolated cortices of fertilized and cleaving eggs were probed with either fluorescent-phalloidin or actin antibodies. Ultrastructural analysis of cortical actin filament organizaton was obtained by examining quick-frozen, deep-etched and rotary shadowed replicas of isolated zygote cortices. These experimental procedures demonstrated the change in spatial and temporal distribution of actin filaments within the dividing egg, leading to the extensive network organization of filaments within the contractile band apparatus. To address the mechanism of molecular regulation of the actin filament network, the structural properties of egg spectrin were analysed with regard to its relatedness to other spectrins, its relatedness to T. gratilla 220K actin binding protein and its presence in the fertilized egg cortex. Finally, the ‘collapsing baby-gate’ model for cytokinesis is presented, describing the genesis of the actin filament component of the contractile apparatus and its © 1988 Zoological Society of Japan potential regulation by actin filament crosslinking proteins. INTRODUCTION The history of research examining the mecha- nism of cell division contains a wealth of exciting and ingenious experimentation (see [1-3] for ex- cellent reviews). In many of these studies, eggs of marine organisms have often served as a primary experimental system. For example, Dan, his col- legues and others have used marine eggs and embryos to probe for cell membrane dynamics [4— 6], cortical cytoplasm dynamics [4-6], spindle- membrane interactions [7, 8], microfilaments and myosin interactions [9, 10], and cytoplasmic rheological changes [11-13] associated with cyto- kinesis. In simplest terms, cytokinesis results from the action of an actin-myosin based contractile system, the contractile ring [1, 2, 9, 10, 14], whose temporal and spatial manifestation is in part deter- Accepted March 17, 1988 ' Reprint requests to: Dr. E. M. Bonder, Department of Biological Sciences, Rutgers University, 101 Warren Street, Newark, NJ 07102, U.S.A. mined by the mitotic spindle [1, 3, 7, 8]. The molecular signals and structural changes modulat- ing the egg’s actin cytoskeleton during cell division still remain largely unresolved. Recently, using fluorescent cytochemistry of whole mounts, frozen sections and isolated cor- tices, in concert with rapid-freeze and deep-etch analysis of cortices, our labs have been able to define actin’s molecular and strucural disposition within the unfertilized egg cortex [15, 16]. In this report, the aforementioned methods were applied to fertilized and dividing sea urchin eggs in an effort to investigate the organizational changes in the actin cytoskeleton during cytokinesis. To begin dissecting the molecular *basis of the egg’s actin microfilament-membrane association, Fish- kind et al. ({17], also see [18]) have isolated and identified sea urchin egg spectrin. Interestingly, the egg spectrin isoform is Ca* *-sensitive in its binding interaction with actin filaments [17]. Here, we present data demonstrating egg spectrin’s pres- ence in the isolated cortex of fertilized eggs and spectrin’s relatedness to another egg actin filament 700 E. M. Bonper, D. J. FIsHKIND et al. crosslinking protein, T. gratilla 220K [19]. Based on our results using light and electron microscopy of fertilized eggs and isolated cortices, as well as the observations of many other investi- gators (for reviews see [1—3]), a model is presented describing the structural dynamics of the cortical actin cytoskeleton, leading to the formation of the ‘contractile apparatus’. Finally, we discuss how Cat *-sensitive actin filament crosslinking mole- cules may participate in modulating the physical characteristics of the actin cytoskeleton during the events of cell division. MATERIALS AND METHODS S. purpuratus (from Marinus, Inc., Westchester, CA), A. punctulata (from Marine Biological Laboratory, Woods Hole, MA) and L. variegatus (from Susan Decker, Hollywood, CA) eggs were obtained and prepared according to Begg and Rebhun [20] and used within two hours of collec- tion. Eggs were fertilized by addition of dilute sperm solutions. Fertilization envelopes were re- moved by mechanical stripping and the eggs were subsequently washed and reared in gently circulat- ing Ca* *-free sea water [21]. Light microscopy Developing zygotes were collected by gentle centrifugation in a hand oper- ated centrifuge. The concentrated cells were fixed by rapid addition of a 20-fold volume excess of Ca* *-free sea water fortified with 50 mM EGTA containing 1-3.7% formaldehyde solution. The stock 37% formaldehyde solution contained 11% methanol or less and these methanol concentra- tions were of utmost importance for preserving visible surface microvilli and rhodamine (rh)- phalloidin (Molecular Probes, Junction City, OR) staining of actin filaments. Higher concentrations of methanol in the formaldehyde solutions resulted in a diminution of phalloidin staining in eggs (personal observation) and tissue culture cells (Dr. M. Rutten, personal communication). The zy- gotes were fixed for | hr on ice before further processing. Fixed cells were washed out of fix by washing in 0.1% Triton X-100 in phosphate buffered saline (TX-PBS) and then incubated in TX-PBS contain- ing 50 nM rh-phalloidin for 1 hr. After incubation, the cells were incubated in TX-PBS to remove excess label, mounted on slides [22] and examined by epifluorescence or Nomarski light microscopy. Frozen sections were prepared according to the methods of Tokuyasu [23] as detailed by Bonder et al. [15]. Briefly, the fixed cells were embedded in gelatin, equilibrated in graded sucrose series and frozen by immersion into liquid nitrogen cooled Freon-22. Semi-thin sections were cut on a Reichert Ultracut E Ultramicrotome equipped with a cryo-stage. The sections were probed either with rh-phalloidin or with an affinity purified anti- actin antibody (generously donated by Drs. K. Fujiwara and S.Hagen). For indirect munofluorescence a commercially available (Hy- clone, Ogden City, Utah, USA) rhodamine labeled goat anti-rabbit antibody was used at a 1:200 dilution [15]. im- Electron microscopy Egg cortices were pre- pared according to the shearing method of Vac- quier [24]. Isolated cortices were quick-frozen, deep-etched and rotary-shadowed using the Heus- er and Kirschner technique [25]. Complete details of cortex preparation, fixation, myosin S-1 decora- tion and light microscopic rh-phalloidin staining are described in Henson and Begg [16]. Replicas of the cortices were examined on a JEOL-100CX at an accelerating voltage of 80 kV. Egg spectrin preparation and analysis Egg spec- trin was prepared according to Fishkind et al. [17]. Chicken erythrocyte spectrin was a gift from Dr. T. Coleman and human brain spectrin was a gift from Dr. A. Harris. Following extensive dialysis against 75 mM KCl, 1 mM MgCh, 0.2 mM CaCl, and 10 mM Imidazole pH 7.2, the various spectrin preparations were rotary shadowed according to Tyler and Branton [26] and examined by electron microscopy. Purified spectrins and egg preparations were analysed by immunoblotting [27] using monoclonal (a gift from Dr. J. Bryan) and polyclonal (a gift from Dr. J. Otto) antibodies raised against T. gratilla 220K actin binding protein [19]. Proteins were run on SDS-PAGE, transferred to Zeta- Probe (Bio-Rad, Richmond, CA) and reacted with Actin in Cytokinesis 701 anti-220K anti-serum (1:250 dilution) followed by immunodetection using a 1: 1000 dilution of alka- line phosphatase conjugated secondary antibody (Hyclone, Ogden City, Utah) according to Dub- reuil et al. [28]. Analysis of spectrin-like proteins in fertilized egg cortices was carried out on cortices isolated according to Begg and Rebhun [20]. Following isolation, the cortices were incubated in Solution A (75 mM KCl, 5mM MgCh, 5mM EGTA, 0.5 mM DTT and 20 mM Hepes pH 7.5) or Solution A containing 250 mM KCI or Solution A containing 0.2 mM CaCl, for 30 min on ice. The cortices were then sedimented in a microfuge for 10 min and the resultant supernates and pellets stoichiometrically analysed on SDS-PAGE. Other methods Myosin subfragment-1 was pre- pared according to Margossian and Lowey [29]. SDS-PAGE was carried out according to Laemmli [30] and gels were stained with Coomassie Blue [31]. Light microscopic observations were per- formed using a Zeiss ICM-405 and images were recorded on Kodak Tri-X 35 mm film. RESULTS Rhodamine-phalloidin staining of fertilized and di- viding sea urchin eggs Staining of cells and tissues with rhodamine (rh)-phalloidin has become a primary method for identifying and localizing F-actin [32]. To investi- gate global changes in cortical (microvillar and sub-membraneous) actin distribution during cyto- kinesis, fertilized and dividing eggs were gently fixed, permeabilized and reacted with rh- phalloidin. As judged by light microscopic (Nomarski optics) comparisons of fixed and un- fixed cells, there were no detectable structural alterations following fixation (Fig. la). Cytochem- ical staining of fertilized eggs with rh-phalloidin resulted in bright cortical fluorescence (Fig. 1b) with readily observable labeling of individual mi- crovilli that are uniformly spread across the surface of the egg (Fig. 1b). Examination of the eggs en face further demonstrated the uniform distribution of microvillar staining across the egg surface (Fig. lc). In these whole mount preparations it was difficult to definitively identify a sub-membraneous staining component (Fig. 1b). Cleaving eggs displayed a dramatic increase in the staining intensity at the cleavage furrow (Fig. ld). The increased fluorescence resulted both from an increase in fluorescence of the microvillar layer as well as from the appearance of a sub- membraneous band approximately 30 ~m wide (Fig. 1d). This sub-membraneous staining repre- sents the nascent stages of contractile apparatus formation within the cleavage furrow region. Additionally, the microvillar staining density is greater within the cleavage furrow than out at the poles (Fig. 1d). This corresponds to the apparent increase in microvillar density within the cleavage furrow that is observed by light microscopy of live and fixed samples of dividing eggs (personal observations, reviewed in [1]). When the same dividing cell (as in Fig. ld) is observed in surface views, the bright cleavage furrow band is com- posed of an extensive anastomosing network of fluorescence (Fig. le). Occassionally, small ‘ran- domly’ oriented filamentous bundles or fascicles are seen within this contractile band (Fig. le, see arrow). Figure If depicts a cell that is 50% cleaved and the furrowing region continues to be com- posed of an extensive meshwork of filaments and filament bundles. In all of our observations on whole dividing cells, an organized circumferential alignment of filaments was never observed with rh-phalloidin staining. To further examine the redistribution of cortical actin during cell division, fertilized and cleaving eggs were processed for frozen sectioning [15]. Examination of semi-thin frozen sections of ferti- lized eggs revealed discrete staining of the indi- vidual microvilli, and a less intense staining of the sub-membraneous actin domain (Fig. 2a, b). Cleaving eggs exhibited an increase in cleavage furrow staining both within the microvillar and sub-membraneous components (Fig. 2c). The dif- ferential distribution of microvillar and sub- membraneous actin within dividing eggs is further substantiated by indirect immunofluorescent stain- ing of frozen sections with an anti-actin antibody (Fig. 2d), which was monospecific for S. purpu- ratus egg actin [15]. = 8 ary IN) Q ra Ms en) 2 aay = a ea) Q Zz fo) —Q = jaa} Actin in Cytokinesis 703 Rapid freeze, deep-etch and_ rotary-shadow observations of isolated cortices from fertilized and cleaving eggs To investigate the ultrastructural organization of the actin filaments within the contractile band, isolated cortices were examined using rapid freeze according to Heuser and Kirschner ([25], see [16] for details). Fertilized cortices isolated 15 min post fertilization demonstrated the presence of elongate microvilli containing actin core bundles (Fig. 3a, b). The cytoplasmic plasma membrane surface is covered with numerous invaginating clathrin coat- ed vesicles (Fig. 3a, b) and a sparse network of actin-like filaments (Fig. 3a, b). Cortices isolated from cleaving eggs resulted in the isolation of ‘butterfly’ shaped cortices having the contractile band located in the constricted region between the two wings (Fig. 3f). Recently, Yonemura and Kinoshita [33] and Maekawa et al. [34] have re- ported the isolation of similar cortices using divid- ing eggs. Myosin S-1 decorated and rapid-frozen cortices from cleaving eggs demonstrate the pre- sence of an extensive three dimensional network of sub-membraneous actin filaments (Fig. 3d, e). The sub-membraneous network in cleaving eggs is clearly more extensive than that found at early times after fertilization (compare Fig. 3a and d). The image depicted in Figure 3d appears to corres- pond to the cleavage furrow because this cortex has a central constriction and long surface micro- villi characteristic of the furrow region. Notewor- thy is a striking absence of any detectable circum- ferential alignment of filaments within the cleavage furrow (Fig. 3d). This organization was indeed surprising since rh-phalloidin staining of the iso- lated cortex suggests the presence of circum- ferentially aligned filament structures (Fig.3f). Certainly, a more extensive characterization of the ultrastructure in the isolated cortices is needed to thoroughly substantiate these exciting prelimi- nary observations. Structural and immunological characterization of egg spectrin The structural conservation of three functionally diverse spectrin molecules — chicken erythrocyte, human brain and sea urchin egg — was examined by rotary shadowing according to Tyler and Bran- ton [26]. Under the experimental conditions em- ployed, all three spectrin isoforms are tetramers composed of two intertwined heterodimers form- ing an elongate (2-4 by 200-210 nm) molecule (Fig. 4, also see [17, 33]). Interestingly, in the presence of Ca **, the sea urchin spectrin mole- cule, at times, exhibits a disruption of the alpha/ beta interaction at the actin binding domain (Fig. 4c, also Fishkind, unpublished observations). The observed disruption at the actin binding domain may, in part, be responsible for the observed Cat* effects on egg spectrin binding to actin filaments [17]. Previously, Bryan and Kane [19] isolated a 220K egg protein, from gelled egg extracts, that demonstrated the same elution profile on size exclusion columns and the same inability to bind actin in the presence of 100 mM KCI as has been observed for egg spectrin [17]. To test the im- munological relatedness of these two molecules, monoclonal and polyclonal antibodies raised Fic. 1. Rhodamine phalloidin staining of fertilized and cleaving A. puntulata eggs. Panel a. Nomarski light micrograph of a fertilized egg probed with rh-phalloidin at the onset of spindle formation. Note the numerous microvilli covering the egg surface. Panel b. Fluorescence light micrograph of an equatorial view of the egg shown in panel a. There is clearly a one-to-one correlation between microvilli and rh-phalloidin fluorescence (compare arrowheads in panels a and b). Panel c. En face image of the egg in panels a and b. Notice the uniform distribution of fluorescence without any observable pattern. Panel d. Equatorial image of an rh-phalloidin stained egg that has started to undergo cytokinesis. There is now an increase in fluorescence intensity in the cleavage furrow as compared to the poles. The increased fluorescence is associated with the microvillae and submembraneous layers. Panel e. En face image of the cleaving egg in panel d. As expected there is an increase in staining intensity within the cleavage furrow which is composed of a filamentous network lacking any readily definable continuous circumferential alignment of filaments. Ocassionally, short fluorescent actin fascicles (see arrowhead) are observed within the contractile band. Panel f. Rh-phalloidin egg that is 50% cleaved. Again, observe the continued presence of a network-type organization lacking any identifiable circumferential actin bundle even at this late stage of cytokinesis. Magnification: x 600. 704 E. M. Bonper, D. J. FISHKIND et al. re ] Fic. 2. Rh-phalloidin and actin immunofluorescence staining of frozen sections of fertilized and cleaving eggs. Egg preparations equivalent to those used in Fig. 1 were prepared for semi-thin (1-2 4m) frozen sections. Panel a. Phase contrast micrograph of a frozen section of fertilized Arbacia eggs probed with rh-phalloidin. In such preparations the elongate microvilli are readily observable. Panel b. Fluorescence micrograph of the egg shown in panel a. The microvilli are highly fluorescent with some indication of a submembraneous actin component. Panel c. Rh-phalloidin stained cleaving egg at approximately the same stage as shown in Fig. 1d. Notice the apparent increase in microvillar staining density at the cleavage furrow as compared to the poles. Additionally, there is the readily identifiable increase in fluorescence associated with the contractile band. Fluorescent staining of the mitotic spindle is never observed. Panel d. Indirect immunofluorescence of a S. purpuratus cleaving egg using monospecific actin antibodies. The actin immunofluorescence staining also demonstrates the increase in microvillar staining density compared to the poles and the increased concentration of actin within the contractile band of the cleavage furrow. Magnifications: a and b. x 700; c. x 600; d. x 1200. Actin in Cytokinesis 705 against the 7. gratilla 220K protein were used for immunoblotting analysis. The monoclonal anti- body cross-reacted with purified egg spectrin and a band that co-migrated with egg spectrin in T. gratilla gelled extract (Fig. 5). The polyclonal antiserum cross-reacted with purified egg spectrin as well as with both subunits of erythrocyte and brain spectrins (Fig. 5). Furthermore, the poly- clonal antiserum is immunoreactive with a closely spaced doublet at 240 kDa in T. gratilla extracts similar to the closely spaced doublet composition of egg spectrin [17]. Recently, Mabuchi and Kane [35] have reported similar observations further demonstrating that T. gratilla 220K actin binding protein is actually a spectrin-related protein. Unfortunately, we have not been successful at using the antibodies for immunocytochemical studies because the monoclonal does not appear to work on frozen sections and the polyclonal is highly immunoreactive, not only with spectrin, but also with a 250K polypeptide in eggs and isolated cortices (Fig. 5). Preliminary data (Fishkind, un- published observations) suggests that the 250K cross-reactive polypeptide is a filamin-like protein and may be related to the 250K proteins isolated by two other groups [34, 35]. Immunoblotting fertilized egg cortices with the spectrin antibody (220K polyclonal) demonstrated the association of spectrin with the fertilized cor- tex. The salt and Ca** sensitive extractability of the cortex associated spectrin was examined by incubating the isolated cortices in either Solution A or Solution A containing either 0.2 mM Ca** or 250 mM KCI and 1 mM EGTA (see Materials and Methods). These experiments demonstrated the presence of a high molecular weight doublet in the isolated egg cortex that co-migrates with egg spectrin. The spectrin molecule is only slightly extracted by Ca** even though greater than 50% of the actin is solubilized (Fig. 6). High salt was much more effective at liberating the egg spectrin from the cortex. Interestingly, the spectrin re- mained associated with the cortex in the presence of Cat * even though the majority of the actin was solubilized (Fig. 6). This cortical association could be indicative of membrane anchorage by other red cell analogs present in the egg, such as ankyrin (see [36] for excellent review). Recent preliminary results (Fishkind, unpublished observations) have shown a high affinity binding interaction (Ky~4 x 10~’) between human red cell ankyrin and sea urchin egg spectrin. DISCUSSION The reorganization of cortical actin filaments during cytokinesis was investigated by examining whole mounts, frozen sections and rapid-freeze deep-etch preparations of fertilized and cleaving eggs. To follow the distribution of actin within the cortex, rh-phalloidin and indirect immunofluores- cence were used for light microscopic observa- tions. Three dimensional ultrastructural organiza- tion of the filaments was determined using rapid freeze techniques [16, 25]. The findings reported indicate that a global change in the distribution of cortical actin (microvillar core and submem- braneous) occurs prior to or concomitant with the formation of the cleavage furrow. Redistribution brings about an increase in microvillar and sub- membraneous actin within the equatorial region as compared to the poles. This initial positional shift in actin probably establishes the actin component of the future actin-myosin contractile machinery [2, 9, 10, 14]. Careful inspection of the submem- braneous actin indicates it is arranged as an exten- sive three-dimensional network of filaments and short bundles. This basic network organization is maintained throughout cytokinesis, and there are no detectable circumferentially aligned filaments. The visualization of circumferential contractile ring filaments [1, 14] by conventional thin section electron microscopy may in fact result from sec- tioning through the small filament fascicles observed by rh-phalloidin staining. The temporal changes in the distribution of actin leading up to and during cytokinesis can be corre- lated to other changes within the cortex and mem- brane of the egg. Dan and collegues, as well as others, have demonstrated that with the onset of cleavage there occurs: a. a ‘stretching’ or ‘relaxa- tion’ of the membrane at the poles [4, 37]; b. a movement of cortically associated echinochrome granules away from the poles and toward the equator ({1, 4, 38, 39], and Begg, unpublished observations); c. a decrease in tension at the poles E. M. Bonper, D. J. FISHKIND et al. 706 Actin in Cytokinesis 707 Ree aN fo 5: BNR ti lee ES, : x : BOT ey | Cc Be ey A a Fic. 4. Rotary shadowed images of three functionally diverse spectrin molecules. Panel a. Chicken erythrocyte spectrin. Panel b. Human brain spectrin (fodrin). Panel c. Sea urchin egg spectrin. Each spectrin species shares an elongate (200-210 nm), interwoven morphology composed of two heterodimer units assembled into functional actin filament crosslinking tetramer. The ends of the tetramers often display an enlarged, globular domain while the interior portions show obvious regions of strand separation (see arrows). Note the disrupted subunit interaction within the egg spectrin molecule demonstrated in the panel c (center molecule, see arrowhead). bth ore ret ¢] ¢ SS 3 f Kes * « in association with the onset of cytokinesis [11-13, 40]. We believe the occurrence of these events is a direct consequence of the underlying cellular re- location of the actin filaments toward the plane of cleavage and away from the poles. For example, how does the actin cytoskeleton modulate cortical tension? Originally, increases and decreases in tension were attributed to an active contraction based on an actin-myosin interaction [40]. Recent- ly, Schroeder [41] has demonstrated that myosin is not localized to the cortex of dividing blastomeres until the cleavage furrow starts to form, suggesting that cortical tension could be generated by an alternate mechanism. Rather than actin-myosin contraction, the cell could be regulating cortical tension by modulating the filament concentration in the cortex ([33, 42] and see Figs. 1-3) in coor- dination with changes in the degree of crosslinking Fic. 3. Quick-freeze, deep-etch ultrastructural analysis of isolated L. variegatus fertilized and cleaving egg cortices. Panel a. Replica of fertilized cortex isolated 15 min post-insemination depicting elongate microvilli at the edge of the cortex (see arrows). Microvillar core actin filament bundles (see arrowheads) can be seen emerging from the bases of the microvillar opening on the inner surface of the plasma membrane. Numerous invaginating clathrin coated vesicles (double arrowheads) are associated with the plasma membrane surface. Note, the relative absence of an actin filament network in association with the isolated cortex. Panel b. Higher magnification image of the microvillar actin filament bundles (arrowheads) observed in panel a. Panel c. Rh-phalloidin stained cortex of an equivalent preparation to panel a. Panel d. Quick-freeze replica of an isolated cortex from a dividing cell in the region of the cleavage furrow after decoration with myosin S-1. Note the presence of an extensive anastomosing network of decorated actin filaments and the apparent absence of a circumferentially aligned filament bundle. Panel e. Higher magnification of panel d illustrating the ‘twisted rope’ appearance (see arrows) of the isotropic network of decorated actin filaments. Panel f. Rh-phalloidin stained cortex from a preparation equivalent to panel d. Note the characteristic ‘butterfly’ shape of the cleaving egg cortex with the intense staining of the furrow region. Magnifications: a. x 20,000; b and e. x 70,000; c and f. x 520; d. x 17,000. 708 E. M. Bonper, D. J. FISHKIND et al. Mab 220K Fic. 5. gratilla 220K actin binding protein. Pab_anti- 220K 23456 789 Immunoblot analysis demonstrating the immunological relatedness of egg, red cell, and brain spectrins to T. A monoclonal antibody (IgG1) raised against T. gratilla 220K protein (Mab 220K) crossreacts with purified S. purpuratus egg spectrin (lane 1), and a 240 kDa polypeptide present in T. gratilla egg extracts (lane 2), A. punctulata eggs (lane 3) and S. purpuratus eggs (lane 4). Additionally, a polyclonal antibody serum made against T. gratilla 220K (Pab 220K) crossreacts with both egg spectrin and a 250 kDa polypeptide in samples of T. gratilla extracts (lane 2°), A. punctulata eggs (lane 3’) and S. purpuratus eggs (lane 4’). Pab 220K also recognizes spectrin and a 250 kDa in preparations of isolated S. purpuratus cortices (lane 5), 230 kDa and 250 kDa polypeptides in S. purpuratus coelomycytes (lane 6) chicken erythrocyte spectrin (lane 7), goat erythrocyte spectrin (lane 8) and human brain spectrin (lane 9). Note, the presence of an immunoreactive doublet at 240 kDa in T. gratilla egg extracts (lane 2”). Interestingly, the coelomycyte sample does not contain a 240 kDa polypeptide immunoreactive with Pab 220K. The asterisk marks the position of the 150 kDa proteolytic fragment of vertebrate spectrins. of filaments to each other and to the membrane. A viable candidate for this crosslinking role is egg spectrin because it crosslinks filaments [17] and is associated with the fertilized egg cortex ([43], and see Fig. 6). For a more detailed discussion of egg spectrin’s role in regulating the rheological prop- erties of the egg cortex, see Fishkind et al. [17]. Coalescence of the cortical actin structure to- ward the future cleavage furrow may result from localized changes in the degree of regional cross- linking. Diminished crosslinking density between the poles and the equator would manifest itself as a gradual ‘migration’ of the cortical actin cyto- skeleton away from the poles and toward the equator. This lower crosslinking at the poles is physically measured as a decrease in cortical ten- sion [11-13, 40]. Additionally, the equatorial translocation of actin would either passively or actively redistribute the echinochrome granules associated with the submembraneous actin net- work [1, 4, 38, 39]. Following the establishment of the actin compo- nent of the contractile band apparatus, myosin is finally positioned in the nascent cleavage furrow to provide the force needed for cell division [9, 10, 41, 44]. The force generation is isotropic, leading to a ‘collapse’ of the contractile band apparatus. Thus as modeled by White and Borisy [45], cell division proceeds by the tightening of an isotropic network of filaments rather than by the use of circumferentially aligned filaments. This revised version for the mechanism of constructing the contractile apparatus for cytokinesis is refered to, by our labs, as the ‘Collapsing Baby-Gate’ hypoth- Actin in Cytokinesis 709 Fic. 6. Solubilization of proteins from isolated fertil- ized egg cortices. Panel 1: Cortices prepared in Solution A (75 mM KCl, 5 mM MgCl, 5 mM EGTA and 20 mM Hepes, pH 7.5). Panel 2: Cortices in Solution A containing 0.2 mM CaCl, final concentration. Panel 3: Cortices in Solution A containing 250mM KCI and 1 mM EGTA. Note, in the presence of Ca** (Panel 2) a small amount of spectrin (240K doublet, see arrows) is released into the supernatant (s). At higher ionic strength (Panel 3) this effect is greatly enhanced. Ft: sample of cortex prior to extraction. P: sample of pellet after extraction. S: sample of supernate after extraction. esis. Hypothetical model for the mechanism of cyto- kinesis Over the years there have been many models proposed to describe the mechanism of cytokinesis [1-3, 11, 37, 40, 45, 46]. The model presented below incorporates our findings on actin dynamics within the egg and the finding and ideas of many previous investigators in an attempt to provide a testable hypothesis for the mechanism of cyto- kinesis. 1. Prior to the onset of cleavage there is a net increase in filament number ((33, 42], also see Figs. 1-3) as well as a net increase in filament- filament and filament-membrane crosslinking with- in the egg cortex. This change in actin content and crosslinking is physically observed as an increase in surface tension [11-13, 40]. Actin binding proteins that may be involved in cortical actin filament crosslinking include spectrin [17, 18], alpha-actinin [47] and filamin ({34, 35], and Fishkind, unpub- lished observations). 2. There is a global migration of cortical actin (microvillar and sub-membraneous) away from the poles and toward the equator prior to or concom- itant with the formation of the contractile appa- ratus (Figs. 1 and 2). The movement of actin away from the poles is detected as polar relaxation [11- 13, 40] and the accumulation of actin at the equator establishes the actin component of the contractile apparatus. The stimulus for relaxation could be elevated Ca** [48-50], which would decrease the crosslinking capabilities of spectrin [17] and/or alpha-actinin [47]. Alternatively, effec- tive crosslinking could be decreased by increasing filament number due to the activation of actin severing proteins [51-53]. 3. Once the actin component is positioned in the plane of future cleavage furrow, myosin is re- cruited into the region [41, 44] and provides the force generating component of the contractile band apparatus [9, 10]. As division progresses, the interaction of actin and myosin is generally iso- tropic, leading to the collapse of the extensive actin network (collapsing baby-gate). For an ex- cellent computer simulation of reorganizing net- works see White and Borisy [45]. Within this isotropic collapse there is the ocassional formation of small filament fascicles (Fig. 1) that by electron microscopy may be interpreted as the circumferen- tial alignment of contractile ring filaments [14]. 4. Active contraction continues until division is completed and the actin cytoskeleton returns to its resting state. Undoubtedly this is an oversimplified model for cell division that will need future adjustments in its fine details as new experimental information is accumulated. However, the model is testable and our labs are currently investigating many of the ideas being presented to gain a more complete 710 understanding of how actin is organized and reg- ulated during cell division. In conclusion, we extend our hearty thanks to Professor Dan whose experiments and insight have stimulated our think- ing and understanding of cytokinesis and the pro- cess of cell division. ACKNOWLEDGMENTS The authors wish to gratefully acknowledge Drs. K. Fujiwara and S. Hagen (actin), J. Otto (220K polyclonal) and J. Bryan (220K monoclonal) for their generosity with the various antibody preparations used in this study. We also warmly thank Dr. M. Neutra for use of the Reichert microtome, Dr. T. Coleman for erythrocyte spectrin, Dr. A. Harris for human brain fodrin, Dr. T. Phillips and Mrs. T. Phillips for their help in starting-up the frozen sectioning, Ms. Carol Bonder for help in manuscript preparation and finally the children in our families whose presence provided the inspiration for the ‘collapsing baby-gate’ hypothesis. This research was supported by National Institutes of Health grants GM 09788 (EB), GM 28307 (DB), GM 07226 and GM 07258 (DF) and HD 07130 (JH). EB was further supported by a Henry Rutgers Research Fel- lowship. REFERENCES 1 Schroeder, T. E. (1981) Interrelations between the cell surface and the cytoskeleton in cleaving sea urchin eggs. In “Cytoskeletal Elements and Plasma Membrane Organization”. Ed. by G. Poste and G. L. Nicolson. Elsevier, North-Holland Biomedical Press, pp. 169-216. 2 Mabuchi, I. (1986) Biochemical aspects of cyto- kinesis. Int. Rev. Cytol., 101: 175-213. 3 Rappaport, R. (1986) Establishment of the mecha- nism of cytokinesis in animal cells. Int. Rev. Cytol., 105: 245-281. 4 Scott, A. (1960) Surface changes during cell divi- sion. Biol. Bull., 119: 260-272. 5 Dan, K. (1954) The cortical movement in Arbacia punctulata eggs through cleavage cycles. Embryolo- gia, 2: 115-122. 6 Dan, K. and Ono, T. (1954) A method of computa- tion of the surface and area of the cell. Embryologia, 2: 87-98. 7 Hiramoto, Y. (1971) Analysis of cleavage stimulus by means of micromanipulation of sea urchin eggs. Exp. Cell Res., 68: 291-298. 8 Rappaport, R. and Rappaport, B. N. (1985) Sur- face contractile activity associated with isolated as- ters in cylindrical sand dollar eggs. J. Exp. Zool., 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 E. M. Bonper, D. J. FISHKIND et al. 235: 217-226. Mabuchi, I. and Okuno, M. (1977) The effect of myosin antibody on the division of starfish blasto- meres. J. Cell Biol., 74: 251-263. Kiehart, D. P., Mabuchi, I. and Inoue, S. (1982) Evidence that myosin does not contribute to force production in chromosome movement. J. Cell Biol., 94: 165-174. Wolpert, L. (1966) The mechanical properties of the membrane of the sea urchin egg during cleavage. Exp. Cell Res., 41: 385-396. Hiramoto, Y. (1970) Rheological properties of sea urchin eggs. Biorheology, 6: 201-234. Yoneda, M. and Dan, K. (1972) Tension at the surface of the dividing sea urchin egg. J. Exp. Biol., 57: 575-587. Schroeder, T. E. (1972) The contractile ring II. Determining its brief existence, volumetric changes and vital role in cleaving Arbacia eggs. J. Cell Biol., 53: 419-438. Bonder, E. M., Cotran, N. M., Fishkind, D. J. and Begg, D. A. (1988) Identification and localization of filamentous actin, non-filamentous actin and spec- trin domains. J. Cell Biol., (in press). Henson, J. H. and Begg, D. A. (1988) Filamentous actin organization in the unfertilized sea urchin egg cortex. Dev. Biol., (in press). Fishkind, D.J., Bonder,E.M. and Begg, D. A. (1987) Isolation and characterization of sea urchin egg spectrin: calcium modulation of the spectrin- actin interaction. Cell Motil. Cytoskel., 7: 304-314. Kuramochi, K., Mabushi, I. and Owaribe, K. (1986) Spectrin from sea urchin eggs. Biomed. Res., 7: 65- 68. Bryan, J. and Kane, R. E. (1978) Separation and interaction of the major components of sea urchin actin gel. J. Mol. Biol., 125: 207-224. Begg, D. A. and Rebhun, L. I. (1979) pH regulates the polymerization of actin in the sea urchin egg cortex. J. Cell Biol., 83: 241-248. Burgess, D. R. and Schroeder, T. E. (1977) Polar- ized bundles of actin filaments within the microvilli of fertilized sea urchin egg. J. Cell Biol., 74: 1032- 1037. Platt, J. L. and Michael, A. F. (1988) Retardation of fading and enhancement of intensity of im- munofluorescence by p-phenylenediamine. J. His- tochem. Cytochem., 31: 840-842. Tokuyasu, K.T. (1980) Immunochemistry on ultrathin frozen sections. Histochem. J., 12: 381- 403. Vacquier, V. D. (1975) The isolation of intact cor- tical granules from sea urchin eggs: calcium ions trigger granular discharge. Dev. Biol., 43: 62-74. Heuser, J. E. and Kirschner, M. W. (1980) Fila- ment organization revealed in platinum replicas of 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Actin in Cytokinesis freeze-dried cytoskeletons. J. Cell Biol., 86: 212- 234. Tyler, J. and Branton, D. (1980) Rotary shadowing of extended molecules dried from glycerol. J. Ultrastruct. Res., 71: 95-102. Towbin, H., Staehelin, H. and Gordon, J. (1979) Electrophoretic transfer of protein from polyacryla- mide gels to nitrocellulose sheets: procedure and some application. Proc. Natl. Acad. Sci. USA, 76: 4350-4354. Dubreuil, R., Byers, T. J., Branton, D., Goldstein, L. S. B. and Kiehart, D. P. (1987) Drosophila spec- trin. I. Characterization of the purified protein. J. Cell Biol., 105: 2095-2102. Margossian, S. S. and Lowey, S. (1973) Substruc- ture of the myosin molecule. III. Preparation of single-headed derivatives of myosin. J. Mol. Biol., 74: 301-311. Laemmli, U. K. (1970) Cleavage of structural pro- teins during the assembly of the head of bacterio- phage T4. Nature, 227: 680-685. Fairbanks, G., Steck, T. L. and Wallach, D. F. H. (1971) Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane. Biochemistry, 10: 2606-2617. Wulf, E., Deboben, A., Bautz, F. A., Faulstich, H. and Wieland, T. (1979) Fluorescent phallotoxin, a tool for the visualization of cellular actin. Proc. Natl. Acad. Sci. USA, 76: 4498-4502. Yonemura, S. and Kinoshita, S. (1986) Actin fila- ment organization in the sand dollar egg cortex. Dev. Biol., 115: 171-183. Maekawa, S., Endo, S. and Sakai, H. (1987) A high molecular weight actin binding protein: Its localiza- tion in the cortex of the sea urchin egg. Exp. Cell Res., 172: 340-353. Mabuchi,I. and Kane,R.E. (1987) A 250K molecular-weight actin-binding protein from the actin-based gels formed in sea urchin egg cytoplas- mic extract. J. Biochem., 102: 947-956. Marchesi, V. T. (1985) Stabilizing infrastructure of cell membranes. Ann. Rev. Cell Biol., 1: 531-561. Swann, M.M. and Mitchison,J.M. (1958) The mechanism of cleavage in animal cells. Biol. Res., 33: 103-135. Dan, K. (1954) Further study on the formation of the “new membrane” in the eggs of the sea urchin, Hemicentrotus pulcherrimus. Embryologia, 2: 99- 114. Rappaport, R. and Rappaport, B. N. (1976) Pre- furrow behavior of the equatorial surface in Arbacia eggs. Dev. Growth Differ., 18: 189-193. Schroeder, T. E. (1981) The origin of the cleavage forces in dividing eggs: A mechanism in two steps. 41 42 43 44 45 46 47 48 49 50 53 53 711 Exp. Cell Res., 134: 231-240. Schroeder, T. E. (1987) Fourth cleavage of the sea urchin blastomeres: Microtubule patterns and myosin localization in equal and unequal cell divi- sions. Dev. Biol., 124: 9-22. Usui, N. and Yoneda, M. (1982) Ultrastructural basis of the tension increase in sea urchin eggs prior to cytokinesis. Dev. Growth Differ., 24: 453-465. Shatten, H., Cheney, R., Balczon, R., Willard, M., Cline, C., Simerly, C. and Shatten, G. (1986) Lo- calization of fodrin during fertilization and early development of the sea urchins and mice. Dev. Biol., 118: 457-466. Yumura, S. and Fukui, Y. (1985) Reversible cyclic AMP-dependent change in distribution of myosin thick filaments in Dictyostelium. Nature, 314: 194- 196. White, J.G. and Borisy,G.G. (1983) On the mechanism of cytokinesis in animal cells. J. Theor. Biol., 101: 289-316. Marsland, D. (1956) Protoplasmic contractility in relation to gel structure: Temperature-pressure ex- periments on cytokinesis and amoeboid movement. Int. Rev. Cytol., 5: 199-227. Mabuchi, I., Hamaguchi, Y., Kobayashi, T., Hosoya, H., Tsukita,S. and Tsukita,S. (1985) Alpha-actinin from sea urchin eggs: Biochemical properties, interactions with actin, and distribution in the cell during fertilization and cleavage. J. Cell Biol., 100: 375-383. Yoshimoto, Y., Iwamatsu,T. and Hiramoto, Y. (1985) Cyclic changes in intracellular free calcium associated with cleavage cycles in echinoderm and Medaka eggs. Biomed. Res., 6: 387-394. Poenie,M.J., Alderton,J., Tsien,R.A. and Steinhardt, R. A. (1985) Changes of free calcium levels with stages of the cell division cycle. Nature, 315: 147-149. Yoshimoto, Y. and Hiramoto, Y. (1985) Cleavage in a saponin model of the sea urchin egg. Cell Struct. Function, 10: 29-36. Wang, L.-L. and Spudich, J. A. (1984) A 45,000- mol-wt protein from unfertilized sea urchin eggs severs actin filaments in a calcium-dependent man- ner and increases the steady-state concentration of nonfilamentous actin. J. Cell Biol., 99: 844-851. Hosoya, H. and Mabuchi, I. €1984) A 45,000-mol- wt protein-actin complex from unfertilized sea urchin egg affects assembly properties of actin. J. Cell Biol., 99: 994-1001. Hosoya, H., Mabuchi, I. and Sakai, H. (1986) A 100-kDa Ca**sensitive actin-fragmenting protein from unfertilized sea urchin eggs. Eur. J. Biochem., 154: 233-239. ZOOLOGICAL SCIENCE 5: 713-725 (1988) Immunofluorescent Analysis of Actin and Myosin in Isolated Contractile Rings of Sea Urchin Eggs Tuomas E. SCHROEDER and JoaNN J. Otto! Friday Harbor Laboratories, University of Washington, 620 University Road, Friday Harbor, WA 98250, and ‘Department of Biological Sciences, Purdue University, West Lafayett, IN 47907, U.S.A. ABSTRACT—Contractile rings attached to coverslips were “isolated” from sea urchin eggs at the time of first cleavage. Actin and myosin were localized by immunofluorescence after various treatments designed to dissolve, dissociate or dislodge these contractile proteins in order to determine how they are associated and which component is attached to the plasma membrane independently of the other. Experiments using high salt (600 mM KCI), ATP or pyrophosphate did not detectably alter the presence or distribution of either contractile ring actin or myosin. Treatment with gelsolin, a protein that severs actin polymers under the conditions used, rapidly reduced actin in the contractile ring to an undetectable level; since myosin was unaffected by gelsolin, these results suggested that contractile ring myosin is attached to the plasma membrane independently of actin. Additional preparations traced the build-up of actin and myosin in the cortex from fertilization onward, as a means of gaining insight into the formation of the contractile ring. The contractile ring appears by local augmentation of actin and myosin in the furrow, rather than at the obvious expense of actin and myosin from other parts of the cortex; hence, we tentatively conclude that the contractile ring does not form by lateral recruitment of polymers already fixed in the cortex but by accretion of other subunits, perhaps from the deeper © 1988 Zoological Society of Japan cytoplasm. INTRODUCTION The contractile ring is an array of microfilaments closely associated with the plasma membrane at the base of the cleavage furrow of dividing cells. It has been shown to be composed of contractile proteins, including actin and myosin, and to oper- ate in a manner somewhat akin to contracting muscle in generating the constriction force of cleavage [1-3]. The intimate structural and func- tional details of actin and myosin molecules in the contractile ring, however, are poorly understood. For example, it is not known how they interact during contraction, what role is played by assem- bly-disassembly kinetics, where the actin and myosin subunits come from as cleavage is initiated, or which component is attached to the plasma membrane. Such topics have been nearly inac- Accepted March 17, 1988 Dedication: The authors dedicate this paper to Professor Katsuma Dan in recognition of his outstanding career in cell division research. cessible to experimentation because the contractile ring is so short-lived and so small relative to the whole cell and because its actin and myosin repre- sent such minor fractions of the cytoplasmic pools of these abundant cytoskeletal proteins. Yonemura and Kinoshita [4] recently isolated cortices from sea urchin eggs that included con- tractile rings and used fluorescently-labeled phallo- toxin to confirm that the contractile rings contain polymeric actin. We were inspired by their success to investigate the relative distributions of actin and myosin by double-label immunofluorescence. Accordingly, in the present study we trace the association of both actin and myosin with the cortical cytoskeleton from fertilization to the formation of the contractile ring and report our initial explorations of the associations between these contractile proteins themselves and with the plasma membrane. Our strategy for determining whether actin or myosin is attached to the plasma membrane independently is to identify which con- tractile protein remains associated with the iso- 714 T. E. SCHROEDER and J. J. Orro lated cortex after the other has been removed, for example by dissociating the actin-myosin com- plexes or by selectively disassembling or dissolving one of the components. MATERIALS AND METHODS Spawning of the sea urchin Strongylocentrotus purpuratus was induced by KCl injection. Proce- dures for fertilization, denuding in urea, and cul- turing as monolayers in artificial calcium-free sea- water (CaFSW) at 12.0-14.0°C have been de- scribed elsewhere [5]. Egg cortices were obtained by attaching eggs to 22 x 22 mm coverslips pre- treated with 1% polylysine and shearing them in a stream of buffer whose composition is described elsewhere [6], except that the detergent Triton X-100 was omitted. Cortices from unfertilized eggs were prepared from mechanically dejellied eggs with intact vitelline envelopes; eggs of all other stages were denuded. Cortices from unferti- lized and post-fertilization stages, except for cleav- age stages which are described below, were obtained from eggs that were allowed to attach to coverslips for 1 min before shearing. These prepa- rations were fixed immediately after shearing. Initial efforts to “isolate” contractile rings re- vealed that the optimum time for attaching cleav- age stage eggs was difficult to predict, at least for the purpose of obtaining several coverslips-full of contractile rings from each culture, and also dem- onstrated that eggs attached for more than 10 min divided abnormally or not at all. Therefore, we developed a special procedure to avoid prolonged attachment while guaranteeing that each coverslip would contain at least some contractile rings. For each culture, several polylysine-covered coverslips were submerged in CaFSW in a Petri dish at the experimental temperature. When the onset of first cleavage was observed in the culture (about 120 min after fertilization), a pipetful of eggs was gently distributed among all of the coverslips; at 2-min intervals, until cleavage was obvious throughout the donor culture, additional pipetfuls were added to the same coverslips. By this proce- dure every coverslip included enough cortices iso- lated from cells at cleavage, although the stages were necessarily somewhat rancomized. After dividing cells were sheared, the coverslips bearing the cortices were either fixed immediately or were immersed in buffered experimental or control solutions of defined ionic, nucleotide or other contents at room temperature. ATP, GTP or sodium pyrophosphate was added to 50 mM KCI, 50mM MgCh, 2mM EGTA and 20mM Hepes buffer at pH 7.4, to which 20 mM glycerol was sometimes added. Additional KCI was added to the above buffer solution to obtain high-salt concentrations. Solutions of the actin-severing protein gelsolin (5 ~M) contained 50 mM KCI, 0.3 mM CaCl, 0.2 mM EGTA, 1mM MgCl, and 20 mM Hepes buffer at pH7.4. Human plasma gelsolin, prepared according to Cooper et al. [7], was generously supplied by Dr. Joseph Bryan (Baylor College of Medicine, Houston, TX). For indirect immunofluorescence microscopy, cortical preparations were fixed with 1% formal- dehyde in phosphate-buffered saline (PBS) for 5 min at room temperature and then with methanol for 10 min at —15°C. They were washed in PBS and rinsed in 1% bovine serum albumin in PBS. They were then double-labeled for actin and myosin by separate 45-min treatments at room temperature with antiactin and antimyosin fol- lowed by a mixture of secondary antibodies for 45 min. The specific primary antibodies were: 0.1 mg/ml of mouse monoclonal antiactin, provided by Dr. James L. Lessard (Children’s Hospital Re- seach Foundation, Cincinnati, OH), that reacts with all known vertebrate isoactins [8]; and 0.5 mg/ml of rabbit polyclonal antimyosin IgG. This antimyosin antiserum was prepared by one of us (JJO) in collaboration with Dr. Joseph Bryan (Baylor College of Medicine, Houston, TX) against sea urchin (Tripneustes gratilla) egg myosin purified as described by Kane [9]. It reacts with a ~200 kDa protein in eggs of the sea urchin S. purpuratus and oocytes of the starfish Pisaster ochraceus in immunoblots (Fig. 1). A ~180 kDa protein is also recognized in sea urchin egg samples and is probably a proteolytic fragment of myosin. This antimyosin antiserum was previously used to stain myosin in isolated cortices from. starfish oocytes [6] and the contractile ring of intact sea urchin cells [5]. In the present experiments affini- ty-purified antimyosin gave similar staining pat- Actin and Myosin in the Contractile Ring 715 Fic. 1. Characterization of antibody against sea urchin egg myosin by immunoblotting. A. Coomassie blue-stained electrophoresis gel. B. Immunoblot of similar gel stained with antimyosin IgG. Lanes: a, S. purpuratus eggs; b, Pisaster ochraceus oocytes; m, molecular weight markers (arrowheads): rabbit muscle myosin (200 kDa), (-galactosidase (116 kDa), phosphorylase b (92.5kDa), and bovine serum albumin (66kDa). The antibody to sea urchin egg myosin primarily recognizes a ~200 kDa protein in both sea urchin eggs and starfish oocytes which is interpreted to be myosin. Occasionally lower molecular weight bands are also recognized and are thought to be proteolytic fragments of myosin. terns as the unfractionated IgG (not shown). The secondary antibodies were rhodamine- labeled goat antimouse (E-Y Laboratories, Inc.) and fluorescein-labeled goat antirabbit IgG (Miles- Yeda, Ltd.), mixed together so that each was diluted 1:10. Control preparations lacking all antibodies or treated only with the secondary antibodies resulted in cortices with extremely low fluorescence (not shown). Stained cortices were observed by epifluores- cence microscopy using a 50 W mercury or 100 M tungsten-halogen lamp and filter sets appropriate for rhodamine or fluorescein, respectively, and photographed with Kodak Tri-X Pan film, routine- ly using a 63x objective lens and exposure times of 5 or 15sec. Figures 3c and d were photographed using a 40x objective. For immunoblotting (Fig. 1), samples separated by SDS-polyacrylamide gel electrophoresis [10] were blotted onto nitrocellulose as described by Burnette [11]. After blocking nonspecific binding sites with 0.25% gelatin and 4% bovine serum albumin in 0.9% NaCl, 15 mM Tris-HCl (block- ing buffer), the blots were stained with 75 g/ml antimyosin IgG followed by 2.5 «g/ml Protein A (Kierkegaard and Perry, Gaithersburg, MD), each diluted in blocking buffer. The substrates for the peroxidase were HO, and chloronaphthol. RESULTS Cortices isolated from sea urchin eggs were 1-2 yam in thickness and were attached to the coverslip by way of their outermost layer, which was the plasma membrane in all cases except for unferti- lized eggs in which the vitelline coat was out- ermost. Since detergents were not used in the present method of isolation, we assume that the plasma membrane was relatively intact; thus, the persistence of any cortical material after isolation implies its attachment to the plasma membrane, either directly or indirectly. Before describing the staining patterns of isolated cortices by antiactin and antimyosin antibodies in detail, we must point out that: a) the polymerization states of actin and myosin cannot be definitively determined by these methods, since the antibodies presumably stain all forms, including monomers, oligomers and poly- 716 T. E. SCHROEDER and J. J. Otro mers; b) the staining methods are qualitative, not quantitative; and c) actin-filled microvilli trapped between the isolated cortex and the coverslip inevitably contribute to the overall pattern of actin staining, confusing the pattern that may be due to actin in the cortex proper. Cortical actin and myosin after fertilization In cortices from unfertilized eggs, the staining pattern for actin was a faint uniform stippling (Figs. 2a and a’). The tiny dot-like structures appear evenly spaced 0.5—0.75 ym apart and prob- ably indicate the presence of small quantities of actin in the papillae that later develop into micro- villi after fertilization [12, 13]. Myosin staining in the same cortices was generally fainter and less regular than the actin staining (Figs. 2b and b’). Sparse myosin-positive spots, somewhat larger than the actin dots, were scattered randomly throughout the cortex against a nearly blank back- ground. Before fertilization no coincidence be- tween actin and myosin in the cortex could be detected and there were no fibrillar configurations. Both the amount and organizational complexity of actin and myosin in isolated cortices from sea urchin eggs increased abruptly after fertilization (Fig. 2), and these changes eventually stabilized before the onset of cleavage. By 22 min after fertilization (Fig. 2c) the actin pattern was still essentially punctate but, in addition, some of the dots were more brightly stained and clustered. Myosin staining at 22 min (Fig. 2d) was also some- what brighter; the pattern was still coarsely punc- tate, with early signs of a slightly fibrillar organiza- tion. Co-localization of actin and myosin was not yet evident. During subsequent development prior to first division (Figs. 2e—h) the punctate patterns of actin and myosin staining gave way to more reticular patterns and the degree of co-localization in- creased. By 61 min (Figs. 2e and f) this conversion was nearly complete. By 100 min (Figs. 2g and h), corresponding to the time of nuclear envelope breakdown of the first division, the patterns of actin and myosin staining were finely meshlike or reticular and there was considerable coincidence between them (Figs. 2g’ and h’). Meshlike patterns of cortical actin and myosin persisted through mitosis with little change up to the onset of cleavage when the contractile ring appears, as described below. Furthermore, once first division was complete, cortices isolated from 2-cell blastomeres at interphase continued to pre- sent finely reticular patterns of actin and myosin that were largely coincident (Figs. 2i and j). The cortex and contractile ring during cell Cortices isolated from unfertilized and early post-fertilization eggs were usually circu- division lar in outline; cortices isolated from cleaving eggs were often dumbbell-shaped, with the site of con- striction (containing the contractile ring) some- times imaging at a plane of focus different from the rest, presumably because the cleavage furrow was not intimately adhered to the coverslip at the time of shearing. Cortices were occasionally isolated as elongate strips, apparently as a result of eggs rolling along the adhesive coverslip surface during the shearing procedure (Fig. 3). Early in cleavage, the actin and myosin patterns were uniform throughout some of these long strips of cortical material, but in other cortical strips the contractile ring was evident as one or two transverse bands in which the intensities of actin and myosin staining were augmented (Figs. 3a-d). When two such transverse bands per strip were observed, they were about 100 um apart (Figs. 3c and d), as would be expected if an 80-um S. purpuratus egg rolled along during cortex isolation, peeling off two “sides” of the same contractile ring. Fic. 2. Pairs of immunofluorescence micrographs of isolated cortices double-stained for actin (a, c, e, g, and 1) and myosin (b, d, f, h, and j). Before fertilization, actin staining (a) is confined to an even field of low-intensity dots (a’, higher magnification), probably representing the papillary precursors of microvilli; myosin staining (b) is represented by occasional dots, (b’, higher magnification). More extensive staining for actin and myosin is present at 22 and 61 min after fertilization (c and d, and e and f). By 100 min after fertilization (g and h) the patterns of actin and myosin staining are reticular; the two patterns do not entirely correspond (g’ and h’, higher magnifications). Similar reticular patterns exist at 160 min (i and j), which is in the interphase between first and second cleavage. Scale bars: 10 um. Actin and Myosin in the Contractile Ring T. E. SCHROEDER and J. J. Otro Fic. 3. Isolated cortices at the time of cleavage. In its earliest form, a contractile ring may appear as a faint transverse band (a and b, between arrowheads) across a strip of isolated cortex in which both actin (a) and myosin (b) are slightly enhanced. Occasionally, a cortical strip from a cleaving cell contains two bright transverse bands (c and d), representing two parts of the same contractile ring isolated as the cell rolled along the coverslip during the shearing process; staining for actin (c) and myosin (d) are considerably enhanced in these bands over the surrounding cortex. The strip of cortex between the bands represents cortex material from the cell poles and is neither enriched or diminished in actin or myosin relative to other parts of the cortex. Scale bars: 10 um. Actin and Myosin in the Contractile Ring 719 Sometimes, a visible transverse band in such a cortical strip was barely detectable (Figs. 3a and b), perhaps because it represented a contractile ring at an early stage of formation. In such images, the concentrations of actin and myosin were only slightly greater in the band than in the adjacent cortex and no distinctive substructure was evident. Of the two components in the band in Figures 3a and b, myosin seems slightly more prominent than actin. As shown in Figures 3-6, fully-formed contrac- tile rings routinely straddled the constrictions in isolated cortices. They were consistently 5-10 ~m in width, were very thin and closely applied to the plasma membrane, and were markedly enriched for both actin and myosin relative to other parts of the cortex. The contrast between the contractile ring and adjacent cortex was consistently greater with myosin staining than with actin staining. At the poleward edges of the contractile ring, actin and myosin appeared to terminate quite abruptly rather than to merge gradually with the cytoskele- tal reticulum of the general cortex. In some instances, regardless of experimental conditions, the contractile ring was sometimes disrupted, as if by the mechanical trauma of isolation (e.g. Fig. 5). The actin and myosin patterns in portions of the cortex outside of the contractile ring at the time of cleavage roughly resembled the patterns before the onset of mitosis (Figs. 2g and h) or in the interphase after first cleavage (Figs. 2i and j), except possibly that the level of myosin staining was slightly reduced during cleavage. Fur- thermore, the cortex farthest from a contractile ring (the polar cortex) stained indistinguishably from the cortex right next to it. This was particu- larly evident in cortical strips where a complete transpolar transect was represented in the cortex between the two portions of the contractile ring (Figs. 3c and d); the cortex immediately adjacent to the contractile ring seemed equivalent to the cortex at the extreme pole in terms of the reticular organization of both actin and myosin. Contractile rings typically exhibited a fibrillar substructure with linear elements aligned along the cell constriction, i.e. circumferentially with respect to the cleaving egg. The character of staining varied somewhat between specimens; in some cases the myosin component appeared more fibril- lar than the actin component (Figs. 4c and d), sometimes the actin appeared more fibrillar (Figs. 4e and f), and sometimes neither was particularly fibrillar. Fibrils stained by antiactin tended to be long and parallel, whereas fibrils stained with antimyosin tended to resemble a mesh composed of interlinked short elements. In most instances there was a fairly close correspondence between the staining patterns of actin and myosin in the contractile ring, but that correspondence was nev- er perfect in all details and must be pursued at higher resolution. Experimental attempts to extract the contractile ring The following experiments were predi- cated on the assumption that actin and myosin in the contractile ring are structurally and functional- ly complexed together and that one of them, but probably not both, is attached to the plasma membrane. Accordingly, if actin-myosin com- plexes could be dissociated we would expect that the unattached component would be removed and that the attached component would persist. Assuming that salt concentration of 600 mM would dissociate actin-myosin complexes by solu- bilizing myosin, coverslips containing isolated cor- tices were immersed into buffered solutions of 50, 150, 300 or 600 mM KCI, to which 20 mM glycerol was sometimes added, for 5 min immediately after shearing and before fixation. In terms of actin and myosin patterns, however, and contrary to ex- pectations, contractile rings after all of these treat- ments (not shown) were consistently indistinguish- able from controls (Figs. 2 and 3) which had been fixed immediately after shearing. Likewise, when isolated cortices were transfer- red to 1, 2.5 or 5 mM ATP or inorganic pyrophos- phate in buffered 50 mM KCl for 5 min, treatments also expected to dissociate actin-myosin com- plexes, the contractile rings again remained fun- damentally intact (Fig. 5). After these treatments the adjacent cortical cytoskeleton frequently appeared slightly disorganized, but 5mM GTP caused a similar disordering, so the effect was interpreted as being a non-specific relative to actin- myosin dissociation. 720 T. E. ScHRoEpDeER and J. J. Orro Fic. Actin and Myosin in the Contractile Ring 721 5. Lack of effect of ATP on staining of contractile rings for actin (a and c) and myosin (b and d) after 5 min in 0 mM ATP (a and b) and 5mM ATP (c and d). Gross fragmentation of the contractile ring, as seen here, was occasionally seen in all preparations. Scale bar: 10 um. Fic. 4. Details of fully-formed contractile rings stained for actin (a, c, and e) and myosin (b, d, and f). The contractile ring is consistently less sharply distinguished from the nearby cortex after actin staining than after myosin staining, perhaps because of actin in the microvilli. In some instances the myosin component of the contractile ring seems to be slightly more fibrillar in organization (c and d) than the actin component, and sometimes the actin is more prominently fibrillar (e and f). Scale bars: 10 «zm. 722 T. E. SCHROEDER and J. J. Orro FiG. 6. Dissolution of actin by gelsolin treatment The actin-severing protein gelsolin [14] was ap- plied to isolated cortices at 1, 2.5 uM for 5-15 min. At the lowest concentration no effect was noticed, and the patterns of actin and myosin were indis- tinguishable from controls in buffer lacking gelso- lin (Figs. 6a and b). Actin staining of the contrac- Reduction of actin staining by treatment with the actin-severing protein gelsolin. Contractile rings were stained for actin (a, c, and e) and myosin (b, d, and f) after 15 min in control buffer lacking gelsolin (a and b), for 5 min in 5 uM gelsolin (c and d), and 15 min in 5 uM gelsolin (e and f). Myosin staining is unaltered despite the dissolution of actin. Scale bar: 10 um. tile ring was somewhat reduced after 2.5 “M gelso- lin and was significantly reduced after 5 uM gelso- lin (Figs. 6c and e); significantly, however, myosin staining was not altered by gelsolin (Figs. 6d and f). Even though actin in the contractile ring apparently disappeared completely after 5 min in 5 LM gelsolin (Fig. 6c), small amorphous aggregates Actin and Myosin in the Contractile Ring 723 of antiactin-staining material still demarcated the cortical fragment even after 15 min (Fig. 6e). DISCUSSION Development of the cortical cytoskeleton We infer that the membrane-associated cortical cytoskeletons shown here are reasonable repre- sentations of the cortices of normal living cells. On the other hand, these studies are still preliminary and therefore certain details may be altered by further study using additional protocols and other species. Our results on the development of the cortical cytoskeleton through the first cell cycle compare favorably with evidence obtained in other ways. Prior to fertilization cortex-associated actin seems to be confined to the short papillary precursors of microvilli, and cortex-associated myosin is even sparser. Comparing our results with phallotoxin- labeled cortices of unfertilized eggs [15] and other related findings [16, 17], the cortical actin at this stage is either largely or entirely in the polymeric form and very limited in amount. Cortical actin before fertilization is believed to be associated with the plasma membrane rather than cortical granules since it persists even after the latter have been selectively removed [18]. Based on the present immunofluorescence stud- ies, the cortical content of actin and myosin in- creases during the first 60 min after fertilization, which coincides with biochemical quantifications during the same period [19] that show an 8-fold increase of cortical actin and a coordinate increase of a 200kDa protein that was interpreted as myosin. After this time, as the cell undergoes mitosis and prepares to divide, no additional change in cortical actin or myosin was detected by either kind of investigation. Origin and organization of the contractile ring The source of contractile ring actin and myosin remains one of the important unsolved mysteries of cell division. Previous biochemical analyses of cortices from sea urchin eggs [3, 19] failed to show any increase in actin or myosin that correlated with the appearance of the contractile ring, thereby encouraging ideas that this structure forms from cytoskeletal elements already present in the cortex. One postulate in this category is that the cortical cytoskeleton builds up during a global contraction in mitosis and that this material is accumulated laterally from subequatorial or polar regions of the cortex to form the contractile ring [20]. Contrary to ultrastructural findings supporting this idea [21], however, the amount or organiza- tion of actin and myosin by immunofluorescent staining does not seem to increase between 60 and 120 min when the global contraction occurs, and there is no obvious gradient of depleted cytoskele- tal material from the polar to the subfurrow cor- tex, as might be expected. In terms of suggested scenarios for the formation of the contractile ring [2, 3], it thus seems improbable that the contractile ring is produced by mere rearrangement or lateral recruitment of pre-existing formed elements of the cortical cytoskeleton. The present evidence seems to be more consist- ent with the proposition that the contractile ring forms as a result of actin and myosin being locally added to the furrow cortex by recruitment of mobile subunits, either from cortical actin and myosin not yet fixed into the cortical cytoskeleton or from the deeper cytoplasm. A similar conclu- sion was reached in recent studies of the appear- ance of myosin in the contractile rings of intact sea urchin cells by immunofluorescence [5] and in cultured cells by microinjection of fluorescent myosin light chain [22], although none of this evidence can as yet be considered conclusive. As expected, actin and myosin appear to be intimately associated within the microfilamentous substructure of the contractile rings of sea urchin eggs, as in other cells. Important details of the association, however, remain unclear and require further study. Some features of the contractile rings shown here conflict with similar data obtained from other cell types; for example, con- tractile ring myosin in Dictyostelium has been shown to exist as discrete rods about 0.6 ~m in length [23], and striations of about this same dimension have been observed in the contractile rings of myosin-stained cultured cells [24] and in sea urchin egg myosin aggregated in vitro [25]. In contractile rings of sea urchin cells, neither dis- 724 T. E. SCHROEDER and J. J. Orto crete myosin rods nor striations have yet been clearly observed. Is contractile ring myosin attached to the plasma membrane independently of actin? In this preliminary study, we unexpectedly failed to ex- tract myosin from the contractile ring by dis- assembling its putative polymers in 600 mM KCl; this treatment was previously shown to be effective in extracting myosin from egg cortices prepared by a different procedure [19, 26], although the fate of contractile ring myosin was not specifically deter- mined. Similarly, we observed no evidence that actin-myosin complexes in the contractile ring were dissociated in 5 mM ATP or 5 mM pyrophos- phate, even though these procedures have been shown to be effective in muscle [27]. In these experiments, immunofluorescent staining of actin or myosin in the contractile ring was not noticeably reduced relative to controls; it remains to be determined if these results are due to some pecu- liarity caused by the preparative technique or if they reflect intrinsic properties of the contractile ring. In our experiments with exogenous gelsolin to dissect the contractile ring, most or all contractile ring actin was removed, since none was detected by immunofluorescence, yet neither the quantity or distribution of myosin was affected. This result opposes the idea that myosin is anchored to the contractile ring solely by its interaction with actin filaments, so we tentatively conclude that myosin is associated with the cortex (and therefore the plas- ma membrane, if only indirectly) independently of actin. On the other hand, because of the nature of gelsolin’s severing action on actin filaments, we cannot yet rule out the posibility that the myosin is attached to undetected residual actin dimers or oligomers persisting in the cortex. In conclusion, isolated cortices from dividing sea urchin eggs present an unusual opportunity to dissect the molecular organization of the contrac- tile ring, and we intend to extend the preliminary results reported here. For example, it will be interesting to explore the roles of the accessory proteins known to influence the polymerization states and association of actin and myosin in sea urchin cells [25, 28-33] and of other endogenous factors such as phosphorylation levels that may modulate the fully-formed contractile ring. These and other approaches should help to resolve how myosin and actin are associated with each other and how they are attached to the cell cortex and the plasma membrane. ACKNOWLEDGMENTS The authors are grateful for research support from the U. S. National Institutes of Health (grant HD/GM 20306) to T. E. S. and from the American Cancer Society (grant CD-108) to J. J. O. We also thank Dr. Joseph Bryan for generously supplying gelsolin and Dr. Issei Mabuchi for suggestions concerning the manuscript. REFERENCES 1 Schroeder, T. E. (1975) Dynamics of the contractile ring. In “Molecules and Cell Movement”. Ed. by S. Inoue and R. E. Stephens, Raven Press, New York, pp. 305-334. 2 Schroeder, T. E. (1987) The origin and action of the contractile ring. In “Biomechanics of Cell Division”. Ed. by N. Akkas, Plenum, New York. pp. 209-230. 3. Mabuchi, I. (1986) Biochemical aspects of cyto- kinesis. Int. Rev. Cytol., 101: 175-213. 4 Yonemura, S. and Kinoshita, S. (1986) Actin fila- ment organization in the sand dollar egg cortex. Dev. Biol., 115: 171-183. 5 Schroeder, T. E. (1987) Fourth cleavage in sea urchin blastomeres: microtubule patterns and myosin localization in equal and unequal cell divi- sions. Dev. Biol., 124: 9-22. 6 Otto, J.J. and Schroeder, T. E. (1984) Assembly- disassembly of actin bundles in starfish oocytes: an analysis of actin-associated proteins in the isolated cortex. Dev. Biol., 101: 262-273. 7 Cooper, J. A., Bryan, J., Schwab, B., Frieden, C., Loftus, D. J. and Elson, E. L. (1987) Microinjec- tion of gelsolin into living cells. J. Cell Biol., 104: 419-S01. 8 Lessard, J. L., Scheffter, S., Engel, L. and Tepper- man, K. (1983) Immunofluorescent localization of actins in differentiating chick myoblasts. J. Cell Biol., 97: 74a. 9 Kane, R. E. (1980) Induction of either contractile or structural actin-based gels in sea urchin egg cytoplasmic extract. J. Cell Biol., 86: 803-809. 10 Laemmli, U.K. (1970) Cleavage of structural pro- teins during assembly of the head of bacteriophage T4. Nature, 277: 680-685. 11 Burnette, W.N. (1981) “Western blotting”: elec- trophoretic transfer of proteins from sodium dodecylsulfate-polyacrylamide gels to unmodified 12 13 14 15 16 17 18 19 20 21 22 23 Actin and Myosin in the Contractile Ring nitrocellulose and radiographic detection with anti- body and _ radioiodinated Protein A. Anal. Biochem., 112: 195-203. Vacquier, V. D. (1981) Dynamic changes of the cell cortex. Dev. Biol., 84: 1-26. Schroeder, T. E. (1986) The egg cortex in early development of sea urchins and starfish. In “De- velopmental Biology”, vol. 2. Ed. by L. D. Brow- der, Plenum, pp. 59-100. Stossel, T. P., Chaponnier, C., Ezzell, R. M., Hart- wig, J.H., Janmey,P.A., Kwiatkowski, D. J., Lind, S. E., Smith, D. B., Southwick, F. S., Yin, H. L. and Zaner,K.S. (1985) Nonmuscle actin- binding proteins. Ann. Rev. Cell Biol., 1: 353-402. Yonemura, S. and Mabuchi, I. (1987) Wave of actin polymerization in the sea urchin egg. Cell Motil. Cytoskel., 7: 46-53. Spudich, A. and Spudich, J. A. (1979) Actin in Triton-treated cortical preparations of fertilized and unfertilized sea urchin eggs. J. Cell Biol, 82: 212- 226. Cline, C. A., Schatten, H., Balczon, R. and Schat- ten, G. (1983) Actin-mediated surface motility dur- ing sea urchin fertilization. Cell Motil., 3: 513-524. Kopf, G.S., Moy,G.W. and Vacquier, V. D. (1982) Isolation and characterization of sea urchin egg cortical granules. J. Cell Biol., 95: 924-932. Mabuchi,I., Hosoya,H. and Sakai,H. (1980) Actin in the cortical layer of the sea urchin egg. Changes in its content during and after fertilization. Biomed. Res., 1: 417-426. Schroeder, T. E. (1981) The origin of cleavage forces in dividing eggs: a mechanism in two steps. Exp. Cell Res., 134: 231-240. Usui, N. and Yoneda, M. (1982) Ultrastructural basis of the tension increase in sea-urchin eggs prior to cytokinesis. Dev. Growth Differ. , 24: 453-465. Mittal, B., Sanger, J. M. and Sanger, J. W. (1987) Visualization of myosin in living cells. J. Cell Biol., 105: 1753-1760. Yumura, S. and Fukui, Y. (1985) Reversible cyclic AMP-dependent change in distribution of myosin 31 es) nN 33 Ws thick filaments in Dictyostelium. Nature, 314: 194- 196. Sanger, J.M. and Sanger, J. W. (1980) Banding and polarity of actin filaments in interphase and cleaving cells. J. Cell Biol., 86: 568-575. Yabkowitz, R. and Burgess, D. R. (1987) Low ionic strength solubility of myosin in sea urchin egg ex- tracts is mediated by a myosin-binding protein. J. Cell Biol., 105: 927-936. Mabuchi, I. (1973) A myosin-like protein in the cortical layer of the sea urchin egg. J. Cell Biol., 59: 542-547. Harrington, W.F. and _ Rodgers, M. E. Myosin. Ann. Rev. Biochem., 53: 35-73. Wang, L.-L. and Spudich, J. (1984) A 45,000-mol- wt protein from unfertilized sea urchin eggs severs actin filaments in a calcium-dependent manner and increases the steady-state concentration of non- filamentous actin. J. Cell Biol., 99: 844-851. Hosoya, H. and Mabuchi, I. (1984) A 45,000-mol- wt protein-actin complex from unfertilized sea urchin egg affects assembly properties of actin. J. Cell Biol., 99: 994-1001. Mabuchi, I., Hamaguchi, Y., Kobayashi, T., Hosoya, H., Tsukita,S. and Tsukita,S. (1985) Alpha-actinin from sea urchin eggs: biochemical properties, interaction with actin, and distribution in the cell during fertilization and cleavage. J. Cell Biol., 100: 375-383. Hosoya, H., Mabuchi, I. and Sakai, H. (1986) An 100-kDa Ca**-sensitive actin-fragmenting protein from unfertilized sea urchin egg. Eur. J. Biochem., 154: 233-239. Maekawa, S., Endo, S. and Sakai, H. (1987) A high molecular weight actin binding protein: its localiza- tion in the cortex of the sea urchin egg. Exp. Cell Res., 172: 340-353. Mabuchi, I. and Kane, R.E. (1987) A 250 K- molecular-weight actin-binding protein from actin- based gels formed in sea urchin egg cytoplasmic extract. J. Biochem., 102: 947-956. (1984) 4 J va in leweoui) Fae rit ONT 7 iP flay My Fin Published by 4, 4 ff : 4 A Gi & erontiation the Japanese Society of Developmental Biologists Papers in Vol. 30, No. 3. (June 1988) 20. REVIEW: S.-I. Ape: Cell culture of spermatogenic cells from amphibians. 21. R.E. Hinkley, Jr. and A. N. NEWMAN: Changes in the distribution of calcium-sequestering membranes during the first cell cycle of the sea urchin, Lythechinus variegatus. 22. H.L.Hosick, Y. Inaguma, M. Kusakabe and T. SAkAKuRA: Morphogenesis of mouse mam- mary epithelium in vivo in response to biomatrix prepared from a stimulatory fetal mesen- chume. 23. K. Takiguchi, S. Yasugi and T. Mizuno: Popsinogen induction in chick stomach epithelia by reaggregated proventricular mesenchymal cells in vitro. 24. S. Kobayashi, Mizuno and M. Okapa: Accumulation and spatial distribution of poly(A)+ RNA in oocytes and early embryos of Drosophila melanogaster. 25. S.Tone, S.Tanaka and Y.Kato: The cell cycle and cell population kinetion in the programmed cell death in the limb-buds of normal and 5-bromodeoxyuridine-treated chick embryos. 26. S.Noda and T. Mitsui: Distributions of actin, vinculin and fibronectin in the duodenum of developing chick embryos: Immunohistochemical studies at the light microscope lebel. 27. A.W. Gibson and R. D. Burke: Localization and characterization of an intergral membrane protein antingen expressed by pigment cells in embryos of the sea urchin Strongylocentrotus purpuratus. 28. S. Ohta, Y. Suzuki, W. Hara, S. Tokiya and T. Suzuki: Fibroin gene transcription in the embryonic stages of the silkworm, Bombyx mori. 29. O.Taguchi, R.M. Bigsby and G.R.Cunua: Estrogen responsiveness and the estrogen receptor during development of the murine reproductive tract. 29. M.L. Wright, S.T. Jorey, Y.M.Meyrs, M.L. Fieldstad, C.M. Paquette and M. B. Clark: Influence of photoperiod, daylength, and feeding schedule on tadpole growth and development. Development, Growth and Differentiation (ISSN 0012-1592) is published bimonthly by The Japanese Society of Developmental Biologists, Department of Biology, School of Education, Waseda University, Tokyo 160, Japan. 1988: Volume 30. Annual subscription U.S. $110.00 including air speed delivery except Japan. Application to mail at second class postage rate is pending at Jamaica, NY 11431, U.S.A. Outside Japan: Send subscription orders and notices of change of address to Academic Press, Inc., Journal Subscription Fulfillment Department, 6277 Sea Harbor Drive, Orlando, FL 32887, U.S. A. Send notices of change of address at least 6-8 weeks in advance. Please include both old and new addresses. U.S. A. POSTMASTER: Send changes of address to Development, Growth and Differentiation, Academic Press. Inc., Journal Subscription Fulfillment Department, 6277 Sea Harbor Drive, Orlando, FL 32887, U.S. A. In Japan: Send nonmember subscription orders and notices of change of address to Business Center for Academic Societies Japan, 16-3, Hongo 6-chome, Bunkyo-ku, Tokyo 113, Japan. Send inquiries about membership to Business Center for Academic Societies Japan,4—16, Yayoi 2-chome, Bunkyo-ku, Tokyo 113, Japan. Air freight and mailing in the U.S. A. by Publications Expediting, Inc., 200 Meacham Avenue, Elmont, NY 11003, U.S.A. THE BOTANICAL MAGAZINE TOKYO An international journal for plant sciences published quarterly by the Botanical Society of Japan. For acentury, the journal has continuously published outstand- ing papers by Japanese as well as foreign botanical scientists. Contributors to the journal are not limited to the members of the Society and their papers are accepted by paying the page charge. Papers in a Recent Issue: Yokota, M., 8. Hica, H. Yosuroka anv K.Sxuimasuku: Viola stoloniflora (Violaceae), a New Species from the Ryukyus Ragu, M.V.S., A. WaLtHEeR anp W.A. Quick: Growth and Development of Embryo Parts during the Germination of Caryopses of the Wild Oat (Avena fatua L.) Fuxupa, Y.: Phyllotaxis in Two Species of Rubia, R. akane and R. sekkimensis Sou, W.Y., S.S. Hone anp D.Y. Coo: The Ontogeny of the Vascular Cambium in Ginkgo biloba Roots AGUINAGALDE, I.: Flavonoids in Brassica nigra (L.) Koch, B. oleracea L., B. campestris L. and Their Natural Amphidiploids Yamamoto, S.: Seedling Recruitment of Chamaecypans obtusa and Scradopitys vertiillata in Different Microenvironments in an Old-Growth Sciadopitys verticillata Forest Mariko, S.: Maintenance and Constructive Respiration in Various Organs of Helianthus annuus L. and Zinma elegans L. Morr, H.: Control of Plant Development: Signals from Without — Signals from Within Order form Send to THE BOTANICAL SOCIETY OF JAPAN Toshin Building Hongo 2-27-2, Bunkyo-ku, Tokyo 113, Japan THE BOTANICAL MAGAZINE, TOKYO | Individuals: ¥ 7,000 p.a. () Institutions: ¥ 17,500 p.a. Name (Please print) : Address : Date : Signature : NARISHIGE THE ULTIMATE NAME IN MICROMANIPULATION OUR NEW MODELS MO-102 and MO-103 MAKE PRECISION MICROMANIPULATION SO EASY! We ay 4 igs 4, ee wi (Photo: by courtesy of Olympus Optical CO., LTD.) - SOME FEATURES of MO-102 and MO-103: « The manipulator head is so small that it can be mounted directly on the microscope stage. There is no need for a bulky stand. * Hydraulic remote control ensures totally vibration-free operation. * 3-D movements achieved with a single joystick. Micromanipulators Microelectrode pullers Stereotaxic instruments NARISHIGE SCIENTIFIC INSTRUMENT (2 / LABORATORY CO.,LTD. 4-9-28, Kasuya, Setagaya-ku, Tokyo 157 JAPAN Telephone: 03-308-8233 Telex: NARISHG J27781 ZOOLOGICAL SCIENCE VOLUME 5 NUMBER 3 JUNE 1988 CONTENTS OF SPECIAL ISSUE ON Advances in Cell Division Research Sakai, H.: General introduction to the spe- cial issue on Advances in Cell Division Re- search Dan, K.: urchin egg: mechanism to constricting mechanism Mazia, D.: nogenesis: Mechanism of equal cleavage of sea astral ... 507 transposition from Mitotic poles in artificial parthe- a letter to Katsuma Dan .....519 Inoué,S.: The living spindle ............. 529 Nakano, Y. and Y. Hiramoto: Measurement of spindle birefringence by the optical in- tegration method Hamaguchi, Y.: Jn vivo cytochemistry in cell division Yoneda, M.: Computed profiles of compressed sea-urchin eggs with elastic membranes. . 553 Yamao, W. and T. Miki-Noumura: Effect of hexyleneglycol on meiotic division of starfish oocytes Longo, F., W. H. Clark, Jr. and G. W. Hinsch: Gamete interactions and sperm incorpo- ration in the nemertean, Cerebratulus lacteus Schatten, H., C. Howard, G. Coffe, C. Simer- ly and G. Schatten: Centrosomes, trioles and post-translationally modified mi- cen- crotubules during fertilization Palazzo,R.E., J.B.Brawley and = LI. Rebhun: Spontaneous aster formation in cytoplasmic extracts from eggs of the surf clam Ohta, K., M. Toriyama, S. Endo and H. Sakai: Mitotic apparatus-associated 51-kD protein in mitosis Sato, H. and J. Bryan: The thermodymanics of molecular association in the mitotic spin- dle with or without heavy water (D2O) .. 623 Harris, P. J.: tion of sea urchin eggs examined in caffeine- induced monasters Kojima, M. K.: Marked elongation of the anaphase spindle by treatments with local Metaphase to anaphase transi- anesthetics in sea urchin eggs Sluder, G.: The role of spindle microtubules in the tim- Control mechanisms of mitosis: ing of mitotic events Sawada, T.: segregation in the ascidian egg The mechanism of ooplasmic Kawamura, K.: The contraction wave in the cortex of dividing neuroblasts of the grass- hopper Sawai, T.: Participation of the subcortical and interior cytoplasm in cleavage division of newt eggs Ohnuma, M. and I. Mabuchi: Partial purification and characterization of a factor which dissociates 45K protein-actin complex from sea urchin egg .................2000 6% 691 Bonder, E. M., D. J. Fishkind, J. H. Henson, N.M.Cotran and D.A. Begg: Actin in cytokinesis: Formation of the contractile apparatus Schroeder, T. E. and J. J. Otto: fluorescent analysis of actin and myosin in Immuno- isolated contractile rings of sea urchin eggs INDEXED IN: Current Contents/LS and AB & ES, Science Citation Index, ISI Online Database, CABS Database Issued on June 15 Printed by Daigaku Printing Co., Ltd., Hiroshima, Japan A SIE ON. - ai Se a er eer Tene |_NVINOSHLIWS $3 bYVeait_ LIBRARI ES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIW: > wT - 0 z ow s RS Sh = n = Ar “Z a” _ = AS a = «. tY% sy = a = oa ANS. No a = a WN rE > ra i = ae on co = o m YW D e E > 4s, Na 4 _ NVINOSHLINS |S 3 tuvudlt _t! BRARI ES |, SMITHSONIAN INSTITUTION, NOILNLILSNI _ NVINOSHLIMS =? = < : = = <= 2 xt z ={ mas ¢ = => xX = GEE >, Zz z E 5 Wy 2 2 \y i G2 2 g 2 SN 8 2 WN HG? ra > = s = ne ras WS S Y f° CZ) = — —, pa Nn — —_ S\S — — = \ Se a 3 ” » 2 7 re bd = S SMITHSONIAN INSTITUTION NOILNLILSNI_ NVINOSHLINS S31YVYAIT LIBRARIES SMITHSONIAR rm Yy z a z if = 4 : 4, g = Z & : z - YY, = Dy fs = = = SS < = x Df. 4 top SG fae: MCL c Mihi fo. a Y = fs = WS co = ae Y ; a 2 a 2 = a = ef | _NVINOSHLINS S3INVYGIT LIBRARIES SMITHSONIAN INSTITUTION NOILMJILSNI_NVINOSHLIWS ~~ ; 7“ ~~ on = feo _ . a fs) — re) a re) — 2 ww OE 2 5 5 5 2 KD = iy = = NK = ra See iE Baa 2 Wa = = oe 2 2 7 m ” m os m oh m n ee wn = n rs wn : 5 SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S3INVYIT_ LIBRARIES. SMITHSONIA g = aft g z g z x g : ESS 2 wy? = ty =E=N 3 g 2S 8 fy By he 2RWE 2 2 =z WMO 2 “YD * 2 “iy = VK 2 2 eS . 5 8 a ~ } ai . 8 || NVINOSHLINS S31¥VYdIT LIBRARIES SMITHSONIAN INSTITUTION) NOLLMLILSNI_ NVINOSHLIW 7 * = ss (ep) aW S as u a rae <” = nw = AS w re x . Wy wy = x = ERR = : Gly: = Rue) = X 2 Gi 2“ 5 : a Wo 3 : 3 X a z 4 ae S~ SMITHSONIAN” INSTITUTION NOILMLILSNI NVINOSHLINS S31YVYGIT LIBRARIES z ‘ - z — 4 z é z iS O° a S) = S) . = 6% 5 2 Ne 5 2 5 £G a Vi, Ss - a WR x 4 iN 2 r Z 5 WE r Z N Ji NVINOSHLINS SSIYVYEIT LIBRARIES SMITHSONIAN INSTITUTION NOILOLILSNI_ NVINOSHLIW SMITHSONIAN NVINOSHLINWS NVINOSHLIAS wt SMITHSONIAN NVINOSHLIWS SMITHSONIAN S SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYVYEIT LIBRARIES SMITHSONIAI LIBRARIES SMITHSONIA Mi S et 2 2 a - 2 : 2 g = ) < iy fame < =. Oe ‘\\ = < Di a x YY fl = a + “N SS eS a ™. por = Fs} © o OSS Fs} oF 2) =z J Zz = a 1] NYVINOSHIIWCO C3AIlMYyUuUgIDT FTIR PARIES CRAITHCONIAN INGTITIITION NOLINITITICNE NYVINOSHLIW! wt .f/ aA i. al Y. US JUGi7 LIBRARIES SMITHSONIAN INSTITUTION NOJLOLILSNI NVINOSHLIWS bo SRINVYUSIT LIBR: =, ee NV SM NV SN NV SM NVI Zz i 2 a 2 &. @ z x a. ii fy = re a XY = a = < PL = < a IN. oY oey q : sp 3 : a YE i 5 a GY 35 ss 5 ae 5 | 2 = 2 - 2 eh ed 2 UTION > NOILALILSNI NVINOSHLIWS 233 IyVvug ig Uh LIBRARI ES_ SMITHSONIAN INSTITUTION NOILN » = (@) sty [e) mat Oo = oO = o : = ow = Yn OD = 5 2 Ne 5 : 5 £G= = = 9 VN = xy = Uf > = - es SNS = . pee = 2 mn SW 8 m 2 = 2 ‘Yai LIBRAR | ES SMITHSONIAN INSTITUTION NOILOLILSNI_ NVINOSHLINS, $3 iyYvadity kl BRE y, ‘ = : = < = Wi a ts 4 oot S = V4 % pa pases “yj _ Vy, 3 Ny 3 82 = YG.s ® ‘La g aN 8 HE ? 8B PLIES ab =z AS z . A f° 4 = = Pi. a - = \ \ ae pes : : Fo NOILNLILSNI NVINOSHLIWS S3IYVYGIT LIBRARIES SMITHSONIAN INSTITUTION NOI/LN. UTION LIBRARIES SMITHSONIAN NOILNLILSNI Ye NOILALILSNI LIBRARIES SMITHSONIAN LIBRARIES NOILALILSNI LIBRARIES Yall LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYNVURIT LIBR: - 5 ~ S cs z - a = f es = o ; 2 oo = > LY y a = Pe) “ > = x 3 = “jy-ip = = = ae > ~ - Oy ia - a = » D 2°)» z mn 2 mn st — Ww : aoe w UTION | NOILOLILSNI_ NVINOSHLINS S31YVYEIT_LIBRARIES SMITHSONIAN INSTITUTION NOLL = = = = Ks = z ‘\ g = z + uf fy = NS = z WKS x : E 2 yy = » 2 z WO 3 2 7) 2 nv » > = .. S g a SMITHSONIAN _INSTITUTION NOILMLILSNINVINOSHLINS S3IMVYaIT LIBR & mi = rt 2 & 2 2 — <« pees w § KY 3 e = ez 4 WS = g 4 xt % a < =] MN = = = oc =| a = ; \. ow Cc = = a 9 oy : ro) a4 > . oo = J 4 =x! —- ay 23 = ITION | NOILALILSNI _NVINOSHLINS 234 lyuvugi 7 LIBRARI ES” SMITHSONIAN” INSTITUTION _ NOILLN. Oo — oO a oS ; a = E o Ye = o = o = 5 2 Wg 5 7 as _ - E i W = > 2 i> = b = W's a r, 2 = Yar LIBRARIES | SMITHSONIAN INSTITUTION NOILOLILSNI_NVINOSHLINS S3/NWY S17 LI RA ‘G < = * = x = ae = ‘. yy Zz = = OY = = WW); = YS ww F 5 Gy, + 2 YG.2 ‘phe \N B ao JE a wn te 6A mee: = oS Vip Ut, =. re) SG fi ; ‘ = >" s >" = on =z wo a= Ww \ LALILSNI_ NVINOSHLIWS HYSIT LIBRARIES SMITHSONIAN INGTITIITION NiQnTAITIGCNE NYWINOQCUTIWNG cal uWwudia) ..cekn. LIBRARIES SMITHSONIAN INSTITUTION NOIiNS LIBRARIES SMITHSONIAN NOILNLILSNI LIBRARIES NOLLNILSNI NOLLALILSNI LIBRARIES il 2693 | | HN | | | ii INSTIT! 3 9088 012 A Nase OT ars eine eect Wetman aerate : Bias tiene teen rome" seveeainnme mosemnicetrans nae vase ee pte se Prensa vepaeteee : me pig anaenmysg ube Tie tenga eg tate TEP AE Seer nett ames “nad pene ee 5 nap EE Seaton Soca sateen apenas We SSO astra SESS ITE apne mages PE wena pol gh as ASSEN BORD pane ts ON eSAATAOd 7 5 “ OE ID AT RII nee ee pie ligh noes Sart Ct arpa Rser PACE SPER. atin en BES ad arava . ae eefi neni ean Fav ew Seat aan yee Pod Perepmnerey Sa pelke STN EE OE NT roars * : UN oie 68s aoraciaein tat perrer trea) ee ops nea Gge was tp VO ieoeaton ale” Aeteseneraie SE e mics Weave equ ved aN ohh dts emit Dette BY |