: 5 q SIT ee ae’ ES Be Nei es Rete ke 2 a Tne e- } J \reda } 4 yeahh wins ivan) Lr vatlenge sagt Want Aye ae ja Nay ace aOR a they dv yn ey m Hyead ay ites Mo aA nee Bor Viet nse eek RA Ae Whee Faninsd obs se Wt eu ie 4h SPA RO tahiti H eine Pes Ph a Paya Marky ys ei th Re ay Sa Stake fat ah AL cy " i Ista ernst ‘| ie NWN ates RSPCA DURES TR AS a Grey rie cA ay in graen eens eta ke teat tea bae ti aa Hed nyt Loaves his Ml Sota r i Vey dey ta whe ce afte ey StS Ate beg eee Narita di “int EE hs a Biter \ Pale tity Tt ae ae Tire yialaty hyve then tg tis, eat By fa! oe ye are Me ona) Car eC ee ne i wt i ayeyu deny tala ead stn itt seat 6 Pope neat et COM chee aN DDC NOIRE Wehgiynd eM anki Pete ME t Hee icy ‘ THlese tak tah OMEN pekeList SepH ey user cult yt ptgecebiby ( ete eee aera ; , wk Porienac i atc COMME Me phe cy 4 wceaey Bie Me syniesypuy coke aves ern ‘as i oli fae Me : tal ie i tiie fale Oya onakt ih She 2 x hae BSN. ine Sih ats Madd Bld yt AY et ax ay POOP Ey Mare ce iazeyted f Heh AalgiCtuAee Wena ea Aeae es pied Pearcy) oy Sefer tacip t o58 {hema tere ea Preah are yea heh ig” Bi dete hut te tee Y fgn geen Qeek eee h Mp SPL ra Aare a yea penne Ieee ft pny r Y Libtesge ee Ate Nd Fata ey i raegeli ert on te Bea he wh bgt ofan rps ety Sealy La BODO ad Hae va Md Panetoty yaa gs sig ne ty haryintyr ets i Yobasetak es eatebgysy teen pg Yan ‘ Ryd ie Noey sd Bee, iF ate ety fiat ey te nenaee ty P aibakgtonctutetye Walt th tana oy hehe Nee clatu tian suses pcarhties tes era ae me ae Se (ond 9 it deat SAN mae ena oa Rte has where BO ma venta fn St Ean Mea tas Phyl b tayo dietatuntll Soe aoe PLT a 4d =H ty anast a COL a ear ae aT. ia eet Cindy tata ert E Ja teiyina wineantrach tebe al ahah MAORI Allaah . i » teas rf NAN, S 4 wih: Ryo ie “. t i i ‘ , ‘ AP i ‘ ij i t , \ t ‘ ‘ u 1 F { “) t aI 5 i ; f ‘ + a MA ¢ i} / } t i H a) : 4 1 2 t i { fl oe ‘ } \ ( \ ZOOLOGICAL SCIENCE An International Journal VOLUME 11 1994 published by The Zoological Society of Japan CONTENTS OF VOLUME 11 NUMBER 1, FEBRUARY 1994 REVIEWS Hanke, W., W. Kloas: Hormonal regulation of osmo- mineral content in amphibia Nassel, D. R., E. Bayraktaroglu, H. Dircksen: Neuropeptides in neurosecretory and efferent neural systems of insect thoracic and abdominal ganglia ....15 ORIGINAL PAPERS Phisiology Takahashi, T., O. Matsushima, F. Morishita, M. Fujimo- to, T. Ikeda, H. Minakata, K. Nomoto: A myomodu- lin-CARP-related peptide isolated from a polychaete annelid Perinereis vancaurica Sugimoto, M., T. Kawamura, R. Fujii: Changes in the responsiveness of melanophores to electrical nervous stimulation after prolonged background adaptation in the medaka, Oryzias latipes .................. see eeee 39 Wilder, M. N., T. Okumura, Y. Suzuki, N. Fusetani, K. Aida: Vitellogenin production induced by eyestalk ablaton in juvenile giant freshwater prawn Macro- brachium rosenbergii and trial methyl farnesoate admi- nistration Cell and Molecular Biology Meyer-Rochow V. B., Y. Ishihara, J. R. Ingram: Cytochemical and Histological details of muscle fibers in the southern smelt Retropinna retropinna (Pisces; Galaxioidei) Genetics Patil, J. G., V. Wong, H. W. Khoo: Assessment of pMTL construct for detection in vivo of luciferase expression and fate of the transgene in the zebrafish, IBTAGIYAGNIONNCHION, saj540 secs sn oe2 ono one sere se s+ ee 63 Developmental Biology Furukawa, T., Y. Maeda: K252a, a potent inhibitor of protein kinases, promotes the transition of Dictyoste- lium cells from growth to differentiation Iwamatsu, T., S. Nakashima, K. Onitake, A. Matsuhisa, Y. Nagahama: Regional differences in granulosa cells of preovulatory medaka follicles Kanno, Y., S. Koike, T. Noumura: Immunohistoche- mical localizations of epidermal growth factor in the developing rat, ZONaAdS MRI II Sw ecoreadcoraete cpereinncre aio eierere « 83 Satoh, Y., T. Shimizu, Y. Sendai, H. Kinoh, N. Suzuki: Nucleotide sequence of the proton ATPase beta- subunit homologue of the sea urchin Hemicentrotus pulcherrimus (RAPID COMMUNICATION) _..... 153 Kanbayashi, H., Y. Fujita, K. Yamasu, T. Suyemitsu. K. Ishihara: Local change of an exogastrula-inducing peptide (EGIP) in the pluteus larva of the sea urchin Anthocidaris crassispina (RAPID COMMUNICA- LON) pitas) erlhe sian nents didinte y sores ee eed 157 Reproductive Biology Hosokawa, K., Y. D. Noda: The acrosome reaction and fertilization in the bivalve, Laternula limicola, in refern- ce to sperm penetration from the posterior region of the mid-piece Endocrinology Iga, C., I. Koshimizu, S. Takahashi, Y. Kobayashi: Experimental manipulation of pituitary hemorrhage induced by intraperitoneal injection of a hypertonic solutionvin’mice: . },.cesmetees th 2.0. ocseia a emtes 101 Yamashita, K., S. Kikuyama: Immunohistochemical study of ontogeny of pituitary prolactin and growth hormone cells in Xenopus laevis (RAPID COM- MUNICATION) &...2ctesadate hh hen. AE. RAs de 149 Asahina, M., H. Fugo, S. Takeda: Ecdysteroid syn- thesis in dissociated cells of the prothoracic gland of the silkworm, Bombyx mori ..........0.0-eeeceeee eee 107 Behavior Biology Naruse, M., T. Oishi: Effects of light and food as zeitgebers on locomotor activity rhythms in the loach, Misgurnus anguillicaudatus ................+-++++++- 113 Suzuki, H., T. Sekiguchi, A. Yamada, A. Mizukami: Sensory preconditioning in the terrestrial mollusk, Limax flavus Environmental Biology and Ecology Ohdachi, S.: Growth, metamorphosis and gape-limited cannibalism and predation on tadpoles in larvae of salamanders Hynobius retardatus .................. 127 Lawrence J. M., M. Bryne: Allocation of resources to body components in Heliocidaris erythrogramma and Heliocidaris tuberculata (Echinodermata: Echinoidea) Systematics and Taxonomy Ishikawa, K.: Two new species of the genus Holaspulus (Acarina: Gamasida: Parholaspidae) from the Ryukyu Islands Wapanite Muerte Rae eee 139 Experimental Animal Kiguchi, K., M. Shimoda: The sweet potato hornworm, Agrius convolvuli, as a new experimental insect: Con- tinuous rearing using artificial diets ................ 143 il Announcements Instructions to Authors NUMBER 2, APRIL 1994 REVIEWS Ward, A., P. Bierke, E. Pettersson, W. Engstrém: In- sulin-like growth factors: Growth, transgenes and im- PHINGING B24. UE. RR I eee dot 167 Mizunami, M: Processing of contrast signals in the insect ocellarisystem). «4.5 eadage sits cerrcia seis siesta sleteecteiacs RS 175 ORIGINAL PAPERS Physiology Aonuma, H., T. Nagayama, M. Hisada: Output effect of identified interneurons upon the abdominal postural system in the crayfish Procambarus clarkii (Girard) 191 Immunology Hirose, E., T. Ishii, Y. Saito, Y. Taneda: Phagocytic activity of tunic cells in the colonial ascidian Aplidum yamazii (Polyclinidae, Aplousobranchia) ........... 203 Biochemistry Harumi, T., K. Hoshino, N. Suzuki: In vitro autophos- phorylation and cyclic nucleotide-dependent dephos- phorylation of sea urchin sperm histone kinase .... 209 Mukai, M., T. Kondo, K. Yoshizato: Rapid and quan- titative detection of aspartic proteinase in animal tissues by radio-labeled pepstatin A ....................... 221 Furukohri, T., S.Okamoto, T. Suzuki: Evolution of phosphagen kinase (III). Amino acid sequence of argi- nine kinase from the shrimp Penaeus japonicus .... 229 Developmental Biology Furuya, H., K. Tsuneki, Y. Koshida: The development of the vermiform embryos of two mesozoans, Dicyema acuticephalum and Dicyema japonicum ............ 235 Kimura, K., K. Usui, T. Tanimura: Female myoblasts can participate in the formation of a male-specific muscle in Drosophilam etre. erate ee 247 Yazaki, I., H. Harashima: Induction of metamorphosis in the sea urchin, Pseudocentrotus depressus, using L-glutamine; ....0.)...caeeeee eee en ee Ch eee 253 Ohya, Y., K. Watanabe: Control of growth and dif- ferentiation of chondrogenic fibroblasts in soft-agar culture: Role of basic fibroblast growth factor and transforming growth factor-B .................0000ee 261 Reproductive Biology Okia, N. O.: Membrane-bound inclusions in the leydig cell cytoplasm of the broad-headed skink, Eumeces laticeps: (Lacertilia:, Scincidae)) |. . . 201.2. . esses 269 Yoshizaki, N.: Identification and localization of a ligand molecule of Xenopus cortical granule lectins ....... 275 Nakamura, M., T. Yamanobe, M. Takase: Localization and purification of serum albumin in the testis of XENOPUS LAEVIS. oo oe on ne cenccico ess siete 285 Endocrinology Ohta, N., T. Mori, S. Kawashima, S. Sakamoto, H. Kobayashi: Spatio-temporal pattern of DNA syn- thesis detected by bromodeoxyuridine labeling in the mouse endometrial stroma during decidualization .. 291 Tsai, P. I., S. S. Madsen, S. D. McCormick, H. A. Bern: Endocrine control of cartilage growth in coho salmon: GH influence in vivo on the response to IGF-I in vitro wie doa die aid wenn e wien ciclo «nese 299 Environmental Biology Takaku, G., H. Katakura, N. Yoshida: Mesostigmatic mites (Acari) associated with ground, burying, roving carrion and dung beetles (Coleoptera) in Sapporo and Tomakomai, Hokkaido, northern Japan ........... 305 Systematics and Taxonomy Amemiya, S., Y. Mizuno, S. Ohta: First fossil record of the family Phormosomatidae (Echinothurioida: Echi- noidea) from the Early Miocene Morozaki Group, central Japan s...... hee ne sig o's ele 313 Grismer, L. L., H. Ota, S. Tanaka: Phylogeny, clas- sification, and biogeography of Goniurosaurus kuroiwae (Squamata: Eublepharidae) from the Ryukyu Archipelago, Japan, with description of a new subspe- CIES. sae eee ceedene cee cess ce ress + 319 Ohtani, H.: Polymorphism of lampbrush chromosomes in Japanese populations of Rana nigromaculata .... 337 Matsuoka, N., K.Fukuda, K. Yoshida, M. Sugawara, M. Inamori: Biochemical systematics of five asteroids of the family Asteriidae based on allozyme variation 2eHHu A. od STi) 08. neha se ee 343 NUMBER 3, JUNE 1994 REVIEWS Tsutsui, K.,S. Kawashima: Regulation of gonadotropin receptors and its physiological significance in higher vertebrates: <...... Mavemien bs. Senne cits cle lsrotehe teres 351 Rastogi, R. K., L. Iela: Gonadotropin-releasing hor- mone: present concepts, future directions .......... 363 ORIGINAL PAPERS Physiology Nakamura, M., M. Tani, T. Kuramoto: Effects of rapid cooling on heart rate of the Japanese lobster in vivo Jas Viewsatradd. AUIS. det. atl Be see ee 375 Naitoh, T., M. Matuura, R. J. Wassersug: Effectiveness of metoclopramide, domperidone and ondansetron as antiemetics in the amphibian, Xenopus laevis ncn ATLA hath ALi bitch CER PAs Sis fae rae ee ae a 381 Sata, O., T. Sato: Electrical responses of non-taste cells in frog tongue and palate to chemical stimuli ...... 385 Cell Biology Arikawa K., A. Matsushita: Immunogold colocalization of opsin and actin in Drosophila photoreceptors that undergo active rhabdomere morphogenesis ........ 391 Ricci, N., F. Verni: Experimental perturbations of the Litonotus-Euplotes predator-prey system ........... 399 Biochemistry Ohtsuka, Y., H. Nakae, H. Abe, T. Obinata: Im- munochemical studies of an actin-binding protein in ascidian body wall smooth muscle ................. 407 Takikawa, S.,M. Nakagoshi: Developmental changes in pteridine biosynthesis in the toad, Bufo vulgaris EE Ne ed EO eee ees 413 Developmental Biology Hou, L., T. Takeuchi: Neural crest development in reptilian embryos, studied with monoclonal antibody, IBINT Kell: | ba oadetoccno Gi wonUiAd nacncd no sen oe eee eens 423 Suzuki, H., A. Kondo: The second maturation division and fertilization in the spider Achaearanea japonica (BOSMOS ELS) ATE LF, WS MSP BE SS 433 Nishiyama, I., T. Oota, M. Ogiso: Neuron-like mor- ill phology expressed by perinatal rat C-cells in vitro Endocrinology Uesaka, T., K. Yano, M. Yamasaki, M. Ando: Gluta- mate substitution for glutamine at position 5 or 6 reduces somatostatin action in the eel intestine (RAPID COMMUNICARION) Heis.0o. (eee eee Ee ae: 491 Takahashi, S., S. Oomizu, Y. Kobayashi: Proliferation of pituitary cells in streptozotocin-induced diabetic mice: effect of insulin and estrogen ................ 445 Takano, M., Y. Sasayama, Y. Takei: Molecular evolu- tion of shark C-type natriuretic peptides ........... 451 Systematics and Taxonomy Aizawa, T., M. Hatsumi, K. Wakahama: Systematic study on the Chaenogobius species (family Gobiidae) by analysis of allozyme polymorphisms ............ 455 Ohtani, H.: Speciation of Japanese pond frogs deduced from lampbrush chromosomes of their diploid and triploid! hybridsdeses . 3). PRs eee oh Soci te 465 Ono, T., Y. Obara: Karyotypes and Ag-NOR variations in Japanese Vespertilionid bats (Mammalia: Chirop- LET Aa) Wamtever parce mteyansr ct tte try Bateane on eve opeloneeveutancaveemnens 473 Matsui, M.,G. Wu: Acoustic characteristics of treefrogs from Sichuan, China, with comments on systematic relationship of Polypedates and Rhacophorus (Anura, NUMBER 4, AUGUST 1994 REVIEWS Larriva-sahd, J., A. Matsumoto: The vomeronasal sys- tem and its connections with sexually dimorphic neural SENUICCUTES ea tTAV TN A oth Harcteh sere tne 8s, oat oierste urd EN 495 Hosoya, H.: Cell-cycle-dependent regulation of myosin iehtichambkanasesy 22-6 el. cise salon. Weaeeeeeee 507 ORIGINAL PAPERS Physiology Shibayama, R., T. Kobayashi, H. Wada, H. Ushitani, J. Inoue, T. Kawakami, H. Sugi: Stiffness changes of holothurian dermis induced by mechanical vibration REDE Ere SM AOR NIE FA SRP oO EEL 2 511 Cell and Molecular Biology Kosaka, T.: Life cycle of Paramecium bursaria syngen 1 in a natural pond Goda, M., J. Toyohara, M. A. Visconti, N. Oshima, R. Fujii: The blue coloration of the common surgeonfish, Paracanthurus hepatus—I. Morphological features of chromatophores! “2268.24. Riek eee nade seven ee 527 Harigaya, T., S. Tsunoda, M. Mizuno, H. Nagasawa: Different gene expression of mouse transforming growth factor a between pregnant mammary glands and Rhacophondac)/ eee ceec Lecce eee eee oy ee 485 mammary tumors in C3H/He mice (RAPID COM- MUNICATION) war ease So trees. 8. Reet 625 Biochemistry Miura, K., M. Nakagawa, Y. Chinzei, T. Shinoda, E. Nagao, H. Numata: Structural and functional studies on biliverdin-associated cyanoprotein from the bean bug, Riptortus clavatus ............. 0c cence cece 537 Developmental Biology Matsuda, M., H. Keino: An open cephalic neural tube reproducibly induced by cytochalasin D in rat embryos LIDAVELTO Wate terse tote arses rene aor cmos reefer chat sie ke vans spatonetetacys lair 547 Endocrinology Shinobu, N., Y. Mugiya: Effects of hypophysectomy and replacement therapy with bovine growth hormone and triiodothyronine on the in vitro uptake of calcium and methionine by scales in the goldfish, Carassius aurautus Sawada, K., T. Noumura: Characterization of androgen receptors for testosterone and 5a-dihydrotestosterone in the mouse submandibular gland ................. 563 Takagi, K., S. Kawashima: Effects of sex steroids on dopamine neurons in cultured hypothalamus and preoptic area cells derived from neonatal rats ...... 571 Itoh, M. T., A. Hattori, Y. Sumi, T. Suzuki: Identifica- tion of melatonin in different organs of the cricket, Gryllus Dirmaculatus ...... cee cece cence eee eee eens S/7/ Wakahara, M., N. Miyashita, A. Sakamoto, T. Arai: Several biochemical alterations from larval to adult types are independent on morphological metamorpho- sis in a salamander, Hynobius retardatus ........... 583 Ecology Nishi, E., M. Nishihara: Colony formation via sexual and asexual reproduction in Salmacina dysteri (Huxley) Polychaetay Serpulidaciar: ae. cere eee ree 589 Phylogeny Masuda, R., M. C. Yoshida, F. Shinyashiki, G. Bando: Molecular phylogenetic status of the Iriomote cat Felis iriomotensis, inferred from mitochondrial DNA sequ- ence analysis. ...,.;P2SS Ae. 3. eee 597 Masuda, R., M. C. Yoshida: A molecular phylogeny of the family Mustelidae (Mammalia, Carnivora), based on comparison of mitochondrial cytochrome b nuc- leotide sequences) 252.222). isece sees eee 605 Fukatsu, T., S. Aoki, U. Kurosu, H. Ishikawa: Phy- logeny of Cerataphidini aphids revealed by their sym- biotic microorganisms and basic structure of their galls: implications for host-symbiont coevolution and evolu- tion of sterile soldier castes) ................2...2-5. 613 Errata ssscig. snes ols h 0S SS eee ee 629 NUMBER 5, OCTOBER 1994 REVIEWS O’Brien, M. A., P. H. Taghert: The genetic analysis of neuropeptide signaling systems ..................... 633 Morisawa, M.: Cell signaling mechanisms for sperm Motility”) 2s...ncaL ee. ci eae eee eee 647 ORIGINAL PAPERS Physiology Kobayashi, M., K. Kitayama, G. Satoh, K. Ishigaki, K. Imai: Relationships between the slope of the oxygen equilibrium curve and the cooperativity of hemoglobin as analyzed using a normalized oxygen pressure scale 0h A EEE, RE Se Ae 663 Karakisawa, H., S. Tamotsu, A. Terakita, K. Ohtsu: Identification of putative photoreceptor cells in the siphon of a clam, Ruditapes philippinarum ......... 667 Nakagawa, T., E. Eguchi: Differences in flicker fusion frequencies of the five spectral photoreceptor types in the swallowtail butterflyss compound eye (RAPID COMMUNICATION) haces. Ace eee eee 759 Cell Biology Yoshikawa, T., Y. Yashiro, T. Oishi, K. Kokame, Y. Fukada: Immunoreactivities to rhodopsin and rod/ cone transducin antisera in the retina, pineal complex and deep brain of the bullfrog, Rana catesbeiana FES eine ASR pn HORDES SR Re ee ee Re 675 Genetics Nakamura, M., M. Sumida, T. Yamanobe, M. Nishioka: Identification of protein C in sera of the frogs, Rana nigromaculata and Rana brevipoda (RAPID COM- MUNICATION) ie. e88..2h.. 2S ee oes 763 Immunology Ohtake, S., T. Abe, F. Shishikura, K. Tanaka: The phagocytes in hemolymph of Halocynthia roretzi and theirsphagocyticractivity ster -\.cheteteer ener eee te 681 Developmental Biology Suzuki, H., A. Kondo: Changes at the egg surface during the first maturation division in the spider Achaearanea japonica (Bos. et Str.) .............2.. 693 Reproductive Biology Mita, M., A. Oguchi, S. Kikuyama, R. de Santis, M. Nakamura: Ultrastructural study of endogenous ener- gy substrates in spermatozoa of the sea urchins Arbacia lixula and Paracentrotus lividus” ..................+- 701 seve The effect of glucocorticoids on the activity of monoamine oxidase and superoxide dismutase in the rat interscapular brown adipose tissue ................. 707 Suzuki, N., Y. Nosé Y. Kasé Y. Sasayama, Y. Takei, H. Nagasawa, T. X. Watanabe, K. Nakajima, S. Sakaki- bara: Amino acid sequence of sardine calcitonin and its hypocalcemic activity in rats .................... 713 Ge, W., R. E. Peter: Evidence for non-steroidal gonad- al regulator(s) of gonadotropin release in the goldfish, Carassius Quratus oo! \). vesasenwaed.. A ids se eee TAG Behavior Biology Nakagawa, A., A. Iwama, A. Mizukami: Age- dependent changes related to reproductive develop- ment in the odor preference of blowflies, Phormia regina, and fleshflies, Boetteherisca peregrina ...... 725 Hongoh, Y., H. Ishikawa: Changes of mycetocyte sym- biosis in response to flying behavior of alatiform aphid (Acyrthosiphon pisum) ...... 0.0.0 cee cece eee e cee nees 731 Morphology Hirose, E., T. Ishii, Y. Saito, Y. Taneda: Seven types of tunic cells in the colonial ascidian Aplidium yamazii (Polyclinidae, Aplousobranchia): morphology, classifi- cation, and possible functions ...................05- 737 Ando, K., N. Okura: Aminergic and acetyl- cholinesterase-positive innervation in the cerebral arte- rial system and choroid plexus of the newt Triturus pyrrhogaster, with special reference to the plexus in- nervation Taxonomy Tanaka, T., M. Matsui, O. Takenaka: Estimation of phylogenetic relationships among Japanese brown frogs from mitochondrial cytochrome b gene (Amphibia: TTIUOSEY eo SUOAAHS SBE Iao 0 Gets OURO OT One cent: Seams ae 753 NUMBER 6, DECEMBER 1994 REVIEWS Sato, T., T. Miyamoto, Y. Okada: Comparison of gustatory transduction mechanisms in vertebrate taste COUSMEE See ncs 5 ceiek adic nosis s be beara Laws 767 Baguna, J., E. Salo R. Romero, J. Garcia-Fernandez, D. Bueno, A. M. Munoz-Marmol, J. R. Bayascas- Ramirez, A. Casali: Regeneration and pattern forma- tion in planarians. Cells, molecules and genes _..... 781 ORIGINAL PAPERS Genetics Miura, I.: Sex chromosome differentiation in the Japanese brown frog, Rana japonica. 1. Sex-related heteromorphism in the distribution pattern of constitu- tive hetero-chromatin in chromosome No.4 of the Wakuya population Miura, I.: Sex chromosome differentiation in the Japanese brown frog, Rana japonica. Il. Sex-linkage analyses of the nucleolar organizer regions in chromo- some No. 4 of the Hiroshima and Saeki populations 2.900 ob oUA aR GORDO EE eer ee eae atone rte OUT, Immunology Sawada, T., S. Ohtake: Mixed-incubation of allogeneic hemocytes in tunicate Halocynthia roretzi .......... 817 Biochemistry Maki, S., S. Kimura, K. Maruyama: Localization of connectin-like proteins in the giant sarcomeres of barmaclesmuscle) s5..ccc desea sews sass cevecece es 821 Developmental Biology Iwamatsu, T.: Stages of normal development in the MEG QAvAIOS DEAS Yoasasasososuessacoonoosoene 825 Harigaya, T., S. Tsunoda, H. Yokoyama, K. Yamamoto, H. Nagasawa: Mouse TGFa gene expression in nor- mal and neoplastic mammary glands and uteri of four strains of mice with different potentials for mammary gland growth and uterine adenomyosis ............. 841 Murakami, R., K. Miyake, I. Yamaoka: Androgen- induced differentiation of the fibrocartilage of os penis CUCU CI BV IE Olen eerste aA ne elena shee oe Seals 847 Reproductive Biology Harada, T.: Adult diapause induced by the loss of water surface in the water strider, Aquarius paludum (Fabri- CHUS) Wg menisci nee Acie tra on aM ORR eM Seats 855 Asahina, K., J. G. D. Lambert, H. J. Th. Goos: Bioconversion of 17a-hydroxyprogesterone into 17a, 20a-dihydroxy-4-pregnen-3-one and 17a, 20f- dihydroxy-4-pregnen-3-one by flounder (Platichthys HSS) GOSTTENOAOA ceo v0c0c0s0c00000000000000000000 859 Endocrinology Horiuchi, T., M. Isobe, M. Suzuki, Y. Kobayashi: Cal- citonin induces hypertrophy and proliferation of pars intermedia cells of the rat pituitary gland .......... 865 Environmental Biology and Ecology Prevot, S., J. Senaud, J. Bohatier, G. Prensier: Varia- tion in the composition of the ruminal bacterial mi- croflora during the adaptation phase in an artificial fenmemtior CROSIMNEC) cococcscocococcccubovuncucce 871 Peng, K. W., Y. K. Ip,: Is the coelomic plasma of Phascolosoma arcuatum (Sipuncula) hyperosmotic and hypoionic in chloride to the external environment? Saulich, A. H., T. A. Volkovich, H. Numata: Control of seasonal development by photoperiod and tempera- ture in the linden bug, Pyrrhocoris apterus in Belgorod, Russia Systematics and Taxonomy Katakura, H., S. Saitoh, K. Nakamura, I. Abbas: Multi- variate analyses of elytral spot patterns in the phytopha- gous ladybird beetle Epilachna vigintioctopunctata (Coleoptera, Coccinellidae) in the province of Sumatera Barat, Indonesia .......................-- 889 RUG OT 1M OG Xe erates caresses aces RSP RPO EN ee re aes ccafolot 895 Ncknowledpmentyy) antniseeer cen serene teresa 899 Contents of ZOOLOGICAL SCIENCE, Vol. 11, Nos. 1-6 D '/ t Published Bimonthly by the Japanese Society of eve opm en Developmental Biologists Distributed by Business Center for Academic Growth & Differentiation Societies Japan, Academic Press, Inc. Papers in Vol. 36, No. 6. (December 1994) 55. T. Yagi: Src Family Kinases Control Neural Development and Function 56. S.H. Keller and V. D. Vacquier: N-Linked Oligosaccharides of Sea Urchin Egg Jelly Induce the Sperm Acrosome Reaction 57. S. K. Satoh, M. T. Oka and Y. Hamaguchi: Asymmetry in the Mitotic Spindle Induced by the Attachment to the Cell Surface during Maturation in the Starfish Oocyte 58. P. Hardman, E. Landels, A. S. Woolf, and B. S. Spooner: TGF-1 Inhibits Growth and Branching Morphogenesis in Embryonic Mouse Submandibular and Sublingual Glands in Vitro 59. A. Obinata, Y. Akimoto, T. Kawamata, and H. Hirano: Induction of Mucous Metaplasia in Chick Embryonic Skin by Retinol-Pretreated Embryonic Chick or Quail Dermal Fibroblasts through Cell-Cell Interaction: Correlation of a Transient Increase in Retinoic Acid Receptor 8 mRNA in Retinol-Treated Dermal Fibroblasts with Their Competence to Induce Epidermal Mucous Metaplasia 60. M. Yoshida, K. Inaba, K. Ishida and M. Morisawa: Calcium and Cyclic AMP Mediate Sperm Activation, but Ca** Alone Contributes Sperm Chemotaxis in the Ascidian, Ciona savignyi 61. T. A. Ferguson, J. Vozenilek and C. M. West: The Differentiation of a Cell Sorting Mutant of Dictyostelium discoideum 62. M. Asaoka, M. Myohara and M. Okada: Digitonin Activates Different Sets of Puff Loci Depending on Developmental Stages in Drosophila melanogaster Salivary Glands 63. T. A. Quill and J. L. Hedrick: Oviductal Localization of the Cortical Granule Lectin Ligand Involved in the Block to Polyspermy of Xenopus Laevis 64. V. Gangji, E. Bastianelli, M. Rooze, and R. Pochet: Transient Calretinin Expression during Interverteb- ral Disc Formation of the Chick Embryo 65. S. Kobayashi, H. Saito and M. Okada: A Simplified and Efficient Method for in situ Hybridization to Whole Drosophila Embryos, Using Electrophoresis for Removing Non-hybridized Probes 66. K. Akasaka, N. Sakamoto, T. Yamamoto, J. Morokuma, N. Fujikawa, K. Takata, S. Eguchi and H. Shimada: Corrected Structure of the 5’ Flanking Region of Arylsulfatase Gene of the Sea Urchin, Hemicentrotus pulcherrimus Development, Growth and Differentiation (ISSN 0012-1592) is published bimonthly by The Japanese Society of Developmental Biologists. Annual subscription for Vol. 35 1993 U.S. $ 191,00, U.S. and Canada; U. S. $ 211,00, all other countries except Japan. A\ll prices include postage, handling and air speed delivery except Japan. Second class postage paid at Jamaica, N.Y. 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 32887-4900, 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 32887-4900, U.S. A. In Japan: Send nonmember subscription orders and notices of change of address to Business Center for Academic Societies Japan, 16-9, Honkomagome 5-chome, Bunkyo-ku, Tokyo 113, Japan. Send inquiries about membership to Business Center for Academic Societies Japan, 16-9, Honkomagome 5-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. ISSN 0289-0003 An_ International Journal OOLOGICAL oo EINGE ZOOLOGICAL SCIENCE The Official Journal of the Zoological Society of Japan Editors-in-Chief: The Zoological Society of Japan: Seiichiro Kawashima (Tokyo) Toshin-building, Hongo 2-27-2, Bunkyo-ku, Tsuneo Yamaguchi (Okayama) Tokyo 113, Japan. Phone 03-3814-5461 Division Editor: Fax 03-3814-5352 Shunsuke Mawatari (Sapporo) Oiicon: Yoshitaka Nagahama (Okazaki) President: Hideo Mohri (Chiba) Takashi Obinata (Chiba) Secretary: Takao Mori (Tokyo) Suguru Ohta (Tokyo) Treasurer: Makoto Okuno (Tokyo) Neils sation (Kyoto) Librarian: Masatsune Takeda (Tokyo) Assistant Editors: Auditors: Hideshi Kobayashi (Tokyo) Akiyoshi Niida (Okayama) Hiromichi Morita (Fukuoka) Masaki Sakai (Okayama) Sumio Takahashi (Okayama) Editorial Board: Kiyoshi Aoki (Tokyo) Makoto Asashima (Tokyo) Howard A. Bern (Berkeley) Walter Bock (New York) Yoshihiko Chiba (Yamaguchi) Aubrey Gorbman (Seattle) Horst Gurnz (Essen) Robert B. Hill (Kingston) Yukio Hiramoto (Chiba) Tetsuya Hirano (Tokyo) Motonori Hoshi (Tokyo) Susumu Ishii (Tokyo) Hajime Ishikawa (Tokyo) Sakae Kikuyama (Tokyo) Makoto Kobayashi (Higashi-Hiroshima) Kiyoaki Kuwasawa (Tokyo) John M. Lawrence (Tampa) Koscak Maruyama (Chiba) Roger Milkman (Iowa) Kazuo Moriwaki (Mishima) Richard S. Nishioka (Berkeley) Chitaru Oguro (Toyama) Masukichi Okada (Tsukuba) Andreas Oksche (Giesen) Hiraku Shimada (Higashi-Hiroshima) Yoshihisa Shirayama Tokyo) Takuji Takeuchi (Sendai) Ryouzo Yamagimachi (Honolulu) ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and rapid communications in the broad field of Zoology. The journal was founded in 1984 as a result of unification of Zoological Magazine (1988-1983) and Annotationes Zoologicae Jasponenses (1897-1983), the former official journals of the Zoological Soceity of Japan. An annual volume consists of six regular numbers and one supplement (abstracts of papers presented at the annual meeting of the Zoological Society of Japan) of more than 850 pages. The regular numbers appear bimonthly. MANUSCRIPTS OFFERED FOR CONSIDERATION AND CORRESPONDENCE CONCERNING EDITORIAL MATTERS should be sent to: Dr. Tsuneo Yamaguchi, Managing Editor, Zoological Science, Department of Biology, Faculty of Science, Okayama University, Okayama 700, 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, 5—16—9 Honkomagome, Bunkyo-ku, Tokyo 113. Subscriptions outside Japan should be ordered from the sole agent, VSP, Godfried van Seystlaan 47, 3703 BR 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. © Copyright 1994 by the Zoological Society of Japan. In the U.S.A., authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by [copyright owner’s name], provided that designated fees are paid directory to Copyright Clearance Center. For those organizations that have been granted a photocopy license by CCC, a separate system of payment has been arranged. Copyright Clearance Center, Inc. 27 Congress St., Salem, MA, U.S.A. (Phone 508-744— 3350; Fax 508-741-2318). Publication of Zoological Science has been supported in part by a Grant-in-Aid for Publication of Scientific Research Results from the Ministry of Education, Science and Culture, Japan. ZOOLOGICAL SCIENCE 11: 1-3 (1994) © 1994 Zoological Society of Japan OBITUARY Masao Sugiyama (1908-1993) It is very sad for us to have to announce that Masao Sugiyama, Professor Emeritas of Nagoya University, passed away at the age of 84 on the 14th of September 1993, following nearly 5 years of illness. His death has deeply grieved both those who knew him and those who worked with him. His death is a great loss to the field of biological science both in Japan and abroad. Masao Sugiyama was born on the 11th of November 1908 in Nagoya City in central Honshu. As a junior high school student, he first encountered a sea urchin shell on the sea shore near Nagoya and was deeply impressed by its beautiful, elaborately carved patterns. At that moment, although he did not suspect it, his long-life connection with the sea urchin might be said to have begun. In 1925, he entered the Eighth High School, Science Course, in Nagoya. Inspired by his high school biology teachers, the late Professors Usaburo Kohno and Noboru Takamine, Sugiyama took an interest in Zoology and was determined to become a Zoologist. After graduating in 1928, he entered the Zoological Institute of the Faculty of Science, the Imperial University of Tokyo (now the University of Tokyo). He became a student of the late Professor Naohide Yatsu and studied the spermatogenesis of Asellus nipponensis, a kind of isopod, crustacean, for his graduation thesis. In 1931, he was enrolled in the Graduate School of the Imperial University of Tokyo and proceeded with studies on the behavior of the sex chromosomes in the spermatogenesis of Japanese earwig Anisolabis marginalis. Thus, during the early period of his post-graduate course, he wrote the two papers concerned with spermatogenesis mentioned above. Soon after, however, he became aware of the importance of Experimental Cytology and began to study the effects of mitogenetic rays on cell division, under the direction of the late Professor Tokusuke Goda. At that time, whether or not mitogenetic rays could induce cell division was an exciting problem which was paid much attention by many scientists with several hundred papers published on the subject. Despite his efforts, Sugiyama could not obtain positive results on induction of cell division, using yeast, root-tip cells and sea urchin eggs. He thus reached the conclusion that mitogenetic rays are not emitted from dividing cells. In 1938, he published three papers denying the presence of such rays. Since the 1940’s, the number of studies concerned with mitogenetic rays has rapidly decreased, with even the term “mitogenetic rays” completely disappearing from recent text-books on Cytology. This means that his conclusion was certainly correct. Then, following a suggestion by Professor Goda, he started to investigate the effects of heparin and hirudine, anticoagulants, on cleavage in sea urchin eggs and found the very interesting fact that nuclear divisions occur without cytoplasmic segmentations, so that multi-nucleated eggs are induced. In 1936, he finished his post-graduate course and for these excellent studies he was awarded the degree of Dr. Sc. from Imperial University of M. SuGIYAMA Tokyo in 1938. On the other hand, he took an interest in the behavior of the cell surface during cleavage in sea urchin eggs. Professor Katsuma Dan (former President of Tokyo Metropolitan University) had the same idea. Therefore, Sugiyama cooperated with Professor Dan and Professor Tomosuke Yanagita in such studies. In 1938, they attained their objective using a very simple but very skilfull technique, namely the measurement of the distance between two kaoline particles attached to the egg surface. As a result of such series of experiments, Prof. Dan proposed his famous hypothesis on cell division in sea urchin eggs in 1943. In 1938, too, he wrote two papers concerned with the fertilization and formation of fertilization membrane in sea urchin eggs. His full-fledged research on fertilization or artificial activation seems to have begun around this period and marks the start of his successful scientific career. It should be noted here that he published a book “Physiological Fertilization” in 1940. In 1939, he became a staff member of the Nagoya Medical College and was then appointed associate professor at Nagoya Imperial University. In those days, there was no Marine Biological Station in central Honshu. Therefore, he planned and made great effort to found a Marine Biological Station available to all for research as well as education in marine biology. In 1942, a new building for such a Station was completed on Sugashima, a small island near Toba City, Mie-Prefecture, a place very rich in fauna. He was nominated Director of the Sugashima Marine Biological Station in 1943 and subsequently served as Director for nearly 30 years. During this period, he devoted his efforts to fitting out the station for modern scientific research and he succeeded in the introduction of both transmission and scanning electron microscopes and facilities for radio-isotopes. In 1962, he was promoted to full professor at the same university and retired in 1972. He successively served as professor of biology at Sugiyama-jogakuen University until 1984. The following is an abbreviated resume of Prof. Sugiyama’s most significant accomplishments since moving to Nagoya University in 1939. Once sea urchin eggs are fertilized by a single spermatozoon, they do not accept other spermatozoa if insemination is repeated. This is the case even when fertilization membrane, formed at the time of the first insemination, is mechanically removed. This fact means that some changes blocking extra sperm-entrance into eggs occur at the time of fertilization; in other words, re-fertilization does not take place. In 1952, however, Sugiyama found that when eggs are washed in Ca-, Mg-free sea water just after insemination and the fertilization membrane is removed, those eggs can accept re-inseminated spermatozoa. This finding clearly demonstrated that a sperm-block mechanism is destroyed under such Ca-, Mg-free circumstances. On the other hand, since 1947, he had done a series of experiments on artificial activation in sea urchin eggs. And, by introducing a very simple but skilfull method, he succeeded in analyzing a mechanism of egg activation following formation of fertilization membrane. From the results obtained by such experiments, he divided cortical changes, provoked by insemination and various activating reagents, in nature into two groups; that is, a propagating and non-propagating one. He called the former propagating cortical change a fertilization wave, according to the definition of activation in Medaka eggs reported by the late Professor Tokio Yamamoto (Nagoya University). Thus, Sugiyama proposed a working hypothesis, the so called “Fertilization Wave Hypothesis”. He supposed that, in the process of fertilization or activation, a chain reaction on the cell surface is involved in which a spermatozoon or an activating reagent triggers stimulation of the egg surface and this stimulation in turn induces the propagating cortical change, fertilization wave, with this wave provoking the formation of fertilization membrane via the breakdown of cortical granules. Later, in 1969, this scheme was improved by the addition of new findings such as insufficient activation without breakdown of cortical granules, published by himself and his co-workers in 1967 and 1969. In 1974 and 1975, Mazia and his collaborators reported that ammonia can induce activation without cortical granule breakdown in sea urchin eggs. It should be noted here that as mentioned above, Sugiyama and his group had already published papers on such phenomena “insufficient activation” nearly 5 years before. In recognition of his pioneering work, he was invited to participate in the “International Symposium of Fertilization in the Sea Urchin Eggs” held at Palermo in Italy in 1955. At the symposium, he played a very important role as chairman as well as a speaker. He read a paper on his “Fertilization Wave Hypothesis” that made a very strong impression on fellow delegates. For this contribution, he was awarded the Zoological Society of Japan Prize in 1957. In 1973, he and his co-worker, the late Mr. Yoshiteru Takahashi, clearly demonstrated the importance of acrosome reaction at the time of fertilization in sea urchin eggs. They found that spermatozoa in which the acrosome reaction has once been induced are fertilizable even in Ca-free sea water. This means that Ca ions are not indispensable for membrane fusion between egg and spermatozoon but only for induction of the acrosome reaction. In his later years following retirement from Nagoya University in 1974, his interests extended to grasping the process of fertilization as one of the processes of cell membrane fusion and spent much time on such studies. In this manner, he continued his efforts as an investigator until the end of his active life. Sugiyama contributed much to the Zoological Society of Japan as an active member of the Board from 1963 to 1972 and from 1975 to 1979. He was also on the Board of the Japan Society for Cell Biology from 1965 to 1974. He was OBITUARY President of the Japanese Society of Developmental Biologists from 1973 to 1976 and, further, he served as Editor-in-Chief of Development, Growth and Differentiation from 1968 to 1973. In 1978, he was elected as a member of the Science Council of Japan and was an active participant until 1984. In addition, he was also on the Board of the National Institute of Basic Biology in Okazaki City. He also served as chairman of the Council of Japanese National Marine and Inland Biological Stations for many years and made a great effort to improve mutual communication and cooperation between stations and to expand their facilities. Thus, Sugiyama’s contributions to the academic world went far indeed. He was warm, gentle, quite and always generously prepared to listen to his students. We never heard him raise his voice. He was not a narrow-minded scientist. He loved literature, art and music. He was especially fond of works by Beethoven. He was an accomplished pianist. Occasionally, he kindly played a part of the first movement of Piano Sonata No.14 op. 27-2 (the so called Moon Light) for us. He also loved taking photographs and 16 mm motion pictures. For some time, his poetic feeling led him to write Japanese verse with thirty one-syllables (Tanka). Here is one of them which was composed by him on the occasion of his retirement. Fumikoeshi yamano kanatani miyuru mine Sarani noboramu chikarano kagiri “A new peak appears beyond the mountain I have just climbed However, Ill try to conquer this peak too as best as I can” This verse vividly highlights Sugiyama’s continuous devotion to, and passion for, scientific research which never ceased even after his official retirement. MANABU K. Koyima Department of Environmental Biology and Chemistry Faculty of Science Toyama University | iad oe ea hove wl oodtwl bon OTE of ECOL angah it tensioned 6 OAR wh disveles ew oaoe Ck pera tos (i) day, veakh od) ahs liv cy tl @nistbye ti ARE : iusnaltn lt nasiancciel Ace dboeniteal) ol We “ie ern a aa ane Osa Tomiie landagtan et ess eacorege, ‘ eies ctl. iio ain dia ae oceania ste bung ¢deeomje® awe Ob) 4 cite: bitig pedal nil may OE ae eo ceoy ataotgetay sitesi pet dlanpinasitl wineiq tadlalgencane (snes ; if i Ath pela not gribhss taperet enlieaiy, a fn Lf iirc Ve stinbbopenae cehoied) Crave nee scene es tie ets Shnine Biolog uitaraniaay prises 4 a ; ‘ ripe c tell on lige hina. n nial we sou ive souk ose ibe aod Oded pipet sua stied lls effyy ia Tee hel b tera lek) Lunebgga keer momAt dine Ae ; Levelymeibw: tue maids Lamnindcien cut) bnotesdor 4] thes i wepeda Hear wise Tsteppira en ebellthices atk sity ti ay : ‘lc teem ire im ET ate | eet ey a? i TT. aa i - = —_ Rikw eorigaaes pel ey Y gee Den p00 onda eure Terie: 21 penta Pa: me 8 Oey oy Ueoinaaian gee ‘ Appts s yi ia pte Pe nan th i Ti , i 1 iw ewe Ving yin inset rt her weather ne fe Dik fe ivubivions be Ge Xu j ‘ f ’ 7 wrt »viluat b - » iio) ectivetiun new ite . | 4 Ligiehy OF Te ae tee — ) er iets, Tao cbeeniehs a j rash ioc eT gators. SNR! A a io : mew ee a mi er voce henge, Ti é “ > (Sogows Liieveniy 2 he Sappesed thats GT E see ~ a S vijt st ‘POSC ret) eee 4 oer oir vi t i wiles \ i —_ : i" fh) fe n wii Th, 6A ty Sianeli al be van i, Mee edie tay with iene F iA3 : | , yer ee hike “Fi int tthe Tegsoy Ln tai ae cael 4 wheats ke atroude =. we ae tot inde * ey aie 1CAC! ia, fin hy te - : ay TRS priente ae bey ihe to na ste re con lial sie 37 ¢ meeribedy @ the ican foie img i . 4) Bala Prem TRS bey 1974, 6 ZOOLOGICAL SCIENCE 11: 5-14 (1994) © 1994 Zoological Society of Japan REVIEW Hormonal Regulation of Osmomineral Content in Amphibia WILFRIED HANKE and WERNER KLOAS Department of Zoology, University of Karlsruhe, Kaiserstr. 12, D-76128 Karlsruhe, Germany INTRODUCTION Osmomineral regulation is among the most important physiological functions in Amphibia. These animals live in fresh water or depend strongly on water in the environment to prevent desiccation. Only one species, Rana cancrivora, is found in sea water where the frogs capture food and are able to survive for a longer time. The larval stages of Amphibia live generally in fresh water and develop during metamorphosis the capability to survive in air except those like the clawed toad, Xenopus laevis, which stay in fresh water for the whole life. Different types of adaptation are found. Most of the Ampibia living in fresh water, larvae or adults, are stenoha- lme animals. They are hyperregulators, but the osmotic concentration is mostly low compared with other vertebrates. It is equivalent to saline of 0.6% which is approximately 180 mOsm. These fresh water animals can be acclimated to 1.2% salt concentration which is mostly the limit for the hyperregulatory capacity, because the animals must adjust the internal osmolality about 10% above the external value to survive. Another type of adaptation is represented by eury- haline toads and Rana cancrivora. These animals can sur- vive in higher external NaCl concentration or under extreme dry conditions. They adjust the internal osmolality by in- creasing the urea concentration. Internal osmotic values of 800-1000 mOsm can be reached. Using this mechanisms, toads living under arid conditions avoid desiccation. The regulation of the processes which enable the animals to survive under osmotic stress is exerted by several hor- mones. Generally, all hormonal systems take part in osmoregulation, but several hormones are especially impor- tant. These are the hormones of the hypophysial- adrenocortical axis, the nonapeptides of the eurohypophy- sis, tissue derived peptide hormones like angiotensin II or natriuretic peptides, catecholamines from the adrenal medul- la, and regulators of energy metabolism, like thyroid hor- mones, insulin, glucagon, or of Ca‘* metabolism, e.g. prolactin, calcitriol (Table 1). In the same way, several organs are involved in osmoregulatory processes. Not only those which directly Received July 14, 1993 take up or release water and electrolytes, like skin, bladder, and kidney, but also organs for general physiological func- tions, e.g. liver, circulatory organs, muscles, intestine etc. These organs are important because of supply with metabo- lites and oxygen, for transport processes etc. (Table 2). This review mainly focusses on the mechanisms working in two representative of Anura and Urodela, the clawed toad and the axolotl. Both represent specific types of adaptation. Larvae and adults of Xenopus laevis live in water. The adults can survive on air only for about half a day. Axolotls are neotenic animals which means that they also stay in water for the whole life. This implies that these species may be especially adapted to the life in water and are not typical for Anura or Urodela. In this review special attention is given to two main questions. Firstly, the cooperation between the different hormones and the target organs is described. Secondly, the regulation of the release of corticosteroids from the interrenal tissue by osmoregulatory hormones is most extensively discussed. 1. Effects and significance of hormonal systems 1.1. Role of the hypothalamo-hypophysial system in osmo- regulation The hormones of the hypothalamo-hypophysial system involved in osmoregulation have two different origins. They derive from the pars nervosa or the adenohypophysis. In both cases the role of the hypothalamus is to regulate the release of the hypophysial hormones and to transmit informa- tion about the environment and the internal milieu to the gland-like structures. This results in changing the titre of the hormones which act on peripheral organs. It is well known for a long time that the nonapeptides of the pars nervosa control the water economy of the Amphibia. They stimulate water uptake by the skin or the bladder (Brunn effect). It is interesting that the effectiveness of these hormones depends on the adaptational type. Amphib- ia living in water do not respond as strong to nonapeptides as those living on land and are subject to changes of water supply or desiccation. On the other hand, it has been reported that those animals living in extremely dehydrating situations, e.g. Rana cancrivora or Scaphiopus couchii, while staying in the holes for aestivation do not respond to nonapeptides at this time. The effects of nonapeptides on 6 W. HANKE AND W. K1ioas TaBLE 1. Hormones effective in osmoregulation of Amphibia Gland Hormone Target Function Neurohypophysis Nonapeptides, Hydrins Skin, Bladder Water uptake Interrenals Steroid secretion Liver cells Glycogenolysis Adenohypophysis ACTH Interrenals Steroid secretion PRL Skin Ca-uptake Interrenal gland Corticosteroids Skin Na-uptake (Corticosterone, Aldosterone) Liver cells Gluconeogenesis Chromaffin (adrenal) tissue Catecholamines Liver cells Glycogenolysis (Adrenaline, Noradrenaline) Heart etc. ANF Kidney Natriuresis Liver, Kidney etc. Angiotensins Kidney Water retention Interrenals Steroid secretion Calcitriol Skin etc. Ca-uptake Thyroid gland T4, T3 different organs Metabolism TasBLE2. Mechanisms of osmotic adaptation in Amphibia Effectors Function Interrenal gland Na-uptake by the skin Corticosterone Glucose-production by the liver Electrolyte shift between organs and blood Aldosterone Chromaffin tissue Glycogenolysis Adrenaline Changes of circulation Noradrenaline Skin Water-uptake Na-uptake Ca-uptake Liver Glucose production Urea production Circulation Blood supply for skin, intestine etc. Bladder and kidney Water and electrolyte homeostasis the regulation of water content have been recently reviewed [2]. The water balance of Amphibia is also affected by special neuropeptides called hydrins [36] which are isolated from Xenopus and Rana neurointermediate lobes. The peptides derive from the pro-vasotocin-neurophysin precursor. It has been suggested that these peptides are specific regulators of water and electrolyte content of amphibians. The water retardation stimulated by neurohypophysial hormones (Brunn effect) is quite clear. The permeability of the skin or bladder epithelium increases which allows water to penetrate rapidly into the animal. It is the type of per- meability called the osmotic permeability which is influenced. The effect still depends on an osmotic gradient between the internal milieu and the environment. Water molecule aggregates follow the gradient. Besides, the short circuit current is also influenced by nonapeptides which means that not only water, but also sodium uptake increase. This is recently reported for Xenopus laevis [17]. The direct actions of hypophysial hormones important for osmoregulation are not limited to skin and bladder. Receptor binding of nonapeptides is also found at the glomeruli of the kidney [21] which is congruent with the observation that AVT reduces the glomerular filtration rate [12]. A vasoconstrictor effect of AVT on glomerular blood vessels is quite common in Amphibia [4, 32]. Receptor binding of AVP was also found at the glomeruli of axolotls (Kloas and Hanke, unpublished). Very interesting effects of the nonapeptides were found at the interrenal tissue and at the liver. At least, these effects support the osmoregulatory processes by providing metabolites. There is a quite strong stimulatory effect of AVT on the corticosteroid release from the interrenal tissue of the clawed toad in vitro (Fig. 5). The stimulation already occurs with very low amounts of AVT added to the medium. 0.2 nM given for 5 min (=1 ml, total 0.2 pM) were already effective. This dose is in the same range as the effective dose of ACTH in molar units. The receptors which are involved in the stimulatory action of nonapeptides cannot be described as V; or V> receptors. Vj, antagonists did not prevent the effect of the nonapeptides and the V> agonists could not mimic the effects of AVT. Therefore, it has been postulated that a special type of Amphibian receptor exists in this case [18]. Nonapeptides induce glycogenolysis in liver cells of Amphibia. This effect has been described in several papers for Xenopus and axolotl [13-16] and also for lungfish tissue [10]. In a recent study, isolated hepatocytes (primary cultures) of the clawed toad have been carefully analyzed. Figure 1 Hormones in Osmoregulation of Amphibia 7 demonstrates a dose-effect relation between the dose of arginine-vasotocin (AVT) and glucose release. It is evident that addition of 1 nM AVT to the medium has already quite clear effects. There is also a good correlation between glycogen content after incubation and dose of AVT. The higher the dose the lower the glycogen that is left in the cells (Fig. 2). To analyze the type of nonapeptide receptor which is involved an antagonist to the V, receptor, (1-8-mercapto-f, B-cyclo-penta methylene propionic acid)-2-(0-methyl)-tyro- sine-arginine vasopressin), was given. ‘The antagonist did not inhibit the response. Furthermore, a V>-agonist, (1- deamino-8-D-arginine)-vasopressin, was as effective as AVT (Figs. 3 and 4). This demonstrates that a V2-type receptor which works via cAMP mediates the effect which is in contrast to that known for rat liver [1]. Summarizing the effects of the nonapeptides, three main targets must be pointed out. The hormones stimulate water and Na-uptake through skin and bladder, they induce corti- costeroid release by the adrenals with some effect on the catecholamine release, too, and they have a glycogenolytic AVT dose—effect—curve 0.8 © Control 0).7/ o AVT 1 uM T o AVT 0,1 uM 0.6 v AVT 10 nM = AVT 1nM lo} o ©. aN ©) iN) Glucose release (mg/mg Protein) ) T T T T T LTS Sse (0) SOR OOmeoOm es ZO DON | Om 2024 O27 Oe SOO S50 Incubation time (min) Fic. 1. Changes of glucose release from hepatocytes of the clawed toad after addition of different doses of AVT (according to [1]). AVT effect on glycogen breakdown =.2 oF q < Ta a as 3 3.0 [0 2 34d 3 A Fe Ss 2 Sel Le & eS = Gel ] a5 E 5 M ‘0 | 2S re) oo 0 ee 2 &§ re aA SOs ie a ws 2 zo | n 4 ° xa Ss 4 5 > o ° 8 left column S Fo] a2, bo | o 1 G& = control ac " S § (lll) = inv =) 2 2 & |=) = lal Co inM 10nM 0,1uM 1M B 3 = 0,1uM MMMM Glucose release (mg/mg Protein) eerezs| = 1uM [J Glycogen content before hormonal treatment Fic. 2. Changes of glycogen breakdown in hepatocytes of the clawed toad after addition of different doses of AVT (according to [1]). AVT effect of the Vi—receptor antagonist ((1—B—mercapto—£,B—cyclopentamethylen propionic acid)—2—(O—methyl)—Tyrosine)—AVP) 1.0 © Control O"S = © AVT 1 uM e AVT 1 uM + Antagonist 10 uM 0.8 | y AVT 1 wM + Antagonist 1 uM oa AVT 1 uM + Antagonist 0,1 uM 0.6 0.5 0.4 0.3 0.2 0.1 Glucose release (mg/mg Protein) 0.0 . T T =p la T SEES Fed Li Fis 0 30 60 90 120 150 180 210 240 270 300 330 Incubation time (min) Fic. 3. Effects of the V,-receptor antagonist, added in different doses to the medium containing 1 ~M AVT (according to [1]). AVT effect of the V2—receptor agonist (1—deamino—8—D—Arginine)—Vasopressin (dDAVP) 1.0 ° Control 0.9 © AVT 1 uM Has no dDAVP 1 uM ° =» dDAVP 0,1 uM ae 4 aDAVP 10 nM Glucose release (mg/mg Protein) (0) 30 60 90 120 150 180 210 240 270 300 330 Incubation time (min) Fic. 4. Effects of the V2-receptor agonist added in different doses to the medium in comparison to the effect of 14“M AVT (according to [1]). effect on the liver. 1.2. Role of the adrenocortical hormones in osmoregulation 1.2.1. Osmoregulatory function of the corticosteroids The two main corticosteroids of Amphibia, corticoster- one and aldosterone, are both involved in osmoregulation. It is well known that the skin is the main target and Na~ transport is strongly enhanced by these hormones. In a general view, it has been suggested that aldosterone as well as corticosterone and cortisol (the latter may not be present in Amphibia) increase sodium transport across the toad bladder and skin [3]. This is mainly based on the work of Crabbé [5, 6] who described this effect at the skin of Bufo marinus and other toads. With respect to this classic work it must be pointed out that the main problems of this effect are not clearly elaborated, especially not in the clawed toad which may not have a very effective Na‘ transport system. The following questions are not answered. Are both corticoster- oids equally effective on a molar basis? Is the transport 8 W. HANKE AND W. KLOoaAs system influenced by the acclimation condition, e.g. desicca- tion, Osmotic stress etc., and does the response on corti- costeroids depend on these conditions? Much more work must be done. The importance of corticosteroids for osmoregulation is clearly demonstrated when the clawed toads are acclimated to salt water. During acclimation the titre of both corticoster- oids increased (see chapter 1.2.2.3.). In connection with these changes of the corticosteroids blood glucose levels also increase and are adjusted to normal levels during next days. Long term acclimation (for about 4 weeks) in hypertonic solutions, 1.2% salt, 340mM urea or 340mM mannitol, resulted in insignificantly elevated levels of aldosterone or corticosterone [17]. The congruence of the changes of corticosterone and aldosterone in response to acclimation conditions clearly points out that the secretion of both hormones is strongly correlated. There is no separate regulation of one hormone. Furthermore, it is evident that glucose levels in the blood are also correlated with the titre of both hormones. The in- crease of glucose levels is likely to be the most important effect of both corticosteroids. The action of corticosteroids is involved in stress re- sponses. Since hypertonicity of the environment or desicca- tion are important stressors in lower vertebrates, it can be suggested that the normal physiological function of osmoregulation is correlated with stress response. Till now it cannot be clearly decided whether the several functions of corticosteroids are separate direct effects. It could still be possible that some of the target organs are only indirectly influenced via the metabolic effects. 1.2.2. Regulation of the secretion of corticosteroids 1.2.2.1. Effects of telocrine factors The discussion of the regulation of corticosteroid secre- tion must begin with the statement that in all interrenal cells of Amphibia one biosynthetic pathway occurs which starts with cholesterol esters and runs via pregnenolone to corticos- terone (Table 3). This is then converted to aldosterone. Both the intermediate product corticosterone and the end- product aldosterone are secreted by the cell in definite amounts. The exact amounts of both products cannot be easily determined. The amount present in the blood de- pends on the turnover rate besides the secretion rate. In the clawed toad, 4-6 ng/ml of corticosterone and 0.8—1.0 ng/ml of aldosterone were found under normal conditions. The C/ A ratio in the blood serum is about 5. The release from interrenal preparations depends strongly on the medium composition. The less polar corticosterone needs a protein carrier to be released, while aldosterone is less dependent on it. Interrenals of the clawed toad produce under the defined conditions used in our laboratory 150-250 pg/min corticos- terone while about 50-100 pg/min aldosterone is produced. The C/A ratio is mostly less than 3 which may show that the corticosterone secretion is reduced under the more artificial conditions described. Stimulaltory signals can change the C/A ratio, because there are two main steps where the biosynthesis can be regulated. The first step is at the level of the 3 OH-steroid dehydrogenase converting pregnenolone, the second one is the 18-hydroxylation changing corticosterone to aldosterone. When the first step is influenced, a change of the C/A ratio is unlikely unless a limitation of the rate of the corticosterone TaBLE 3. Regulation of hormone production in the Amphibian interrenal cell Biosynthesis Regulatory steps Effectors CHOLESTEROL PREGNENOLONE PROGESTERONE CORTICOSTERONE ALDOSTERONE Release: C/A 3-OH-Steroiddehydrogenase 18-Hydroxylation Stimulation <10nM ACTH AVT UROT. II >10 nM UROT. I ANG II Substance P Inhibition ANF Modulation minus: Met-enkephaline plus: Endorphin Adrenaline plus or const.-stimulation of 3-OH-SDH minus-stimulation of 18-Hydroxylase Hormones in Osmoregulation of Amphibia 9 conversion occurs. The stimulation of the second step de- ANGIOTENSIN Il — Response creases the C/A ratio by increasing the aldosterone secretion pg/ml (Table 3). It is not known whether a change in the C/A ratio has Wy 400 any physiological impact. One assumption may be that Fe 3 aldosterone is less active in inducing glucose release by the = $ liver but this is only supported by results on the specificity of 1S =e) SOO) the receptors in the liver and not by the amount of glucose Zz < produced. = ; d é : x There are two effective groups of peptides which stimu- = 200 late the corticosteroid release [9]. The first group stimulates W in low doses (less than 10 nM must be added for 5 min, total 5) of 10 pM/mlin1 ml). ACTH, AVT and urotensin II belong Gi 100 “ae to this group (Figs. 5 and 6). The second group of peptides thi we act only in higher doses (more than 10nM for 5 min). EE Setes | . . . Q S252) Urotensin I, substance P and angiotensin II are representa- e KY | tive of this group (Table 3). Some of the mentioned pep- o a | tides may act on a paracrine way and shall be discussed in the ALDO BX® CORT! Fic. 7. Changes of corticosteroid release in vitro after addition of | AVT — Response different doses of angiotensin II. Differences in pg/ml between Jal the highest value (ca. 30 min after addition) and the basal value 1600. before angiotensin II. Abscissa nM angiotensin II given for 5 ” min (=1 ml). Ld =) 3 < 800 En Z 2 600 UROTENSIN | — Response z ss pg/ml 400 8 Q SNS E Zz Kos bes W200 wa RY NS oz z B NG 2S 250 a o L sks. BY NS: Elec OQ. 0.2 0) 50 nM 4 Y 200 ALDO B® CORTI =o Fic. 5. Changes of corticosteroid release in vitro after addition of x< different doses of AVT. Differences in pg/ml between the = 150 highest value (ca. 30 min after addition) and the basal value @p) before AVT. Abscissa nM AVT given for 5 min (=1 ml). S 100 Ro ZZ RS Lo seses UROTENSIN II— Response e5 | pg/ml Re 590 1000 = 0 oO 3 0 : =, 800 1.0 10 100 500 nM 2 NSYALDO RI CORT! a Fic. 8. Changes of corticosteroid release in vitro after addition of different doses of urotensin I. Differences in pg/ml between the highest value (ca. 30 min after addition) and the basal value before urotensinI. Abscissa nM urotensin I given for 5 min (= 1 ml). DIFFERENCES MAXIMAL minus obi Petetetetetex next chapter (Figs. 7-9). In most of the experiments, the changes of aldosterone were larger than those of corticos- [o) (e) ol [o) ol [e) [e) wW le) ol (o) a = ALDO [&X# CORT! Fic. 6. Changes of corticosteroid release in vitro after addition of different doses of urotensin II. Differences in pg/ml between the highest value (ca. 30 min after addition) and the basal value before urotensin II. Abscissa nM urotensin II given for 5 min terone. 1.2.2.2. Effects of paracrine factors There is a lot of evidence that the adrenal gland, (=1 ml). including the interrenal and the chromaffin tissue, contains a 10 W. HANKE AND W. VX90 EFFECTS OF SUBSTANCE P ng/ml Q 2 8 Ss 0.65 WP 0.51 Qo ig io © / 0.4 Oo Loe fe) 0-O-(f N po) 0.3 oO O 0-0 e Oo y, e 0.24 ° e \° ‘ ike ot | x i \ eo e-@ 0.1 e® @ 0.0+ — a (0) 50 100 150 200 250 300 O—O CORTICOSTERONE @——®@ ALDOSTERONE Fic. 9. Changes of corticosteroid release in vitro after addition of 4 different doses of substance P. Normal course for 5S hrs. The doses are given in 60 min intervals. high amount of different factors, peptides as well as catechol- amines. Using immunohistochemistry, it has been shown that the gland of several anuran species contain opioid peptide, natriuretic peptides, tachykinines, substance P, angiotensins and others besides catecholamines {27, 28, 34, 35]. The peptide content of the adrenal gland of Xenopus has not been determined. Differences between the species certainly exist, because the gland of the water frogs contains Stilling cells which are not present in the gland of grass frogs or the clawed toad. Despite of not knowing which sub- stances are produced in the Xenopus adrenal gland, it must be expected that several peptides can act on corticosteroid secretion. The influence of several peptids and catecholamines has been tested on corticosteroid release from Xenopus interrenal preparation in vitro. Substance P has already been men- tioned as stimulator in quite high doses, which suggests that paracrine application of low doses might be effective. Opioid peptides clearly modulate the steroid release. Met- enkephaline has no clear effects per se, but it reduces the response on ACTH. There is a very clear and significant reduction of the amount of aldosterone secreted under the stimulatory influence of ACTH (Fig. 10), when met- enkephalin is present. In the same type of experiment a-endorphin enhances the response on ACTH. Especially, aldosterone secretion is much larger, when a-endorphin is present at the time of stimulation (Fig. 11). In the same way, catecholamines may influence the corticosteroid release. It is difficult to find a direct effect on the secretion rate. For dopamine, inhibiton has been reported [30], but the situation is not yet clear. In the case of adrenaline, we found good evidence in our preparations that the release of corticoster- one under the influence of ACTH was reduced. But, the aldosterone secretion rate stimulated by ACTH was higher when adrenaline was present. This effect would explain that adrenaline regulates the biosynthesis in such a way that stimulation of the more potent corticosteroid by ACTH is CONTROL DIFFERENCES ACTH—EFFECT minus Fic. 10. DIFFERENCES ACTH—EFFECT minus Fic. 11. DIFFERENCES ACTH—EFFECT minus Fic. 12. KLoas 100 ACTH—RESPONSE under METENKEPHALIN + + + CORTICOSTERONE ALDOSTERONE ON 4 S x OO x 2 2 S LQ ras OOO x xs x 6; vat xXx ose 3. 4. 5. 1. 02, ORS Tae 5 min INTERVALS after ACTH CONTR ACTH BSI MET.+ACTH Influence of 1 ~M met-enkephalin on the response to ACTH. Comparison of the response of controls and ex- perimentals with added met-enkephalin. 5 successive 15 min intervals are indicated. ACTH—RESPONSE under alpha—ENDORPHIN + + + 4+ CORTICOSTERONE ALDOSTERONE oH OOH =e S OD S x x x Xx x x x x rr S S © © D 2 © © D> x o> OO x Xx > Z © <6 Xx D x x © vas © 2X > xX 6, > > xX Xx > © D > x =e XX er D XS 5 Ls 628 x Ne Ne NINN: ne NEN: OL ete LO 5 ae 2 SS ‘Sy. 15 min INTERVALS after ACTH CONTR ACTH BY} ENDO+ACTH Influence of 100nM a-endorphin on the response to ACTH. Comparison of the response of controls and ex- perimentals with added a-endorphin. 5 successive 15 min inter- vals are indicated. ACTH—RESPONSE under ADRENALINE + + CORTICOSTERONE ALDOSTERONE S > So x © > ¢, x x o x > D os x © S o S yj ES x © 5 > x x x x x x x x © xX > x x S, o > x © o x > x x x > x © © ¢ © > x © 4 © x ro > x o xX © x (CLLLLSLLLL LLL LLL LL OO LLL IISA xX CLLLLLLMLLMMM LL 08 KO ‘Zs 5O COLLTLLLLLLLL LLL LLL . OO min INTERVALS after ACTH CONTR ACTH BSI ADR + ACTH Ow » 2 NO GW uo Influence of 1 «M adrenaline on the response to ACTH. Comparison of the response of controls and experimentals with added adrenaline. 5 successive 15 min intervals are indicated. Hormones in Osmoregulation of Amphibia 11 more pronounced (Fig. 12) [7]. It is difficult to incorporate all these evidences for paracrine effects of peptides and catecholamines on corticos- teroid production and secretion to the current picture for Osmomineral regulation. But it can be concluded that all these factors are involved when the Amphibia are osmotically stressed. 1.2.2.3. Effects of electrolyte concentrations Osmotic stress for the amphibians is mostly followed by changes of the osmotic and electroyte concentrations in the blood serum. It is well known, that clawed toads can be acclimated to 1.2% salt solution, to quite high concentrations of 500 mM urea, however only to about 300 mM mannitol. These three osmolytes act differently. Salt solution stimu- lates an increase of Na* concentration. Urea can penetrate into the animal and increase the osmotic value by accumula- tion of urea. Mannitol acts by desiccation which also in- creases the electrolyte levels. It has been shown that after 3 weeks of acclimation a steady state has been reached when the three osmolytes are applied with 340 mOsm [17]. In all three cases the osmolal- ity is increased to 340-380 mOsm instead of 230 mOsm. The Na* concentration is elevated from 110 mM to approx- imately 120-130 mM, and it is a bit higher in salt water acclimated toads. But, the urea content increases to 80 mM in the urea or mannitol treated toads in contrast to 10 mM which is characteristic for the clawed toad in tap water. The urea concentration reaches about 60 mM when the osmotic stress is due to 1.2% NaCl. Together with the level of Na* which is about 20mM higher the same osmotic pressure is adjusted. It is known from former experiments [37] that acclima- tion to 1.2% NaCl resulted in a strong increase of Na* within the serum till the higher production of urea after about 3 days joins the necessary increase of the osmotic value and allows a reduction of the Na‘ concentration. The electrolyte changes under osmotic stress suggest that there might be a correlation between changes of Na* or K~ concentrations and steroid production. It has been shown that elevation of K* in the medium increases the corticosteroid release as well as reduction of Na‘ concentration in Rana temporaria [29]. To check the influence of Na* and K* concentrations in the surrounding medium more carefully in Xenopus laevis the following combination of in vivo and in vitro experiments were designed; 1. in vivo; The toads were kept in 1.2% sea salt solution (SW) (approx. 1/3 of sea water) or in 0.5% KCI solution. The electrolyte content and the corticosteroid concentra- tion were followed for 3 days. The changes of glucose concentration were also determined to decide whether a stress like correlation is present. 2. in vitro; The interrenals were subjected to media with different Na* and K* concentrations and the direct influence was measured. The results of the in vivo experiment are demonstrated in Figures 13-17. In 1.2% sea salt, the increase of the serum Na‘ concentration during 3 days is obvious. At the same time the K* level decreases but the effect is not statistically significant because of a large variation of the values. Some- times haemolysis occurs which increases the K* concentra- tion. In 0.5% KCl, a trend of increasing K* concentration is observed, while the Na* level is lowered. The opposite changes of Na‘ and K* levels under both acclimation conditions demonstrate that the levels of the both monova- lent ions are inversely correlated (Figs. 13 and 14). ELECTROLYTE CONCENTRATION in XENOPUS Kt Nat mM TREATMENT with 1.2% SEA SALT 20 200 Lee BY Kt 850% 5% C2325 °. C2505 e; e2eZe| 2; $2525 6252525 sores <3 SE 25 OR $ 225 ererere! 25252595 o, e res re > © vas o, LLL: rat ararae Le ererereres 6 $35 ° O05 b > e525 0506 6 5 2 O "es oe o 5 52626 "as ‘ones Se -o5 > G $35 2 PS Me b2 p> 0.0. co 1 6 1 24 72 ~~ hrs Fic. 13. Changes of serum electrolyte concentrations when the clawed toads are kept in 1.2% salt solution. Increase of Na‘ concentration and trend of decrease of K* levels are found. ELECTROLYTE CONCENTRATION in XENOPUS Kt Nat mM TREATMENT with 0.5% KCI 20 2007 SSY Nat Be Kt 15 150} 10 100+ 5050505 PPPS ore: <1 & %s oS oe es 2 es es <5 S508 25 o oR 625 0% res $03 25 O oe res o res res 5 oe Re% Me <> RX? be bX le, "es %, PS % Be PS (5 Ke oO CO 1 6 12 24 72 ~ ‘rs Fic. 14. Changes of serum electrolyte concentrations when the : : k clawed toads are kept in 0.5% KCl solution. Increase of K concentration and trend of Na* decrease are found. At the same time glucose concentration increases reaching maximal levels at 12-24 hrs (Fig. 15). The glucose level is adjusted to normal after 72 hr, which indicates that the acclimation status is approached. ‘The elevation of glu- cose is stronger in toads acclimated to 1.2% sea salt than those in 0.5% KCl. This may indicate that the higher concentration (1.2%) is more effective than the lower (0.5%). The type of the ions is not so important despite of the toxicity of high K* concentration. The concentration of both corticosteroids is increased. 12 W. HANKE AND W. KLoas GLUCOSE CONCENTRATION IN XENOPUS ug/ml SERUM TREATMENT with sea salt or KCI 1500 7 NN 1.2% sea salt N BB 0.5% KCI 1200 \ | \ 900+ N | NIN Stale fC SoBNGS Ne Ni we Na NE Ne Ne NG Ne Ne Ne Na Ne 0 N=: N=: NEN: co 1 6 24 72 ‘hrs Fic. 15. Changes of the glucose levels in the serum of the clawed toads treated with 1.2% salt solution or 0.5% KCI. Corticosterone is raised from about 4mg/ml to 14 mg/ml after 24 hr of acclimation. It decreases afterwards to nearly normal levels after about 7 days (final values not indicated in Fig. 16). Aldosterone increases from less than 1 ng/ml to 2.5 ng/ml after 12 hr and reaches normal values after 3 days. The same trends are seen during acclimation to 0.5% KCl, but again the extent of the increase is smaller (Fig. 17). CORTICOSTEROID CONCENTRATION in XENOPUS TREATMENT with 1.2% SEA SALT ng/ml! SERUM 15 CORTICO BX ALDO co 1 6 12 24 72ers Fic. 16. Changes of the corticosteroid concentration in the serum of the clawed toads treated with 1.2% salt solution. CORTICOSTEROID CONCENTRATION in XENOPUS ng/ml SERUM TREATMENT with 0.5% KCI 1S 7 CORTICO Ry ALDO CO 1 6 12 24 72 ~=‘hrs Fic. 17. Changes of the corticosteroid concentration in the serum of the clawed toads treated with 0.5% KCl. In in vitro experiments, the Na* and K* concentrations of the media were changed. The normal medium contains 100 mM Nat and4mM K*. This was changed to Na*/K* 95/15, 90/20, 85/25, 78/35 mM. In all cases, corticosterone was not significantly changed. But the release of aldoster- one increased when the Na* concentration was lowered and the K* level increased. The extent of the release depends on the concentration (Fig. 18). EFFECTS of MEDIUM CHANGES — ALDOSTERONE RELEASE 3500 HM 54 0505005050! SKK Oe RXR % 4 oS 5 S O re 050; e e, a BRAK? or NSS OOO OXI 050 e er oe OD ORXS Oy ORK SS DIFFERENCES VALUES minus CONTROL Y : Y & Kee) ote 100 gs OS a \Z ee RY \Z= Gs Ses} ee, G os} 20 eS SUNG NG: o> OY Nees Key \Z Rx N Y Neca 7S YS NAS NZ Re IWR INZ2% NB NZ PERIODS: 1. 2. 3. 4. MM Na/K 95/15 Na/K 90/20 15 min Na/K 85/25 BY Na/K 78/35 Fic. 18. Changes of aldosterone release from in vitro preparations of Xenopus interrenals after changes of the medium. The difference between the value after change minus the basal value is given for 4 periods of 15 min. 1.3. Role of the tissue derived peptides in osmoregulation A very strong influence on osmoregulation comes from two tissue derived peptides, the effector of the renin- angiotensin system, angiotensin II, and the most important natriuretic factor, the atrial natriuretic peptide (ANF). Both peptides and systems are found in Amphibia and affect mainly the kidney function. Both systems have very often opposite effects and can be regarded as systems balancing excretion by the kidney. Generally, angiotensin II causes vasoconstriction and retention of sodium and water, while ANF promotes sodium secretion and diuresis. They also have opposite effects on the level of corticosteroid produc- tion. Angiotensin II stimulates the release of aldosterone. ANF is an inhibitor of aldosterone production. These effects are well known from mammals but there is also good evidence that similar effects occur in Amphibia. 1.3.1. Role of the renin-angiotensin system Angiotensin II reduces primarily the glomerular filtra- tion rate and increases blood pressure. Indirectly the urine production is reduced in Amphibia. The decrease of urine production after AII injection has been shown in Bufo arenarum [33]. Stimulatory influence on aldosterone and corticosterone secretion was demonstrated in Rana ridibunda [31] and Rana temporaria [11]. The situation within the clawed toad, Xenopus laevis, is complicated. Quite large Hormones in Osmoregulation of Amphibia 13 doses of AII are necessary to stimulate the corticosteroid release (Fig. 14). Binding sites for AII are found in the structure of the glomeruli of the kidney and at the adrenal tissue of Rana temporaria [25], but are only found at the glomeruli and not at the adrenals in Xenopus. This points to the fact that the aquatic living clawed toad does not respond as well to the renin-angiotensin system at the adrenal level [23]. The physiological significance for AII binding at the glomeruli in Xenopus can be seen when the animals are acclimated to 1.2% salt water. During acclimation a much higher binding capacity of AII at the glomeruli is found 12 hrs after the start of the experiment. The binding is reduced to normal levels, when the animals are acclimated for 7 days. It can be suggested that the amount of urine is reduced at the beginning of the salt water influence. When the acclimation is completed, the urine production returns back to normal. 1.3.2. Role of the atrial natriuretic peptide Several natriuretic peptides are meanwhile found, but the atrial one is surely the most important of the Amphibia. In mammals, specific receptors have been found in the kidney and in the adrenal tissue (for reference see [20]). This suggests that ANF antagonizes the effects of AII and vaso- pressin in the main target organs. Frog cardiac extract has diuretic and natriuretic effects in rats. ANF binding sites in Xenopus laevis have been localized in the glomeruli and the adrenal tissue. The binding sites have been characterized with respect to specificity, association and dissociation con- stants etc. Again acclimation to 1.2% increases the amount of bound ANF in glomerular region. This does not fit into the picture of osmoregulatory influence and, therefore, the physiological significance of this is not clear[20, 24]. In vitro experiments measuring corticosteroid and catecholamine re- lease from preparation of adrenal tissue, have shown that ANF clearly inhibits the release of corticosteroids. The secretion rate of catecholamines is not affected [22]. 1.4. Role of catecholamines in osmoregulation Hyperosmotic conditions of the environment are a very important stress situation for Amphibia. An important hor- monal response versus stress is the increase of circulating catecholamines. In the case of most Amphibia, adrenaline is secreted in higher amounts than noradrenaline. There are several indications that these catecholamines are involved in osmoregulation. Firstly, it has been reported that catecho- lamines increase the uptake of Na* by the skin. But, this effect is not clearly analyzed. Secondly, glycogenolysis is induced by catecholamines. The process of osmoregulation depends on energy, provided by glucose the level of which increases in the blood after gluconeogenesis. Thirdly, there is a connection between catecholamine secretion from the chromaffin tissue and the stimulation of corticosteroid release from the interrenal cell. In this report, the effect of catechol- amines on liver tissue and cells should be discussed. There are several papers dealing with the effect of adrenaline on liver pieces of Xenopus or axolotol [13, 15, 16]. These glycogenolytic effects of adrenaline in Amphibia are medi- ated by a f-adrenergic receptor mechanism which involves cAMP as mediator. The same effects have been clearly elaborated using Xenopus hepatocytes. Adrenaline works via receptors which can be blocked by f-antagonist, e.g. propranolol. Noradrenaline is not effective. Forskoline, the stimulator of adenylate cyclase, is as effcetive as adrenaline [1]. The question how catecholamine and corticosteroid secretion affect each other has been often discussed. In- fluence in both directions, catecholamine stimulation of corti- costeroid secretion and vice versa, is proposed in the litera- ture. The aldosterone secretion of the interrenals of Xeno- pus laevis is significantly stimulated by adrenaline (Fig. 12) [7]. But, reduction of corticosteroid secretion by catecholam- ines has been found in Rana ridibunda [30]. On the other hand, stimulators of corticosteroid release in Xenopus laevis, like ACTH or AVT, induce an increase of catecholamine release [19]. All these findings indicate that catecholamines support the mechanisms of osmoregulation at different levels, the target organs, the metabolic support and effects at the level of the adrenal gland, the mixture of interrenal and chromaffin tissue. 2. Cooperation at the level of peripheral organs The different hormonal systems discussed in the preceed- ing chapter act together at different target organs. The cooperation at the level of the hypophysis, the liver, the skin and the kidney shall be discussed briefly. Other targets like intestine and muscle may also be of interest in this respect. Binding sites for ANF and angiotensin II have been found at the pituitary [26]. This suggests that these hor- mones cooperate with AVT and adenohypophysial hormones like ACTH. It is possible that ANF inhibits the release of AVT, because both hormones have opposite effects at the renal level. On the other hand, angiotensin II may mod- ulate the release of nonapeptides from the pituitary. Both peptide systems act synergistically at the renal level. Catecholamines, corticosteroids and nonapeptides act on the liver and stimulate at least the release of glucose. The mechanisms are different. Catecholamines induce glu- coneogenesis via f-receptors. The effect of nonapeptides is mediated by cAMP and might be a V>-receptor mechanism. It is not clear whether besides glycogenolytic also glu- coneogenetic pathways are involved. Corticosteroids act on gluconeogenesis to elevate glucose release. In all cases glucose is provided to support the process with energy. Nonapeptides induce water retardation by the skin and bladder. They also affect the glomerular filtration rate and reduce the urine volume. At the level of the skin, Na uptake is also stimulated by nonapeptides. The mechanism is not clear. These effects are in accordance with those of the corticosteroids on the skin. It has to be investigated if catecholamines support these actions. The mechanisms of these hormones at the the skin, bladder and glomeruli are still 14 not exactly known. W. HANKE AND W. It must be studied whether the mecha- nisms of the different hormones are similar. 10 11 12 13 14 REFERENCES Ade T, Hanke W (1992) Proc Int Symp on Amphibian Endo- crinology, Tokyo Barker-Jorgensen C (1993) Comp Biochem Physiol 104A: 1- 21 Bentley PJ (1971) Endocrines and Osmoregulation, Springer Verlag, Berlin, Heidelberg, New York Boyd SK, Moore FL (190) Gen Comp Endocrinol 78: 344-350 Crabbé J (1963) Effects of adrenocortical steroids on active sodium transport by the urinary bladder and ventral skin of amphibia. In “Hormones and the Kidney” Memoirs Soc En- docrinol 13: 75-87 Crabbé J (1964) Endocrinology 75: 809-811 Gussetti C, Miller R, Hanke W (1993) Exp Clin Endocrinol 101: 111 Hanke W (1985) A comparison of endocrine function in osmotic and ionic adaptation in amphibians and teleost fish. In “The endocrine system and the environment” Ed by BK Follett, S Ishii, A Chandola, Springer Verlag, Tokyo, Berlin, pp 33-43 Hanke W (1990) Corticosteroid function: evoluionary aspects. In “Progress in Comparative Endocrinolgy” Ed by A Epple, CG Scanes, MM Stetson, Wiley-Liss, Inc., New York, pp 445-452 Hanke W, Janssens PA (1983) Gen Comp Enocrinol 51: 364- 369 Hanke W, Maser C (1985) Regulation of interrenal function in amphibians. In “Current Trends in Comparative Endocrinol- ogy” Ed by B Lofts, WN Holmes, Honkong Univ Press, Honkong, pp 447-449 Henderson IW, Edwards BR, Garland HO, Chester Jones I (1972) Gen Comp Endocrinol, Suppl 3: 350-359 Janssens PA, Caine AG, Dixon JE (1983) Gen Comp Endocri- nol 49: 477-483 Janssens PA, Grigg JA (1984) Comp Biochem Physiol 77C: 403-408 15 16 il7/ 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 KLoas Janssens PA, Maher F (1986) Gen Comp Endocrinol 61: 64-70 Janssens PA, Grigg JA (1984) Gen Comp Endocrinol 67: 227— 233 Katz U, Hanke W (1993) J Comp Physiol B 163: 189-195 Kloas W, Hanke W (1990) Gen Comp Endocrinol 80: 321-330 Kloas W, Hanke W (1990) Proc Soc Comp Endocrinol, Leuven, p 146 Kloas W, Hanke W (1992) Gen Comp Endocrinol 85: 26-35 Kloas W, Hanke W (1992) Gen Comp Endocrinol 85: 71-78 Kloas W, Hanke W (1992) Gen Comp Endocrinol 85: 269-277 Kloas W, Hanke W (1992) Gen Comp Endocrinol 86: 173-183 Kloas W, Hanke W (1992) Peptides 13: 297-303 Kloas W, Hanke W (1992) Peptides 13: 349-354 Kloas W, Hanke W (1992) Cell Tissue Res 267: 365-373 Larcher A, Delarue C, Idres S, Lefebvre M, Feuilloley M, Vandesande F, Pelletier G, Vaudry H (1989) Endocrinology 125: 2691-2700 Leboulenger F, Leroux P, Delarue C, Tonon MC, Charnay Y, Dubois PH, Coy DH, Vaudry H (1983) Life Sci 32: 375-384 Maser C, Janssens PA, Hanke W (1982) Gen Comp Endocri- nol 47: 458-466 Morra M, Leboulenger F, Vaudry H (1990) Endocrinology 127: 218-226 Perroteau I, Netchitailo P, Homo-Delarche F, Delarue C, Lihrmann I, Leboulenger F, Vaudry H (1984) Endocrinology 115: 1765-1773 Pang PKT, Galli-Gallardo SM, Collie N, Sawyer WH (1980) Am J Physiol 239: R156-R160 Reboreda JC, Segura ET (1989) Comp Biochem Physiol 93A: 505-509 Reinecke M, Heym C, Forssmann WG (1992) Cell Tissue Res 268: 247-256 Rossier J, Dean MD, Livett BG, Udenfriend S (1981) Life Sci 28: 781-789 Rouillé Y, Michel G, Chauvet MT, Chauvet J, Acher R (1989) Proc Natl Acad Sci USA 86: 5272-5275 Shréck H, Hanke W (1979) Comp Biochem Physiol 63A, 393- 397 ZOOLOGICAL SCIENCE 11: 15-31 (1994) © 1994 Zoological Society of Japan REVIEW Neuropeptides in Neurosecretory and Efferent Neural Systems of Insect Thoracic and Abdominal Ganglia Dick R. NAsset!, EMINE BAYRAKTAROGLU~ and HEINRICH DIRCKSEN® ‘Department of Zoology, Stockholm University, S-10691 Stockholm, Sweden, *Department of Biological Sciences, Middle East Technical University, Ankara, Turkey, and *Department of Zoophysiology, Rheinische Friedrich-Wilhelms-Universitat, Bonn, FRG INTRODUCTION Over the last ten years there has been a dramatic increase in the number of identified insect neuropeptides [38, 39, 55, 56, 63, 94]. Many of these peptides have been attributed biological actions in a variety of assays, and it is clear that insect neuropeptides play important roles in differ- ent regulatory processes involved in development, reproduc- tion, diapause, metabolism, osmoregulation, muscle activity and behavior [38, 55, 56, 86, 90, 95, 105]. With the develop- ment of further bioassays and extensive testing of novel neuropeptides even in heterologous bioassays, it has become increasingly apparent that neuropeptides have more actions that those ascribed to them at the time of isolation [22, 56, 62, 94]. In fact, it may well be that it is a rule rather than an exception that neuropeptides have several functions and thus many names given to neuropeptides may become misleading as novel important functions are revealed. Classically the isolation of peptides involved ablation of endocrine organs followed by reconstitution of regulatory functions by injections of purified “factors” extracted from these organs. Most insect neuropeptides have, however, been isolated with the aid of in vitro assays of actions on peripheral target organs. In either case it is likely that in vivo many physiological actions of these peptides are of hormonal nature. Hence, it is not surprising that a large number of insect neuropeptides have been demonstrated by immunocytochemistry in neurosecretory cells and neurohe- mal release organs [40, 66, 86]. In addition many of the same neuropeptides are present in central neurons and in endocrine cells of the gastro-intestinal tract (reviewed in Refs [18, 66, 94, 110]) indicating that also in insects peptides can act as neuromodulators and local neurohormones [80, 90]. Immunocytochemistry has proved to be a powerful technique for the localization of storage and release sites of both neuropeptides and other neuroactive compounds such as monoamines and amino acid transmitters. By now quite a number of neuropeptides have been mapped in the nervous Received August 25, 1993 and neuroendocrine systems of different insects and it is apparent that the complexity in peptidergic signalling is staggering [18, 66, 86, 94]. The number of known insect neuropeptide sequences by far exceed 100 [56, 66, 94]. Only from a single species of insect, Locusta migratoria, not less than 32 different neuropeptides have been isolated and sequenced as of mid 1993 and many more are under way [94]. Just a fraction of these have been mapped by immuno- cytochemistry and almost nothing is known about their receptors and physiological actions in the nervous system and at peripheral targets. For most peptides suspected to be acting as neurohormones it remains to demonstrated that they are actually released into the circulation. Actions of neuropeptides can be assayed more conve- niently at peripheral targets than in the central nervous system. To draw attention to possible peripheral targets for studies of peptide action, the present review focuses on neuropeptides in the neurosecretory and efferent systems of the thoracic and abdominal ganglia of insects, with special emphasis on blowflies. The distribution of neuropeptides in the insect brain [66] and intestine [110] has recently been reviewed. ORGANIZATION OF INSECT NEUROSECRETORY SYSTEMS Neurosecretory cells have been classified into a few main groups [86]: (i) neurosecretory cells with cell bodies in central ganglia and axon terminals in peripheral neurohemal organs or release areas, (ii) central neurosecretory cells with peripheral so called neuroeffector junctions (secretory- motorneurons and other efferents with peripheral innervation areas), (iii) central neurosecretory cells with arborizations (“neurosecretory endings”) in neuropils of the central nerv- ous system and (iv) peripheral neurosecretory cells with peripheral release sites. Some of this organization is illus- trated in Figures 1 and 2. The first type of neurosecretory cells are found in a few locations of the cephalic ganglia and in the subesophageal, thoracic and abdominal ganglia and send axons to neurohe- 16 D. R. NAsSsEL, E. BAYRAKTAROGLU AND H. DiIRCKSEN Fic. 1. Distribution of CCAP immunoreactive neurons (black circles) and some neurohemal areas (shaded) in the nervous system of the locust. The neurohemal areas are not drawn for all segments. Arrows point at neurohemal areas in junctions of nerves 1 and6. Abbreviations: CC=corpora cardiaca, CA=corpora allata, SEG=subesophageal ganglion, T1-T3=thoracic ganglia, Al-A7=abdominal ganglia, TAG=terminal abdominal ganglion, PSO=perisympathetic organ, LHN=Iateral heart nerve. From Dircksen et al. [19] with permission from Springer Verlag. mal organs or release areas associated with these ganglia [86]. The cephalic neurosecretory cells supply axons to the neurohemal organs termed corpora cardiaca and corpora allata (Fig. 1) and in some insect species to neurohemal areas in the wall of the anterior aorta (see Fig. 4), in the so called antennal heart and at the surface of certain cranial nerves. The neurosecretory cells of the ventral cord supply segmental neurohemal organs, termed perisympathetic or perivisceral organs, located in the median and/or transverse nerves (Figs. 1-3). Other release sites of ventral cord neurosecretory cells are found in neurohemal areas in the pericardial septum of the abdominal aorta, the lateral cardiac nerve, the dorsal diaphragm and the intestine (Figs. 2, 3). The “neurosecretory cells” with arborizations in central neuropil and the efferents with targets such as glands and different types of muscle were originally identified with classical neurosecretory staining methods (see Ref. [86]). Some of these cells have later been identified by immuno- cytochemistry as peptidergic and/or monoaminergic neurons. An example of this kind of cells is the pair of vasopressin immunoreactive neurons of the locust subesophageal gang- lion with extensive arborizations restricted to the central nervous system and the core of some peripheral nerve roots [89, 100]. Peripheral neurosecretory cells have been found at several locations: in nerves (link nerves, connecting trans- verse and segmental nerves) of the thoracic and abdominal ganglia and in nerves associated with the heart and alary muscles and the intestinal tract [28, 86]. Neurosecretory cells have also been reported in the frontal and hypocerebral ganglia. We shall be concerned here with the neurohemal release sites of the body segments only since the cephalic neurohemal organs have been more frequently dealt with in the literature [33, 40, 56, 86]. In the less evolved insects, such as the locust, the distinct segmental neurohemal organs associated with the dorsal median and transverse nerves are easily distinguished in the larval and adult stages [86] (Figs. 1, 2). In blowflies and other higher diptera, however, these neurohemal organs can only be clearly distinguished before metamorphosis in the larval stages [72] (see Fig. 10). In the adult flies the terminals of the neurosecretory cells are located in the dorsal neural sheath of the fused thoracic and abdomi- nal ganglia [31, 68, 72] (Fig. 4). These “centralized” neurohemal structures represent the most evolved type of neurohemal perisympathetic organs associated with the ven- tral nerve cord of insects. Intermediate types of organs derived from median or transverse nerve types that anasto- mose with connectives or segmental nerves are found in hymentopterans,’some coleopterans and orthopterans [32, 86]. NEUROPEPTIDES DEMONSTRATED IN THE THORACIC AND ABDOMINAL GANGLIA Most insect neuropeptides known today have been iso- lated from whole heads, whole insects, dissected brains and corpora cardiaca-corpora allata complexes or from dissected entire nervous systems [39, 55, 63, 94, 97]. Some neuropeptides have, however, specifically been isolated from dissected thoracic-abdominal ganglia as is the case for some blowfly peptides: thirteen different FMRFamide-related pep- tides (FaRPs) and two callatostatins, peptides structurally closely related to the cockroach allatostatins [22, 24]. As Neuropeptides in Neurosecretory and Efferent Neural Systems Se We 17 0G Fic. 2. Semischematic drawing of pathways of CCAP immunoreactive (CCAP-IR) neurons and neurohemal release sites in the locust abdomen. 1 and 2 (N1 and N2). Of interest here are the segmentally repeated neurons 1 and 4 which send axons to segmental nerves From these nerves CCAP-IR supply terminals to neurohemal distal perisympathetic organs (dPSO 5-8), to stigmata of tracheal system (S4-8; inset circle b), ventral diaphagm muscles (VDM; inset circle a), alary muscles (AM5-8; inset circle c) and lateral heart nerve (LHN). these structures are fused in the higher dipteran insects. indicated by immunocytochemistry it appears that most pep- tides isolated from whole insects, whole heads, whole CNS or dissected brain-corpora cardiaca are present in the ventral cord ganglia. The neuropeptides indicated by immuno- cytochemistry in the thoracic-abdominal ganglion of blowflies (Calliphora vomitoria and Phormia terraenovae) are listed in Table 1. In other insect species the presence of some addi- tional native neuropeptides have been indicated in the ventral ganglia by immunocytochemistry. In Locusta migratoria: crustacean cardioactive peptide (CCAP) [19], locustamyotro- pin [94]; male accessory gland myotropin [81], ovary maturat- ing neurohormone [89]. In cockroaches: proctolin [25, 80], leucokinins [70]. In moths: pheromone-biosynthesis- activating neuropeptide (PBAN) in Helicoverpa zea [43] and eclosion hormone in Manduca sexta larvae [103]. It should be noted that the above mentioned neuropeptides are com- No CCAP-IR fibers were seen in median PSOs (asterisks). drawing is useful as a basis for the organization of primitive and segmental neurohemal structures in insects. This Many of From Dicksen et al. [19] with permission from Springer Verlag. monly found not only in neurosecretory cells, but also in different types of interneurons of the thoracic and abdominal ganglia (Table 1). In addition proctolin has been demon- strated in motorneurons [80] and leucokinin in putative sensory neurons (intestinal stretch receptors) [70]. GENERAL ORGANIZATION OF THORACIC- ABDOMINAL NEUROSECRETORY SYSTEMS AND NEUROHEMAL ORGANS AND RELEASE SITES Before turning to the neurosecretory systems of the fused ventral cord of higher Diptera it may be useful to present the organization of the less evolved neurosecretory system of the locust where the segmental peptidergic neurons and neurohemal structures are clearly discernible. The Fic. Fic. D. R. NASSEL, E. BAYRAKTAROGLU AND H. DircKsSEN 3. CCAP immunoreactive structures in sections and whole mount preparations of Locusta migratoria. A. labeled semithin cross-section through seventh distal perisympathetic organ (dPSO7), showing axon profiles and terminals. Note lack of label in motor axons of the transverse nerve (ITN); Ventral diaphragm muscles are labeled by asterisks. Tr= trachea. B. Axon terminal in a dPSO7 containing neurosecretory granules (pre-embedding immunocytochemistry; peroxidase labeling). Note unlabeled axon profile (asterisk). BL=basal lamina. C. Jn situ whole mount immuno- fluorescence preparation of a dPSOS showing and axon originating in the link nerve (LN; arrow) that gives rise to fine terminals at the surface of the dPSO and the paramedian nerve (PMN). D. Cross section through the neurohemal lateral heart nerve showing three labeled central axons and axon profiles next to the surface of the nerve. Note the granule contents of almost all profiles. BL=basal lamina. Scales: A=10 ~m, B=S500 nm, C=100 ~m, D=S500 nm. a aorta A1-8 4 MES 4. Schematic dissected view of the nervous system, intestinal tract and aorta of the blowfly. The cross hatched areas are putative release sites of peptidegic neurons. In most cases these areas probably represent neurohemal release sites. Release sites are found in the anterior aorta (a. aorta), corpora cardiaca (triangular cross hatched structure below anterior aorta), dorsal sheath of thoracic-abdominal ganglion (T1-3, Al—8), pericardial septum at posterior aorta (abd aorta) and hindgut. Cell bodies (filled circles; not accurate numbers) of neurons are shown in one hemisphere only. The systems displayed are (1) protocerebral neurosecretory cells with axons to corpora cardiaca, anterior aorta and crop duct (CD); (2) subesophageal system (serotonergic) with axons to thoracic-abdominal dorsal neural sheath and several other targets not shown here; (3) thoracic system with terminals in dorsal neural sheath; (4) Lateral abdominal system with axons to pericardial septum of abdominal aorta; (5) median abdominal system with axons to hindgut and sometimes rectal pouch (RP) and its papillae. MT=Malpighian tubules. SEG=subesophageal ganglion. 18 Neuropeptides in Neurosecretory and Efferent Neural Systems 19 TaBLe 1. Neuropeptides indicated in the blowfly thoracic-abdominal ganglia‘ Antisera to native neuropeptide” distribution FMRFamide CalliFMRFamide 1-13 IN, NC, EF pigment-dispersing hormone PDH-like? IN, EF proctolin proctolin* IN, NC, EF, MN leucokinin I leucokinin-like IN, NC locustatachykinin I locustatachykinin-like° IN crustacean cardioactive peptide CCAP-like EF, IN corazonin corazonin-like IN allatostatin Callatostatins 1-5 EF° adipokinetic hormone AKH-like or AKH*~'°-like’ IN, EF, NC? myomodulin (Aplysia) locustamyotropin-like’ NC galanin (mammalian) ? IN, NC galanin message associated peptide 2 INSNCGSEE enkephalins (mammalian) ? IN, NC substance P (mammalian) ? (not locustatachykinin-like) NC gastrin/CCK (mammalian) FaRPs or drosulfakinin-like IN, NC Abbreviations: IN=interneurons, NC=neurosecretory cells, EF=efferent neurons, MN=motorneurons 1. Literature references in text. Further neuropeptides have been indicated in ganglia of Drosophila (see text) 2. The native peptides in italics have been isolated from Calliphora. Others are suggested by analogy with peptides isolated from other insect species. 3. Strong indication for peptide homologous to PDH in Calliphora (partial sequence obtained). 4. Proctolin isolated from different arthropods so far is identical (see Ref. [66]) 5. Peptides with strong sequence homologies to locustatachykinins have been isolated from Calliphora (Lundquist, Holman, Nichols, Nachman, Clottens, Nassel, in press) 6. In Drosophila allatostatin immunoreactivity was detected in neurons and neurosecretory cells throughout the CNS. 7. See Schools et al. (Ref. [94]) schematic diagram of Figure 2 highlights the morphology of two types of typical segmental peptidergic neurosecretory cells with terminals in peripheral neurohemal organs. These are the CCAP-immunoreactive (CCAP-LI) type 1 and type 4 neurons of the locust (L. migratoria) abdominal ganglia described by Dircksen et al. [19]. Terminals of these neurons occur in the distal perisympathetic organs, the lateral heart nerves and the alary muscles associated with the dorsal diaphragm (Figs. 2, 3). Electron microscopy of the CCAP- LI terminals show that they contain large granular vesicles typical of neurosecretory neurons [19] as shown in Figure 3B. The Type 1 and 4 neurons reach the periphery via the lateral segmental nerves (N1, N2) of the abdominal ganglia, which is also the case for segmental leucokinin-like immunoreactive neurons [20]. Other putative neurosecretory cells of the locust have axons running dorsally via the perisympathetic organs in the median and transverse nerves to the peiphery: locustamyotropin-, FMRFamide- and pancreatic polypep- tide-like immunoreactive neurons [27, 60, 94]. In the adult blowfly there are two major neurohemal release sites in the body segments (Fig. 4). One is located in the neural sheath of the dorsal part of the fused thoracic- abdominal ganglion and may correspond to the neurohemal organs of the median and transverse nerves mentioned above. The other is in the pericardial septum or dorsal diaphragm surrounding the abdominal aorta, possibly corresponding to the release sites in the lateral heart nerve and alary muscles of the locust. NEUROPEPTIDES IN THE NEUROHEMAL AREA IN THE DORSAL NEURAL SHEATH OF THE BLOWFLY THORACIC-ABDOMINAL GANGLION Although neurohemal areas were known in the neural sheath of thoracic-abdominal ganglia of higher diptera [2, 31], the full extent of the release area in the blowfly neural sheath was first recognized when serotonin immunoreactive (5- HTIR) terminals were revealed in Calliphora by immuno- cytochemistry [68]. The 5-HTIR fibers supply the entire dorsal surface of the thoracic-abdominal ganglion (see Fig. 6D) and also the neural sheath of many of the nerve roots of the ganglion. It was found that these arborizing 5-HTIR fibers are derived from four large neurons in the sub- esophageal ganglion [68], earlier shown to supply fibers to the neural sheath of ventral nerve roots of the subesophageal ganglion [67]. Later it became apparent that the dorsal neural sheath of the blowfly and fruitfly thoracic-abdominal ganglion also is the termination area of different systems of peptide containing neurons (Figs, 5, 6A, C, Table 2). The first system to be outlined in some detail is formed by six large gastrin/cholecystokinin-like immunoreactive (CCK-LI) ven- tral neurosecretory cells forming an extensive plexus of fibers in the dorsal sheath of Calliphora [65, 72]. This CCK-LI system formed by the ventral thoracic neurosecretory cells 20 D. R. NASsEL, E. BAYRAKTAROGLU AND H. DirckKsEN Fic. 5. Myomodulin-like immunoreactive neurons (neurosecretory cells) in the fused thoracic-abdominal ganglion of the adult blowfly. A. Ventral view of immunoreactive cell bodies. The six large VINCs are located in the thoracic neuromeres (T1—3). Four pairs of VANCs are found in the abdominal neuromeres (A1—4). Cc=cervical connective. B. Sagittal view of the ganglion displaying the same neurons with their processes. Note terminals in the neurohemal release area dorsally in the neural sheath (NH and arrows). Immunoreactive processes are also found in central neuropil and in axons projecting to the subesophageal ganglion via the cervical connective. From Nassel et al. [74] with permission. TaBLE 2. Neuropeptides in neurons and neurohemal areas of blowflies Putative release site Location of cell bodies Thoracic neuromeres (T1-3) Abdominal neuromeres hindgut pericardial septum area GMAP’ proctolin, FaRPs, PDH, CCAP, CavAS proctolin, FaRPs, LK, LVP thoracic neurohemal area FaRPs, SP, MM, GAL MM, FaRPs? proctolin, FaRPs, LK, LVP abdominal neurohemal area fibers in thoracic-abdominal nerves AKH!', GAL/GMAP! Abbreviations: FaRPs=FMRFaminde related peptides, PDH=pigment dispersing-hormone-like peptide, CCAP= crustacean cardioactive peptide, CavAS=callatostatins, LK=leucokinin-like peptide, GMAP=galanin message associ- ated peptide-like peptide, LVP=lysine vasopressin-like peptide (?), SP=substance P-like peptide, MM=myomodulin- like peptide, GAL=galanin-like peptide, AKH=adipokinetic hormone-like peptide 1. The location of the cell bodies is not clear. 2. The “specific” FMRFamide antisera raised in Guinea pig do not label these cells (VINCs) was studied in more detail and turned out to be immunopositive with a number of antisera to non-insect peptides: bovine pancreatic polypeptide, CCK, methionine enkephalin, methionine enkephalin-Arg-Phe [21], FMRFa- mide, molluscan small cardioactive peptide (SCP) (Fig. 6), substance P [50] and the molluscan peptide myomodulin [74] (Fig. 5). It is likely that the VITNCs contain peptides of the CalliFMRFamide series which have been isolated from Cal- liphora thoracic-abdominal ganglia [22] and it is possible that many of the antisera listed above cross react with different epitopes of the CalliFMRFamides. In Drosophila the homologs of the VTNCs can also be identified by antisera against FMRFamide [50, 106, 109]. Several FMRFamide- related peptides (FaPRs) have been isolated from Drosophila tissue or deduced from isolated cDNAs [64, 75-77, 92] and the FaRPs appear to be the products of three different genes [76]. It was shown by in situ hybridization histochemistry that the message encoding one of the FMRFamide precursors is expressed in the Drosophila VINCs [78, 93]. In addition to FaRPs it is possible that the WITNCs contain colocalized Neuropeptides in Neurosecretory and Efferent Neural Systems Fic. 6. Fluorescence micrographs of thoracic-abdominal neurons in wholemounts of the blowfly. A. The neurohemal plexus in the dorsal neural sheath of the thoracic-abdominal ganglion labeled with monoclonal antibody toSCPg._ B. The six cell bodies (T1-3) and segmental neurohemal release sites (arrows) of the VITNCs of a pupal blowfly (48h pupa). At this stage the segmental organization of the release sites is still apparent. Antiserum toSCPg. C and D. Double labeling of the same wholemount with antisera against FMRFamide (C) and serotonin (D) viewed with filters for fluorescein (FITC) and Texas red (using biotin-streptavidin detection). The fibers in the FMRFamide immunoreactive plexus are closely adjacent to the serotonergic plexus (see arrows for correlation) along the ganglion midline. Scales A, C, D=50 um; B= 100 zm. D. R. NAssEL, E. BAYRAKTAROGLU AND H. DircKSEN Fic. 7. Electron microscopy of peptidergic ter Calliphora. A. Immuno-gold labeling to display cluster of FMRFamide immunoreactive terminals in the sheath just below the acellular basal lamina (BL). B. Detail in higher magnification of FMRFamide immunoreactive terminals with large granular vesicles. CC. Conventional electron microscopy of exocytotic profile (arrow) in one of the peptidergic terminals in the sheath. Two peptidergic vesicles are being released. D and E. With the TARI method exocytosis is more easily detected. Here three vesicles (arrows) are being released by a peptidergic terminal. With the TARI method the exocytosing vesicles become osmophilic and hence stand out (arrows in D). BL=the acellular basal lamina. E isa higher magnification of terminal in D. Magnifications: A=14.000x , B=46.000x , C=86.000 x , D=36.000x, E= 72.000 X . Neuropeptides in Neurosecretory and Efferent Neural Systems 23 peptides related to substance P and myomodulin (or locus- tamyotropins). A recent study of larval and adult Drosophi- la has indicated the presence in the VINC homologs of peptides reacting with antisera to Manduca allatotropin and allatostatin [111]. The same authors showed similar cells with terminals in the dorsal neurohemal area reacting with antisera to Bombyx PTTH and Manduca diuretic hormone. The peptidergic neurohemal plexus formed by the VTINCs is not as extensive as the one formed by the 5-HTIR cells. The plexus of the VINCs is restricted to a median portion of the sheath of the dorsal thoracic-abdominal gang- lion, whereas the 5-HTIR plexus covers the entire dorsal surface (Fig.6C, D). Double labeling experiments with 5-HT- and FMRFamide- or SCP, antisera revealed that the fibers of the VINCs and the 5-HTIR fibers are located adjacent to each other in the plexus of the median region of the dorsal neural sheath (Fig. 6C, D). The VINCs also arborize extensively within the thoracic neuropils (Fig. 5) and each cell sends an axonal process anteriorly to the subesophageal ganglion [50]. It has not been determined whether any of the arbors represent input regions of the neurons. The possibility exists that the den- dritic arbors of the VINCs were not immunolabeled, by analogy with the vasopressin-like immunoreactive neurons of the locust subesophageal ganglion where the exclusion of such immunostaining was determined by intracellular dye injection [100]. Additional neurosecretory cells forming terminals in the dorsal neural sheath are found in the abdominal ganglion. Most clearly this was seen for ventral abdominal.neurosecre- tory cells (VANCs) labeled with an antiserum to the mollus- can neuropeptide myomodulin [74] (Fig. 5). Myomodulin [15] shares the C-terminus -RLamide with the locustamyotro- pins I-IV [94] and probably the myomodulin antiserum recog- nizes native blowfly peptide(s) of myotropin type. There- fore it is not surprising that in the locust locustamyotropin antiserum labels abdominal neurosecretory cells with termin- als in median nerve perisympathetic organs [94]. In Cal- liphora and Phormia double labeling experiments with anti- sera to myomodulin and FMRFamide (raised in Guinea pig) and SCP, reveal that the myomodulin immunoreactive VANCs in the abdominal ganglion, do not contain epitopes recognized by the SCP, and the more specific FMRFamide antiserum. A dense myomodulin-like immunoreactive ple- xus extends over the entire abdominal portion of the neural sheath, whereas in the same specimens the plexus labeled with SCPg or FMRFamide antisera is more insignificant in the abdominal portion. Thus it is possible to determine that most of the peptide containing fibers in the most caudal portion of the neurohemal plexus in blowflies are derived from the abdominal neurosecretory cells. A separate plexus of fibers in the thoracic-abdominal sheath derived from cell bodies distinct from the VINCs was labeled with an anti- serum against the mammalian neuropeptide galanin [51]. The only substance that so far has been shown to be released from the thoracic-abdominal ganglion of Calliphora by high potassium depolarization is serotonin; the likely role of the released serotonin is to induce secretion in the salivary glands [101]. No experimental evidence for peptide release from thoracic-abdominal neurohemal areas are available for blowflies, but the peptide immunoreactive terminals in the neural sheath are located outside the blood brain barrier [21, 72| (Fig. 7), like the serotonergic terminals [68]. For in- direct demonstration of peptide release we applied the tannic acid ringer incubation (TARI) method [7] on living ganglia in vitro to augment the detection of exocytosis (diagnostic of release) in peptidergic terminals in the sheath of the thoracic- abdominal ganglion. The dissected ganglia were left in a dish with 0.5% tannic acid in insect saline for 2 hr followed by fixation in glutaraldehyde and osmium [3]. The TARI method renders the core of extruded peptidergic vesicles osmophilic (Figs. 7D, E) which facilitated the detection of numerous exocytosis profiles in peptidergic fibers of the Calliphora thoracic-abdominal neurohemal release site (cf. Fig. 7C). What are the actions of neurohormones released from the neurohemal area in the thoracic-abdominal ganglion? It is presumed that the serotonin released from this area induces fluid secretion in the blowfly salivary gland [10]. It is also known that serotonin can induce diuresis and modulate activity of heart, visceral and oviduct muscles [13, 14, 54, 58]. For the FaRPs some clues have been obtained from in vitro studies on Calliphora salivary glands: three of the CalliF- MRFamides induce fluid secretion in the salivary glands [22]. If this action is physiological it is likely that the peptide(s) reach the salivary glands via the circulation like serotonin does. Duve et al. [23] have also demonstrated that two of the CalliFMRFamides increase the spontaneous activity of the abdominal heart. Since a direct innervation of the abdominal heart has been demonstrated (see below), it is not clear whether hormonally released peptide is involved in this action. The FaRPs, if released into the circulation, can reach a whole host of peripheral targets all of which need to be tested for their response. PEPTIDES IN THE NEUROHEMAL AREA IN THE PERICARDIAL SEPTUM OF THE BLOWFLY The organization of the neurohemal area of the peri- cardial septum (dorsal diaphragm) was first described in the stable fly by electron microscopy [57]. The first identified neurons innervating this neurohemal area were detected in the blowfly with antiserum against the cockroach myotropic peptide leucokinin I [10]. The leucokinin-like immunoreac- tive (LK-LI) fibers reach the septum via the segmental abdominal nerves and are derived from a set of about 20 cell bodies in the abdominal ganglion (Figs. 4, 8A, 9A, B). These cell bodies also form central processes within the median portion of the abdominal neuropil. It is likely that these central processes represent release sites within the neuropil, but it cannot be excluded that they also receive synaptic inputs in this region. Interestingly the cell bodies 24 D. R. NAssEL, E. BAYRAKTAROGLU AND H. DirRcCKSEN 100um Fic. 8. The abdominal portion of the blowfly thoracic-abdominal ganglion with efferent peptidergic neurons. A. Antiserum to leucokinin I labels about 20 neurons (AN) with axons to pericardial septum (altered from [10]). B. Antiserum to proctolin labels a similar (but not identical) set of 20 efferents with axons to the pericardial septum and a set of four caudal abdominal efferent neurons (AE) with axons to the hindgut (altered from [69]). C. An antiserum to CCAP labels four cells posteriorly in the abdominal neuromeres. At least two of these send axons to the hindgut (Nassel and Dircksen, unpublished). and fibers of the 20 LK-IR efferents could also be immuno- labeled with an antiserum to lysine vasopressin [71]. Similar abdominal LK-LI efferents in the cockroach Leucophaea maderae were also found lys-vasopressin immunoreactive [70] (see also Ref. [17]). The native blowfly peptide(s) related to the cockroach leucokinins have not yet been isolated, but radioimmunoassays of blowfly tissue extract indicate the presence of leucokinin-immunoreactive material that eluted in HPLC in two zones, one of which has the same retention time as synthetic leucokinin I [53]. The biological actions of leucokinin-related peptides in blowflies are not known. Although present in fibers in the lateral cardiac nerves of L. maderae [70], none of the leucokinins are cardioactive in this cockroach [39]. Possibly these peptides are instead released in into the circulation via the pericardial septum in L. maderae and blowflies, and have targets elsewhere. Clues to such targets have been obtained from studies of the actions of some of the leucokinins and achetakinins in some other insects by in vitro assays. It is clear that apart from myotro- pic actions of the leucokinins at the cockroach foregut, hindgut and oviduct [39], some of the achetakinins and leucokinins can regulate fluid secretion in the Malpighian tubules of the mosquito Aedes aegypti and the cricket Acheta domesticus [30, 35]. It is to be expected that the leucokinin- related peptides from different insect species have diverse physiological roles (as neurohormones, as well as modulators within the CNS). A calcium dependent release of leucoki- nins could be induced in vitro from the corpora cardiaca of L. maderae by potassium depolarization [59], but was not yet tested for abdominal neurohemal areas. The hemolymph of L. maderae was also shown to contain leucokinin immuno- reactive material in the nanomolar range [59], indicating that leucokinins may be released as hormones in vivo. FMRFamide-like immunoreactive (F-LI) fibers have been detected in the wall of the abdominal aorta and the pericardial septum of Calliphora. The origin of these fibers was not determined, but they are probably originating from cell bodies in the abdominal neuromeres. As noted above, two of the CalliMRFamides increase the spontaneous heart rhythm [23] and it is possible that this effect is by peptide released from the F-LI efferents innervating the heart. Proctolin-like immunoreactive (P-LI) material has been detected in a relatively large number of neurons in the blowfly thoracic-abdominal ganglion [69]. Two sets of efferent P-LI neurons have cell bodies located in the abdominal neuro- meres (Fig. 8B). They send axons via the median and lateral abdominal nerves respectively. The approximately 10 P-LI neurons located laterally in each side of the abdomin- al ganglion send axons via the lateral abdominal nerves to the pericardial septum. Proctolin is known to increase the heart beat in a number of insect species [79], but no records of activity in flies exist. In different dipteran insects myogenic actions of proctolin have, however, been noted in hindgut and oviduct, and in larval body wall muscle [79]. Recent experiments reveal that an antiserum to the mammalian peptide galanin message associated peptide (GMAP) labels an extensive plexus of varicose fibers in the pericardial septum of Calliphora. So far, no biological ac- tion for this peptide has been recorded for any organism (Lundquist et al., 1992), but the action of GMAP on the blowfly heart can be probed quite easily. The cellular origin of the GMAP immunoreactive fibers in the septum has not been determined. Antisera to three more peptides, crustacean cardioactive peptide (CCAP), corazonin and pigment-dispersing hormone (PDH), were tested on the dorsal abdominal diaphragm and pericardial septum. Although the first two of these are cardioactive in other insects [11, 19, 41, 48, 107], no immuno- labeling was obtained with any of the three antisera in the blowfly pericardial septum. Serotonin immunoreactive fibers could, however, be detected in the pericardial septum of Phormia. These fibers are likely to be derived from the four large subesophageal cell bodies, via the superficial fiber system in the neural sheath emerging also through the sheath of the lateral abdominal nerves. A direct action of serotonin on the heart or alary muscle as indicated in other insects is Neuropeptides in Neurosecretory and Efferent Neural Systems 25 possible [58]. In conclusion it is not clear whether the substances indicated by immunocytochemistry in fibers in the blowfly pericardial septum act on the heart muscle (and alary mus- cles) or if they are released into the circulation for action elsewhere (or both). An exception may be the F-LI fibers since some of the CalliF MRFamides have been shown to be cardiactive in Calliphora [23]. VARICOSE AXONS IN THE NEURAL SHEATH OF PERIPHERAL NERVES In several insect there are plexuses of varicose fibers in the neural sheath of several of the peripheral nerve roots. For instance antisera against FMRFamide, pancreatic polypeptide, glucagon, adipokinetic hormone (AKH) and CCA label plexuses of varicose fibers in the perineurium of segmental and link nerves of crickets and locusts [19, 61, 86, 96, 99] (see also Fig. 2). Some of these fibers are derived from cell bodies in the CNS, others from peripherally located neurosecretory cells. In blowfies there are also systems of varicose fibers in the neural sheath of some of the peripheral nerves. These fibers have in some cases been traced from abdominal cell bodies to the periphery where they form terminals in neurohemal release areas: fibers in sheath of the lateral abdominal nerves, destined for the pericardial septum, react with antisera against antisera against leucokinin [10], proctolin, FMRFamide and lysine vasopressin [9, 69, 61]. In the sheath of the anterior prothoracic-, anterior dorsal mesothoracic- and haltere nerves superficial fibers react with antisera to AKH and the mammalian neuropeptides galanin and galanin message associated peptide (GMAP) [51, 52]. The origin of these fibers has not been determined. Further- more, 5-HTIR fibers from cell bodies in the subesophageal ganglion were seen in the sheath of most cephalic and thoracic-abdominal nerve roots [67, 68]. NEUROPEPTIDES IN EFFERENT NEURONS TO THE BLOWFLY HINDGUT The blowfly hindgut is innervated by abdominal neurons via the median abdominal nerve (Fig. 4). As will be shown below, several neuropeptides have been indicated in some of these efferent abdominal neurons by immunocytochemistry with antisera to proctolin, FaRPs, PDH, callatostatin and CCAP. It is not clear whether these peptides act on hindgut motility, water and ion balance or are released as neuro- hormones into the circulation around the intestine (or have several functions). Proctolin. Proctolin was isolated on basis of its myotro- pic action on the cockroach hindgut [98], and early on proctolin was detected immunocytochemically in six efferent abdominal neurons with terminals in muscle of the hindgut of Periplaneta americana [25]. It has also been shown that proctolin increases the frequency and amplitude of myogenic contractions of the hindgut of a dipteran insect, the stable fly Stomoxys calcitrans [37]. In blowflies four abdominal proc- tolin immunoreactive (P-LI) neurons (Fig. 8B) send axons via the median narve to the hindgut where P-LI terminals could be found on the hindgut, rectal valve, rectum and rectal papillae [9]. Some of the P-LI fibers innervate the muscular is of the intestine and in the rectal papillae the fibers invade the medullary region. The P-LI fibers may hence mediate control of muscle activity as well as regulation of water and ion balance. FMRFamide related peptides. ©Efferent F-LI nuerons of the abdominal ganglion send axons via the median abdominal nerve to the hindgut where they form an innervation pattern very similar to that of the proctolin containing neurons, but with larger number of arborizations [9, 50]. The Fa-LI fibers innervate the hindgut (Fig. 9C), rectal valve, rectum and rectal papillae. Also FaRPs may have myotropic actions on the hindgut and/or be involved in the regulation of water and ion balance. Pigment dispersing hormone (PDH). Members of the pigment dispersing hormone family of peptides have been isolated from a number of crustacean and insect species [87] and a partial amino acid sequence (12 of 18 amino acids) of a Calliphora peptide with strong homologies to $-PDH has been obtained (Lundquist et al., in prep.; see also Ref. [73]). In blowflies there are neurons reacting with PDH antiserum in the brain and thoracic abdominal ganglion [73]. In the abdominal ganglion six PDH-LI neurons send axons to the hindgut where they form varicose terminals in the posterior region of the midgut and anterior portion of the hindgut (Fig. 9D), the rectum and the rectal papillae [73]. The six PDH- LI neurons also have varicose arborizations within abdominal neuropil. An additional release site for PDH-related pep- tide may be the wall of the anterior aorta where PDH-LI terminals, derived from cephalic neurosecretory neurosn, are found [73]. Crustacean cardioactive peptide (CCAP). The nona- peptide CCAP was first isolated from the crab Carcinus maenas and later in identical form from the insects Locusta migratoria and Manduca sexta [11, 48, 97]. In the locust CCAP immunoreactivity (CCAP-LI) was demonstrated in numerous neurons and neurosecretory cells as seen in Figures 1 and 2 [19]. Some of these supply fibers to segmental distal perisympathetic organs and neurohemal release sites in the dorsal diaphragm including the alary muscles and lateral heart nerve [19]. In adult blowflies there are only four CCAP-LI cells in the entire fused thoracic-abdominal gang- lion, two of which are only weakly immunoreactive (Fig. 8C). These cells are located posteriorly in the abdominal portion of the ganglion and send axons to the hindgut via the median abdominal nerve (Fig. 8C). The CCAP-LI fibers supply only the hindgut and rectum, but not the rectal papillae or pouch. Similar CCAP-LI neurons occur in adult Drosophila (Dircksen and Breidbach, unpublished). Allatostatin-related peptides (callatostatins). Five neuro- peptides termed callatostatins 1-5 have been isolated from adults of the blowfly Calliphora vomitoria [24]. Two of 26 D. R. NAsSSsEL, E. BAYRAKTAROGLU AND H. DiRCKSEN Fic. 9. Fluorescence micrographs of peripheral peptidergic axons in blowfly. A. Leucokinin immunoreactive terminals in the pericardial septum (PS). The fibers arrive by the segmental nerves (one indicated by arrow). B. Leucokinin immunoreactive varicose fibers at the surface of the lateral abdominal nerves (LN) on route to the pericardial septum. C. FMRFamide immunoreactive fibers in the hindgut anterior to rectal valve. The axons arrive from the abdominal nerve at arrow. D. PDH immunoreactive fibers in first part of the rectum (Re) of the hindgut (rectal valve at arrows). Scales: A-D=50 um. A-D altered from [9, 10, 73]. these were specifically isolated from dissected thoracic- abdominal ganglia. In the cockroach Diploptera punctata the cockroach allatostatins as well as the fly callatostatins inhibit juvenile hormone production in the corpora allata in vitro, whereas in the adult blowfly these peptides have no action on the production of juvenile hormone in the corpora allata [24]. In accordance with this Duve er al. [24] found no callatostatin immunoreactive material in the corpora allata or Neuropeptides in Neurosecretory and Efferent Neural Systems 27 any of the neurosecretory cell systems of the blowfly nervous system. Instead, callatostatin immunoreactivity was found in neurons of the abdominal ganglion with axons emerging through the median abdominal nerve to terminals in the hindgut, rectum, rectal papillae and oviduct. These authors therefore suggest that functions other than allatostatic ones may have to be sought for the callatostatins in adult blowflies. In sharp contrast to the findings of Duve et al. [24], allatosta- tin-like immunoreactivity was, however, found in neurons and neurosecretory cells both in the brain and thoracic- abdominal ganglion of Drosophila by Zitnan et al. [111]. It is not clear whether this discrepancy is caused by species differences or methodological differences. LARVAL NEUROHEMAL ORGANS AND PERIPHERAL RELEASE SITES IN BLOWFLIES In the larvae of higher diptera the corpora cardiaca, corpora allata and prothoracic gland form a composite organ, the ring gland or Weissman’s ring [108]. In Calliphora serotonin and gastrin/CCK immunoreactive cell bodies and fibers were detected in the ring gland [8]. Additionally cephalic neurosecretory cells reacting with antisera against gastrin/CCK [8], FMRFamide, PDH (Nassel, unpublished) and corazonin (Cantera, Veenstra, Nassel, in prep.) form plexuses of varicose axons in the wall of the anterior portion of the aorta. The presence of neurohemal release sites in the larval thoracic-abdominal nervous system of Drosophlia and Cal- liphora was first indicated by immunocytochemistry with antisera against FMRFamide and gastrin/CCK [65, 72, 101, 109]. In these flies each of the three thoracic segments have a dorsal unpaired median nerve which contributes to an extended bulb-like neurohemal organ [65, 72] (Fig. 10). In some specimens there is one neurohemal organ on each dorsal unpaired nerve, in others these fuse to one or two organs. FMRFamide-, myomodulin- and CCK-like immuno- reactive fibers invade these thoracic neurohemal structures via the dorsal unpaired nerves (Fig. 11A) and peptidergic terminals could be revealed in the neural sheath outside the blood brain barrier [72]. The immunoreactive fibers are derived from two large cell bodies ventrally in each thoracic segment (Fig. 11A). By analysis of the postembryonic de- velopment of these immunoreactive cells it could be demons- trated that they persist throughout metamorphosis (see Fig. 6B) and form the six VI'NCs of the adult blowfly and fruitfly [72, 101]. Interestingly, it can thus be concluded that the segmentally organized larval thoracic neurohemal organs transform into a large fused neurohemal area. There are also varicose myomodulin immunoreactive fibers emerging through the first four dorsal unpaired nerves (A1-4) of the abdominal ganglion in larvae of Calliphora (Fig. 11A) and Phormia (Nassel, unpublished). These nerves bifurcate laterally and the immunoreactive fibers continue along the length of the branches. The fibers entering the first four abdominal dorsal median nerves are derived from five pairs of ventral myomodulin immunoreactive cell bodies in abdomin- al neuromeres Al-5 (Fig. 11A). Four of these probably correspond to the myomodulin immunoreactive VANCs seen in the adults. As mentioned earlier the larval VINCs of Drosophila were shown to contain peptide reacting with antisera to allatostatin and allatotropin [111]. Peptide containing cells with abdominal cell bodies and peripheral axonal projections can also be detected in the larval blowflies. Some of them have been followed through metamorphosis. The neurons of interest react with antisera Fic. 10. The larval nervous system of Calliphora. Scanning electron micrograph of the larval nervous system in dorsal view. The three thoracic neurohemal organs (TNOs) are seen as spherical structures. A neurohemal organ (ANO) is also formed by the first abdominal dorsal unpaired nerve. The remaining abdominal median unpaired dorsal nerves (MUDNs) also contain some varicose fibers from neurosecretory cells (myomodulin immunoreactive). Altered from [72]. BH=brain hemishere. RG=portion of the ring gland. Scale=100 pm. 28 D. R. NASSEL, E. BAYRAKTAROGLU AND H. DiIRCKSEN 100 um Fic. 11. Peptidergic neurons in the larval blowfly nervous system. A. Myomodulin immunoreactive neurons located ventrally (some dorsal cells in the brain are not shown). Neurons are found in two of the three subesophageal segments (S1 and S3). The three VINCs are seen in segments T1-T3. These cells send their axons to the three median neurohemal organs of the thoracic neuromeres (location indicated by asterisks). Five pairs of ventral abdominal neurosecretory cells (A1—5) send their axons into the dorsal unpaired median nerves (location of one indicated by asterisk). B. Leucokinin immunoreactive neurons in the brain and abdominal neuromeres. The seven pairs of abdominal neurons (AE1-7) send axons through the segmental nerves 1-7 (arrows). The axons terminate on segmental abdominal body wall muscles. C. Pigment dispersing-hormone immunoreactive neurons in brain and thoracic-abdominal ganglia. The 8 pairs of thoracic and abdominal neurons (T1-3 and Al—5) appear to be interneurons, whereas three pairs of posterior abdominal neurons (AE) are efferents probably destined for the hindgut. A-C (Nassel, unpublished). against proctolin, lysinevasopressin [71], leucokinin [10] and PDH (Nassel, unpublished). The leucokinin (and lysine vasopressin) immunoreactive cells of the larva are segmentally distributed laterally in the abdominal neuromeres A1-7 (one pair of cells per neuro- mere, except in A8; Fig. 11B). These cells form efferent axons projecting to segmental muscle of the abdominal body wall of the larva [10]. The abdominal LK-LI and vasopres- sin immunoreactive cells survive metamorphosis and their peripheral axons innervate the pericardial septum as des- cribed above [10, 71]. A proctolin antiserum labels four cells with axons in the median abdominal nerves (A8) and seven pairs of lateral neurons with axons emerging through the lateral nerves (A1-7) in the blowfly and fruitfly larvae [1, 71]. In Dro- sophila the lateral cells send axons to body wall musculature and in both insects the median one innervate the intestine [1, 71]. Both sets of abdominal P-LI cells survive metamorpho- sis and in adults supply the pericardial septum and hindgut respectively. In the blowfly larva six efferent PDH-LI neurons can be seen in the caudal portion of the abdominal ganglion (Fig. 11C). It has not yet been determined whether these cells survive metamorphosis and transform into the six cells of the adults that innervate the hindgut [73]. In summary, it can be proposed that many of the neurosecretory and efferent cell systems of the blowfly are present already in the larva. The cells survive metamorpho- sis, attain slightly altered morphologies and form novel release sites. It is not known whether the functions of the larval neurosecretory systems and the peptides they release are retained into the adult organism or if they obtain new actions. CONCLUSIONS Insects possess an impressive set of multiple specialized neurohemal release sites in different portions of the head and body compartments. Raabe [86] suggested that the multiple release sites are necessary due to the poorly developed circulatory system of insects. Thus it is not unusual to detect putative release sites for the same neuropeptide in the corpora cardiaca, the thoracic and abdominal perisympathe- tic organs and in the abdominal heart region (pericardial septum and alary muscles) [19, 49]. In this account we have shown that also in flies such as Calliphora, Phormia and Drosophilla there are several neurohemal release sites in addition to the corpora cardiaca. In fact, in these flies the Neuropeptides in Neurosecretory and Efferent Neural Systems 29 neurohemal part of the corpora cardiaca is insignificant in size compared to the neurohemal release sites in the neural sheath of the thoracic-abdominal ganglion and the pericardium. Thus the release of peptides from corpora cardiaca and the attending neurohemal plexus in the anterior aorta may be small in comparison with that of neurohemal release sites in structures derived from the body segments. Many of the neurosecretory cells in addition to periphe- ral release sites have putative release sites within neuropils of the central nervous system. An interesting example is pro- vided by the eclosion hormone containing neurons of larval moths (Manduca sexta). These neurons have varicose axons running through the entire length of the cephalic and tho- racic-abdominal ganglia before they reach their neurohemal release sites on the hindgut [103]. Another example are the blowfly and Drosophila VYNCs which have extensive central arborizations in the subesophageal and thoracic neromeres as well as terminals in a substantial neurohemal release site [50]. Release of peptides or monoamines by the same neurons at different peripheral neurohemal release sites in addition to central neuropil regions would ensure synchronous action on (regulation of) peripheral targets and central circuits and thus enable orchestration of behavioral routines or other physiolo- gical functions [5, 45, 104, 105]. Some peptidergic neurons, that were originally classified as neurosecretory cells have all their known processes within the central nervous system [88, 100]. Are they still to be considered as neurosecretory cells? With the data available today on the distribution of a large number of different neuropeptides in many types of interneurons it may be called for to be cautious about terminology. Scharrer [90] noted that neurons employing chemical messenger substances have secretory capacity and may, in case of interneurons, release their regulatory substances in certain distinct neuropils with- out forming typical synaptic contacts. However, with the emergence of the concept of colocalized neuropeptides and classical transmitters [49] the classification into neurosecre- tory cells and neurons may be obsolete. Neuropeptides in central neurons may have a host of actions as neurotransmit- ters, cotransmitters, neuromodulators or even as trophic factors [36] and in insects we are only just starting to learn about central functions. Neuropeptide release often is epi- sodic [36, 90, 104, 105] and many neuropeptides act in pacemaker circuits [34] indicating that central roles of neuropeptides are in regulation of rhythmic events or trigger- ing of innate behaviors. A large number of neuropeptides have been isolated from crustaceans and insects [38, 42, 55, 56, 94], and probably most of them have some actions at peripheral targets. Why are sO many peptides needed in a “simple” organism like an insect? It is likely that many regulatory processes are finely tuned and that this requires several chemical messengers. Studies on diuresis in insects have for example provided evidence that more than one peptide may be involved only at the level of the Malpighian tubules [4, 30]. In the cricket A. domesticus fluid secretion can be induced by two different types of neuropeptides, achetakinin I and the 46 amino acid diuretic peptide [30]. The two peptides act via different receptors and second messenger systems in the Malpighian tubules, achetakinin by an unknown pathway and diuretic peptide via cAMP. Additionally serotonin acts on cricket Malpighian tubules [12] indicating that diuresis is a finely tuned proces in insects, notwithstanding further medhanisms involved in water reabsorption in the hindgud (See Ref. [4]). We also know that a variety of neuropeptides in addition to monoamines act on skeletal, visceral and heart muscle in insects [16, 82, 94]. For instance the spontaneous contrac- tions of locust oviduct muscle appear regulated by a number of neuropeptides as well as by octopamine [46, 47, 82, 94] and a very large number of neuropeptides and serotonin act on the Leucophaea hindgut [39, 94]. The role of insect hor- monal neuropeptides in regulation and initiation of behaviors has been explored for eclosion hormone and cardioactive peptides during development of the moth Manduca [104, 105]. It is to be expected that for instance feeding behavior is under peptide hormone control since salivary glands, intestinal muscle and skeletal muscle are regulated by differ- ent peptides and central circuits associated with feeding are innervated by peptidergic neurons [16, 22, 66, 80, 94]. Peptides originating from cells outside the nervous system have also been implicated in behavior regulation. As an example the Drosophila 36 amino acid sex peptide of the male accessory glands is transferred to females during copula- tion and elicit rejection of further males as well as an ovulation and oviposition response [91]. Characterization of receptors mediating the action of different neuropeptides and studies of the structure of the active cores of neuropeptides have been initiated [30, 44, 56, 62, 85], but much is to be learned about interactions of different peptides at the receptor and second messenger level. We also need to know more about the degradation of peptides at their target organs such as Malpighian tubules, hindgut and ovaries [83, 84] or within the hemolymph [29]. These studies are necessary to be able to determine whether peptides released from neurohemal organs will be available in physiological concentrations for actions at their targets (at threshold concentrations indicated in in vitro assays) or if they are likely to act by direct release from neuronal terminals supplying the targets. It is also critical to employ sensitive assays such as radioimmunoassay or ELISA to determine the content of peptides in the hemolymph in in vivo experiments to corroborate claimed hormonal actions. Analysis of neuroendocrine systems at the molecular and genetic level will no doubt be helpful in filling many gaps in the understanding of hormonal control of development, behaviour and homeostasis [6, 95]. It is to be expected that some of the diversity in neuropeptides and peptides in non-neural cells will be explained by a certain redundancy of mediators in regulatory systems which is partly caused by a need for subtle control mechanisms. It is also likely that peptides are involved in processes other than classical neuro- transmission, hormonal actions or neuromodulatory roles, 30 D. R. NASsEL, E. BAYRAKTAROGLU such as being cytokines or factors in development and mainte- nance of the nervous system. ACKNOWLEDGMENTS Our colleagues in Bonn and Stockholm are thanked for their contributions to much of the original research presented. The technical assistance of Anne Karlsson, Ylwa Lilliemarck and Bibbi Meyrhofer is gratefully acknowledged. Financial support by the Swedish Natural Science Research Council is also acknowledged. REFERENCES Anderson MDS, Halpern ME, Keshishian H (1987) J Neuros- ci 8: 283-294 Baudry-Partiaoglu N (1975) CR Acad Sci Paris 281: 921-924 Bayrahtaroglu E, Golding DW, Whittle AC (1989) Acta Zool (Stockholm) 70: 87-94 Beyenback KW (Ed.) (1993) “Structure and function of prim- ary messengers in invertebrates: insect diuretic and antidiuretic peptides”. Mol Comp Physiol, Vol. 12, Karger, Basel Bicker G, Menzel R (1989) Nature 337. 33-39 Buchner E (1991) J Neurogenet 7: 153-192 Buma P, Roubos EW, Buijs RM (1984) Histochemistry 80: 247-257 Cantera R (1988) Cell Tissue Res 253: 425-433 Cantera R, Nassel DR (1991) Comp Biochem Physiol 99C: 517-525 Cantera R, Nassel DR (1992) Cell Tissue Res 269: 459-471 Cheung CC, Loi PK, Sylwester AW, Lee TD, Tublitz NJ (1992) FEBS Lett 313: 165-168 Coast G (1989) Physiol Entomol 14: 21-20 Cook BJ, Holman GM (1979) J Gen Physiol 74: AS—A6 Cook BJ, Meola S (1978) Physiol Entomol 3: 273-280 Cropper EC, Tenenbaum R, Kolks MAG, Kupferman I, Weiss KR (1987) Proc Natl Acad Sci USA 84: 5483-5486 Cuthbert BA, Evans PD (1989) J Exp Biol 144: 395-415 Davis NT, Hildebrand JG (1992) J Comp Neurol 320: 381- 393 De Loof A (1987) Entomol Exp Appl 45: 105-113 Dircksen H, Miiller A, Keller R (1991) Cell Tissue Res 263: 439-457 Dircksen H, Nassel DR (1993) In “Gene -Brain- Behaviour” Ed by N Elsner, M Heisenberg, Thieme, Stuttgart, p 519 Duve H, Thorpe A, Nassel DR (1988) Cell Tissue Res 253: 583-595 Duve H, Johnsen AH, Sewell JC, Scott AG, Orchard, I, Rehfeld JF, Thorpe A (1992) Proc Natl Acad Sci USA 89: 2326-2330 Duve H, Elia AJ, Orchard I, Johnsen AH, Thorpe A (1993) J Insect Physiol 39: 31-40 Duve H, Johnsen AH, Scott AG, Yu CG, Yagi KJ, Tobe SS, Thorpe A (1993) Proc Natl Acad Sci USA 90: 2456-2460 Eckert M, Agricola H, Penzlin H (1981) Cell Tissue Res 217: 633-645 Elia AJ, Tebrugge VA, Orchard I (1993) J Insect Physiol 39: 459-469 Ferber M, Pfltiger HJ (1992) Cell Tissue Res 267: 85-98 Finlayson LH, Osborne (1968) J Insect Physiol 14: 1793-1801 Fox AM, Reynolds SE (1991) Peptides 12: 937-944 Goldsworthy G, Coast G, Wheeler C, Cusinato O, Kay I, Khambay B (1992) In “Insect molecular science” Ed. by JM Crampton, P Eggleston, Academic Press, London, pp 205-225 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 Sl 52 53 54 55 56 57 AND H. DIRCKSEN Grillot JP (1977) Int J Insect Morphol Embryol 6: 303-343 Grillot JP (1983) In “Neurohemal organs of arthropods” Ed. by AP Gupta, CC Thomas, Spring-field, Illinois, pp 481-512 Gupta AP (Ed.) (1983) “Neurohemal organs of arthropods”. CC Tomas, Spring-field, Illinois Harris-Warrick R, Marder E (1991) Annue Rev Neurosci 14: 39-57 Hayes TK, Pannabecker TL, Hinckley DJ, Holman GM, Nachman RJ (1989) Life Sci. 44: 1259-1266 HO6kfelt T (1991) Neuron 7: 867-879 Holman GM, Cook BJ (1979) Comp Biochem Physiol 62C: 231-235 Holman GM, Nachman RJ, Wright MS (1990b) Annu Rev Entomol 35: 201-217 Holman GM, Nachman RJ, Schoofs L, Hayes TK, Wright MS, De Loof A (1991) Insect Biochem 21: 107-112 Homberg U, Davis NT, Hildebrand JG (1991) J Comp Neurol 303: 35-52 Jahn G, Kauser G, Koolman J (1991) Gen Comp endocrinol 82: 287 Keller R (1992) Experientia 48: 439-448 Kingan TG, Blackburn MB, Raina AK (1992) Cell Tissue Res 270: 229-240 Konopinska D, Rosinski G, Sobdtka W (1992) Int J Pept Protein Res 39: 1-11 Kravitz EA (1988) Science 241: 1775-1781 Lange AB, Orchard I, Adams ME (1986) J Comp Neurol 254: 279-286 Lange AB, Orchard I, Te Brugge V (1991) J Comp Physiol A 168: 383-391 Lehman HK, Murgiuc CM, Miller TA, Lee TD, Hildebrand JG (1993) Peptides 14: 735-741 Lundberg JM, Hokfelt T (1983) Trends Neurosci 6: 325-333 Lundquist CT, Nassel DR (1990) J Comp Neurol 294: 161- 178 Lundquist CT, Rokaeus A, Nassel DR (1991) J Comp Neurol 312: 77-96. Lundquist CT, Rékaeus A, Nassel DR (1992) J Neuroendoc- rinol 4: 605-616 Lundquist CT, Brodin E, Muren JE, Nassel R (1993) Pep- tides 14: 651-663 Maddrell S, Herman WS, Mooney RL, Overton JA (1991) J Exp Biol 156: 557-566 Masler EP, Kelly TJ, Menn JJ (1993) Arch Insect Biochem Physiol 22: 87-111 Menn JJ, Kelly TJ, Masler EP (Eds.) (1991) Insect Neuropeptides: Chemistry, Biology and Actions, ACS Sympo- sium Series 453. American Chemical Society, Washington Meola S, Cook BJ (1986) Int J Insect Morphol Embryol 15: 269-281 Miller TA (1979) Am Zool 19: 77-86 Muren JE, Lundquist CT, Nassel DR (1993) J Exp Biol 179: 289-300 Myers CM, Evans PD (1985) J Comp Neurol 234: 1-16 Myers CM, Evans PD (1988) J Morphol 195: 45-58 Nachman RJ, Holman GM, Haddon WF (1993) Arch Insect Biochem Physiol 22: 181-197 Nagasawa H (1992) Experientia 48: 425-430 Nambu JR, Murphy-Erdosh C, Andrews PC, Feistner GJ, Scheller RH (1988) Neuron 1: 55-61 Nassel DR (1987) In “Nervous systems of Invertebrates” Ed. by MA Ali, Plenum Press, New York, pp 171-212 Nassel DR (1993) Cell Tissue Res 273: 1-29 Niassel DR, Elekes K (1984) Neurosci Lett 48: 203-210 68 69 70 71 2 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 Neuropeptides in Neurosecretory and Efferent Neural Systems 31 Nassel DR, Elekes K (1985) Neurosci 15: 293-307 Nassel DR, O’Shea M (1987) J Comp Neurol 265: 437-454 Nassel DR, Cantera R, Karlsson A (1992) J Comp Neurol 322: 45-67 Nassel DR, Holmqvist BI, Movérus BJA (1989) J Comp Neurol 283: 450-463 Nassel DR, Ohlsson LG, Cantera R (1988) J Comp Neurol 267: 343-356 Nassel DR, Shiga S, Mohrherr CJ, Rao KR (1993) J Comp Neurol 331: 183-198 Nassel DR, Cantera R, Johard HAD, Lundquist CT, Muren JE, Shiga S (1992) In “Nervous systems: principles of design and function”. Ed. by RN Singh, Wiley Eastern, New Dehli, pp 189-212 Nichols R (1992a) Molec Cell Neurosci 3: 342-347 Nichols R (1992b) J Mol Neurosci 3: 213-218 Nichols R, Schneuwly SA, Dixon JE (1988) J Biol Chem 263: 12167-12170 O’Brien MA, Schneider LE, Taghert PH (1991) J Comp Neurol 304: 623-638 Orchard I, Belanger JH, Lange AB (1989) J Neurobiol 20: 470-496 O’Shea M, Schaffer M (1985) Ann Rev Neurosci 8: 171-198 Paemen L, Schoofs L, De Loof A (1992) Cell Tissue Res 268: 91-97 Peeff NM, Orchard I, Lange AB (1993) J Insect Physiol 39: 207-215 Picquot M, Proux J (1980) Regul Pept 31: 139-156 Puiroux J, Loughton BG (1992) Arch Insect Biochem Physiol 19: 193-202 Puiroux J, Pedelaborde A, Loughton BG (1992) Insect Biochem 22: 547-551 Raabe M (1989) Recent developments in insect neurohor- mones. Plenum, New York Rao KR, Riehm JP (1989) Biol Bull 177: 225-229 Remy C, Girardie J (1980) Gen Comp Endocrinol 40: 27-35 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 Richard O, Girardie J (1992) Cell Tissue Res 270: 587-596 Scharrer B (1987) Ann Rev Neurosci 10: 1-17 Schmidt T, Choffat Y, Klauser S, Kubli E (1993) J Insect Physiol 39: 361-368 Schneider LE, Taghert PH (1988) Proc Natl Acad Sci USA 85: 1993-1997 Schneider LE, Roberts MS, Taghert PH (1993) Neuron 10: 279-291 Schoofs L, Vanden Broeck J, De Loof A (1993) Insect Biochem Molec Biol 23: 859-881 Segal D (1993) Arch Insect Biochem Physiol 22: 199-231 Sp6rhase-Eichman U, Dircksen H, Hect T, Helle J, Schiirmann FW (1991) In “Synapse —Transmision— Modulation” Ed. by N Elsner, H Penzlin, Thieme, Stuttgart, p 339 Stangier J, Hilbich C, Keller R (1989) J Comp Physiol B 159: 5-11 Starrat AN, Brown BE (1975) Life Sci 17: 1253-1256 Taghert PH, Carr JN, Wall JB (1988) Adv Insect Physiol 20: 87-117 Thompson KSJ, Tyrer NM, May ST, Bacon JP (1991) J Comp Physiol A 168: 605-617 Trimmer BA (1985) J Exp Biol 114: 307-328 Truman JW (1990) J Neurobiol 21: 1072-1084 Truman JW, Copenhaver PF (1989) J Exp Biol 147: 457-470 Tublitz NJ, Copenhauer PF, Taghert PH, Truman JW (1986) Trends Neurosci 9: 359-363 Tublitz N, Brink D, Broadie KS, Loi P, Sylwester AW (1991) Trends Neurosci 14: 254-259 Veenstra JA (1987) Neurosci. Lett. 73: 33-37 Veenstra JA (1989) FEBS Lett 250: 231-234 Vogt M (1943) Biol Zentralbl 63. 56-71 White K, Hurteau T, Punsal P (1986) J Comp Neurol 247: 430-438 Zitnan D, Sauman I, Sehnal F (1993) Arch Insect Biochem Physiol 22: 113-132 Zitnan D, Sehnal F, Bryant PJ (1993) Dev Biol 156: 117-135 Os TRG al poet BaD : bis 17 ix t. deed ; = #& a 42 é = wes - +h . v2.7 oH a aayary J ne) GE: Veet sti mmsigne kong ttn (St) | imarai 19 ease, POETS: ortngl 2 Ff intiodD "F tae 1S OE atleast ’ | oe ia Wi gealed LA vsihaadted Te |..007 8 a, rs iv be 2A eeepea 4S brite? by: sl ee 7 reel ivshiwv (RAARBILS ine oshitAGpdiedy the SOF orn Rak | iE wi sitiess, UN ohesandge Se Debate ana oy r ce asi pedo] aria} seurietege ro a7) ety eal (LO Wd a? ites H sett Hd Qantibabsemee,2 [dit *¢3 ia Pea ht ihe Jove GHG, ol we sth produaT y crite ol AAT Ane, tr tO. 891 AA a fh AML Ad: hah? ueaX Wed G2 ower] i Ansa + Wl. nemasy 1. iL poqualesind). nui ol fi a KaPviuse aires | A-mltun® AI gente wait .\~ et £e ek pt eat | Achaia, ten [ pran CHAS 108M Dal Deke tee, & wil Stier BA ibe) 4 it 4 % A aw 4 ws Aires Cl mews een jel j AmITTLS le -) © Lah, he een POE qe - SRE si eae af ve Gant h, pa) te tatiery 4 = wera 0 : tae Bel eae innayitt ey hom ropa ah Lalit PE polis mrs A i NOD, : be Sal A ea e ht ths lon yopef qr? wat (0% ol hive JE 5.2 i “a bY yi whiney ny » } t).. Mew J as 2 5 a e ‘a a i Miler FF ; ‘ i r; ting eee 5A < > LUC cok ¢ “rw _ 2) on ) ied Ti ® r ite ‘i TE Mtv) ORY tn CSOT 4 Cuan ab 2 ot i Morphot 1a nlertsg VME Piaddon “WE HGR 17.4 ee omy” / tink pina Acdww BE wh Mina! saa ‘Gla 5 (te ae Breas Ws evo Of t i" ip l‘y~er New You. ry (Us ‘ ies FV 4A tuten 6) Fed) Novroacs Lest & L —_ ? J "| @.of lbe@ f ZOOLOGICAL SCIENCE 11: 33-38 (1994) A Myomodulin-CARP-related Peptide Isolated from a Polychaete Annelid, Perinereis vancaurica TosHio TAKAHASHI’, OSAMU MATSUSHIMA2*, FUMIHIRO MorisHITA~, Masaaki Fuyimoto*, TETSUYA IKEDA?, H1royYUKI MINAKATA® and KyosukE Nomoto* ‘Faculty of Integrated Arts and Sciences, Hiroshima University, Higashi-Hiroshima 724, Department of Biological Science, Faculty of Science, Hiroshima University, Higashi-Hiroshima 724, and *Suntory Institute for Bioorganic Research, Osaka 618, Japan ABSTRACT—Myomodulin-CARP-family peptides have been isolated only from molluscs. In the present study, a heptapeptide, Ala-Met-Gly-Met-Leu-Arg-Met-NH>, termed Pev-myomodulin, was isolated from a polychaete annelid, Perinereis vancaurica using the esophagus of the animal as the bioassay system. The sequence of the annelid peptide is highly homologous with those of the myomodulin-CARP-family peptides found in molluscs. The annelid peptide is regarded as a member of the myomodulin-CARP family, though all the molluscan peptides have a Leu-NH) at their C-termini. The annelid peptide showed a potnet contractile action on the esophagus of the annelid. The peptide may be an excitatory neuromediator involved in the regulation of the esophagus. Among various myomodulin-CARP-family peptides and their analogues, the annelid peptide showed the most potent contractile action on the esophagus. Replacement of the C-terminal Met-NH)p of the annelid peptide with a Leu-NH> decreased its contractile potency, while replacement of the C-terminal Leu-NH2 of myomodulin and CARP with a Met-NH)p increased their potency. The C-terminal Met-NH, of the annelid peptide seems to be important, but not essential, for exhibiting its contractile activity on the esophagus. On the anterior byssus retractor muscle of the bivalve mollusc Mytilus edulis, the annelid peptide showed catch-relaxing and contraction-modulating effects qualitatively similar to those of the authentic peptide CARP, © 1994 Zoological Society of Japan though the annelid peptide was less potent than CARP. INTRODUCTION Over the past two decades, a large number of bioactive peptides have been isolated from invertebrates, especially from arthropods and molluscs [5, 9, 12, 18, 20. 21]. As to annelids, however, only several FMRFamide-related pep- tides have been identified using radioimmunoassay for detec- tion of the peptides. Krajniak and Price [13] found FMRFa- mide in the polychaete Nereis virens. Baratte et al. [1] isolated FMRFamide and its analogue from Nereis diversico- lor. Evans et al. [4] identified FMRFamide and four analo- gues in the medicinal leech Hirudo medicinalis. In the previous study, we isolated two S-Iamide-family peptides, AKSGFVRIamide and VSSFVRIamide, from the polychaete annelid, Perinereis vancaurica [16]. The pep- tides show a potent contractile effect on the esophagus of the annelid [16], while FMRFamide shows an inhibitory effect [14]. This was the first finding of annelid peptides which were not considered to be FMRFamide-family peptides. Since this finding, we have continued to search for bioactive peptides in the annelid P. vancaurica using its esophagus as the bioassay system. In the present study, we found a novel heptapeptide, Ala-Met-Gly-Met-Leu-Arg-Met-NH>, _ that showed a potent excitatory action on the esophagus. The Accepted October 27, 1993 Received October 4, 1993 * To whom correspondence should be addressed. sequence of the peptide is highly homologous to myomodu- lin, catch-relaxing peptide (CARP) and their related pep- tides, all of which have been isolated from molluscs ([2, 3, 8, 10, 11], See also Table 1). That is, the annelid peptide is regarded as a member of myomodulin-CARP family. Here, we report purification, structure determination and pharma- cological characterization of the annelid myomodulin-CARP- family peptide. MATERIALS AND METHODS Animals The marine polychaete annelid worms, Perinereis vancaurica, were purchased from a fishing-bait store and the sea mussels, Mytilus edulis, were collected in the Hiroshima Bay. These animals were kept in laboratory tanks filled with aerated seawater at 15°C. Extraction and purification Extraction procedures for bioactive peptides in P. vancaurica were essentially the same as those reported previously [16]. Briefly, 0.5 kg of the animals were boiled for 10 min at 100°C in 21 of water containing 4% acetic acid and homogenized with a Waring blender and then with a Polytron homogenizer. The homogenate was centrifuged at 15,000Xg for 40 min at 4°C. The supernatant was concentrated with a rotary evaporator, and 1/10 volume of 1 N HCl was added to the concentrated solution with constant agitation. The concentrated solution was centrifuged again. The supernatant was applied to two C-18 cartridges (Mega Bond Elut, Varian) in series. After the cartridges were washed with 0.1% trifluoroacetic acid (TFA), the retained material was eluted with 50% methanol. The 34 T. TAKAHASHI, O. MATSUSHIMA et al. eluate was applied to four steps of reversed-phase and cation- exchange high performance liquid chromatography (HPLC). The eluting substances were monitored with an UV detector at 220 nm. Each of the fractions obtained at each HPLC step was bioassayed using the isolated esophagus of P. vancaurica. At the first HPLC-purification step, the eluate from the C-18 cartridges was applied to a Capcell-Pak C-18 reversed-phase column (Shiseido, 10x250mm). The column was eluted with a 120-min linear gradient of 0-60% acetonitrile (ACN) in 0.1% TFA (pH 2.2). The flow rate was 1ml/min. An aliquot (1/250) of each 2-ml fraction was evaporated to dryness, dissolved in artificial seawater (ASW) and subjected to bioassay. A bioactive peak eluted at around 20% ACN was then subjected to the second step of HPLC purification. At this step, another C-18 reversed-phase column (ODS-80TM, Tosoh, 4.6150 mm) was used. The column was eluted with a 50-min linear gradient of 15-25% ACN in 0.1% TFA. A bioactive peak was observed at around 16% ACN. At the third step, the bioactive substance was applied to a cation-exchange column (SP-5PW, Tosoh, 7.575 mm), and the column was eluted with a 70-min linear gradient of 0-0.7 M NaCl in 10 mM phosphate buffer (pH 7.1) at a flow rate of 0.5 ml/min. The bioactive sub- stance was eluted at around 0.22 M NaCl. At the fourth step (final step), the bioactive substance was applied to the reversed-phase column (ODS-80TM), and the column was eluted with a 25-min linear gradient of 15-25% ACN in 0.1% TFA at a flow rate of 0.5 ml/min. The substance was eluted at around 20% ACN as a single absorbance peak (Fig. 1). The purified bioactive substance was subjected to peptide se- quence analysis (Shimadzu PSQ-1 Protein Sequencer) and fast atom bombardment mass spectrometric (FAB-MS) analysis (JEOL JMS- HX 110/110A). These analyses suggested that the substance is a pentapeptide with an amidated C-terminus. The peptide having the suggested structure was then synthesized by a solid-phase peptide synthesizer (Applied Biosystems 430A) and purified by HPLC. The structure of the synthesized peptide was confirmed by amino acid sequence analysis and FAB-MS analysis. The synthetic peptide was compared with the native one in the behavior on HPLC. Bioassay The anterior half of the body was cut open ventrally and the esophagus (about 15mm long) was excised. Both ends of the isolated esophagus were ligated with cotton threads. One of the thread was tied to a fixed support of a trough (2 ml) filled with aerated ASW, and the other was connected to a force-displacement transducer attached to a manipulator. Changes in the esophagus tension in response to bioactive substances were recorded on a chart recorder through an amplifier. Aliquots (1/250-1/100) of the fractions obtained at each HPLC purification step were evaporated to dryness, dissolved in 0.1 ml ASW and injected into the aerated trough in which the isolated esophagus was mounted. After each recording of the effects of the fractions, the esophagus was washed with ASW. The next test was started 10 min after the recording. Pharmacology The contractile actions of 10 myomodulin-CARP-family pep- tides including the annelid peptide isolated in the present study and nine synthetic analogues were examined on the isolated esophagus of P. vancaurica. It has been shown that CARP powerfully relaxes catch tension in the anterior byssus retractor muscle (ABRM) of M. edulis and that the peptide potentiates phasic contraction of the ABRM in response to repetitive electrical stimulation at_ lower concentrations and inhibits the contraction at higher concentrations [6, 7, 18]. Therefore, the actions of the annelid peptide on catch tension and phasic contraction of the ABRM were also examined to compare them with those of CARP. Catch contraction was pro- duced by applying 10-* M acetylcholine (ACh) to the ABRM for 2 min at 20min intervals and phasic contractions was elicited by stimulating the muscle with repetitive electrical pulses (15 V, 3 msec, 10 Hz, 50 pulses) at 10 min intervals. This procedure was basically the same as that of Muneoka and Twarog [19]. Saline The saline used for the esophagus of P. vancaurica and the ABRM of M. edulis was ASW of the following composition: 445 mM NaCl, 10 mM KCl, 10 mM CaCl, 55 mM MgCl and 10 mM Tris- HCI (pH 7.6). RESULTS After the three steps of HPLC purification, the final purification was performed with the reversed-phase column (Fig. 1). The single absorbance peak obtained was found to coincide with an active peak. At the first step of HPLC purification on a Capcell-Pak C-18 reversed-phase column, the active substance was recovered in the same fraction (No. 34) as a_ previously reported S-Iamide peptide, VSSFVRIamide [16]. At the second step of HPLC using another C-18 reversed-phase column (ODS-80TM), these two substances were separated from each other: VSSFVRIamide was recovered in fractions 7, 8, and 9, while the present substance was in fractions 12 and 13. o A Absorbance at 220 nm jo) Oo WN [oe] De) (on CH3CN (°) 0 20 40 Time (min) a4 lmin iv ‘| 0.39 ri ESS mI — Ln A Fic. 1. HPLC profile at the final step of purification of the bioactive substance on the C-18 ODS-80TM column (upper) and the activity of the purified substance on the esophagus of P. van- caurica (lower). The elution was performed with a linear gradient of 15-25% ACN (0.1% TFA) at a flow rate of 0.5 ml/ min. The absorbance was monitored at 220nm. The broken line shows the concentration of ACN. An aliquot (1/100) of the purified substance was applied at the time indicated by an arrow head. The determined sequence and detected amount (pico- moles) of each amino acid in the amino acid sequence analysis of the purified substance were as follows: Ala (820.6)-Met Myomodulin-CARP-related Annelid Peptide 35 (892.2)-Gly (508.6)-Met (808.2)-Leu (729.2)-Arg (114.3)- Met (293.1). In the FAB-MS spectrum of the purified substance, a molecular ion peak was observed at 810.1 m/z (M+H)*. The results of these analyses suggested that the purified substance is an amidated heptapeptide having the following primary structure: Ala-Met-Gly-Met-Leu-Arg- Met-NH>. The yield of this peptide from 500 g of the worms was roughly estimated to be 3 nmoles using the data on amino acid sequence analysis. The value was comparable to the yield of an S-Iamide peptide, AKSGFVRIamide, and larger than the yield of another S-Iamide peptide, VSSFVRIamide [16]. A mixture of the synthetic peptide with the suggested structure and the native peptide (purified substance) showed a single absorbance peak when applied to a C-18 reversed- phase column and a cation-exchange column (Fig. 2). We were not able to compare the bioactivity between the synthe- tic and native peptides, because the quantity of the native peptide obtained was not enough for the comparison. However, we confirmed contraction-eliciting activity of the synthetic peptide at concentrations of 10~® M or higher (Fig. Wl tL SS eS 0 (0) 0 10 20 30 (oe) (oy) Absorbance at 220 nm (o) wo Time (min) Fic. 2. HPLC profiles of mixtures of the native and synthetic peptides on a C-18 ODS-80TM column with an isocratic elution of 21% ACN (0.1% TFA) at a flow rate of 0.3 ml/min (left) and on the cation-exchange SP-SPW column with an isocratic elution of 0.2 M NaCl (10 mM phosphate buffer, pH 7.1) at a flow rate of 0.5 ml/min (right). 10°®M 4—*>—~— | et 5 2 min Fic. 3. Actions of the synthetic peptide on the isolated esophagus at different concentrations. The peptide was applied at the times indicated by arrow heads. The sequence of the annelid peptide is highly homolo- gous to those of myomodulin-CARP-family peptides found in several molluscs (Table 1). We examined the effects of these molluscan peptides and several synthetic analogues on the esophagus of P. vancaurica, and compared the effects with those of the annelid peptide which we designated Pev-myomodulin (Table 2). Pev-myomodulin showed the most potent contractile effect on the esophagus, and was approximately 10 times more potent than myomodulin and CARP. In Figure 4, the effects of 10-° M Pev-myomodulin and 10-°M CARP ona preparation of the esophagus are shown. Met’-CARP and Met’-myomodulin were also more potent than the respective authentic molluscan peptides, though they were slightly less potent than Pev-myomodulin. Phe’-CARP and Phe’- myomodulin were far less potent than Pev-myomodulin. Leu’-Pev-myomodulin was also far less potent than the authentic annelid peptide. C-terminal-free CARP did not show any effect even at 10-° M. Met-Leu-Arg-Leu-NHp, a common C-terminal fragment of the most molluscan myomo- dulin-CARP-family peptides, showed a considerable effect, TABLE 1. Myomodulin-CARP-family peptides Phyla Animals Structures References Annelida Perinereis AMGMLRMamide this study Mollusca Mytilus AMPMLRLamide Hirata et al. (1987) Aplysia PMSMLRLamide Cropper ef al. (1987) GSYRMMRLamide Cropper et al. (1991) Fusinus PMSMLRLamide Kanda et al. (1990) PMNMLRLamide Kanda et al. (1990) Helix PMSMLRLamide Ikeda et al. (1993) SLGMLRLamide Ikeda et al. (1993) GLNMLRLamide Ikeda et al. (1993) pOLSMLRLamide Ikeda et al. (1993) pOQLPMLRLamide Ikeda et al. (1993) Achatina SLGMLRLamide Ikeda et al. (unpublished) GLHMLRLamide Ikeda et al. (unpublished) pQ, pyroglutamic acid. 36 T. TAKAHASHI, O. MATSUSHIMA et al. TaBLE2. Contractile effects of myomodulin-CARP-family peptides and synthetic analogues on the isolated esophagus of Perinereis vancaurica Concentrations (M) Peptides ? 10-8 10-7 10-° 10m Myomodulin-CARP-family peptides AMGMLRMamide + feat See NTT AMPMLRLamide NT — db ab 6 ae ie PMSMLRLamide — ab SL aL NPT GSYRMMRLamide _ + 4+ 4 seas PMNMLRLamide NT - ++ db abot SLGMLRLamide — + ab ft ae fe tL GLNMLRLamide — + ++ eee pQLSMLRLamide - + ++ fe hh pOLPMLRLamide — = = SL te aie GLHMLRLamide — + ab dk ae fee Analogue peptides AMPMLRMamide — +4 +++ NT PMSMLRMamide — ++ JLab sk NPT AMGMLRLamide NT = — ab ab AMPMLRFamide NT NT = ar PMSMLRFamide — — + ++ AMPMLRL NT NT NT — MLRLamide NT 4f ++ +++ WLRLamide NT — - +44 LRLamide NT — = 2 +++, strong effect. ++, moderate effect. +, weak effect. —, no effect. NT, not tested. pQ, pyroglutamic acid. but Leu-Arg-Leu-NH) did not show any effect at 10-° M or lower. Replacement of the Met residue of Met-Leu-Arg- Leu-NH) with a more hydrophobic residue Trp decreased the contractile potency. The effects of Pev-myomodulin on catch tension and Ai A2 feds 10°°M AMPMLRLa B Bo 10°*M Ach 10°°M AMPMLRLa 4 10°°M AMGMLRMa B N 10°°M AMPMLRLa 2 min Fic. 4. Actions of 10-° M AMGMLRMamide (Pev-myomodulin) and 10-°M AMPMLRLamide (CARP) on the isolated esopha- gus of P. vancaurica. A, Pev-myomodulin. B, CARP. phasic contraction of the ABRM were also examined, and compared with those of CARP. The effects of 10-°M of these peptides on a preparation of the ABRM are shown in Figure 5. The actions of CARP and Pev-myomodulin (107° M) on the phasic contractions were apparently opposite. However, those actions were found to be dose-dependent (Fig. 6), indicating that actions of these peptides on the phasic contractions were not qualitatively different. Dose- response relationships of these peptides showed that CARP was much more potent than Pev-myomodulin in exerting both modulatory effects on the phasic contraction and relaxing effects on the catch tension (Fig. 6). DISCUSSION The novel heptapeptide Pev-myomodulin isolated from the annelid, P. vancaurica, in the current study is apparently a member of myomodulin-CARP family. The members of the myomodulin-CARP family have so far been identified only in molluscs (Table 1). Thus, this is the first report on me 10°°M AMGMLRMa 2 min 25g Ach 10°°M AMGMLRMa Fic. 5. Effects of AMPMLRLamide (CARP) and AMGMLRMamide (Pev-myomodulin) on phasic contractions (A;, A;) in response to repetitive electrical stimulation (15 V, 3 msec, 10 Hz, 50 pulses) and on catch contractions (B,;, B>) induced by 10°>*M ACh. Myomodulin-CARP-related Annelid Peptide 37 2007 se —S 150+ S — = Cc cA 6 ) ae = igen 100 =e te 2 x S ie) ® Oo. 50F 0 aes are as a CE a or 10° 10°” 10°° Relaxation (°/) 1 sates sc ee % oO 607 J = ae sip fe Sa Caen Once On enue On Og Concentration (M) Fic. 6. Dose-response relationships of AMPMLRLamide (CARP) and AMGMLRMamide (Pev-myomodulin) for modulat- ing actions on the phasic contractions (left) and for relaxing actions on the catch contractions (right) of the Mytilus ABRM. the occurrence of a member of the family in the Annelida. Since Pev-myomodulin showed a potent contractile effect on the esophagus of P. vancaurica, this peptide may be involved in the regulation of the esophagus of the animal. We have isolated previously two S-Iamide peptides from P. vancauri- ca, both of which elicit spontaneous contraction of the esophagus of the same animal as did Pev-myomodulin [16]. It is probable that some other peptides, which affect gut motility, may be found in P. vancaurica, and in other annelids as well, in future. A conspicuous difference in structure between the anne- lid peptide and the molluscan peptides is at their C-termini. The annelid peptide has a Met-NH) at its C-terminus, while all the molluscan peptides have a Leu-NH>. Among the myomodulin-CARP-family peptides including Pev- myomodulin, the annelid peptide shows the most potent contractile effect on the esophagus of P. vancaurica. The C-terminal Met-NH), of the peptide seems to be important in eliciting the potent effect. This notion is supported by the facts that Leu’-Pev-myomodulin is far less potent than Pev- myomodulin and that Met’-CARP and Met’-myomodulin are more potent than CARP and myomodiulin, respectively. The synthetic analogues Phe’-CARP and Phe’- myomodulin show less potent effects than CARP and myomodulin, respectively. This may be partly due to FMRFamide-like actions of the analogue peptides, because it has been known that FMRFamide has an inhibitory action on the esophagus of a polychaete annelid, N. virens [14]. FMRFamide has been shown to be present in some annelids [1, 4, 13]. C-terminus-free CARP, Ala-Met-Pro-Met-Leu-Arg- Leu-OH, does not show any contractile effect on the esopha- gus, suggesting that the C-terminal amide is essential for exertion of the contractile effects of the myomodulin-CARP- family peptides. It has been reported that the C-terminal amide is essential for catch-relaxing action of CARP on the ABRM of M. edulis [6]. Another example for essentiality of C-terminal amide for bioactivity is the RPCH-related pep- tide, APGWamide, which has been isolated from ganglia of the gastropod mollusc Fusinus ferrugineus [15]. On electri- cally-elicited phasic contractions of the Mytilus ABRM, the C-terminal dipeptide fragment of APGWamide, GWamide, shows a comparable activity to the native tetrapeptide, while C-terminal-free dipeptide, GW, does not show any activity [17]. Therefore, the C-terminal amide of small bioactive peptides having amidated C-termini may be generally essen- tial for exertion of their bioactivity. The tetrapeptide fragment, Met-Leu-Arg-Leu-NH), shows a considerable effect on the esophagus, but the tripeptide fragment, Leu-Arg-Leu-NH), does not show any effect. These facts suggest that the C-terminal tetrapeptide sequence is the minimum structure required for expression of the contractile effect of the myomodulin-CARP-family pep- tides on the esophagus, though the Met residue can be substituted at least with Trp. It has been shown that CARP potentiates phasic contrac- tion of the ABRM of M. edulis in response to repetitive electrical stimulation at lower doses and inhibits at higher doses [7, 18]. The annelid peptide Pev-myomodulin also shows a potentiating effect at lower doses and an inhibitory effect at higher doses on phasic contraction of the ABRM. However, CARP exerts these effects at about 100 times lower 38 T. TAKAHASHI, O. MATSUSHIMA et al. concentrations than Pev-myomodulin. As shown in Figure 5A, therefore, 10-° M CARP inhibits the phasic contraction while 10-°M Pev-myomodulin potentiates it. The reason why the myomodulin-CARP-family peptides show such dual actions is obscure at present. The annelid peptide, as well as CARP, relaxes catch tension of the ABRM. However, the annelid peptide is approximately 10 times less potent than CARP. In the ABRM, therefore, the authentic peptide CARP is more potent than the foreigh peptide Pev- myomodulin. In conclusion, we isolated a novel member of myomodu- lin-CARP family from the marine polychaete, P. vancaurica. It has been shown that S-Iamide-family peptides [16] and FMRFamide-family peptides [1, 4, 13] are present not only in molluscs but also in annelids. Annelids seem to have many neuropeptides closely related to molluscan neuropeptides. ACKNOWLEDGMENTS The authors wish to express their sincere thanks to Professor Y. Muneoka, Hiroshima University, for his valuable comments through- out the present study and critical review of the manuscript. REFERENCES 1 Baratt B, Gras-Masse H, Ricart G, Bulet P, Dhainaut-Courtois N (1991) Isolation and characterization of authentic Phe-Met- Arg-Phe-NH, and the novel Phe-Thr-Arg-Phe-NH> peptide from Nereis diversicolor. Eur J Biochem 198: 627-633 2 Cropper EC, Tenenbaum R, Kolks MAG, Kupfermann I, Weiss KR (1987) Myomodulin: a bioactive neuropeptide present in an identified cholinergic buccal motor neuron of Aplysia. Proc Natl Acad Sci USA 84: 5483-5486 3 Cropper EC, Vilim FS, Alevizos A, Tenenbaum R, Kolks MAG, Rosen S, Kupfermann I, Weiss KR (1991) Structure, bioactivity, and cellular localization of myomodulin B: a novel Aplysia peptide. Peptides 12: 683-690 4 Evans BD, Pohl J, Kartsonis NA, Calabrese RL (1991) Iden- tification of RFamide neuropeptides in the medicinal leech. Peptides 12: 897-908 5 Greenberg MJ, Price DA (1992) Relationships among the FMRFamide-like peptides. In “Progress in Brain Research Vol 92” Ed by J Joose, RM Buijis, FJH Tilders, Elsevier Science Publishers BV, pp 25-37 6 Hirata T, Kubota I, Imada M, Muneoka Y (1989) Pharmacolo- gy of relaxing response of Mytilus smooth muscle to the catch- relaxing peptide. Comp Biochem Physiol 92C: 289-295 7 Hirata T, Kubota I, Imada M, Muneoka Y, Kobayashi M (1989) 10 11 12 13 14 15 16 17 18 20 21 Effects of the catch-relaxing peptide on molluscan muscles. Comp Biochem Physiol 92C: 283-288 Hirata T, Kubota I, Takabatake I, Kawahara A, Shimamoto N, Muneoka Y (1987) Catch-relaxing peptide isolated from Myti- lus pedal ganglia. Brain Res 422: 374-376 Holman GR, Nachman RJ, Wright MS, Schoofs L, Hayes TK, DeLoof A (1991) Insect myotropic peptides. Isolation, structural characterization, and biological activities. In “Insect Neuropeptides. Chemistry, Biology and Action” Ed by JJ Menn, TJ Kelly, EP Masler, American Chemical Society, Washington, DC. pp 40-50 Ikeda T, Minakata H, Fujita T, Muneoka Y, Kiss T, Hiripi L, Nomoto K (1993) Neuropeptides isolated from Helix pomatia. I. Peptides related to MIP, buccalin, myomodulin-CARP and SCP. In “Peptide Chemistry 1992” Ed by N Yanaihara, ESCOM Science Publishers BV, pp 576-578 Kanda T, Kuroki Y, Kubota I, Muneoka Y, Kobayashi M (1990) Neuropeptides isolated from the ganglia of a prosobranch mol- lusc, Fusinus ferrugineus. In “Peptide Chemistry 1989” Ed by N Yanaihara, Protein Research Foundation, Osaka, pp 39-44 Keller R (1992) Crustacean neuropeptides: structure, function and comparative aspects. Experientia 48: 439-448 Krajniak KG, Price DA (1990) Authentic FMRFamide is present in the polychaete Nereis virens. Peptides 11: 75-77 Krajniak KG, Greenberg MJ (1992) The localization of FMRFamide in the nervous and somatic tissues of Nereis virens and its effects upon the isolated esophagus. Comp Biochem Physiol 101C: 93-100 Kuroki Y, Kanda T, Kubota I, Fujisawa Y, Ikeda T, Miura A, Minamitake Y, Muneoka Y (1990) A molluscan neuropeptide related to the crustacean hormone, RPCH. Biochem Biophys Res Commun 167: 273-279 Matsushima O, Takahashi T, Morishita F, Fujimoto M, Ikeda T, Kubota I, Nose T, Miki W (1993) Two S-Iamide peptides, AKSGFVRIamide and VSSFVRIamide, isolated from an anne- lid, Perinereis vancaurica. Biol Bull 184: 216-222 Minakata H, Kuroki Y, Ikeda T, Fujisawa Y, Nomoto K, Kubota I, Muneoka Y (1991) Effects of the neuropeptide APGWamide and related compounds on molluscan muscles —GWamide shows potent modulatory effects. Comp Biochem Physiol 100C: 565-571 Muneoka Y, Kobayashi M (1992) Comparative aspects of structure and action of molluscan neuropeptides. Experientia 48: 448-456 Muneoka Y, Twarog BM (1977) Lanthanum block of contrac- tion and of relaxation in response to serotonin and dopamine in molluscan catch muscle. J Pharmac Exp Ther 202: 601-609 Penzlin H (1989) Neuropeptides—occurrence and function in insects. Naturwissenschaften 76: 243-252 Walker RJ (1992) Neuroactive peptides with an RFamide or Famide carboxyl terminal. Comp Biochem Physiol 102C: 213- 222 ZOOLOGICAL SCIENCE 11: 39-44 (1994) Changes in the Responsiveness of Melanophores to Electrical Nervous Stimulation after Prolonged Background Adaptation in the Medaka, Oryzias latipes MASAZUMI SUGIMOTO, TAKAYUKI, KAWAMURA, RYOZO FUJII and Noriko OSHIMA Department of Biomolecular Science, Faculty of Science, Toho University, Miyama, Funabashi, Chiba 274, Japan ABSTRACT—The effects of prolonged background adaptation on the responses of melanophores were studied using electrical nervous stimulation. Electrical stimulation at various intensities and frequencies and for various periods of time was applied to scales isolated from B and W fish and the responses of melanophores in the scales were recorded photoelectrically. When electrical stimulation at enough intensity to induce maximal melanosome-agpregation response (“maximal stimulus”) was applied, there was no significant difference, in terms of the relationship between the magnitude of the aggregation response and the frequency or period of stimulation, between melanophores of B and W fish. However, the minimum effective voltage necessary to provoke the discernible aggregation of melanosomes in B fish was lower than that in W fish. With application of stimulation at intensities less than intensity of maximal stimulus, the response of melanophores in B fish was greater than that in W fish. These results suggest that prolonged background adaptation may induce changes in the excitability or in the density of distribution of chromatic nerve fibers, with a resultant © 1994 Zoological Society of Japan change in the concentration of released norepinephrine, upon electrical stimulation. INTRODUCTION Color changes in teleost fish are under both neural and hormonal control, and melanophores in the skin play an important role in such changes. Bidirectional movements of melanosomes within melanophores, which are responsible for the physiological color change, are known to be regulated mainly by neural control. Norepinephrine, the neurotrans- mitter liberated from sympathetic postganglionic fibers, causes the aggregation of melanosomes via stimulation of a-adrenoceptors on the melanophore membrane [7, 10, 11]. In addition, various hormonal principles, such as melatonin, alpha melanophore-stimulating hormone (a-MSH) and mela- nin-concentrating hormone (MCH), have the ability to cause translocation of melanosomes [4, 5, 9, 15, 16]. Changes in the number and in the size of melanophores, which give rise to the morphological color change, are considered to be controlled by hormones. Several investigators have re- ported that the pituitary hormones, namely, a-MSH and MCH, are responsible for the morphological color change [2, 3, 22]. The two types of color change may be interrelated, but the interrelationship is not well understood. Background adaptation is considered to be composed of two phases; an initial physiological color change and a subsequent morphological change [1, 24]. Therefore, an analysis of background adaptation may contribute to clarifica- tion of the relationship between the two types of color change. Recently, using chemically denervated medaka, Sugimoto [19] found that innervation also influenced the Accepted October 29, 1993 Received September 2, 1993 density of melanophores during background adaptation. Moreover, a change in the responsiveness of melanophores to exogenous norepinephrine occurred after prolonged back- ground adaptation [19]. In the present study, therefore, we used electrical stimulation of chromatic nerves to examine, at the tissue level, whether a change in the responsiveness of melanophore to endogenous neurotransmitter liberated from sympathetic nerve fibers could be recognized after prolonged background adaptation. On the basis of our results, we discuss the effects of prolonged background adaptation on the nerves that control the aggregation of pigment in melano- phores. MATERIALS AND METHODS Wild-type medaka, Oryzias latipes, of both sexes, 25-35 mm in total length, were purchased from local commercial sources. For background adaptation, fish were maintained for 10 days in a black- or a white-background aquarium under continuous illumina- tion at 2000 Ix at the surface of the water. In the present paper, these fish are referred to as B fish and W fish, respectively. Scales isolated from the dorsal trunk of B or W fish were immersed for 10 min at 4°C in Ca**- and Mg’*-free physiological saline solution (CME-PSS) of the following composition: NaCl, 129.8 mM; KCl, 2.7 mM; D-glucose, 5.6 mM; EDTA, 1 mM; Tris-HCl buffer (pH 7.4), 5.0mM. The epidermal layer of the scales was then ripped off with forceps, and the epidermis-free scales were kept in physiological saline solution (PSS) of the following composition: NaCl, 125.3 mM; KCl, 2.7 mM; CaCh, 1.8 mM; MgCh, 1.8 mM; D-glucose, 5.6 mM; Tris-HCl buffer (pH7.4), 5.0mM. These scales were used for electrophysiological studies within 3 hr of their isolation from fish. The system employed for the perfusion of experimental solutions and the electrical stimulation of chromatic nerves was similar to that 40 M. Sucimoto, T. KAwAMUuRA et al. described by Fujii & Novales [6] and Fujii & Miyashita [8]. A platinum wire of 300 ~m in diameter was fully insulated, except at its tips, and used as the stimulating electrode. The electrode was placed vertically on the central part of a scale in order to stimulate melanosome-aggregating nerves. As an indifferent electrode, another platinum wire was dipped in the solution in the perfusion chamber at a distance of 5mm from the stimulating electrode. Electrical stimulation at various intensities and frequencies and for various periods of time was delivered as a volley of negative pulses of 1-msec duration by means of an electric stimulator, SEN-2101 (Nihon Kohden, Tokyo). A synchroscope, SS-5116 (Iwatsu, Tokyo) was used to monitor the stimulating pulses. Response of each single melanophore, located at a distance of 100-200 4m from the stimulating electrode, was recorded photoelec- trically. The method used for recording the motile response of a melanophore was identical to that described by Oshima and Fujii [17]. Since there are many xanthophores in the scales of the wild-type medaka, a red filter (R-60; Toshiba, Tokyo), which block- ed light with wavelengths below 600 nm, was placed across the light path in the optical part of the recording system to eliminate any influence from the motile activity of these cells. In each series of measurements, the extent of the response of a melanophore was expressed as a percentage of the complete aggregation of melano- somes produced by norepinephrine hydrochloride at 10 ~M (NE; Sankyo, Tokyo). All measurements were performed at a room temperature, which fluctuated between 18° and 26°C. RESULTS After background adaptation for 10 days, which was taken as “prolonged background adaptation”, remarkable differences in terms of the number and the size of melano- phores were apparent between the scales of B and W fish. In both fish, electrical nervous stimulation at 5 V to central portions of isolated scales brought about the rapid aggrega- tion of melanosomes within the melanophores (Fig. 1). For as much as 3 hr after the isolation of scales, melanophores responded repeatedly to electrical nervous stimulation with similar aggregation of pigment. Therefore, one and the same melanophore was observed in a series of experiments in which electrical stimulation with various parameters was applied to the scale. The melanosomes redispersed within 4 min after cessation of electrical stimulation, and the next stimulus was applied after complete redispersion. Effects of changes in stimulus intensity on the responses of melanophores Figure2 is a typical recording of melanosome- aggregating responses to electrical nervous stimulation at various intensities in a melanophore from a B fish and that from a W fish. In these experiments, the frequency and duration were kept constant at 1 Hz and 1 msec, respectively. Stimulation was applied for 30 sec since the medaka really adapted to a white background within 30 sec. In both B and W fish, melanin-aggregating responses started 5-10 sec after the initiation of the stimulus, and redispersion began im- mediately after cessation of the stimulation. The rela- tionship between the intensity of the stimulus and the magni- tude of the response of melanophores is summarized in Fic. 1. Photomicrographs showing the melanosome-aggregating responses of melanophores in scales to electrical nervous stimulation. The upper series (A-D) shows the responses in a scale from a B fish and the lower series (E-H) shows those in a scale from a W fish. In each series, photographs were taken with transmission optics of the same part of a given scale from the dorsal trunk. A and E: equilibrated in PSS. Melanosomes in melanophores are fully dispersed. Note that larger melanophores are present at higher density in the B fish. B and F: 15 sec after the application of stimulating pulses at an intensity of 5 V, at a frequency of 1 Hz and with a duration of 1 msec. Melanosomes are in the process of aggregating. C and G: 1 min after the start of electrical stimulation. The melanosome-aggregating responses have reached the maximal level. D and H: 4 min after treatment with 10 «NE, which was applied at the end of the electrical stimulation. Melanosomes are completely aggregated. Note that the extent of melanosome aggregation caused by NE is greater than that induced by the electrical stimulation (x 160). Melanophore Responsiveness in Fish 41 100/- 5 oO =y Aggregation (%) oad ( m oO Ss 4 [o) So Aggregation (%) ES (V) 0.5 Fic. 2. Typical recordings of the responses of melanophores to electrical stimulation at various intensities. ttlut ae cual = 10 uM NE 10 uM NE The upper and lower recordings show the responses of a melanophore from a B fish and those of a melanophore from a W fish, respectively. The frequency and period of stimulation were 1 Hz and 30sec, respectively. After the cessation of electrical stimulation, 10 ~M NE was applied for 4 min to bring about complete (100%) aggregation of melanosomes. Although there was a difference in the threshold intensity between melanophores from B and W fish, the magnitude of the response increased depending on the intensity of the stimulus in both cases. Figure 3. In B fish, melanosomes in melanophores slightly aggregated in response to 1-V pulses and considerably in response to the intensities above 2 V. In melanophores of W fish, however, no distinct aggregation response was seen at intensities of 2 V and lower. Thus, the response of melano- phores to 2-V pulses in B fish was significantly greater than that in W fish (P<0.001). In melanophores of both fish, stimulation at 5 V induced the maximal response, which was about 70% of the full response caused by 10 ~M NE, and a further increase in the intensity of the stimulus caused a gradual decrease in the magnitude of the response. Effects of changes in the period of stimulation on responses of melanophores Electrical stimulation was applied for various periods of time at an intensity of 5 V, a frequency of 1 Hz and a duration of 1msec. The relationship between the period of stimula- tion and the response of melanophores is shown in Figure 4. Five-sec stimulation, namely, five pulses of 1-msec duration over the course of 5 sec, induced discernible aggregation of melanosomes in both B and W fish. Until the responses of melanophores reached a plateau at about 30-sec stimulation, increases in the period of stimulation were accompanied by gradual increases in the magnitude of the response. As indicated in Figure 4, the curves for B and W fish were similar, and the response in both cases did not exceed 80% of the full response caused by NE at 10 uM. ES, electrical stimulation. 80 Ff T 4 & 60 + 1 c 4 2 [ > 40 + | 2 D r 4 < 20 7 | 0 | 1 0.5 1 2 5 10 20 Intensity of stimulation (v) Fic. 3. Relationship between the intensity of the stimulus and the magnitude of the melanosome-aggregating response in a single melanophore. The curves drawn through solid and open circles indicate the relationships in B and W fish, respectively. Sti- mulation was applied at a frequency of 1 Hz (1-msec duration) for 30sec. Complete (100%) aggregation of pigment was caused by treatment with 10 ~M NE for 4 min. Each point represents the mean of measurements on scales from five different fish. Vertical lines indicate standard errors. Effects of changes in the frequency of the stimulus on re- sponses of melanophores The frequency of pulses also affected the responses of 42 M. Sucimoro, T. Kawamura et al. Aggregation (%) 80 | 4 60 + 5| 40 | 20 + 0 4 1 2 5 10 20 50 100 Period of stimulation (sec) Fic. 4. Relationship between the period of stimulation and the magnitude of the melanosome-aggregating response in a single melanophore. The curves through solid and open circles indi- cate the relationships in B and W fish, respectively. Stimulation at 5V and 1Hz was applied for various periods of time. Complete (100%) aggregation of pigment was caused by the treatment with 10 ~M NE for 4 min. Each point represents the mean of measurements on scales from five different fish. Ver- tical lines indicate standard errors. Aggregation (%) 2 5 10 20 Frequency of stimulation (Hz) Relationship between the frequency of the stimulus and the magnitude of the melanosome-aggregating reaction in a single 0.1 0.2 0.5 1 Fic. 5. melanophore. The curves through solid and open circles indi- cate the relationships in B and W fish, respectively. Stimulation was applied at an intensity of 2 V (dashed line) or 5 V (uninter- rupted line) for 30sec. Complete (100%) aggregation of pig- ment was caused by treatment with 10 ~M NE for4 min. Each point represents the mean of measurements on scales from five different fish. Vertical lines indicate standard errors. melanophores. Figure5 shows the frequency-response The frequency was changed from 0.1 Hz to 20 Hz and other parameters were kept constant (5 V, 1-msec dura- tion, 30-sec stimulation). As the frequency of the stimula- tion was increased from 0.1 to 5 Hz, the magnitude of the response increased by degrees. The response reached a plateau at more than 5 Hz, and the maximal responses were curves. about 90% of the full response to 10 ~M NE in both B and W fish. There was no significant difference between the curves for B and W fish. However, frequencies below 1 Hz had a tendency to provoke a greater response in B fish than in W fish. Next, the frequency of stimulation was reduced succes- sively without a cessation of stimulus for 4min among measurements at each frequency, unlike the experiments mentioned above. The magnitude of the response also decreased, reaching a stable value at each frequency (Fig. 6). The effective frequencies of the stimulation for the induction of 50% aggregation of melanosomes [EFso (95% confidence limits)] in B and W fish, that were estimated by experimental results shown in Figure 5, were 0.66 Hz (0.22—2.00 Hz) and 1.29 Hz (0.56-2.97 Hz), respectively. 70 1 min Aggregation (%) 1min Aggregation (%) 0 ES (Hz) | 2 4 0.5 | 2028) Fic. 6. Typical recordings showing the changes in the magnitude of the melanophore response with changes in the frequency. The upper panel shows a recording from a B fish melanophore at frequencies of 1, 0.5, 0.2 and 0.1 Hz, inturn. The lower panel shows a recording from a W fish melanophore at frequencies of 2,1,0.5 and 0.2 Hz, inturn. The intensity of the stimulus was 5 V and the duration was 1 msec. Complete (100%) aggregation of pigment was caused by treatment with 10 «M NE for 4 min. ES, electrical stimulation. When a volley of pulses at 2 V was applied, which induced only a small response at 1 Hz in W fish (see Fig. 3), the threshold frequency for the induction of the response in B fish was lower than that in W fish (dashed line in Fig. 5). Within the range from 0.5 to 5 Hz, the magnitude of the response in B fish was significantly different from that in W fish (P<0.01) and EFs 9 (95% confidence limits) was 1.23 Hz (0.36-4.22 Hz) and 17.67 Hz (4.5-68.18 Hz), respectively. The response of melanophores in B fish reached a plateau value at 5 Hz, whereas that in W fish gradually increased with increases in the frequency. The maximal responses caused by 2-V pulses were 74% and 51% of the maximal aggregation Melanophore Responsiveness in Fish 43 at 5 V in B and W fish, respectively. DISCUSSION Since the melanophore itself is a nonexcitable type of cell [5], neural control of its motility has been effectively studied by electrical nervous stimulation [6, 7, 8, 11, 18, 21]. Fujii and Novales [6], using split tail-fin preparations of Fundulus, showed that even a single stimulating pulse to chromatic nerves causes moderate aggregation of melanosomes in a melanophore. They also suggested that multiple chromatic nerve fibers might control the motility of a single melano- phore since the melanosome-aggregation response in- creased with increases in the intensity of electrical stimula- tion. In the present study, we used melanophores in scales isolated from the medaka and obtained similar results for the relationship between the extent of the response and the intensity of the stimulus (see Fig. 3). More than two nerve fibers seem to be present in the scale, as well as in the tail fin. Our results are also consistent with morphological observa- tions on the innervation of teleost melanophores reported by Yamada et al. [25] and Miyata & Yamada [14], who found networks of adrenergic varicose fibers labeled with [>H]-NE in the scale and the tail fin, respectively. In general, an indication of the excitability of nerve fibers is proved by the extent of effector responses caused by stimulation at various intensities [23]. We compared re- sponses of melanophore from B fish to electrical stimulation at various intensities with those of W fish. At 5V, all chromatic fibers in the scales of W fish seemed to participate in the response because stimulation at this intensity elicited the maximal response. In B fish, by contrast, the maximal response was induced at 3 V, and 5-V stimulation did not further augment the magnitude of the response, which was not the full response, being only 70% of the complete aggregation of pigment caused by 10“M NE. These findings suggest a change in the excitability of chromatic nerve fibers in response to the electrical stimulation. These data may also suggest that the density of distribution of chromatic nerve fibers in the scales of B fish might be higher than that in the scales of W fish. It is unclear why the magnitude of the response was reduced at intensities above 8 V in both fish but polarization of stimulating electrode might be involved in this phenomenon, at least to some extent. The magnitude of the response also depended on the period of stimulation (see Fig. 4). In both B and W fish, however, the period of stimulation required for the maximal response was the same. Melanophores in scales from B fish possessed many thick dendritic processes, and dendrites of W fish melanophores were thin and small in number (see Fig. 1). Therefore, the area occupied by each melanophore of B fish was larger than that of W fish [19]. However, melanophores of W fish seemed to be roughly equal to these of B fish in the total length of a cell: the distance between the tips of two dendrites which project in an opposite direction each other with the cell body between. Thus, the distances that melanosomes would migrate within melanophores of both B and W fish might be almost the same. About 30sec were required to induce the maximal response by stimulation at 5 V and 1 Hz. In order to investigate the effect of prolonged back- ground adaptation on the mechanism of release of the neurotransmitter, melanophore responses of B and W fish to various frequencies of stimulation were examined. Stimula- tion at 5 V was applied for 30sec to activate all chromatic fibers and to achieve the maximal response at a given frequency. It is generally known that, when electrical sti- mulation at an intensity that activates all fibers is used to stimulate the peripheral neuroeffector system, the amount of NE released per stimulating pulse is constant. An increase in frequency increases the concentration of the released neurotransmitter that diffuses in the synaptic cleft [23]. Thus, the frequency-response curves in the present study indicate the relationships between the amount of NE released and the extent of the melanosome-aggregating response (see Fig. 5). Since there was no significant difference between the curves for B and W fish at an intensity of 5 V, no difference in the sensitivity of melanophores to the endog- enous neurotransmitter was demonstrated. However, Sugi- moto [20] reported that, as judged from the ECso, the responsiveness of melanophores to exogenous NE in B fish was about 16 times higher than that in W fish. Such dis- agreement with the present result implies that the action of neurotransmitter released in situ from chromatic nerves dif- fers from that of exogenous NE in vitro. In the former case, released NE is rapidly inactivated by neuronal re-uptake and metabolizing enzymes [13]. Moreover, Kumazawa and Fujii [12] demonstrated the concomitant release of ATP together with the true transmitter, NE, and they suggested that ATP is successively dephosphorylated to adenosine and functions as a cotransmitter to promote the effective redispersion of pigment. An increase in levels of released NE by high- frequency stimulation is also accompanied by increased re- lease of ATP, with resultant prompt recovery from the effects of NE. In fact, as shown in the Figure 6, the extent of melanosome aggregation decreased quickly upon reduction of the frequency of stimulation, whereas a decrease in the concentration of exogenous NE did not rapidly reduce the extent of the responses of melanophores in the reaction chamber (data not shown). Thus, the change in the sensitiv- ity of melanophores to NE, which was observed by Sugimoto [20] of the addition of exogenous NE, was not clear in the present experiments at the tissue level. When electrical stimulation at 2 V was applied, the EF so in B fish was about 14 times lower than that in W fish, suggesting the enhanced excitability of the chromatic fibers of B fish at various frequencies of electrical stimulation. The response in B fish reached a plateau at a frequency of 5 Hz, but the aggregation of pigment was far from complete. Probably, increases in frequency above 5 Hz failed to in- crease the concentration of liberated transmitter. From the present results, we conclude that the prolonged 44 M. Sucimoto, T. KAWAMURA et al. background adaptation affects the responses of melanophores to electrical nervous stimulation in situ, as a result of the changes in the excitability of nerves to stimulation and/or in the density of distribution of nerve fibers. In addition, a change in the sensitivity to exogenous NE of melanophores themselves has also been pointed out [20]. It seems likely that the change in the responsiveness of melanophores contri- butes to adaptation to a new background color after pro- longed background adaptation. As an illustration, when fish that have fully adapted to a dark background are transferred to a new background that is brighter than the previous one, the increased excitability of chromatic nerve fibers and the increased sensitivity of melanophores to melanosome- aggregating agents may be advantageous since the fish can adapt to the new background rapidly. By contrast, de- creased excitability and sensitivity may be of advantage to fish that are fully adapted to a bright background. In our next report, patterns of innervation of melanophores in B and W fish will be described. REFERENCES 1 Bagnara JT, Hadley ME (1973) Chromatophores and Color Change. Prentice-Hall, Englewood Cliffs, New Jersey 2 Baker BI, Wilson JF, Bowley TJ (1984) Changes in pituitary and plasma levels of a-MSH in teleost during physiological color change. Gen Comp Endocrinol 55: 142-149 3 Baker BI, Bird DJ (1992) The biosynthesis of melanin- concentrating hormone in trout kept under different conditions of background colour and stress, as determined by an in vitro method. J Neuroendocrinol 4: 673-679 4 Fain WB, Hadley ME (1966) In vitro response of mela- nophores of Fundulus heteroclitus to melatonin, adrenaline and noradrenaline. Am Zool 6: 596 5 Fujii R (1961) Demonstration of the adrenergic nature of transmission at the junction between melanophore- concentrating nerve and melanophore in bony fish. J Fac Sci Univ Tokyo, IV, 9: 171-196 6 Fujii R, Novales RR (1969) The nervous mechanism controll- ing pigment aggregation in Fundulus melanophores. Comp Biochem Physiol 29: 109-124 7 Fujii R, Miyashita Y (1975) Receptor mechanisms in fish chromatophores—I. Alpha nature of adrenoceptors mediating melanosome aggregation in guppy melanophores. Comp Biochem Physiol 51C: 171-178 8 Fujii R, Miyashita Y (1976) Receptor mechanisms in fish chromatophores—III. Neurally controlled melanosome aggregation in a siluroid (Parasilurus asotus) is strangely medi- 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 DS ated by cholinoceptors. Comp Biochem Physiol 55C: 43-49 Fujii R, Miyashita Y (1978) Receptor mechanisms in fish chromatophores—IV. Effects of melatonin and related subst- ances on dermal and epidermal melanophores of the siluroid, Parasilurus asotus. Comp Biochem Physiol 59C: 59-63 Fujii R, Oshima N (1986) Control of chromatophore move- ments in teleost fishes. Zool Sci 3: 13-47 Iwata KS, Fukuda H (1973) Central control of color changes in fish. In “Responses of Fish to Environmental Changes” Ed by W Chavin, Springfield, Illinois, pp 316-341 Kumazawa T, Fujii R (1984) Concurrent release of nore- pinephrine and purines by potassium from adrenergic melano- some-aggregating nerve in tilapia. Comp Biochem Physiol 78C: 263-266 Langer SZ (1981) Presynaptic regulation of the release of catecholamines. Pharmacol Rev 32: 337-362 Miyata S, Yamada K (1987) Innervation pattern and respon- siveness of melanophores in tail fins of teleosts. J Exp Zool 241: 31-39 Nagai M, Oshima N, Fujii R (1986) A comparative study of melanin-concentrating hormone (MCH) action on teleost mela- nophores. Biol Bull 171: 360-370 Negishi S, Obika M (1980) The effects of melanophore- stimulating hormone and cyclic nucleotides on teleost fish chro- matophores. Gen Comp Endocrinol 42: 471-476 Oshima N, Fujii R (1984) A precision photoelectric method for recording chromatophore responses in vitro. Zool Sci 1: 545— 552 Parker GH (1948) Animal color change and their neurohu- mours. Cambridge Univ Press, London Sugimoto M (1993) Morphological color changes in the meda- ka, Oryzias latipes, after prolonged background adaptation—I. Changes in the population and morphology of melanophores. Comp Biochem Physiol 104A: 513-518 Sugimoto M (1993) Morphological color changes in the meda- ka, Oryzias latipes, after prolonged background adaptation—II. Changes in the responsiveness of melanophores. Comp Biochem Physiol 104A: 519-523 Ueda K (1955) Stimulation experiment on fish melanophores. Annot Zool Jap 28: 194-205 van Eys GJJM, Peters PTW (1981) Evidence for a direct role on a-MSH in morphological background adaptation of the skin in Sarotherodon mossambicus. Cell Tissue Res 217: 361-372 von Euler US (1959) Autonomic neuroeffector transmission. In “Neurophysiology”. Ed by Am Physiol Soc, Washington, D.C., pp 215-237 Waring H (1963) Color Change Mechanisms of Cold-blooded Vertebrates. Academic Press, New York London Yamada K, Miyata S, Katayama H (1984) Autoradiographic demonstration of adrenergic innervation to scale melanophores of a teleost fish, Oryzias latipes. J Exp Zool 229: 73-80 ZOOLOGICAL SCIENCE 11: 45-53 (1994) Vitellogenin Production Induced by Eyestalk Ablation in Juvenile Giant Freshwater Prawn Macrobrachium rosenbergii and Trial Methyl Farnesoate Administration Marcy N. WiLper, TAkusI OkumurRA!, YuUzURU SUZUKI, NOoBUHIRO FUSETANI and Katsumi AIDA Department of Fisheries, Faculty of Agriculture, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113, Japan ABSTRACT—Juvenile Macrobrachium rosenbergii were bilaterally eyestalk-ablated and hemolymph-sampled at 3 to 6 day-intervals for three weeks. In males, vitellogenin appeared several days after ablation, increased slightly for 1-2 weeks, and then decreased. Quantification by enzyme immunoassay indicated peak vitellogenin levels ranging from 0.1 to 0.8mg/ml. Females showed a similar profile, but peak levels ranged from 0.5 to 3.0 mg/ml; one individual was however exceptional with titers reaching nearly 30 mg/ml. Vitellogenin was not detectable in non-ablated animals. Juvenile male and female vitellogenin was shown by SDS-PAGE/Western blotting to consist of a single polypeptide component of 199K; however, in the exceptional female only, vitellogenin was composed of three polypeptides of 199, 102 and 90K as in adult female vitellogenin. Nevertheless, all ablated juvenile females exhibited increased gonadosomatic index and vitellogenic oocytes, as demonstrated by immunocytochemical techniques. Subsequently, juvenile males were employed to examine the effects of methyl farnesoate (MF) administration on vitellogenin production. In ablated animals, no significant differences in vitellogenin production were observed between the MF-injected and saline-injected groups. MF administration could not induce vitellogenin production in non-ablated animals. The physiological significance of the appearance of vitellogenin in juveniles of both sexes in context of the above results is discussed. © 1994 Zoological Society of Japan INTRODUCTION In decapod crustaceans, the processes of molting and reproduction are inextricably linked; these are under the negative control of the peptide hormones, vitellogenesis- inhibiting hormone (VIH) and molt-inhibiting hormone (MIH). MIH and VIH originate in the X-organ/sinus gland complex of the eyestalk and have been extensively characte- rized in several species [5, 9, 29, 35]. Stimulatory factors also exist but their identities are not fully known; the thoracic ganglia have been demonstrated by several investigators [15, 33, 39] to be a source of a vitellogenesis-stimulating hormone (VSH) or gonad-stimulating hormone (GSH), and the ovary to be the putative site of vitellogenesis-stimulating ovarian hormone (VSOH) and factors considered to induce secon- dary sexual characteristics [20, 22, 32]. Also of much in- terest regarding crustacean reproduction, is the role of methyl farnesoate (MF), which in insects is the unepoxidated precur- sor of juvenile hormone (JH) III. MF, a product of the mandibular organs, has been detected in the hemolymph of adult Macrobrachium rosenbergii (27, 37] and has been demonstrated to be at high levels during active vitellogenesis in crabs [19]. It has been observed that eyestalk ablation results in the production of vitellogenin and the acceleration of ovarian Accepted December 27, 1993 Received November 5, 1993 " Present address: Japan Sea National Fisheries Research Institute 5939-22, Suido-cho 1-chome Niigata City, Niigata 951, Japan maturation and spawning in adult female Macrobrachium rosenbergii [23], and in mature female Crustacea in general. However, little information has been available concerning the likelihood that juvenile crustaceans can synthesize vitel- logenin and whether such factors as MIH and VIH are present at the juvenile stage. Therefore, the main aim of this investigation was to examine the effects of eyestalk ablation on vitellogenin production in juvenile M. rosenbergii of both sexes and to gain more understanding of what factors may be involved in the control of its production. The appearance of vitellogenin after eyestalk removal was fol- lowed, and the nature of juvenile vitellogenin was examined immunologically and electrophoretically. Given the current implications that MF may be involved in vitellogenin synthesis in Crustacea as JH is in insects [21, 38], it was considered here that juvenile males could be used as a system for testing the effects of methyl farnesoate administration in M. rosenbergii. For trial purposes, MF was prepared from farnesoic acid and injected intramuscu- larly. MATERIALS AND METHODS Sampling: Vitellogenin production in ablated and non-ablated juve- niles Thirty juveniles (body weight, BW=2-5 g) were divided into two groups to be hemolymph-sampled for three weeks at intervals of several days irrespective of sex. In the first group (eyestalkless), bilateral ablation was carried out by simply holding the animal half-submerged in water, and snipping both eyestalks using scissors rinsed in crustacean saline. After allowing hemolymph flow to stop, 46 M. N. WILDER AND T. Okumura et al. the animal was fully returned to its tank. In the second group (intact) no treatment was employed. Hemolymph samples (less than 20 jl for each occasion) were taken by 25G needle and syringe, quick frozen at —80°C and stored at —30°C until analysis by enzyme immunoassay (EIA) and sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). At the end of the sampling period, animals were dissected for sex determination. Enzyme immunoassay (EIA) Antiserum raised against purified vitellin from Macrobrachium nipponense [14, 24] shown to be specific for vitellin in M. rosenbergit [23] was employed in EIA in this investigation. The details of the development and validation of this EIA have been reported previous- ly [23]; this system utilizes M. rosenbergii vitellin for expression of the standard curve and M. rosenbergii hemolymph in the preparation of m-CB (0.01% male hemolymph-containing carbonate buffer used in the dilution of standards and samples). Otherwise, procedures are identical to those of the M. nipponense EIA [24] in which standards or samples are adsorbed onto wells, blocked, and subsequently incubated with vitellin antiserum followed by goat anti-rabbit IgG alkaline phosphatase conjugate. In this investigation, after adding substrate (nitrophenylphos- phate disodium salt) to wells, absorbance was measured at 405 nm; sample concentrations were calculated from the standard curve and expressed in terms of milligram equivalents of vitellin per milliliter hemolymph. The lower limit of detection was 0.03 mg/ml. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS- PAGE) and immunoblotting SDS-PAGE was carried out on 7.5% polyacrylamide separating gels with 3% polyacrylamide stacking gels. Sample preparation was done by mixing sample dilution buffer (20mM Tris-HCl pH6.8, containing 4% SDS, 40% glycerol, 2% mercaptoethanol, and 0.005% bromophenol blue) and hemolymph diluted 10-fold with distilled water ata1:1ratio. These were heat-treated at 100°C for 5 min. For purposes of comparison, vitellin (Vn) purified by gel filtration and ion exchange chromatography from M. rosenbergii ovary [14] and female hemloymph from adult female M. rosenbergii with mature ovaries (reproductive molt) were also run concurrently. For molecular weight determination, high molecular weight (200, 116, 97, 66, 45K) markers (Bio-Rad molecular weight marker kit) were employed. For subsequent immunoblotting procedures, proteins were firstly transferred electrophoretically from gels to Immunobilon PVDF Transfer Membranes (Millipore). Membranes were soaked in 10% Block Ace (Yukijirushi Nyugyou K.K.)/TBS (Tris-HCl pH 8.0 containing 0.9M NaCl), washed with TBS-Tween (TBS contain- ing Tween 20) and transferred to a dilution of protein A purified and biotinylated anti-Vn IgG (approx. 3 ug/ml) in TBS-Tween. Mem- branes were stained using the Vectastain Avidin-Biotin Complex kit (Vector Laboratories) and 3,3’-diaminobenzidine (25 mg DAB, 4 ul 31% H202, 20 ml TBS-Tween). After saturating the membranes in DAB solution, staining was enhanced with the addition of 40 «l each 200mM CoCl, and NiCl,. Reference gels were stained with Coomassie brilliant blue. Histology Histological studies were employed to confirm the extent of vitellogenesis in ovaries of eyestalk-ablated juvenile females. Ovar- ian tissue was fixed for 24—48 hrs in Bouin’s solution and dehydrated through an alcohol gradient as described previously [36]. Tissues embedded in paraffin were sectioned to 5-7 um and stained with hematoxylin-eosin to examine ovarian stages. Additionally, immunocytochemical techniques were undertaken to confirm the presence in oocytes of vitellin-immunoreactive mate- rial. Sections were treated with 3% hydrogen peroxide to block endogenous peroxidase activity, and incubated in anti-Vn IgG (Pro- tein A purified) diluted to 1.3 ug/ml in 0.1 M phosphate-buffered saline (0.9% NaCl, pH 7.5) overnight at 4°C. Immunocytochemical procedures were done using the Histofine immunostaining kit (Nichirei): biotinylated goat anti-rabbit IgG was firstly applied to sections, followed by the application of peroxidase-conjugated strept- avidin. Final staining was done using 3,3’-diaminobenzidine (12.5 mg DAB in 50 ml 0.1 M phosphate buffer, pH 7.4, containing 750 yl 0.3% H2O2). As a control, the same anti-Vn IgG dilution (1 ml) was preincubated overnight with 2 ug purified vitellin. Methyl farnesoate (MF) injection experimental protocol Forty juvenile males already showing development of male gonopores and petasma were chosen. These were divided into four groups to be subjected to the following protocol: 1) eyestalk ablation, MEF injection; 2) eyestalk ablation, saline injection; 3) no ablation, MF injection; 4) no ablation, saline injection. MF was prepared from farnesoic acid (gift of Kuraray Co., Ltd.) by methylation with diazomethane; confirmation of identity was done via gas chroma- tography-mass spectrometry (GC-MS) as in Laufer eral. [19]. Stock solutions in ethanol at 4 mg/ml were made up and MF for injection was prepared as a suspension of 5% ethanol in crustacean saline. This gave 200 ~g MF per ml and injections were carried out intramus- cularly (at the base of the fifth pleiopod) using 25 ul, or5 ~g MF. In MF-injected animals, injections were carried out every day after ablation for 5 days, and then blood-sampled. In control animals, injection was done with 5% ethanol in saline. At the end of the experiment, experimental animals were dissected to confirm the presence of testes; hemolymph samples were taken and quick-frozen at —80°C and stored at —30°C until use in EIA and SDS-PAGE. Initial samples were taken in some cases to confirm non-detectable (ND) values at the outset. The Student’s t-test was employed to examine final vitellogenin levels. RESULTS 1. Vitellogenin production and molt frequency in juveniles A. Enzyme immunoassay (EIA) and molting In the first phase of this investigation, 20 individuals were ablated, and 10 were left intact and followed for a three-week period. Eyestalk ablation led to increased molting and rapid development of the gonads. Of 20 eyestalk-ablated indi- viduals, 5 females and 6 males survived for the duration of the experiment. In addition, 2 females showing developed ovar- ies apparent through the carapace and 1 male showing gonopores survived into the second week. Other individuals did not survive after ablation. Of 10 intact individuals, 8 survived the three-week duration; animals were dissected, but sex could not be determined. Ovaries of eyestalk- ablated females were additionally examined histologically (see below). Vitellogenin titers of individuals followed for the three week duration are shown for males M1-6 (Fig. 1), females F1-5 (Fig. 2a, b) and intact animals, L1—8 (Fig. 3). In males Vitellogenin in Juvenile Prawns 47 Vitellin Equivalents (mg/ml) 9-8] M4 y M 0.0 CI T= Salen a M5 1.0 M 0.5 0.0 0 2 4 6 8 10 12 I4 JO 18 2O Days Fic. 1. Vitellogenin titers in ablated juvenile males M1—6 as quan- tified by enzyme immunoassay (expressed in vitellin equiva- lents). Peak levels ranged from 0.1 to 0.8mg/ml. Vitel- logenin appeared several days after ablation, increased slightly for 1-2 weeks, and then decreased. Eyestalk removal was performed on Day 0, and the first hemolymph sample was taken several hours later or within 24hr of this. Individuals were then followed for the days indicated. Arrows marked by “M” indicate occurrence of molting. (body weight, BW=4.02 £0.45 g), no immunoreactive mate- rial was detectable at initial sampling occasions, performed within 24hr of ablation. Vitellogenin generally appeared several days after ablation, increased slightly for 1-2 weeks, and then decreased. Molts are indicated by arrows; all but one male individual molted twice during the twenty-day period. Maximum titers were between 0.1 and 0.8 mg/ml. Females (BW=3.80+0.29 g) also molted twice during the 3-week duration, but peak levels were greater than those of males, ranging between 0.5 to 3.0mg/ml. F5 was however exceptional, with levels reaching almost 30 mg/ml. It should be noted that in this species, normally maturing adult females exhibit peak Vg levels around 10 mg/ml [23]. Intact animals (BW=3.10+0.18g) were non-detectable (ND) throughout the experiment and molted 0-1 times. 3 Fi Vitellin Equivalents (mg/ml) pea T 0 2 4 6 3s I@ 12 14 16 18 2O Days Fic. 2a. Vitellogenin titers in ablated juvenile females F1—F4 as quantified by enzyme immunoassay (expressed in vitellin equiva- lents). Peak levels ranged from 0.5 to3 mg/ml. Vitellogenin appeared several days after ablation, increased slightly for 1-2 weeks, and then decreased. Eyestalk removal was performed on Day 0, and the first hemolymph sample was taken several hours later. Individuals were then followed for the days indi- cated. Gonadosomatic index ranged from 0.54 to 1.02%. Arrows marked by “M7” indicate occurrence of molting. 48 M. N. WILDER AND T. Okumura et al. 3075 F5 20 Vitellin Equivalents (mg/ml) 0 2 4 6 3 IO i 4 16 Ws 2o Days Fic. 2b. Vitellogenin titers in ablated female F5 as quantified by enzyme immunoassay (expressed in vitellin equivalents). Vitel- logenin concentrations initially reached levels of about 3 mg/ml; thereafter, titers increased exceptionally to nearly 30 mg/ml. Eyestalk removal was performed on Day 0. Final gonadosoma- tic index was 4.10%. Arrows marked by “M” indicate occur- rence of molting. B. SDS-PAGE and Western blotting Representative Western blots are shown for ablated males and females using M3 (Fig. 4a) and F1 (Fig. 4b) representatively. Exception F5 (Fig. 4c) is also shown. Numbers 1-5 or 1-6 correspond to sampling occasions done at 3-6 day intervals (Days 0, 3, 8, 13, 19 for M3 and Days 0, 4,8, 12,16, 20 for Fl and F5). Vitellin purified from eggs of M. rosenbergii eggs (102, 90K) and adult female vitellogenin (199, 102, 90K) are shown for reference in lanes A and B, respectively. Individuals M3 and F1 which are representa- tive of all males and females except F5, show only the 199K band. Its appearance and apparent quantity at each sam- pling occasion reflect vitellogenin titers as determined by EIA. In F5, after only the 199K band appeared on Day 12, the full three-band pattern was observed on Days 16 and 20. C. Ovarian development in females Females Fl, F2, F3, and F4 exhibited gonadosomatic indices (GSI) of 1.02, 0.54, 0.60, and 0.65%. These values are higher than those corresponding to immature ovaries which average about 0.4% during the common molt cycle of the adult female, but are low compared to GSI values that Vitellin Equivalents (mg/ml) L8 = 0) SS no Se ae a eee coe ee ee 0 2 4 6 8° 10 124 a Cm ee Days Fic. 3. Vitellogenin titers in intact individuals L1-8 as determined by enzyme immunoassay (expressed in vitellin equivalents). Non-ablated individuals were followed as in Figs. 1 and 2a-b for a three-week duration. No immunoreactive material was de- tected throughout the observation period. Occurrence of molt- ing is indicated by arrows marked by “M”. reach above 9 percent in fully mature ovaries [36]. FS with exceptionally high vitellogenin levels also had a much greater GSI, of 4.10%. In histological examination using hematoxy- lin-eosin, it was observed that ovaries in females Fl—F4 were in secondary vitellogenesis, exhibiting oocytes partially filled with yolk globules. In these oocytes, the nucleoli were clearly visible within the nucleus, which was not condensed in appearance (Fig. 5a). Ovaries in female F5 were quite Fic. 5. Extent of oocyte development in ablated juvenile females. (a) Fl was typical of the degree of ovarian maturity in females Fl-F4. Oocytes are enlarged, contain yolk globules, and can be considered to be in secondary vitellogenesis. The nucleoli are clearly visible within the nucleus. nucleus is not condensed in appearance. Gonadosomatic index (GSI) ranged from 0.54 to 1.02%. Germinal vesicle breakdown (GVBD) has not yet begun, as the (b) FS, exhibiting extremely high titers of full vitellogenin, possessed oocytes which appeared to have undergone GVBD as the nucleus is no longer observable. However, F5 did not undergo spawning. GSI was above 4%. Scale bar: 50 ~m. 200K 116K 97K 66K 45K 200K 116K - 97K 66K 45K 200K 116K 97K 66K 45K Vitellogenin in Juvenile Prawns 49 a AB 123456 — oy ee ee ee —< AB 123456 : — wy ee ge = > > different in appearance (Fig. 5b). Oocytes were extremely enlarged in size in comparison to those of the other indi- viduals. The nucleus was not present, and oocytes gave the appearance of having undergone germinal vesicle breakdown (GVBD) [36]. However, this individual did not spawn eggs within the duration of the experiment. Employing immunocytochemical techniques, we were able to confirm the accumulation of vitellin-immunoreactive material in oocytes in the latter stages of secondary vitel- logenesis. Figure 6 shows histology for female F1. Matur- ing oocytes partially filled with yolk globules indicate an immunocytochemical reaction (a), but immature oocytes from the center of the ovary exhibit no reaction (b). In the control version of the same individual, mature oocytes in sections which were incubated with pre-absorbed antibodies did not exhibit a reaction (c). Likewise, immature oocytes also showed no indication of staining (d). Methyl farnesoate (MF) administration in males Male juveniles were divided into four groups of ten individuals each and received combinations of ablation/ non-ablation and MF injection/saline injection. Results are shown in Table 1, along with average body weights for each Fic. 4. Examination of juvenile vitellogenin by SDS-PAGE and Western blotting. (a) Vitellogenin was shown to consist of a single polypeptide component of 199K in juvenile males (M1- M6) (ex. M3; lanes 1-5 correspond to Days 0, 5, 8, 13,19). In comparison, adult female vitellogenin (lane B) was composed of three polypeptides of 199, 102, and 90K, and purified vitellin (lane A) of 102 and 90K. Western blotting results were in agreement with enzyme immunoassay results. (b) Vitellogenin consisted of a single polypeptide component of 199K also in juvenile females (F1—F4) (ex. F1; lanes 1-6 correspond to Days 0, 4, 8, 12, 16, 20). Lanes A and B are asin (a). (c) Juvenile female F5, however, was exceptional. Vitellogenin first be- came detectable by enzyme immunoassay (lane 4—Day 12), showing only the 199K component, but thereafter, full vitel- logenin became very abundant, and all three peptide compo- nents, e.g. 199, 102 and 90K, were observed (lanes 5, 6—Days 16, 20), as in adult female vitellogenin. Lanes A and B are as in (a). 50 Fic. 6. Immunohistology, and examination of uptake of vitellin-immunoreactive material in oocytes (ex. F1). oocytes, occurrence of immunocytochemical reaction. control; absence of immunocytochentical reaction. M. N. WILDER AND T. Okumura et al. (a) Mature (b) Immature oocytes from core of ovary; absence of immunocy- tochemical reaction. (c) and (d) Same oocytes as in (a) and (b) respectively, incubated with pre-absorbed antiserum as a Scale bar: 50 um. TABLE 1. Experimental protocol for methyl farnesoate (MF) administration in eyestalk-ablated (ES) and non-ablated (intact) juvenile males. Data for final vitellogenin levels, average body weight, and average number of molts pare shown Schedule (Days) Treatment No. ind. MF inj. N=10 + ES Saline N=10 + ES MF inj. N=10 + Intact Saline N=10 + Intact ' Difference not significant. * 5 ug MF injected as suspension in 25 ul 5% ethanol/saline. o Blank injection as 25 yl ethanol saline. a Eyestalk ablation performed. h Hemolymph samples taken. 2 3 4 BW(e) * * * 3.24+0.38 0 Co) ) 3.82+0.11 * * * 3.21+0.21 Co) Co) fC) 2.37+0.24 No. molts 0.6+0.2 1.0+0.0 0.3+0.2 0.3+0.2 Vg (mg/ml) 0.15+0.04' 0.14+0.04' ND ND Vitellogenin in Juvenile Prawns 51 group. For ablated animals, number of molts was 0.6+0.2 and 1.0+0.0 for MF-injected, and saline-injected groups, respectively, but was 0.3+0.2 for both groups of intact individuals. At the end of the experiment, only 3 individuals in the MF-injected, and 2 individuals in the saline-injected ablated groups exhibited non-detectable (ND) values. There was no significant difference in average value between MF-injected animals (0.15+0.04 mg/ml) and saline-injected animals (0.14+0.04 mg/ml). Some of the ablated animals were examined by SDS-PAGE; these showed only the 199K banding pattern (data not shown). In non-ablated groups, all final EIA values were ND, irrespective of whether animals were injected with MF or with saline. DISCUSSION In this investigation, it was revealed that eyestalk remov- al in juvenile male and female M. rosenbergii results in the production of a vitellin-immunoreactive substance which con- sists of a 199K peptide component. Adult female vitel- logenin has been shown by Okumura [23] to be comprised of the 199K component as well as of lighter molecular weight components of 102K and 90K. In eyestalk-ablated juveniles of both sexes, 199K vitellogenin was seen to increase for 1—2 weeks and then decrease; in general, females reached higher overall levels (0.5—3 mg/ml) than did males (0.1-0.8 mg/ml). One female individual was exceptional; initially only 199K vitellogenin appeared, but thereafter, full three-component vitellogenin was manifested with levels in the hemolymph reaching nearly 30 mg/ml. The advancement of ovarian development in eyestalk- ablated juvenile females was confirmed for all individuals. In females F1—F4 expressing only the 199K component, gonadosomatic indices (GSI) ranged from 0.54 to 1.02%. In comparison to the state of the immature ovary in the adult female, ovaries in these females were enlarged in appearance and bright orange in color, and thus similar to normally maturing ovaries in the adult female. Histological examina- tion revealed that oocytes were in secondary vitellogenesis, with the nucleus uncondensed in appearance. In the excep- tional female F5, GSI was 4.10% and oocytes appeared to have already undergone germinal vesicle breakdown (GVBD). These results suggested a correlation between the extent of ovarian development and the ability to produce full vitellogenin; whether this is related to an ovarian factor such as vitellogenesis-stimulating ovarian hormone (VSOH), is as of yet unclear. The accumulation of yolk proteins in oocytes of females expressing only the 199K band was further confirmed using immunocytochemical techniques. In females F1—F4, matur- ing oocytes in secondary vitellogenesis stained via immunocy- tochemical reaction, and immature oocytes in the core of the ovary exhibited no reaction. This indicated that although juvenile vitellogenin lacked the 102 and 90K components of adult vitellogenin, increases in the hemolymph of the 199K peptide were correlated with oocyte development, providing further support that juveniles, not only females, but also males have the ability to produce vitellogenin to a certain extent. The small size of the ovaries of the above individuals did not permit analysis on SDS-PAGE in parallel to the above histological studies; this will be investigated subse- quently. While little information is available concerning vitel- logenin in juveniles, vitellogenin and vitellin structure have been examined in a number of adult crustacean species [8, 25, 26, 34]. Adult M. rosenbergii seems similar to many of these; for example, in Penaeus monodon, vitellogenin and vitellin are comprised of subunits of 74, 83, 104 and 168K [26]. Less is known however, concerning how crustacean vitellogenins and vitellins are processed while becoming sequestered in eggs. In the terrestrial isopod, Armadilli- dium vulgare [30], four forms of vitellin are initially accumu- lated in oocytes, but the higher molecular weight ones undergo proteolytic processing, leaving only the lightest component. Komatsu and Ando [18] have reported a low density lipoprotein (LDL) present in the egg yolk of the sand crayfish, [bacus ciliatus, which degrades vitellogenin and may be involved in vitellogenin processing in this species. The native form of vitellin in M. nipponense has been estimated by gel filtration as 350K [14], and preliminary work indicates that this is similar in M. rosenbergii (Okumura, unpublished data). In M. rosenbergii, as vitellin yields the 102K and 90K components on SDS-PAGE, it seems plausible that the 199K protein is initially synthesized as a precursor vitellogenin and undergoes further processing in the mature female in relation to a factor perhaps produced by the ovary. Vitellin in its final form may be a multimer association of the 102K and 90K proteins subunits, thus the 199K protein would not appear in eggs, but occurs in the hemolymph. Of note, in Derelle et al. [12], M. rosenbergii female vitellogenin and vitellin examined on SDS-PAGE was comprised of subunits of 84 and 92.2K. Regarding vitellin, these results are similar to those obtained in our investigation. However, these authors did not observe a higher molecular weight component in hemolymph vitellogenin, such as the 199K component obtained here. Discrepancies in results between their inves- tigation and this study could be due to differences in methods of antisera preparation. This is the first report to our knowledge concerning vitellogenin production in male decapods, but in isopods, this can occur as a result of loss of androgenic gland function due to natural or artificial circumstances [28, 31], In A. vulgare andrectomized males, vitellogenin levels are in fact higher than those of normal females; male vitellogenin does not differ electrophoretically from that of females [31], although the ovary is suggested to be necessary to maintain vitellogenin titers during the molt cycle. In Porcellio dilatatus, fat bodies from surgically-untreated males synthesize vitellogenin in vitro [13]. In M. rosenbergii, vitellogenin production can be induced by simply removing the eyestalks. As some sinus gland peptides are thought to be involved in the maintenance of the male reproductive system and the control of the 52 M. N. WILDER AND T. OxkuMuRA et al. androgenic gland in other decapod species [3, 10], in this investigation, whether the appearance of vitellogenin in M. rosenbergii is due to the absence of a putative VIH or is more related to the removal of other eyestalk factors, is at present unclear. However, vitellogenin levels in males do not reach those of juvenile females, suggesting that there may be additional factors of female origin involved in further stimula- tion of early vitellogenin production. On the other hand, in adult males, eyestalk ablation results in the appearance of vitellogenin in only occasional individuals (Yang et al., un- published data); this suggests some involvement of the androgenic gland in that it is expected to be less developed and therefore less inhibitory in juveniles than in adults. In male insects, several authors have studied by immuno- logical and electrophoretic methods the induction of vitel- logenin synthesis by juvenile hormone and/or ecdysone treat- ment [2, 16, 21]. The actions of juvenoids and ecdysteroids vary with species. Agui et al. [4] have examined production of vitellogenin mRNA in male and female housefly Musca domestica, revealing that in males, 20-hydroxyecdysone, but not juvenile hormone can stimulate transcription of the vitellogenin genes; this has indicated the necessity of the ovary. In Diploptera punctata, vitellogenin synthesis is directly inducible by application of juvenile hormone ana- logue [21]. Induced male vitellogenin levels are often lower than levels in normal females (similar to M. rosenbergii). Along these lines, it was postulated that methyl farneso- ate (MF) may be necessary for further stimulation and processing of vitellogenin in M. rosenbergii. MF being highly insoluble in water, was firstly dissolved in stock solutions of ethanol, which were adjusted in crustacean saline. This formed a cloudy suspension. However, as discussed in the results, MF injection did not produce signif- icantly higher levels of vitellogenin (0.15+0.04 mg/ml) above those of ablated animals injected with saline (0.14+ 0.04 mg/ml). MF injection also did not induce vitellogenin synthesis in non-ablated animals. In preliminary work, other methods of injection were attempted, for example, cardiac injection with either MF (and farnesoic acid) in saline or in purified sesame oil, or similar intramuscular injections using oil instead, but these were also ineffective in stimulating vitellogenin production in either sex. Thus, eyestalk remov- al was observed to be a prerequisite for induction of vitel- logenin production in males, but MF administration seemed to have no influence. From these results, it was not possible to obtain positive evidence that MF has any connection to vitellogenin produc- tion in M. rosenbergii, but it also can not be conclusively stated that MF has no physiological role whatsoever. MEF has been shown to activate Na/K-ATPase in Artemia, in- dicating its potential role in osmoregulation and molting [6, 7]. MF has been detected in both sexes in M. rosenbergii by Sagi et ai. [27]; we have additionally determined that MF is present in females during both the reproductive molt and common molt cycles, and seems highest during the early premolt stages [37]. Additionally, it is not ruled out that MF has no involvement in the process of vitellogenin uptake. In insects, juvenoids affect increased membrane Na/K-ATPase activity in the ovarian follicles, causing cell volume to shrink, and creating spaces through which vitellogenin can pass through to access the oocytes [1, 11, 17]. This is known as “patency”. The results of this study have suggested that vitellogene- sis- and molt-inhibiting eyestalk factors are present in juve- niles as well, but this is not sufficient to explain the means by which precursor vitellogenin is processed into the full vitel- logenin observed in the adult female and how vitellogenin production is possible in males. In subsequent studies, it will be important to investigate the involvement of ovarian factors in concert with MF and the relationship between androgenic gland functioning and male vitellogenin produc- tion. ACKNOWLEDGMENTS We thank Kuraray Co., Ltd. for the farnesoic acid sample used to prepare methyl farnesoate (MF), and S. Okada, The University of Tokyo, for GC-MS of sample MF. This study was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Educa- tion, Science, and Culture of Japan. Additionally, M.N. Wilder acknowledges the support of a Japan Society for the Promotion of Science Postdoctoral Fellowship. REFERENCES 1 Abu-Hakima R, Davey KG (1977) The action of juvenile hormone on the follicle cells of Rhodnius prolixus: The import- ance of volume changes. J Exp Biol 69: 33-44 2 Adams TS, Filippi PA, Kelly TJ (1989) Effect of 20- hydroxyecdysone and a juvenile hormone analogue on vitel- logenin production in male houseflies, Musca domestica. J Insect Physiol 35: 765-773 3 Adiyodi R (1984) Seasonal changes and the role of eyestalks in the activity of the androgenic gland of the crab, Paratelphusa hydrodromous (Herbst). Comp Physiol Ecol 9: 427-431 4 Agui N, Shimada T, Izumi S, Tomino S (1991) Hormonal control of vitellogenin mRNA levels in the male and female housefly, Musca domestica. J Insect Physiol 37: 383-390 5 Aguilar MB, Quackenbush LS, Hunt DT, Shabanowitz J, Huberman A (1992) Identification, purification and initial characterization of the vitellogenesis-inhibiting hormone from the Mexican crayfish Procambarus bouvieri (Ortmann). Comp Biochem Physiol 102B: 491-498 6 Ahl JSB, Brown JJ (1990) Salt-dependent effects of juvenile hormone and related compounds in larvae of the brine shrimp, Artemia. Comp Biochem Physiol 95A: 491-496 7 Ahl JSB, Brown JJ (1991) The effect of juvenile hormone III, methyl farnesoate, and methoprene on Na/K-ATPase activity in larvae of the brine shrimp, Artemia. Comp Biochem Physiol 100A: 155-158 8 Andrieux N, de Frescheville J (1992) Caractérisation de la vitelline chez le Crustacé Brachyoure Carcinus maenus. CR Acad Sci t314, ser III: 227-230 9 Chang ES, Prestwich GD, Bruce MJ (1990) Amino acid sequ- ence of a peptide with both molt-inhibiting and hyperglycemic activites in the lobster, Homarus americanus. Biochem Bio- phys Res Commun 171: 818-826 10 Charniaux-Cotton H, Payen GG (1988) Crustacean reproduc- 11 12 ile) 14 15 16 17 18 19 20 21 22 23 24 25 Vitellogenin in Juvenile Prawns 33) tion. In “Endocrinology of Selected Invertebrate Types” Ed by H Laufer and RGH Downer, Alan R. Liss, Inc. pp 279-303. Davey KG, Huebner E (1974) The response of the follicle cells of Rhodnius prolixus to juvenile hormone and antigonadotropin in vitro. Can J Zool 52: 1407-1412 Derelle E, Grosclaude J, Meusy J-J, Junéra H, Martin M (1986) ELISA titration of vitellogenin and vitellin in the freshwater prawn, Macrobrachium rosenbergii, with monoclonal antibody. Comp Biochem Physiol 85B: 1-4 Gohar M, Souty C (1983) Mise en évidence in vitro d’une synthése et d’une libération de vitellogénine dans le tissu adipeux male de Porcellio dilatatus Brandt (Crustacé Isopode terrestre). CR Acad Sci Paris, t297: 145-148 Han C-H (1988) Physiological and reproductive studies on a freshwater prawn, Macrobrachium nipponense (De Haan). Ph. D. Dissertation, The University of Tokyo, Japan. Hinsch GW, Bennett DC (1979) Vitellogenesis stimulated by thoracic ganglion implants into destalked immature spider crabs, Libinia emarginata. Tissue Cell 11: 345-351. Huybrechts R, De Loof A (1977) Induction of vitellogenin synthesis in male Sarcophaga bullata by ecdysterone. J Insect Physiol 23: 1359-1362 Ilenchuk TT, Davey KG (1987) The development of respon- siveness to juvenile hormone in the follicle cells of Rhodnius prolixus. Insect Biochem 17: 525-529 Komatsu M, Ando S (1992) A novel low-density lipoprotein with large amounts of phospholipid found in the egg yolk of crustacea sand crayfish Jbacus ciliatus: Its function as vitel- logenin-degrading proteinase. Biochem Biophys Res Commun 186: 498-502 Laufer H, Borst D, Baker FC, Carrasco C, Sinkus M, Reuter CC, Tsai LW, Schooley DA (1987) Identification of a juvenile hormone-like compound in a crustacean. Science 235: 202-205 Meusy J-J, Payen GG (1988) Female reproduction in malacos- tracan Crustacea. Zool Sci 5: 217-265 Mundall EC, Szibbo CM, Tobe SS (1983) witellorentn induced in adult male Diplotera punctata by juvenile hormone and juvenile hormone analogue: Identification and quantitative aspects. J Insect Physiol 29: 201-207 Nagamine C, Knight AW (1987) Induction of female breeding characteristics by ovarian tissue implants in androgenic gland ablated male freshwater prawns Macrobrachium rosenbergii (de Man) (Decapoda, Palaemonidae). Int J Invertebr Reprod Dev 11: 225-234 Okumura T (1992) Physiological studies on molting and gonad- al maturation in prawns. Ph.D. Dissertation, The University of Tokyo, Japan Okumura T, Han C-H, Suzuki Y, Aida K, Hanyu I (1992) Changes in hemolymph vitellogenin and ecdysteroid levels dur- ing the reproductive and non-reproductive molt cycles in the freshwater prawn, Macrobrachium nipponense. Zool Sci 9: 37- 45 Quinitio ET, Hara A, Yamauchi K, Mizushima T, Fuji A (1989) Identification and characterization of vitellin in a hermaphrodite 26 27 28 29 30 31 32 33 34 35 36 37 38 39 shrimp, Pandalus kessleri. Comp Biochem Physiol 94B: 445— 451 Quinitio ET, Hara A, Yamauchi K, Fuji A (1990) Isolation and characterization of vitellin from the ovary of Penaeus monodon. Invertebr Reprod Dev 17: 221-227 Sagi A, Homola E, Laufer H (1991) Methyl farnesoate in the prawn Macrobrachium rosenbergii: Synthesis by the mandibular organ in vitro, and titers in the hemolymph. Comp Biochem Physiol 99B: 879-882 Souty-Grosset C, Juchault P (1987) Etude de la synthése de la vitellogénine chez les males intersexués d’Armadillidium vulgare Latreille (Crustacé Isopode Oniscoide): Comparison avec les males et les femelles intactes ou ovariectomisées. Gen Comp Endocrinol 66: 163-170 Soyez D, Le Caer JP, Noel PY, Rossier J (1991) Primary structure of two isoforms of the vitellogenesis inhibiting hor- mone from the lobster Homarus americanus. Neuropeptides 20: 25-32 Suzuki S (1987) Vitellins and vitellogenins of the terrestrial isopod, Armadillidium vulgare. Biol Bull 173: 345-354 Suzuki S, Yamasaki K, Katakura Y (1990) Vitellogenin synth- esis in andrectomized males of the terrestrial isopod, Armadilli- dium vulgare (Malacostracan Crustacea). Gen Comp Endocri- nol 77: 283-291 Suzuki S, Yamasaki K (1991) Ovarian control of oostegite formation in the terrestrial isopod, Armadillidium vulgare (Malacostraca, Crustacea). Gen Comp Endocrinol 84: 381- 388 Takayanagi H, Yamamoto Y, Takeda N (1986) An ovary- stimulating factor in the shrimp, Paratya compressa. J Exp Zool 240: 203-209 Tirumalai R, Subramoniam T (1992) Purification and charac- terization of vitellogenin and lipovitellins of the sand crab Emerita asiatica: Molecular aspects of crab yolk proteins. Mol Reprod Dev 33: 16-26 Webster SG, Keller R (1986) Purification, characterisation and amino acid composition of the putative moult-inhibiting hor- mone (MIH) of Carcinus maenas (Crustacea, Decapoda). J Comp Physiol B 156: 617-624 Wilder MN, Okumura T, Aida K (1991) Accumulation of ovarian ecdysteroids in synchronization with gonadal develop- ment in the giant freshwater prawn, Macrobrachium rosenbergii. Zool Science, 8: 919-927 Wilder MN (1993) Physiochemical studies on ecdysteroids and juvenoids in relation to molting, reproduction and embryo- genesis in the giant freshwater prawn, Macrobrachium rosenber- gli. Ph. D. Dissertation, The University of Tokyo, Japan. Wyatt GR (1988) Vitellogenin synthesis and the analysis of juvenile hormone action in locust fat body. Can J Zool 66: 2600-2610 Yano I (1992) Effect of thoracic ganglion on vitellogenin secretion in kuruma prawn, Penaeus japonicus. Bull Natl Res Inst Aqacult 0: 9-14 i are ey A coil INET Te = A A ageale ColbccT ia to edt, rhuchmgeamall OR, * sen 1s a A eh ARO watt i ails ia" mobicinae vity Gd a Heeak Ruphaare™ i « leery WHA | Unhan a ta ty of ii (eee connie Acie airs I Bh fs . ligitwlnett. tle lcs apenas ea TREY |) | “ - Whe Ga AYR Lola sh Hatsl , AOE) 4 heli, ob kcsenene tale Pe oy AY bea Toiyil est lady gal te ’ insnaltyy ae) Et mind oP hey yy) mille T 7" a reruTITTVen eG LEE Dy ee tRb i ID Setithyy the WES Ad | ptt behead hoin Ghiedvo? 0 - : trate 1 a] ae - ? ong yi, re ae ' i+ “1 Hassett yo ew he ated . é e “is tf bmn oe tae a water wate o a { bce ib’ Oye ees} . = hye iA ridtes Ne ae x 1 Cr eeyyieeg tas J y x mI , . y P ft: ‘ania yea iia filers Tolan autaomn eo] -noapibt 4 weep er te * aoe ped bey “irihalt ey Osesmeceyaim andy eae, aeaieey ber a 6 below’. parser omilegN, ere! maagedotey le auetietal con! | selina @ Zug eSya eO = Poa Lon « a a hae sTeL a Rates and a . td we eae lei Fa Weal A atall 7 Thuis ott IN cep rohreH Melee fe wise] «le mefenah PES SER naiemerids: Oahu OTe ale yp ee vatat.* >a aes ghar Y pres hi yD bap ‘ature perenne Sisiergeas SS Laquegy boi (esto maieotal ¢ Goce Tienes? sur he Prgyre” whi deaelath ae} ihe pirreliagh (Ragyy: Aili rhe spree ae (nt ane nace eee a Li) Orptytieatee rats haga array) tds commen Sl Ae ay tdereewel pred | hageps Qaceapeigy aHogse ae / 2 Ra@loumnd esdivindeorgione : om) weds 0.05) in Ey production was measured between the animal and the vegetal hemispheres. The production of 17a,206-diOHp by granulosa cells was also stimulated by the presence of PMSG or progesterone. The concentrations (more than 0.4 pg/ml; [9]) of 17a,208-diOHp were sufficient to induce oocyte maturation, and significantly greater in the vegetal hemisphere than in the animal hemisphere in the presence of PMSG. DISCUSSION The granulosa cells are in contact with the oocyte via cytoplasmic processes that pass through the pore canals in the chorion in the medaka [2] and the pipefish [1]. The present observations revealed morphological differences in granulosa cells localized at special regions (VPA) surrounding the oocyte. These differences seem to reflect the different in- teractions of the granulosa cells with the oocyte, the polarity of which is established during oogenesis. In the early stage of oogenesis, the VPA is determined by the position of the Balbiani body in the ooplasm, and the granulosa cells aligned compactly on the VPA (Iwamatsu, unpublished data). The attaching filaments differentiate and elongate on a restricted region of the chorion in the VPA, and then spirally wind around the VPA probably due to rotation of the oocyte and granulosa cells [6]. We recently found that growing oocytes with granulosa cells may rotate within the basement membrane [5]. The granulosa cells in the animal pole region possess many microvilli on their apical surfaces, but those in the remaining area do not. The difference in the distribution of the 82 T. Iwamatsu, S. NAKASHIMA et al. granulosa cells with microvilli may be related to a difference in the interactions with the basement membrane, or move- ment of granulosa cells beneath the basement membrane. Whether the morphological differences in the granulosa cells depend on regional differences in the chorion regions with which they are in contact is not clear from the present study, but it was suggested that the patterns of spiral structures on the chorion are due to the movement of follicular cells during oogenesis [6]. The steroidogenic response to exogenous hormones of granulosa cells from the animal pole hemisphere differed from that of cells from the vegetal hemisphere. The produc- tion of 17a,208-diOHp by granulosa cells that were stimu- lated by gonadotropin was greater in the vegetal hemisphere than in the animal hemisphere. A similar tendency was also observed in the steroid production by progesterone- stimulated follicles, although it was not significant statistical- ly. The difference may be due to the difference in cell number, since there are more cells in the vegetal hemisphere than in the animal hemisphere. On the other hand, another cause of the difference may relate to the high ability of the tall granulosa cells with crowded mitochondria to produce ster- oids in the vegetal pole area. The cells that produce steroids are generally filled with mitochondria with well-developed tubular cristae and tubular or dilated ER (see [12]). _Howev- er, no difference in Ey production was recognized between the two hemispheres, in spite of the difference in cell number. This may indicate that the E, production by each granulosa cell in the vegetal hemisphere is not different from that in the animal hemisphere during the maturation period. There- fore, the granulosa cells with gonadotropin receptors seem to differ physiologically or distributively along the animal- vegetal axis of the oocyte. ACKNOWLEDGMENTS The authors thank Dr. Cherrie A. Brown, California Regional Primate Research Center, University of California, Davis, for critical reading of the manuscript. REFERENCES 1 Begovac PC, Wallace RA (1989) Major vitelline envelope 10 11 12 13 14 15 16 proteins in pipefish oocytes originate within the follicle and are associated with the Z3 layer. J Exp Zool 251: 56-73 Hirose K (1972) The ultrastructure of the ovarian follicle of medaka, Oryzias latipes. Z Zellforsch 123: 316-329 Iwamatsu T (1975) Medaka as a teaching material. II. Matura- tion and fertilization of oocytes. Bull. Aichi Univ. Educ., 24 (Nat. Sci.): 113-144 (in Japanese) Iwamatsu T (1980) Studies on oocyte maturation of the meda- ka, Oryzias latipes. III. Role of follicle cells in gonadotropi- nand steroid-induced maturation of oocytes in vitro. J Exp Zool 206: 355-364 Iwamatsu T (1992) Morphology of filaments on the chorion of oocytes and eggs in the medaka. Zool Sci 9: 589-599 Iwamatsu T, Nakashima S, Onitake K (1993) Spiral patterns in the micropylar wall and filaments on the chorion in eggs of the medaka, Oryzias latipes. J Exp Zool 267: 225-232 Iwamatsu T, Ohta T (1981) On a relationship between oocyte and follicle cells around the time of ovulation in the medaka, Oryzias latipes. Annot Zool Japon 54: 17-29 Iwamatsu T, Ohta T, Oshima E, Sakai N (1988) Oogenesis in the medaka Oryzias latipes. —Stages of oocyte development. Zool Sci 5: 353-373 Iwamatsu T, Takahashi SY, Sakai N, Asai K (1987) Inductive and inhibitory actions of a low molecular weight serum factor on in vitro maturation of oocytes of the medaka. Biomed Res 8: 313-322 Iwamatsu T, Takahashi SY, Sakai N, Onitake K (1987) Induc- tion and inhibition of in vitro oocyte maturation and production of steroids in fish follicles by forskolin. J Exp Zool 241: 101- 111 Masui Y, Clarke HJ (1979) Oocyte maturation. Cytol 57: 185-282 Nagahama Y (1983) Functional morphology of teleost gonads. In “Fish Physiology”, vol. IXA (WS Hoar, AJ Randall and EM Donaldson, eds), pp. 223-275, Academic Press, New York Nagahama Y, Adachi S (1985) Identification of maturation- inducing steroid in a teleost, the amago salmon (Oncorhynchus rhodurus). Develop Biol 109: 428-435 Riehl R, Appelbaum S (1991) A unique adhesion apparatus on the eggs of the catfish Clarias gariepinus (Teleostei, Clariidae). Jap J Ichthyol 38: 191-197 Sakai N, Iwamatsu T, Yamauchi K, Suzuki N, Nagahama Y (1988) Influence of follicular development on steroid produc- tion of the medaka (Oryzias latipes) ovarian follicle in response to exogenous substances. Gen Comp Endocrinol 71: 516-523 Schuetz AW (1967) Action of hormoness on germinal vesicle breakdown in frog (Rana pipiens) oocytes. J Exp Zool 166: 347-354 Intern Rev ZOOLOGICAL SCIENCE 11: 83-87 (1994) © 1994 Zoological Society of Japan Immunohistochemical Localization of Epidermal Growth Factor in the Developing Rat Gonads YASUHIKO KANNO, SATOSHI Koike! and Tetsuo NoUMURA~ Department of Regulation Biology, Faculty of Science, Saitama University, Urawa, Saitama 338, and Upjohn Pharmaceuticals Limited, Wadai, Tsukuba, Ibaraki 300-42, Japan ABSTRACT—Fpidermal growth factor (EGF) is known to have various endocrine, paracrine and autocrine roles in adult mammalian tissues. In order to clarify the participation of EGF in rat gonadal differentiation, immunohistochemical localization of EGF as chronologically studied in perinatal rat gonads. Sprague-Dawley rat gonads from gestational day (GD) 13 to postnatal day (PD) 21 were fixed in acetic acid-free Bouin’s solution and stained with a polyclonal antibody against mouse EGF by using avidin-biotin complex technique. Immunohistochemical reactivity was positive in almost all cell types in the prenatal male gonads. Male germ and Leydig/interstitial cells showed a positive reactivity from GD 15 to 21. Slight and moderate staining were seen in the Sertoli/supporting cells from GD 13 to 21. After birth, positive expression was not seen in any types of cells in male gonads except for germ cells on PD 21. On the other hand, in prenatal female gonads positive signs were found in the interstitial cells from GD 14 to 21 and in the granulosa cells on GD 21. During the postnatal period from PD 5 to 21, the granulosa and theca cells were slightly positive and the interstitial cells moderately positive. Wolffian ducts in males and Miillerian ducts in females were stained during the prenatal period. These results indicate that EGF shows stage-specific patterns of expression in the developing rat gonadal cells and may be involved in the rat gonadal development and diffferentiation. INTRODUCTION Epidermal growth factor (EGF), a single-chain poly- peptide of 53 amino acid residues, was first isolated from the submandibular gland of male mice [8]. This peptide is a potent mitogen for a wide variety of cell types in vivo and in vitro and distributes in various tissues and fluids in mammals [7]. EGF is mainly synthesized in the submandibular gland of mice [5] and rats [22], and its synthesis is under the control of androgens [6] and thyroid hormones [9]. EGF is involved in many physiological functions in the adult gonads [24]. EGF reduced follicle-stimulating hor- mone (FSH)-stimulated testosterone production in rat Leydig cells in vitro [15], which was due to a regulation in 17f- hydroxylase activity [25]. EGF also attenuated FSH- stimulated aromatase activity in the Sertoli cells [2] and the granulosa cells [15], but stimulated Sertoli cell proliferation [16]. SH-stimulated inhibin production in the Sertoli cells [20] and plasminogen activator production in the granulosa cells were increased by EGF [10]. Although basic properties and functions of EGF have been reported in adults, little is known about its contribution to the fetal gonadal develop- ment and diffferentiation. In order to clarify the participation of EGF in developing rat gonads, immunohistochemical expression of EGF was chronologically studied in the fetal and prepubertal rat gonads from gestational day (GD) 13 to postnatal day (PD) 721M Accepted December 17, 1993 Received October 21, 1993 2 To whom reprint requests should be addressed. MATERIALS AND METHODS Aminals Crj: CD (Sprague-Dawley) rats from 13 to 20 weeks of age were housed in constant temperature (23+1°C), relative humidity (55+ 10%) and light-dark cycle (light on 7:00-19:00). Animals were given a commercial diet (CRF-1, Charles River Japan Co.) and tap water ad libitum. Cohabitation was done in the evening of vaginal proestrus in the 1:1 basis of male: female. In the next morning, copulation was checked by the presence of sperm in the vaginal smear. The day when sperm-positive smear was found was desig- nated as GD 0, and the day when litter was found was designated as PD 0. Preparation of tissues for immunostaining Dams were sacrificed from GD 13 to 21 and neonates on PDs 5, 11 and 21 by diethyl ether anesthesia. The gonads and genital ducts dissected from more than three fetuses and pups in different litters were fixed in acetic acid-free Bouin’s solution for a few hours at 4°C. The sexes of fetuses were determined as described by Agelopoulou et al. [1]. Then the tissues were dehydrated through a series of graded concentrations of ethanol and xylene, embedded in paraffin and sectioned in 5 wm thickness. The male rat submandibular glands on PD 35 were also dissected and fixed in the same fixative, for checking the specificity of the antibody used. Immunohistochemistry Sections were deparaffinized with xylene and hydrated in de- creasing concentrations of ethanol. The sections were incubated with 0.5% periodic acid (Sigma Chemical Co.) in 50mM Tris- buffered saline (TBS, pH 7.6, Wako Pure Chemical Industries, Ltd.) at room temperature for 30 min to block endogenous peroxidase. Sections were subsequently rinsed with 50mM TBS for 20 min, blocked non-specific staining with 1.5% normal goat serum in 50 mM 84 Y. Kanno, S. KOIKE AND T. NOUMURA TBS for 20 min. Then, sections incubated overnight at 4°C with the polyclonal antibody against mouse EGF raised in the rabbit (Col- laborative Research Inc.) at 0.02 mg/ml in 10mM PBS. Dose response study indicated that this concentration of the antibody gave optimal labelling results. Follwing this incubation the sections were rinsed with TBS and then treated with 0.5% biotinylated goat anti-rabbit secondary antibody (Vector Lab, Inc. ABC kit Elite) diluted in 50mM TBS for 30 min at room temperature. Sections were again washed in TBS and subsequently incubated with 2% avidin-biotin complex (Vector Lab, Inc, ABC kit Elite) in 50 mM TBS for 40min at room temperature. Avidin and biotin were prepared at least 30 min before applied to the sections to allow the complex to form. The sections were again washed in TBS, and the bound antibody was visualized with 0.05% 3,3’-diaminobenzidine tetrachloride (Dojindo Laboratories) in 50mM TBS (pH7.2) and 0.02% hydrogen peroxide for 5 min. These sections were counter- stained with hematoxylin. Controls included (a) replacing the primary antibody with nor- mal rabbit serum, (b) using the primary antibody that had been pre-incubated overnight at 4°C with 0.04 mg/ml mouse EGF (Chemi- con International, Inc.) before this mixture was applied to the section to check specificity of the primary antibody, and (c) omitting the primary antibody to check the specificity of the secondary antibody. RESULTS Specificity of antibody Preparations which were stained with the antibody to EGF, and with the immunoneutralized antibody by pre- incubated with the antigen were shown in Fig.1. EGF antibody stained the granular convoluted tubule cells of submandibular gland in male rats on PD 35, but the neutral- ized antibody did not stain any cells. Therefore, these results showed that this polyclonal antibody specifically stained EGF-containing cells, because the rat submandibular gland was confirmed to contain EGF-positive cells [22]. Immunohistochemical localization Immunohistochemical expression of EGF in the develop- ing rat gonads was summarized in Table 1. In male gonads, many germ cells were moderately stained from GD 15 to 21 and PD 21 (Fig. 2.A-C). A few Sertoli/supporting cells and almost all Leydig/interstitial cells showed positive reactivity; the staining intensity was moderate or slight in Sertoli/ supporting cells from GD 13, the time when the supporting cells were firstly recognized by histological examination, to GD 21 and marked in Leydig/interstitial cells from GD 15 to 21 (Fig. 2.A-C). On the other hand, in female gonads, a few interstitial cells expressed slight reactivity from GD 14 to 21 and moderate from GD 21 to PD 21 (Fig. 2.E-G). A few theca cells slightly stained from PD 5 to 21 and a few granulosa cells slightly stained from GD 21 to PD 21 (Fig. 2.E-G). But the germ cells in females were not shown any positive reactivity during the experimental period (Fig. 2.E- G). The epithelial cells of the genital ducts were stained in sex-specific manner during the prenatal period: the Wolffian ducts in males were moderately or markedly stained from GD 16 to 21 (Fig. 2.D) and the Miillelrian ducts in females were slightly or moderately stained from GD 16 to PD 21 (Fig. 2.H). (A), or with the neutralized antibody (B). Staining is seen in the granular convoluted tubule cells of A, but those of B. Bar: 50 um. Fic. 2. Immunohistochemical localization of EGF in the gonad of perinatal rats. (A) Male gonad on GD 14. The Sertoli/ supporting cells (arrow) show slight staining. (B) Male gonad on GD 17. In the Leydig/interstitial cells (arrow head), and the moderate stainings in the germ (thick arrow) and the Sertoli/supporting cells (arrow) express marked staining. (C) Male gonad on PD 5. Gonadal cells do not express any positive sign. (D) Male Wolffian duct andon GD 17. The epithelial cells (arrow) show marked staining. (E) and (F) Female gonads on GDs 14 and 17. The interstitial cells (arrow) express slight staining. (G) Female gonad on PD 5. The interstitial cells (arrow head) are stained moderately and the granulosa (arrow) and theca (thich arrow) cells slightly. (H) Female Miillerian duct on GD 17. The epithelial cells (arrow) are moderately stained. All sections are shown in the same magnification. Bar: 50 «~m. aed GF TPES Gs Fearn oe & 86 Y. Kanno, S. KoIKE AND T. NOUMURA TasLe 1. Immunohistochemical localization of EGF in the developing rat gonad Male Female Gonad Duct Gonad Duct G S P L Ww M G Gr T I Ww M GD 13 = A fj * * = * = * * * = * 14 = ap = * * al = a = 2 +/— — — 15 apap ae = * aF ar ar a = = = * ae = = = 16 ++ +/—- © +44 ++ = ~ - * a —- ++ 17 apap apap 8 aP aP ar ar SP AF * = = + =F = Sar 18 ++ +4+/—- — +++ deat * = = * 4 — + 19 ES ERPS SO abt ++ “ - — ‘ + * + 20 apart are SP aR aF SP ar * = = * AF * =F 21 ++ 4/—- = +44 ++ * —- +/- * ++ * ++ PD 5 — = = = = * = se ap aR SE = * Far 11 _ — = — - * — +/— +/— ++ * + 21 ++ 4/- = — = * — +/—= +/—= +4 * + GD: Gestational day, PD: Postnatal day, G: Germ cell, S: Sertoli/Supporting cell, P: Peritubular cell, L: Leydig cell, W: Wolffian duct, M: Miullerian duct, Gr: Granulosa/Supporting cell, T: Theca cell, I: Interstitial/Stromal cell, Grade, —: Not detectable, +: Slight but above background levels, + +: Moderate, + ++: Marked staining, *: The cells or tissues were not found in that day. DISCUSSION Differentiation and development of gonads are complex events which involve various cell-cell interactions. EGFisa potent mitogen [12, 13] and affects to gonadal steroid and peptide syntheses [15, 26]. Male germ cells showed a posi- tive reactivity from GD 15. Germ cells in males are encom- passed by the somatic cells from GD 14 ([{11] and this experiment) and thereafter undergo mitotic division. Therefore, positive expression from GD 15 indicates that EGF may have a function in the germ cell proliferation during the fetal period. The positive reactivity to anti-EGF antiserum was shown from GD 13 in the somatic cells of the developing gonads. The supporting and interstitial cells in the gonads proliferate rapidly during the fetal period [19]. Therefore, the express- ion in the Sertoli/supporting and Leydig/interstitial cells may indicate that EGF stimulates the proliferation of these gonad- al somatic cells as well as the germ cells. IGF decreased FSH-stimulated testosterone production by regulating 17/- hydroxylase activity in Leydig cells in vitro [15, 25]. Fetal rat testes have steroidogenic activity during the prenatal period in vitro [21]. So positive reactivity in the Leydig/ interstitial cells indicates that EGF may contribute to the fetal Leydig cell steroidogenesis. But the marked staining of the Leydig/interstitial cells during the prenatal period was lost after birth. This may be due to a regression of fetal Leydig cells which begin shortly before birth and replace to adult- type Leydig cells after birth in the rat [11]. The interstitial cells in female gonads have not been known to participate in physiological roles on the ovary, at least, during the prenatal and early postnatal period. The positive staining in these cells suggests that EGF is likely to mediate the ovarian functions in autocrine and/or paracrine fashions. EGF stimulates or inhibits the immature rat gra- nulosa cell proliferation depending on the presence or abs- ence of FSH [3]. In addition, the theca cells produce an EGF-like substance, which may regulate the granulosa cell functions [23]. Together with these results, positive staining in the granulosa and theca cells during the perinatal period suggests that EGF contributes to the follicle formation through the granulosa and theca cell poliferation. The immuno-reactivity was seen in the Wolffian ducts in males and Miillerian ducts in females. EGF plays important roles in the male and female reproductive-tract development during the perinatal or critical period [4, 14]. Therefore, the immuno-positive expression in the gonoducts in this stage supports the previous findings that EGF has been reported to be concerning with the development and differentiation in the gonoducts. Other growth factors have also been reported to partici- pate in the gonadal development in rat fetuses. Transform- ing growth factor-@ (TGF-f) localized in many types of cells in developing fetal gonads [18], like EGF. However, in- hibin-a, a member of TGF- superfamily, localized in Sertoli cells at the time of seminiferous tubule formation and in the Leydig cells during the late gestational period [17]. Together with these results, growth factors could play an important role in the gonadal development during the peri- natal period. The mechanisms and roles of these growth EGF Expression in Developing Rat Gonads 87 factors on gonadal development are under investigation. 10 11 12 REFERENCES Agelopoulou R, Magre S, Patsavoudi E, Jost A (1984) Initial phases of the rat testis differentiation in vitro. J Embryol Exp Morph 83: 15-31 Ascoli M, Euffa J, Segaloff DL (1986) Epidermal growth factor activates steroid biosynthesis in cultured Leydig tumor cells without affecting the levels of cAMP and potentiates the activation of steroid biosynthesis by choriogonadotropin and cAMP. J Biol Chem 262: 9196-9203 Bendell JJ, Dorrington JH (1990) Epidermal growth factor influences growth and differentiation of rat granulosa cells. Endocrinology 127: 533-540 Bossert NL, Nelson KG, Ross KA, Takahashi T, MacLachlan JA (1990) Epidermal growth factor binding and receptor dis- tribution in the mouse reproductive tract during development. Dev Biol 142: 75-85 Byyny RL, Orth DN, Cohen S (1972) Radioimmunoassay of epidermal growth factor. Endocrinology 90: 1261-1266 Byyny RL, Orth DN, Cohen S, Doyne ES (1974) Epidermal growth factor: effects of androgens and adrenergic agents. Endocrinology 95: 776-782 Carpenter G, Wahl MI (1991) The epidermal growth factor family. In “Peptide Growth Factors and Their Receptors II” Ed by MB Sporn and AB Roberts, Springer-Verlag, New York, pp 70-171 Cohen S (1962) Isolation of mouse submaxillarly gland protein accelerating incisor eruption and eyelid opening in the newborn animal. J Biol Chem 237: 1555-1562 Fisher DA (1990) Hormone epidermal growth factor interac- tions in development. Hormone Research 33: 69-75 Galway AB, Oikawa M, Tor NY, Hasueh AJW (1989) Epidermal growth factor stimulates tissue plasminogen activator activity and messenger ribonucleic acid levels in cultured rat granulosa cells: mediation by pathways independent of protein kinases-A and -C. Endocrinology 125: 126-135 Gondos B (1977) Testicular development. In “The Testis” Vol. 4, Ed by AD Johnson, WR Gomes, Academic Press, New York, pp 1-37 Gospodarowicz D. Ill CR, Birdwell CR (1977) Effects of fibroblast and epidermal growth factors on ovarian cell prolifra- tion in vitro. 1. Characterization of the response of granulosa cells to FGF and EGF. II. Proliferative response of luteal cells to FGF but not EGF. Endocrinology 100: 1108-1120; 1121- 1128 13 15 16 17 18 19 20 21 22 23 24 25 26 Gospodarowicz D, Bialecki H (1979) Fibroblast and epidermal growth factors are mitogenic agents for cultured granulosa cells of rodent, porcine, and human origin. Endocrinology 104: 757-764 Gupta C, Siegel S, Ellis D (1991) The role in testosterone- induced reproductive tract differentiation. Dev Biol 146: 106- 116 Hsueh AJW, Welsh TH, Jones PJC (1981) Inhibition of ova- rian and testicular steroidogenesis by epidermal growth factor. Endocrinology 108: 2002-2004 Jaillard, C, Chatelain PG, Saez JM (1987) Jn vitro regulation of pig Sertoli cell growth and function: effects of fibroblast growth factor and somatomedin-C. Biol Reprod 37: 665-674 Koike S, Noumura T (1993a) Immunohistochemical express- ion of inhibin-a subunit in the developing rat gonads. Zool Sci 10: 449-454 Koike S, Noumura T (1993b) Immunohistochemical localiza- tions of TGF- in the developing rat gonads. Zool Sci 10: 671- 677 Loading DW, de Kretser DM (1972) Comparative ultras- tructural and histochemical studies of the interstitial cells of the rat testis during fetal and postnatal development. J Reprod Fertil 29: 216-269 Morris PL, Vale WW, Cappel S, Bardin CW (1988) Inhibin production by primary Sertoli cell-enriched cultures: regulation by follicle-stimulating hormone, androgens, and epidermal growth factor. Endocrinology 122: 717-725 Noumura T, Weisz J, Lloyd GW (1966) Jn vitro conversion of 7--H-progesterone to androgens in the rat testis during the second half of fetal life. Endocrinology 78: 245-253 Poulsen SS, Nexo E, Olsen PS, Hess J, Kirkegaard P (1986) Immunohistochemical localization of epidermal growth factor in rat and man. Histochemistry 85: 389-394 Skinner MK, Lobb D, Dorrington JH (1987) Ovarian thecal/ interstitial cells produce an epidermal growth factor-like subst- ance. Endocrinology 121: 1892-1899 Tsutsumi O, Kurachi H, Oka T (1986) A physiological role of epidermal growth factor in male reproductive function. Scien- ce 233: 975-977 Welsh THJr, Hsueh AJW (1982) Mechanism of the inhibitory action of epidermal growth factor on testicular androgen biosynthesis in vitro. Endocrinology 110: 1498-1506 Zhiwen, Z, Carson RS, Herington AC, Lee VWK, Burger HG (1987) Follicle-stimulating hormone and somatomedin-C stimulate inhibin production by granulosa cells in vitro. Endoc- rinology 120: 1633-1638 ole las) Ter? oy ve enteaipe OY STENT wt Pear aay o am nun Piers neve” fT 7) : bit PEF (immer) GC) ait taser 9 rede) ai Ls sr ltevllits We! syvilsuboryay hosyettenl ” atl mt 4E weleW! VELA Asarh 37 hie Va homieea? Gee WA 4 {ret evttos an. - St if rantormca hight 9 tar bietlied. «dl “ ~ » fies rise Sy i it CTW mb) 2 sia +45 virleirm aoe te a ad . | Y sworn 4 oa. DG ha D pt tenga Vg f j I “o) Sein f i 30 4 igre! wa) tet ig ja? “4 w~ ’ 5e2 aed Gebel ‘ i strife os : | al ha Wiesel ry 4 4 i Se roe lier bi +) stas? = o-, ‘ =| - ° ' Per ele | i “hace t ‘ pen) i } a: f o . " 0 } : ry wite (tere _— (i innit Uns ee: “i ral aA j ty (Sooty tit pmael rf UHL pA, tenet Ene a “ah 6, hPa SF ieghdial Aas BLES Lea te Mactan) aay ie ery 1 Krone 2 Tore ie 7) iy hishat oh ” AAW - ed) ve ‘ pet ioe are ii ‘tdoen oes ie 4 ark: mate rnesen* At HoOct seg FS) pies. hi a Re. Perna ‘ fro 397.1+3.7 (9) Number of animals is in parentheses. 1), 2), 3) and 4) are different from each control value (p<0.01). No significant difference between 1 and 2, and significant difference between 3 and 4 (p<0.01). Enhancement of pituitary hemorrhage Pituitary hemorrhage was induced by a low dose of 0.02 ml/g bw (a sub-threshold dose) of a 35% glucose solution in combination with pretreatment of following drugs. The results indicated that bleeding occurred in 8 out of 10 mice (80%) with norepinephrine, in 5 out of 7 mice (71.4%) with epinephrine, and in 16 out of 20 mice (80%) with bromocrip- tine, and in 9 out of 10 mice (90%) with thiamazole, respec- tively (Table 6). No effects were observed with injections of adrenocorticotropin, vasopressin, serotonin and serotonin precursor. TABLE 6. Enhancement of pituitary hemorrhage by agents given subcutaneously prior to the intraperitoneal injection of 35% glucose at a low dose of 0.02 ml/g bw No. of mice Dose hemorrhaged Bents (ug/g) /no. of mice tested (%) Control — 0/27 (0) Phenylephrine 0/8 (0) Isoproterenol 0/6 (0) Phenoxybenzamine 20 0/8 (0) Propranolol 1 0/6 (0) Adrenocorticotropin 40 mU 1/10 (10) Vasopressin 5 mU 0/10 (0) 5-Hydroxytryptophan 253 days 0/7 (0) 5-Hydroxytrypatamine 2 1/10 (10) Norepinephrine 3 8/10 (80)* Epinephrine 2 5/7 (71.4)* Bromocriptine 5 16/20 (80)* Thiamazole 100 9/10 (90)* * p<0.01 vs control Suppression of pituitary hemorrhage Pituitary hemorrhage was suppressed by the sc treatment of following drugs prior to the ip injection of a 35% glucose solution at a high dose of 0.03 ml/g bw. In this experiment, the rate of suppression was 0 out of 55 mice in the control (0%) (Table 7). Then, pituitary hemorrhage was suppres- sed in 8 out of 17 mice (47.1%) with ether inhalation, in 16 out of 20 mice (80%) with pentobarbital, in 3 out of 14 mice TABLE 7. Suppression of pituitary hemorrhage by agents given subcutaneously prior to the intraperitoneal injection of 35% glucose at a high dose of 0.03 ml/g bw No. of mice Dose suppressed Agents (ug/g bw) /no. of mice tested (%) Control = 0/55 (0)” Phenylephrine 1 0/6 (0) Isoproterenol 1 0/6 (0) Phenoxybenzamine 20 0/6 (0) Propranolol 1 0/8 (0) Chlorpheniramine 50 0/8 (0) maleate p-Chlorophenylalanine 100 0/6 (0) p-Chlorophenylalanine 1003 days 0/6 (0) Haloperidol 5x3 days 3/14 (21.4)* Sulpiride 150 5/23 (21.7)* Ether vapor 20 min 8/17 (47.1)* Pentobarbital 50 16/20 (80.0)* Metoclopramide 30 0/13 (0) Metoclopramide 60 4/21 (19.0)* Metoclopramide 303 days 10/16 (62.5)* Ether vapor (20 min) 16/20 (80.0)* + Metoclopramide 303 days Pentobarbital _ 50 39/39 (100)”* + Metoclopramide 303 days Pentobarbital _ 50 11/11 (100)* + Metoclopramide 60 *, p<0.01 vs control. Hematocrit of 1) and 2) was 51.7+0.5 and 50.5+0.5, respectively. (21.4%) with haloperidol, in 5 out of 23 mice (21.7%) with sulpiride, in 4 out of 21 mice (19.0%) with a single dose of metoclopramide and, further, in 10 out of 16 mice (62.5%) with metoclopramide for consecutive 3 days (Table 7). Further, metoclopramide (30 ug/g bw) for 3 days in combination with a single exposure of ether vapor was highly suppressive in 16 of 20 mice (80%). Again, three consecu- tive metoclopramide (30 ug/g bw) in combination with a single injection of pentobarbital (50 ug/g bw) completely inhibited the pituitary bleeding in 39 out of 39 mice (100%). The combined pretreatment of a single high dose of metoc- lopramide (60 “g/g bw) and pentobarbital (50 ug/g bw) also resulted in the complete suppression of pituitary hemorrhage in 11 out of 11 mice (100%) (Table 7). Both a histamine receptor antagonist and a serotonin synthesis inhibitor had no effects on the bleeding. Neither stimulative nor inhibitory effects on pituitary hemorrhage were observed in agents of all adrenergic recep- tor agonists except for norepinephrine and epinephrine, and antagonists (Tables 6 and 7). DISCUSSION The present study demonstrated new data concerning experimental manipulations and pharmacological agents 104 C. Ica, I. Kosuimizu et al. which stimulated or suppressed the occurrence of pituitary hemorrhage induced by the ip injection of a 35% glucose solution in mice. Physiological study An efficacious dose of a 35% glucose solution to induce pituitary bleeding is in a narrow range of a sublethal dose, and it depends on age of animals. In our previous studies, a 9% NaCl solution was given to induce pituitary hemorrhage in mice and a sufficient dose for the induction was 0.03 ml/g bw [13]. When a 8.5% NaCl solution of the same dose was used, no pituitary bleeding had been observed in ddY mice [22, 23]. In the present study, a dose of 0.02 ml/g bw of a 35% glucose solution was ineffective to induce pituitary bleeding in young male mice. In old mice, however, a low dose (0.02 ml/g bw) of a 35% glucose solution was sufficient to produce pituitary hemorrhage. This is probably due to either age-associated attenuation in the physiological tole- rance to acute dehydration caused by the ip injection of a hypertonic solution or an absolute increase in the volume of the glucose solution injected in old mice (ca. 40-45 g). Therefore, male mice at 5-6 weeks of age and weighing about 30-35 g were used in the present experiments. Water deprivation for 3 days significantly elicited the occurrence of this pituitary hemorrhage. Gradual dehydra- tion by water deprivation and a subsequent acute dehydration by the ip injection of a hypertonic solution (a sub-threshold dose) could exert their effects on the pituitary bleeding additively in mice. Under this pretreatment of water dep- rivation, the pituitary seemed to be highly susceptible to subsequent changes in plasma osmolality. Water depri- vation-associated rise in osmolality [1], a large increase in the anterior pituitary dopamine content [6] and also an increase in dopamine receptors of human pituitary adenomas [14] are assumed to be involved in the possible mechanism of this pituitary hemorrhage. In contrast, the decline in the incidence of pituitary hemorrhage in nursing dams may relate to attenuation of inhibitory control by dopamine over prolactin secretion dur- ing nursing behavior [6,17]. Thus, it is likely that a reduced dopamine activity could arrest this pituitary bleeding in mice. Hematocrit values were not different between the control dams and nursing dams each other, so alterations in the condition of body fluid might be ruled out in this case. An acute rise in osmolality seems to be one of factors necessary to produce pituitary hemorrhage, because pituitary hemorrhage occurred after the injection of a 35% glucose solution in the present study. A gradual increase in osmolal- ity (mOsm/1) by water deprivation did not induce pituitary bleeding [1]. The ip injection of hypertonic saline is assumed to increase serum viscosity by rapid transport of water from the blood into the peritoneal cavity, resulting in high osmotic pressure of the blood [23]. Heavy congestion of hypertonic solution revealed by electron microscopy [13] might reflect the increased serum viscosity. Pharmacological study Enhancement of pituitary hemorrhage was obtained by the pretreatment of norepinephrine, epinephrine, bromocrip- tine and thiamazole in mice given an insufficient low dose (0.02 ml/g bw) of a 35% glucose solution. At present, it is hard to deduce a common action of these agents to explain a possible mechanism of this pituitary bleeding. However, vascular excitatory action of these agents may be partly involved in the pituitary hemorrhage. Epinephrine and norepinephrine are potent vasopressor drugs [10, 24], and epinephrine has long been known to accelerate blood coagulation in animals and man [9]. Both a and adrenoreceptor agonists and antagonists had neither stimula- tive nor inhibitory effects on pituitary hemorrhage, although epinephrine and norepinephrine enhanced a hypertonic solu- tion-induced-hemorrhage. Association of bundles of un- myelinated nerve fibers with blood vessels in the anterior pituitary was often observed, and these nerve fibers are assumed to be not vasomotor nerves in rats [16]. Epinephrine and norepinephrine may act at vascular systems other than the pituitary portal one. Furthermore, bromocriptine is a potent dopaminergic agonist at D, receptors [18]. Bromocriptine is well known to regress the bulk of tumors such as prolactin- and growth hormone- secreting adenomas. Pituitary apoplexy occurring in the course of chronic bromocriptine therapy has been reported in man [25], suggesting that bromocriptine suppres- sion of the growth of pituitary adenomas resulted from a necrosis of the tumor tissue followed by hemorrhage into adenomas. Although pituitary apoplexy has not been de- scribed as a complication of bromocriptine therapy in man, the pretreatment of bromocriptine has induced significant enhancement of the pituitary hemorrhage in mice. Bromoc- riptine must have acted as a dopaminergic agonist at the pituitary level. The use of thiamazole, an anti-thyroid gland drug, was based on the previous observation that thyroidectomy in rats usually resulted in apparent congestion of the anterior pitui- tary (unpublished data). These actions of thiamazole may be indirect ones, because TRH raised blood pressure in hypothyroidal animals [11]. A significant increase in portal plasma flow (140%) in hypothyroid rats rendered by pro- pylthiouracil has been demonstrated [7]. Although it is uncertain that responses of TRH secretion and of blood flow could occur immediately after the administration of thiama- zole, either direct or indirect actions of thiamazole might be involved in the mechanism of this pituitary bleeding. Pituitary apoplexy has been documented in patients given a triple bolus test of TRH, LRH and insulin [5]. Thus, various hormones and agents are known to cause incidental pituitary apoplexy in man. In the present study, catechol- amines, dopamine agonists and anti-thyroidal drugs were shown to enhance the incidence of experimental pituitary hemorrhage through an unestablished mechanism(s) in mice. On the contrary, agents that arrested pituitary hemor- rhage were haloperidol, metoclopramide, ether vapor and Pituitary Hemorrhage pentobarbital. Haloperidol, metoclopramide and sulpiride were used as dopaminergic blockers [8, 19]. There is a good contrasting relation between effects of dopamine receptor agonists as an enhancer of the pituitary hemorrhage in mice and those of dopamine receptor antagonists as a suppressor. The rates of suppression of pituitary hemorrhage with haloperidol, sulpride and metoclopramide were, however, rather low (20%), although pharmacologically over-doses were used. In this connection, the present study indicated that the combined pretreatment of dopamine antagonists and anesthetics resulted in complete suppression of the pituitary hemorrhage in some cases. Anesthetics such as pentobarbital and ether vapor alone showed a highly suppressive effect on the bleeding by 80.0% and 47.1%, respectively. In our previous experiments, neither intestinal peristalsis nor pituitary bleeding occurred although a rather excess volume of a hypertonic solution was infused into the opened peritoneal cavity of mice after laparotomy under ether anesthesia (unpublished data). Thus, inhibition of pituitary bleeding by anesthetics may be related to deterioration of the intestinal peristalsis, but at least to the reported action of brain ischemia by barbital [20, 21]. When peristalsis was enfeebled by anesthetics, the intestines could not respond normally to hypertonic solutions and the incidence of pituitary hemorrhage would be lowered. If the site of action of dopamine antagonists and anesthe- tics is different, combined treatment with dopamine antagon- ists and anesthetics would give additive or synergistic effects on pituitary hemorrhage. As speculated, combination of pentobarbital (50 “g/g bw) and metoclopramide (60 ug/g bw) resulted in complete suppression of this pituitary bleed- ing. However, EDs» should be matched between two agents of pentobarbital and metoclopramide. The present study is the first report on experimental manipulations as to stimula- tion and inhibition of the hypertonic solution-induced pitui- tary hemorrhage in mice. The exact mechanism of this bleeding remains to be disclosed. As to the organ specificity of this pituitary hemorrhage, the specific vascular architecture of the anterior pituitary may be concerned. The capacity of the venous connections draining the adenohypophysis to the cavernous sinus appeared small when compared to that of the long portal vessels supplying the adenohypophysis [3, 4]. Thus, it is supposed that venous hyperemia would easily take place in the anterior pituitary when the inflow of the blood into the gland is larger than the outflow in mice given the ip injection of a hypertonic solution. No enhancement of the pituitary bleeding was observed by the pretreatment of vasopressin, adrenocorticotropin, 5-hydroxytryptophan and 5-hydroxytryptamine, respectively, probably due to lack of interaction of these agents to the sinusoidal capillaries of the anterior pituitary in mice. 10 11 12 13 14 15 16 17 18 19 20 21 22 in Mice 105 REFERENCES Aguilera G, Lightman SI, Kiss A (1993) Regulation of the hypothalamic-pituitary-adrenal axis during water deprivation. Endocrinology 132: 241-248 Alper RH, Demarest KT, Moore KE (1980) Dehydration selectively increases dopamine synthesis in tuberohypophyseal dopaminergic neurons. Neuroendocrinology 31: 112-115 Bergland RM, Page RB (1978) Can the pituitary secrete directly to the brain? (Affirmative anatomical evidence). En- docrinology 102: 1325-1338 Bergland RM, Page RB (1979) Pituitary-brain vascular rela- tions: A new paradigm. Science 204: 18-24 Bernstein M, Hegele RA, Gentili F, Brothers M, Horgate R, Sturtridge WC, Deck J (1984) Pituitary apoplexy associated with a triple bolus test. Case report. J Neurosurg 61: 586-590 Demarest KE, Riegle GD, Moore KE (1984) Adenohypoph- ysis dopamine content during physiological changes in prolactin secretion. Endocrinology 115: 2091-2097 Eckland DJA, Lightman SL (1987) Hypothalamo-hypophyseal blood flow: A novel control mechanism in pituitary function? J Endocrinol 113: R1-2 Fielding S, Lal H (1978) Behavioral actions of neuroleptics. In “Handbook of Psychopharmacology Vol 10” Ed by LL Iver- sin, SD Iversien, SH Snyder, Plenum Press, New York, pp 91- 128 Forwell GD, Ingram GIC (1957) The effect of adrenaline infusion on human blood coagulation. J Physiol Lond 135: 371-383 Goldengerg M, Aranow H Jr, Smith AA, Faber M (1950) Pheochromocytoma and essential hypertensive vascular disease. Arch Intern Med 86: 823-836 Horita A, Carino MA, Lai H (1987) Pharmacology of thyro- tropin-releasing hormone. Annu Rev Pharmacol Toxicol 26: 311-332 Kobayashi Y, Iga C (1989) Erythrocyte diapedesis in anterior pituitary hemorrhage after intraperitoneal injection of hyper- tonic solution in mice. Zool Sci 6: 359-365 Kobayashi Y, Masuda A, Kumazawa T (1982) Hypertonic solutions induce hemorrhage in the anterior pituitary in mice. Endocrinol Japon 29: 647-652 Koga M, Nakano H, Arai M, Sato B, Noma M, Morimoto Y, Kishimoto S, Mori S, Uozumi T (1987) Demonstration of specific dopamine receptors on human pituitary adenomas. Acta Endocrinol 114: 595-602 Koshimizu I, Awamura N, Takeuchi S, Kobayashi Y (1992) Structural changes and morphometrical analysis of the pituitary gland after hemorrhage induced by intraperitoneal injection of hypertonic solution in mice. Biomed Res 13: 253-258 Kurosumi K, Kobayashi Y (1980) Nerve fibers and terminals in the rat anterior pituitary gland as revealed by electron micro- scopy. Arch Histol Jpn 43: 141-155 Leong DA, Frawley LS, Neill JD (1983) Neuroendocrine control of prolactin secretion. Annu Rev Physiol 45: 109-127 Markstein R (1981) Neurochemical effects of some ergot de- rivatives: A basis for their antiparkinson actions. J Neural Trans 51: 49-59 McCallum RW, Albibi R (1983) Metoclopramide: Pharmacol- ogy and clinical application. Ann Intern Med 98: 86-95 Shapiro HM (1985) Barbiturates in brain ischemia. Anaesth 57: 82-95 Steer CR (1982) Barbiturate therapy in the management of cerebral ischemia. Dev Med Child Neurol 24: 219-231 Takeshita M, Doi k, Mitsuoka T (1988) Brain lesions induced by hypertonic saline in mice: Dose and injection route and Bie 106 C. Ica, I. Kosuimizu et al. incidence of lesions. Exp Anim 37: 191-194 limb blood vessels. Br Med Bull 19: 125-131 23 Takeshita M, Doi K, Imaizumi Mitsuoka T (1989) Initial 25 Yamaji T, Ishibashi M, Kosaka K, Fukushima T, Hori T, Lesions in the mouse brain induced by intraperitoneal injection Manaka S, Sano K (1981) Pituitary apoplexy in acromegaly of hypertonic saline. Exp Anim 38: 31-39 during bromocriptine therapy. Acta Endocrinol 98: 171-177 24 Whelan RF, De La Lande IS (1936) Action of adrenalin on ZOOLOGICAL SCIENCE 11: 107-111 (1994) © 1994 Zoological Society of Japan Ecdysteroid Synthesis in Dissociated Cells of the Prothoracic Gland of the Silkworm, Bombyx mori Masako Asauina!, HayiMe Fuco!* and SatosHt TAKEDA2 "Department of Environmental Science and Resources, Faculty of Agriculture, Tokyo University of Agriculture and Technology, Fuchu-shi, Tokyo 183 and *National Institute of Sericultural and Entomological Science, Tsukuba, Ibaraki 305, Japan ABSTRACT—A method for the preparation of viable cells of prothoracic glands of Bombyx mori was developed. Dissected prothoracic glands (PGs) from larvae at the spinning stage were incubated in the Grace’s insect culture medium containing dispase (6,000 PU/ml of medium at pH 6.5) at 37°C for 15min. After centrifugation for 2 min at 250 rpm, resultant cells were incubated with Grace’s insect medium for designated times. After this treatment 87+7% of the cells remained viable as judged by exclusion of trypan blue. The amounts of ecdysteroid secreted into medium by these cells were approximately 70 to 80% of the amounts secreted by intact PGs. The rates of ecdysteroid secretion over 6 hr'in organ culture and in cell culture were quite similar, although the amounts of ecdysteroid detected in the latter medium were somewhat lower. INTRODUCTION The synthesis of radioactive ecdysone made possible the development of the radioimmunoassay (RIA) which provided an extremely sensitive, simple and cheap method of detecting ecdysteroid [2, 5, 17]. By using these techinques, it is now well established that ecdysone is the product of the PGs and that its production and release is controlled by the prothoraci- cotropic hormone (PTTH) [1, 3, 8, 13, 18]. In Bombyx the studies on the mode of action of PITH on the PGs and biochemical processes on the regulation of ecdysteroid synthesis by PITH are quite few, although identification and isolation of PTTH have been developed [6, 7, 9]. In the present study, we demonstrate an improved method for preparing the dispersed cells of PG for the investigation of mode of action of PITH. MATERIALS AND METHODS Insects Larvae of the silkworm, Bombyx mori (J122XxC115), were reared with mulberry leaves or artificial diet (Nihon Nosan Kogyo, Yokohama) in a rearing room of our laboratory at 25+1°C in a 16 hr light: 8 hr dark photoperiod. Larvae were staged on the day of 4th ecdysis, and this day was designed as Day 0 of 5th instar. Only female animals were used throughout the experiments. Preparation of larval prothoracic gland After immersing the larvae into 75% of ethanol for 2 to 3 min, the prothoracic glands (PGs) of the staged larvae were removed. The PGs were rinsed with physiological saline (0.85% NaCl) for Bombyx and then with Grace’s medium (GIBCO, USA) two to three Accepted December 20, 1993 Received October 7, 1993 * To whom all correspondence should be addressed. times to avoid contamination by haemolymph. Thereafter, PGs were incubated in Grace’s medium or were treated with a medium containing dispase as described below. Preparation of incubation medium containing dispase To dissociate the cells of the PG, we used a proteolytic enzyme, dispase, (Godo Shusei, Tokyo). The optimal pH for this enzyme is between 7.5 and 8.0 and the optimal temperature is 25 to37°C. The pH of the silkworm haemolymph is about 6.5. Accordingly, the pH of the medium was adjusted with 1 N NaOH to either pH 6.5 or pH 8.0 before use. The dispase was then dissolved in this medium. Activity of this enzyme was represented as protease units per ml of medium. Incubation temperature was 25 or 37°C. Incubation of prothoracic gland or its cells PGs were dissected from the staged animals and incubated immediately. A pair of glands was incubated in 300 pl of Grace’s medium at 25°C. Thirty microliter aliquots of the medium were taken for RIA at various times and were replaced by 30 yl of fresh medium. In the case of dissociated cells, larval PGs were incubated in various concentrations of dispase in Grace’s medium for 15-60 min at either 25 or 37°C. After several rinses with Grace’s medium, one pair glands equivalent of dissociated cells was incubated in 300 sl of medium. During the incubation, the reaction mixture was shaken gently with a micromixer (TAITEC: EM 33, Japan) and cells were dissociated by drawing the tissue in and out of a siliconized Pasteur pipette about 80 times at 15 min intervals for up to 1 hr. Viability of the dissociated cells of prothoracic gland After dispase treatment, the viability of the dissociated cells was checked using 0.1% trypan blue. Viable cells excluded this dye, whereas dead cells became blue. Estimation of ecdysteroids in haemolymph and in culture medium Haemolymph (10 wl) or medium (30 1) was subjected to an ecdysteroid radioimmunoassay (RIA) [15]. Ecdysteroids were ex- tracted with 300 41 of absolute methanol, then aliquots of the 108 M. AsaHiIna, H. FuGo AND S. TAKEDA supernatant were assayed for ecdysteroid by RIA. Since 20- hydroxyecdysone was used as a standard, the average values of the amount of ecdysteroids are expressed as ng of 20-hydroxy-ecdysone equivalents + standard deviation. The tritiated ligand[23, 24-*H(N)] ecdysone (82.8 Ci/mmol) was purchased from New England Nuc- lear. RESULTS Dissociation of prothoracic gland cells by a proteolytic en- zyme, dispase For studies of ecdysteroid synthesis in dissociated cells of the PG in vitro, the following prerequisites must be fulfilled by the experimental procedure: low variability among repli- cate samples, rapid and easy methods for handling of large numbers of cells, homogeneity among the dispersed cells, and maintenance of viability for at least 6 to 8hr during the incubation. The PG cells in Bombyx are compact and surrounded by a thick basal lamina (Fig. 1 [16]). Therefore, we tried various enzymes to obtain viable preparations of single PG cells. Preliminary incubations with either trypsin or collage- nase were not satisfactory since most cells were not viable (data not shown). By contrast, with dispase treatment, PG cells partially dissociated and remained viable. In order to obtain viable cells of PG consistently, we sought to establish procedures for preparing single cells of PG. Firstly, PGs from day 9, Sth instar Bombyx larvae, were incubated for various times in Grace’s medium containing several concentrations of dispase at different pHs and temperatures. Five pairs of PG were removed and rinsed first with physiological saline for Bombyx, then with Grace’s medium. These organs were incubated in a disposable plas- tic dish (Corning, 35 x 10 mm) with 2 ml of dispase solution (0 to 1,500 PU/ml Grace’s medium). At 25°C the cells of PG showed little dissociation even at the highest enzyme concentration used (1,500 PU/ml for 60 min) irrespective of pH (Table 1). Under these conditions, the cells were dispersed gradually with time and concentra- tion of enzyme but some clusters of cells remained, and the dissociated cells were significantly damaged. Accordingly, these conditions were not sufficient for preparation of viable cells from PG. On the other hand, when the PGs were incubated with dispase at 37°C, quantities of dispersed cells could be obtained both at pH 6.5 (Fig. 2) and pH 8.0 (Table 1). However some precipitates of cellular debris were observed at higher enzyme concentrations at pH 8.0. Thus, it seemed that the treatment with dispase at pH 8.0 was unsatisfactory for preparing viable cells of PG. Secondly, the effects of dispase concentration on the dissociation of PG cells were investigated. A pair of PG from a larva in the spinning stage (day-9) was treated with various concentration of dispase in Grace’s medium (pH 6.5) at 37°C. As shown in Figure 3, there was no correlation between the enzyme concentration and the percentage of viable cells. The yield of viable cells was 86.8+7.2% for all concentrations from 1.5 to 10 10° PU/ml. Lastly, we separated a pair of PGs and treated each individual PG with dispase (pH 6.5, 6000 PU/ml, 15 min, 37°C) to determine the cell number in the PG. There was no difference between the cell number in the night and the left PG throughout development in the Sth instar with the aver- age number of cells between 160 to 230 in both right and left PG (data not shown). Fic. 1. Prothoracic gland in Bombyx mori. Bar, 100 «m. Viable cells (%) Dissociation of Bombyx PG to Viable Cells 109 TaBLE 1. Dissociation of prothoracic glands by dispase under different conditions Concentration of Incubation time (min) Incubation time (min) dispase (PU/ml) 15 30 45 60 15 30 45 60 [pH 6.5 at 25°C] [pH 6.5 at 37°C] 0 bs -s pe = 500 = = 36 aE ae + ++ + 1,000 + + + ++ + + ++ eteate 1,500 a + +f sar qPap SPARSE SPSPSE AR SRAP [pH 8.0 at 25°C] [pH 8.0 at 37°C] 0 ND ND ND ND = = = = 500 ND ND ND ND ND ND + + 1,000 ND ND ND ND + + dL tt aL 1,500 ND ND + 4p 4p tees pe tee The grades of the dissociation of prothoracic glands were as follows; —: not dissociated, +: basal lamina was removed but cells were not dissociated. +: some clusters of prothoracic gland cells seen, + +: partial dissociation of prothoracic glands with a mixture of single cells and groups of 2 to 5 cells, +++: complete dissociation of the prothoracic gland yielding dispersed cells as shown in Fig. 2. -ND: not determined. Fic. 2. Dissociated cells of prothoracic glands. The prothoracic glands were treated with 1,500 PU of dispase/ml Grace’s medium (pH 6.5) at 37°C for 30 min. Bar, 350 pm. Secretory activity of ecdysteroid by intact PGs and dissociated PG cells Intact PGs from larvae in the spinning stage (day-8 to day-10 of Sth instar) were incubated in Grace’s medium for 6 hr to determine their secretory ability in comparison to those of dispersed cells. As shown in Figure 4, the secretory activity of PGs on day 8 (1 day after the onset of spinning) was relatively low (10.9+1.8 ng per 6hr). By day 9, secre- tory activity increased about 3-fold to 28.6+4.1 ng per 6 hr Fic. 3. Relationship between the concentration of dispase in Grace’s medium (pH 6.5) and the percentage of viable cells as Dispase concentration (x 10°PU/ml) judged by exclusion of the trypan blue dye. 15 2 3 4 5 6 7 8 9 10 110 M. ASAHINA, H. FuGo AND S. TAKEDA Day-8 —@®— Organs —Oo-— cells Ecdysteroids (ng/glands) pre- O 2 4 6 pre- O incu- incu- bation Incubation time (hr) bation —®-— organs —O— cells Incubation time (hr) —@®-— organs —O— cells 0 4 6 pre- O 2 4 6 bation Incubation time(hr) Fic. 4. Ecdysteroid synthesis following incubation of the dissociated cells and intact prothoracic glands of Bombyx mori on different days of the final instar. The cells of the prothoracic gland were dispersed with dispase (6,000 PU/ml of Grace’s medium, pH 6.5) at 37°C for 15 min. The enzyme was removed by centrifugation at 250 rpm for 2 min. The amount of ecdysteroids in this enzyme solution was estimated by RIA, and was plotted as “pre-incubation” on each abscissa. The resultant cells of prothoracic glands were immediately rinsed with 300 41 Grace’s medium. After centrifugation (250 rpm for 2 min), the cells were resuspended in 300 1 of medium at time “0” of incubation. A pair of intact prothoracic glands or a pair of one gland’s equivalent of dissociated cells was used. The stage of prothoracic gland is indicated in the upper right side of each figure. incubation. Ecdysteroid produced by the glands decreased significantly by day 10 (pharate pupal stage). To measure the amount of ecdysteroid produced by the dispersed PG cells, we used a centrifugation tube with filter (Ultrafree, C3SV, 5 um Millipore Co. Ltd) to reduce the handling time necessary for the dissociation of PG. Cells from PGs of day 8 to day 10 larvae secreted ecdysteroids into the medium (Fig. 4). However, the amounts of ecdysteroids were about 30 to 35% lower than those produced by the intact PGs. The rate of ecdysteroid production in the dissociated cells was similar to that of intact PG (Fig. 4). Since the mortality of cells by the dispase treatment as described above ranged between about 14 to 25%, the reduced amounts of ecdysteroid in the dissociated cells may be due to the loss of viable cells during the treatment of the glands. DISCUSSION The PG cells in Bombyx are surrounded by a thick basal lamina (Fig. 1), while the cells in Hyalophora cecropia are connected only by a thin strand [4] and the morphology of the glands of Manduca sexta is intermediate between these two extremes [3, 11]. Manduca PG cells have been successfully dissociated by either 0.4% trypsin/chymotrypsin/elastase [14] or 0.4% elastase [12]. In these experiments, the yield of viable cells was 295% and the cells remained viable for at least 4hr [12, 14]. In the present experiment, a technique was developed whereby the intact PG of Bombyx could be dispersed into viable cells by using dispase in Grace’s The yield of viable cells ranged between 80 to 94% and these cells remained viable at least 6 hr. To minimize the handling of the dissociated cells, we medium. The day of 4th ecdysis was designated as “day 0” of Sth instar. used an Ultrafree centrifugation tube. Thus, after incuba- tion the supernatant could be readily removed and assayed for ecdysteroids. With this technique we showed that the dissociated cells were similar to intact glands in their linear production of ecdysteroids over a 2—6hr period in vitro. Previous studies by Okuda ef al. [10] have shown that Bombyx PGs show different rates of ecdysteroids synthesis around the time of gut purge and spinning. Our studies show a similar increase on day 9 for both the intact glands and the dissociated cells. Thus, although the dissociated cells produced less ecdysteroid than the intact gland, their rate of production is similar to that of the intact gland at a particular time. Dissociated PG cells from Manduca lacking the basal lamina respond to PTTH in vitro by an increase in ecdysteroid production [12, 14]. Now that PTTH in Bombyx has been purified and sequenced [6, 7, 9], studies of its mode of action on the PG are needed. Our dispersed PG cell preparation should be ideal for such a study. ACKNOWLEDGMENTS We would like to thank Prof. L. M. Riddiford, University of Washington, for her critical reading and valuable comments on this manuscript. We also thank Mr. Skarlatos Dedos for his help on a preparing the draft of this manuscript. REFERENCES 1 Bollenbacher WE, Granger NA (1985) Endocrinology of the prothoracicotropic hormone. In “Comprehensive Insect Phy- siology, Biochemistry and Pharmacology Vol7” Ed by GA Kerkut, LI Gilbert, Pergamon Press, Oxford, pp 109-152 10 11 Dissociation of Bombyx PG to Viable Cells 111 Borst DW, O’Connor JD (1972) Arthopod molting hormone: radioimmune assay. Science 178: 418-419 Gilbert LI, Combest WL, Smith WA, Meller VH, Rountree DB (1988) Neuropeptides, second messengers and insect molting. Bio Essays 8: 153-157 Herman WS, Gilbert LI (1966) The neuroendocrine system of Hyalophora cecropia (L) (Lepidoptera: Saturniidae), I: anatomy and histology of the ecdysial glands. Gen Comp Endocrinol 7: 275-291 Horn DHS (1989) Historical Introduction. In “Ecdysone” Ed by J Koolman, Thieme Verlag, Stuttgart, pp 8-19 Kataoka H, Nagasawa H, Isogai A, Tamura S, Mizoguchi A, Fujiwara Y, Suzuki C, Ishizaki H, Suzuki A (1987) Isolation and partial characterization of a prothoracicotropic hormone of the silkworm, Bombyx mori. Agric Biol Chem 51: 1067-1076 Kawakami A, Kataoka H, Oka T, Mizoguchi A, Kawakami KM, Adachi T, Iwami M, Nagasawa H, Suzuki A, Ishizaki H (1990) Molecular cloning of the Bombyx mori prothoracicotro- pic hormone. Science 247: 1333-1335 Koolman J (1989) Ecdysone. Thieme Verlag, Stuttgart, pp 482 Matsuo N, Aizono Y, Funatsu G, Funatsu M, Kobayashi M (1985) Purification and some propeties of prothoracicotropic hormone in the silkworm, Bombyx mori. Insect Biochem 15: 189-195 Okuda M, Sakurai S, Ohtaki T (1985) Activity of the prothor- acic gland and its sensitivity to prothoracicotropic hormone in the penultimate and last larval instar of Bombyx mori. J Insect Physiol 31: 455-461 Sakurai S, Warren JT, Gilbert LI (1989) Mediation of ecdy- 12 13 14 15 16 17 18 sone synthesis in Manduca sexta by a hemolymph enzyme. Arch Insect Biochem Physiol 10: 179-197 Sakurai S, Gilbert LI (1990) Biosynthesis and Secretion of Ecdysteroids by the Prothoracic glands. In “Molting and Meta- morphosis” Ed by E Ohnishi, H Ishizaki, Japan Sci Soc Press, Tokyo, Springer-Verlag, Berlin, pp 83-106 Smith SL (1985) Regulation of ecdysteroid titer: synthesis. In “Comprehensive Insect Physiology, Biochemistry and Pharma- cology Vol 7” Ed by GA Kerkut, LI Gilbert, Pergamon Press, Oxford, pp 295-341 Smith, WA, Rountree DB, Bollenbacher WE, Gilbert LI (1986) Dissociation of the prothoracic glands of Manduca sexta into hormone-responsive cells. In “Insect Neurochemistry and Neurophysiology” Ed by AB Borkovec, DB Gelman, pp 319- 322 Takeda S, Kiuchi M, Ueda S (1986) Preparation of anti-20- hydroxyecdysone serum and its application for radioimmunoas- say of ecdysteroids in silkworm hemolymph. Bull Seric Sci Stn 30: 361-374 Toyama K (1902) Contributions to the study of silk-worm, I. on the embryology of the silkworm. Bull Coll Agric Tokyo Imp Univ 5: 73-117 Warren JT, Gilbert LI (1988) Ecdysteroids. In “Immunolo- gical Techniques in Insect Biology” Ed by LI Gilbert, TA Miller, Springer-Verlag, New York, pp 181-214 Watson RD, Spaziani E, Bollenbacher WE (1989) Regulation of ecdysone biosynthesis in insects and crustaceans: a comarison. In “Ecdysone” Ed by J Koolman, Thieme Verlag, Stuttgart, pp 188-210 gait ctepereytovnrsn a oped ener e ~ ynavlar pe aba er em ena | hid 01 ee na jpoa! [Pua ee oat ‘| treeeahT = pOt) 0. tei e ‘peVadb2eeRd aa ~ miiLos wulit4’ of ohxele oapeedlient wii Hout bat Sia dicewhel Ft cehiomG? SE yi et fecacwlq itary aq athe got: Sede fe: wyeaul go) a dune ry i“ \ pz Wa peat ‘ fe ‘5 os Ter Nuh vealelievdd reral ae Uli ~qmnG ye 4 hs: WO OT sure A AD vd bo Tv ane - | Lat UT w7 teeth } f } WV toyfoetnslliodt , He Titty h wv “iting” bh d ss . sath rei sory t Gry ita ie AN Reva : WF toaen! [so (Sviagoqegy-ncetoe seen me ! f tel “vee oie deporteyt ity SQGlirrOd , j oy J stol-Taur is ic Lied } j\ Hawi 4 iiead tle a Homie, Ei z e loqe 24h brs mira anoryh amas - : HM Aeximal A. sea vi : 1 peas tienvwdltted chi Vesey Ge icy ed} seit «ase Wh bet noe the re , PR ; ; , y bin Ltr 4A pale otk As ait 2 i uso Wp J ry IS Ay epee AT p 8 ee : i if mr Ai Oe - u HHOamA + 0G. ere Vey wqly? fag eo 7o poaptri gs iat i iafiborieGA Riz bhi denied Ve ertael a ope Spnoiticlayt tp hd pode Atiemetl: hdv o Aaa ay nee ie oe sere dance : CAG. “= tn yell apa gd ened plainly a alee nk é pe miata tapane: Avyarts ype \ > ‘ aM 4A, oniee vi Mist ie vit: ne wet ant fs veanhicee, a S earl 7 oli ol engages CSE 04 1A dey path at be Bip da: cael j Sry aon6 1 Oesalenuie nevboe Wo ob scrim siyvorpaksspodtery Bl a « ' Fain ilo aA i “sucevtod” wf veal | tvisets thE Wr ser A (iA ~ ie “et oe. “ena =a Wi tis r o6 veciie be 4 te Py iyi wen Creed | ke . ; rt: Cae To : TP Dov 6 ‘or Gt 2s, paar a J ats ie (Gers / mete : aan th i t. : eceatewnered ot Miele a : o ha re hs ie Tee ; ¥ ; j f Fed i, z } UN > 7 4 ee = t= vi *) bc LOry ; et m , ACKNOWL - A t E Pea, t= wy 4 : . y wt ad act advice id ae | # Oe ae Wi ipager A, (UNES ‘Lin * ee &yovw - A “ae i res [ Vem Maree meng * Nolet : Lie ao vemos Pyows Oivtand, j ‘i 7 a, wey aay ZOOLOGICAL SCIENCE 11: 113-119 (1994) Effects of Light and Food as Zeitgebers on Locomotor Activity Rhythms in the Loach, Misgurnus anguillicaudatus Mayumi Naruse!* and Tapasui Otsu’ ‘Division of Human Life and Environmental Science, Graduate School of Human Culture, and *Department of Biology, Faculty of Science, Nara Women’s University, Nara 630, Japan ABSTRACT—Four experiments were performed to examine the synchronizing effects of light and food on the locomotor activity rhythm in the loach ( Misgurnus anguillicaudatus). In Exp. 1, food was given once a day at the scheduled time in the middle of light (L) or dark (D) phase under LD 12:12. When the fish were fed in the L phase, they showed three types of activity pattern (light-active, dark-active and light-dark-active). On the other hand, most of the fish became dark-active when they were fed inthe D phase. In both conditions, the fish were entrained well to the scheduled feeding. In Exp. 2, the scheduled feeding cycle was removed from the previous Exp. 1. Under these conditions, the loach remained to show the same pattern as that in Exp. 1 or tended to become dark-active. The feeding-anticipatory activity peak gradually disappeared during these experiments. In Exp. 3, the fish were exposed to the scheduled feeding and constant darkness (DD). Since almost all fish were entrained to the scheduled feeding, scheduled feeding as well as LD cycles is effective as a zeitgeber for the locomotor activity rhythm in the loach. Under constant conditions (Fig. 4), free-running rhythms were observed, although the duration of the rhythm was short and the ratio of the individuals that showed free-running rhythms was low (7-50%). Therefore, the locomotor activity rhythm in the loach is an endogenous rhythm but the coupling between the oscillator and the locomotor activity seems to be weak and different depending on © 1994 Zoological Society of Japan individuals. INTRODUCTION One of the characteristics of the circadian rhythms in fishes is variability of the rhythms compared with those in higher vertebrates [15]. The appearance of the circadian rhythms varies inter- [24] and intra-specifically [5, 7, 27], and even intra-individually [12]. Activity patterns showed diur- nal, nocturnal, crepuscular and intermediate types depending on fishes [24]. In the perch (Perca fluviatilis), juvenile fish showed a nocturnal activity pattern but adult fish changed the activity pattern to diurnal [7]. Several activity patterns appeared simultaneously such as diurnal and nocturnal in the juvenile pink salmon (Oncorhynchus gorbuscha) [5] and diurnal, nocturnal, light-change-active and arrhythmic in the medaka (Oryzias latipes) [27|. Locomotor activity changes seasonally in the minnow (Phoxinus phoxinus), the sculpin (Cottus poecilopus) and the burbot (Lota lota) [12], and the medaka [31]. These reports suggest that a plastic reactivity to various zeitgebers may induce the variability of the rhythm. In order to clarify this probability, we selected two environmental factors, light and food. Light is a major environmental factor as a zeitgeber in fishes as well as in other organisms. Food has been reported to act as a zeitgeber in some fishes, such as the goldfish (Carassius auratus) [14] and the medaka [29]. Accepted October 15, 1993 Received July 27, 1993 * Present address: Department of Hygiene, Akita University School of Medicine, Akita 010, Japan 2 To whom reprint requests should be addressed. The loach (Misgurnus anguillicaudatus) is one of the common freshwater fishes in Japan and inhabits paddy fields and small streams. Although there are several studies con- cerning the spawning behavior and season in the loach [19, 23], there is only one report by Yanagishima and Mori [30] on the locomotor activity rhythm of the loach. They reported that the fish showed the exogenously controlled nocturnal activity rhythm because the rhythmicity immediately dis- appeared under a constant condition. In this paper, we investigated (1) effects of single environmental factor, light (LD cycle) or food (scheduled feeding cycle), on the locomo- tor activity rhythm, (2) effects of the phase relationship between two environmental factors, light and food, on the rhythm, and (3) whether the locomotor activity rhythm of the loach is an endogenous circadian rhythm or not. MATERIALS AND METHODS Adult loaches (Misgurnus anguillicaudatus) including males and females of about 9-16 cm in total length were used. We caught them at a small stream along the paddy fields in Kyoto Prefecture in May, 1988, or obtained cultured fish in the Shikoku districts from a fish shop in April to October, 1987. The fish were kept individually in a plastic water tank (31.517 21H cm) with sand at the bottom and placed in a bioclimatic chamber at 25+1°C at least one month before experiments. Fluorescent lamps (40 W) were used as the light source. Light intensities at the water surface were measured by a radiometer (UDT161, United Detector Technology Inc., California) and ad- justed at 460-600 lux (0.26-0.39 mW/cm7) or 5 lux (0.002 mW/cm?) in the light (L) phase and 0 lux in the dark (D) phase. We fed about 0.5-1.0 g wet weight of live tubifexes as food by hand or an auto 114 M. NARUSE AND T. OISHI feeding instrument (SP-10A, Nippon Denshi Kagaku, Kyoto) during scheduled feeding experiments. Tubifexes with water were put in a small plastic tube and thrown into a water tank by overturning this tube at the feeding time. As a control, we gave only water using the tube for 3 days just after the scheduled feeding experiment, but none of the fish reacted to this sham-feeding regimen. Water was always aerated and filtered. The locomotor activity of the loach was measured by a pair of infrared photocells (JU-33P, -33R or PL3-E, -FL, Hokuyo Denki, Osaka) at both sides of the water tank. The light source and receiver of photocells were set at about 1 cm above the sand, since the loach is a benthic fish. Main activity including searching and feeding behavior and most swimming behavior was observed at the bottom layer (see the results of Exp. 3). When the fish interrupted the beam, it was recorded by an actograph (Fuji Denki, Tokyo) and the number of interruptions were counted by a digital data recorder (DDR3010, Sanyo Denki, Osaka). Data from the actograph were used for visual examination, and total counts per one hour from the digital data recorder were used for the statistical analyses such as ¥-test and periodogram analysis. In these analy- ses, the significance level we adopted was 95% confidence limit. In the present study, we selected four experimental conditions to examine the synchronizing effects of light-dark (LD) cycles and scheduled feeding in the locomotor activity rhythm of the loach and compared the results individually by using the same members of fish throughout these experiments. Before each experiment, loaches were subjected to constant darkness (DD) without food for at least 10 days to eliminate the after-effect of the previous condition. Experiment I Locomotor activity rhythms under LD cycles and scheduled feeding. (A) Food was given at the scheduled time (12:00) in the middle of the L phase under LD 12:12 (L: 06: 00-18 : 00, D: 18 : 00-06 :00; L= 460-600 lux, D=0 lux) for 10 (n=6) or 15 (n=14) days. (B) Food was given at the scheduled time (12:00) in the middle of the D phase under reversed LD 12:12 (L: 18: 00-06 :00, D: 06: 00- 18:00) for 12 (n=6) or 15 (n=14) days. Experiment 2 Effects of LD cycles on the locomotor activity rhythms without food after Exp. 1. (A) Fish (n=6) were placed under LD 12:12 without food after Exp. 1-A for 10 days. (B) Fish (n=6) were placed under LD 12:12 without food after Exp. 1-B for 10 days. Experiment 3 Effects of scheduled feeding on the locomotor activity rhythms under constant darkness (DD). Fish (n=6) were fed once a day at the scheduled time (12:00) under DD (D=0 lux) for 11 days. In order to analyze the behavior of the loach under the scheduled feeding and constant dim light (dim LL, L =5 lux), we recorded the behavior of a female by VTR (TV camera; WV-1550, Matsushita Tsushin, Osaka, Time Lapse VTR; NV-720, Matsushita Denki, Osaka). Experiment 4 Free-running rhythms under constant conditions after Exp. 1 and 3. (A) Fourteen fish were subjected to DD without food after Exp. 1-A for 15 days. (B) Fourteen fish were placed under DD without food after Exp. 1-B for 15 days. (C) Six fish were placed under DD without food after Exp. 3 for 14 days. All experiments were performed during November, 1987 to September, 1988. RESULTS Experiment 1 Locomotor activity rhythms under LD cycles and scheduled feeding. (A) Food was given at the scheduled time (12:00) in the middle of the L phase under LD 12:12. The locomotor activity of the loach was classified into four patterns in relation to LD cycle, i.e., dark-active (nocturnal) (ex. Fig. 1A-a, B), light-dark-active (ex. Fig. 1A-b, c and Fig. 2a), light-active (diurnal) (ex. Fig. 1A-d) and arrhythmic. The term “light-dark-active” denotes that there are peaks of activity in both L and D phases. Among 20 fish used in this experiment, one fish was dark-active, 12 were light-dark- active and seven were light-active, and the difference was statistically significant (P<0.05, y?-test for one sample) (Table la). It took at least 3-5 days for the loaches to establish the stable locomotor activity pattern (see Fig. 2a). Four out of six light-active fish tended to change their activity patterns from dark- to light-active during this experiment. All fish excluding one dark-active fish (Fig. 1A-a) were entrained to scheduled feeding (P<0.01, y-test for one sample), and the activity was classified into two types as follows. Type 1 (E;): the peak of activity lasted several hours before the feeding time, which probably reflects the anticipation for food (n=11), and among these 11 fish, eight remained inactive for several hours after feeding (ex. Fig. 1A-b) and three continued to be active for several hours after feeding (ex. Fig. 1A-d). Type 2 (E,): the activity peak lasted several hours after the feeding time without the anticipatory peak (n=8) (ex. Fig. 1A-c). When Exp. 1-A regimen was repeated after Exp. 2-A using the same indi- viduals, a similar tendency of activity patterns was observed for both LD cycle and scheduled feeding. (B) Food was given at the scheduled time (12:00) in the middle of the D phase under LD 12:12. In this experiment, all fish except one light-dark-active fish showed a dark-active pattern (ex. Fig. 1B and Fig. 2b) (P< 0.01, x?-test for one sample) (Table 1a). A tendency to- ward the change of activity pattern from light- to dark-active was observed in four dark-active fish during this experiment. All of the 19 -dark-active fish were synchronous to the scheduled feeding (P<0.01, y-test for one sample) (Table la). The two types of entrainment to scheduled feeding as the previous experiment were also observed in this experi- ment. The number of fish for type 1 and 2 was 18 (ex. Fig. 1B) and 1, respectively. When Exp. 1-A was compared with 1-B, various activity patterns appeared under the scheduled feeding in the L phase, while the loach became exclusively dark-active under the scheduled feeding in the D phase, and the difference was highly significant (P<0.01, X?-test for multiple samples) (Table 1a). When the activity patterns were compared indi- vidually between these two conditions, three types were detected in the entrainment to LD cycles; (1) fish strongly affected by the scheduled feeding, as they tended to be light-active when they were fed in the L phase and dark-active Locomotor Activity Rhythms in the Loach Aa 200 200 100 D-N 100 0 0 O biG oi Mein ste mdea TOL Te mh) 7) d a a thorig 300 = LD - E2 200 z 100 100 = 0 0 5 OG 9312 “By 2 OPMcRNNDNte 2a < B r 600 EEE SOOM AAVVOQVKAMIM: A 400 D-E1 200 0 0 SOAR TIME OF DAY (HR) Fic. 1. Examples of locomotor activity rhythms of the loach in Experiment 1. Exp. 1-B (LD 12:12, Food in D). LD cycles, and E;, E> and N for the entrainment to scheduled feeding. D: dark-active. light-active. E,: entrained with the anticipatory activity peak (type 1, see text). peak (type 2, see text). N: not entrained. Horizontal bar: light condition (LD 12:12, white Triangle: feeding time. Data were shown as mean+standard error (SE) for 10 or 12 days. (A) Exp. 1-A (LD 12:12, Food in L). Activity patterns were represented as D, LD and L for the entrainment in relation to LD: light-dark-active. E,: entrained without the anticipatory 115 LD-E1 (B) L: bar: L, dotted bar: D). TABLE 1. The number of individual fish for each activity pattern (a) LD cycle Scheduled feeding Expemmems Dark-active Eight dark: Light-active Arrhythmic Entrained Nore 1-A (LD 12:12, Food in L) 1 12 7 0 * 19 1 + 1-B (LD 12:12, Food in D) 19 1 0 0 sett | 19 1 = 2-A (LD 12:12, Food removed in L) 4 0 2 0 = — — 2-B (LD 12:12, Food removed in D) 6 0 0 0 — — 3 (DD, Food at 12:00) — — — — 5 1 (b) Constant condition Experiments Free-run Not free-run 4-A (after LD 12:12, Food in L) 1 13 4-B (after LD 12:12, Food in D) 4 10 4-C (after DD, Food at 12:00) 3 3 *: P<0.05, X?-test for one sample **: P<0.01, x7-test for one sample : P<0.01, 7?-test for multiple samples 116 M. NARUSE AND T. OIsHI when they were fed in the D phase (n=7), (2) fish affected only by LD cycles, as they were always dark-active irrelevant to the phase of feeding (n=1), and (3) the intermediate type between types (1) and (2) (n=12). In the case of entrain- ment to scheduled feeding, there were also three types, (1) fish entrained with an anticipatory activity peak (type 1) irrelevant to the phase of feeding (n=11), (2) fish changed their activity patterns to type 1 when they were fed in the D phase (n=7), and (3) fish did not show the anticipatory activity peak (n=2). Experiment 2 Effects of LD cycles on the locomotor activity rhythms without food after Exp. 1. (A) Fish were placed under LD 12:12 without food after Exp. 1-A. In six fish observed, four were dark-active (ex. Fig. 2a) and two were light-active (Table 1a). Under LD 12:12 and the scheduled feeding at the L phase in Exp. 1-A, all of the four dark-active fish showed light-dark-active pattern, while two light-active fish were light-dark- or light-active. A signi- ficant reduction in the amount of activity was observed in the two light-active fish (ex. Fig. 2a) and one dark-active fish in a few days after the beginning of this experiment probably because of starvation. The anticipatory activity peak observed before the feeding time disappeared within a few days. (B) Fish were placed under LD 12:12 without food after Exp. 1-B. In this experiment, all six fish were dark-active (ex. Fig. 2b) (Table 1a). They were dark-active in the previous Exp. 1-B. Three of them decreased the amount of activity in a few days. When the results of Exp. 2 was compared with that of Exp. 1, in five fish that were light-dark-active under the scheduled feeding in the L phase (Exp. 1-A), four fish changed their activity patterns to dark-active, and only one fish changed to light-active. One fish that were light-active in Exp. 1-A kept the same pattern. The difference between Exp. 1-A and 2-A was highly significant (P<0.01, 7-test for multiple samples) (Table 1a). On the other hand, the dark- active pattern in all of the six fish under the scheduled feeding in the D phase (Exp. 1-B) was maintained in the condition without food. Therefore, the loach is mainly a nocturnal species and the activity peak in the L phase seems to depend on the scheduled feeding in the L phase. Experiment 3 Effects of scheduled feeding on the locomotor activity rhythms under constant darkness (DD). Five out of six fish were considered to be entrained by the scheduled feeding (Table la). They showed a peak of anticipatory activity prior to the feeding time (type 1) (Fig. 3). The amount of activity reduced just after the feeding time and increased gradually until the next feeding time. 0 a LD 12:12 Food in L (Exp. 1-A) 10 LD 12:12 without Food 20 ep) > Lp 12:12 a —— _ ue SS SGa | without Food 99 a ar an ee rears rn ae ee SE aL TIME OF DAY (HR) Fic. 2. performed after Exp. 1-A (LD 12:12, Food in L). Horizontal bar: light condition (LD 12:12). Examples of locomotor activity rhythms of the loach in Experiment 2 (LD 12:12, without Food). (a) Exp. 2-A was (b) Exp. 2-B was performed after Exp. 1-B (LD 12:12, Food in D). Locomotor Activity Rhythms in the Loach 117 We recorded and analyzed the behavior of feeding- entrained activity pattern of a female loach under dim LL by aVTR. This female loach showed the type 1 in the feeding- entrained activity pattern. We divided the behavior of the SS) b ro) ro) Ba loach into searching and feeding behavior, and other swim- + 300 ming behaviors. Feeding crawl specified as crawling ex- = 200 ploration with plowing, plowing ahead and gulping (including ra dig and twist) was regarded as searching and feeding behavior > 100 [28]. From the results observed every one hour, only swim- o ming behavior at the upper and/or the bottom layer was BS 0 0 X 1 18 yi observed till 16:00 (two hours before the scheduled feeding TIME OF DAY (HR) time). Searching behavior and rest were added to swimming Fic. 3. An example of locomotor activity rhythms of the loach in at 16:00=17 200! Swimming at the upper layer decreased Experiment 3 (DD, Food at 12:00). Horizontal bar: light and searching behavior relatively increased, and swimming condition (DD). Triangle: feeding time. Data were shown as near the place where the fish was always fed was frequently mean+SE for 11 days. 0 LD 12:12 a Food in L (Exp. 1-A) 15 DD without Food 30 0 LD 12:12 b Food in D (ap) (Exp. 1-B) 215 =) DD without Food 30 0 DD c Food at 12:00 Se (Exp. 3) DD without Food | 25 0 12 24 12 24 TIME OF DAY (HR) Fic. 4. Free-running locomotor activity rhythms of the loach in Experiment 4 (DD, without Food). (a) Exp. 4-A was performed after Exp. 1-A (LD 12:12, FoodinL). Free-running period ( t)=23.5hr. (b) Exp. 4-B was performed after Exp. 1-B (LD 12:12, FoodinD). 7c=24.1hr. (c) Exp. 4-C was performed after Exp. 3 (DD, Food at 12:00). 7=24.7 hr. Data were double-plotted. 118 M. NARUSE AND T. OISHI observed at 17:00-18:00. When the food was given at 18:00, the fish fed it within a few minutes and gradually became inactive. Whenever the fish was active, it was usually at the bottom layer. There were swimming and rest at the bottom from 19:00 to 22:00. Swimming at the upper layer appeared at 22:00-23:00 and it was increased at 23:00-0:00. This series of behavior corresponded to the feeding-entrained activity pattern of the female loach re- corded by the actograph. In this experiment, it was shown that the scheduled feeding could induce the locomotor activ- ity rhythm of the loach. Experiment 4 Free-running rhythms under constant condi- tions after Exp. 1 and 3. (A) Fish were subjected to DD without food after Exp. 1-A. Only one of the 14 fish showed free-running rhythms for about 15 days (r=23.5 hr) (Fig. 4a, Table 1b). This fish was light-active and entrained to scheduled feeding as type 1 in the Exp. 1-A. (B) Fish were placed under DD without food after Exp. 1-B. Four of the 14 fish showed free-running rhythms for about 7— 13 days. One of them showed shorter free-running period than 24.0 hr (t=22.0 hr) and three fish showed longer free- running periods than 24.0 hr (c=24.1, 24.1 and 25.6 hr) (Fig. 4b, Table 1b). All of these fish were dark-active and en- trained to scheduled feeding as type 1 in the previous Exp. 1-B. (C) Fish were placed under DD without food after Exp. 3. Three of the six fish showed free-running rhythms for about 5-9 days. One fish showed the free-running period of 22.1— 23.7 hr and two fish showed free-running periods of 24.7 (Fig. 4c) and 28.2 hr (Table 1b). The fish with r=28.2 hr showed free-running rhythms throughout Exp. 4-A to C. The free- running period in Exp. 4-A and B was 23.5 and 22.0 hr, respectively. Two fish with longer free-running period than 24 hr were entrained to scheduled feeding as type 1 in the previous Exp. 3 and one fish with shorter free-running period than 24 hr was not entrained to scheduled feeding. Since free-running rhythms were observed after entrain- ment to the scheduled feeding, feeding can be considered as a zeitgeber in the loach. DISCUSSION Reports on the effect of scheduled feeding cycles on the locomotor activity rhythm have been increasing. In the goldfish (Carassius auratus), the scheduled feeding with LD cycle affected the locomotor activity rhythm, growth rate, and serum-cortisol and -thyroxine concentrations [14, 21]. In the medaka (Oryzias latipes), scheduled feeding cycles had different influences on the different types of behavior, that is, the agonistic behavior was entrained to the feeding cycle, but the egg laying and courtship behavior were entrained rather to LD cycles than to feeding cycles [29]. In the mudskipper (Periophthalmus cantonensis), the locomotor activity rhythm could be entrained to the 12-hour feeding cycle [13]. The channel catfish (Ictalurus punctatus) showed subjective feed- ing time under ad-lib feeding regimen [17]. In contrast, the scheduled feeding did not affect the locomotor activity rhythm in the blenny (Blennius pholis) [4]. Feeding both in the L phase and in the D phase could entrain the activity rhythm of the loach (Exp. 1). The activity pattern varied widely when the fish were fed in the L phase, but they were consistently dark-active when they were fedin the D phase. In Exp. 2, the activity patterns tended to change to dark-active under the regimen without food. Therefore, fundamentally the loach seems to be a nocturnal species. Benthic fishes, such as the eel, catfish and loach, have been considered as nocturnal or light-dark-active species because benthic fish approach to their food horizontally and rely on not only vision but also other senses to find food [20]. However, since cone cells and cone visual pigments (iodop- sin-like substances) existed in the retina of the loach [11], they seem to have an ability to be active during daytime. In the field, probably the availability of food for the loach do not change diurnally because the loach is an omnivorous detritus feeder. In the present study, however, loaches were en- trained to the daily feeding cycle, although food (tubifexes) was not taken away, and thus, the fish could feed these tubifexes alive in the bottom sand whenever they want, and this was supposed to provide a weaker influence on the entrainment than the food-removed regimen. Thus, the importance of feeding in the circadian structure of the loach should not be neglected. Davis and Bardach [3] indicated the importance of the anticipatory activity peak that appeared prior to the sche- duled feeding time. This activity peak was also mentioned by Aschoff [1], in which it was suggested to appear under both LD cycles and constant conditions. In the present study, the loach showed the anticipatory activity peak in both LD cycle and constant darkness (DD). The loach showed conspicuous resting periods of several hours after the feeding time. There were no reports about this type of resting period in the study of the feeding- entrained rhythm. Thus, this resting period may be unique to the locomotor activity rhythm of the loach entrained to the feeding time. The reason why the loach needs this period may be due to the fact that they are benthic fish and they have to spend many hours to digest food [26]. It has been suggested in fishes that free-running rhythms are labile and do not last for a long time [15, 18], and the ratio’ of individuals with free-running rhythms is lower than those of higher vertebrates, although there are some exceptions such as the lake chub (Couesius plumbeus) [9], the goldfish [10], two species of the hagfish (Eptatretus burgeri, Paramyx- ine atami) [8, 16] and the catfish ( Silurus asotus) [25]. In the loach, free-running rhythms lasted for 5-15 days and thus, the locomotor activity rhythm of the loach is an endogenous circadian rhythm. However, the ratio of individuals which showed free-running rhythms was low and varied from 7 to Locomotor Activity Rhythms in the Loach 119 50% depending on experiments, and the rhythm persisted only for short periods. This indicates that the extent of coupling between the oscillator and the locomotor activity seems to be weak and might differ depending on individuals. Since the periodic feeding in the rat caused to uncouple the feeding-anticipatory peak from the component of free- running rhythm [2, 6, 22], it is suggested that the feeding- entrained oscillator is different from the LD cycle-entrained oscillator. The SCN-lesioned animal that showed arrhyth- mic locomotor activity also showed the anticipatory activity prior to feeding [2, 22]. Different oscillator systems for feeding and LD cycles might also exist in the loach, because the uncoupled anticipatory peak was observed and gradually disappeared in Exp. 1-A and 2-A. ACKNOWLEDGMENTS We are grateful to Prof. Yoshihiko Chiba, Dr. Kenji Tomioka, Department of Biology, Yamaguchi University, and Mr. Sou Miyake, Department of Electrical Engineering, Kyoto University, for providing us actographs and the program of the actogram. REFERENCES 1 Aschoff J (1987) Effects of periodic availability of food on circadian rhythms. In “Comparative Aspects of Circadian Clocks” Ed by T Hiroshige, K Honma, Hokkaido Univ Press, Sapporo, pp 19-41 2 Boulos Z, Rosenwasser AM, Terman M (1980) Feeding sche- dules and the circadian organization of behavior in the rat. Behav Brain Res 1: 39-65 3 Davis RE, Bardach JE (1965) ‘Time-co-ordinated prefeeding activity in fish. Anim Behav 13: 154-162 4 Gibson RN (1971) Factors affecting the rhythmic activity of Blennius pholis L. (Teleostei). Anim behav 19: 336-343 5 Godin J-GJ (1981) Circadian rhythm of swimming activity in juvenile pink salmon (Oncorhynchus gorbuscha). Mar Biol 64: 341-349 6 Honma K, von Goetz Ch, Aschoff J (1983) Effects of restricted daily feeding on freerunning circadian rhythms in rats. Z vergl Physiol 62: 93-110 7 Johnson T, Miller K (1978) Different phase position of activity in juvenile and adult perch. Naturwissenschaften 65: 392-393 8 Kabasawa H, Ooka-Souda S (1991) Circadian rhythms of locomotor activity in the hagfish and the effect of reversal of the light-dark cycle. Bull Japan Soc Sci Fish 57: 1845-1849 9 Kavaliers M (1978) Seasonal changes in the circadian period of the lake chub, Couesius plumbeus. Can J Zool 56: 2591-2596 10 Kavaliers M (1981) Period lengthening and disruption of socially facilitated activity rhythms of goldfish by lithium. Phy- siol Behav 27: 625-628 11 Kawata A, Oishi T, Fukada Y, Shichida Y, Yoshizawa T (1992) Photoreceptor cell types in the retina of various vertebrate species : Immunocytochemistry with antibodies against rhodop- sin and iodopsin. Photochem Photobiol 56: 1157-1166 12 Miller K (1978) Locomotor activity of fish and environmental 14 15 16 17 18 19 20 21 22 23 24 25 26 AT 28 29 30 oscillations. In “Rhythmic Activity of Fishes” Ed by JE Thorpe, Academic Press, London, pp 1-19 Nishikawa M, Ishibashi T (1975) Entrainment of the activity rhythm by the cycle of feeding in the mud-skipper, Periophthal- mus cantonensis (Osbeck). Zool Mag 84: 184-189 Noeske TA, Spieler RE (1984) Circadian feeding time affects growth of fish. Trans Am Fish Soc 113: 540-544 Oishi T (1991) Fishes. In “Handbook of Chronobiology (In Japanese)” Ed by Y Chiba, K Takahashi, Asakura Shoten, Tokyo, pp 69-78 Ooka-Souda S, Kabasawa H, 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 Randolph KN, Clemens HP (1976) Some factors influencing the feeding behavior of channel catfish in culture ponds. Trans Am Fish Soc 105: 718-724 Richardson NE, McCleave JD (1974) Locomotor activity rhythms of juvenile Atlantic salmon (Salmo salar) in various light conditions. Biol Bull 147: 422-432 Saitou K, Katano O, Koizumi A (1988) Movement and spawn- ing of several freshwater fishes in temporary waters around paddy field. Jpn J Ecol 38: 35-47 Sawara Y (1989) The ecology of rhythmic activity in fishes (I). Biol Sci 41: 57-67 Spieler RE, Noeske TA (1984) Effects of photoperiod and feeding schedule on diel variations of locomotor activity, corti- sol, and thyroxine in goldfish. Trans Am Fish Soc 113: 528-539 Stephan FK (1981) Limits of entrainment to periodic feeding in rats with suprachiasmatic lesions. J Comp Physiol A 143: 401- 410 Suzuki R (1983) Multiple spawning of the cyprinid loach, Misgurnus anguillicaudatus. Aquaculture 31: 233-243 Tabata M (1988) 6. Diel and circadian locomotor activity in fishes. In “Daily Rhythmic Activities in Aquatic Animals (In Japanese)” Ed by I Hanyu, M Tabata, Koseikaku Koseisha, Tokyo, pp 79-100 Tabata M, Minh-Nyo M, Niwa H, Oguri M (1989) Circadian rhythm of locomotor activity in a teleost, Silurus asotus. Zool Sci 6: 367-375 Tanaka K (1955) Observation on the length of the digestive time of feed by the mud loach, Misgurnus anguillicaudatus. Japan J Ichthyol 4: 34-39 Ueda M, Oishi T (1982) Circadian oviposition rhythm and locomotor activity in the Medaka, Oryzias latipes. J interdis- cipl Cycle Res 13: 97-104 Watanabe K, Hidaka T (1983) Feeding behaviour of the Japanese loach, Misgurnus anguillicaudatus (Cobitididae). J Ethol 1: 86-90 Wever DN, Spieler RE (1987) Effects of the light-dark cycle and scheduled feeding on behavioral and reproductive rhythms of the cyprinodont fish, Medaka, Oryzias latipes. Experientia 43: 621-624 Yanagishima S, Mori S (1951) Relation between activity and glycogen contents of the Japanese loach. Mem Coll Sci Univ Kyoto, Ser B 20, 1-6 Yokota T, Oishi T (1992) Seasonal change in the locomotor activity rhythm of the medaka, Oryzias latipes. Int J Biometeorol 36: 39-44 i 2 hh yhoo” | 4 3} =. pca ey ay GRAD m1 i cD TESPT) St saben : ‘ AL = A) wet : an rey atve ¥- 7 wonaT -TTVeTT Paty 4 eee yf Deg Fae M:, ) Ay é ae: eee i ‘fa j 71 fO taal Ath 3 an vs Naive if ce i : foewey 2 mi ‘2 7 - Eisai ey Y r ‘ 2 F iy ¥ . as eh i { a ay i ¥ + ’ ; 4 2 r " [ sentiilnds 2 re Realodag Pye h wat an ogee Prid A UY bet idel Beret ree 1 CN ORL ee Ta yale Mf lacy (Pyaet th) ah sictadver ya ry ‘shes ” ns coil ™ iq) a niga, ‘ nlentrnledy + ecriberts ye om R afesl by iF an 4 wt, tbls nt eet ieee / a ash wa) weil veel Bisse niente sic RFR idee 2 ey it “winbiee2 Seach " “tines jekontlr GARE wind sacar tees tas.) Dotine Maes ati’ ’ tod eee erenyGrset ci! oP eee riba 'd 5. wid hea sta caves -- a 9 fe 7 Yn Wi, slve's 2 eS “Or VR? ay ert) wy Lab ivebl> aay het ' ~ beak roy ite + ied “dilbat hd sean ig “4 Gt Do | wes eon . ny aa ur ™ ‘ a ~~ a 7 . ’ és a eri sucha tit Soe ea ae otal vtech artearpst tii iq é sinh iy ie) Chel lo naieiohy sili pl “ L £F hers é be Sh AE Vi L ine “wis “uty, degre ard 2 reali ol abonies em aon ‘ * J ny * A Tugs * oc ; thy , iV bo? eerie ddr pape wilh ware Recee di hah et Gas be sh iy wi vont id Ts Ba nil (ass 4 : : re . iat pewrarenh epi aii . 3 ri ll oe sero Aico Oey 2 Sau nkgan ate Het ac rly ‘ey j : T . ‘ wes ar CAN Ee ae ek ” ha. it Ay ‘Ya . aN nag é : ia tot : 1 pariah edi Wh phen law bos tees ag = or { >a Larcic ays ot Biinyhs fame A ‘Oger fv 7, ; fe 1) tars THES Javed ett Selpetod ry sbelltiiat aera ys! he tenlogived nes nites } 5* F i pd ut r Wa tery. " vanmaanned Ma sul ruce’ bonginbie iq pen) tee UNE val ' , fi i on " f — ' = (kaa f , t. ‘ ’ at : ad is: Pa T Sogeme > oe (ime _— f (ies, Sue bie. alee ee ‘ve. (erry OR how yg Meee eugl ZOOLOGICAL SCIENCE 11: 133-137 (1994) © 1994 Zoological Society of Japan Allocation of Resources to Body Components in Heliocidaris erythrogramma and Heliocidaris tuberculata (Echinodermata: Echinoidea) JoHn M. LAwrence! and Maria ByRNEZ2 "Department of Biology, University of South Florida, Tampa, Florida 33620, U.S.A. and *Department of Histology and Embryology, University of Sydney, Sydney, N.S.W. 2006, Australia ABSTRACT— Adult Heliocidaris erythrogramma (He) (50 g wet body weight) and H. tuberculata (Ht) (200 g wet body weight) are compared. The relative sizes of the test and spines are the same for both species, but that of the Aristotle’s lantern is 2-fold larger in He. The gonad index shows no sexual differences and is similar for both species (He: 10, Ht: 12). The proximate composition of the gut, Aristotle’s lantern, test, and spines of both species are similar. The percent organic material in the ovaries of He is greater than in the testes and in the ovaries and testes of Ht. Lipid is 50% of the dry weight of the ovaries of He, twice that in the ovaries and testes of Ht. DNA is 6% of the dry weight of the testes of He, half that in the testes of Ht. The direct development in He and indirect development in Ht is correlated with differences in body size and proximate composition of the gonads. INTRODUCTION Grime [11] pointed out the desirability of dtudying both the established and regenerative phases in the life-history of an Organism in considering adaptive strategies, as they may be subject to different forms of natural selection or respond differently to the same selecltive force. Grime stated these phases are uncoupled, noting the radical differences between larval and adult phases of invertebrates. He suggested a particular regenerative strategy may be modified by two important variables: the strategy adopted during the estab- lished phase and the breeding system. A case in point is the genus Heliocidaris. Clark [3] stated the genus Heliocidaris A. Agassiz and Desor 1846 contains only two species that differ obviously in general appearance but not in specifics. He separated the species on the basis of number of pore-pairs, a characteristics not expected to cause major functional differences. Among the gross differences is the smaller size of H. erythrogramma Valenciennes 1846 (to 86 mm horizontal diameter) compared to H. tuberculata Lamarck 1816 (to 106 mm horizontal dia- meter). The two species differ in habitat. In general, Heliocidaris erythrogramma is a rock-burrowing species at low-tide level and H. tuberculata is subtidal on the Australian coast and on the reef flat on Lowe Howe Island [3, 4]. The two species probably have different levels of resource availa- bility as well as exposure. Considerable interest in the two species has resulted from the differences in their development noted by Morten- sen [19, 20] and documented by Williams and Anderson [30]. H. erythrogramma has direct development in contrast to H. tuberculata. Their divergent ontogenies have provided a Accepted January 4, 1994 Received October 7, 1993 model for studying the evolution of development [22]. Stud- ies of the established adult phase have not accompanied these studies. The ways resorces are allocated to structures and func- tions are important lief-history characteristics [2]. Amont the most important is the trade-off associated with increased age and size involving a transition to increased allocation to reproduction and decreased allocation to growth. Strath- mann & Strathmann [27] noted the association of brooding with small adult-size in marine invertebrates, and pointed out the smaller of co-occurring species are the brooders. None of the hypotheses considered (allometry of egg protection and brood care, longevity with iteroparity, variable recruitment, dispersal) involved the allometry of acquisition and allocation of resources. Being aware of the difference in size of the established phase of the two species and the trade-off models involving growth and reproduction, we hypothesized allocation of resources to body components in Heliocidaris erythrogramma and H. tuberculata would differ and that the difference could be correlated with their biology and ecology and with the differences in their regenerative phases. We believe an understanding of the characterisitics of the established phase can assist in understanding those of the regenerative phase. MATERIALS AND METHODS Heliocidaris erythrogramma were collected at Little Bay on 21 November 1991 and H. tuberculata were collected at Shelley Beach, New South Wales, Australia on 7 June and 2 August 1992. These dates are at the maximal gonad indexes of the species’ broad reproductive cycles [13]. Individuals were dissected into body com- partments that were lyophilized and weighed. The components were homogenized in a Wiley mill for proximate analysis. Ovaries and testes were analyzed separately, but no sexual distinction was made for the somatic compartments as evidence indicates no sexual 134 J. M. LAWRENCE AND M. BYRNE differences in echinoids [18]. Lanterns were treated with hydrogen peroxide to obtain the demi-pyramids and rotules by the method of Ebert [5]. Pigmented growth lines in the lantern were counted by the method of Jensen [12] of three specimens of each species. Proximate analysis and calculation of energy equivalents was done by the methods used by Lawrence [14] and Watts and Lawrence [29]. Significant differences were tested by ANOVA except for concentra- tion of DNA, which was compared by Student’s t-test (P<0.05). Data were transformed as necessary to satisfy conditions of normality and equal variances. RESULTS The sized of the individuals and weights of the body compartments are given in Table 1. Neigher Heliocidaris erythrogramma or H. tuberculata showed sexual dimorphism in size. Heliocidaris erythrogramma was smaller than H. tuberculata,, the difference in weight (4-fold less) being greater than the difference in diameter (1.4-fold less). The weights of the body compartments were correspondingly smaller in H. erythrogramma than in H. tuberculata. The relative sizes, indexes based on dry weight, did not differ significantly except for the Aristotle’s lantern, which was 2-fold greater in H. erythrogramma (Table 2). The demi- pyramids of specimens of both species had 4 to 6 pigmented lines, indicating equivalent age and faster growth in H. tuberculata. The proximate composition of the body compartments of the two species (Table 3) are similar except for the gonads. TABLE 1. Heliocidaris erythrogramma and _ Heliocidaris tuberculata: Horizontal diameter (mm); wet body weight and compartment dry weights (g) TaBLeE2. Heliocidaris tuberculata: Gonad index (% wet body weight); test, spine, and lantern indexes (% dry weight of somatic compartments). erythrogramma and Heliocidaris Sex Heliocidaris Heliocidaris erythrogramma tuberculata Gonad index 9.9 (1.5) 13.3 (0.6) [9] [10] F 9.1 (0.7) 11.1 (0.8) [14] [8] Test index 45 (1) 47 (1) [23] [16] Spines index 50 (1) 50 (1) [23] [16] Lantern index 5.3 (0.1) A 2.7 (0.1) B [23] [16] Sexes are combined except for gonads. (in parenthesis), and n (in brackets) are given. not significantly different (P>0.05) except for the lantern index. TABLE 3. Heliocidaris erythrogramma and Means, standard error The indexes are Heliocidaris tuberculata: Proximate composition (% dry weight) Sex Heliocidaris Heliocidaris erythrogramma tuberculata Horizontal diameter 67 (1) 94 (1) [23] [30] Wet body weight 53 (2) 201 (6) [23] [30] Gonads dry weight M 1.00 (0.32) 6.56 (0.61) [9] [14] F 1.22 (0.15) 5.59 (1.76) [14] [9] Test dry weight 14 (1) 57 (3) [23] [19] Spines dry weight 16 (1) 63 (3) [23] [20] Lantern dry weight 1.6 (0.01) 3.2 (0.01) [23] [19] 0.027 (0.001) 0.063 (0.003) [19] [18] 0.096 (0.004) 0.182 (0.007) [19] [18] Rotule dry weight Demi-pyramid dry weight Sexes are combined except for gonads. Means, standard error (in parenthesis), and n (in brackets) are given. All values for the two species are significantly different (P<0.05). Sex Heliocidaris Heliocidaris erythrogramma tuberculata Gonads % organic material M 91 (1) A 91 (1) A [9] [14] F 96 (0) B- 92 (1) A [13] [8] % carbohydrate M 11 (1) A 11 (1) A [8] [14] F 8 (1) B12 A [14] [8] % lipid M 26 (2) A 22 (1) A [9] [13] F 50 (2) B27 (1) A [14] [9] % soluble protein M 37 (3) AB 42 (2) BC [9] [13] F 29 (2) A 45 (3) B [14] [9] % insoluble protein M 11 2 F 9 9 % DNA M 5.5 (1.9) A 14.1 (1.2) B [4] [5] Test % organic material 18 (1) 22 (2) [20] [18] % carbohydrate 3.21 (0.18) 3.68 (0.35) [9] [9] % lipid 1.28 (0.04) 1.38 (0.08) [12] [10] % soluble protein 4.85 (0.28) 4.97 (0.30) [12] [10] % insoluble protein 9 12 Allocation in Heliocidaris 135 Spines % organic material 16 (1) 17 (1) [27] [18] % carbohydrate 0.36 (0.10) 0.27 (0.02) [10] [10] % lipid 1.14 (0.10) 0.02 (0.07) [12] [10] % soluble protein 3.98 (0.30) 2.51 (0.13) [12] [10] % insoluble protein 11 14 Lantern % organic material 20 (1) 24 (2) [20] [19] % carbohydrate 4.09 (0.32) A 0.78 (0.15) B % \ipid 1.55 (0.10) 1.38 (0.23) [12] [10] % soluble protein 5.95 (0.68) 1.38 (0.23) [12] [10] % insoluble protein 9 15 Means, standard error (in parentheses), and n (in brackets) are given. Insoluble protein values were calculated from means. Values with the same letter or no letter are not significantly different. The ovaries of H. erythrogramma had a slight but significantly higher concentration of organic material than those of H. tuberculata. This was due to a 2-fold greater concentration of lipid in the ovaries compared to the testes of H. erythro- gramma and both the ovaries and testes of H. tuberculata. The concentrations of the other proximate components in the ovaries of H. erythrogramma were correspondingly reduced. The concentration of DNA in the testes of H. erythrogramma was one-third that in the testes of H. tuberculata. Most energy in the body compartments was in the form of protein, even in the gonads, except in the ovaries of Heliocidaris erythrogramma where most was lipid (Table 4). Most energy in the somatic compartments was in the form of insoluble protein. The ratios of the weights of the wet body and dry somatic components of Heliocidaris erythrogramma to H. tuberculata were similar, indicating isometry. Like- wise the somatic component indexes were similar but the gonad indexes were ca. 2-fold greater in energy units than in gravimetric units for both species. Five to six growth lines were present on the demi- pyramids of three specimens of each species. DISCUSSION The established, mature phases of Heliocidaris erythro- gramma and H. tuberculata differ in three major ways: the dimensions and weights of the body and body compartments are smaller in H. erythrogramma; the concentration of lipid in the ovaries is greater in H. erythrogramma and the concentration of DNA in the testes is less in H. erythrogramma ; the relative sizes of the body compartments are similar except the Aristotle’s lantern index is greater in H. TaBLeE 4. Heliocidaris erythrogramma and _ Heliocidaris tuberculata: Kilojoules in body compartments of females Heliocidaris Heliocidaris erythrogramma tuberculata Ovaries Carbohydrate iL iU7/ 10.63 Soluble protein 8.03 54.56 Insoluble protein 3.54 19.37 Lipid 23.30 54.91 Sum 36.41 139.47 Test Carbohydrate 1.37 7.88 Soluble protein 2.83 14.64 Insoluble protein 53.61 266.64 Lipid 1.19 6.72 Sum 59.00 295.88 Spines Carbohydrate 0.17 0.51 Soluble protein 2.36 6.38 Insoluble protein 56.92 245.38 Lipid 0.19 <0.01 Sum 60.64 252.27 Lantern Carbohydrate 0.17 0.17 Soluble protein 0.47 1.18 Insoluble protein 6.61 16.30 Lipid <0.01 0.40 Sum V2) 18.05 Sum all somatic tissues 126.89 566.20 Sum all somatic tissues 163.30 705.67 and ovaries Ovaries/total somatic 22 20 tissue and ovaries erythrogramma. The high concentration of lipid in the ovaries of Helio- cidaris erythrogramma is at the upper limit reported for the ovaries and eggs of other direct developing echinoderms [8, 15]. Wray & Raff [in 24] reported the eggs of H. erythro- gramma were ca. 50% lipid by volume. Energy investment per egg would be greater in H. erythrogramma than in H. tuberculata as a result of the higher concentration of lipid and the larger egg-size. This is the first report of the concentration of DNA in echinoderm testes. As the sperm-head length of Heliocida- ris erythrogramma is ca. two-fold greater than that of H. tuberculata while the DNA content of the sperm is only ca. 1.3-fold [13, 23], the difference in concentration seems to result from a packing phenomenon. The difference in sperm size affects both energy investment per sperm and sperm number per individual. Heliocidaris erythrogramma invests more energy per sperm, and produces fewer sperm per individual as a results of its larger sperm and smaller body size. 136 J. M. LAwRENCE AND M. ByRNE Allometry is the general rule with increase in body size [21]. consequently, the isometry of the test, spine, and particularly the gonad indexes of the asymptotic sized of Heliocidaris erythrogramma and H. tuberculata was not ex- pected. Allometry of the gonads has been reported in several strongylocentrotid species up to a body size at which isometry occurs [9, 10, 28]. The continued increase in relative gonadal production in Strongylocentrotus francisca- nus beyond the body size at which it ceases to increase in S. intermedius and S. purpuratus is correlated with lantern size and allometric coefficient (Lawrence et al., unpub.). Con- sumption is the largest term in the energy budget and sets the upper limit for all other variables [21]. Assuming the rela- tive size of the lantern is indicative of feeding capacity [5], H. erythrogramma should have a greater relative capacity for resource acquisition than H. tuberculata at asymptotic body sized. The lantern sizes of similarly-sized (horizontal dia- meter=60-70 mm) H. tuberculata and H. erythrogramma are similar (ca. 1.5 g dry weight, Lawrence and Byrne, unpub.). It seems the conclusion of Strathmann & Strathmann [27] that the species of related co-occurring marine inverte- brates that broods has the smaller adults can be extended to species with direct development. This generalization does not address the question of absolute size. Heliocidaris erythrogramma is no smaller than the strongylocentrotid species that have indirect development. If stressful condi- tions (decreased potential for production) tend to result in the evolution of species that invest more energy per propagule [16, 17], H. erythrogramma may have less capacity for production so that it ceases somatic growth at a smaller size, and makes up for decreased fecundity by direct development. These considerations fit Ebert’s [6] conclusion that H. erythrogramma is a very long-lived species. The longevity of H. tuberculata is not known. The similar number of growth lines but different body sizes indicates H. tuberculata grows more rapidly than H. erythrogramma. Overall, the differ- ences in asymptotic body size, growth rates, and fecundity between the two species seem related to their potential for production. Growth has both genetic and environmental controls [25]. Although echinoids fit the indeterminate growth pat- tern (plastic asymptotic, growth), in which the growth trajec- tory and adult size are determined by energy intake and costs, a genetic component exists as seen here with the two species of Heliocidaris. We assume the differenc in growth rates and asymptotic sizes of Heliocidaris erythrogramma and H. tuberculata involved the evolution of genetic control of growth patterns. The difference in developmental pattern in H. erythrogramma involved the evolution of genetic control of developmental processes [22]. Did the evolution of both the regenerative and established phase occur simultaneously or sequentially? If indirect development is lost more fre- quently than gained evolutionarily [7, 8, 26], the direction of evolution was from H. tuberculata to H. erythrogramma. ACKNOWLEDGMENTS We thank Daniel Bishop and Steven Beddingfield (University of Alabama at Birmingham) for analyzing the DNA in the testes and growth lines in the lantern, and Carolyn Yates (USF) for proximate analysis. REFERENCES 1 Brody S (1945) Bioenergetics and Growth. Hafner Pub- lishing Company, New York, pp 1023 2 Calow P (1984) Economics of ontogeny—adaptational aspects. In “Evolutionary Ecology” Ed by B Shorrocks, Black- well Scientific Publications, Oxford, pp 81-104 3 Clark HL (1946) The echinoderm fauna of Australia: its com- position and origin. Carnegie Inst Wash Pub 566, pp 567 4 Dakin WJ (1952) Australian Seashores. Angus and Robert- son, Sydney, p 372 5 Ebert TA (1980) Relative growth of sea urichin jaws: an example of plastic resource allocation. Bull Mar Sci 30: 467— 474 6 Ebert TA (1982) Longevity, life history, and relative body wall size in sea urchins. Ecol Monogr 52: 353-394 7 Emlet RB (1990) World patterns of developmental mode in echinoid echinoderms. Adv Invert Reprod 5: 329-335 8 Emlet RB, McEdward LR, Strathmann RR (1987) Echi- noderm larval ecology viewed from the egg. Echinoderm Stu- dies 2: 55-136 9 Fuji A (1967) Ecological studies on the growth and food consumption of Japanese common littoral sea urchin, (Strongy- locentrotus intermedius) (A. Agassiz). Mem Fac Fish Hokkaido Univ 15: 83-160. 10 Gonor JJ (1972) Gonad growth in the sea urchin (Strongy- locentrotus purpuratus) (Stimpson) (Echinodermata: Echi- noidea) and the assumptions of the gonad index methods. J Exp Mar Biol Ecol 10: 89-103 11 Grime JP (1979) Plant strategies and vegetation. John Wiley & Sons, Chichester 12 Jensen M (1969) Age determination of echinoids. Sarsia 37: 41-44 13 Laegdsgaard P, Byrne M, Anderson DT (1991) Reproduction of sympatric populations of Heliocidaris erythrogramma and H. tuberculata (Echinoidea) in New South Wales. Mar Biol 110: 359-374 14 Lawrence JM (1973) Level, content, and caloric equivalents of the lipid, carbohydrate, and protein in the body components of Luidia clathrata (Echinodermata: Asteroidea: Platyasterida) in Tampa Bay. J Exp Mar Biol Ecol 11: 263-274 15 Lawrence JM (1987) A Functional Biology of Echinoderms. Crooms-Helm, London, p 340 16 Lawrence JM (1990) The effect of stress and disturbance on echinoderms. Zool Sci 7: 17-28 17 Lawrence JM (1991) Analysis of characteristics of echinoderms associated with stress. In “Bioology of Echinodermata” Ed by T Yanagisawa, I Yasumasu, C Oguro, N Suzuki, T Motokawa, Balkema, Rotterdam, pp 11-26 18 Lawrence JM, Lane JM (1982) The utilization of nutrients by postmetamorphic echinoderms. In “Echinoderm Nutrition” Ed by M Jangoux, JM Lawrence, Balkema, Rotterdam, pp 331- 371 19 Mortensen TH (1915) Preliminary note on the remarkable, shortened development of an Australian sea-urchin, Toxocidaris rtythrogramma. Proc Linn Soc N S Wales 40: 203-206 20 Mortensen TH (1921) Studies of the development and larval 21 22 23 24 25 Allocation in Heliocidaris 137 forms of echinoderms. IV. K Danske Vidensk Selsk (Naturvid Math Afd Ser 9) 7(3): 1-59 Peters RH (1983) The Ecological Inrplications of Body Size. Cambridge Univ. Press, Cambridge Raff RA (1992) Direct-developing sea urchins and the evolu- tionary reorganization of early development. Bio Essays 14: 211-218 Raff RA, Herlands L, Morris VB, Healy J (1990) Evolutionary modification of echinoid sperm correlates with developmental mode. Dev Growth Diff 32: 283-291 Scott LB, Lennarz WJ, Raff RA, Wray GA (1990) The “lecithotrophic” sea urchin Heliocidaris erythrogramma lacks typical yolk platelets and yolk glycoproteins. Develop Biol 138, 188-193 Sebens KP (1987) The ecology of indeterminate growth in animals. Ann Rev Ecol Syst 18: 371-407 26 “it 28 29 30 Strathmann RR (1987) The evolution and loss of larval feeding stages of marine invertebrates. Evolution 32: 894-906 Strathmann RR, Strathmann MF (1982) The relationship be- tween adult size and brooding in marine invertebrates. Am Nat 119: 91-101 Tegner MJ, Levin LA (1983) Spiny lobsters and sea urchins: analysis of a predator-prey interaction. J Exp Mar Biol Ecol 73: 125-150 Watts SA, Lawrence JM (1985) The effect of feeding and starvation on the level and content of nucleic acids in the pyloric caeca of Luidia clathrata (Say). In “Echinodermata” Ed by BF Keegan & BDS O’Connor, Balkema, Rotterdam, pp 571-576 Williams DHC, Anderson DT (1975) The reproductive sys- tem, embryonic development, larval development and meta- morphosis of the sea urchin Heliocidaris erythorogramma. Aust J Zool 23: 371-403 j iam oe ae - = dad Wow tealidere sae ee barney ae wi te aseg sale Mir ty! rT} ‘SioPticly Paee ] ba os Rl ij : by LY GD Dan inet SEP eRe ka Pret 9 (Gp aT atuaag od thllivdides SMe Ce tabi | at @ ! . » Sek TMD fp. gebieasanie Fee ie arirael cow Meme tetatfrnbeat tate i (RMR Eph hee Ly th, loves a Mater vi a ‘ { Peat Toavhgader nity Pou & A rie 9 i i e iD . ve ans Es a pteeiieen “(et ea (ral) Mat BousiWwed-, Ae re awe” tae Ma nearer ss: dite 23} At: * Thos? Wa be ede es vada Ee), sae i sien ot Paced Menken, lee BRN Aer Pe oy peariberetiel Lira nonetkey (a aC doritgect - ~~ upset Went: bi at mat Hiee ibis} yale ur Cie rast Beane 1. nt ly >ireov eles. spade an * Laat cean Egn he ; sdncaiey inet ny wa By ili ORI ; a o' Se f rhi.laere rh ae ; “Was pay 2068 ia aie sa Canveige bal wei Pié ¢ (Ang Honea f = th? Part Tay TR” iesiaitne A * of ror Pes nee ad ; - 4. ey A y ime Pare ; : an f eg wy J a ms ae = — — | 5 a - a & io a Fa i= ‘ : \ ‘ 4 7 2 eT “ ‘ ¥ i WeReethy he « 4 is ce e ‘ Ta | . i oo SB, — Fe : fl F S: my NE On pee ; ; ‘ ly ‘ 1 - — * ah ‘ a a : : rt. ee * A ie eSetahe alli cm C * 4 ‘Soi ; ic is 7 “iy ; 1 Strnly he op of Pam .e wT < y's 1 1 or ikea? Vagal Bae). ee 7) mort”: i ‘

30.0» Agar 7.5 10.0 10.0 10.0 10.0 10.0 Potato starch 7.5 10.0 10.0 12.0 13.5 15.0 Sucrose 8.0 10.0 10.0 12.0 13.5 15.0 Cellulose powder 20.8 29.6 29.6 34.6 34.6 34.6 Soybean oil,not refined 1.5 2.0 2.0 ; 2.0 2.0 2.0 B-Sitosterol 0.2 0.2 0.2 0.2 0.2 0.2 Sorbic acid 0.2 0.2 0.2 0.2 0.2 0.2 Ascorbic acid 2.0 2.0 2.0 2.0 2.0 2.0 Citric acid 4.0 4.0 4.0 4.0 4.0 4.0 Wesson’s salt mixture 3.0 3.0 3.0 3.0 3.0 3.0 Total 115.7 116.0 116.0 116.0 116.0 116.0 Vitamin B mixture” Added Added Added Added Added Added Antiseptics” Added Added Added Added Added Added Distilled water(ml/g dry diet) for 1-4th instar diet 3.0 4.0 4.0 4.0 4.0 4.0 for Sth instar diet 2.2 3.0 3.0 3.0 3.0 3.0 1) Diet A: Standard diet for Bombyx mori developed by Horie et al. [7]; 2) Crude soybean meal, defatted; 3)Soybean meal, high nitrogen content (Solpea 600); 4)See Horie et al. [6]; 5)Antiseptics consisted of 0.015% (dry matter) of chloramphenicol and 0.75% (dry matter) of propionic acid. Agrius as a New Experimental Insect 145 Preparation of artificial diets Attificial diets were developed after slight modifications of the diet for the silkworm, Bombyx mori, described by Horie et al. [7]. The diet composition is shown in Table 1. The main change was the substitution of sweet potato leaf powder for mulberry leaf powder. Five different diets designated as SPLP-0, SPLP-5, SPLP-10, SPLP- 20 and SPLP-25 were prepared, each of which contained different amounts of sweet potato leaf powder. Mixed ingredients were stored in a refrigerator at 5(C. For the preparation of the diets, 300-400 ml of distilled water was added to 100g of the mixed dry powder. The diet with a high water content (400 ml) was supplied to the first four larval instars, and that with a low content (300 ml) to the final 5th instar. After blending in the prescribed amount of water, the mixture was steamed for about 50 minutes at 100°C, then the diet was cooled down to room temperature. Each wet diet was covered with a wrapping film (Krewrap) and stored at 5°C until use. Rearing method on artificial diets A rearing system was developed after slight modifications of the system for Manduca adopted by L. M. Riddiford and J. W. Truman, University of Washington, Seattle. Namely, each newly hatched larva was individually confined in a plastic cup (50 ml) with a small piece of food (ca. 6g). The diet was changed to a fresh one 7 days after feeding. When larvae molted into the 5th instar, each larva was transferred to a larger plastic cup (200 ml) with a lid on which several small circular holes were made to provide adequate aeration for the growing larva, and given a larger amount of food (ca. 25 g). Wandering larvae were transferred to another cup of the same size only with a piece of tissue paper, where they pupated. Throughout this experiment, eggs, larvae and pupae were kept in an environmen- tal room maintained at 25+1°C, 50-60% RH, and under a long day photoperiod (16L : 8D). RESULTS Rearing on fresh host plant leaves Mean pupal weights of the field-collected Agrius convol- vuli were 5.13 £0.83 g for males (n=53) and 5.47+1.01 g for females (n=50). Hatched larvae from the obtained eggs were reared on fresh harvested leaves of sweet potato without difficulty (Table 2). Percentage of survival to the adult stage exceeded 90%. However this method required much space, large quantities of fresh sweet potato leaves, and time to handle the hornworms. The life cycle from egg to egg was TasLe2. Growth and development of Agrius larvae on the fresh sweet potato leaves No. of eggs collected 255 No. of hatched larvae 207 (82%) No. of larvae placed on diets 207 No. of larvae that molted into 5th instar 196 (95%) No. of Sth instar larvae reared 163 No. of individuals that became pupae 152 (93%) No. of individuals that became adults 148 (91%) Days from hatching to 4th larval ecdysis 12.5+0.8 (92)* Days from 4th larval ecdysis to wandering 5.8+0.8 (65) Days from wandering to pupation 4.2+0.5 (64) 17.0+0.8 (60) 3.51+0.51 (89) 4.01+0.72 (63) Days from pupation to adult emergence Pupal weight (g) Male Female The rearing experiment was conducted in October 1989. * Values are mean+S.D., and number of insects observed is indicated in parentheses. TABLE 3. Growth and development of Agrius larvae on various artificial diets Diets used No. of Pupation Duration of 5th Pupal weight larvae instar (day) g 1—4th instar Sth instar tested* (%) (Mean+S.D.) (Sex, Mean+S.D.) Early generation (July, 1990) SPLP-25 SPLP-25 36 VD 5.7+0.8 M 4.70+0.34 (20) ** F 5.26+0.45 (18) SPLP-25 SPLP-20 36 94.4 5.7+0.8 M 4.72+0.41 (16) F 5.18+0.45 (15) SPLP-25 SPLP-10 36 100.0 5.8+0.7 M 4.77+0.46 (19) F 5.23+0.49 (17) SPLP-25 SPLP- 5 35 91.4 5.8+0.6 M 4.62+0.56 (10) F 5.18+0.62 (22) SPLP-25 SPLP- 0 36 91.7 6.9+1.2 M 4.46+0.53 (18) F 4.79+0.93 (15) Advanced generation (July, 1993) SPLP-25 SPLP-25 59 100.0 5.2+0.4 M 4.54+0.40 (34) F 5.11+0.30 (25) SPLP-20 SPLP- 5 52 100.0 5.0+0.5 M 4.55+0.34 (23) F 4.96+0.55 (29) * Larvae were selected on day O of the Sth instar ( see text for detail). ** Number of insects observed is indicated in parentheses. 146 K. KiGucHt AND M. SxHIMoDA about 42 days at 25+1°C and 65-75% RH (roughly, egg: 4, larva: 18, pupa: 18, preoviposition: 2 days). Mean pupal weights were 3.51 g for males and 4.01 g for females, which were significantly smaller than those of the field-collected individuals. Rearing on artificial diets We prepared five different diets (Table 1) after slight modifications of an artificial diet for Bombyx larvae (Diet-A in Table 1). Preliminary experiments revealed that newly hatched larvae showed a high feeding activity on the SPLP-25 and SPLP-20 diets. Moreover, we could successfully rear the hornworm on the SPLP-25 diet throughout the larval stage and obtained two further generations of larvae. Also, we observed that the Sth instar larvae fed and grew well, even on the diets containing smaller amounts of host plant leaf powder. To determine the optimum regimen for the least expense, we carried out combination experiments of two different diets. In July 1990, a total of 300 larvae were reared on the SPLP-25 diet until the molt to the 5th instar, then the 5th instar larvae were divided into several groups which were given different diets. As shown in Table 3, insects reared on the SPLP-25 diet throughout the larval stage showed a considerably higher performance in three developmental parameters examined than the individuals reared on fresh host plant leaves (Table 2). For example, the values for the mean pupal weights of the former were significantly larger than those of the latter in both sexes, although they were slightly smaller than those of the wild population. When 5th instar larvae were reared on the SPLP-O diet, larval duration was prolonged by 1 to 2 days, and the resultant pupae were smaller. By contrast, the other groups reared on the SPLP- 5, SPLP-10 and SPLP-20 diets did not show significant differences in the duration of larval life and in pupal weights when compared with those of the insects reared on the SPLP-25 diet. Although the initial colony consisted of less than 50 pupae at the start, and no wild individuals were introduced thereafter, we have not yet observed any serious inbreeding depression. For example, the hatchability in the first gen- eration which was 81% (Table 2) remained at the level of 70- 90% after 20 generations. Also, the pupal weights were nearly the same in the early and the recent generations as shown in Table 3. Thus, viability has remained stable. DISCUSSION We initially intended to identify an experimental insect suitable for comparative studies with the silkworm, Bombyx mori, which has been an important research target in our institute. Our attention was first directed to the tobacco hornworm, Manduca sexta, which has played an important role as an experimental insect. Since Manduca does not occur in Japan, we selected the sweet potato hornworm, Agrius convolvuli, a species closely related to Manduca sexta. As the morphology and life cycle of Agrius are similar to those of Manduca, the information accumulated on Manduca has been extremely useful in our attempt to develop an egg collection and rearing system. Actually, the system adopted for Agrius is basically the same as that for Manduca. Yet the composition of the artificial diet is very different from that developed for Manduca. We simply modified the diet for the commercial silkworm, Bombyx mori, by substituting sweet potato leaf powder for mulberry leaf powder. Our diets are satisfactory in nutritional requirements since we were able to rear the hornworm on the diet for over 20 generations during the past 3 years. During this time the hatchability of the eggs and survival rates did not change significantly, indicating that the diet is satisfactory for con- tinuous rearing procedures. However, there are a few problems which required further attention. First, it is preferable to develop a diet without host plant leaf powder. Manduca diets do not contain any leaf powder [1-3, 5, 18], while leaf powder is necessary for the current Agrius diet. At present we use SPLP-20 or SPLP-25 (ca. 17-22% leaf powder)for the 1st to 4th instar larvae, and SPLP-5 or SPLP-10 (ca. 4-9% leaf powder) for the last 5th instar. Yet it is noteworthy that most of the Sth instar Agrius larvae fed on the SPLP-O diet lacking leaf powder survived, although their larval develop- ment was prolonged and the resultant pupae were smaller (Table 3). Therefore, we consider that an artificial diet without leaf powder could be developed through changes in the diet composition compatible with feeding activity. The second problem concerns the number of changes of diet necessary. In the Manduca rearing system, the diet is usual- ly changed only once after a feeding larva reaches the 5th instar. By contrast, food must be changed twice in the Agrius rearing system to maintain an adequate larval de- velopment: first on the 7th day after hatching and second on the day when the larva molts to the 5th instar. To eliminate the first diet change, further improvement of the diets is required. Needless to say, research on insects can be greatly facilitated by the development of a year-round rearing sys- tem. However, this system provides only one of the neces- sary conditions for a suitable experimental insect, and it is also important to accumulate fundamental information on the physiology and behavior of the insect. Unfortunately the sweet potato hornworm, Agrius convolvuli, had not been studied thoroughly until now. Yet the insect could be a suitable experimental insect as it is closely related to Man- duca sexta. Presumably, the information accumulated on Manduca for the past 25 years will be useful for studies on the sweet potato hornworm. Studies on Agrius may also contri- bute to gain further insights into Manduca physiology and development. Although both Bombyx and Agrius are large lepidopter- ans that are similar in many characteristics, they are also very different in various aspects. For example, the silkworm uses ingested nitrogen both for growth and synthesis of the silk Agrius as a New Experimental Insect 147 proteins, and its diapause occurs at the embryonic stage. By contrast, the hornworm does not make silk, only builds a pupal chamber in soil, and enters diapause at the pupal stage. A comparative study of these developmental and behavioral differences would be most significant. We hope that such comparative studies among Bombyx, Agrius and Manduca will mutually contribute to a better understanding of the insect bio-mechanisms and functions. ACKNOWLEDGMENTS The authors would like to express their sincere thanks to Professor L. M. Riddiford, University of Washington, Seattle, for her advice and critical reading and comments on this paper. We thank Dr. I. Tarumoto and Mr. H. Ishikawa, Sweet Potato Breeding Laboratory, National Agriculture Research Center,Japan, for their assistance in the field collection of Agrius and suggestions on the insect. We also thank Dr. S. Kimura, National Institute of Seri- cultural and Entomological Science, for his encouragement through- out this work. Deep thanks are due to Mrs. F. Karube and Mrs. Y. Yagihashi for their cooperation in rearing the insect. REFERENCES 1 Ahmad IM, Waldbauer GP, Friedman S (1989) A defined artificial diet for the larvae of Manduca sexta. Entomol exp appl 53: 189-191 2 Baumhover AH (1985) Manduca sexta. In “Handbook of Insect Rearing Vol II” Ed by P Singh, RF Moore, Elsevier, Amsterdam, pp 387-400 3 Baumhover AH, Cantelo WW, Hobgod Jr JM, Knott CM, Lam Jr JJ (14977) An improved method for mass rearing the tobacco hornworm. U S Dep Agric, ARS-S-167, pp 1-13 4 Bell RA, Joachim FG (1976) Techniques for rearing labora- tory colonies of tobacco hornworm and pink bollworms. Ann entomol Soc Amer 69: 365-373 5 Hoffman JD, Lawson FR, Yamamoto RT (1966) Tobacco hormworms. In “Insect Colonization and Mass Production” Ed by CN Smith, Academic Press, New York, pp 479-486 6 Horie Y, Watanabe K, Ito T (1966) Nutrition of the silkworm, Bombyx mori XIV Further studies on the requirements for B vitamines. Bull Sericul Exp Sta 20: 393-409 7 Horie Y, Inokuchi T, Watanabe K, Nakasone S, Yanagawa H (1973) Food efficiency and the composition of artificial diets 10 11 12 13 14 15 16 17 18 19 for the silkworm, Bombyx mori. 96, 41-55 (In Japanese) Kerkut GA, Gilbert LI Ed (1985) Comprehensive Insect Phy- siology, Biochemistry and Pharmacology, Vol. 1-13, Pergamon Press, Oxford Nakagawa K, Setokuchi O, Kobayashi M, Oashi K (1986) Ecological studies on the defoliators of sweet potato. II. Sea- sonal occurrence of adults of four major pests, Aedia leucomelas Linne, Agrius convolvuli Linne, Brachmia triannulella (Herrich- Schaffer) and Spodoptera litura Fabricius. Proc Assoc Pl! Prot Kyushu 32: 136-139 Riddiford LM (1985) Hormone action at the cellular level. In “Comprehensive Insect Physiology, Biochemistry and Pharma- cology Vol 8” Ed by GA Kerkut, LI Gilbert, Pergamon Press, Oxford, pp 37-84 Riddiford LM, Hiruma K (1990) Hormonal control of sequen- tial gene expression in lepidopteran epidermis. In “Molting and Metamorphosis” Ed by E Ohnishi, H Ishizaki, Japan Scientific Societies Press, Tokyo and Springer-Verlag, Berlin, pp 207-222 Setokuchi O, Nakagawa K, Kobayashi M (1985) Ecological studies on the defoliators of sweet potato. I. Development process in the larval stage of three major pests, Aedia leucomelas Linne, Agrius convolvuli Linne and Brachmia trian nulella (Herrich-Schaffer). Proc Assoc Pl Prot Kyushu 31: 143-147 Setokuchi O, Nakagawa K, Kobayashi M (1986) Food con- sumption of three major sweet potato defoliators, Aedia leucomelas Linne, Agrius convolvuli Linne and Brachmia trian- nulella (Herrich-Schaffer). Jpn J Appl Ent Zool 30: 93-98 Tazima Y (1978) Preface. In “The Silkworm: An Important Laboratory Tool” Ed by Y Tazima, Kodansha LTD, Tokyo, pp Vii—Vili Tech Bull Sericul Exp Sta No. Truman JW (1992) Developmental neuroethology of insect metamorphosis. J Neurobiol 23: 1404-1422 Truman JW, Riddiford LM (1989) Development of the insect neuroendocrine system. In “Development, Maturation, and Senescence of the Neuroendocrine system: A Comparative Approach” Ed by MP Schreibman, CG Scanes, Academic Press, New York, pp 9-22 Yamamoto RT (1968) Mass rearing of the tobacco hornworm. I. Egg production. J Econ Entomol 61: 170-174 Yamamoto RT (1969) Mass rearing of the tobacco hornworm. II. Larval rearing and pupation. J Econ Entomol 62: 1427- 1431 Zhai Y-J (1977) Preliminary studies on Herse convolvuli Lin- naeus. Acta Ent Sin 20: 352-354 pre! = vA nat rel ; . law ot i> tre Xt dered in bark hag * pornagak pl) eM K SLR LAR BA Pee AG? datvnly, rare tore yt mean in KRONE 5, r Faia ’ ; a dae ee foe aie anti te a) mq eothaite fi veqotcs 1° AT RNERE” Bind ei ; F Huber i SETI) tale wit bart Goal betaid ill ‘ Soe Plast anny OS eam fis Hitan (26 Awe ia we gw " ~ > rr ee ni dat howe ? ; y = J eS wis ' ate aw vA eed a aalial Tas ee = Yaihedin. a i, ee ee ' iit stirs 9 es Aa fulbes! i) WHR Gerla bod aa . rey. Chee @ oti ver thay reine AY qaieporitT Rea MibbeoRbis aa _ OE aha ach 0 seni Diet f WporKa th } Ff nha ss Teas cas Ha =P Naa rah byes nly far reed A ie flicae wit lO eae Myal hia bal: 16 Ta aa COUT UNA Ch ihee P| gna 4 ph . Ggher ‘heh Mae seni pedal , nay Y ah aewie A dA Pade ’ “ i o Lomita agp ; . we syed | = the am iat ~~ 2 latuitot f Ver ii pein i § Tet Wael 7 ee , eaten ‘\ ya (eit Prsigeraett) ty , tii pe x whew OP shai Seidman Ren aah ins 6 A rigger ae He as i Gee iJ AO Repeal bol beg MOM 7 . ' i Pie : area ea , ae ae if é ' Weel Sriset 1 we nie av! ‘ , 1.” eealeekertane ad | Wid hate Vo : j a (ri ny AE AR fe 4 iheel) Wtretpetenes i ia | ; ‘ i ; a ; nuiesrtlevte aad Dae might hiehgge ee | “Pou sw ¥ “pi avTouw saee lly Peat , jatqyiere cobs bil -m | ¥ ‘ ' iu, 42 ‘ nz r i COO rE finhbordee a ad : woud, ' - t ie ri ne Vr Qath a eilie ae Ha Oet Mik ie bin CKAniphe, onl : whi end ay min it ZOOLOGICAL SCIENCE 11: 149-152 (1994) [RAPID COMMUNICATION] © 1994 Zoological Society of Japan Immunohistochemical Study of Ontogeny of Pituitary Prolactin and Growth Hormone Cells in Xenopus laevis Kaoru YAMASHITA and SAKAE KikuyAma! Department of Biology, School of Eduction, Waseda University, Nishiwaseda 1-6-1, Tokyo 169-50, Japan ABSTRACT—The ontogeny of prolactin (PRL) and growth hor- mone (GH) cells in the pars distalis of Xenopus laevis was examined immunohistochemically using anti-bullfrog PRL serum and anti- bullfrog GH serum. Immunoreactive PRL and GH cells first appeared simultaneously at Nieuwkoop and Faber (NF) stage 42, in the anterodorsal region and in the central region of the pars distalis, respectively. Immunoreactive PRL cells increased moderately as metamorphosis progressed. They were distributed mainly in the anterior portion of the pars distalis. Immunoreactive GH cells showed a marked increase in number at NF stages 50-52 and NF stages 62-64 and a slight decrease at the end of metamorphosis. Throughout late premetamorphosis, prometamorphosis and climax, the GH cell number always exceeded the prolactin cell number. GH cells were situated in the posterior portion of the pars distalis. Examination of consecutive sections stained alternately with anti- PRL and anti-GH did not reveal colocalization of PRL and GH at any stage of development. INTRODUCTION Prolactin (PRL) and growth hormone (GH) belong to a family of hormones that are functionally and structurally related [8]. In amphibians, PRL stimulates growth of larval organs such as gills and tail and GH stimulates somatic growth [10]. Amino acid sequences of bullfrog PRL and GH have been determined by direct protein sequencing [12, 24] or deduced from their cDNAs [20, 21]. The two proteins exhibit a considerable sequence homology. Recently, co- localization of PRL and GH in the pituitary of bullfrog larvae at early developmental stages has been reported [9, 13]. Ontogenic differentiation of pituitary GH and/or PRL cells in several species of amphibians has been studied immunohis- tochemically using antisera against GH and/or PRL of mammalian origin [3, 4, 7, 14, 18, 29]. Recently, however, antisera against PRL [25] and GH [11] of amphibian origin have also become available. The antiserum against bullfrog PRL stained PRL cells in adult Rana ridibunda, Pleurodeles waltlii, Ambystoma mexicanum, Xenopus laevis, Bufo vulgar- is and Triturus cristatus [1, 2, 16]. The antiserum against bullfrog GH has been applied to Rana ridibunda [28], Bufo Accepted November 4, 1993 Received October 14, 1993 " To whom requests for reprints should be addressed. vulgaris, Bufo japonicus and Xenopus laevis [17]. However, ontogenic studies of amphibian PRL and GH cells using these antisera have been limited to only two species, namely, Rana catesbeiana [13] and Rana dalmatina [5]. The present study was carried out to study the development of GH and PRL cells in Xenopus larvae, paying particular attention to coexistence of PRL and GH within the same cell. MATERIALS AND METHODS Animals Fertilized eggs of Xenopus laevis were obtained by PN PI b Fic. 1. The consecutive mid-sagittal sections of the pituitary gland of larval Xenopus (NF stages 42) stained with anti-bullfrog RPL serum (a) and anti-bullfrog GH serum (b). ME, median emi- nence; PD, pars distalis; PI, pars intermedia; PN, pars nervosa. Bar, 20 um. 150 K. YAMASHITA AND S. KIKUYAMA injection of 200-400 IU human chorionic gonadotropin (Teikokuzoki Co, Tokyo) into mature male and female animals. The hatched embryos were reared under laboratory conditions until use. The larvae were staged according to Nieuwkoop and Faber (NF) [15]. In addition to larvae, several juvenile animlas (one month after metamorphosis) were used. Immunohistochemistry The whole brains were fixed for 24 hr in Bouin’s solution. After dehydration and embedding in paraplast, serial sagittal sections (5 zm) were cut and mounted on gelatin- The deparaffinized sections were incubated in a After rinsing 3 times coated slides. solution of 0.3% H>O> in methanol for 30 min. with phosphate-buffered saline (PBS) (pH7.2), the slides were treated with normal swine serum (1:20) for 1hr. After washing with PBS, the sections were immunostained by the peroxidase anti-peroxidase (PAP) method [19]. Sections were incubated se- quentially with the following: rabbit anti-bullfrog PRL serum (1:2900) [25] or anti-bullfrog GH serum (1:2000) [12], swine anti-rabbit IgG (1:20) (Dako Japan, Kyoto) for 2 hr and rabbit PAP complex (1:50) (Dako Japan, Kyoto) for 1.5 hr. The section were stained with 10 mg of 3,3’-diaminobenzidine tetrahydrochloride and 0.005% HO, in 100 ml of Tris-HCl buffer (pH 7.6), rinsed with distiled water, stained with 1% methyl green, dehydrated in 100% isopropanol and xylol and mounted in Bioleit. The number of Fic. 2. Mid-sagittal sections of pituitary gland showing localization of immunoreactive PRL (a, c and e) and GH (b, d and f) cells in NF stages 55 (a and b), NF stages 62 (c and d) and juvenile (e and f) Xenopus laevis. Bar, 50 «zm. pars distalis; PI, pars intermedia; PN, pars nervosa. ME, median eminence; PD, Xenopus Prolactin and GH Cells 151 immunoreactive cells with a visible nucleus in mid-sagittal section was counted to use as an index of the cell population [24]. The values from five specimens of each group were expressed as mean + standard error of the mean (SEM). Student’s ¢ test was used for statistical analysis. Control sections were incubated with normal rabbit serum or antisera preadsorbed with an excess of the corresponding antigen instead of the specific antiserum. RESULTS The intensity of the immunoreaction was considerable with the antisera against bullfrog PRL and bullfrog GH. No reaction was observed when sections were incubated with normal rabbit serum instead of the antiserm against bullfrog PRL or bullfrog GH. Immunostaining was completely abolished when sections were incubated with the primary antisera preadsorbed with corresponding antigens (data not shown). At NF stage 42 (embryonic nonfeeding stage), im- munoreactive PRL cells first appeared in the anterodorsal region of the pars distalis (Fig. 1a). Almost simultaneously, GH-immunoreactive cells appeared more caudally than PRL cells (Fig. 1b). There was an apparent segregation of PRL and GH groups. Comparison of two consecutive sections stained with anti-bullfrog PRL and anti-bullfrog GH, respec- tively, did not show colocalization of PRL and GH (Fig. 1, a and b). At the subsequent premetamorphic stages (NF stages 43-54), the increase in PRL cell population was not so marked 150 100 50 Number of immunoreactive cells 42 44 46 48 505254 56 5860626466 Juvenile NF stage Fic. 3. Population of PRL (clear circles) and GH (solid circles) cells detected in mid-sagittal sections of Xenopus larvae at various developmental stages and of juveniles. The values are express- ed as mean of 5 determinations+S.E.M. Significant differ- ences at *P<0.05, **P<0.01 or ***P<0.001 versus preceding stage (Student’s ¢ test). (Fig. 3). During prometamorphosis (NF stages 55-61) and climax (NF stages 62-64), PRL cells increased in number moderately. They were located mainly in the anterior portion of the pars distalis (Fig. 2, a and c). On the other hand, a marked increase of GH cell number was observed during the late premetamorphic period (NF stages 50-52). Thereafter, the population of GH cells became 2-3 times larger than that of PRL cells. Again, the increase of GH cell number occurred during early climax (NF stages 62-64). At the end of metamorphosis, a slight but significant decrease of the number of GH cells was observed (Fig.3). Im- munoreactive GH cells were abundant in the caudal portion of the pars distalis (Fig. 2, b and d). In juveniles, localiza- tion of PRL and GH cells was fundamentally the same as that in larvae. PRL cells were situated mainly in the rostral portion and GH cells were in the caudal portion of the pituitary gland (Fig. 2,c ande). The presence of PRL cells with occasional long processes was noted. No apparent coexistence of PRL and GH was observed in the pituitary gland of larvae at advanced metamorphic stages or of juveniles. DISCUSSION Moriceau-Hay et al. [14] have studied the development of PRL and GH cells in Xenopus tadpoles using antisera against bovine PRL and GH. According to them, PRL cells first appeared at stage 42. This is consistent with the present result. However, they were able to first recognize im- munoreactive GH cells only at stage 44, whereas we detected them at stage 42. This discrepancy may be due to the difference in sensitivity of the antisera used in these two experiments. Moriceau-Hay ef al. [14] stated that the cross-reactivity of the anti-bovine GH serum they used was quite low. This is often the case when antisera against GH of mammalian origin are used for the detection of GH cells in amphibian hypophyses [5]. In this study, we observed that PRL cells were less abundant than GH cells throughout prometamorphosis and climax. This persisted even one month after metamorph- osis. However, with the same antisera as those used in this study, we have confirmed that the population of PRL cells exceeds that of GH cells when the toads become adult [17]. It has been reported that, in Rana esculenta, an anti-ovine GH serum stains both PRL and GH cells [23]. Immunolo- gical studies by Hayashida [6] also demonstrated cross- reactivity between frog PRL and antiserum against primate GH. In the present study, we used antisera against RPL and GH of amphibian origin. The specificity of these antisera has been confirmed by radioimmunoassay [1, 11, 25] and immunoblotting [13, 17, 28]. Using these anti-PRL and anti-GH sera, coexistence of PRL and GH in secretory granules within the same cells in the pituitary gland of embryonic bullfrogs has been demonstrated [9, 13]. In this study, however, colocalization of PRL and GH in Xenopus pituitary cells was not observed. Failure to demonstrate the 152 K. YAMASHITA AND S. KIKUYAMA coexistence of PRL and GH within the same cell of Rana dalmatia pituitaries has also been reported by Guastalla et al. [5]. Recently, we have isolated two molecular forms of both PRL and GH [26, 27] from the Xenopus pituitary gland. These two forms of PRL and GH showed considerable cross-reactivity with the anti-bullfrog PRL and anti-bullfrog GH, respectively. Production of specific antiserum against each hormonal molecule is under way. If these specific antisera become available and are applied to immunohis- tochemistry and radioimmunoassay, more precise informa- tion about PRL and GH cell function in Xenopus larvae could be obtained. ACKNOWLEDGMENTS This work was supported by a grant from Waseda University and Grants-in-Aid from the Japanese Ministry of Education to S. K. The authors express their appliciation to Dr. K. Yamamoto, Dr. K. Kawamura and Dr. T. Kobayashi for their help and advice during the course of the experiment. REFERENCES 1 Andersen AC, Kawamura K, Pelletier G, Kikuyama S, Vaudry H (1989) Gen Comp Endocrinol 72: 299-307 2 Campantico E, Guastalla A (1992) Gen Comp Endocrinol 86: 197-202 3 Eagleson GW, McKeown BA (1978) Cell Tissue Res 189: 53- 66 4 Garcia-Navarro S, Maria MM, Gracia-Navarro F (1988) Gen Comp Endocrinol 69: 188-196 5 Guastalla A, Campantico E, Yamamoto K, Kobayashi T Kikuyama S (1993) Gen Comp Endocrinol 89: 364-377 6 Hayashida T (1970) Gen Comp Endocrinol 15: 432-452 Kar S, Naik DR (1986) Anat Embryol 175: 137-146 8 Kawauchi H, Yasuda A (1988) In “Prolactin Gene Family ana Its Receptors” Ed by K Hoshino, Elsevier, Amsterdam, pp 61- 70 > ~ 10 25 26 27 28 29 Kikuyama S (1994) Can J Zool, in press Kikuyama S, Kawamura K, Tanaka S, Yamamoto K (1993) Int Rev Cytol 145: 105-148 Kobayashi T, Kikuyama S (1991) Gen Comp Endocrinol 82: 14-22 Kobayashi T, Yasuda A, Yamaguchi K, Kawauchi H, Kikuyama S (1991) Biochim Biophys Acta 1078: 383-387 Kobayashi T, Tanaka S, Matsuda K, Yamamoto K, Kikuyama S (1992) Proc Int Symp Amphibian Endocrinol, Tokyo, pp 45 Moriceau-Hay D, Doerr-Schott J, Dubois MP (1979) Gen Comp Endocrinol 39: 322-326 Nieuwkoop PD, Faber J (1956) Normal Table of Xenopus larvis Daudin, North Holland Publ, Amsterdam Olivereau M, Olivereau JM, Kikuyama S, Yamamoto K (1990) Fortschr Zool 38: 371-383 Olivereau M, Olivereau JM, Yamashita K, Matsuda K, Kikuyama S (1993) Cell Tissue Res, 274: 627-630 Remy C, Dubois MP (1973) CR Soc Biol 167: 1581-1584 Sternberger LA (1979) Immunocytochemistry, Wiley, New York, 2nd ed Takahashi N, Yoshihama K, Kikuyama S, Yamamoto K, Wakabayashi K, Kato Y (1990) J Mol Endocrinol 5: 281-287 Takahashi N, Kikuyama S, Gen K, Maruyama O, Kato Y (1992) J Mol Endocrinol 9: 283-289 Thanaka S, Sakai M, Park MK, Kurosumi K (1991) Gen Comp Endocrinol 84: 318-327 Van Kemenade JAM (1974) Fortschr Zool 22: 228 Yasuda A, Yamaguchi K, Kobayashi T, Yamamoto K, Kikuyama S, Kawauchi H (1991) Gen Comp Endocrinol 83: 218-226 Yamamoto K, Kikuyama S (1982) Endocrinol Japon 29: 159-167 Yamashita K, Matsuda K, Hayashi H, Hanaoka Y, Tanaka S, Yamamoto K, Kikuyama S (1993) Gen Comp Endocrinol 91: 307-317 Yamashita K, Yamamoto K, Hayashi H, Kikuyama S (1993) Zool Sci 10 (Suppl): p 132 Yon L, Feuilloley M, Kobayashi T, Pelletier G, Kikuyama S, Vaudry H (1991) Gen Comp Endocrinol 83: 142-151 Zuber M, Dubois M (1975) CR Acad Sci 280 D: 1595-1598 ZOOLOGICAL SCIENCE 11: 153-156 (1994) [RAPID COMMUNICATION] © 1994 Zoological Society of Japan Nucleotide Sequence of the Proton ATPase Beta-Subunit Homologue of the Sea Urchin Hemicentrotus pulcherrimus' Yu-Icui SATOH, TAKESHI SHIMIZU, YUTAKA SENDAI, HIROAKI KINOH and Norio Suzuki? Noto Marine Laboratory, Kanazawa University, Ogi, Uchiura, Ishikawa 927-05, Japan ABSTRACT—A cDNA with 2.3 kb encoding F,-Fo ATP synthase (proton ATPase) beta-subunit homologue was isolated from a testis cDNA library of the sea urchin, Hemicentrotus pulcherrimus. The deduced amino acid sequence consisted of 523 residues which contained a 19-residue amino-terminal signal peptide and a 8-residue glycine-rich consensus sequences. Analysis of poly(A) ‘RNA and/ or total RNA from H. pulcherrimus testis, ovary, unfertilized eggs, and embryos by Northern blot revealed a 2.4 kb RNA. INTRODUCTION A sperm-activating peptide (SAP-I: GLy-Phe-Asp-Leu- Asn-Gly-Gly-Gly-Val-Gly), isolated from the egg jelly of sea urchins, Hemicentrotus pulcherrimus [13] and Strongy- locentrotus purpuratus [3], increases sea urchin sperm respira- tion rate and motility. It induces a Na‘-dependent net proton efflux and raises the intracellular pH [10]. As the result SAP-I stimulates sperm energy metabolism which depends on the oxidation of endogenous phosphatidylcholine [8]. ATP synthesis by oxidative phosphorylation is a multis- tep membrane-located process that occurs in the inner membranes of mitochondria. Fo-F; ATP synthase (proton ATPase) in membranes of mitochondria synthesizes ATP coupled with an electrochemical gradient of protons gener- ated by the electron transfer chain. The enzyme from many different sources have been studied extensively at the molecular biological level [2]. However, no molecular biological study has been made on the enzyme from sperma- tozoa of any kind of animals. In this study, we screened a H. pulcherrimus testis CDNA library with oligonucleotide probes synthesized based on the amino acid sequence of peptide obtained from the protease V8 digest of wheat germ agglutinin (WGA)-binding protein of H. pulcherrimus spermatozoa and isolated a cDNA encoding the beta-subunit homologue of mitochondrial F;-Fo Accepted December 28, 1993 Received December, 1, 1993 " The nucleotide sequence data reported in this paper will appear in the DDBJ, GenBank and EMBL Nucleotide Sequence Databases with the following accession number D17361. 2 To whom correspondence should be addressed. ATP synthase. Here, we report that the cDNA is 2259 bp long and an open reading frame predicts a protein 523 amino acids. MATERIALS AND METHODS Cloning and sequencing of cDNA A cDNA library (4.9X10° pfu) from poly(A)*RNA isolated from growing testes of the sea urchin H. pulcherrimus was con- structed in A gt10 using the cDNA Synthesis System and the cDNA Cloning System A gt10 (Amersham International plc., Amersham, UK). A 220kDa WGA-binding protein was purified from H. pulcherrimus spermatozoa by affinity chromatography on a WGA- Sepharose 4B column as described previously [12], and digested by protease V8. The partial amino acid sequence of a peptide purified from the digest by preparative SDS-gel electrophoresis was deter- mined to be V-S-S-J-D-N-I-F-R-V. The sequence indicated by italics was the same as the conserved sequence found in F,-Fo ATP synthase beta-subunit from various sources. Based on the sequence of the decapeptide, the mixed oligonucleotides (5’- GACACGGAAGATGTTGTCGATGCTGCTGAC-37/5--GACAC- GGAAGATGTTGTCGATAGAGGAGAC-3’) were synthesized and used to screen. Forty-six positive hybridizing clones were isolated from approximately 6 x 10* recombinants. Restriction endo- nuclease mapping of the inserts indicated that five different types of clones had been isolated. The insert of 2.3 kb from one member of the largest group in which fifteen clones belong was subcloned into the plasmid vector Bluescript II KS(+) (Stratagene, La Jolla, CA, USA) for further analysis. Serial deletion mutants of subclones were made according to Yanisch-Perron et al [16]. Nucleotide sequences were determined by the dideoxy chain termination method [11] using the Sequenase Kit (United States Biochemical Co., Cleaveland, OH, USA) and the 7-DEAZA Sequencing Kit (Takara Shuzo Co., Kyoto, Japan) analyzed on DANASIS software (Hitachi Software Engineering Co., Yokohama, Japan). Northern blot analysis Total RNA was prepared from testes, ovaryies, unfertilized eggs, and embryos of H. pulcherrimus by the LiCl method of Cathala etal[1]. Poly(A)*RNA was prep ared by two passage of the total RNA over a column of oligo(dT)-cellulose (Pharmacia LKB Biotech- nology, Uppsala, Sweden). Northern blot analysis was carried out as follows: 2-5 yg of poly(A)*RNA or total RNA was denatured 154 Y. Satou, T. SHimizu et al. 5 'CGTGACCCCTGGAAGAATT TCACATCGCCATGT T TAGCAGGGT TGCAAAGACGAGT TT T TCGGCCGT AAGGGCT GCAAAAT CACAAT TT ¥ A F bs BONS oa od TCACACTCATTATCACAACAGACGAGT AAAACATGGGT ACCAGCAGCAACT TGT AGCAAAAGATCATATGCT GCT GAGGCAAAGACGT CG S HS RES Qa 1) ES SKE oT WV Ps OAR BA POPU! ES2RIK BIR: SRE LYAR AMEE AGRIE SPAS AKGR au ees GCAGCCCCAGTTTCGGGTCAGATCGTAGCTGTCATTGGAGCTGTCGTCGACGTT CAGT TCGAGGATGACCT CCCACCCAT TCTCAATGCC A A P V S GQ TT VA V YG A OVEN DOV ORE es Dee Dee eelee eee es TTGGAGGT T CAGGGAAGGACAT CCAGGCTGGTGT TGGAAGT TGCACAGCATCT TGGTGAGAACACAGT CAGGACAAT T GCCATGGACGGT bE Me GQ aG.SR Ths SoRealikV ol cE VAUD pat AL Bair Ean et eligi Peel nese ACAGAAGGT CTGATCCGAGGCCAGAAGTGCGT T GACACTGGCT CCCCCATCAGCAT CCCCGT CGGCCCCGAGACGCT GGGACGCATCATC TE GE A ORG GP Ve MDLTOMGHESHE Bale Sielie AVG (Ga oP Es ils Ue) Glace AATGTCATTGGTGAACCCAT T GACGAGAGAGGACCAAT T GGAACAGACAGGAGAT CAGCAATCCAT GCAGAAGCT CCAGAGT T TGTAGAG NV IT GE P Y D ER GP IG 7 D R R S A DOH A ED AD Pp OE ee ATGAGTGTAAACCAGGAAATCCTTGT TACTGGAATCAAGGT TGTAGATCTACTCGCCCCATACGCCAAGGGAGGAAAGAT TGGTCTGTTT Me 8S-Vo ON Qi TE SCT LIV Se Ge Sei Ve DL ESA PY MAVIK Ge (GU KU eS Grea GGCGGTGCTGGTGTAGGAAAGACT GT ACTCATCATGGAGCTGAT T AACAACGT AGCCAAGGCCCACGGAGGT T ACTCTGTGTT 1GCCGGT GoGUA G.VeG KTV & IT MM EL EON NPV APS KOA SH, SGirG SV eee GTAGGAGAGAGGACCCGTGAGGGT AACGATCT IT TACCATGAGATGAT I GAAGGAGGT GTCATCT CCCTCAAGGATGACACAT CAAAGGTA VG ER T R E GN D EL Y H € M I E°SG) SG) Ve oe [SOL SKyy DeDP See sSeeaKGee GCGTTGGTGTACGGACAGAT GAACGAGCCT CCCGGCGCCCGT GCCCGT GT CGCCT T GACCGGACT GACCGT T GCCGAAT ACT TCCGTGAC AE OV OY" GeQh OM ONE E SP SP Ge AD Roa UR VO ACE Gok) TV ALE Vee ee CAAGAGGGACAGGATGTGCTGCTCTTCATTGACAACATCT TCCGCT T CACACAGGCTGGAT CAGAGGT ATCT GCTCTGCTGGGACGTATC Qe ES sGecQeiDo eV mlboalorer) SnD INE IE WEYMORE EF TOF SA Gus. Vey 2Vig SevAi eG eeee CCATCTGCCGT AGGAT ACCAGCCAACCCTGGCCACTGACATGGGT ACT AJ GCAGGAGCGTAT TACCACCACCAAGAAGGGAT CCAT CACT PoS- Aye G Yo QuPood -cliocA Ta eiDieMaeGe T MOPS ORS Pee Te Ken Kee Grasmere TCCGT ACAGGCCAT CT ACGTGCCTGCT GACGATCTCACTGACCCTGCCCCTGCCACCACCT T CGCCCATT T GGACGCCACCACTGTGCTG $V QA or Y VPA OD DLT (DOP OA OP) AY TetniigiFa AV shliles Die Alysia in amma TCCCGTGGTATCGCTGAGCTGGGTATCTACCCTGCTGTGGATCCTCTGGAT TCCTCCTCCCGTATCATGGACCCCAACGT CGT CGGAGAG SOR Ge TO APOE LPG MIS yooR AVEO IP LE: (DEES SHESEERMSIG MADER Navies Veen Gee CGTCACTACAGCATCGCTCGTGGAGT ACAGAAAATCCT TCAGGACAACAAGACCCTGCAGGACATCATCGCCATCT TGGGTATGGACGAG RH OY -SoT A SReG Wo@rKrybijE iQerbs NK of UL 6D UL A IG TTGTCTGAGGACGACAAACT GACCGT GT CCCGAGCCAGGAAGAT CCAGAGGT T CT TGTCCCAACCCT T CCAGGT T GCCGAGGTCT TCACC L's ED DK LE T V SRA R K I Q ReeShestenSiia) MP IRAQ) \VayAG RE aman el GGCAGTCCAGGCAAGCTCGTCT CAATGGCGGAGACCATCGATGGAT T CGAGT CCAT TAT CAAGGGCGAGT GCGACCAT CT ACCAGAGATT GieiS oP! aiGi Kem VerST MFA ET I eDPG? E ES ll Ko "Gh (G10 his pan GCTTTCTACATGGTAGGCAACAT TCAAGATGTCAAGGATAAGGCCGACAGGCT CGCAGAAGAACTATCATAAAT TATCCCCCCTCTCCCA Ay Rc OM) OV Gig INE Ta) Deine Ky Die ai KcmavA Gin oti i EA Gin Eo Es [eee S es AACAAT GAAGT T TAGAGCT GGCATGGCT ACGGGT CAGAGACACCCCTCTTGATTGTTGTTAT TCAGGGCT AGT TGTCT AACACTACCCGT GCCTGGGCCCAAAGAATT TATGT TCAGAGTTATAACT TATATCAAGAT TGTTT TCTAAAT TGTAAT TGTGAAAAAT TGAGAGCAAGGGAA TTCCAACCTAGCGTACTTTTGTCATATGAATCTGTCGTTTTCCTICTITTTT1 1 TGCTTGTTATCCACCATAGAT TGTAAATGCACAAACA GCTTGGCAAAGTTTGTAAATT TGATCATAACCAAT TATCCCAAT T TAAGGCAGT ACCT TT AGCACAT TGGTGTGTCACCGATGCCTGATT TCATGTTTATTGTCTGATCTGATCT TACAAGAAAT 1 GGCCGATGTCCAAACATT TCCAATGTAGATATAGACATATATCTTCACT IGATT TCTGTGTAGAGCCGT T CACGTATGACAGATGAT TGGCATTTATTT TGAATGGATGTTT TAGAGCTT TACTGAACCCAGT TGCGATTGTGA TTTCCTTGTGTGAACAGAATCGCAACTGGCCT T GAAAAAGAAAAACAAGT GT AT TAAAAAT TAT T GGAAGGT TCAAGAACCAAAAAAAAA AAAAAAAAAA 3' Fic. 1. Nucleotide sequence and deduced amino acid sequence of the 2.3 kb insert. The shadowed box indicates predicted signal peptide sequence and the open box denotes glycine-rich consensus sequence. The amino acid sequence deisgnated by an underline is the same as partial sequence of the decapeptide used for synthesis of oligonucleotide probes. * denotes start or stop codon. Sea Urchin Proton ATPase Beta-Subunit 155 with 2.1 M formaldehyde, electrophoresed on a 1% agarose gel in the presence of 2.2 M formaldehyde, and transferred onto a Hybond-N- membrane. The RNA on the membrane was hybridized to the random-primed ECL labelled (Amersham International ple., Amer- sham, UK) or random-primed [a-**P]dCTP-labelled 2.3 kb cDNA insert at 65°C for 18 hr. The membrane was washed with 0.5 x SSC and 0.1% SDS at 65°C for 30min. The size of the RNA was estimated using a 0.24-9.5 kb RNA Ladder (GIBCO BRL, Baith- ersburg, MD, USA) as a marker. RESULTS AND DISCUSSION The 2.3 kb insert contained DNA sequences encoding an open reading frame of 523 amino acids including I-D-N-I-F-R 10 20 which is the same as the partial sequence of the peptide used for synthesis of oligonucleotide probes (Fig. 1). The de- duced amino acid sequence suggests that the protein contains a 19-residue amito terminal signal peptide which has the potential to form amphipathic helix being characteristic of mitochondrial signal peptide sequence [5] and a 8-residue (residues 201-208) glycine-rich consensus sequence (G-X-X- X-X-G-K-T/S) found in the F;-Fo ATP synthase beta- subunit, adenylate kinase, p21 ras protein, and other nu- cleotide-binding proteins [14]. The deduced amino acid sequence has 68% homology with those of chloroplast F,-Fo ATP synthase beta-subunits and 85% with those of mitochondrial F;-Fp ATP synthase beta-subunits from various 30 40 50 60 Spermatozoa (sea urchin) MFSRVAKTSFSAVRAAKSQFSHSLSQQTSKTWVPAAT CSKRSY AAEAKT SA--APVSGQI VAVIGAVVDV Mitochondria (human) MLGFVG. ..AAPA.GALRRLTPSASLPPA.LLLRAA.T.VHPV.D...QTSP.PKAGAAT.R........... Mitochondria (rat) MLSLVG. ..SA.A.GALRGLNPLAALPQAHLLLRTA. .GVHPA.D...QSSAAPKAGTAT............. Chloroplast (potato) Chloroplast (spinach) MRINPTTSGS.VS.VE--KKNL.R..KI..P.L.. MRINPTTSDPGVS.LE--KKNL.R.AQI..P.LN. 70 80 90 100 110 120 130 140 150 QF-EDDLPPILNALEVQGR----TS---RLVLEVAQHLGENTVRT I AMDGTEGLIRGQKCVDTGSPISIPVGPETLGRI INVIGEPIDER SDEG reds aussesetyore one =| Soa aes ane oS Sica vttronpaten seen Vi resi SEtAw ic Ke, sence acess Megas a3s-o ees Brig OE Gis, opened } et A iw > ] \ dd af EPG iG wos ‘ P| ged ad ap i> Ua Peete (28 Prete “te se ‘wi rotten Bohanseth | te anh ott haw abek ba = oJ ra wpuree shontd-ae oe letters aad alinth== ia reprodertian Tone Ky apie a 7? AIOE Beneath doh may ' Be ‘Bh ah, rata, ‘Tiay Jawa end gregh. aul apiwerk af high quake a PAPApPCibe Pea ie a0 (ieee © ted 1O: Signenietion “i 4 pis Faw 1 Solid tif - be npn Tad F _ iw (ewer ight cong pier sai piu 7 ‘ e ==: :) Ae Ti Twa (Cn: ¢ = ? ob eas Ft 67 « aomtpenind: by 8 Te tepiit: Toe sovdral § hy crweh oj Tinley Suloant Lo roa este without ff ? “ie «= a — Development Published Bimonthly by the Japanese Society of Developmental Biologists Gr owth & Differentiati on Distributed by Business Center for Academic Societies Japan, Academic Press, Inc. Papers in Vol. 36, No. 1. (February 1994) 1. REVIEW: T. Muramatsu: The Midkine Family of Growth/Differentiation Factors 2. H. Yasuo and N. Satoh: An Ascidian Homolog of the Mouse Brachyury (T) Gene is Expressed Excusively in Notochord Cells at the Fate Restricted Stage 3. I. Nagata and N. Nakatsuji: Migration Behavior of Granule Cell Neurons in Cerebellar Cultures. I. A PKH26 Labeling Study in Microexplant and Organotypic Cultures 4. K. Ono, N. Nakatsuji and I. Nagata: Migration Behavior or Granule Cell Neurons in Cerebellar Cultures. II. An Electron Microscopic Study 5. T. Miya, K. W. Makabe and N. Satoh: Expression of a Gene for Major Mitochondrial Protein, ADP/ ATP Translocase, during Embryogenesis in the Ascidian Halocynthia roretzi 6. U.A. O. Heinlein, S. Wallat, A, Senftleben and L. Lemaire: Male Germ Cell-Expressed Mouse Gene TAZ&3 Encodes a Putative, Cysteine-Rich Transmembrane Protein (Cyritestin) Sharing Homologies with Snake Toxins and Sperm-Egg Fusion Proteins 7. A. Nagafuchi and S. Tsukita: The Loss of the Expression of a Catenin, the 102 kD Cadherin-Associated Protein, in Central Nervous Tissues during Development 8. E.M. del Pino, I. Alcocer and H. Grunz: Urea is Necessary for the Culture of Embryos of the Marsupial Frog Gastrotheca riobambae, and is Tolerated by Embryos of the Aquatic Frog Xenopus laevis 9. §S. Fukada, N. Sakai, S. Adachi and Y. Nagahama: Steroidogenesis in the Ovarian Follicle of Medaka (Oryzias latipes, a Daily Spawner) during Oocyte Maturation 10. A. H. Wikramanayake and W. H. Clark, Jr.: Two Extracellular Matrices from Oocytes of the Marine Shrimp Sicyonia ingentis That Idenpendently Mediate Only Primary or Secondary Sperm Binding 11. S. Furuya, Y. Kamata and I. Yasumasu: ADP-Ribosylation of Histones in Nuclei Isolated from Embryos of the Sea Urchin, Hemicentrotus pulcherrimus 12. C. Cirotto, L. Barberini and I. Arangi: The Wavy Erythropoiesis of Developing Chick Embryos. Isolation of Each Wave by a Differential Lysis and Identification of the Constituent Erythroid Types Development, Growth and Differentiation (ISSN 0012-1592) is published bimonthly by The Japanese Society of Developmental Biologists. Annual subscription for Vol. 35 1993 U.S. $ 191,00, U.S. and Canada; U. S. $ 211,00, all other countries except Japan. All prices include postage, handling and air speed delivery except Japan. Second class postage paid at Jamaica, N.Y. 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 32887-4900, 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 32887-4900, U.S. A. In Japan: Send nonmember subscription orders and notices of change of address to Business Center for Academic Societies Japan, 16-9, Honkomagome 5-chome, Bunkyo-ku, Tokyo 113, Japan. Send inquiries about membership to Business Center for Academic Societies Japan, 16-9, Honkomagome 5-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. |N_USA AND EUROPE Narishige. The complete range for micromanipulation BRANCHES NOW OPEN or over 30 years Narishige have been developing their This large range includes micro- manipulators, microelectrode ~ pullers and micro- ¥ tel . forges, micro- EF q injectors, microgrinders Shown here is a small \ )} and stereotaxic selection of instruments from ’ instruments. the extensive Narishige range. MF-83 Patch Clamp Forge MM-31 Miniature Micromanipulator SN-3 Stereotaxic Instrument extensive range of precision - — : instruments for Physiology, Ne US-1 Miniature Pharmacology, Zoology and IM4 Microinjector Stand Psychology research. Narishige are pleased to announce that repair facilities have been opened in Europe and the USA. EG-6 Microgrinder This enables us to provide an improved service to the many users of our quality products throughout the world, including the upgrading of previous types of water MX-1 3 Axis filled hydraulic Please contact us for PN-3 Microelectrode Puller * Micromanipulator micromanipulators. more information. REPAIRS, AFTER SALES SERVICE AND TECHNICAL SUPPORT Narishige Scientific Instrument Laboratory 9-28 Kasuya 4-Chrome, Setagaya-Ku, Tokyo 157, Japan. Tel: +81 (0) 3 3308 8233 Fax: +81 (0) 3 3308 2005 WoO Narishige International London Branch Unit 7 Willow Business Park, Willow Way, London SE26 4QP U.K. Tel: +44 (0) 81 699 9696 Fax: +44 (0) 81 291 9678 EUR W«U.S. Narishige International Inc 104 Glen Cove Avenue, Sea Cliff, New York 11579 U.S.A. Tel: +1 (516) 759 6167 Fax: +1 (516) 759 6138 (Contents continued from back cover) Experimental Animal Kiguchi, K., M. Shimoda: The sweet potato hornworm, Agrius convolvuli, as a new experimental insect: Con- tinuous rearing using artificial diets ................ 143 /NMOOWINGS 100 Sr 161 ZOOLOGICAL SCIENCE VOLUME 11 NUMBER 1 FEBRUARY 1994 CONTENTS Obituary © osc eds jos ve cc atone ote eee eee ee 1 subunit homologue of the sea urchin Hemicentrotus REVIEWS pulcherrimus (RAPID COMMUNICATION) sleleiehy 153 Hanke, W., W. Kloas: Hormonal regulation of osmo- mineral content in amphibia.......................... Nassel, D. R.,. E. Bayraktaroglu, H. Dircksen: Neuropeptides in neurosecretory and efferent neural systems of insect thoracic and abdominal ganglia ORIGINAL PAPERS Physiology Takahashi, T., O. Matsushima, F. Morishita, M. Fujimo- to, T. Ikeda, H. Minakata, K. Nomoto: A Myomodu- lin-CARP-related peptide isolated from a polychaete annelid Perinereis vancaurica Sugimoto, M., T. Kawamura, R. Fujii: Changes in the responsiveness of melanophores to electrical nervous stimulation after prolonged background adaptation in the medaka, Oryzias latipes Wilder, M. N., T. Okumura, Y. Suzuki, N. Fusetani, K. Aida: Vitellogenin production induced by eyestalk ablaton in juvenile giant freshwater prawn Macro- brachium rosenbergii and trial methyl farnesoate admi- nistration Cell and Molecular Biology Meyer-Rochow V. B., Y. Ishihara, J. R. Ingram: Cytochemical and Histological details of muscle fibers in the southern smelt Retropinna retropinna (Pisces; Galaxioidei) Genetics Patil, J. G., V. Wong, H. W. Khoo: Assessment of pMTL construct for detection in vivo of luciferase expression and fate of the transgene in the zebrafish, Brachydanio rerio Developmental Biology Furukawa, T., Y. Maeda: K252a, a potent inhibitor of protein kinases, promotes the transition of Dictyoste- lium cells from growth to differentiation Iwamatsu, T., S. Nakashima, K. Onitake, A. Matsuhisa, Y. Nagahama: Regional differences in granulosa cells of prevulatory medaka follicles Kanno, Y., S. Koike, T. Noumura: Immunohistoche- mical localizations of epidermal growth factor in the developing rat gonads Satoh, Y., T. Shimizu, Y. Sendai, H. Kinoh, N. Suzuki: Nucleotide sequence of the proton ATPase beta- Kanbayashi, H., Y. Fujita, K. Yamasu, T. Suyemitsu, K. Ishihara: Local change of an exogastrula-inducing peptide (EGIP) in the pluteus larva of the sea urchin Anthocidaris crassispina (RAPID COMMUNICA- TION) Reproductive Biology Hosokawa, K., Y. D. Noda: The acrosome reaction and fertilization in the bivalve, Laternula limicola, in refer- ence to sperm penetration from the posterior region of the mid-piece Endocrinology Iga, C., I. Koshimizu, S. Takahashi, Y. Kobayashi: Experimental manipulation of pituitary hemorrhage induced by intraperitoneal injection of a hypertonic solution in mice Yamashita, K., S. Kikuyama: Immunohistochemical study of ontogeny of pituitary prolactin and growth hormone cells in Xenopus laevis (RAPID COM- MUNICATION). -..0.5..66.0 00-200 eee Asahina, M., H. Fugo, S. Takeda: LEcdysteroid syn- thesis in dissociated cells of the prothoracic gland of the silkworm, Bombyx mori 101 149 107 Behavior Biology Naruse, M., T. Oishi: Effects of light and food as zeitgebers on locomotor activity rhythms in the loach, Misgurnus anguillicaudatus Suzuki, H., T. Sekiguchi, A.. Yamada, A. Mizukami: Sensory preconditioning in the terrestrial mollusk, Limax flavus Environmental Biology and Ecology Ohdachi, S.: Growth, metamorphosis and gape-limited cannibalism and predation on tadpoles in larvae of salamanders Hynobius retardatus 127 Lawrence J. M., M. Bryne: Allocation of resources to body components in Heliocidaris erythrogramma and Heliocidaris tuberculata (Echinodermata: Echinoidea) Systematics and Taxonomy Ishikawa, K.: Two new species of the genus Holaspulus (Acarina: Gamasida: Parholaspidae) from the Ryukyu Islands, Japan 139 (Contents continued on inside back cover) INDEXED IN: Current Contents/LS and AB & ES, Science Citation Index, ISI Online Database, CABS Database, INFOBIB Issued on February 15 Front cover designed by Saori Yasutomi Printed by Daigaku Letterpress Co., Ltd., Hiroshima, Japan a ISSN 0289-0003 An_International OOLOGICAL ah ZOOLOGICAL SCIENCE The Official Journal of the Zoological Society of Japan Editors-in-Chief: The Zoological Society of Japan: Seiichiro Kawashima (Tokyo) Toshin-building, Hongo 2—27-2, Bunkyo-ku, Tsuneo Yamaguchi (Okayama) Tokyo 113, Japan. Phone 03-3814-5461 Division Editor: Fax 03-3814-5352 Shunsuke Mawatari (Sapporo) Officers: Yoshitaka Nagahama (Okazaki) Takashi Obinata (Chiba) Suguru Ohta (Tokyo) Noriyuki Satoh (Kyoto) President: Hideo Mohri (Chiba) Secretary: Takao Mori (Tokyo) Treasurer: Makoto Okuno (Tokyo) : , Librarian: Masatsune Takeda (Tokyo) Assistant Editors: Auditors: Hideshi Kobayashi (Tokyo) Akiyoshi Niida (Okayama) Hiromichi Morita (Fukuoka) Masaki Sakai (Okayama) Sumio Takahashi (Okayama) Editorial Board: Kiyoshi Aoki (Tokyo) Makoto Asashima (Tokyo) Howard A. Bern (Berkeley) Walter Bock (New York) Yoshihiko Chiba (Yamaguchi) Aubrey Gorbman (Seattle) Horst Gurnz (Essen) Robert B. Hill (Kingston) Yukio Hiramoto (Chiba) Tetsuya Hirano (Tokyo) Motonori Hoshi (Tokyo) Susumu Ishii (Tokyo) Hajime Ishikawa (Tokyo) Sakae Kikuyama (Tokyo) Makoto Kobayashi (Higashi-Hiroshima) Kiyoaki Kuwasawa (Tokyo) John M. Lawrence (Tampa) Koscak Maruyama (Chiba) Roger Milkman (Iowa) Kazuo Moriwaki (Mishima) Richard S. Nishioka (Berkeley) Chitaru Oguro (Toyama) Masukichi Okada (Tsukuba) Andreas Oksche (Giesen) Hiraku Shimada (Higashi-Hiroshima) Yoshihisa Shirayama Tokyo) Takuji Takeuchi (Sendai) Ryouzo Yamagimachi (Honolulu) ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and rapid 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 Jasponenses (1897-1983), the former official journals of the Zoological Soceity of Japan. An annual volume consists of six regular numbers and one supplement (abstracts of papers presented at the annual meeting of the Zoological Society of Japan) of more than 850 pages. The regular numbers appear bimonthly. MANUSCRIPTS OFFERED FOR CONSIDERATION AND CORRESPONDENCE CONCERNING EDITORIAL MATTERS should be sent to: Dr. Tsuneo Yamaguchi, Editor-in-Chief, Zoological Science, Department of Biology, Faculty of Science, Okayama University, Okayama 700, 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, 5—16—9 Honkomagome, Bunkyo-ku, Tokyo 113. Subscriptions outside Japan should be ordered from the sole agent, VSP, Godfried van Seystlaan 47, 3703 BR 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. (©) Copyright 1994 by the Zoological Society of Japan. In the U.S.A., authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by [copyright owner’s name], provided that designated fees are paid directory to Copyright Clearance Center. For those organizations that have been granted a photocopy license by CCC, a separate system of payment has been arranged. Copyright Clearance Center, Inc. 27 Congress St., Salem, MA, U.S.A. (Phone 508—744— 3350; Fax 508-741-2318). Publication of Zoological Science has been supported in part by a Grant-in-Aid for Publication of Scientific Research Results from the Ministry of Education, Science and Culture, Japan. ZOOLOGICAL SCIENCE 11: 167-174 (1994) © 1994 Zoological Society of Japan REVIEW Insulin-like Growth Factors: Growth, Transgenes and Imprinting ANDREW Warp!, PAR BIERKE~*, EvA PETTERSSON”, WILHELM ENGSTROM2* "Cancer Research Campaign Growth Factors, Department of Zoology, University of Oxford, South Parks Road, Oxford OX13PS, UK and *Department of Pathology Faculty of Veterinary Medicine Swedish University of Agricultural Sciences S-750 07 Uppsala, Sweden *National Defenee Research Institute, S-90182 Umea Sweden I BACKGROUND The idea of integrating stable gene mutations into the germ line has long fascinated developmental biologists. Ex- perimental gene modifications that influence subsequent de- velopmental patterns have therefore been at the top of the list of priorities. The field of experimental embryology took off in the early 1960s when Krystof Tarkowski [86] and Beatrice Mintz [59] succeeded in fusing genetically distinct 8-cell embryos into stable chimaeras. In 1968 Richard Gardner [28] was able to show that by moving the inner cell mass from one embryo to another a chimaera could be formed. But it was not until 1974 when Jaenisch and Mintz [41] could report the first deliberate genetic modification of the mouse embryo. By infecting embryos at the blastocyst stage with SV40-virus they were able to demonstrate a stable integration of viral DNA in the live born progeny. This breakthrough was rapidly followed by Moloney murine leukemia virus mediated germ line transfer of foreign nucleic acid sequences [42]. Since that time the field has exploded, it appears pertinent to briefly discuss some key experimental landmarks during the last 20-year period. It was in 1980 [30] that the pronuclear microinjection technique was devised and used to demonstrate the possibility of inserting foreign DNA from almost any source into the murine embryo. This technique can be applied to almost any animal species. The next important step forward was the demonstration that integrated foreign DNA can also be expressed. In 1981 Palmiter er al. [64] were able to ligate the 5’ regulatory sequence of the murine metallothionein (MT1) gene with the core sequence of the herpes virus thymidine kinase gene. This transgene was injected into pronuclei and properly integrated. In the adult progeny, the herpes thymi- dine kinase gene was expressed in a pattern as expected from the MT1 gene with abundant expression in kidney and liver. Furthermore, it was concluded from this and subsequent studies that the actual integration site plays an inferior role in determining the level of expression. It is rather the cis- Received February 10, 1994 * Reprint requests should be addressed to Dr. W. Engstrém. acting regulatory sequences in and adjacent to genes that determine spatial and temporal transcription patterns [32, 63]. There are however important exceptions to this rule. In the case of the beta globin gene the absence of key enhancer elements makes it relatively more dependent on the integration site for adequate expression [13]. This implies that the interplay between structural regula- tory elements is imperative in determining stage- and tissue- specific expression. If an isolated promoter sequence is fused with a heterologous coding region and integrated in the germ line, the expression pattern will in many cases differ from that of an integrated genomic clone containing the regulatory and structural elements in intact order. There are also transacting factors that affect expression of inte- grated foreign genes. The presence or absence of such transactivators indeed differ between species. When a hu- man gene is integrated into transgenic mice, the transgene may therefore be expressed in organs that do normally not express the murine counterpart, but which are sites of ex- pression in man. Foreign genes normally integrate as head to tail con- catamers into one unique site in the mouse genome. However at this single site a considerable number (from one to 50 copies) of transgenes can be integrated. There is ample evidence that this large scale integration is not always a straight forward process. It predisposes delays and recom- binations leading to mosaic patterns. It can also lead to strange structural alterations of the mouse host genome including deletions, duplications, translocations and inter- spersion of islands of genomic DNA within the transgene sequences. The disruption of endogenous gene function upon transgene integration is also an alarmingly frequent event [58] Il METHODOLOGY 1. Pronuclear injection Broadly, two approaches are currently employed in generating transgenic mice, leading to either over-expression or targeted alteration of the gene of interest. In the first approach, exogenous copies of the gene are incorporated into 168 A. Warp, P. BIERKE ef al. the genome with the aim of elevating its level of expression in vivo. This is usually achieved by injecting fertilised eggs, collected approximately 12 hr after mating has taken place, with linear cloned DNA fragments [39]. The injected embryos are then introduced into the uterus of receptive, pseudopregnant females, so that normal development can proceed. Liveborn mice are tested for the presence of the transgene by assaying biopsy material using DNA (Southern) blotting or polymerase chain reaction (PCR) techniques. Although the proportion of transgenic offspring which result is very much dependant on the nature of the injected DNA construct, frequencies of transgenesis over 20% can be achieved, representing one major advantage of this approach. This is offset against the principle disadvantage of this scheme, in that incorporation of the transgene into the genome occurs at essentially randomly determined positions. Since expression of the transgene can be influenced by sequences surrouding the integration site (so called position effects) and since transgene insertion can disrupt the function of a resident gene, the phenotype associated with any given DNA construct must be qualified by studying lines of trans- genic mice representing at least two independent integration events. 2. Embryonic stem cells In the second approach to producing transgenic mice, more recently developed technology allows the targeted alteration of specific endogenous genes by homologous re- combination in embryonic stem (ES) cells. ES cells are derived from blastocyst embryos and can be maintained in culture for many generations before being microinjected into host blastocysts and returned to pseudopregnant females. Subsequently, they can colonise all tissues of the resulting animals, which develop to form chimeric mice. Chimeric animals in which the ES cells have given rise to germ cells can be bred to produce transgenic offspring which are entirely ES cell derived [40, 43]. Compared with the one cell embryos used for pronuclear microinjection, ES cells represent a more abundant source of starting material which can be manipu- lated in vitro over a much longer time period, greatly increasing the scope for the manipulation of the genome. This is essential to the success of the gene targeting scheme. Gene targeting requires a construct comprising sequences of the target gene, bearing a mutation which will alter or ablate its function, together with at least one selectable marker gene (such as neo’, which confers resistance to the drug G418). Often the neo’ gene is positioned within the construct such that it disrupts or replaces part of the coding region of the target gene. This construct can be introduced rapidly into many thousands of ES cells by electroporation. These cells are subsequently grown in the presence of G418 to identify resistant colonies which represent those cells in which the targeting construct was stably incorporated into the genome. The sequences of the target gene which flank the neo’ gene provide regions of homology which allows recombination to take place such that the cloned DNA can replace the target gene following its introduction into embryonic stem cells. However, since homologous recombination events are usual- ly much less frequent than random integration events, the G418 resistant ES cell colonies must be screened to establish which should be used to generate chimeras. This screening process is one of the rate limiting steps of this transgenic route since the homologous recombination events typically repre- sent less than 0.1% of all stable transformants. However, this figure is constantly being improved with the development of methods to enrich for homologous recombination over random integration. These improvements include the use of a second selectable marker which must be lost from the targeting vector during homologous recombination to allow cell survival [53] and the use of DNA within the targeting construct which derives from the same mouse strain as the host ES cells [87, 88]. 3. Strategies for increased expression Perhaps the crudest overexpression strategy relies upon the use of an entire genomic sequence which contains the full coding sequence with cognate upstream regulatory elements as the transgene. This approach has had a practical advan- tage in compensating for defects in mutant animals lacking a specific gene. A classical experiment succeeded in restoring fertility in hypogonadal mice by introducing the gonadotropin releasing factor gene [55]. In addition to its simplicity, this method can be advantageous where it is important to repro- duce the expression patterns of the gene of interest. For instance following cloning of the sex-determining gene Sry, its own regulatory sequences were relied upon to create trans- genic mice which were XX males [46]. In this case accumu- lated molecular and genetic evidence indicated that the testis determining gene was normally expressed transiently in spe- cific cells in the developing gonad [47]. A more widely appreciated approach is based upon the fusion of a coding sequence with a well characterized promo- ter sequence that operates in a multitude of tissues. Com- monly used promoters that induce widespread expression are the actin, H2 and SV40 promoters. Other promoter se- quences induce expression in a more limited spectrum of cells and tissues. The mouse Mammary Tumour Virus Long Terminal Repeat (MMTV-LTR) operates mainly in secretory epithelial cells, and as a result transgenes containing this promoter sequence are expressed specifically in this cell type. When the MMTV-LTR was fused with the Granulocyte Macrophage Colony Stimulating Factor gene a strange subset of phenotypes were observed amongst the offspring [48]. Other promoters that allow cell specific expression include the Immunoglobulin promoter. When this sequence was fused with the interleukin-6 (IL-6) gene, mice harbouring this transgene developed plasmocytosis, with IL-6 expressing B-cells circulating throughout the body giving rise to high circulating levels of IL-6 protein [82]. The options can be further narrowed by using tissue specific promoters. The first of its kind to be utilized for this purpose was the rat insulin promoter (RIP) which is only Insulin-like Growth Factors 169 active in pancreatic beta cells. By fusing RIP with a nerve growth factor (NGF) coding sequence a unique pattern of hyperinnervation of the islets of Langerhans was achieved [20]. The great advantage of this approach is that the phenotypic effects on a single organ is enabled. Now a plethora of organ specific promoters are used for the purpose of site directed expression of transgenes. The bovine kera- tin 10 promoter directs the expression of transgenes to the supra basal layer of the skin and to the forestomach [2]. Gene expression can be directed to the mammary gland by using either the Whey Acidic Protein (WAP) promoter [35] or the beta lactoglobulin promoter [91]. The latter promo- ter limits expression to the secretory epithelium of the mammary gland. Moreover, induction of beta lactoglobulin expression coincides with that of beta casein in the tissue, indicating that the pattern of expression is determined by the differentiated state of the mammary cells [35, 91]. Using this promoter, protein products of the transgene can be secreted into milk and form up to 10% of the total milk protein or 30% of whey protein [1]. To achieve kidney specific expression, much attention has been devoted to the three murine renin genes Ren-IC, Ren-ld and Ren-2. However, the three renin genes exhibit distinct expression profiles at a number of extra renal sites [27, 75, 76]. III TRANGENIC STUDIES INVOLVING INSULIN AND THE INSULIN-LIKE GROWTH FACTORS Insulin and the insulin-like growth factors (IGF-I and IGF-II) are structurally related and share affinities for the same set of trans-membrane receptors (reviewed in [72]). Each factor exhibits a preference for the appropriately named receptor (insulin, IGF type 1 and IGF type 2) but can interact with the others. This family of growth factors and receptors represent probably the most intensively studied using mouse transgenesis; insulin, IGF-I and IGF-II have all been analy- sed using overexpression strategies and IGF-I, IGF-II and the IGF type 1 receptor have each been the subject of gene targeting experiments to disrupt their function in mice. These experiments have revealed much about the in vivo functions of these factors and collectively they amply demon- strate the effectiveness of applying transgenic techniques to the analysis of mammalian growth factors. 1. Insulin In several reports a common strategy has been adopted to elevate insulin expression at the natural sites of synthesis [11, 26, 54, 74]. By introducing copies of the human insulin gene into mice it was possible to distinguish mRNA and protein of transgene and endogenous origin while harnessing the cognate promoter to ensure expression was obtained in the pancreatic islets. The efficacy of this approach is clear since in all cases transgene expression was restricted to the pancreas and in one case, by comparing various lengths of promoter sequence, it was established that as little as 168 bp upsiream of the transcript initiation site are sufficient to ensure correct tissue-specificity of expression from the insulin gene promoter [26]. In mice with detectable levels of human insulin, but no significant increase in total serum insulin levels, appropriate regulation of the human insulin transgene was established by challenging mice with glucose, amino acids or the hypoglycemic drug tolbutamide [11, 74]. In these cases normal pancreatic histology was reported and there were no obvious effects on growth. More recently it was shown that higher than normal serum insulin levels can result from similar transgenic experiments, consequently these mice are hyperglycemic in response to a glucose challenge and their normal fasting glucose levels suggest some insulin tolerance [54]. To date, however, transgenic experiments have not suggested any direct effects on growth in vivo, although insulin can promote growth in vitro [79] and is a routine addition in culture media. While it may be that insulin has no significant effect on growth in vivo further experiments, perhaps involving the targeted mutation of the insulin gene or its overexpression at ectopic sites and during earlier developmental stages, might be more revealing in this context. 2. IGF-I A key role of IGF-I, in mediating the effects of growth hormone (GH) on postnatal growth, was proposed over ten years ago when adminstration of purified IGF-I was shown to stimulate growth in hypophysectomised rats [70]. This rela- tionship was subsequently bome-out by experiments in which GH or IGF-I over-expression was achieved in transgenic mice. When rat or human GH genes were placed under the control of the mouse metallothionein I (mTM-I) gene promo- ter serum GH levels could be raised several hundred-fold, typically resulting in a 1.5-fold increase in adult body weight [64, 65]. Growth enhancement was detectable from about 3 weeks of age and was associated with elevated levels of circulating IGF-I. The rise in serum IGF-I was shown to preceed the acceleration in growth by about one week [56]. Although these data are consistent with the proposed role for IGF-I, the findings of Stewart et al. [80] suggest the increased circulating IGF-I might not be necessary to mediate GH action. These workers expressed human GH in mice using a mammary tumour virus LTR and these transgenic mice displayed supernormal growth kinetics similar to those de- scribed previously but did not always have high serum IGF-I levels. They suggest the extra growth results from direct GH action through the somatogenic and PRL receptors. Alternatively, stimulation of IGF-I production local to its sites of action might be the important factor as this could occur independently of changes in serum levels. In fact a combination fo direct and IGF-I mediated mechanisms of GH action remains likely and is consistent with the results of IGF-I over-expression in transgenic mice [55]. Although only one of two lines of mice established using an mMT-I/ hIGF-I construct were found to express the introduced gene these mice were 1.3 fold larger than controls, which is within the range of the effect obtained with mMT-I/GH consiructs. 170 A. Warp, P. BIeRKE et al. However, the phenotype of the IGF-I transgenic mice dif- fered from that of GH trangenics, notably in the range of organs which were enlarged or exhibited histopathological changes, and that the onset of the effect on live weights was delayed until at least six weeks after birth [7, 16, 57, 66]. Mice lacking a functional IGF-I gene have recently been created by gene targeting [3,52]. The homozygous mutants were approximately 60% normal size at birth and _ their survival into adulthood varied with genetic background. Postnatally, the surviving animals continued to grow slowly, eventually attaning only about 30% the weight of normal littermates. These experiments confirm unequivocally that IGF-I has a significant effect on growth during embry- ogenesis, before any effects of GH are manifest. 3. IGF-II In contrast with studies involving IGF-I, attempts to produce mice over-expressing IGF-II using promoters with relatively broad fields of activity have been largely unsuccess- ful. Atleast three groups of researchers found that, at best, only low levels of transgene expression were obtained when using promoters from the mMT-I (expression reported in 2/ 17 transgenic lines (63)), rat IGF-II (0/4; (49)) and human H-2k (0/8; (21)) genes. In all of these cases no significant effects on growth were reported. This can be attributed to an embryonic lethal effect which is thought likely to result from the presence of excess IGF-II, perhaps at critical sites and/or periods during development. Recent genetic evi- dence supports this theory as mice lacking the type 2 IGF receptor (T"? mutant mice) usually die at mid-gestation, but this phenotype is rescued when they also lack a functional Igf-2 gene [25]. The principle role for the type 2 receptor seems to be one of targeting its ligands (IGF-II and mannose- 6-phosphate) for lysosomal degradation, and not mitogenic signalling. Thus, the lethality observed in T"? mice is seen as an absence of this important regulator of IGF-II action. Very recently the problems associated with producing IGF-II transgenic mice were circumvented by using a keratin gene promoter which had previously provided tissue res- tricted expression of a h-ras transgene [2]. IGF-II transgene expression was seen in all of the transgenic lines established using this promotor (Ward ef a/., unpublished). These mice exhibited dramatic local effects on growth, consistent both with the paracrine/autocrine mode of action suggested for IGF-II by in situ expression studies (reviewed in [90]) and with proposed roles for IGF-II in growth related diseases, including Beckwith-Wiedemann syndrome and certain can- cers [51, 73, 89]. These over-expression experiments com- plement studies of mice in which the IGF-II gene was disrupted by gene targeting. Growth of IGF-II null mice was compromised during development, but not postnatally as recorded for mice with- out an intact IGF-I gene, and their size at birth was about 60% that of their normal littermates. Further differences between IGF-I and IGF-II null mice demonstrated that IGF-II acts at least two embryonic days earlier than IGF-I and that only IGF-II has an effect on placental growth [3, 73, 89]. The type 1 receptor was also mutated as part of this spectacular series of experiments and IGF-I-R null mice were more severely dwarfed at birth than either of those lacking individual functioning IGF genes. The availability of these three types of gene targeted mice, together with TP mutants, allowed the breeding of various double mutants [3, 25, 52]; some of their characteristics are summarized in Table 1. Comparison of all the resulting single and double mutants yeilded several important conclusions, including that IGF-I acts solely through the type | receptor (since the phenotypes of mice null for either J/gf-Jr or Igf-1 and Igf-Ir were indistinguishable) while IGF-II acts through both the type 1 receptor and at least one other (since similar phenotypes were Tas_e 1. Gross effects on growth of mice null for IGF and/or IGF receptor functions Nullitype Birth weight Placenta Onset Notes (% normal) lef-1 mee 60% : 100% E13.5 Survival to adulthood varies with genetic background. Some distortion in prop- ortionality of features. De- lays in ossification. Igf-2 60% 15% E11 Proportionally dwarfed. De- lays in ossification. Igflr or 45% 100% Ell Neonatal death (ieseHaLOny Ief-1/1eflr failure). Marked hypopla- sia in muscle and skin. X- tended delays in ossification. Igflr/Igf-2 30% 75% Ell Neonatal death (respiratory Igf-1/Igfr-2 failure). Marked hypopla- sia in muscle and skin. X- tended delays in ossification. Igf2r = — - Mid-gestational lethality. Igf-2/Igf2r 60% 75% - Neonatal lethality. These results are implicit and all quantities are approximate. Data from (3, 14, 15, 25,52). Note that Igf-2 results were obtained with heterozygous mutations, following paternal transmission of the null allele and the intact maternal allele is known to be expressed in some tissues. Insulin-like Growth Factors 171 displayed by mice null for either /gf-/ and /gf-2 or Igf-Jr and Igf-2 and in both cases these phenotypes were more severe than that of /gf-Jr mutants). The identity of the second IGF-II receptor remains ambiguous, although it could be the insulin receptor, and this receptor was deemed the sole mediator of IGF-II action on placental growth. IV TRANSCRIPTION FACTORS One alternative approach to studying growth factor production in vivo, involves manipulation of relevant trans- cription factor genes and introducing them into transgenic animals. It has been suggested that many growth factor defects arise as a consequence of a defect in transcriptional regulation rather than mutations in the growth factor gene itself. This is particularly important in embryogenesis, and it was recently shown that deregulation of the pax-2 gene had a dramatic impact on kidney development [19]. This gene belongs to a family of transcription factors that display organ and stage specific activity. pax-2 is expressed in the embryonic kidney after mesenchymal induction, in the ureter epithelium and early epithelial structures [17]. Furth- ermore, its expression is repressed upon terminal differentia- tion but persists at increased levels in Wilms tumours [18]. In this respect it mirrors the transcriptional activity of the IGF II gene which makes studies of persistent pax-2 gene express- ion in transgenic animals particularly interesting. Dereg- ulated pax-2 expression in transgenic mice gave rise to histologically abnormal and dysfunctional renal epithelium reminiscent of the congenital nephrotic syndrome. The in- terrelationship between pax-2 and IGF II and their combined detrimental impact on development however remains to be further clarified. It has been shown that both the IGF-I and IGF-II genes contain AP-1 binding sites [12, 45]. Therefore studies which aim at altering the availability of Fos and Jun proteins ought to influence the expression of the IGF genes and would be of particular interest from a developmental point of view. When both copies of c-jun were inactivated by homologous recombination, perfectly viable ES-cells were obtained [37]. When these were integrated into the germ line it was found that heterozygous mutant mice were normal, but embryos lacking c-jun, died at mid or late gestation, and displayed impaired hepatogenesis, altered fetal liver erythropoesis and generalized oedema. Moreover, it was found that c-jun -/- ES cells were capable of participation in development of all somatic cells in chimaeras except liver cells [38]. An exam- ple of the opposite approach, namely an overexpression of the AP-1 component fos yielded an unexpected series of results. Expression of c-fos has for more than a decade been considered an early step in the chain of events between growth stimulation and onset of DNA-synthesis. However, embryonic development is a fine balance between prolifera- tion, differentiation and programmed cell death. By study- ing fos-lacZ transgenic mice it was possible to demonstrate that continuous fos expression begins hours or days before the morphological demise of a condemned cell. Expression, therefore appears to be a hall mark of terminal differentiation and a harbinger of death [77]. These examples show that alteration of transcription factors expression lead to dramatic biological effects in transgenic animals. Therefore they should be taken into account when growth factors are to be considered as prime candidates for developmental processes in transgenic animals. V GROWTH FACTORS, TRANSGENESIS AND GENOMIC IMPRINTING One surprising outcome of the IGF-II gene knock-out experiments was the finding that the growth deficient phe- notype occurred in heterozygous mutant mice when the mutation was inherited from a male but not from a female [15]. This led to the discovery that IGF-II is subject to genomic imprinting, with only the allele inherited from the male being expressed in most tissues [14]. _The phenomenon of genomic imprinting was known before this, since nuclear transplantation experiments had demonstrated the none- quivalence of the male and female genetic contribution in mammalian development [84, 85], but /gf-2 was the first gene known to be influenced by this peculiar form of regulation. Soon afterwards the Jgf-2r gene was also shown to be imprinted after it was mapped within deletions of chromo- some 17 harboured by the T?? and T"”? mutants [4]. In this case, in accord with the lethality associated with female but not male transmission of the T'? and T'? mutations, the maternal allele was found to be active. Opposite imprinting of the /gf-2 and Igf-2r genes im- mediately brought to life an hypothesis of parent-offspring conflict predicted in plant and animal species in which the growth of progeny is dependent on a maternal nutient supply (e.g. the mammalian placenta) and in which multiple paterni- ty is likely among offspring of individual females [45]. Under these circumstances, and given that larger offspring are the most well equipped to survive, the interest of the father is best served if each of his progeny are as large and strong as possible. As there is a chance that he may be usurped the father makes the most of each opportunity to reproduce although this might involve an increased burden on the mother. Obviously, she is guaranteed to be equally related to each of her offspring, irrespective of any changes in mate choice and it follows that it is in the mothers interest to distribute resources evenly among her unborn young whether within a litter or in conserving the capacity to produce later litters. Since this parental conflict of interest requires that a “choice” is made regarding growth during embryonic life, Haig and Westoby [34] suggested that this would be reflected at the molecular level. That /gf-2 and the Jgf-2r can be viewed as molecular weapons in this conflict is entirely in keeping with their antagonistic effects on growth and, : course, that it is respectively the paternally- and materna' y- derived copies of each that are expressed [33]. Genomic imprinting of human IGF-II was recent!y con- 172 A. Warp, P. BIERKE et al. firmed [29] and is implicated in the ontogeny of the fetal overgrowth disease Beckwith Wiedemann syndrome [36]. Furthermore, it has been suggested that mis-regulation of imprinted genes might be important in a range of neoplastic diseases since, for instance, as only one allele is normally expressed then a single mutation would effectively result in a nullisomy. Conversely, activation of the normally silent allele would result in the gene product being overrepresented [23, 44]. Indeed relaxation of imprinting was recently de- monstrated for IGF-II in a proportion of Wilms’ tumours [60, 67]. There is one further intriguing link between IGF-II and imprinting in that the H/9 gene, located only about 80 kb from IGF-II [92], is also imprinted with the maternally derived copy being active in mouse and man [6, 93]. This gene has no ascribed function in vivo and might exert its effect as an RNA product since it does not encode an open reading frame [8]. However, the influence of H/9 on de- velopment is evidenced by transgenic experiments in which overexpression of intact H19 genes resulted in mid- gestational lethality [10]. It has been suggested that the regulation of IGF-I and H19 might be linked in some fashion since these two oppositely imprinted genes are such near neighbours [5, 83, 92]. The perceived importance of geno- mic imprinting in both normal development and growth related disease has fuelled an intense interest in achieving an understanding of the mechanism which underlies this phe- nomenon. An epigenetic signal which can be appropriately modified during passage of genes through either the male or female germline is required to establish and maintain allele specific expression of imprinted genes. Studies of the /gf-2 [69], Zgf-2r [81] and H19 [5, 24] genes indicate that DNA methylation might be involved in this process, and this was recently confirmed through the discovery that all three loci were lost in mice with a methyltransferase gene deficiency [50]. However, the race towards a full understanding is far from over. Early evidence suggests that not all of the signals required for properly imprinted expression have yet been discovered as only a minor proportion of /gf-2 [49] and H19 [5] transgenes behave in accord with their parent of origin, but it is without doubt that transgenesis will play a key role in unravelling this biological conundrum. ACKNOWLEDGMENTS The authors have been generously supported by the Cancer Research Campaign of Great Britain as well as by Cancerfonden and Barncan- cerfonden, Sweden. REFERENCES 1 Archibald AL, McClenaghan M, Hornsey V, Simons JP, Clark AJ (1990) High levels of expression of biologically active human alphal-antitrypsin in the milk of transgenic mice. Proc Natl Acad Sci USA 87: 5178-5182 2 Bailleul B, Surani MA, White S, Barton SC, Brown K, Blessing M, Jorcano J, Balmain A (1990) Skin hyperkeratosis and 10 11 13 14 17 19 20 papilloma formationin transgenic mice expressing a ras oncogene from a suprabasal keratin promoter. Cell 62: 697— 728 Baker J, Liu J, Robertson LJ, Efstriatiadis A (1993) Role of insulin-like growth factor type 2 receptor in embryonic and postnatal growth. Cell 75: 73-82 Barlow DP, Stoger R, Herrmann BG, Saito K, Schweifer N (1991) The mouse insulin-like growth factor type-2 receptor is imprinted and closely linked to the Tme locus. Nature 349: 84— 87 Bartolomei MS, Webber AL, Brunkow ME, Tilghman SM (1993) Epigenetic mechanisms underlying the imprinting of the mouse H19 gene. Genes Dev 7: 166-167 Bartolomei MS, Zemel S, Tilghman SM (1991) Parental im- printing of the mouse H19 gene. Nature 351: 153-155 Behringer RR, Lewin TM, Quaife CJ, Palmiter RD, Brinster RL, d@ErcoLe AJ (1990) Expression of insulin-like growth factor I stimulates normal somatic growth in growth hormone- deficient transgenic mice. Endocrinology 127: 1033-1040 Brannan CI, Dees EC, Ingram RS, Tilghman SM (1990) The product of the H19 gene may function as an RNA. Mol Cell Biol 10: 28-36 Brinster RL, Chen HY, Trumbauer ME, Senear AW, Warren R, Palmiter RD (1981) Somatic expression of herpes thymi- dine kinase in mice following injection of a fusion gene into eggs. Cell 27: 223-231 Brunkow ME, Tilghman SM (1991) Ectopic expression of the H19 gene in mice causes prenatal lethality. Genes Dev 5: 1092-1101 Bucchini D, Ripoche M-A, Stinnakre M-G, Desbois P, Lores P, Monthioux E, Absil J, Lepesant J-A, Picter R, Jami J (1986) Pancreatic expression of human insulin gene in transgenic mice. Proc Natl Acad Sci USA 83: 2511-2515 Caricasole A, Ward A (1993) Transactivation of mouse IGF II gene promoters by the AP-1 complex Nucl Acids Res 21: 1873- 1879 Chada K, Magram J, Raphael K, Radice G, Lacy E, Constantini F (1985) Specific expression of foreign beta globin gene in erythroid cells of transgenic animals. Nature 314: 377-380 deChiara TM, Efstratiadis A, Robertson EJ (1991) Parental imprinting of the mouse insulin-like growth factor gene. Cell 64: 849-859 deChiara TM, Robertson EJ, Efstratiadis A (1990) A growth deficiency phenotype in heterozygous mice carrying an insulin- like growth factor II gene disrupted by targetting. Nature 345: 78-80 Doi T, Striker LJ, Gibson CC, Agodoa LY, Brinster RL, Striker GE (1990) Glomerular lesions in mice transgenic for growth hormone and insulin like factor-I. Am J Path 137: 541-552 Dressler GR, Deutsch U, Chowdbury K, Nornes HO Gruss P (1990) Pax-2, a new murine paired box containing gene and its expression in the developing secretory system. Development 109: 787-795 Dressler GK, Douglas EC (1992) Pax-2 is a ANA-binding protein expressed in embryonic kidney and Wilms tumour. Proc Natl Acad Sci USA 89: 1179-1183 Dressler GR, Wilkinsson JE, Rothenpieler UW, Patterson LT, Williams L, Westphal H (1993) Deregulation of pax-2 express- ion in transgenic mice generates severe kidney abnormalities. Nature 362: 65-67 Edwards RH, Rutter WJ, Hanahan D (1989) Directed express- ion of NGF to pancreatic beta cells in transgenic mice leads to selective hyperinnervation of the islets. Cell 58: 161-170 Ellis C (1990) Studies of the roles of IGF-II during mouse development. DPhil Thesis Oxford University 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Insulin-like Growth Engstr6m W, Lindham S, Schofield PN, Wiedemann Beckwith Syndrome. Eur J Pediatr 147: 450-457 Feinberg AP (1993) Genomic imprinting and gwnw activation in cancer. Nature Genet 4: 110-113 Ferguson-Smith AC, Sasaki H, Cattanach BM, Surani MA (1993) Parental-origin-specific epigenetic modification of the mouse H19 gene. Nature 362: 751-755 Fitson A, Louvi A, Efstradiatis A, Robertson EJ (1993) Res- cue of the T-associated maternal effect in mice carrying null mutations in IGF-2 and IGF-2-r, two reciprocally imprinted genes. Development 118: 731-736 Fromont-Racine M, Bucchini D, Madsen O, Desbois D, Linde S, Nielsen J, Saulnir C, Ripoche M-A, Jami J, Picet R (1990) Effect of 5’-flanking sequence deletions on expression of the human insulin gene in transgenic mice. Mol Endocrinol 4: 669— 677 Fukamizu A, Seo MS, Hatae T, Yokohama M, Nomura T, Katsuki K and Murakami K (1990) Tissue specific expression of the human renin gene in transgenic mice. Biochem Biophys Comm 165: 826-832 Gardner RL (1968) Mouse chimaeras obtained by the injection of cells into the blastocyst. Nature 220: 596-597 Giannourakis N, Deal C, Paquette J, Goodyear CG, Polychro- nakas C (1993). Parental genomic. imprinting of the human IGF2 gene. Nature Genet 4: 98-101 Gordon JW, Scangos GA, Plotkin DJ, Barbariosa JA, Ruddle FH (1980) Genetic transformation of mouse embryos by mic- roinjection of foreign DNA. Proc Natl Acad Sci USA 77: 7380-7384 Gordon JW, Ruddle FH (1981) Intergration and stable germ line transmission of genes injected into mouse pronuclei. Sci- ence 214: 1244-1246 Gordon JW (1989) 171-229 Haig D, Graham C (1991) Genomic imprinting and the strange case of the insulin-like growth factor II receptor. Cell 64: 1045-1046 Haig D, Westoby M (1989) Parent-specific gene expression and the triploid endosperm. Am. Nat 134: 147-155 Harris S, McClenaghan JP, Simons JP, Ali S, Clark AJ (1990) Gene expression in the mammary gland. J Reprod Fert 88: 707-715 Henry B-PC, Chenbensse V, Beldjord C, Schwartz C, Uter- mann G, Junien C (1991) Uniparental patenal disomy in a genetic cancer disposing syndrome. Nature 351: 665-667 Hilberg F, Wagner EF (1992) Embryonic stem cells lacking functional c-jun. Consequences for growth and differentiation, AP-1 activity and tumorigenicity. Oncogene 7: 2371-2380 Hiberg F, Aguzzi A, Howells N, Wagner EF (1993) C-jun is essential for normal mouse development and hepatogenesis. Nature 365: 179-181 Hogan B, Constantini F, Lacy E (1986) Manipulating the mose embryo a laboratory manual. Cold Spring Harbor Press, Cold Spring Harbor, New York Hooper ML (1992) Embryonal stem cells: introducing planned chances into the animal germ line. In “Modern Genetics” Harwood Academic Publishers, Switzerland Jaenisch R, Mintz B (1974) Simian virus 40 DNA sequences in DNA of healthy adult mice derived from preimplantation blasto- cysts injected with viral DNA. Proc Natl Acad Sci USA 71: 1250-1255 Jaenisch R (1976) Germ line integration and mendelian trans- mission of the exogenous Moloney leukemia virus. Proc Natl Acad Sci USA 73: 1260-1264 Joyner AL (1993) Gene targeting: a practical approach. Ox- Transgenic Animals. Int Rev Cytol 115: 46 47 48 53 56 57 58 59 60 61 62 63 64 173 Factors ford University Press, Oxford Junien C (1992) Beckwith-Wiedemann syndrome, tumouri- genesis and imprinting. Curr Opin Gen Dev 2: 431-438 Kajimoto Y, Kawamori R, Umayahara Y, Iwama N, Imano E, Morishima T, Yamasaki Y, Kamada T (1993) An AP-1 enhan- cer mediates TPA-induced transcriptional activation of the chicken insulin-like growth factor I gene. Biochem Biophys Res Comm 190: 767-773 Koopman P, Gubbay J, Collignon J, Lovell-Badge R (1989) Zfy gene expression patterns are not compatible with a primary role in mouse sex determination. Nature 342: 940-942 Koopman P, Gubbay J, Vivian N, Goodfellow P, Lovell-Badge R (1991) Male development of chromosomally female mice transgenic for Sry. Nature 351: 117-121 Lang RA, Metcalf D, Cuthbertson RA, Lyons I, Stanley E, Kelso A, Kannourakis G, Williamsson DJ, Klintworth GK, Gonda TJ, Dunn A (1987) Transgenic animals expressing a hematopoetic growth factor gene develop accumulation of mac- rophages, blindness and a fatal syndrome of tissue damage. Cell 51: 675-686 Lee JE, Tantravahi U, Boyle AL, Efstratiadis A (1993) Paren- tal impriting of an IGF-2 transgene. Mol Rep Dev 35: 382-390 Li JE, Beard C, Jaenish R (1993) Role for DNA methylation in genomic impriting. Nature 366: 362-365 Little M, van Heyningen, Hastie N (1991) and disease. Nature 351: 609-610 Liu J, Baker J, Perkins AS, Robertson LJ, Efstratiadis A (1993) Mice carrying null mutations of the genes encoding insulin like growth factor I and type 1 IGF receptor. Cell 75: 59-72 Mansour SL, Thomas KR, Capecchi M (1988) Disruption of the proto-oncogene int-2 in mouse embryo-derived stem cells: a general strategy for targeting mutations to non-selectable genes. Nature 336: 348-353 Marban SL, DeLoia JA, Gearhart JD (1988) Hyperinsu- linemia in transgenic mice carrying multiple copies of the human insulin gene Dev Gene 10: 356-364 Mason AJ, Pitts SL, Nikolics K, Szonyi E, Wilcox JN, Seeberg PH, Stewart TA (1986) The hypogonadal mouse. Reproduc- tive functions restored by gene therapy. Science 234: 1372- 1378 Mathews LS, Hammer RE, Behringer R, d’Ercole AJ, Bell Cl, Brinstcr R, Palmiter RD (1988) Growth enhancement of trans- genic mice expressing human insulin-like growth factor I. En- docrinology 123: 2827-2833 Mathews LS, Hammer RE, Brinster RL Palmiter RD (1988) Expression of insulin-like growth factor I in transgenic mice with elevated levels of growth hormone is correlated with growth. Endocrinology 123: 433-437 Meisler M (1992) Dads and disomy Insertional mutations of classical and novel genes in transgenic mice. Trends Genet 8: 341-344 Mintz B (1962) Formation of genotypically mosaic mouse embryos. Am Zool 2: 432 (Abstract) Ogawa O, Eccles J, Szeto J, McNoe LA, Dun K, Maw MA, Smith PJ, Reeve AE (1993) Relaxation of insulin like growth factor II gene imprinting implicated in Wilms’ tumour. Nature 366: 749-751 Ohlsson R, Nystrém A, Pfeifer-Ohlsson S, Tohonen V, Hedberg F, Schofield PN, Flarn F, Ekstr6m T (1993) /GF2 is parentally imprinted during human embryogenesis and in the Beckwith- Wiedemann syndrome. Nature Genet 4: 95-97 Palmiter RD, Brinster RL (1985) Transgenic mice. 343-345 Palmiter RD, Brinster R (1986) Germ line transformation of mice. Ann Rev Genet 29: 233-236 Palmiter RD, Brinster RL, Hammer RE, Trumbauer MG, Cell 41: 174 65 66 67 68 69 70 71 72 73 74 75 76 WI 78 A. Warp, P. BIERKE et al. Rosenfeld MG, Birnberg NC, Evans RM (1982) Dramatic growth of mice that develop from eggs microinjected with metallothionein-growth hormone fusion genes. Nature 300: 611-615 Palmiter RD, Norstedt G, Gelinas RE, Hammer RE, Brinster RL (1983) Meallothionein-human GH fusion genes stimulate growth of mice. Science 222: 809-814 Quaife CJ, Mathwes LS, Pinkel C, Hammer RE, Brinster RL, Palmiter RD (1989) Histopathology associated with elevated levels of growth hormone and insulin like growth factor I in transgenic mice. Endocrinology 124: 40-48 Rainier S, Johnson LA, Dobry CJ, Ping AJ, Grundy PE, and Feinberg AP (1993) Relaxation of imprinted genes in human cancer. Nature 362: 747-749 Robertson EJ (1991) Using embryonic stem cells to introduce mutations into the mouse germ line. Biol Reprod 44: 238-245 Sasaki H, Jones PA, Chaillet JR, Ferguson-Smith AC, Barton SC, Reik W, Surani MA (1992) Parental imprinting; potential- ly active chromatin of the repressed maternal allele of the mouse insulin-like growth factor II gene. Genes Dev 6: 1843-1856 Schoenle E, Zapf J, Humbel R, Froesch E (1982) IGF-I stimulates growth in hypophysectomized rats. Nature 296: 252-253 Schofield PN, Lindham S, Engstr6m W (1989) Analysis of gene dosage on chromosome 11 in children suffering from Beckwith Wiedemann Syndrome. Eur J Pediatr 148: 320-324 Schofield PN (1992) The insulin-like growth factors structure and biological functions. Oxford University Press, Oxford Schofield PN, Engstro6m W (1992) Insulin like growth factors in human cancer. In “The insulin-like growth factors structure and biological functions” Oxford University Press, Oxford pp. 240-257 Selden RF, Skoskicwicz MZ, Burke Howie K, Russell PS, Goodman HM (1986) Regulation of human insulin gene ex- pression in transgenic mice. Nature 321: 525-528 Sigmund CD, Jones CA, Mullins JJ, Kim U, Gross K (1990) Expression of murine cutaneous renin genes in subcutaneous connective tissue. Proc Natl Acad Sci USA 87: 7993-7997 Sigmund CD, Jones CA, Fabian JR, Mullins JJ, Gross KW (1990) Tissue and cell specific expression of a renin promoter reporter gene construct in transgenic mice. Biochem Biophys Comm 170: 344-350 Smeyne RJ, Vendrell M, Hayward M, Baker SJ, Miao GG, Schilling K, Robertson LM, Curran T, Morgan JI (1993) Con- tinuous c-fos expression precedes programmed cell death in vivo. Nature 363: 166-169 Sola C, Tronik D, Dreyfus M, Babinet C, Rougeon F (1989) Renin promoter SV40 Large T antigen transgenes induce tumors 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 irrespective of normal cellular expression of renin genes. Oncogene Res 5: 149-153 Spaventi R, Antica M, Pavclic K (1990) Insulin and insulin-like growth factor I (IGF-I) in early mouse embryogenesis. De- velopment 108: 491-495 Stewart TA, Clift S, Pitts-Meek S, Martin L, Terrell G, Liggitt D, Oakley H (1992) An evaluation of the functions of the 22-kilodalton (kDa) the 20-kDa, and the N-terminal polypeptide forms of human growth hormone using transgenic mice. En- docrinology 130: 405-414 Stoger R, Kubicka P, Liu C-G, Kafri T, Razin A, Cedar H, Barlow DP (1993) Maternal-specific methylation of the im- printed mouse locus identifies the expressed locus as carrying the imprinted signal. Cell 73: 61-71 Suematsu S, Matsuda T, Aozasa K, Akira S, Nakano N, Ohno S, Yamamura J, Hirnao T, Kishimoto T (1989) IgG plasmocy- tosis in IL-6 transgenic mice. Proc Natl Acad Sci USA 86: 7547-7551 Surani MA (1993) Silence of the genes. Nature 366: 302-303 Surani MA, Kothary R, Allen ND, Singh P, Fundele R, Ferguson-Smith AC, Barton SC (1990) Genomic imprinting and development in the mouse. Development (suppl) 89-98 Surani MA, Reik W, Norris MI, Barton SC (1986) Influence of germline modifications of homologous chromosomes on mouse development. J Embryol Exp Morphol 97 (suppl) 123-126 Tarkowski K (1961) Mosue chimaeras developed from fused eggs. Nature 190: 857-860 te Riele H, Maandag ER, Berns M (1992) Highly efficient gene targeting in embryonic stem cells through homologous recom- bination. Proc Natl Acad Sci USA 89: 5128-5132 Van Deursen J, Wieringa B (1992) Targeting of creatin kinase M gene in embryonic stem cells uisng isogenic and nonisogenic vectors. Nucleic Acids Res 20: 3815-3820 Van Heyningen V, Hastie ND (1992) Wilms’ tumour: reconcil- ing genetics and biology. Trends Genet 8: 16-21 Ward A, Ellis CJ (1992) The insulin like growth factor genes. In “Insulin like growth factors - Structure and function” Oxford University Prees Wilde CJ, Clark AJ, Kerr MA, Knight CH, McCleanaghan M, Simons P (1992) Mammary development and milk secretion in transgenic mice expressing the sheep beta lactoglobulin gene. Biochem J 284: 717-720 Zemel S, Bartolomei MS, Tilghman SM (1992) Physical link- age of two mammallan imprinted genes, H19 and insulin-like growth factor 2. Nature Genet 2: 61-65 Zhang Y, Tycko B (1992) Monoallelic expression of the human H19 gene. Nature Genet 1: 40-44 ZOOLOGICAL SCIENCE 11: 175-190 (1994) © 1994 Zoological Society of Japan REVIEW Processing of Contrast Signals in the Insect Ocellar System MAKoTo MIZUNAMI Laboratory of Neuro-Cybernetics, Research Institute for Electronic Science, Hokkaido University, Sapporo 060, Japan INTRODUCTION Most visual systems are designed to analyze space, and each array of their photoreceptors encodes light intensity at its narrow receptive field directed at different parts of the surrounding environment. The intensity signals are then converted into temporal and spatial contrast signals, i.e., the ratio of changes in light intensity in time and space, as a necessary pre-processing for detection of visual images. In- deed, all biologically significant visual information, such as movement, shape, and color of objects can be extracted from temporal and spatial contrasts but not directly from absolute light intensities. Although neural mechanisms for detecting spatial contrast are well examined, as reviewed by Laughlin [41], neural mechanisms for detecting temporal contrast are not well understood. This is mainly because the mechanisms for detecting temporal contrast signals are incorporated into complicated neural mechanisms for detecting more specific visual features such as spatial and color contrasts and move- ment, thus, it is almost impossible to separate neural circuits for detecting temporal contrast signals from those subserving advanced visual functions. Insect ocelli are simple photoreceptive organs incapable of detecting shape, motion or color of small objects; they detect only intensity changes averaged over their large visual field. The ocelli, therefore, provide excellent material for studying neural mechanisms underlying the processing of temporal contrast signals. In this review I will summarize our present knowledge of the mechanisms underlying the processing of contrast signals in insect ocellar systems. This will include a discussion of spatial, temporal, and spectral properties of insect ocellar systems with special reference to their behavioral roles, as well as a review of neural organiza- tion of the insect ocellar system. I will examine the mechan- isms to detect and process temporal contrast signals and summarize some neural principles underlying information processing in insect ocellar systems. I will then show that simple modifications and duplications of neural circuits for detecting temporal contrast signals are sufficient to explain neural circuits for detecting more elaborate visual features, such as the direction of motion. Finally, I will discuss Received February 25, 1994 possible evolutionary pathways by which advanced neural networks subserving advanced visual functions have emerged from simpler neural networks subserving simpler visual func- tions. FUNCTIONAL PROPERTIES OF INSECT OCELLI Most adult insects possess two or three ocelli in addition to a pair of compound eyes. Three ocelli is the usual number but some insects like cockroaches and moths have two ocelli. The compound eyes are sophisticated visual organs responsible for functions that require good spatial resolution, such as the perception of movement, fixation of objects and pattern recognition. In contrast, the ocelli are simple photoreceptive systems with very poor spatial resolu- tion [21]. Why then do insects need simple photoreceptors, i.e., ocelli, even though they are equipped with sophisticated compound eyes? Although this question has not been fully resolved, some answers have emerged recently from exten- sive anatomical, physiological and behavioral studies [51]. Insect ocelli possess high-aperture dioptrics which ex- hibit wide visual fields [9, 21, 82]. Measurements of the focal length of the ocellar lens show that the image plane lies well behind the retinal receptory layer. Due to underfocus- sing, an object entering the ocellar visual field results in a change in the light intensity impinging on the photoreceptor layers rather than the formation of animage. The output of a large number of photoreceptor cells converges upon a small number of large second-order neurons, called L-neurons [21]. Such a system is best suited for the detection of small changes in light intensity integrated over its wide visual field [111]. Thus, the ocelli are specialized for the effective capture of photons, at the cost of spatial resolution. The spectral properties of ocellar L-neurons are char- acterized by broad spectral tuning with sensitivity to both ultraviolet (UV) and visible light. Ocellar spectral sensitiv- ity curves have a marked UV peak with an additional peak in either the blue (in flies [31, 34]) or green (in dragonflies [6], locusts [111], moths [14, 69], and honeybees [19]). The combined UV-blue or UV-green sensitivity in the ocelli appears to be an adaptation for increased sensitivity. Although the ocelli of most insects have sensitivity to both UV and visible light, a few insects living in extreme light 176 M. Mizunami environments have ocelli with only UV or green sensitivity. Mote and Wehner [59] found that the ocelli of desert ants, which live under bright and UV-rich sunlight, have only a UV peak and are not sensitive at wavelengths longer than 445 nm. The ocelli of some nocturnal insects which rarely encounter UV-rich sunlight such as field crickets, Gryllus firmus [40] and cockroaches, Periplaneta americana [19] have only single receptor systems maximally sensitive to green and not sensi- tive to UV. These examples nicely fit the theory that visual pigments have evolved in accordance with environmental light conditions. Measurement of the absolute sensitivity of large second- order ocellar neurons, called L-neurons, of the locust sug- gests that the sensitivity of ocelli is at least several times as high as that in compound eyes. Wilson [111] concluded that L-neurons of locust ocelli are 5 times more sensitive to a point source than lamina monopolar cells, a major class of second- order neurons of compound eyes. Wilson [111] also con- cluded that L-neurons of locusts are 5000 times more sensitive to extended sources than the lamina monopolar cells, although pooling of signals from monopolar cells could improve the sensitivity of higher visual neurons. What are the advantages of having high sensitivity? Light intensity is defined as the average number of photons per unit time, and the fewer the photons, the larger is the uncertainty about the true average. The absolute sensitiv- ity, i.e., the effectiveness in catching a photon and converting it to a voltage signal, is an essential limiting factor for visual systems. In any arbitrary visual system, increased absolute sensitivity can be used to see at lower intensities, to see faster events, or to detect objects of less contrast. There is evidence to show that ocelli contribute to low light intensity perception. Schricker [81], for example, ex- amined phototactic runs of honeybees where bees could choose between two lights. He noted that a significantly greater number of them chose the brighter light. Below 1 Lux, only a two-fold difference of intensity was needed for this to occur. Bees with one ocellus blinded needed a four-fold difference, bees with two ocelli blinded needed a six-fold difference and bees with all their ocelli blinded needed an eight-fold difference before a significantly greater number of the bees chose the brighter of the two lights. Initiation and cessation of diurnal activities of insects often depend on light intensity levels [12, 81]. Under natu- ral environmental conditions where the light-dark cycle con- sists of dawn and dusk ramps, it would be advantageous for insects to be sensitive to minute changes in illumination. There is evidence to show that signals from ocelli are utilized to determine the threshold light intensity for diurnal activities in bees [23, 81], moths [15, 92] and crickets [74]. In the bee, the first and last of its daily flights is dependent upon the intensity level [81]. If the ocelli are occluded, foraging bees behave normally in most respects [81]. Occlusion of the ocelli, however, does interfere with the timing of the first and last foraging flights. Bees with one, two or three ocelli occluded start to collect food later in the morning and cease collecting earlier in the evening than normal workers. The light intensity required for the first and last collecting flight is increased by a factor of two if one ocellus is covered, 3.3 if two ocelli are covered and 4.5 if all the ocelli are covered. Further evidence of an ocellar contribution to low light intensity perception is reviewed by Mizunami [51]. Another notable advantage of ocelli over compound eyes is the higher speed of signal transmission. In bees, the latency of ocellar and compound eye pathways were mea- sured in descending multimodal neurons which receive inputs from both the compound eyes and ocelli [25]. The descend- ing neurons received ocellar input before the arrival of the delayed compound eye input. The latency of ocellar path- ways was 9 msec, while that of the compound eye pathway was 25-35 msec. For fast-flying insects, the ability to perform rapid visual steering is essential for survival. There is evidence to show that insect ocelli contribute to rapid visual steering in flight. Taylor [101] measured the delay of the head motion response of a tethered locust after the motion of an artificial horizon. Both the ocelli and compound eyes were involved in this response. The delay was 45.4+4 msec when all eyes were intact. The delay was almost unchanged when the com- pound eyes were ablated (47.4+12 msec) but significantly increased when the ocelli were cauterized (103.4+7 msec), thereby demonstrating that the high speed of signal transmis- sion by the ocelli contributes to shortening the latency of the locust’s response to motion of horizon. The perfect suitability of insect ocelli for detecting movement of the horizon, 1.e., the contrast between the earth and sky, and for stability control in flight have been discussed in detail [29, 111]. Wilson [111] argued that locust ocelli have a large receptive field directed horizontally, providing the animal with heavily blurred neural images of the skyline, where unwanted information about structural details are eliminated. Ocellar sensitivity to UV facilitates horizon detection since the contrast between bright sky and dark ground is highestin UV. The high speed of signal detection and transmission in the ocellar system is ideal for rapid course control. Pitch and roll deviation of the flight course are independently detectable by the combination of signals from three ocelli. A roll (turning around the long axis of the body) will cause no change in signal from the median ocellus but will tend to cause a decrease of illumination in one lateral ocellus and an increase in the other lateral ocellus. Detec- tion of pitch can be achieved through measurement of the output of the median ocellus. This hypothesis was con- firmed by behavioral studies in dragonflies [93, 94] and locusts [20, 100, 101]. The neural mechanisms underlying the ocel- lar contribution to the flight steering are well established in locusts [24, 73, 75, 85]. In conclusion, although the ocelli can detect only intensi- ty changes averaged over its large receptive field, they are superior to the co-existing compound eyes in sensitivity and speed. Thus, the ocelli play a major role in insects where functions such as stability control in flight and low light Signal Processing in Insect Ocelli 177 intensity perception require for high speed or high sensitivity. Insects have successfully extended the range of visual stimuli to which they can respond by having two fundamentally different visual systems, i.e., compound eyes and ocelli, the former designed to attain a high spatial resolution and the latter to attain a fast signal transmission and a high absolute sensitivity. The specific functional design of insect ocelli makes them a suitable model system for examining the neural basis of the processing of temporal contrast signals. NEURAL ORGANIZATION OF THE COCKROACH OCELLAR SYSTEM The neural circuits of the insect ocellar system have been studied in a number of species including locusts [22, 44, 90] and bees [25, 68]. Here I will briefly discuss the neural organization of the cockroach ocellar system, as a representa- tive example. Comparative aspects of the neural circuits of insect ocellar systems have been discussed (Mizunami, sub- mitted). Cockroaches have two ocelli, each of which con- tains about 10,000 photoreceptor cells [108], which converge synaptically upon four large second-order neurons in the ocellar plexus [10, 55, 104]. The large second-order neurons, called L-neurons, exit the ocellus and project into the ocellar tract neuropil of the protocerebrum, through the ocellar nerve (Fig. 1) [52, 55]. Electron microscopic studies show that the ocellar nerve and the ocellar tract are a continuous neuropil area where second-order neurons make synaptic interactions with third-order and efferent neurons [102; 104]. The morphology and physiology of neurons 200um Fic. 1. Morphology of an large second-order neuron (L-neuron) of the cockroach ocellus. The L-neuron extends their dendritic branches into the ocellar plexus, and its axon projects into the ocellar tract neuropil of the protocerebrum, through the ocellar nerve. OC, ocellus; ON, ocellar nerve; PC, protocerebrum; OT, ocellar tract neuropil; OL, optic lobe. Viewed postero- dorsally. Modified from Mizunami et al. [56]. involved in the cockroach ocellar system have been success- fully identified by combining extracellular and intracellular staining techniques [52, 55, 63, 64, 65, 105]. Recently, the number and gross morphologies of interneurons in the ocellar tract neuropils were estimated by injecting cobalt ions into the ocellar tract neuropil via microelectrodes. Cobalt ions were taken up and transported by neurons in the ocellar tract; thus, the morphologies of these neurons could be revealed by subsequent histological treatment. By comparing the mor- phologies of neurons in more than 50 cobalt-filled prepara- tions, it was concluded that each ocellar tract neuropil contains at least 25 interneurons. Twenty-two out of 25 neurons have been characterized anatomically and physiolo- gically by intracellular recordings and stainings [52, 55, 64, 65; Mizunami, submitted], which are: (1) four large second- order neurons projecting into the ocellar tract [55], (2) 16 third-order neurons [52, 55; Mizunami, submitted] which receive monosynaptic inputs from second-order neurons in the ocellar nerve and the ocellar tract [54, 102, 104] and project into various neuropil areas of the brain, (3) two possible efferent neurons modulating the activity of large second-order neurons [64]. Morphology of some of the third-order neurons are shown in Fig.2. The remaining three neurons still to be characterized. The projection areas of the third-order ocellar neurons include: (1) visual, olfac- tory and mechanosensory centers, (2) the mushroom body (a higher associative center) [58], (3) premotor centers, includ- ing the posterior slope, from which descending brain neurons originate [97, 98] and (4) the thoracic motor systems. The abundance of target neuropil areas in the cockroach ocellar system suggests a multiplicity of ocellar functions: (1) to modulate the activity of visual, olfactory and mechanosensory systems, (2) to influence the activity of higher associative centers, (3) to modulate the activity of descending brain neurons and (4) to produce direct behavioral actions. The projection areas of ocellar neurons of the cockroach are more or less similar to those reported for other insects including locusts [22], bees [68], and crickets [39]. In conclusion, the neural organization of the cockroach ocellar system is summarized as follows: (1) the large number of photorecep- tors converge onto only a very small number (four) of second-order neurons in the ocellar plexus and (2) in the ocellar tract neuropil, second-order neurons synapse onto a large number of third-order neurons which project into a number of target neuropils. LINEAR PERIPHERAL MECHANISMS FOR THE DETECTION OF CONTRAST SIGNALS An effective approach for understanding information processing in the ocellar system is to analyze the dynamics and sensitivity of the responses of its interneurons to changes in light intensity around a mean level, because this system is characterized by a high sensitivity for detecting changes in intensity and high speed for transmitting them to various target neuropils including thoracic motor centers. Analysis 178 Fic. 2. Examples of third-order ocellar neurons of the cockroach ocellar system. Viewed postero-dorsally. neuron which arborizes in an ocellar tract and projects into the ipsilateral posterior slope. arborizes in bilateral ocellar tracts and projects into bilateral posterior slopes. lobula and medulla of the optic lobe. tritocerebrum. Modified from Mizunami and Tateda [52]. of intracellularly-recorded responses of ocellar neurons to changes in light intensity was pioneered by Chappell and Dowing [7] who studied the responses of photoreceptors and the second-order neurons of dragonfly ocelli, and discussed signal processing at synapses between them, which I refer to as first synapses. Subsequently, Mizunami and his col- leagues [49, 53, 54, 56] have analyzed response properties of second- and third-order neurons of cockroach ocelli and examined information processing at synapses between second and third-order ocellar neurons, which I refer to as second synapses, by using white-noise and sinusoidally modulated light. photoreceptors and second-order neurons. This section deals with information processing by Further proces- sing of contrast signals by higher-order neurons will be C M. MIZUNAMI (A) A PS-I (B). A PS-III neuron which (C) A OL-I neuron which projects into the (D) A D-I neuron which arborizes in the ocellar tract, posterior slope and Its axon descends contralaterally toward the thoracic ganglia. neurons receive monosynaptic inputs from L-neurons [54]. There is evidence to show that these OC, ocellus; PC, protocerebrum; OT, oceilar tract neuropil; PS, posterior slope; OL, optic lobe; M, medulla; LO, lobula; TC, tritocerebrum; SO, suboesophageal ganglion. discussed in the next section. Signal processing at first synapses Insect ocelli contain a large number of photoreceptors which make synapses with a small number of large second- order neurons, called L-neurons. Both the photoreceptors and L-neurons generate graded, slow potential responses to light stimuli. Photoreceptors exhibit a depolarizing re- sponse to a step of light given in the dark, whereas L-neurons exhibit a hyperpolarizing response (Fig. 3A, B) [7, 35, 70, 88]. The hyperpolarizing response of L-neurons is due to an increase in membrane permeability to ions whose equilibrium potential is negative to the membrane potential in the dark [111]. Signal Processing in Insect Ocelli 179 ; Receptor Response > ree eee oa I. 2 / (0) LEO Fv Rec! a & ae "es ae = o—$_ 729 Postsynaptic v4 SoS Response e ae ye transient -4 -10 =a neaaaad e sustained fv -5 4 age SOLE 72 ie Say LOG I Fic. 3. Responses of a photoreceptor and a large second-order neuron of the median ocellus of the dragonfly, Aeschna. In- tracellularly recorded responses of a photoreceptor (A) and a large second-order cell (B) to illumination of the median ocellus with pulses of white light at different intensities Log I, logy relative intensity of illumination. Scales: 200 ms; 20mV. (C) Intensity-response relationships for receptor and postsynaptic units. The amplitude of the peak of the transient wave and the sustained component of the receptor response and the post- synaptic response are plotted as a function of intensity. The sustained component of the receptor response was measured 3 sec after the start of illumination. A and B are from Patterson and Chappell [70]; C is from Chappell and Dowling [7]. Chappell and Dowling [7] have compared responses of photoreceptors and L-neurons of dragonfly ocelli to a step of light given in the dark, and have noted two important differences between the responses of photoreceptors and L-neurons. First, the sensitivity of L-neurons is larger than that of photoreceptors by 1-2 log units (Fig. 3A, B). This is explained by a high ratio of convergence of photoreceptors onto L-neurons [11]. Second, the waveform of the response of L-neurons is much more transient (Fig. 3A, B), and the sustained component of the response of L-neurons is much less prominent (Fig. 3C) than that of photoreceptors. The functional significance of the enhancement of transi- ence was clarified by further examinating the responses of photoreceptors and second-order neurons to changes in light intensity around a mean level [7, 56]. The photic inputs that visual systems receive naturally is a modulation of light intensity around a mean illuminance. The mean illuminance changes slowly but covers a large range in the course of one day. The depth of fluctuation around the mean is moderate and remains roughly constant. The photoreceptor response to changes in intensity consists of two components, a steady mean potential and a time-varying component, which signal mean illuminance levels and intensity changes, respectively. The steady response component of L-neurons, however, is highly compressed (in dragonflies [7, 56, 88]) or completely eliminated (in bees [3]), indicating that L-neurons signal relative intensity changes rather than absolute light intensity level. It is concluded that, at the first synapses, the photore- ceptor responses signaling absolute intensity level is high-pass filtered to produce a postsynaptic response signaling about relative intensity change. The enhancement of the response to intensity changes by temporal highpass filtering have been noted at first synapses of most visual systems studied so far, including barnacle ocelli [28, 95, 99], insect compound eyes [42, 84, 109], Limulus lateral eyes [72] and vertebrate retinas [4, 60, 62, 110]. This apparent common principle among visual systems reflects the fact that visual systems are de- signed to detect the contrast between objects and background in the presence of background illumination, rather than the light intensity itself. Dowling and Chappell [11] performed an electron mic- roscopic study of the dragonfly ocellar plexus and found that L-neurons make observed feedback synapses onto photore- ceptors. They [11] proposed that these feedback synapses play major roles in the temporal highpass filtering. Kling- man and Chappell [35] have studied the effects of various drugs on the response of L-neurons in dragonflies and sug- gested that the receptor synapses are inhibitory (sign- inverting) and curare-sensitive, whereas there are excitatory (sign-conserving) GABAergic feedback synapses from L- neurons which facilitate photoreceptor transmitter release. Stone and Chappell [96] have suggested that hyperpolarizing oscillation at the off-set of illumination in dragonfly photore- ceptors reflects GABAergic feedback synapses onto photore- ceptor terminals. Subsequent studies, however, have failed to confirm the feedback hypothesis for the enhancement of response transience. Simmons [87] made simultaneous in- tracellular recordings from a photoreceptor and an L-neuron of a dragonfly ocellar retina; no response was evoked in the photoreceptors when depolarizing or hyperpolarizing cur- rents were injected into L-neurons, although L-neurons pro- duced responses when current was injected into photorecep- tors. In addition, no immunoreactivity to GABA was observed from L-neurons, at least in locusts [2] and bees [80]. Unfortunately, GABA-immunocytochemical studies have not been performed on dragonfly L-neurons. The response of bee L-neurons is very phasic [25], although no feedback synapses have been observed between L-neurons and recep- tors in the bee [103]. There should be a mechanism to enhance the transience other than the feedback from L- neurons, at least in the bee. Simmons [87] proposed that the intrinsic membrane properties of the photoreceptor terminals and L-neurons, and excitatory synapses made by small second-order neurons on L-neurons are included in the mechanisms to enhance the response transience of dragonfly L-neurons. Ammermiler and Weiler [85] discussed the possible contribution of GABAergic small second-order neurons in a feedback system of the locust. In summary, neural mechanisms underlying the enhancement of response 180 M. MIZuUNAMI transience at first synapses of insect ocelli still to be estab- lished. Dynamics of slow potential responses of second-order neurons The most extensive analysis of responses of ocellar L-neurons have been made in the cockroach by using white- noise modulated light with various mean illuminances (Fig. 4) [56]. The kernels, obtained by cross-correlating the white- noise input against the resulting response, provided a mea- sure of incremental sensitivity as well as of response dynamics (Fig. 5). The first-order kernels, the product of first-order cross-correlation, represent the linear component of the cell’s response, and second- and higher-order kernels, products of second- and higher-order cross-correlation, represent the CONTRAST SENSITIVITY : CROSS-CORRELATOR Madre etegbra’- oom ‘ i WHITE-NOISE LIGHT = FILTER lo te 1 (8) < Vo+ V(t) INCREMENTAL SENSITIVITY Fic. 4. Schematic drawing of experimental procedure to study incremental response of cockroach ocellar L-neurons. A series of ND filters were interposed between the light source and the preparation to attenuate both the mean illuminance and white- noise modulation by the same proportion, so that the “contrast” of the stimulus was kept unchanged. The light signal was monitored before it was attenuated by filters and a correlation was made between the unattenuated light signal and the cellular response. The correlation produces kernels on a contrast sensi- tivity scale. Kernels were converted to an incremental sensitiv- ity scale by multiplying the kernel’s amplitude by the attenuation factor. From Mizunami et al. [56]. Nodal Y 0 Ved t sicibin -20 (mV) ~Vp jt. ne = —— 1 ————E————EE 0 4 8 30 sec Fic. 5. Responses from a cockroach ocellar L-neuron evoked either by steps of light given in the dark or by white-noise-modulated light. The relationship between I, and V, or V, is the cell’s DC (static) sensitivity and the relationship between I(t) and V(t) is incremental sensitivity. Spike potentials are seen at the offset of step stimulation as well as during white-noise stimulation. From Mizunami et al. [56]. nonlinear components of the response. A brief step of light given in the dark produced step-like hyperpolarization in L-neurons. At the off-set of light sti- mulation, L-neurons generate solitary spikes. Here I will concentrate on the slow potential response of L-neurons; their spike response will be discussed in the next section. With continued white-noise stimulation, the membrane potential reaches a steady level within 30sec (Fig.5). At this dynamic steady state, the actual responses of L-neurons to white-noise stimulus can be predicted by a linear model (Fig. 6) with mean-square-errors of about 11%. This indi- cates that the response is practically linear, i.e., the magni- tude of the response is proportional to the depth of modula- tion. The linear nature of the response to intensity changes has been reported from peripheral visual neurons in a variety of visual systems including the photoreceptors of the Limulus compound eye [18], vertebrate retina [5, 61] and insect compound eye [71], as well as the second-order neurons of vertebrate retina [8, 60, 107], the Limulus compound eye [38] and insect compound eye [17]. Linear coding of photic signals thus appear to be a general principle of peripheral visual systems. In Fig. 7, first-order kernels which represent the linear A TIME RECORD PDF 15mV 1.8 0 B POWER (20 dB) 0.1 1 10 100 FREQUENCY (Hz) Fic. 6. Responses of a cockroach ocellar L-neuron to a white-noise modulated light. (A) Time records of part of a white-noise stimulus and the resulting response of a cockroach L-neuron (continuous line). Superimposed on the response trace is the linear model (broken line). Probability distribution function (PDF) for the light stimulus and the recorded response are also shown. The light PDF is also superimposed on the response PDF. (B) Power spectra of the light stimulus, response (con- tinuous line), and linear model (broken line). The mean illumi- nance of the stimulus is 20 «W/cm?. From Mizunami ef al. [56]. Signal Processing in Insect Ocelli 181 A 10 1) = Zz = 0 —| Ww = a WwW < -20 B ] w es a0 — Ww Zz oc uw x -2 Re ee eee eee | 0 0.1 0.2 03 TIME (s) Fic. 7. Comparison of first-order kernels of second-order neurons of cockroach ocelli and vertebrate retinas. (A) First-order kernels from a cockroach ocellar L-neuron, plotted on a contrast sensitivity scale, obtained at five mean illuminance levels. The first-order kernels were calculated by cross-correlating the white- noise light stimuli with the recorded responses. Kernels are labeled 0 through —4 to indicate the log density of the filters interposed. Note that the amplitudes of the kernels did not differ by more than 30% and the peak response times were constant at 50 msec for all kernels, although the mean levels covered a range of 1:10,000. Stimuli dimmer than —4 log units did not produce any reliable results. (B) First-order kernels from a turtle retinal horizontal cell, plotted asin A. The peak response times, waveforms, and amplitudes differed at various levels of mean illuminance. Kernel units are in millivolts per microwatt per square centimeter per second. The larger in- cremental sensitivity for ocellar kernels was due to a dimmer mean illuminance (204W/cm? at 0 log) of the white-noise stimu- lus than in the turtle experiment (S0”W/cm? at Olog). From Mizunami et al. [56]. component of the response obtained at four log ranges of mean luminance levels have been plotted on a contrast sensitivity scale. The waveforms are almost identical, with constant peak response times of about 50 msec, while the amplitudes differ only by 30%. The results show that the response is an exact Weber function, i.e., contrast sensitivity remains unchanged over at least a four log range of mean illuminance levels and that the response dynamics remains unchanged at the same range of mean illuminance levels. For comparison, an example of kernels from a horizontal cell of turtle retina obtained under comparable conditions is shown in Fig. 7, where the amplitude of kernels differs at different levels of mean illuminance, the peak response times shorten from 100 to 50 msec, and the waveforms become more biphasic (differentiating) with an increase in mean luminance. These studies show that (1) the cockroach L- neurons are ideal contrast detectors since their response amplitude is exactly proportional to the contrast of intensity change for at least 4log ranges of mean illuminance levels and (2) signal processing in the cockroach ocellus differs from other visual systems, including Limulus lateral eyes [18], insect compound eyes [13, 71], vertebrate retinas [5, 60], and the human visual system [33], in that the system’s dynamics change depending on the levels of mean illuminance. Fre- quency-response characteristics of L-neurons have also been studied using sinusoidally-modulated light [49, 54]. The response of L-neurons is bandpass with an optimal frequency of 1-3 Hz, a lower cut-off frequency (—3 dB) of 0.05—0.1 Hz and a higher cut-off frequency of 12-15 Hz, indicating that this neuron can respond to slowly occurring events. Interestingly, there are marked differences in in- cremental responses of light-adapted L-neurons among diffe- rent insects. Contrast sensitivity of bee L-neurons [3] is much lower than that of the cockroach [56], i.e., they can not respond to stimuli of small contrast. However, some L- neurons of the bee exhibit a higher cut-off frequency of about 30 Hz [3], and thus can respond to stimuli of much higher frequency than can cockroach L-neurons. These findings are consistent with the hypothesis proposed by Mizunami (in preparation) that the ocellar system of the bee is more concerned with speed than sensitivity, while that of the cockroach is more concerned with sensitivity than speed. In addition, the waveform of the response of bee L-neurons to sinusoidally-modulated light highly deviates from sinusoid [3], indicating that the response contains a high degree of nonlinearity. The bee L-neurons are not suited to faithfully monitoring the stimulus contrast, and are perhaps more related to the initiation of direct behavioral reactions. Furthermore, Simmons [89] has recently measured frequency response characteristics of locust L-neurons using sinusoidally modulated light, and has noted that the optimal frequency, where the contrast sensitivity is highest, changes from 2 Hz to 10 Hz with an increase in mean illuminance levels. Since such changes have not been observed in the responses of cockroach L-neurons [56], ocellar systems of locusts and cockroaches must adopt different strategies to adjust their response dynamics to environmental light intensity levels. Such differences may imply that there are fundamental differences between the behavioral roles of locust and cock- roach ocelli, a possibility which should be examined further. NONLINEAR CENTRAL MECHANISMS FOR THE PROCESSING OF CONTRAST SIGNALS The signals about temporal contrast encoded in the graded responses of ocellar L-neurons are further processed to detect specific visual features. In the following section, I will first discuss signal conversion at the spike initiation process in L-neurons, and then discuss signal processing at synapses from second- to third-order neurons (second synapses) of cockroach ocellar system. 182 Dynamics of the spike initiation process in L-neurons L-neurons of most insects respond to light stimuli with regenerative spikes, in addition to graded slow potentials (cockroaches [55, 57]; bees [45, 46]; locusts [88, 112]). The relationship between the slow potential and spikes of L- neurons have been analyzed in the cockroach using a sinu- soidally modulated light with various mean illuminances [53]. L-neurons generated a solitary spike at the depolarizing phase of the modulation response (Fig. 8). The relationship between the peak-to-peak amplitude of the slow potential response and the rate of spike generation was sigmoidal, with a linear part covering the spike rate ranging from 10 to 90% (Fig.9A). In Fig.9B, the spike rate has been plotted against the amplitude of the slow potential response at five different frequencies. The results at a spike rate of between 10 to 90% are shown. The extrapolated regression lines for each frequency cross the vertical axis at almost the same point, suggesting that the nonlinear threshold is frequency independent. Incontrast, the slope of the lines changes with frequency, which indicates that the spike initiation process contains dynamic linearity. Subsequent studies have shown that the spike threshold at optimal frequency (0.5-5 Hz) remains unchanged over a mean illuminance range of 3.6 log units, whereas the spike threshold at frequencies of <0.5 Hz was lower at a dimmer mean illuminance (Fig. 10A), where the L-neurons exhibited a larger voltage noise and their mean membrane potential levels were more positive (Fig. 10B). Steady or noise current injection during sinusoidal light stimulation showed that (1) the decrease in the spike A 100 ( */o) 50 SPIKE RATE 0 1 2 3 4 3) SLOW POTENTIAL (mV) Fic. 9. SPIKE RATE (%) Dynamics of the spike initiation process in a cockroach ocellar L-neuron. M. MIzuNAMI i / ee 0 3 30 Ss Fic. 8. Responses of a cockroach ocellar L-neuron evoked either by a step light stimulus given in the dark or by a sinusoidally modulated light stimulus. The L-neuron responded to the sinusoidal stimulation with a sinusoidal voltage modulation, V(f), around a mean voltage, V,. A spontaneous voltage fluctuation (voltage noise), V,,, was superimposed on the mod- ulation response. Spikes were seen at the offset of step stimula- tion and at the peak of the voltage modulation. The mean illuminance of the stimulus, J,, was 20 pW-em?; the modula- tion frequency, f, was 2 Hz; the depth of modulation of the stimulus, J (f), was 60%. From Mizunami and Tateda [53]. threshold at a dimmer mean illuminance was due to an increase in the noise variance, i.e., the noise had a facilitatory effect on the spike initiation and (2) the change in the mean potential level had little effect on the spike threshold (Fig. 11). In summary, the spike response of L-neurons is repre- 100 50 3 1 Pikd DB v2 le aye, SLOW POTENTIAL Gil ian (mV) / a TELA a, Gye Yt ee LAD? “5 fo’ (A) The rate of spike generation plotted against the peak-to-peak amplitude of the slow potential response of a cockroach L-neuron, obtained at a frequency of 1 Hz. The form of the curve was sigmoidal, with the linear part covering the range of spike rate from ~ 10-90%. (B) The spike rate plotted against the amplitude of the slow response of a cockroach L-neuron, obtained at five different frequencies. The extrapolated broken lines are the regression lines for each frequency. These lines cross the vertical axis at almost the same point. The stimulus had a mean illuminance of 2 7«W-m~?. From Mizunami and Tateda [53]. Signal Processing in Insect Ocelli 183 A E Qa = fo) 25 ip) WwW a 3c | coal 5 ive} 34 a= T a a T T a 0.2 0.5 1 2 5 10 FREQUENCY (Hz) B 5 mV 10s Dark =) -2 0 Dark Fic. 10. Effects of mean illuminance changes on the dynamics of the spike initiation process in a cockroach L-neuron. (A) 50% threshold of spike response, defined as the peak-to-peak ampli- tude of the potential modulation at a spike rate of 50%, plotted against the modulation frequency. The plots are from the responses of a cockroach L-neuron to sinusoidal lights with a mean illuminance of 0.005 (—3.6log), 0.2 (—21log), and 20 uW-cm * (log). (B) Responses of a cockroach ocellar L- neuron to prolonged illuminations. The light intensities are indicated as logy) attenuation (Olog units=20 ~W-cm ’). From Mizunami and Tateda [53]. ‘i 50% THRESHOLD (mv) w —l 02 ea 2 5 FREQUENCY (Hz) Fic. 11. Effects of noise current and steady current injection on the 50% threshold of a cockroach L-neuron. A noise current having a peak-to-peak amplitude of ~4 nA or a steady depolar- izing current of 4 nA was injected during sinusoidal light stimula- tion. The light stimulus had a mean illuminance of 20 pW-cm~’. The estimated potential change produced by the injection of a 4nA current was ~3 mV (the estimated input resistance at that mean illuminance was ~0.8MQ). The 50% threshold at a low-frequency range decreased when the noise current was injected. The steady depolarizing current had little effect on the 50% threshold. From Mizunami and Tateda [53]. = 10 A “igur | BANDPASS BANDPASS STATIC 5} UINEAR LINEAR |—3| NON- LINEAR PIEMER FILTER FILTER 1s Fic. 12. (A) A model for the spike response of cockroach ocellar L-neurons. Light signals are passed through a bandpass linear filter, producing a slow potential response in L-neurons. The slow potential contains a noise which reflects that contained in the synaptic potential from photoreceptors. The slow potential is further passed through a linear/nonlinear cascade and pro- duces a spike discharge. The linear filter is bandpass, and the nonlinear filter is a static threshold. (B, C) Typical responses of two types of third-order ocellar neurons of the cockroach to sinusoidal light stimulation. One type of third-order neuron, the OL-I neuron (type 1 neuron projecting into the optic lobe), showed sinusoidal modulation of spike frequency (B), whereas the other type, the D-I neuron (type 1 neuron descending to the thoracic ganglia), generated solitary spikes at the decremental phase of light modulation (C). The lower traces indicate the stimulus light, monitored by a photodiode. The stimulus had a modulation frequency of 1 Hz and a modulation depth of 50%. The mean illuminance was 2.W-cm *. From Mizunami and Tateda [53]. sented by a simple cascade model (Fig. 12A). A light stimu- lus is passed through a bandpass linear filter and produces a slow potential response in L-neurons. The slow potential is passed through a cascade of a linear filter followed by a nonlinear filter and produces a spike discharge in L-neurons. The linear filter is bandpass containing both a differentiating and an integrative nature. The nonlinear filter is a static threshold with a sigmoidal (probablistic) input/output rela- tionship. It is concluded that fundamental signal modifica- tions occur during spike initiation in cockroach L-neurons, a finding which differs from the spike initiation process in other visual systems, including the Limulus compound eye [36, 37] and catfish retina [78], in that it is presumed that little signal modification occurs at the analog-to-digital conversion pro- cess. It would be interesting to see if the slow potential or the spike signals of L-neurons, or both, are encoded in the response of third-order neurons. Figs. 12B and C show typical examples of responses of third-order ocellar neurons to sinusoidally modulated light. A type of third-order neuron, OL-I neuron (type I neuron projecting into the optic lobe), had spontaneous spike activity and exhibited a modula- 184 M. MIzuUNAMI tion of the spike frequency around a mean (Fig. 12B). The pattern of the response was similar to the slow potential response of L-neurons. The other type, the D-I neuron (type I neuron descending to the thoracic ganglia), had no spontaneous spike activity and exhibited single spikes at the decremental phase of light modulation (Fig. 12C). The pat- tern of the response was similar to that of the spike response of L-neurons. Thus, it is concluded that both graded and spike signals of L-neurons are encoded in the spike responses of third-order neurons [48, 53]. The graded response of L-neurons appears to continuously monitor the contrast of intensity changes, whereas the spike response possibly signals an urgent event which requires rapid behavioral reactions. Nonlinear signal processing at second synapses Analysis of the signal transmission at synapses between second- and third-order neurons (second synapses) of the insect ocellar system was initiated by Simmons [86] who examined the operations of synapses which L-neurons make with a pair of large descending third-order ocellar neurons, DNI (an ipsilaterally descending neuron), of the locust. Both L-neurons and DNI neurons hyperpolarize when their ocellus is illuminated. L-neurons and DNI neurons produce sharply rising regenerative responses when a bright light is switched off. L-neurons make excitatory (sign-conserving) chemical synapses with the DNI neurons. Under steady daylight illumination, L-neurons continually release transmit- ters onto the DNI neurons. The hyperpolarizing responses of DNI neurons to an increase in illumination are due to a decrease in the rate of release of transmitter from the L-neurons. Further detailed analyses of transfer characteristics of the second synapses of ocellar systems have been made in the cockroach. Mizunami and Tateda [52] identified nine types of interneurons in the cockroach with arborizations in the ocellar tract neuropil. When recordings were made in the ocellar tract, all types of neurons exhibited a similar response to step stimulus given in the dark, 1.e., a tonic hyperpolariza- tion during illumination and one or a few transient depolar- izations at the end of illumination. These neurons can be classified into several physiological types from responses recorded in their axons or terminal regions. Some neurons exhibited spontaneous spike discharge, some neurons had no spontaneous discharge and others generated no spikes to either ocellar illuminations or extrinsic current injections into the neurons [52, Mizunami, personal observation]. Mizuna- mi and Tateda [54] made simultaneous intracellular record- ings from these neurons and L-neurons and found that these neurons receive monosynaptic inputs from L-neurons. The synapses made from L-neurons to these neurons had similar properties to those reported for the synapses made from L-neurons to DNI neurons of locusts. Excitatory (sign- conserving) synaptic transmission was tonically maintained under normal resting potentials, so that hyperpolarizing responses of L-neurons produced hyperpolarizations in third- order neurons. Mizunami [49] further studied the transfer characteristics of synapses made from L-neurons to third-order neurons of the cockroach ocellar system using simultaneous microelec- trode penetrations and the application of tetrodotoxin. The stimulus used was a sinusoidally- modulated light around a mean illuminance or an extrinsic current applied to the L-neurons. The waveform of the response of L-neurons to sinusoidally-modulated light is almost sinusoidal, which indi- cates that the response is linear, but the waveform of the response of third-order neurons deviates from sinusoid and exhibits a half-wave rectification, i.e., the depolarizing re- sponse to light decrement is much larger than the hyperpolar- izing response to light increment (Fig. 13). Analysis of the synaptic transfer curve relating pre- and postsynaptic voltages showed that the synapses made from L-neurons to third-order neurons operate at an exponentially rising part of the overall sigmoidal transfer curve (Fig. 14). Due to the nonlinear characteristics of the synaptic transfer, the linear responses of presynaptic neurons are converted into half-wave rectified responses of postsynaptic neurons (Fig. 15A). 3rd 5 mV 3 mV Fic. 13. Responses of a second- and a third-order ocellar neuron of the cockroach evoked either by step-stimuli given in the dark or by a sinusoidally modulated stimulus around a mean illumi- nance. The sinusoidal stimulus has a modulation depth of 0.7 and a modulation frequency of 2 Hz. Horizontal lines in the records are the steady (DC) potential levels maintained during steady illumination. The lowest trace indicates the stimulus light, monitored by a photodiode. Calibration: 5 mV for the third-order neuron; 9 mV for the second-order ocellar neuron. From Mizunami [49]. The properties of the second synapses of cockroach ocelli were compared to those reported for first synapses of other visual systems. In most visual systems studied thus far, both photoreceptors and second-order neurons exhibit linear re- sponses to changes in intensity, thereby suggesting the linear nature of signal transmission at first synapses. Indeed, stu- dies of first synapses in turtle retina [62], barnacle ocelli [28], dragonfly ocelli [87] and fly compound eyes [43] show that signal transmission occurs at the mid-region of the sigmoidal transfer curve where the transmission is linear. It is con- cluded, therefore, that operation ranges over the synaptic Signal Processing in Insect Ocelli 185 4 = ; r = | | | | | | | | | oe = 4mV = 10 mV 10 nA 9nA 1s 1s (C 6 D > 10 E 000 va ~ {5 a = / =e —€ 74 a = |/ fo} 2 L2 7 5 a Veee (mV) Vpre (mV) Fic. 14. Measurements of transfer characteristics of second synapses in the cockroach ocellar system. (A) Typical records of a current/voltage relationship of a second-order ocellar neuron (L-neuron) measured by impaling the neuron with two electrodes. (B) Responses of a third-order ocellar neuron evoked by current stimuli applied to an L-neuron. Averaged current-voltage relationships from six L-neurons were used to estimate presynaptic potential changes during current stimuli for the experiments in C. Actual input resistance of L-neurons deviates slightly from cell to cell (~ +13%), thus, the estimated synaptic transfer curve may have slight errors. Lower traces in A and B indicate the magnitude of the stimulus current. (C) Input/output voltage relationship of the synaptic transmission. Measurements were made at the steady-state value for 0.5 sec current pulses. V, is the synaptic potential maintained in the dark. V,,- and V,.s are the potentials of the second- and third-order neuron, respectively. The potentials were mea- sured from the dark level, thus, V,os.+V-< is the actual post- synaptic potential. (D) Semilogarithmic plot of the input/ output voltage relationship of the synaptic transmission. From Mizunami [49]. transfer curve differ between first synapses and second synapses (Fig. 15B), thus resulting in fundamental differences in signal transmission, i.e., transmission is nonlinear and half-wave rectifying at second synapses whereas it is linear at first synapses. Rectified responses have been noted in some third-order neurons of a variety of visual systems including ganglion cells of goldfish retina [91] and cat retina [16, 30], third-order neurons of locust compound eyes [32, 66, 67], locust ocelli [86], and barnacle ocelli [99]. The rectified responses seen in third-order neurons of these visual systems can be ex- plained if their second synapses have a nonlinear rectifying nature, as do second synapses of the cockroach ocellar system. The response of cockroach ocellar third-order neurons to brightening is much smaller than that to dimming. In barna- cle ocelli, Stuart and Oertel [99] noted that the response to dimming is enhanced as signals are passed from second- to third-order neurons. Because the major function of barna- cle ocelli is to detect dimming to facilitate a shadow-induced withdrawal of the animal into the shell [26], it is reasonable to A OFF response ON response FIRST SYNAPSE ON OFF response response 4 V pre Fic. 15. Nonlinear signal transmission at second synapses. (A) Signal rectification by nonlinear synaptic transmission from second- to third-order ocellar neurons of the cockroach. Synaptic transmission occurs using an exponentially-rising part of the over-all sigmoidal transfer curve, thus, the depolarizing response to light decrement of presynaptic neurons is amplified, while the hyperpolarizing response to light increment is com- pressed. From Mizunami ((51]). (B) Linear and nonlinear signal transmission at graded synapses. The synapse between second- and third-order neurons (second synapse) of cockroach ocelli operates at an initially rising part of the S-shaped input/ output relationship; thus, the transmission is nonlinear and rectifying. The synapse between photoreceptors and second- order neurons (first synapse) of visual systems presumably operates at a mid-region of the S-curve, where the transmission is essentially linear. From Mizunami [49]. use a large part of the dynamic range to code signal about dimming. In insect ocelli, Stange [93] studied the role of ocelli in visual steering behavior of dragonflies in flight and concluded that a decrease in illuminance of the ocelli has a strong effect on inducing steering behavior to avoid nose- diving toward the ground, whereas an increase in illumination is less effective. It appears that in these simple visual systems, the detection of dimming is more important than that of brightening, and thus dimming-specific responses are formed by removing redundant signals. NEURAL PRINCIPLES FOR THE DETECTION AND PROCESSING OF CONTRAST SIGNALS In this section, I will briefly summarize information 186 M. MIzuNAMI Third-Order Neurons Second-Order Neurons Second Synapses Light Photoreceptors Transduction First Synapses ; Sign- Linear Inverting Se es LOM OBES Linear Filter pioigees= Filter I, R O R(t) = Ry + R(t) R(t) =k log I'(t) R(t) = KI(t/Ig I'(t) =I, + I(t) Light Intensity Temporal Contrast S(t) T(t) S'(t) =S_ + S(t) S'®)= SW S(t) = -k'I(/Io T(0) =To + TY T(t) = T(t) T(t) = a-(exp(-b I(t)/Ip)- 1) Dimming Fic. 16. Summary diagram of signal processing in the cockroach ocellar system. The light-to-voltage transduction process in photoreceptors has a linear lowpass filtering property. The photoreceptor responses encode absolute light intensity and feed into first synapses, which have a sign-inverting, linear lowpass filtering property. The response of second-order neurons, which encodes contrast of intensity changes, feeds into second synapses. The second synapse is static (frequency-independent) and has an exponential input/output relationship. The response of third-order neurons encodes dimming. From Mizunami [51]. processing in the cockroach ocellar system (Fig. 16). The light inputs which enter the ocellar photoreceptors, I’(t), consist of two components, a time-varying component, I(t), and a steady mean, I,. The light-to-voltage transduction process in ocellar photoreceptors is linear and has a lowpass filtering property. Its output, i.e., photoreceptor response R’(t), consists of two components, a time-varying component R(t), and a steady component R,. R(t) is related to I(t), and R, is related to I,. The photoreceptor response then feeds into first synapses, 1.e., synapses between receptors and second-order neurons. The first synapses are linear, sign- inverting, and have a highpass filtering property. The out- put, i.e., the response of second-order neurons S’(t), can be divided into two components, a time-varying part S(t), and a steady mean S,. S(t) is linearly related to the stimulus contrast I(t)/I,. Because of the highpass nature of first synapses, the S, is small. Thus, S’(t) can be written as: $'(t)=S(t)=—K I(t)/Io, (1) where k’(f)>0. To simplify the discussion, the spike re- sponse of L-neurons is not considered here. The response of second-order neurons S’(t), feeds into second synapses. The second synapses are static (frequency-independent) and have an exponential input/output relationship. The response of third-order neurons T’(t), can be written as: T’(t)=T(t)=a(exp (S’(t)/k’”)—1) =a(exp (—b I(t)/I,)—1), (2) where T(t) is a time-varying component of the response of third-order neurons; k”(f)>0; a>0; b(f)>0. Because the exponential filter enhances the response to dimming and compresses the response to brightening, the response of third-order neurons mainly codes for dimming. This model suggests that dimming detection in the insect ocellar system is performed by a cascade of a few processing steps. Each step extracts an aspect of visual signals by removing redundant signals, so that the system can finally detect specific, biologically significant features. It can be argued that this cascade organization may reflect the path- ways by which insect ocellar systems have evolved. The ocellar systems may have originated from simple photorecep- tors whose functions were to detect the distribution of light intensity around the animal, and then first-order interneurons followed, enabling detection of temporal contrast signals representing the movement of self or large objects. Finally, second-order interneurons followed to specifically code for dimming, enabling the evaluation of an approach of large objects to perform an appropriate avoidance behavior. In short, the present neural circuits of the cockroach ocellar system may involve ancestral neural circuits from which the ocellar system has evolved. DETECTION OF TEMPORAL CONTRAST AS A BASIS OF HIGHER VISUAL FUNCTIONS Detection of temporal contrast is a basic pre-processing necessary for the extraction of biologically significant visual Signal Processing in Insect Ocelli 187 information such as color, motion, and shape of objects. Knowledge of the neural mechanisms for the detection of contrast signals obtained in the cockroach ocellar system, therefore, provides a basis for understanding neural mechan- isms subserving advanced visual functions. Mizunami [49, 50] discussed two examples in which neural mechanisms of higher visual function can be explained by simple duplications and modifications of dimming detection circuits which repre- sent the cockroach ocellar system. One is the neural circuit for segregating contrast signals into ON, OFF and ON-OFF channels [49], and the other is the neural circuit for detecting the direction of motion [50], both of which are major components of visual processing in advanced visual systems including insect compound eyes and vertebrate visual sys- tems. Formation of ON and OFF channels Some classes of third-order neurons of vertebrates (retin- al ganglion cells [76-78, 91]) and insect compound eyes [66] exhibit half-wave rectified, ON-depolarizing or OFF- depolarizing responses. These neurons form separate ON and OFF channels which specifically code for light increments and decrements, respectively. Some third-order neurons of vertebrate retinas [76, 77] and insect compound eyes [32, 66, 67] also show full-wave rectified, ON- and OFF-depolarizing responses, which form flicker-sensitive, ON-OFF channels. The reasons that signals for intensity changes are not trans- mitted by linear contrast detectors but by separate ON and OFF channels have been discussed by Shiller et al. [83], who pointed out that by having both ON and OFF channels, signals for both dimming and brightening can be transmitted without maintaining a high rate of spike discharges which require a high rate of metabolic activity. This allows for an economical coding of contrast. Mizunami [50] pointed out that the segregation of contrast signals into ON, OFF and ON-OFF channels can be explained if the synapses from second- to third-order neurons (second synapses) of these visual systems have a rectifying nature, as do those of cockroach ocelli (Fig. 17). Indeed, Toyoda [106] and Miller [47] proposed that the full-wave rectified response of ON- OFF amacrine cells of vertebrate retina can be explained if the cells receive half-wave rectified synaptic input from both ON and OFF bipolar cells. It is concluded that simple modification of dimming detection circuits is sufficient to explain the mechanisms to form ON and OFF channels. Directionally-selective motion detection Advanced visual systems such as insect compound eyes A: Dimming Detection (Formation of OFF-channel) Light Linear Filter OFF-channel B: A Model of Formation of ON, OFF, ON-OFF Channels Light Linear Filter on OFF-channel ON-OFF channel (+)—_- ON-channel Fic. 17. Neural circuits for dimming detection and for formation of ON and OFF channels. (A) A circuitry model of dimming detection in the cockroach ocellar system, consisting of a sign-inverting bandpass linear filter followed by a synaptic rectifier with an exponential input-output relationship. The output of the linear filter, L°*, and that of the rectifier, OFF-channel, represents the response of second- and third-order ocellar neurons, respectively. (B) A model of a contrast detector which comprises the main structure of the movement detection circuit of Fig. 18A. L° and L° are on- and off-depolarizing linear responses, respectively. The outputs of the circuitry consist of three classes: on-, off-, and on-off depolarizing rectified responses. These specifically encode light increment, light decrement and flicker, respective- ly, and are indicated as ON, OFF and ON-OFF channels. Modified from Mizunami [50]. 188 and vertebrate visual systems have movement detectors which code visual motion in a directionally selective manner. Behavioral and psychophysical studies show that movement detection by insects and humans can be represented by a mathematical algorithm referred to as a correlation model (Fig. 18C) [1, 27, 79]. The neural mechanism of this motion computation, however, has not been established as yet. Mizunami [50] described a model mathematically equivalent to the correlation-type movement detector (Fig. 18A). The main structure of the model is comprised of contrast detectors in Fig. 17B which consist of bandpass linear filters followed by synaptic rectifiers. Linear, one-directional, lateral in- teractions are assumed among the contrast detectors. Thus, the basic assumption of this model is that synapses between second- and third-order neurons of movement detection systems are nonlinear and rectifying, as are those of cock- roach ocellar systems. Mizunami [50] showed that synaptic rectifiers convert linear spatial interactions into a multiplica- tion-like (quadratic) interaction, which is the core of the correlation-type movement detector. One of the neural models, which contains both excitatory (additive) and inhibi- tory (subtractive) lateral interactions among contrast detec- M. MIZuUNAMI tors (Fig. 18B), well approximates the correlation model (Fig. 18C, D) in both time-averaged and dynamic (instan- taneous) responses. Some of the basic features of the model agree with those of actual movement detector neurons of insects [50]. Evolutionary perspective I have shown that: (1) step-by-step modifications of simple photoreceptors are sufficient to explain the evolution of dimming-detection circuits of cockroach ocelli; (2) simple modifications and duplications of dimming-detection circuits are sufficient to explain contrast coding circuits with separate ON and OFF channels; and (3) simple modifications and duplications of contrast coding circuits with separate ON and OFF channels are sufficient to explain neural circuits for motion detection. These findings imply that complex neural circuits subserving advanced visual functions have evolved through step-by-step modifications of simper neural circuits subserving simpler visual functions, and that the present complex neural circuits contain, at least in part, simple neural circuits from which the present neural circuits have evolved. I conclude that careful comparison of neural circuits of L, and J, are inputs to left and right detectors. L- is a sign-inverting bandpass linear filter, the output of which approximates the response of lamina monopo- lar cells. eis a sign-conserving lowpass linear filter. R is a synaptic rectifier with an exponential input/ output relationship. Lateral interaction and rec- tification may take place in the medulla neuropil. A B Stationary Preferred Null I,_ 1 I, I E-] model Ne ae s POOF FSGS. Ik | 16 wy \_) RETINA po" coe pe eke . . MEDULLA C D [R] Preferred Ina I LOBULA D PLATE Null I, Ip IPAS GRY N D D Fic. 18. Computational and neural models for motion detection. tional-selective motion detection. (right). directional linear interactions (A) A model of neural circuits for direc- Possible location of each process in a fly’s compound eye is illustrated The model (E-I model) consists of one- among contrast- detection circuits of Fig. 17B. The model involves both excitatory and inhibitory lateral interactions. On- and off-depolarizing linear responses feed into rectifiers, producing on- and off-depolarizing rec- tified responses, R°° and R°. R°" and R™ are assumed to represent responses of ON- and OFF- EMDs (elementary movement detectors). The final outputs of the models consist of three classes, D°”, D° and D°™°", The model outputs may represent responses of motion-sensitive neurons of the lobula plate. P is the preferred direction of motion. (B) Responses of the E-I model to a stationary flickering light and to a sinusoidal grating moving in the preferred and null directions. The ON-OFF motion detéctor (D°"°" in A) exhibits a steady response to motion, whereas responses of ON and OFF motion detectors (D°" and D°") oscillate, depending on the spatial phase of the stimulus pattern. The stimulus parameters are: 1) for response to motion in the preferred direction, the phase lag due to the spatial separation of left and right input channels, ¢,, is 2/2 and the phase lag due to the delay in lateral interac- tion, ¢, , is 7/2; 2) for response to motion in the null direction, &=2/2 and ¢,=— 7/2; and 3) for re- sponse to a stationary flickering light, 8 =0 and ¢,= m/2. (C) The correlation model proposed by Has- senstein and Reichardt [27]. L* and « are sign- conserving, lowpass linear filters. M is a multiplier. D is the final output. (D) Responses of the correla- tion model to a sinusoidal grating moving in the preferred and null directions. The stimulus para- meters are the same as for (B). From Mizunami [SO]. Signal Processing in Insect Ocelli 189 different animals is an effective approach to understanding possible evolutionary pathways by which complicated neural circuits of animal brains have formed. I thank Ms. Janet Kramer for helpful comments. ACKNOWLEDGMENTS My work is supported in part by the Uehara Memorial Foundation, the Naitoh Science Foundation and by grants from the Ministry of Education of Japan. ONDA N REFERENCES Adelson EH, Bergen JR (1985) J Opt Soc Am A2: 284-299 Ammermiiler J, Weiler R (1985) J Comp Physiol 157: 779- 788 Baader A (1989) J Neurobiol 20: 519-529 Baylor DA, Fuortes MGF, O’Bryan PM (1971) J Physiol 214: 256-294 Baylor DA, Hodgkin AL (1974) J Physiol 242: 729-758 Chappell RL, DeVoe RD (1975) J Gen Physiol 65: 399-419 Chappell RL, Dowling JE (1972) J Gen Physiol 60: 121-147 Chappell RL, Naka K -I, Sakuranaga M (1985) J Gen Physiol 86: 423-453 Cornwell PB (1955) J Exp Biol 32: 217-237 Cooter RJ (1975) Int J Insect Morphol Embryol 4: 273-288 Dowling JE, Chappell RL (1972) J Gen Physiol 60: 148-165 Dreisig H (1980) Physiol Entomol 5: 327-342 Dubs A (1981) J Comp Physiol 144: 53-59 Eaton JL (1976) J Comp Physiol 109: 17-24 Eaton JL, Tignor KR, Holtzman GI (1983) Physiol Entomol 8: 371-375 Enroth-Cugell C, Freeman AW (1987) J Physiol 384: 49-79 French AS, Jarvilehto M (1978) J Comp Physiol 126: 87-96 Fuortes MGF, Hodgkin AL (1964) J Physiol 172: 239-263 Goldsmith TH, Ruck PR (1958) J Gen Physiol 41: 1171-1185 Goodman LJ (1965) J Exp Biol 42: 385-407 Goodman LJ (1981) In “Handbook of sensory physiology”. vol VII 6C. Eds by H Autrum, Springer-Verlag, Berlin, Heidelberg, New York, pp 201-286 Goodman CS, Williams JLD (1976) Cell tissue Res 175: 203- 226 Gould JL (1975) Science 189: 685-693 Griss C, Rowell CHF (1986) J Comp Physiol 158: 765-774 Guy RG, Goodman LJ, Mobbs PG (1979) J Comp Physiol 134: 253-264 Gwilliam GF, Stuart AE (1990) J Exp Biol 151: 83-107 Hassenstein B, Reichardt W (1956) Z Naturforsch 11b: 513- 524 Hayashi JH, Moore JW, Stuart AE (1985) J Physiol 368: 175- 195 Hesse R (1908) Das Sehen der neideren Tiere. Jena Hochstein S, Shapley RM (1976) J Physiol 262: 265-284 Hu KG, Reichert H, Stark WS (1978) J Comp Physiol 126: 15-24 Jansonius NM, Hateren JH van (1993) J Comp Physiol 172: 467-471 Kelly DH (1971) J Optical Soc Am 61: 537-546 Kirschfeld K, Lutz B (1977) Z Naturforsch 32c: 439-441 Klingman A, Chappell RL (1978) J Gen Physiol 71: 157-175 Knight BW (1972a) J Gen Physiol 59: 734-766 G Fischer, 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 Knight BW (1972b) J Gen Physiol 59: 767-778 Knight BW, Toyoda J -I, Dodge FA (1970) J Gen Physiol 56: 421-437 Koontz MA, Edwards JS (1984) Cell Tissue Res 236: 133-146 Lall AB, Trouth CO (1989) J Insect Physiol 35: 805-808 Laughlin SB (1981) In “Handbook of sensory physiology”, vol. VII/6B. Ed by H Autrum, Springer-Verlag, Berlin, Heidelberg, New York, pp 135-280 Laughlin SB, Hardie RC (1978) J Comp Physiol 128: 319-340 Laughlin SB, Howard J, Blakeslee B (1987) Proc R Soc Lond B 231: 437-467 Littlewood PMH, Simmons PJ (1992) J Comp Neurol 325: 493-513 Milde JJ (1981) J Comp Physiol 143: 427-434 Milde JJ (1984) J Comp Physiol 154: 683-693 Miller RF (1979) In “The Neuroscience: Fourth Study Prog- ram”. Eds by FC Schmitt, FG Worden, The MIT press, Cambridge, pp 227-245 Mizunami M (1989) In “Neurobiology of Sensory Systems” Eds by NR Singh, NJ Strausfeld, Plenum Press, New York, pp 71-84 Mizunami M (1990a) J Gen Physiol 95: 297-317 Mizunami M (1990b) Biol Cybern 64: 1-6 Mizunami M: Adv Insect Physiol (in press) Mizunami M, Tateda H (1986) J Exp Biol 125: 57-70 Mizunami M, Tateda H (1988a) J Gen Physiol 91: 703-723 Mizunami M, Tateda H (1988b) J Exp Biol 140: 557-561 Mizunami M, Yamashita S, Tateda H (1982) J Comp Physiol 149: 215-219 Mizunami M, Tateda H, Naka K -I (1986) 275-292 Mizunami M, Yamashita S, Tateda H (1987) J Exp Biol, 130: 259-274 Mizunami M, Weibrecht JM, Strausfeld NJ (1993) In “Biolo- gical neural networks in invertebrate neuroethology and robo- tics” Eds by RD Beer, R Ritzmann, T McKenna, Academic Press, San Diego, pp 199-225 Mote MI, Wehner R (1980) J Comp Physiol 137: 63-71 Naka K -I, Chan RY, Yasui S (1979) J Neurophysiol 42: 441- 454 Naka K -I, Itoh M -A, Chappell RL (1987) J Gen Physiol 89: 321-337 Normann RA, Perlman I (1979) J Physiol 286: 509-524 Ohyama T, Toh Y (1986) J Exp Biol 125: 405-409 Ohyama T, Toh Y (1990a) J Comp Neurol 301: 501-510 Ohyama T, Toh Y (1990b) J Comp Neurol 301: 511-519 Osorio D (1987) J Comp Physiol 161: 431-440 Osorio D (1991) Visual Neurosci 7: 345-355 Pan KC, Goodman LJ (1977) Cell Tissue Res 176: 505-577 Pappas LG, Eaton JL (1977) J Insect Physiol 23: 1355-1358 Patterson JA, Chappell RL (1980) J Comp Physiol 139: 25-39 Pinter RB (1972) J Comp Physiol 77: 383-397 Purple RL, Dodge FA (1965) Cold Spr Harb Sym. Quant Biol 30: 529-537 Reichert H, Rowell CHF, Griss C (1985) Nature 315: 142- 144 : Rence BG, Lisy MT, Garves BR, Quinlan BJ (1988) Physiol Entomol 13: 201-212 Rowell CHF, Reichert H (1986) J Comp Physiol 158: 775-794 Sakai H, Naka K -I (1987a) J Neurophysiol 58: 1307-1328 Sakai H, Naka K -I (1987b) J Neurophysiol 58: 1329-1350 Sakuranaga M, Ando Y -I, Naka K -I (1987) J Gen Physiol 90: 229-259 J Gen Physiol 88: 190 84 85 86 87 88 89 90 91 92 93 94 M. MIZuUNAMI Santen DPH van, Sperling G (1985) J Opt Soc Am A2: 300- 321 Schafer S, Bicker G (1986) J Comp Neurol 246: 287-300 Schricker B (1965) Z Vergl Physiol 49: 420-458 Schuppe H, Hengstenberg R (1993) J Comp Physiol 173: 143-149 Shiller PH, Sandell JH, Maunsell JHR (1986) Nature 322: 824-825 Shaw SR (1979) In “The Neurosciences: Fourth study prog- ram”. Eds by FO Schmitt, FG Worden, MIT press, Cam- bridge, pp 275-295 Simmons PJ (1980) J Exp Biol 85: 281-294 Simmons PJ (1981) J Comp Physiol 145: 265-276 Simmons PJ (1982a) J Comp Physiol 147: 401-414 Simmons PJ (1982b) J Comp Physiol 149: 389-398 Simmons PJ J Comp Physiol (1993) 173: 635-648 Simmons PJ, Littlewood PMH (1989) J Comp Neurol 283: 129-142 Spekreijse H (1969) Vision Res 9: 1461-1472 Sprint MM, Eaton JL (1987) Ann Entomol Soc Am 80: 468- 471 Stange G (1981) J Comp Physiol 141: 335-347 Stange G, Howard, J (1979) J Exp Biol 83: 351-355 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 Stockbridge N, Ross WN (1984) Nature 309: 266-268 Stone SL, Chappell RL (1981) Brain Res 221: 374-381 Strausfeld NJ (1976) Atlas of an insect brain. Springer- Verlag, Berlin Heidelberg New York Strausfeld NJ, Bassemir UK (1985) Cell Tissue Res 240: 617- 640 Stuart AE, Oertel D (1978) Nature 275: 287-290 Taylor CP (1981a) J Exp Biol 93: 1-18 Taylor CP (1981b) J Exp Biol, 93: 19-31 Toh Y, Hara S (1984) J Ultrastruct Res 86: 135-148 Toh Y, Kuwabara M (1975) J Neurocytol 4: 271-287 Toh Y, Sagara H (1984) J Ultrastruct Res 86: 119-134 Toh Y, Tateda H (1991) Zool Sci 8: 395-413 Toyada J -I (1974) J Gen Physiol 63: 214-234 Tranchina D, Gordon J, Shapley R (1983) J Gen Physiol 82: 573-598 Weber G, Renner M (1976) Cell Tissue Res 168: 209-222 Weckstrém M, Juusola M, Laughlin SB (1992) Proc R Soc Lond B 250: 83-89 Werblin FS (1972) Ann NY Acad Sci 193: 75-85 Wilson M (1978a) J Comp Physiol 124: 297-316 Wilson M (1978b) J Comp Physiol 124: 317-331 ZOOLOGICAL SCIENCE 11: 191-202 (1994) Output Effect of Identified Ascending Interneurons upon the Abdominal Postural System in the Crayfish Procambarus Clarkii (Girard) Hitosu1 Aonuma’, TosHIKI Nacayama! and MiruHiko HIsaDA ‘Animal Behaviour and Intelligence, Division of Biological Sciences, Graduate School of Science, Hokkaido University, Sapporo, 060 and, *Faculty of Industrial Science and Technology, Science University of Tokyo, Oshamanbe, Hokkaido, 049-35, Japan ABSTRACT—The output effects of 16 identified ascending interneurons originating in the terminal abdominal ganglion were examined with intracellular recording and stimulating techniques in isolated nerve cord composed of 6 abdominal ganglia (from the 1st to 6th ganglion). The activity of the uropod closer and opener motor neurons was recorded extracellularly with pin electrodes from the terminal abdominal (A6) ganglion and that of the abdominal extensor and flexor motor neurons was recorded from the 1st to 5th abdominal (A1-A5) ganglion. This technique allowed us to monitor the activity of the motor neurons innervating different muscles of more than 10 simultaneously in the same preparation. Majority of ascending interneurons had output effects upon not only the uropod motor neurons but also the abdominal postural motor neurons. The premotor effects of the ascending interneurons were the same in all of abdominal ganglia (A1-A5). Some ascending interneurons also affected the abdominal postural motor neurons on both sides of each ganglion with a similar fashion. Neurobiotin staining revealed that the ascending axons spreaded their branches in each abdominal ganglion. Their branches were extended within the side ipsilateral to their axons. The possible function of ascending interneurons as multi-functional units in the sensory-motor system of crayfish was discussed. Since they received sensory inputs from the tailfan and affected the activity of both uropod and abdominal postural motor neurons simultaneously, they would coordinate the behavioural sequence controlling both the uropod motor system and © 1994 Zoological Society of Japan abdominal postural system. INTRODUCTION The central nervous system (CNS) of arthropod animals, like insects and crustacean, consists of a series of segmental ganglion chained by a pair of connectives [e.g. 1]. Each ganglion is bilaterally symmetrical and contains most of motor neurons for its relevant segment. A segmental move- ment is basically controlled by local circuit within its relevant ganglion [2], then a series of movement and postural changes must be coordinated by intra- and intersegmental activation of different muscles [16, 31, 32]. For example, crayfish avoidance reaction consists of serially ordered behavioural acts [26]. Unilateral mechanic- al stimulation of the tailfan elicited a rapid closing movement of uropods followed by the completion of locomotor acts with a change in abdominal posture and pattern generation of walking legs. This assembly of elementary acts must be activated sequentially at the level of central neurons, since few afferents projected anteriorly through the abdominal nerve cord [11]. Neural elements controlling uropod motor system and abdominal postural system have been so far analyzed respectively. Motor neurons in each system have been identified [24, 36 in uropod motor neurons; 8, 17, 40 in abdominal postural motor neurons]. A vast number of premotor interneurons which affected the activity of motor Accepted March 25, 1994 Received Feburary 3, 1994 neurons has been also characterized both physiologically and morphologically [20—22, 25 in uropod system; 13,18, 19, 37 in abdominal postural system]. Many of earlier behavioural and physiological works have expected that certain inter- neurons would contribute to the segmental linking between uropod and abdominal movement, though no attempt has been carried out to clarify this point. About 65 pairs of ascending interneurons originating in the terminal abdominal ganglion of the crayfish [10, 33] have ascending axons through the anterior abdominal connective and receive sensory inputs directly from the tailfan [23, 34]. Twenty-four ascending interneurons are identified as unique individuals, and many of them affect the activity of the antagonistic sets of uropod motor neurons [21]. These iden- tified ascending interneurons should be, therefore, the most possible candidate to mediate intersegmental coordination between uropod and abdominal movement. The present study examined this hypothesis to characterize their output effects upon the abdominal postural motor neurons. Our results show the majority of ascending interneurons have premotor effects upon the abdominal postural motor neurons from the 1st to Sth abdominal ganglion as well as upon the uropod motor neurons. The homologous postural motor neurons throughout the anterior abdominal ganglia are affected in a similar fashion by ascending interneurons. 192 H. AonuMA, T. NAGAYAMA AND M. HISADA MATERIALS AND METHODS Animals and preparations Adult male and female crayfish, Procambarus clarkii (Girard) (5-9 cm body length from rostrum to telson) were used in all experiments. They were obtained commercially and maintained in laboratory tanks before use. Abdominal nerve chain including all abdominal (A1-A6) ganglia with relevant nerve roots was isolated from the abdomen. This preparation was pinned, ventral side up, to the floor of a Sylgard- lined Petri dish and perfused continuously with cooled physiological saline [38]. Extracellular recording and stimulation The activity of motor neurons innervating closer and opener muscles of exopodite was recorded extracellularly by using pin electrodes from the 2nd and 3rd motor root in the terminal (6th) abdominal ganglion respectively [25]. In the anterior abdominal ganglia (from the Ist to 5th ganglion), the activity of tonic extensor motor neurons was recorded from the 2nd root and the activity of tonic flexor motor neurons was recorded from the superficial 3rd root of each ganglion. The pin electrodes were placed contact with each nerve root and insulated with petroleum jelly (Vaseline : liquid paraffin=3:1 in the volume) to perform multi-recording from a single preparation (Fig. 1). Fic. 1. The picture of experimental arrangement with extracellular and intracellular electrodes. The abdominal nerve cord was dissected out from the 1st to terminal abdominal (A6) ganglion. The motor activities of the uropod and abdominal postural system were monitored from relevant motor nerve root with pin electrodes. Electrical stimulation was delivered to the 2nd root afferents of the terminal abdominal ganglion by the bipolar electrode. The intracellular recording and stimulation was made from the terminal abdominal ganglion. Each motor neuron of the tonic abdominal postural system was identified by its spike amplitude and firing pattern of the extracellular recordings [4, 7, 8, 40]. Six tonic extensor motor neurons (EX) and 6 tonic flexor motor neurons (FL) were numbered in order of increasing spike size (i. e. No.1 is the smallest and No. 6 is the largest). The second largest spike in each extensor or flexor motor neuron was the peripheral inhibitor, No.5. In in vitro preparation, the peripheral inhibitor of the extensor motor neurons (No. 5) showed tonic discharge, although the largest unit of the extensor excitors (No. 6) did not fire spontaneously [37]. On both of the extensor and flexor excitor units, No. 3 and 4 and No. 1 and 2 were similar in amplitude within each grouping and thus were distinguish- able as single unit, i. e., there were two No. 3/4 and two No. 1/2 spikes [15]. The premotor effect of ascending interneurons upon the abdominal postural motor neurons was judged to be excitatory if depolarizing current injected into the interneuron elicited or in- creased tonic spikes of excitors (No. 1/2, No. 3/4, and No. 6) and/or decreased tonic spikes of the inhibitor (No. 5). Inhibitory interac- tion was identified by the decrease in tonic spikes of excitors and/or the increase in tonic spikes of inhibitor. When the current injection caused no change in activity of motor neurons, the ascending interneuron was judged to have no obvious output effect. The mechanosensory afferents (2nd root) innervating the right exopodite [3] were stimulated electrically by the bipolar stimulating pin electrode. Square pulses of 0.1ms duration at 20 Hz were delivered to the stimulating electrode. Stimulus intensity was deter- mined to be appeared that spikes of closer motor neuron increased and spikes of opener motor neurons decreased. Intracellular recording and staining Intracellular recordings were made with glass microelectrodes filled either with a 3% solution of Lucifer yellow CH [35] with 0.1 M lithium chloride (100 to 150 MQ resistance) or a 3% solution of neurobiotin [9] with 1M potassium chloride (40 to 80 MQ resist- ance). Interneurons were always impaled in the right half of the neuropiler processes or axons in the terminal abdominal ganglion. A constant polarizing current could be injected into interneurons through the recording electrode by a bridge circuit to characterize their premotor effects upon both the uropod and abdominal postural motor neurons. In most records, the monitor of membrane poten- tial of ascending interneurons during current injection, especially that of depolarization, was difficult because of extremely high resistance, so only the monitor of current was displayed. After physiological examination, Lucifer yellow was injected into the interneuron with hyperpolarizing current pulses of 10 nA of 500 msec in duration at 1 Hz for 20min. The ganglion was then fixed in a 10% formalin for 20 min, dehydrated with an alcohol series and cleared with methyl salicylate. Ascending interneurons were observed using a fluorescence microscope in whole mount, and photographed for subsequent reconstruction. Neurobiotin was injected into the interneuron with depolarizing current pulse of 10 nA of 500 msec duration at 1 Hz for 3hr. The preparations were then diffused for 10 hr at room temperature and fixed in a 10% formalin over 1 hr. They were dehydrated and immersed in methyl salicylate for 30 min to increase staining intensity and reduce background staining. They were rehydrated with an alcohol down series, rinsed with detergent A: 0.01% Triton X-100 and 0.01% Tween 20 in 0.15 M sodium phosphate buffer solution (pH 7.4) (PBS), and immersed in detergent B (2% Triton X-100 and 2% Tween 20 in PBS) for 90 min. After rinsed with detergent A 3 times for 5 min each, they were incubated in HRP conjugated streptavidin for 30 min. They were rinsed with detergent A 3 times for 15 min each, immersed in 0.025% diaminobenzidine (DAB) in PBS for 20 min and reacted with DAB and H,O, (0.003%) in PBS. Crayfish Ascending Interneurons 193 After rinsed with DW several times, they were dehydrated and cleared. Ventral up views of the dye-filled cells were drawn with the aid of a camera lucid. Each ascending interneuron in this study was identified as the criteria described previously [21]. According to their gross mor- phology (including soma position, number of main branches and axonal projection in the connective) and physiological properties (input from sensory afferents and output to uropod, closer and Opener motor neurons), 24 ascending interneurons were divided into 6 classes; co-activating (CA), co-inhibiting (CI), reciprocally closing (RC), reciprocally opening (RO), no effective (NE) and variably effective (VE) interneurons. In total, 103 crayfish were studied in this paper and 78 ascending interneurons were identified and analy- zed their output effect upon abdominal postural motor neurons. Other types of ascending interneurons were encountered only once and were not described here. All the recordings were stored on a digital tape recorder A BeAceGce Occurrence dart turn tail flip no Reaction C Extensor motor neurons Occurrence 0 20) 4K) BO» -feHO) 100% Al A2 A3 A4 LLL LLL MLL Excitatory response Neutral response (Biologic DTR-1801) and displayed using a chart recorder (Gould TA240S). RESULTS Serially ordered motor pattern elicited by stimulation of the uropod Mechanical stimulation of the exopodite in small crayfish (less than 8 cm in length from rostrum to telson) frequently produced avoidance “dart” response (P<0.0001 with Chi square test) (Fig. 2A). The animals showed bilateral closing of uropods and forward walking with abdominal extension. Repetitive electrical stimulation of the 2nd root afferents which innervated hairs on the surface of the exopodite increased the spike frequency of the closer motor neurons and decreased that of the opener motor neurons on the A4 EX lL ASC INT A Yo j1omv 0.2sec D Flexor motor neurons Occurrence 0» B@ 4OorE0 SO We LLL a Inhibitory response [_]No response Fic. 2. Effect of sensory inputs upon the abdominal postural motor neurons. A. Patterns of crayfish reactions in response to unilateral mechanical stimulation of the exopodite. Twenty-five animals were used in the test and 20 trials were made in each. Observed reactions of the animals were categorized as one of four types: dart, turn, tailflip and no response [26]. B. Typical pattern of spike activity of motor and interneurons in response to the repetitive electrical stimulation (20 Hz) of the 2nd root afferent innervating the exopodite (indicated by arrowheads). The 1st and 2nd traces were extracellular recordings from the uropod, closer and opener motor neurons in the terminal abdominal ganglion. The 3rd and 4th traces were from the extensor and flexor motor neurons in the 4th abdominal ganglion. The spike unit indicated by asterisks in the 3rd trace was tonic extensor inhibitor (No. 5). The 5th trace was intracellular recording of the ascending interneuron (ASC INT) from the terminal ganglion. C and D. Pattern of the activity of the abdominal postural motor neurons (C: tonic extensor motor neurons, D: tonic flexor motor neurons) from the 1st to 5th abdominal ganglion (A1-A5) in response to electrical stimulation of the 2nd root afferents. The results were based on 21 preparations. The pattern of motor response was judged to be excitatory if stimulation of the afferents elicited or increased the spikes of excitors and/or decreased the spikes of the inhibitor. Inhibitory response was identified by the decrease in the spikes of excitors and/or the increase in the spikes of inhibitor. Neutral response was defined as when both the excitatory and inhibitory motor neurons were excited simultaneously. 194 TABLE 1. motor neurons and uropod motor neurons H. AonuMA, T. NAGAYAMA AND M. HiIsaDa Summary of output effects of the identified ascending interneurons upon the abdominal postual ASC INT n = FL EX rm FL EX ‘ad FL EX zo FL EX “ Bere ‘ op INPUT CA-1 i Us A as OP i Op Aiperm: — BPSt Cl-1 a BE Ee OE ea aa I ‘EPSP RC-2 6 eB ot, BA TR. iE asetl iee need ie RC-3 3. FE. BE Bo Eo E/E sottbebeE ence ei EueeeEese RC-4 3. I Bol. Ee. .. © ototoxic (rie rePaN ieee RC-6 6. .E.. B .BoosE EF EB eee beri eee Geers RC-7 aaa i Oe Se Ge A AN Se kee Ns kt | ESE RC-8 Pe BPE EEE! EE ep Ee eerie eI RO-2 SET Me UE I | ee ee ee eee ee ee RO3 2 PE BSE ae pee Ee tn ee RO-4 niles eat Bi illness = TEST RO-5 Dial Ziel) Sinise! Steel ell ain SSMS iE . =EPST RO-6 Soe of aE Ee ee ek ee rT I. , Eo @eBRSE VE-1 [OsabalonTSichsb-nbbel abd SieeladeraTiaome Ey aT IG Bic. yecki-ge Ol con ylllu eV N TES NE-1 7 EB. Eo es O68 TE Beeps die 05 wees sin a oe aN ee ESE NE-2 3 NN NN ONO NGA ON coh NicowN GdNae gNing oN DaN ESE n: Sampling number of ascending interneurons E: Excitatory effect; output effect upon the abdominal postural motor neurons was judged to be excitatory if current injection elicited or increased the spikes of excitors and/or decreased the spikes of inhibitor. I: Inhibitory effect; inhibitory effect was identified by the decrease in the spikes of excitors and/or the increase in the spikes of inhibitor. V: Variable effect; when the current injection caused inconstant change in the activity of motor neurons, the ascending interneuron was judged to be variable effect. N: No effect; when the current injection caused no change in the activity of motor neurons, the ascending ?: No record interneuron was judged to be no effect. electrical stimulation (20 Hz) was applied. *: Interneurons showed antifacilitation when repetitive Input of each interneuron was characterized by the electrical stimulation of the 2nd root afferents of terminal abdominal ganglion on the side ipsilateral to axons of interneurons. stimulating side (Fig. 2B). At the same time, the activity of the abdominal postural motoneurons was also changed by the electrical stimulation. On the extensor motor neurons from the 1st to Sth abdominal ganglion, inhibitory motor neuron (No. 5) decreased the spike frequency and excitatory motor neurons increased the spike frequency (P<0.0001 with Chi square test) (Fig. 2C). The response of the flexor motor neurons was somewhat variable, but the spike frequency of the flexor inhibitor (No. 5) was usually increased (Fig. 2D). Thus, mechanosensory stimulation elicited abdominal exten- sion-like motor pattern in the anterior abdominal ganglia as well as closing pattern of the uropod in the terminal abdomin- al ganglion. Output effects of ascending interneurons upon abdominal postural motor neurons Since the majority of branches in the sensory afferents ended in the terminal abdominal (A6) ganglion [11], the change in the activity of the abdominal postural motor neurons induced by sensory stimulation (Fig. 2C and D) was mediated by certain interneurons with intersegmental projec- tion. Ascending interneurons originating in the terminal abdominal ganglion had intersegmental ascending axon and received sensory inputs directly from the afferents [21]. In this study, 16 identified ascending interneurons were char- acterized by their premotor effects upon both the extensor and flexor motor neurons from the Ist to 5th abdominal (A1-AS5) ganglion (Table 1). Of 16 identified ascending interneurons, 15 interneurons had output effect upon the abdominal postural motor neurons. Seven interneurons excited both the extensor and flexor motor neurons co-actively. Four interneurons had excitatory effects upon the extensor motor neurons and inhibitory effects upon the antagonistic flexor motor neurons. Another four interneurons had excitatory effects upon the flexor motor neurons and inhibitory effects upon the extensor motor neurons. One identified interneuron (NE-2) had no obvious effect upon the abdominal postural motor neurons. Interneurons with co-activating effect Ascending interneurons identified as CA-1, CI-1, RC-3, RC-6, RC-8, RO-5 and NE-1 [21] activated both the extensor and flexor motor neurons (Table 1). For example, CA-1 was identified by its soma location of rostralateral region and two main neurites projected laterally (Fig. 3A). Physiologically, this interneuron excited both the closer and opener motor neurons (not shown). In the Ist abdominal ganglion, tonic spikes of extensor inhibitor (No. 5) were inhibited by the passage of depolarizing current (top in Fig. 3B). At the same time, this interneuron increased the Crayfish Ascending Interneurons 195 200um B Al EX FL A2 EX FL A3 EX jane ae cr neon A4 EX AL CUR 5nA 0.5sec Fic. 3. Output effect of CA-1. A. Morphology of CA-1 in the terminal ganglion. Interneuron was drawn within the outline of the ganglion to show its relative position in the ganglion. B. Output effect of CA-1 upon both the extensor (EX) and flexor (FL) motor neurons in the abdominal (A1-A4) ganglia. In this paper, the spike unit indicated by asterisks was tonic extensor or flexor inhibitor (No. 5) and circles was the largest spike unit of excitor (No. 6). spike frequency of the flexor excitors (No. 1/2 and 3/4). In the 2nd abdominal ganglion, this interneuron increased the spike frequency of the extensor excitors (No. 3/4). At the same time, the flexor excitors spiked and tonic spikes of the flexor inhibitor were suppressed during current injection. In the 3rd abdominal ganglion, tonic spikes of extensor inhibitor (No. 5) were inhibited while those of flexor excitors (No. 1/2 and 3/4) were increased. In the 4th abdominal ganglion, this interneuron increased tonic spikes of excitors of both the extensor and flexor motor neurons. Thus, CA-1 excited the excitors of both the extensor and flexor motor neurons and inhibited the extensor inhibitors of from the Ist to 4th abdominal ganglion, though we could not test the effect of the interneuron upon the 5th abdominal ganglion (Fig. 3B). NE-1 was characterized by its thick axon and the limited extent of main branches (Fig. 4A). Neurobiotin staining showed that the ascending axon ran through, at least, the 4th abdominal ganglion. In the 5th abdominal ganglion, several small branches projected mainly medially from the axon. No axonal branches, however, crossed the midline. Phy- siologically, NE-1 had no obvious output to the uropod motor neurons. The spike frequency of either closer or opener motor neurons did not change significantly even if depolariz- ing current of high intensity was injected into the interneuron (12 nA in Fig. 4B). This interneuron, however, affected the abdominal postural motor neurons in the anterior abdominal ganglia. In each abdominal ganglion (A1-A5), the passage of depolarizing current increased the spike activity of both the extensor and flexor excitors (Fig.4C). Thus, NE-1 had output effect upon the abdominal postural motor neurons, though it had no effect upon the uropod motor neurons. NE-2 (not shown) was another identifiable ascending inter- neuron which had no output effect upon the uropod motor neurons [21]. In this study, NE-2 was obtained three times but no output effect upon the abdominal postural motor neurons was recognized. The other interneurons, CI-2, RC-3, RC-6, RC-8 and RO-5 had similar co-activating effect upon the antagonistic extensor and flexor motor neurons, though their effects upon the uropod motor neurons were variable (Table 1). Interneurons with extension effect Four identified interneurons, RC-2, RC-7, RO-6, and VE-1 [21], elicited abdominal extension-like motor pattern. For example, RC-2 was characterized by its looped primary neurite from the soma (A6 in Fig. 5A) and the output effect of reciprocally closing pattern upon the uropod motor neurons (Fig. 5B). This interneuron could produce closing movement of exopodite. Neurobiotin staining revealed the morphology of ascending axon through the 3rd abdominal ganglion (Fig. 5A). Many small branches extended both medially and laterally within the axon side of the interneuron in each anterior ganglion. When the depolarizing current was injected into RC-2, this interneuron increased the spike frequency of the flexor inhibitor (No.5) in all anterior abdominal ganglia (A1-A5) with the inhibition of the spikes in the flexor excitors (Fig.5C). At the same time, this interneuron significantly increased the spike frequency of the extensor excitors, especially those in the posterior ganglia (A4 and A5). In the 1st abdominal ganglion, this inter- neuron decreased the frequency of small extracellular spikes (No. 1/2) in the extensor excitors but increased that of intermediate spikes (No. 3/4) (top in Fig. 5C). RC-7 had similar output effect to RC-2 upon both the uropod and abdominal postural motor neurons, while RO-6 had reversed effect upon the uropod motor neurons (Table 1). VE-1 was identified by its characteristic shape of the 196 H. AonumMA, T. NAGAYAMA AND M. HISADA A6 200um B A6CL OP CUR 12nA A2 EX promt 0.5sec 12nA 0.5sec Fic. 4. Output effect of NE-1. A. Morphology of NE-1 in the Sth and terminal (A6) abdominal ganglion. The projection of the axonal branches (A5) was in the ipsilateral neuropil only. B. Output effect of NE-1 upon both the closer (1st trace) and opener (2nd trace) motor neurons. C. Output effect of NE-1 upon both the extensor and flexor motor neurons in the abdominal (A1-A5) ganglia. most posterior secondary neurite crossing the midline (A6 in Fig. 6). Neurobiotin staining revealed that the ascending axon of VE-1 projected into at least the 2nd to 3rd abdominal connective. In each anterior abdominal ganglion (AS, 4 or 3), several small branches extended from the axon within the unilateral half of the neuropil (Fig.6A). The extent and number of axonal branches in each ganglion were similar. The anterior, medial and posterior branches projected both medially and laterally. These axonal branches were usually extended within the ventral half of the neuropil (Fig. 6B, C). One of the physiological characteristics of VE-1 was variable output effect upon the uropod motor neurons [21]. In this study, VE-1 was impaled 19 times but their effects upon the uropod motor neurons were inconsistent: 3 interneurons activated both the closer and opener motor neurons while 13 interneurons activated the closer motor neurons and inhibited the antagonistic opener motor neurons. The remaining 3 interneurons had no effect upon the uropod motoneurons. The output effect upon the abdominal postural motor neurons was, by contrast, consistent in almost all prepara- tions. On 17 occasions, this interneuron decreased spike frequency of the extensor inhibitor with the increase in the activity of the extensor excitors (No. 3/4) in all the anterior ganglia (Fig. 7A). At the same time, tonic spikes of the flexor excitors (No. 3/4) in all the anterior ganglia were inhibited with the increase in the spikes of the flexor inhibi- tor. Another 2 VE-1 had no effect upon the postural motor neurons in any anterior ganglia. In many cases, VE-1 also affected the postural motor neurons on the contralateral side. The output effect was the same as that upon the ipsilateral postural motor neurons and elicited reciprocal extension-like pattern from the Ist to Sth abdominal ganglion (Fig. 7B). Interneurons with flexion effect Four identified ascending interneurons, RC-4, RO-2, RO-3 and RO-4 [21], elicited abdominal flexion-like motor pattern. For example, figure 8A showed the morphology of RO-2 in the terminal abdominal ganglion. Physiologically, RO-2 showed antifacilitation in responses to repetitive elec- trical stimulation of the afferents. Spikes followed only the Crayfish Ascending Interneurons 197 B n6 isn AY R i: ey . 200um “4 7 SS Fic. 5. Output effect of RC-2. CL 7nA EX ae EX al EX EE A. Morphology of RC-2 from the 3th to terminal (A3-A6) abdominal ganglion. kk CUR 7nA — f 0.5sec The projection of the axonal branches in each abdominal ganglion (A1-A5) was in the ipsilateral neuropil only. B. Output effect of RC-2 upon both the closer and opener motor neurons. C. Output effect of RC-2 upon both the extensor and flexor motor neurons in the anterior abdominal (A1-A5) ganglia. 1st stimulus and then continuously depressed after the 2nd stimulus (Fig. 8B). When the depolarizing current was in- jected into RO-2, tonic spikes of the opener motor neurons increased while those of the closer motor neurons decreased (not shown). At the same time, this interneuron increased the spike frequency of extensor inhibitor (No. 5) and also activated that of the flexor excitors (No. 3/4 and 6) in all the anterior abdominal ganglia (Fig. 8C). Another 3 inter- neurons, i.e., RC-4, RO-3, RO-4, had similar output effects upon the abdominal postural motor neurons, though the effect of RC-4 upon the uropod motor neurons was opposite from other interneurons (Table 1). DISCUSSION Output effect of ascending interneurons upon abdominal post- ural system This study has demonstrated that many identified ascending interneurons originating in the terminal abdominal ganglion of the crayfish [21] have premotor effects upon the abdominal postural motor neurons in the anterior abdominal ganglia. Each interneuron controlled the homologous post- ural motor neurons from the Ist to 5th abdominal ganglia simultaneously in the same way (Table 1). For example, if the extensor excitors in the Sth abdominal ganglion were excited by a particular interneuron, the activity of extensor excitors in the remaining abdominal ganglia (A1-A4) were also increased. These physiological results suggested that the axons of ascending interneurons ran through the abdo- minal connective and projected into thoracic or more anterior ganglia. Neurobiotin staining could reveal the intersegmen- tal structure of ascending interneurons, since this dye spreaded rapidly for a long distance (about 3cm). The ascending interneurons extended several axonal branches with similar projection in each anterior ganglion (e.g. Fig. 6), though we could not trace the whole structure of inter- neurons. The patterns of output effects of ascending interneurons upon the abdominal postural motor neurons were divided into three types. About half of interneurons encountered in this study (7 out of 16 interneurons) co-actively excited both 198 H. Aonuma, T. NAGAYAMA AND M. HISADA } : pa | 200um 200um 200um Fic. 6. Morphology of VE-1. A. Morphology of ascending interneuron VE-1 in the 3rd (A3), 4th (A4), Sth (AS), and terminal (A6) ganglion. The projections of the axonal branches in the anterior abdominal ganglia (A3-A5) were in the ipsilateral neuropil only. B,C. Drawing of sections (20 «m) of the 4th and Sth abdominal ganglion. Transverse sections arranged from the homologous part of each ganglion. The axon of VE-1 in each ganglion ran through ventral intermediate tract and project axonal branches both medially and laterally. DC II,dorsal commissure II; DIT, dorsal intermediate tract; DMT, dorsal medial tract; LG, lateral giant; MDT, medial dorsal tract; MG, medial giant; VIT, ventral intermediate tract; VMT, ventral medial tract. Crayfish Ascending Interneurons 199 A. |psilateral to axon * * Al EX A2 EX FL A3 EX ae A4 EX FL AS EX FL CUR 12nA 0.5sec B. Contralateral to axon 9nA — 0.5sec Fic. 7. Output effect of VE-1 upon abdominal postural motor neurons. A. Activity change of extensor (1st trace in each ganglion) and flexor (2nd trace in each ganglion) motor neurons on the side ipsilateral to the axonal branches of VE-1. The response of motor neurons from the Ist to 5th abdominal ganglion was drawn succeedingly. B. The response of postural motor neurons on the opposite side. On the flexor motor neurons in the 3rd abdominal ganglion (2nd trace in A3), the largest tonic flexor excitor (No. 6) spiked spontaneously and the current injection decreased the spike (No. 6) and increased the flexor inhibitor (No. 5). the extensor and flexor excitors. Four interneurons elicited reciprocally extension-like motor pattern while another 4 interneurons elicited reversed flexion-like motor pattern. In both the crayfish and lobster, abdominal positioning inter- neurons have been known to produce abdominal movement [5, 6, 13, 14]. Many of them had somata in the anterior abdominal ganglia (A2-A5) and had ascending and/or de- scending axons through the abdominal connective. Parts of them were originated from the terminal abdominal ganglion and some of them might be similar to the interneurons described in this study [5]. In the previous works, abdomin- al positioning interneurons were categorized as one of either FPIs (flexion producing interneurons), EPIs (extension pro- ducing interneurons) or inhibitory interneurons [e.g. 6]. By contrast, no interneurons in this study inhibited both the extensor and flexor excitors, but many interneurons co- actively excited both the antagonistic motor neurons. Multiple function of ascending interneurons Ascending interneurons in the terminal ganglion re- ceived excitatory input directly from the mechanosensory afferents innervating hairs on the surface of the tailfan [23] and/or from the proprioceptive afferents innervating the chordotonal organ of the uropod [28]. They, in turn, prop- agated encoded sensory signals into anterior segments and affected abdominal postural motor neurons in all the anterior (A1-A5) abdominal ganglia [this study]. At the same time, many interneurons also affected the uropod motor neurons in a various fashion [21]. Some interneurons further recruited the unidentified motor neurons of swimmerets (1st root) that were homologous appendages with uropods, though rhythmic period of power- and return-stroke was not affected [Aonu- ma, unpublished data]. The ascending interneurons, there- fore, acted as multifunctional units that controlled the diffe- rent motor systems simultaneously. 200 H. AoNuMA, T. NAGAYAMA AND M. HIsADA C at EX Fk EX FL EX le EX Re EX cepmmnenel ices. caatanmnannaae FL A y __|10mv CUR 0.2sec 6nA O5eec Fic. 8. Output effect of RO-2. A. Morphology of RO-2. B. Repetitive electrical stimulation (indicated by arrowheads) of the 2nd root afferent (20 Hz) elicited antifacilitation in RO-2. C. Output effect of RO-2 upon both the extensor and flexor motor neurons in the abdominal (A1-A5) ganglia. There was, however, no close correlation between the output effect upon the uropod motor neurons and that upon the abdominal postural motor neurons (Table 1). Inter- neurons that elicited reciprocally closing pattern of the uro- pod produced either extension-like or flexion-like motor pattern of the abdomen. Furthermore, interneurons that activated both the extensor and flexor excitors had different effect upon the uropod motor neurons. CA-1 excited both the closer and opener motor neurons while CI-1 inhibited both motor neurons. Interneurons of RC group (RC-3, -6, -8) produced closing pattern of the uropod, while RO-5 produced reversed opening pattern. Only interneurons with flexion effect tended to elicit the reciprocally opening pattern of the uropod (3 out of 4 interneurons). This variety of combination of output to uropod and abdominal postural system in ascending interneurons would partly be derived from the complexity of movement of the abdominal seg- ments. During equilibrium reactions [41], leg reflexes [27], avoidance reaction [26], escape swimming [39], backward walking [12] and defensive reaction [29], coordinated move- ment with a various pattern between abdomen and uropod was performed. The intersegmental coordination between uropod and abdomen was also essential for posture and locomotion of animals. For example, the flexion producing interneurons (e.g. RO-2, -3 and -4) would be related to the LG mediated tail-flip that was initiated by the rapid flexion of the abdomen [39]. On the other hand, RC-2 and RC-7 had potential to mediate avoidance “dart” response that consisted of immediate closing of uropods followed by the forward walking with abdominal extension [26]. These interneurons produced a train of spikes in response to the repetitive sensory stimulation, elicited the closing pattern of the uropod and produced the abdominal extension (Fig. 5). The phy- siological result that interneurons with flexion effect failed to respond spikes continuously by the repetitive sensory stimula- tion of the tailfan as a result of antifacilitation (Table 1) might be consistent with the behavioural observation that the mechanical stimulation of the tailfan preferably elicited the “dart” response (Fig.2A). Furthermore, slow postural movement of the abdomen was usually accompanied with the activation of both the extensor and flexor muscles [30]. The interneurons such as CA-1, RC-3 and RC-6 would contribute Crayfish Ascending Interneurons 201 to increase the tonus of the postural muscles, since they co-actively excited both the extensor and flexor excitors and/ or inhibited the inhibitors. Thus, each ascending inter- neuron will contribute to one or more specific behavioural act(s) and the coordinated sequence of movement between uropod and abdominal posture will be completed by the activation of these ascending interneurons as multisegmental integrators. A particular behavioural act might be triggered through selective activation and certain central interaction between these ascending interneurons. ACKNOWLEDGMENTS T. N. was supported by a Grants (04740394) from the Ministry of Education, Science and Culture. M. H. was supported by a Grant of Human Frontiers Science Program. 10 11 12 13 14 15 REFERENCES Bullock TH, Horridge GA (1965) Structure and function in the nervous systems of invertebrate. WH Freeman San Francisco Burrows M (1992) Local circuits for the control of leg move- ments in an insect. TINS 15. 6: 226-227 Calabres R (1976) Crayfish mechanoreceptive interneurons; I. The nature of ipsilateral excitatory inputs. J Comp Physiol 105: 83-102 Fields HL, Evoy WH, Kennedy D (1967) Reflex role played by efferent control of an invertebrate and vertebrate stretch recep- tor. J Neurophysiol 30: 859-874 Jellies J, Larimer JL (1986) Activity of crayfish abdominal- positioning interneurons during spontaneous and sensory- evoked movements. J Exp Biol 120: 173-188 Jones KA, Page CH (1986) Postural interneurons in the abdo- minal nervous system of lobster. III. Pathways mediating in- tersegmental spread of excitation. J Comp Physiol A 158: 281- 290 Kennedy D, Evoy WH, Fields HL (1966) The unit basis of some crustacean reflexes. Symp Soc Exp Biol 20: 75-109 Kennedy D, Takeda K (1965) Reflex control of abdominal flexor muscles in the crayfish. II. The tonic system. J Exp Biol 43: 229-246 Kita H, Armstrong W (1991) A biotin-containing compound N-(2-aminoethyl) biotinamide for intercellular labeling and neuronal tracing studies: comparison with biocytin. J Neuro Meth 37: 141-150 Kondoh Y, Hisada M (1986) Neuroanatomy of the terminal (sixth abdominal) ganglion of the crayfish, Procambarus clarkii (Girard). Cell Tissue Res 243: 273-288 Kondoh Y, Hisada M (1987) The topological organization of primary afferents in the terminal ganglion of crayfish, Procam- barus clarkii. Cell Tissue Res 247: 17-24 Kovac M (1974) Abdominal movements during backward walking in crayfish. I. Properties of the motor program. J Comp Physiol 95: 61-78 Larimer JL, Jellie J (1983) The organization of flexion-evoking interneurons in the abdominal nerve cord of the crayfish, Procambarus clarkii. J Exp Zool 226: 341-351 Larimer JL, Moore D (1984) Abdominal positioning inter- neurons in crayfish: projections to and synaptic activation by higher CNS centers. J Exp Zool 230: 1-10 Ma PM, Beltz BS, Kravitz EA (1992) Serotonin-containing neurons in lobsters: their role as gain-setters in postural control 16 18 19 20 21 22 23 24 25 26 27 28 30 31 32 33 34 35 mechanisms. J Neurophysol 68: 36-54 Macmillan DL, Altman JS, Kein J (1983) dination in the crayfish swimmeret system reconsidered. Zool 228: 157-162 Miall RC, Larimer JL (1982a) Central organization of crusta- cean abdominal posture motoneurons: connectivity and com- mand fiber inputs. J Exp Zool 224: 45-56 Miall RC, Larimer JL (1982b) Interneurons involved in abdo- minal posture in crayfish: structure, function and command fiber responses. J Comp Physiol 148: 159-173 Murchison D, Larimer JL (1990) Dual motor output inter- neurons in the abdominal ganglia of the crayfish Procambarus Clarkii: synaptic activation of motor outputs in both the swim- meret and abdominal positioning systems by single interneurons. J Exp Biol 150: 269-293 Nagayama T, Hisada M (1987) Opposing parallel connections through crayfish local nonspiking interneurons. J Comp Neurol 257: 347-358 Nagayama T, Iosogai Y, Sato M, Hisada M (1993) Interseg- mental ascending interneurones controlling uropod movements of the crayfish Procambarus clarkii. J Comp Neurol 332: 155- 174 Nagayama T, Isogai Y, Namba H (1993) Physiology and morphology of spiking local interneurons in the terminal abdo- minal ganglion of the crayfish. J Comp Neurol 337: 584-599 Nagayama T, Sato M (1993) The organization of exteroceptive information from the uropod to ascending interneurons of the crayfish. J Comp Physiol 172: 281-294 Nagayama T, Takahata M, Hisada M (1983) Local spikless interaction of motoneurons dendrites in the crayfish Procambar- us clarkii. J Comp Physiol 152: 335-345 Nagayama T, Takahata M, Hisada M (1984) Functional char- acteristics of local non-spiking interneurons as the premotor elements in crayfish. J Comp Physiol 154: 499-510 Nagayama T, Takahata M, Hisada M (1986) Behavioral transi- tion of crayfish avoidance reaction in response to uropod sti- mulation. J Exp Biol 46: 75-82 Newland PL (1989) The uropod righting reaction of the crayfish Procambarus clarkii (Girard): an equilibrium response driven by two largely independent reflex pathways. J Comp Physiol A 164: 685-696 Newland PL, Nagayama T (1993) Parallel processing of prop- rioceptive information in the terminal abdominal ganglion of the crayfish. J Comp Physiol A 172: 389-400 Notvest RR, Page CH (1982) Role of the hemigiant neurons in the crayfish defense response. Brain Research 292: 57-62 Page CH (1975) Command fiber control of crayfish abdominal movement. II. Generic differences in the extension reflexes of Orconectes and Procambarus. J Comp Physiol 102: 77-84 Intersegmental coor- J Exp Pearson KG, Iles JF (1973) Nervous mechanisms underlying intersegmental co-ordination of leg movements during walking in the cockroach. J Exp Biol 58: 725-744 Reichert H, Wine JJ (1982) Neural mechanisms for serial order in a stereotyped behaviour sequence. Nature 296: 86-87 Reichert H, Plummer MR, Hagiwara G, Roth RL, Wine JJ (1982) Local interneurons in the terminal abdominal ganglion of the crayfish. J Comp Physiol 149: 145-162 Sigvardt KA, Hagiwara G, Wine JJ (1982) Mechanosensory integration in the crayfish abdominal nervous system: structural and physiological differences between interneurons with single and multiple spike initiating sites. J Comp Physiol 148: 143- 157 Stewart W (1978) Functional connections between cells as revealed by dye-coupling with the highly fluorescent naphthali- mide tracer. Cell 14: 741-759 202 36 3H 38 H. AonuMA, T. NAGAYAMA AND M. Hisapa Takahata M, Yoshino M, Hisada M (1985) Neuronal mechan- isms underlying crayfish steering behaviour as an equilibrium response. J Exp Biol 114: 599-617 Toga T, Takahata M, Hisada M (1990) An identified set of local nonspiking interneurons which control the activity of abdominal postural motoneurons in crayfish. J Exp Biol 148: 477-482 van Harreveld A (1936) A physiological solution for freshwa- ter crustaceans. Proc Soc Exp Biol 34: 428-432 39 40 41 Wine JJ, Krasne FB (1972) The organization of escape be- haviour in the crayfish. J Exp Biol 56: 1-18 Wine JJ, Mittenthal JE, Kennedy D (1974) The structure of tonic flexor motoneurons in crayfish abdominal ganglia. J Comp Physiol 93: 315-335 Yoshino M, Takahata M, Hisada M (1980) Statocyst control of the uropod movement in response to body rolling in crayfish. J Comp Physiol 139: 243-250 ZOOLOGICAL SCIENCE 11: 203-208 (1994) Phagocytic Activity of Tunic Cells in the Colonial Ascidian Aplidium yamazii (Polyclinidae, Aplousobranchia) Euicut Hirose!, TERUHISA IsHm?, YASUNORI SAITo* and YASUHO TANEDA? 1Biological Laboratory, College of Agriculture and Veterinary Medicine, Nihon University, Fujisawa, Kanagawa 252, *Shimoda Marine Research Center, University of Tsukuba, Shimoda, Shizuoka 415, and *Department of Biology, Faculty of Education, Yokohama National University, Yokohama, Kanagawa 240, Japan ABSTRACT—The phagocytic activity of tunic cells of the colonial ascidian Aplidium yamazii was assessed by incubation of thin tunic slices including these cells with fluorescent microparticles. Only one type of tunic cell engulfed the microparticles. These phagocytic tunic cells are irregularly shaped, motile, and often contain phagosomes. Many of them also contain vesicles laden with round granules. Occasionally, they engulf another tunic cell. Because this type of tunic cell is always found in the tunic of histological sections prepared from whole (unsliced) colony specimens, these cells are probably distributed in the tunic under normal conditions. Peroxidase activity was demonstrated exclusively within granule-containing vesicles of some phagocytic tunic cells. This finding indicated that the phagocytic tunic cells might possess an oxygen-dependent microbicidal system. It is presumed that the phagocytic tunic cells migrate throughout the tunic matrix, engulf extraneous substances (including bacteria) and also function as scavengers to keep the tunic free of discarded tunic cells and other debris, such as from wounds. © 1994 Zoological Society of Japan INTRODUCTION The tunic is an integumentary tissue in urochordates, such as ascidians; it is a gelatinous or leathery matrix covering the outer surface of epidermis. The tunic is a unique tissue in animals because of its cellulosic component [1,12]. The main function of the tunic is protection of the body, although little is known about this defense system. ‘The tunic is an attractive material for studying defense systems from the viewpoint of comparative immunology because of its pecu- liarity and the phylogenetic position of ascidians, whose ancestors may be the same as those of the vertebrates. Ascidians have two types of free cells: hemocytes and tunic cells. Hemocytes circulate in the blood vessels and in the mesenchymal space, and tunic cells are distributed in the tunic; both cell types are presumed to have immunological activity. The phagocytic activity of these free cells was mainly studied by injection or insertion of foreign substances (reviewed in [13]). These studies could not determine whether one or both of these cell types was involved in phagocytosis, because hemocytes seem to infiltrate the tunic that is responding to the experimental operations. Recent studies have demonstrated that particular types of hemocytes have phagocytic activity in vitro [9, 10, 14]. As for tunic cells, De Leo et al. [2] described the “phagocyte” that is characterized by one or two “heterolysosomic vacuoles” as a type of tunic cell in the solitary ascidian Ciona intestinalis. Parrinello et al. [5, 6,7] reported on the phagocytic activity and inflammatory reaction in the tunic by introducing foreign Accepted January 7, 1994 Received November 1, 1993 " To whom all correspondences should be addressed. substances into tunic. However, dealing with tunic cells has following difficulties: It is almost impossible to isolate the tunic cells from tunic matrix, and the specimens are usually contaminated by leakage or infiltration of hemocytes. Aplidium yamazii, a colonial ascidian belonging to the family Polyclinidae and the suborder Aplousobranchia, forms a white, sheet-like colony, and its elongated zooids are separately embedded in a transparent, gelatinous tunic. When a live colony is cut into slices of about 0.5 mm thick, the tunic cells in the slices are observable under a light microscope and remain alive for several hours to a few days. Using these live colony slices, the present study experimental- ly demonstrated phagocytosis of fluorescent microparticles in the tunic. In this study, the possibility of hemocyte con- tamination was almost eliminated, because this polyclinid species has no blood vessels in the tunic. We also performed cytochemical determinations for peroxidase, which may be involved in microbicidal activity. MATERIALS AND METHODS Animals Colonies of Aplidium yamazii were collected in Nabeta Bay, Shimoda (Shizuoka Pref., Japan). They were attached to glass slides with cotton thread and were reared in culture boxes immersed in Nabeta Bay. The sheet-like colonies grew and spread on the glass slides (Fig. 1A). Assay for phagocytic activity A growing part of a live colony was cross-sectioned into slices about 0.5 mm thick using a razor blade (Fig. 1B). Each colony slice consisted of the gelatinous, transparent tunic, the tunic cells, and some fragments of zooids. A solution of fluorescent microparticles (fluoresbrite carboxylate microspheres, 0.5 ~m in diameter, 2.5% 204 E. Hirose, T. Isuu et al. D Fic. 1. The procedures of the assay for phagocytic activity. t, tunic; z, zooids. A: Colonies of A. yamazii growing on a glass slide. B: Sectioning of a live colony. C: Incubation of colony slices in filtered seawater containing fluorescent microparticles. D: Mounting a colony slice on a glass slide. solid latex; Polyscience) was diluted in 1:5 with filtered seawater (FSW). The colony slices were incubated in this solution for 30 min at room temperature, allowing the microparticles to permeate the tunic (Fig. 1C). After extensive washing with FSW, the slices were incubated in FSW or FSW containing penicillin (1001U/ml) and streptomycin (1 mg/ml) for 1-2 hr, 24 hr, or 48 hr at room tempera- ture. Because the specimens that were incubated for 24 hr or 48 hr markedly shrank, they were sliced again for microscopic observation. Each slice was observed under a light microscope equipped with epifluorescence and Nomarski differential interference contrast (DIC) optics (Fig. 1D). For electron microscopy, some colony slices incubated with microparticles were fixed in 2.5% glutaraldehyde-0.1M sodium cacodylate-0.45 M sucrose (pH 7.4) for 2hr on ice. They were rinsed in 0.1M sodium cacodylate-0.45 M sucrose (pH7.4), and postfixed in 1% osmium tetroxide-0.1 M sodium cacodylate (pH 7.4) for 1-1.5hr on ice. They were dehydrated through a graded ethanol series, cleared with n-butyl glycidyl ether, and embedded in low viscosity epoxy resin. Thin sections were stained with uranyl acetate and lead citrate and examined using a Hitachi HS-9 transmis- sion electron microscope at 75 kV. We also fixed colony pieces of about 5 mm x 10 mm (containing more than 30 zooids) and processed them for electron microscopy as described above. We wanted to determine whether the phagocytic cells observed in the above experiment were always distributed in the tunic or were contaminating hemocytes from the slicing process. Cytochemistry for peroxidase activity The colony pieces were fixed in 2.5% glutaraldehyde-2% NaCl- 0.1 M Millonig’s phosphate buffer (pH 7.4) on ice for 15 min, and the fixed pieces were cut into slices with a razor blade. After washing in the same buffer, they were pre-incubated in 0.1% diaminobenzidine (DAB; Sigma)-0.1 M phosphate buffer (pH 6.8) for 30 min at room temperature and then incubated in 0.1% DAB-0.3% H>;O,-0.1M phosphate buffer for 15 min. The specimens were washed with 0.1 M phosphate buffer and postfixed with 1% osmium tetroxide. They were dehydrated and embedded in epoxy resin as described above for electron microscopy. Thick sections were stained with toluidine blue. In negative controls, we omitted H,O, from the incubation medium or added 50mM 3-amino-1,2,4-triazole (Sigma) in the pre-incubation and incubation media as an inhibitor. RESULTS In the tunic of colony slices of Aplidium yamazii, there were tunic cells of various types; some types have protruding filopodia, some types have an elongated cell shape, some contain many granules, and some form multicellular vesicles. Particular types of tunic cells showed phagocytic activity (Fig. 2). Although these phagocytic cells were found throughout the tunic matrix, they did not appear to be evenly distributed. This was caused by the uneven distribution of microparticles within the tunic slices; that is, there were more microparticles per unit area at the surface of the slices than at the core region. For this reason, we could not make a quantitative description of the distribution of phagocytic tunic cells in this study. The tunic cells phagocytizing microparticles were essentially irregularly shaped cells with extending filopodia. There was, however, some variation in their appearance. For instance, Figure 3 shows three tunic cells, all of which engulfed the microparticles. Cell “a” was a thin, flattened cell with numerous filopodia; cell “b” was almost round; and cell “c” was elliptical. Cells “b” and “c” had thicker cell bodies and fewer filopodia than cell “a”. Because many tunic cells that exhibit phagocytosis showed intermediate appearances among these three, we classified them as a single cell type, namely, phagocytic tunic cells. These cells occa- sionally engulfed another tunic cell, and thus might be considered scavengers. A tunic cell engulfing a relatively large cell did not have prominent filopodia, and its cell body became a thin sheet that wrapped around the engulfed cell. With respect to the characteristics of phagocytic cells (i.e. cell Phagocytic tunic cells in ascidians 205 4B Fic. 2. Paired images of a live colony slice incubated for 2hr in FSW. A: Nomarski DIC. B: Epifluorescence. Arrowheads indicate tunic cells phagocytizing fluorescent microparticles. Note round granular cells and some thin filopodial cells do not contain the microparticles. (Scale bar=25 um) Fic. Fic. 3. Light micrograph (Nomarski DIC) of three phagocytic tunic cells (a, b, c) in live colony slice incubated for 2 hr in FSW. Each cell has a different appearance from the others with respect to cell shape, thickness of cell body, and number of filopodia. (Scale bar=10 4m) 4. Cytochemistry for peroxidase in the tunic. A) Dark reaction product indicates peroxidase activity in phagocytic tunic cells (arrowheads). B) A negative control in which peroxidase inhibitor (3-amino-1,2,4-triazole) was added. The reaction product is not found in any tunic cells. The granular inclusions are only stained with toluidine blue (arrowheads). (Scale bar=10 ~m) shape, granular inclusions, and distribution of engulfed microparticles within the cells), prominent differences were not found among the specimens incubated for 2 hr, 24 hr, and 48 hr. In preliminary observations using time-lapse video recording, these phagocytic tunic cells actively migrated within the tunic matrix and some non-phagocytic ones did not. The ability to migrate may be indispensable for pha- gocytic activities. 206 E. Hirose, T. Isum et al. In the cytochemical study, only some of the phagocytic tunic cells showed peroxidase activity. In thick sections for light microscopy, the dark product indicating peroxidase activity is uniquely localized within the phagolysosome-like vesicles and/or vesicles carrying round granules (Fig. 4A). No specific activity is demonstrated in the other cell compart- ments or in the other types of tunic cells. or example, a phagocytic cell engulfing another cell has peroxidase activity in its vesicles that contain round granules, but not in the large phagosome that engulfs the cell. No peroxidase activity was demonstrated in the negative controls (Fig. 4B). In electron microscopic observations, the microparticles were recognized as electron-lucent rounds or ellipsoids whose margins were moderately electron dense. Some microparti- cles adhered to exposed surface of the tunic matrix, and some permeated the tunic slices. In phagocytic tunic cells en- gulfing the microparticles, the latter were found within the vesicles, and these cells also had some vesicles laden with round, electron-dense granules (Figs. 5 and 6). The largest of these granules were about 1.5 ~m in diameter, with some Fic. 5. A phagocytic tunic cell in the colony slice incubated for 2hr in FSW. The cell contains granules (g), the engulfed microparticles (arrows) and a phagolysosome-like vesicle (arrowhead). This cell also carries a vesicle containing a bacteria-like structure (double arrowheads). (Scale bar=2 um) Fic. 6. A phagocytic tunic cell in the colony slice incubated for 24 hr in FSW. The cell engulfs microparticles (arrow) and another tunic cell (c). (Scale bar=2 «m) Fic. 7. A phagocytic tunic cell in a specimen fixed as a whole colony piece. The cell contains electron-dense granules (arrowheads) and phagolysosome-like vesicles (p). (Scale bar=2 ~m) Phagocytic tunic cells in ascidians 207 being discernible under the light microscope. Some of the phagocytic tunic cells also had phagolysosome-like vacuoles containing disorganized structures that appeared to be dis- integrating cellular components. In addition, a bacteria-like structure was occasionally found in a vesicle (double arrow- heads in Fig.5). In Figure 6, the phagocytic tunic cell carrying microparticles has engulfed another tunic cell. Tunic cells with the same characteristics described above were also found in specimens fixed as whole colony pieces (Fig. 7). DISCUSSION When colony slices of Aplidium yamazii are incubated with microparticles, particular tunic cells show phagocytic activity. These phagocytic cells usually have protruding filopodia, often contain some round granules, and occasional- ly engulf another cell. It is presumed that phagocytic tunic cells migrate throughout the tunic matrix, engulf extraneous substances including bacteria, and also function as scav- engers, thereby keeping the tunic free of discarded tunic cells and wound debris. Cellular defense reactions in the tunic involve two groups of free cells: infiltrating hemocytes that respond to infections or trauma, and tunic cells that always “stand by” in the tunic. In the allogeneic rejection reaction of botryllid ascidians (colonial species belonging to the family Botryllidae, the suborder Stolidobranchiata), many hemocytes infiltrate the tunic from the blood vessels and participate in the necrotic reactions [3, 11]. In Ciona intestinalis (a solitary species), when particulate or soluble agents are injected in the tunic, a capsule and/or tissue injury is produced, depending on the dose of the irritant [5, 6, 7]. In this tunic reaction of C. intestinalis, the hemocyte infiltration may also occur, induced by the irritant injection, because some of the cells appear around the wound are different from tunic cells [6]. In Aplidium yamazii, the phagocytic cells observed here are tunic cells that are always present in the tunic, because they are usually observed in the tunic part of the specimens fixed as whole colony pieces that were not sliced before fixation. In C. intestinalis, De Leo et al. [2] studied the fine structure of the tunic and described the “phagocyte” as one type of tunic cell. The characteristics of the “phagocyte” in C. intestinalis are very similar to those of phagocytic tunic cells in A. yamazii: protruding thin cytoplasmic extensions, phagolysosome-like vesicles, and occasional engulfment of other cells. These phagocytic cells of C. intestinalis and A. yamazi probably belong to a homologous cell type. In botryllid ascidians, although we have already found amoeboid tunic cells containing bacteria in their vesicles [4], we have not found cells that engulf other cells. The phagocytic tunic cells in A. yamazii have some morphological variations in their cell shape and in the number of filopodia (see Fig. 3). These variations are probably caused by differ- ent states of phagocytosis, migration, and/or cell differentia- tion in the same cell type. It is noteworthy that Sawada et al. [10] described two types of phagocytes among the hemocytes of the solitary ascidian, Halocynthia roretzi: one type (p1- cell) spreads as thin, flat sheets, and the other type (p2-cell) is thicker than the former. P2-cells in H. roretzi are similar in morphology to some of phagocytic tunic cells, such as cell “c” in Figure 3, in A. yamazii. In the phagocytic tunic cells, peroxidase activity was demonstrated within the phagolysosomes and/or the vesicles carrying the round granules. Peroxidase is known to inacti- vate peroxide ions generated in the oxygen-dependent micro- bicidal pathway. This suggests the following two possibil- ities: The phagocytic tunic cells may have this microbicidal activity, and the granules in the phagocytic tunic cells are possibly the engulfed materials processed in the phagolyso- somes. In contrast to our findings, a cytochemical study using C. intestinalis showed that all of the amoebocytes among the hemocytes are peroxidase-negative [8], though the cytochemical methods were different from those of the pre- sent study. It is possible that oxygen-dependent microbi cidal activities are not ubiquitous in ascidian tissues or ascidian species. Based on our observations, we propose a hypothesis on the differentiation of phagocytic tunic cells: 1) The cells with a thin cell body and numerous protruding filopodia are at a young stage. 2) Phagolysosomes are formed after phagocytosis, and their contents subsequently become round granules. The oxygen-dependent microbicidal activity may be carried out during this process. 3) An increase in the number of granules makes the cell body thicker and roundish. It is thought that almost all tunic cells originate from hemocytes passing through the epidermis. If phagocytic tunic cells originate from hemocytes, the question arises as to whether phagocytic hemocytes migrate into the tunic or hemoblasts differentiate into phagocytic tunic cells within the tunic. For further understanding of the immunological activity and differentiation of tunic cells, a precise description and classification of the tunic cells in A. yamazii is required. ACKNOWLEDGMENTS This study was supported by a Grant-in-Aid for General Scien- tific Research (no. 03455008) to Y. S. from the Ministry of Educa- tion, Science and Culture of Japan. This is contribution no. 564 from the Shimoda Marine Research Center, University of Tsukuba. REFERENCES 1 De Leo G, Patricolo E and D’Ancona Lunetta G (1977) Studies on the fibrous components of the test of Ciona intestinalis Linnaeus. I. Cellulose-like polysaccharide. Acta Zool 58: 135-141 2 De Leo G, Patricolo E , Frittita GG (1981) Fine structure of the tunic of Ciona intestinalis L. II. Tunic morphology, cell dis- tribution and their functional importance. Acta Zool 62: 259- 271 3 Hirose E, Saito Y, Watanabe H (1990) Allogeneic rejection induced by cut surface contact in the compound ascidian, Botrylloides simodensis. Invert. Reprod Dev 17: 159-164 208 E. Hirose, T. Isum et al. Hirose E, Saito Y, Watanabe H (1991) Tunic cell morphology and classification in botryllid ascidians. Zool Sci 8: 951-958 Parrinello N, Patricolo E (1984) Inflammatory-like reaction in the tunic of Ciona intestinalis (Tunicata). II. Capsule compo- nent. Biol Bull 167: 238-250 Parrinello N, Patricolo E, Canicatti C (1977) Tunicate im- munobiology. I. Tunic reaction of Ciona intestinalis L. to erythrocyte injection. Boll Zool 44: 373-381 Parrinello N, Patricolo E, Canicatti C (1984) Inflammatory- like reaction in the tunic of Ciona intestinalis (Tunicata). I. Encapsulation and tissue injury. Biol Bull 167: 229-237 Rowley AF (1982) Ultrastructural and cytochemical studies on the blood cells of the sea squirt, Ciona intestinalis. 1. Stem cells and amoebocytes. Cell Tissue Res 223: 403-414 Rowley AF (1983) Preliminary investigations on the possible antimicrobial properties of tunicate blood cell vanadium. J Exp Zool 227: 319-322 10 11 12 13 14 Sawada T, Fujikura Y, Tomonaga S, Fukumoto T (1991) Classification and characterization of ten hemocytes types in the tunicate Halocynthia roretzi. Zool Sci 8: 939-950 Taneda Y, Watanabe H (1982) Studies on colony specificity in the compound ascidian, Botryllus primigenus Oka. I. Initiation of “nonfusion“ reaction with special reference to blood cell infiltration. Dev Comp Immunol 6: 43-52 Van Daele Y, Revol J-F, Gaill F, Goffinet G (1992) Charac- terization and supramolecular architecture of the cellulose- protein fibrils in the tunic of the sea peach ( Halocynthia papillo- sa, Ascidiacea, Urochordata). Biol Cell 76: 87-96 Wright RK, Cooper EL (1975) Immunological maturation in the tunicate Ciona intestinalis. Amer Zool 15: 21-27 Zhan H, Sawada T, Cooper EL, Tomonaga, S (1992) Electron microscopic analysis of tunicate (Halocynthia roretzi) hemo- cytes. Zool Sci 9: 551-562 ZOOLOGICAL SCIENCE 11: 209-219 (1994) In vitro Autophosphorylation and Cyclic Nucleotide-Dependent Dephosphorylation of Sea Urchin Sperm Histone Kinase Tatsuo Harumi, KATsuAki HosHINo AND Norio Suzukr! Noto Marine Laboratory, Kanazawa University, Ogi, Uchiura, Ishikawa 927-05, Japan ABSTRACT—We identified two phosphorylatable proteins (33 kDa and 48 kDa) in crude extractes of spermatozoa from the sea urchin, Hemicentrotus pulcherrimus. While the 33 kDa protein was phosphorylated in 100 mM Mg**-containing medium with cAMP or cGMP, phosphorylation of the 48 kDa protein occurred irrespective of the Mg *-concentration if it was more than 10mM. Physiological concentration if it was more than 10 mM. Physiological concentrations of cAMP or cGMP triggered dephosphorylation of the [°*P]-phosphorylated 48 kDa protein in a concentration dependent manner. The dephosphorylation of 48 kDa protein was inhibited with calyculin A or okadaic acid. The 48 kDa protein was associated with a 39 kDa protein to form a larger oligomer with about 400 kDa. The purified 400 kDa protein showed cyclic nucleotide-dependent histone kinase activity. The histone kinase activity shifted from 400 kDa to 39 kDa on a Superose 6 HR in the presence of 5x10-°M cAMP. Phosphorylation of the 48 kDa protein in the purified 400 kDa protein required only Mg[**P]ATP. A photoaffinity cAMP analogue, 8-N3-[**P]cAMP, was incorporated into the 48 kDa protein. These results suggest that the 48kDa protein which is a reguratory subunit of the histone kinase is autophosphorylated, and binding of cAMP to the phosphorylated 48 kDa subunit of the kinase results in dissociation of the © 1994 Zoological Society of Japan 48 kDa subunit(s) from the 39 kDa subunit(s) to allow the 48 kDa subunit accessible to a protein phosphatase. INTRODUCTION In sea urchin fertilization, before contacting an egg surface a spermatozoon must pass through the jelly coat which surrounds the egg. The jelly coat contains two high molecular weight glycoconjugates and sperm-activating pep- tides (SAPs). A fucose sulfate glycoconjugate (FSG), one of the glycoconjugates in the jelly coat has been reported to be a major substance responsible for induction of the acro- some reaction [33-34]. FSG has profound effects on sea urchin spermatozoa such as increases in adenylate cyclase activity [45], cAMP concentrations [9], cAMP-dependent protein kinase (A-kinase) acitivty [10], and induction of phosphorylation of sperm histone H1 [27-28]. Sea urchin spermatozoa contain high A-kinase activity [10, 15, 20-21]. SAPs, which were originally identified as factors which acti- vate sperm respiration and motility [11, 37], also increase sperm cAMP concentrations [9]. In addition, SAPs activate membrane-bound guanylate cyclase [2] and increase cGMP concentrations in sea urchin spermatozoa [11]. These observations show that cyclic nucleotides accumulate in re- sponse to FSG and SAPs, which implies an important func- tion in spermatozoa at fertilization. Cyclic nucleotide-dependent protein phosphorylation is one of the major signal transduction pathways in cells. cAMP-dependent protein phosphorylation in sea urchin sper- matozoa was reported by Porter et al. [28]. They showed cAMP-dependent histone H1 phosphorylation in Strongy- locentrotus purpuratus. The mature sea urchin sperm nuc- Accepted March 1, 1994 Received january 14, 1993 * To whom correspondence showed be addressed. leus contains two sperm-specific histone variants, H1 and H2B, which are larger than their somatic counterparts. Within 10 min after the sperm nucleus enters the egg these histones are lost from the sperm chromatin, being replaced by cleavage stage histones contributed by the egg [12]. These histones are characterized by reversibly phosphorylated N- terminal regions consisting largely of multiple clustered “SPKK” tetrapeptides. The SPKK domains bind the ex- traodinarily long linker DNA in sea urchin sperm chromatin to help stabilize the highly condensed DNA and phosphoryla- tion of these domains appears to weaken the DNA binding [13, 26]. In the present study, we show that cyclic nuc- leotides induce not only histone H1 phosphorylation but also dephosphorylation of a 48 kDa protein in the Hemicentrotus pulcherrimus sperm homogenate. We also demonstrate that the purified 48 kDa protein-associated protein phosphory- lates histone, particularily H1 and H2B, in the presence of cyclic nucleotides. MATERIALS AND METHODS Materials Sea urchins (H. pulcherrimus) were collected along the coast of Toyama Bay near Noto Marine Laboraotry. Spermatozoa were obtained by intracoelomic injection of 0.5 M KCl, and collected as “dry sperm” at room temperature and stored on ice until use. Just before use, the dry sperm were washed with Ca*t- and Mg’t-free artificial sea water (CaMgFASW: 454 mM NaCl, 9.7 mM KCl, 34.5 mM (CH;)3;NCICH,CH,OH, 27.1mM Na,SO,, 4.4mM NaHCO; and 10 mM Tris-HCl, pH 7.5). Reagents [y-°P]ATP (111 TBq/mmol) was purchased from New England Nuclear (Boston, MA). 8-N3-[°*?P]cAMP (2.22 TBq/mmol) and 210 T. Harumi, K. HosHIno anp N. Suzuki 8-N3-[>7P]cGMP (370 GBq/mmol) were products of ICN Radioche- micals (Irvine, CA). Phosphoserine (P-Ser), phosphothreonine (P- Thr), phosphotyrosine (P-Tyr), cAMP, cGMP, calf thymus total histone, histones H1, H2A, H2B, H3 and H4, histone-agarose, catalytic subunit of bovine heart A-kinase, a synthetic protein kinase inhibitor (PKI, rabbit sequence) and DEAE-Sephacel were obtained from Sigma (St. Louis, MO). Calyculin A was from LC Services (Woburn, MA). Okadaic acid was a generous gift from Professor T. Yasumoto (Tohoku University). All other reagents and solvent used were of analytical grade. Phosphorylation and dephosphorylation of sperm proteins Dry sperm (0.5 ml) were suspended in 5 ml CaMgFASW and centrifuged at 10,000g for 10min at 4°C. The resulting sperm pellet was suspended in 5 ml of a solution containing 0.5 M NaCl, 10 mM (CH3;COO),Mg, 1 mM 3-isobutyl-methylxanthine (IBMX), 1 mM dithiothreitol (DTT), 1% 3-[(3-cholamidopropyl) dimethylam- monio]-1-propanesulfate (CHAPS) and 10mM Tris-HCl (pH 7.5) and homogenized by ten strokes of a Teflon-glass homogenizer on ice, followed by sonication with a Branson sonifier model 200 for 30 sec onice. The homogenate was centrifuged at 200g for 5 min at O°C. The resulting supernatant fluid (Sup. A) was used for the following experiments. Since Takai et al. [41] reported that 100 mM Mg** concentration was needed for maximal stimulation of cGMP- dependent protein kinase, we also used a solution containing 100 mM (CH3;COO) Mg in some experiments. Sup. A (100s) was incubated at 20°C with 0.83 pmol of [y- ?2P]ATP (92.5 KBq) for 5 min and then 21 of distilled water (DW) or various agents was added to the sample. At the indicated time, the reaction was terminated by addition of 20u1 of 60% (w/v) trichloroacetic acid (TCA). In experiments using protein phospha- tase inhibitors, inhibitors were added to Sup. A containing known concentrations of cAMP or cGMP. After addition of 201 of 60% (w/v) TCA, the resulting precipitate was rinsed with ice-cold acetone and dried under reduced pressure. The residue was dissolved in 1001 of 4% sodium dodecyl sulfate (SDS) and then subjected to SDS-polyacrylamid gel electrophoresis (SDS-PAGE) [19]. The gel was silver-stained according to the method of Morrissey [24]. Radiolabeled protein bands were detected by exposing the gel to a Kodak X-Omat film at —70°C using Kodak intensifying screens. To purify the 48 kDa phosphorylatable protein, fractions (50 or 1001) obtained from each chromatography were incubated with 0.83 pmol of [y-’P]ATP (92.5 KBq) in the presence of 10 mM (CH;COO).Mg for 10 min at 20°C, and the reaction was stopped by addition of final 10% TCA, followed by SDS-PAGE and autoradiography as de- scribed above. Identification of phosphoamino acids in the [**P]-labeled sperm protein A sperm protein phosphorylated with 1 pmol of [y-?7P]ATP (111 KBq) was separated by SDS-PAGE. The gel was stained with Coomassie brilliant blue [46]. The radioactive protein band (moni- tored by autoradiography) was cut out of the gel and homogenized. Then, the protein was retrieved with a Max- Yield Protein Concentra- tor (ATTO, Tokyo), dialysed against DW, and lyophilized. The residue was dissolved in 200u1 of DW (7.1 mg protein/ml). A sample of 420ug protein was mixed with authentic P-Ser (20umol), P-Thr (10uzmol) and P-Tyr (0.1mol) and lyophilized. The residue was hydrolysed in 400 of 5.7 N HCI for 1 or 4 hr at 110°C according to Yang et al. (44) The hydrolysate was lyophilized and the residue was dissolved in 100l of a solution containing 12.5% methanol in 10 mM potassium phosphate (pH 3.0). The sample (101) was analy- zed by a shimadzu LC-6A HPLC system equipped with a Whatman Partisil-10 SAX anion exchange column (26250 mm;particle size 104m). The colum was equilibrated with the sample-dissolving solution and eluted with the same solution at a flow rate of 1 ml/min at 40°C. The absorbance at 210nm of the colum effluent was monitored. Fractions of 1 ml were collected and the radioactivity in the fractions was measured with an Aloka LSC-1000 liquid scintila- tion counter. Gel filtration of sperm proteins Sup. A was centrifuged at 30,000 g for 30min at 4°C. The resulting supernatant solution (Sup. B, 120ml) was incubated with 1 pmol of [y-**P] (111 KBq) for 10 min at 20°C and filtered with a Millipore Ultrafree-C3HV filter. The filtrate was applied to a TSK/GEL G3000SW column equipped to a Hitachi L-6200 HPLC system. The colum was equilibrated with a solution containing 0.1% CHAPS and 0.1 M sodium phosphate (pH 6.8) and eluted with the same solution at a flow rate of 0.5 ml/min at room temperature. The absorbance at 280 nm of the column effluent was monitored, and fractions of 0.5 ml were collected. For determination of the dephos- phorylation activity in these fractions, Sup. B was subjected to chromatography on a TSK-GEL G3000SW without incubation of 1 pmol of [y-*?P] (111 KBaq). To prepare the substrate for determining the cyclic nucleotide- dependent protein dephosphorylating activity, Sup. B incubated with 1 pmol of [y-*’P]ATP (111 KBq) was subjected to chromatography on the TSK column, and the fraction (No. 13) which contained the [°?P]-phosphorylated 48 kDa protein was saved and purified further by chromatography on the same column. Fractions (No. 13 and No. 14) containing radioactivity were pooled and dialysed against 10 mM Tris (pH 7.5) containing 0.1% CHAPS at 4C. The [*P]-labeled protein in the dialysate was used as the substrate for determining the dephosphorylating activity of the fractions obtained from the TSK- GEI G3000SW column chromatography of Sup. B. The substrate (1001) was incubated with 100u1 of each fraction obtained from a TSK-GEl G3000SW column for 5 min at 20°C and then 21 of 10 mM cAMP or cGMP were added to the reaction mixture. At5 min after addition of cAMP or cGMP, the reaction was terminated by addition of 4041 of 60% TCA (w/v). The sample was processed for SDS- PAGE as described above, and analyzed by SDS-PAGE and then autoradiography. As control experiments incubation without cyclic nucleotides was performed. Determination of protein phosphatase activity Protein phosphatase (PPase) activity was determined with [*?P]- phosphohistones as substrate by methods described by Meisler and Langan [23], Swarup and Garbers [39], and Shacter [35] with slight modifications. The reaction mixture in a total volume of 50zl consisted of 50mM Tris (pH7.5), 1mM DTT, 100mM NaCl, [°?P]-phosphohistones containing 60M [*?P]-serine and Syl of sam- ple solution. The reaction was initiated by addition of the sample solution, incubated for 10 min at 30°C, and terminated by addition of 10041 of 10 mM silicotungstic acid in 0.01 N H,SO,. The mixture was centrifuged at 10,000Xg for 5 min. To the resulting deprotei- nized supernatant fluid (100u1), a 251 of 5% (w/v) ammonium molybdate in 4 N H,SO, was added. The resulting phosphomoly- bdate complex was extracted with 2001 of isobutyl alcohol/toluene (1/1, v/v). After centrifugation of the mixture at 1,000 X g for 5 min to separate the organic and aqueous phases, a 160 aliquot of the organic phase was measured for radioactivity with the liquid scintilla- Sea Urchin Sperm Histone Kinase 211 tion counter. Preparation of [°?P|-phosphorylated histones Phosphorylation of histones was carried out using the catalytc subunit of bovine haert A-kinase as described by Meisler and Langan [23] and Swarup et al. [40] with slight modifications. The reaction mixture in a total volume of 4.0 ml contained 40 mg of calf thymus total histones, 400uM [y-’P]ATP (37 MBq), 1mM DTT, 25 mM MgCl, 50mM Tris-HCl (pH7.5) and 250 units of the catalytic subunit. After incubation at 30°C for 12 hrs, phosphorylated his- tones were precipitated with 25% (w/v) TCA, collected by centri- fugation, dissolved in DW, and re-precipitated with 25% (w/v) TCA. The resulting precipitate was washed twice with ethanol/ether (1/4, v/v) and twice with 0.1 N HCl in ethanol/ether (1/4, v/v), and lyophilized. The residue was dissolved in DW to give a final concentration of 100 nmol of incorporated Pi/ml and stored at —20°C until use. Purification of the [?*P|-phosphorylatable 48 kDa protein All experiments were carried out at 4°C unless otherwise men- tioned. CaMgFASW-washed dry sperm (50 ml) were suspended in 500 ml of ice-cold Buffer A [10 mA (CH;COO),.Mg, 1 mM DTT, 10 mM benzamidine-HCl, 5 mM 6-amino-n-caproic acid, 1 mM pheny- Imethylsulfonyl fluoride and 20 mM Tris-HCl, pH 7.5] containing 1% CHAPS, homogenized with a Teflon-glass homogenizer, and soni- cated with a Branson sonifier model 200. The homogenate was centrifuged at 30,000 g for 30min. The resulting supernatant fluid was applied to a DEAE-Sephacel column (2.518 cm) equilibrated with Bffer A containing 0.1% CHAPS, and eluted with a linear gradient of NaCl from 0M to 0.5M in Buffer A containing 0.1% CHAPS at a flow rate of 50 ml/hr. Fractions containing the [°’P]- phosphorylatable 48kDa protein were pooled and applied to a Toyopearl HW-55 column (1.658 cm) equilibrated with Buffer B [10mM (CH;COO),Mg, 1mM DTT, 0.1% CHAPS and 10mM Tris-HCl, pH 7.5] containing 0.1 M NaCl and 10 mM benzamindine- HCl, and eluted with Buffer B containing 0.1 M NaCl at a flow rate of 17.5 ml/hr. Fractions containing the [**P]-phosphorylatable 48 kDa protein were pooled and concentrated with an Amicon Diaflo Cell. The resulting sample was mixed with an equal volume of Buffer B, applied to a Mono Q HRS/5 column (Pharmacia) equilibrated with Buffer B, and eluted with a linear gradient of NaCl from 0 M to 0.5 M in Buffer B at a flow rate of 1 ml/min. Fractions containing the [°°P]-phosphorylatable 48 kDa protein were pooled and mixed with an equal volume of Buffer B. The sample was applied to a histone-agarose column (0.92.4 cm) equilibrated with Buffer B, and eluted with a linear gradient of NaCl from 0 M to 1 M in Buffer B at a flow rate of 10ml/hr. Fractions containing the [°°P]- phosphorylatable 48 kDa protein were pooled and concentrated with an Amicon Diaflo Cell. Then, the sample was applied to a Superose 6 HR10/30 column (Pharmacia) equilibrated with Buffer B contain- ing 0.1 M NaCl, and eluted with the solution at a flow rate of 0.4 ml/ min. Fractions containing the [**P]-phosphorylatable 48 kDa pro- tein were pooled and used for the following experiments. Photoaffinity incorporation of 8-N3-[*°P|cAMP and 8-N;-[°*P|cGMP into the 48 kDa protein A fifty microliter of Sup. B or the fractions obtained from a TSK-GE]l G3000SW column was incubated with 3.3 pmol of 8-Ns- [°°P]cAMP (7.4 KBq) or 20 pmol of 8-N3-[*?P]cGMP (7.4 KBq) in the presence of 4 mM IBMX for 10 min at 20°C, and then irradiated at 0.3 J/cm? at 254 nm with a Funa-UV-linker (Funakoshi Chemical Co., Tokyo). Thereafter, the sample was mixed with 12.51 of a SDS-PAGE sample buffer (50% glycerol, 25% 2-mercaptoethanol, 5 mM EDTA, 0.5% bromophenol blue and 0.31 M Tris-HCl, pH 6.8), heated at 90°C for 30 sec and subjected to SDS-PAGE, followed by autoradiography. Assay of protein kinase Protein kinase was assayed as described by Roskoski [30] with slight modifications. The reaction mixture (1001) contained 50 mM MES (pH 6.5), 10 mM magnesium acetate, 1 mg/ml calf-thymus or sea urchin sperm histone, 200M [y-**P]ATP, 2“M cAMP, and varying amounts of kinase. After incubation at 30°C for 10 min, a 25pl-aliquot was withdrawn and spotted onto a 12cm piece of Whatman P81 cellulose, and dropped into a beaker containing 75 mM phosphoric acid. The paper was subjected to four washes of 75 mM phosphoric acid for 2 min each and then air-dried, followed by measuring the radioactivity in the liquid scintillation counter. Isolation of sea urchin sperm histones Sperm heads were obtained from H. pulcherrimus spermatozoa as described previously [38]. Total histones were extracted from the sperm heads with 0.2 N H,SOx,, precipitated with final 20% (w/v) TCA, and washed with acetone, as described by Green et al. [13]. Total sperm histones were dissolved in 10 mM Tris-HCl, pH 8.0 and perchloric acid (PCA) was added to final 5%, followed by incubation on ice for 1hr. The precipitate containing core histones was re- moved by centrifugation at 10,000g for 10min, and sperm H1 histones were precipitated from the clear supernatant by the addition of final 20% (w/v) TCA and incubation on ice for 1 hr. Sperm histone H1, which migrated at 33 kDa in SDS-PAGE, were collected by centrifugation at 10,000 xg for 10 min, washed three times with acetone, and dried in air. Other methods Intensity of darkness of the band on the autoradiogram was measured quantitatively by a Shimadzu CS-9000 dual wavelength flying spot scanner at an absorbance of 600nm. Protein concentra- tion was determined by the method of Lowry et al. [22] modified by Schacterle and Pollack [32]. RESULTS Phosphorylation and dephosphorylation of sperm proteins When CHAPS-solubilized sperm proteins were incu- bated with [y-’P]ATP in 10mM Mg** medium, a 48 kDa protein was phosphorylated within 1 min and no other pro- teins appeared to be phosphorylated even after incubation for up to 10 min (Fig. 1). Addition of cAMP or cGMP to the reaction mixture resulted in a rapid dephosphorylation of the protein (Fig. 1). The half-maximal effective concentrations of cAMP and cGMP to trigger dephosphorylation of the protein were 0.34M and 4uM, respectively (Fig. 2). When Mg*" in the incubation mixture was replaced by Mn?", the 48 kDa protein was less phosphorylated (data not shown). PKI (94M) which is known to be a potent inhibitor for A-kinase [5] did not inhibit phosphorylation of the 48 kDa protein (Fig. 3). On the other hand, when CHAPS-solubilized sperm proteins were incubated with [y--’-P]ATP in 100 mM Mg?* -medium, addition of cAMP or cGMP to the reaction mixture T. Harumi, K. HosHIno AnD N. Suzuki 212 Control e 205K= 116K= | 66k= & pa | @Se@eeeee@e@ ec @ =<48K 29K= = - 14.2K= +cGVP 8 eeoeee — =A8K +camp & @eeoee - ~48K Fic. Fic. Ss | 23°45 6 78,9 40 Time (min) 1. Time course of phosphphorylation and cyclic nucleotide- dependent dephosphorylation of a sperm protein. Sup. A was incubated with [y-*’P]ATP as described in MATERIALS AND METHODS. At 5 min of the incubation, cAMP or cGMP was added to the reaction mixture to give a final 100uM. Then, TCA (final 10%, w/v) was added at 6 through 10 min of the incubation to stop the reaction. About a 30ug of protein sample was applied per a lane of the SDS-gel. The gel, after being dried, was exposed to a Kodak X-Omat film. S; silver- stained gel. 1-10; autoradiograms. 100 80 e 26 (e) O ia Ps 20 : 8 Tl 6 8 4 —loglcyclic nucleotide] (M) 2. Concentration-dependent effects of cAMP and cGMP on the dephosphorylation of the [*?P]-phosphorylated 48 kDa pro- tein. At Smin of the incubation of Sup. A (100s1) with [y-°P]ATP (0.83 pmol), cAMP or cGMP at the indicated con- centration was added to the reaction mixture and the incubation was continued for another 5 min. The reaction was terminated by addition of TCA (final 10%, w/v) and the resulting protein precipitate, after being rinsed with ice-cold acetone, was analy- zed by SDS-PAGE and then autoradiography. Intensity of darkness of the band on the autoradiogram was determined by scanning densitometry and expressed as the ratio to the darkenss of the band obtained in the absence of cyclic nucleotides. Fic. 4. cAMP -+- — fer =S GNP === =P Seer eee ee a LJ 205K- © oe 116K- = _ 97.4K~ — hs 66K Rh... i a Kee = o .Fe =48K 29K~ — -= = - =33K 14.2K- _# Liomm Mg2+) —100mm mg2+——! . 3. Effects of cAMP, cGMP, and PKI on protein phosphoryla- tion in sea urchin spermatozoa. CHAPS-solubilized sperm proteins (100) in 10 mM or 100 mM Mg’t-containing medium were incubated with 0.83 pmol of [y--’P]ATP (92.5 KBq) in the presence or absence of PKI at 20°C. At5 min of the incubation 2] of 5 mM cAMP or cGMP were added to the reaction mixture and the incubation was continued for another 5min. The reaction was stopped by the addition of 201 of 60% (w/v) TCA, and the resulting protein precipitate, after being rinsed with ice-cold acetone, was subjected to SDS-PAGE and auto- radiography. 1253-4556 noc AY ex. @ = 97. 4K> — = m& ite 66K> oe 4 a 29K == - —-33K 14.2K> Phosphorylation of a 33 kDa protein in H. pulcherrimus spermatozoa. CHAPS-solubilized sperm proteins in 100ml of 100 mM Mg?** -containing medium was incubated with 0.83 pmol of [y-*’P]ATP (92.5 KBq) at 20°C for 5 min, and then 2yl of 5 mM cAMP were added to the reaction mixture. At 5 min after addition, the reaction was stopped by addition of final 10% (w/ v) TCA or 5% (v/v) PCA. The PCA-soluble protein was pelleted by the addition of final 20% (w/v) TCA. Protein samples were analyzed by SDS-PAGE and autoradiography. 1, 2 and 3, siver-stained gels. 4, 5 and 6, its autoradiogram. 1 and 2, 10% TCA-insoluble proteins; 2 and 5, 5% PCA-insoluble pro- teins; 3 and 6, 5% PCA-soluble and 20% TCA-insoluble pro- teins. Sea Urchin Sperm Histone Kinase resulted in dephosphorylation of the 48 kDa protein and phosphorylation of a 33 kDa protein, and this phosphoryla- tion of the 33 kDa protein was blocked by 94M of PKI (Fig. 3). The [**P]-phosphorylatable 33 kDa protein was soluble in 5% (w/v) PCA but insoluble in 20% (w/v) TCA (Fig. 4). Identification of the [*°P|-phosphoamino acid in the 48 kDa protein To identify [*’P]-phosphoamino acid in the 48 kDa pro- tein, the [°*P]-phosphorylated 48 kDa protein band was ex- cised from the gel and hydrolysed in 5.7 N HCI. Since it has been reported that a long term hydrolysis resulted in degrada- tion of P-Tyr [29], the [°’P]-phosphorylated 48 kDa protein was hydrolyzed for 1 hr or 4 hrs at 110°C and the hydrolysate was analysed by HPLC on an anion exchange column (What- man Partisil 10 SAX). Two radioactive peaks were obtained in the sample from 1 hr hydrolysis: one peak (Peak 1 ) prior to P-Tyr and another peak (Peak 2) corresponding to P-Ser. The peak corresponding to P-Thr did not have radioactivity. P-ThrP-Tyr P-Ser An a a 03 E 300 = S N 02 E aa 2005 — ® oO > ul 2 o 01 a 5 100 2 Q fe) < o OL tea) (0) jag co 0 2 0 4 So Go 7 Elution volume (ml) Fic. 5. HPLC profile of the hydrolysate of the [°*P]-phosphorylated 48 kDa protein with P-Thr, P-Tyr and P-Ser on a Whatman Partisil 10 SAX column. The [**P]-phosphorylated 48 kDa protein excised from the SDS-gel and authentic P-Ser, P-Thr and P-Tyr were hydrolysed in constant-boiling HCl for 4hrs. The hydrolysate was applied to the column as decribed in MATE- RIALS AND METHODS. 213 As hydrolysis was continued for 4 hrs, Peak 1 became smaller while Peak 2 became larger. Fig. 5 shows the chromatogram of 4hrs hydrolyzed sample. Since Peak 1 did not overlap with the peak of P-Tyr in both samples from 1 hr and 4 hrs hydrolysis, we concluded that serine residues in the 48 kDa protein were phosphorylated by [y-P]ATP. Effects of PPase inhibitors on cyclic nucleotide-dependent dephosphorylation of the [?*P|-phosphorylated 48 kDa protein Fluoride, zinc and vanadate inhibit PPase activity. Table 1 shows the effects of other agents including these on cyclic nucleotide-dependent dephosphorylation of the [*’P]- phosphorylated 48 kDa protein. In the experiments without cyclic nucleotides, addition of 20 mM molybdate or vanadate resulted in dephosphorylation of the [°*P]-phosphorylated 48 % of control 8 7 6 5 —log[Okadaic acid] (M) Fic. 6. Dose-dependent effects of calyculin A and okadaic acid on the cyclic nucleotide-dependent dephosphorylation of [*P]- phosphorylated 48kDa protein. After 5 min incubation of Sup. A (1001) with [y-*P]ATP, various concentrations of calyculin A (left panel) with final 100~M of cAMP (@) or cGMP (4), or okadaic acid (right panel) with final 10042M of cAMP (@) or cGMP (4) were added to the reaction mixture and the incubation was continued for another 5 min. The reaction was stopped by the addition of final 10% (w/v) TCA. The proteins in the precipitate was analyzed by SDS-PAGE and autoradiogra- phy. Intensity of darkness of the band on the autoradiogram was quantitatively measured by scanning densitometry, and expressed as the ratio to the value for the band obtained in the absebce of cyclic nucleotides and inhibitors. 8 7 6 5 -log[Calyculin A] (M) TABLE 1. Effect of various inhibitors on cyclic nucleotide-dependent dephosphorylation of the [??P]-phosphorylated 48 kDa protein” Control +cAMP + cGMP None 100% 0% 1% (CH3COO),Zn 20 mM 79 78 49 2 56 11 6 0.2 68 0 0 Na,MoO, 20 50 0 0 Na3VO, 20 52 18 6 NaF 100 80 5 21 50 69 1 2) 20 74 0 0 B-glycero- 50 89 0 5 phosphate 1) Values are expressed as the ratio to the remaining [*”P]-radioactivity in the 48 kDa protein as 100%. 214 kDa protein up to about 50% of no addition. EDTA also induced dephosphorylation of the protein up to 15% of control. In the experiments with cyclic nucleotides, zinc inhibited cAMP-dependent dephosphoryration of the Pub phosphorylated 48kDa protein completely and cGMP- dependent dephosphorylation of the protein partially (49% of control). High concentration (100 mM) of fluoride or vana- date (20mM) slightly inhibited both cAMP- and cGMP- dependent dephosphorylation of the [°?P]-phosphorylated 48 kDa protein. (-Glycerophosphate (SO mM) exhibited no T. Harumi, K. HosHINo AND N. SUZUKI significant inhibitory effect on the dephosphorylation. Calyculin A and okadaic acid obtained from sea sponges are potent and specific inhibitors for many PPases [3, 7, 16]. Fig. 6 shows the effects of varying concentrations of calyculin A and okadaic acid on cyclic nucleotide-dependent dephos- phorylation of the [**P]-phosphorylated 48 kDa protein. Calyculin A inhibited both cAMP- and cGMP-dependent dephosphorylations at the same concentration (ICso= 0.24M). Okadaic acid also inhibited both cAMP- and cGMP-dependent dephosphorylations, but the apparent ICs A aes 158K eM fs 7 An 3 E ue 5 256 ©. o ~— 2 < = 05 o£ 8 ge 5 Me fa} i \ aE as c (o} S 2 — g a 15 [20 25 30 a 205K— = a ek 116K— — = 97.4K— [= ss 66K— c.f LS BSeeS=s = 45K— aes =i 2?22S=_ 29K——— a fore) —48K | Control a= @ @ @ @ @ @ @ @ @ —48K +cAMP — = mone 10 11 12 13 14 15 16 17 18 19 Fraction No. Fic. 7. HPLC profile of Sup. B on a TSK-GEL G3000SW column and the SDS-PAGE pattern. Fractions obtained from the column were determined for PPase activity (A), 48 kDa protein-phosphorylating activity (B), and cyclic nucleotide- dependent dephosphorylation activity for the [*’P]-phosphorylated 48 kDa protein (C) as described in MATERIALS AND METHODS. Sea Urchin Sperm Histone Kinase 215 of okadaic acid was 4-6 fold higher (0.8-1.24.M) than that of calycylin A. Purification of the [°°P|-phosphorytable 48 kDa protein In preliminary experiments, we estimated the molecular mass of the [*’P]-phosphorylatable 48 kDa protein by gel filtration. Sup. B was incubated with [y-*°P]ATP and then subjected to TSK-GEL G3000SW gel chromatography. As shown in Fig. 7A and 7B, most of the [*’P]-phosphorylated 48 kDa protein was recovered in the fraction (No. 13) in which proteins with molecular masses from 250 kDa to 400 kDa were eluted. PPase activity was recovered in the fraction (No. 10) in which proteins with molecular masses of over 400 kDa were eluted. In different experiments, dephosphorylat- ing activity for the [*°P]-phosphorylated 48 kDa protein in all the fractions obtained from the TSK-GEL G3000SW column was examined with the fraction (No. 13) as the substrate (Fig. 7C). In the absence of cyclic nucleotides, none of the fractions tested showed dephosphorylating activity for the [°°P]-phosphorylated 48 kDa protein. However, addition of cGMP or cAMP to the fractions which contained proteins with the molecular masses of over 400 kDa induced dephos- phorylation of the [°*P]-phosphorylated 48 kDa protein. To purify the [*’P]-phosphorylatable 48 kDa protein, CHAPS-solubilized sperm protein prepared from 50 ml of dry sperm was centrifuged at 30,000xg and the resulting supernatant fluid was applied to a DEAE-Sephacel column. The [*P]-phosphorylatable 48 kDa protein was eluted at 150-240 mM NaCl. PPase activity was also eluted at the same concentration of NaCl. Fractions containing the 48 kDa protein also contained PPase activity. Further purifica- tion of the protein was carried out by ion exchange chroma- tography on a Mono Q HRS/S column. The 48 kDa protein was eluted at 280-330 mM NaCl, in which a major PPase activity was also eluted. To remove the PPase activity, the fractions containing the 48 kDa protein were subjected to chromatography on a histone-agarose column. In this chro- 2 3 4 Sl Absorbance at 280nm (——) Protein phosphatase activity (nmol Pi/min/ml) (—<-) (2) 0 20 30 40 50 Fraction No. Fic. 8. HPLC profile of the 48 kDa protein-containing fraction on a Superose 6 HR10/30column. Fractions obtained from the column were assayed for phosphorylating activity of the 48 kDa protein. Dotted area indicates the fractions which contained practically pure 400 kDa protein consisting of the 48 kDa and 39 kDa proteins. The 400 kDa protein was phosphorylated with [y-°P]ATP and then analyzed by SDS-PAGE, followed by autoradiography. Silver-stained gel (S) and its autoradiogram (AR) of the [°*P]-phosphorylated 48 kDa protein are shown in insert in the figure. Triangle points on the left side of the gel in insert indicate standard proteins as the same as in Fig. 3. Molecular weight standards used for the gel chromatography were thyroglobulin (669 kDa), ferritin (440 kDa), catalase (232 kDa), bovine serum albumin (67 kDa), and chymotrypsinogen A (25 kDa). matography, the 48 kDa protein was obtained mainly in the fractions eluted at 320-440 mM NaCl. However, these frac- tions still contained PPase activity, although a major PPase activity was eluted after the 48 kDa protein. The 48kDa protein and PPase activity were separated by chromatography on a Superose 6 HR10/30 column (Fig. 8). In this chroma- Sl we 3 4 205K- 116K— 97.4K— 66K— —48K 29K— Fic. 9. SDS-PAGE and its autoradiogram of 8-N*-[?*P]cAMP or 8-N3-[*7P]cGMP-treated CHAPS-solubilized sperm proteins (left panel) or the purified 400 kDa protein (right panel). Protein samples were incubated with 1: 66nM 8-N;- [P]cAMP, 2: 66 nM 8-N;-[**P]cAMP and 204M cAMP, 3: 0.4”M 8-N;-[°7P]cGMP, 4: 0.4uM 8-N;-[°2P]cGMP and 204M cGMP. S; silver stained gel. 216 tography, the 48 kDa protein was eluted as a single protein peak at molecular mass of approximately 400kDa. The protein with the molecular mass of 400 kDa was separated to 48 kDa and 39 kDa proteins by SDS-PAGE. Incubation of the 400 kDa protein with [y-*’P]ATP resulted in phosphoryla- tion of only the 48 kDa protein. Incorporation of 8-N3-[*”?P|cAMP and -[??P|cGMP into sperm proteins When Sup. B or the purified 400kDa protein was incubated with a photoaffinity cAMP analogue 8-N3- [°°P]c AMP or cGMP analogue 8-N3-[°-P]cGMP, cAMP ana- logue was incorporated into the 48 kDa protein (Fig. 9, left panel). When the purified 400 kDa protein was incubated with the cyclic nucleotide analogues, both of the 48 kDa and the 39 kDa proteins were labeled (Fig. 9, right panel). The binding of 8-N3[°*P]cAMP to these proteins was mostly blocked in the presence of 204M cAMP. Properties of the autophosphorylatable sperm protein The purified 400 kDa protein which consisted of the [°*P]-phosphorylatable 48 kDa protein and the 39 kDa pro- tein showed histone kinase activity. Table 2 shows the abil- ity of the histone kinase to phosphorylate various histone types. The lysine rich histones (H1 and H2B) were the best substrates, while the arginine rich histones (H2A and H3) were poor substrates. The kinase did not phosphorylate histone H4. Total histones, core histones and histone H1 isolated from sea urchin sperm heads were also tested for substrates. The sea urchin sperm-derived histones were also good substrates for the kinase (Table 2). Phosphorylation of these histones by the kinase in the presence of cAMP was T. Harumi, K. HosHINo AND N. Suzuki 2 oO ro) re) = SO rc 5 10 Cyclic nucleotide (uM) Fic. 10. Concentration-dependent effects of cAMP and cGMP on activation of sea urchin sperm histone kinase to phosphorylate H. pulcherrimus sperm histone H1. Reaction was started by addition of 41.9 ng of purified sperm histone kinase to 18.51 of reaction mixture containing 1 mg/ml sea urchin sperm histone H1, 10 mM (CH3;COO),Mg, 0.2 mM [7-**P]ATP, indicated con- centration of cAMP/cGMP and 50 mM MES, pH 6.5). After incubation at 30°C for 10 min, the reaction was terminated by heating at 100°C for Smin. A 10yl-aliquot was withdrawn, spotted onto a 1X2 cm strip of Whatman p 81 phsophocellulose and then immersed in 75 mM phosphoric acid. The strip was subjected to four washes of 75 mM phosphoric acid and then air-dried, followed by measuring the radioactivity by the liquid scintillation counter. cAMP (©), cGMP (@). Activity (nmol incorporated Pi / jug protein) as well as cAMP in a concentration-dependent manner, but cGMP was less potent thancAMP. Maximal response of the blocked by 94M of PKI. The kinase was activated by cGMP kinase to cGMP was seen at 10M (Fig. 10). cAMP- or TABLE 2. Substrate specificity of cAMP-dependent histon kinase Calf thymus histones Total histone H1 H2A H2B H3 H4 cAMP inhibitor nmol Pi incorporated/g protein + a 20.6 17.2 5.5 257) 1.4 0.0 + + 0.1 0.1 0.1 “0.1 0.0 0.0 — - 0.9 0.6 0.2 0.9 0.0 0.0 = ar 0.0 0.0 0.1 0.0 0.0 0.0 Sea urchin sperm histones Total histone Core histone H1 cAMP inhibitor nmol Pi incorporated/«g protein - _ 6.8 4.5 7.1 + _ 34.0 37.5 5357 ap 37 ito 7 0.6 Purified enzyme (56 ng in Sl) was added to 951 of reaction mixture containing 1 mg/ml calf thymus or sea urchin sperm histones, 0.2 mM ATP (74 kBq [y-’ P]ATP), 10 mM (CH,COO),Mg and 50 mM MES (pH 6.5) in the presence or absence of 24M cAMP and/or 9M protein kinase inhibitor (TTY ADFIA- SGRTGRRNAIHD), and incubated at 30°C for 10 min. An aliquot of the reaction mixture (251) was transferred onto a Whatman P81 The filter was washed with 75 mM phosphoric acid five times, dried in air, for 5 min. phosphocellulose filter paper. and the radioactivity was measured. The reaction was stopped by heating at 100°C Sea Urchin Sperm Histone Kinase 217 cGMP-dependent histone kinase activity required Mg** more than 5mM. The histone kinase was incubated for 2 min on ice in the presence of 1~M cAMP or cGMP, and the reaction mixture was then subjected to gel chromatography on a Superose 6 HR10/30 column equilibrated with Buffer B containing SuM cAMP. Chromatography with 5uM cAMP gave a peak histone kinase activity at molecular mass of about 39 kDa which was active even in the presence of cAMP less than 0.254M (Fig. 11). However, treatment of the kinase with 10“M cGMP did not result in shift of kinase activity from 400 kDa to 49 kDa. 669k 440k 232k 67k 25k “tt ae e ps e \ xe) 3 | . 5 AOR iba or / \ } Oo e oe \ aaa) “eo ‘e ai Eh ST ea a 0 5 500 30 40 5 = e s N e r g 2 = e L ‘ / \° i oo? @ OL ecoe 0-09 30 40 50 Fraction No. Fic. 11. Gel filtration profiles of sea urchin histone kinase on a Superose 6 HR10/30 column in the presence or absence of cAMP. Autophosphorylatable 48 kDa protein-containing frac- tions obtained from chromatography on a histone-agarose col- umn were pooled and concentrated, and then subjected to a Superose 6 HR10/30 column equilibrated with Buffer B (upper panel). Purified sea urchin histone kinase was incubated in Buffer B containing 14M cAMP at 0°C for 20 min and then subjected to chromatography on the Superose column equili- brated with Buffer B containing 5uM cAMP (lower panel). An aliquot of each fraction was assayed for histone kinase activity using sea urchin sperm histone H1 as substrate in the presence of 2uM cAMP (upper panel) or 0.254M cAMP (lower panel). Molecular weight standards used for the gel chromatography were thyroglobulin (669 kDa), ferritin (440 kDa), catalase (232 kDa), bovine serum albumin (67 kDa) and chymotrpsinogen A (25 kDa). DISCUSSION When detergent-solubilized sea urchin sperm proteins are incubated with [y-*’P]ATP, a 42 kDa protein and histone H1 in S. purpuratus [27-28] and a 45 kDa protein in Arbacia punctulata [43] were phosphorylated. Furthermore, phos- phorylation of histone H1 in S. purpuratus has been reported to be cAMP-dependent [27-28]. Here, we demonstrated that in H. pulcherrimus spermatozoa a 48 kDa protein and a 33 kDa protein were phosphorylated but phosphorylation of the latter protein by an endogenous kinase was observed only in 100 mM Mg*t-containing medium while phosphorylation of the 48 kDa protein occurred irrespective of the concentra- tion of Mg**. The 33kDa protein could be histone H1 based on the following properties: 1) it is PCA-soluble and TCA-incoluble, 2) it is cAMP-dependently phosphoryltable, and 3) it has correct apparent molecular size [17, 25, 27, this study]. Unlike the phosphorylation of the 33 kDa protein, which was PKI-inhibitable, phosphorylation of the 48 kDa protein which was not inhibited by PKI, occurred only in the absence of cyclic nucleotides. Even after the 400 kDa pro- tein was purified, 48 kDa protein was phosphorylated by incubation with [y---P]ATP. Therefore, the phosphoryla- tion of the 48 kDa protein may be autophosphorylation. Addition of cAMP or cGMP triggered dephosphorylation of the [°’P]-phosphorylated 48 kDa protein in a concentration- dependent manner. Since calyculin A and okadaic acid, which are potent inhibitors specific for serine/threonine PPases, inhibited the cyclic nucleotide-dependent dephos- phorylation of the [°’P]-phosphorylated 48 kDa protein, such a type of PPase may be involved in dephosphorylation of the protein. Gel filtration of the [*’P]-phosphorylatable 48 kDa pro- tein demonstrated that it was associated with a 39kDa protein to form a 400 kDa protein molecule. After SDS- PAGE of the 400kDa protein, the 48kDa and 39kDa protein bands were stained by Coomassie brilliant blue to almost the same darkness on the gel, suggesting that the 400 kDa oligomer consists of an equal number of the 48 kDa and 39 kDa subunits. However, the 48 kDa protein was stained less intensively with silver than the 39 kDa protein, so careful quantitative analysis will be required to confirm this idea. Incorporation of a photoaffinity cGMP analogue 8-N3- [°°P]cGMP to sperm protein was not detected probably because of low specific radioactivity of the analogue. However, a photoaffinity cAMP analogue 8-N3-[*’7P]cAMP was incorporated into the same molecular mass protein as the protein phosphorylated with [y-’P]ATP. Similarly, 8-N3- [°?P]cAMP was incorporated into the 48 kDa protein in the purified 400 kDa protein. These results indicate that cAMP binds to the 48 kDa protein. In this regard, it should be mentioned that several enzymes and certain mambrane re- ceptor proteins have been known to be regulated by subs- trate-directedly [1, 4, 8, 18]. Thomas et al. [42] proposed that in the case of phosphorylation of bovine lung cGMP- binding cGMP-specific phosphodiesterase (CG-BPDE), the 218 T. Harumi, K. HosHino AND N. Suzuki binding of cGMP to cG-BPDE induces a conformational change in cG-BPDE to expose the phosphorylation site for the protein kinases. In the present study, we demonstrated that the protein consisting of [°’P]-phosphorylatable 48 kDa protein(s) and 39 kDa protein(s) phosphorylates calf thymus histones and also sea urchin sperm histones in cyclic nuc- leotide-dependent manner. In this regard, it should be mentioned that certain A-kinase activity is regulated by cAMP-dependent dephosphorylation of autophosphorylated enzyme [6, 31]. Type II A-kinase consists of an autophos- phorylatable regulatory (R) dimer and two catalytic (C) subunits. Binding of cAMP to R subunit triggers the dis- sociation of R, and C results in the elevation of protein kinase activity. Considering these results, our results presented here may be explained by the idea that the binding of cAMP to the 48 kDa subunit(s) in the 400 kDa kinase results in dissociation of the 48 kDa regulatory subunit(s) and 39 kDa catalytic subunit(s), which allows the autophosphorylated 48 kDa subunit(s) accessible to a protein phosphatase which is abundant in the detergent-solublized sperm fraction. An alternative explanation may be possible: sea urchin spermato- zoa may contain a novel cyclic nucleotide-dependent protein phosphatase. When we first obtained the results to show cyclic nucleotide-dependent dephosphorylation of the [°*P]- phosphorylated 48 kDa protein we thought that it might be explained by the latter intriguing idea. However, so far we only obtained three different protein phosphatases and none of phosphatases could be activated by cyclic nucleotides (unpublished observation). Therefore, at this moment we prefer the former idea. Sperm histone kinase isolated in this study was activated by both cAMP and cGMP, and phosphorylated sea urchin sperm histone H1 as well as calf thymus histones (H1 and H2B). It has been reported that the egg jelly of H. pulcher- rimus contains sperm-activating peptide I (SAP- I:GFDLNGGGVG) and FSG which elevate sperm cGMP and/or cAMP upto some uM levels (14, 36-37). Green and Poccia [12] reported that phosphorylation of sperm histones H1 and H2B of S. purpuratus precedes sperm chromatin decondensation and histone H1 exchange during pronuclear formation. Therefore, we presume that upon fertilization sperm histone kinase isolated here may be activated in the spermatozoon by SAP-I-elevated cGMP and/or cAMP or FSG-elevated cAMP and phosphorylates these histones be- fore entering sperm nuclei into the egg. ACKNOWLEDGMENTS We would like to thank the staff members of the Radioisotope Center, Kanazawa University, for technical supports throughout this work, and Dr. Daniel Hardy (University of Texas Southwestern Medical Center) for critical reading of this manuscript. We also thank Mr. M. Matada for collecting and culturing sea urchins. This work was supported by Grants-in-Aid 02404006 and 02044059 from the Ministry of Education, Science and Culture of Japan. 10 11 12 13 14 15 REFERENCES Benovic JL, Regan JW, Matsui H, Mayor FJr, Cotecchia S, Leeb-Lundberg LMF, Caron MG, Lefkowitz RJ (1987) Agon- ist-dependent phosphorylation of the aj-adrenergic receptor by the f-adrenergic receptor kinase. J Biol Chem 262: 17251- 17253 Bentley JK, Tubb DJ, Garbers DL (1986) Receptor-mediated activation of spermatozoan guanylate cyclase. J Biol Chem 261: 14859-14862 Bialojan C, Takai A (1988) Inhibitory effect of a marine- sponge toxin, okadaic acid, on protein phosphatases. Biochem J 256: 283-290 Bownds D, Dawes J, Miller J, Stahlman M (1972) Phosphory- lation of frog photoreceptor membranes induced by light. Na- ture 237: 125-127 Cheng H-C, Kemp BE, Pearson RB, Smith AJ, Misconi L, Van Patten SM, Walsh DA (1986) A potent synthetic peptide inhibitor of the cAMP-dependent protein kinase. J Biol Chem 261: 989-992 Chou C-K, Alfano J, Rosen OM (1977) Purification of phos- phoprotein phosphatase from bovine cardiac muscle that cataly- zes dephosphorylation of cyclic AMP-binding protein compo- nent of protein kinase. J Biol Chem 252: 2855-2859 Cohen P (1989) The structure and regulation of protein phos- phatases. Annu Rev Biochem 58: 453-508 El-Maghrabi MR, Claus TH, Pilks SJ (1983) Substrate- directed regulation of cAMP-dependent phosphorylation. Methdos Enzymol 99: 212-219 Garbers DL, Kopf GS (1980) The regulation of spermatozoa by calcium and cyclic nucleotides. Ady Cyclic Nucleotide Res 13: 251-306 Garbers DL, Tubb DJ, Kopf GS (1980) Regulation of sea urchin sperm cyclic AMP-dependent protein kinases by an egg associated factor. Biol Reprod 22: 526-532 Garbers DL, Watkins HD, Hansbrough JR, Smith A, Misono KS (1982) The amino acid sequence and chemical synthesis of speract and of speract analogues. J Biol Chem 257: 2734-2737 Green GR, Poccia DL (1985) Phosphorylation of sea urchin sperm H1 and H2B histones precedes chromatin decondensation and H1 exchange during pronuclear formation. Dev Biol 108: 235-245 Green GR, Lee H-J, Poccia DL (1993) Phosphorylation weakens DNA binding by peptides containing multiple “SPKK” sequences. J Biol Chem 268: 11247-11255 Harumi T, Hoshino K, Suzuki N (1992) Effects of sperm- activating peptide I on Hemicentrotus pulcherrimus spermatozoa in high potassium sea water. Develop Grwth Differ 34: 163- 172 Ishiguro K, Murofushi H, Sakai H (1982) Evidence that cAMP-dependent protein kinase and a protein factor are in- volved in reactivation of Triton X-100 models of sea urchin and starfish spermatozoa. J Cell Biol 92: 777-782 Ishihara H, Martin BL, Brautigan DL, Karaki H, Ozaki H, Kato Y, Fusetani N, Watabe S, Hashimoto K, Uemura D, Hartshorne DJ (1989) Calyculin A and okadaic acid: inhibitors of protein phosphatase activity. Biochem Biophys Res Commun 159; 871-877 Johns EW (1977) The isolation and purification of histones. Methods Cell Biol 16: 183-203 Kuhn H, Dreyer WJ (1972) Light dependent phosphorylation of rhodopsin by ATP. FEBS lett 20: 1-6 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680- 685. 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Sea Urchin Sperm Histone Kinase 219 Lee MYW, Iverson RM (1972) Protein kinase in sea urchin gamets and embryos. Exptl Cell Res 75: 300-304 Lee MYW, Iverson RM (1976) An adenosine 3’ :5’ monophos- phate dependent protein kinase from sea urchin spermatozoa. Biochim Biophys Acta 429:123-136 Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the folin phenol reagent. J Biol Chem 193: 265-275 Meisler MH, Langan TA (1969) Characterization of a phos- phatase specific for phosphorylated histones and protamine. J Biol Chem 244: 4961-4968 Morrissey JH (1981) Silver stain for proteins in polyacrylamide gels: A modified procedure with enhanced uniform sensitivity. Anal Biochem 117: 307-310 Ozaki H (1971) Developmental studies of sea urchin chroma- tin. Chromatin isolated from spermatozoa of the sea urchin Strongylocentrotus pururatus. Dev Biol 26: 209-219 Poccia DL, Green GR (1992) Packaging and unpackaging the sea urchin sperm genome. TIBS 17: 223-227 Porter DC, Vacquier VD (1986) Phosphorylation of sperm histone H1 is induced by the egg jelly layer in the sea urchin Strongylocentrotus purpuratus. Dev Biol 116: 203-212 Porter DC, Moy GW, Vacquier VD (1988) cAMP-dependent protein kinase of sea urchin sperm phosphorylates sperm histone H1 on a single site. J Biol Chem 263: 2750-2755 Robert JC, Soumarmon A, Lewin (1985) Determination of o-phosphothreonine, o-phosphoserine, o-phosphotyrosine and phosphate by high-performance liquid chromatography. J Chromatogr 338: 315-324 Roskoski RJr (1983) Assays of protein kinase. zymol 99: 3-6 Scott CW, Mumby MC (1985) Phosphorylation of type II reguratory subunit of cAMP-dependent protein kinase in intact smooth muscle. J Biol Chem 260: 2274-2280 Schacterle GR, Pollack RL (1973) A simplified method for the quantitative assay of small amounts of protein in biologic material. Anal Biochem 51: 654-655 SeGall GK, Lennarz WJ (1979) Chemical characterization of the component of the jelly coat from sea urchin eggs responsible for induction of the acrosome reaction. Dev Biol 71: 33-48 SeGall GK, Lennarz WJ (1981) Jelly coat and induction of the Methods En- 35 36 37 38 39 40 41 42 43 44 45 46 acrosome reaction in echinoid sperm. Dev Biol 86: 87-93 Shacter E (1984) Organic extraction of P; with isobutanol/ toluene. Anal Biochem 138: 416-420 Shimizu T, Kinoh H, Yamaguchi M, Suzuki N (1990) Purifica- tion and characterization of the egg jelly macromolecules, sialglycoprotein and fucose sulfate glycoconjugate, of the sea urchin Hemicentrotus pulcherrimus. Develop Growth Differ 32: 473-487 Suzuki N, Nomura K, Ohtake H, Isaka S (1981) Purification and the primary structure of sperm-activating peptides from the jelly coat of sea urchin eggs. Biochem Biophys Res Commun 99: 1238-1244 Suzuki N, Kurita M, Yoshino K, Yamaguchi M (1987) Speract binds exclusively to sperm tails and causes and electrophoretic mobility shift in a major sperm tail protein of sea urchins. Zool Sci 4: 641-648 Swarup G, Garbers DL (1982) Phosphoprotein phosphatase activity of sea urchin spermatozoa. Biol Reprod 26: 953-960 Swarup G, Cohen S, Garbers DL (1981) Selective dephos- phorylation of proteins containing phosphotyrosine by alkaline phosphatases. J Biol Chem 256: 8197-8201 Takai Y, Nishiyama K, Yamamura H, Nishizuka Y (1975) Guanosine 3’:5’-monophosphate-dependent protein kinase frrom bovine cerebellum. Purification and characterization. J Biol Chem 250: 4690-4695 Thomas MK, Francis SH, Corbin JD (1990) Substrate- and kinase-directed regulation of phosphorylation of a cGMP- binding phosphodiesterase by cGMP. J Biol Chem 265: 14971- 14978 Vacquier VD, Porter DC, Keller SH, Aukerman M (1989) Egg jelly induces the phosphorylation of histone H3 in spermatozoa of the sea urchin Arbacia punctulata. Dev Biol 133: 111-118 Yang JC, Fujitaki JM, Smith RA (1982) Separation of phos- phohydroxyamino acids by high-performance liquid chroma- tography. Anal Biochem 122: 360-363 Watkins HD, Kopf GS, Garbers DL (1978) Activation of sperm adenylate cyclase by factors associated with eggs. Biol Reprod 19: 890-894 Weber K, Osborn M (1969) The reliability of molecular weight determination by dodecyl sulfate polyacrylamide gel elec- trophoresis. J Biol Chem 244: 4406-4412 - Baan NET a pa aealved We Ni edeaol L°YT Few it Pibicsvovi ae ® Slrivihenchoa dip Regd | iene: ena nee, Lvs ae goal % vise late a : > = ¥ - by ‘ - (tye ane) J ase E inte i 4 . bry ae ae We | Autry “7 BT pi pyol tas | rif at F a, yy lines q | 2 a oticanieesteeil Nts cents uth shyhtuagpterms VCE cebtideny uN: stead At hall et farang pot oe at lmprtoode: tyra f tia abit soma. El Bowe ibe Wk one nuke Rares ae vasibia hs Mo OST CE Big» 46 wih cs we NE Ps os waa " vet ih sarees ee > ca - phived dey ie i Le Topics ous 4 & ‘a Rivals wet 1 ' 4 Ls 4 ees 2 E i ey p: = if : 4 ata 1 oa La ve i . m iA “ ‘ sie a> Hel ¢ oy fey 3 be { 0 ay Mi , gis j St ratve rte aang 4 4 buy duc d 4 e's - ws TA LC sie ss f bs J i ; £ te Luang ast Sa mechan: | iba a wr oF i" a hye ’ a? Utley amd purihca | ; weary tig dicts Tee hy APT PPSNS hat MY, PG les Oech th Av Mphatas “4 | \ 1 Cece obitrantin'nd i i) | ZOOLOGICAL SCIENCE 11: 221-227 (1994) © 1994 Zoological Society of Japan Rapid and Quantitative Detection of Aspartic Proteinase in Animal Tissues by Radio-labeled Pepstatin A Masanori Mukat!, TosHiHiko Konpo” and Katsutosu! Yosuizato!? 'Molecular Cell Science Laboratory, Department of Biological Science, Faculty of Science, Hiroshima University, 1-3-1 Kagamiyama, Higashihiroshima-shi, Hiroshima 724 and *Department of Physical Biochemistry, Institute of Endocrinology, Gunma University, Maebashi 371, Japan ABSTRACT—A new radio-derivative of pepstatin A was developed and was shown to be used as a probe for rapidly and quantitatively detecting aspartic proteinases in animal tissues. The carboxyl group of pepstatin A was activated by the water soluble carbodiimide, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride and was then coupled with N-hydroxysulfosuccinimide (Sulfo NHS). [°°S]-Methionyl-pepstatin with a relatively high specific activity was obtained by coupling the Sulfo NHS-pepstatin with L-[*°S]-methionine. Binding specificity of the pepstatin A derivative was characterized using pepsin as a test enzyme. Binding experiments showed that the radio-labeled pepstatin can be used as a probe which binds specifically to aspartic proteinases and detects them. The probe could detect as low as 0.1 nmoles of pepsin. To know whether the radio-labeled probe can be actually used to detect aspartic proteinases in animal tissues, it was applied to the tail tissue of metamorphosing amphibian tadpole which had been known to show high activity of cathepsin D. The assay demonstrated marked increase in aspartic proteinases at the climax stage. It was concluded that the radio-labeled pepstatin derivative developed by the present study is useful for quick and quantitative determination of pepstatin-reactive enzymes in animal tissues. INTRODUCTION The aspartic proteinase is a group of proteinases (EC 3.4. group 23) which has an optimum pH at an acidic region. The family contains pepsin, cathepsin D, cathepsin E, renin, chymosin and gastricsin. Amino acid sequences of mamma- lian aspartic proteinases and the corresponding nucleotide sequences of their cDNA clones revealed that they have the active center, which contains two aspartic residues and are highly conserved among aspartic proteinases [4, 8, 10, 12, 32, 33, 35]. Pepstatin A is produced by actinomyces [37] and is a potent inhibitor of pepsin (Ki=4.5x10~'! M), cathepsin D (Ki=10~'°M) and other aspartic proteinases [3, 27]. Pepstatin A binds to the active center of the enzyme in a stoichiometric manner which is surrounded by two aspartic moieties [3, 19, 37]. Radio-labeled pepstatin derivatives have been chemical- ly synthesized. Pepstatin was coupled with [°H]-glycine and used for determining Kd values of a complex of pepstatinyl- [H]-glycine-cathepsin D [15]. A radio-iodinated derivative of pepstatin was prepared by introducing a tyrosine residue which was then iodinated with '*I. This was utilized for the determination of Kd of pepsin-pepstatin-[’*I]-monoiodo- tyrosine methyl ester complex [41]. However, *H-labeled compound is chemically unstable because tritium is exchange- able with surrounding hydrogen. '°I-Labeling has dis- advantages because of its short half life of radioactivity. Accepted March 11, 1994 Received January 27, 1994 Chemical modification of pepstatin has been designed to utilize the inhibitor as an experimental probe which tightly binds to the active site of aspartic proteinases and carries reporter groups [16]. Bimane-labeled pepstatin [21] and dinitrophenyl-pepstatin [22] have been synthesized and used for the determination of the subcellular location of cathepsin D in cultured human synovial cells. These methods of modifications require relatively complicated techniques of organic chemistry and contain multiple steps including techni- ques of column chromatography. Because both pepstatin derivatives described above are detected by fluorescence, these derivatives cannot be used as a probe in cases where samples to be analyzed contain fluorescent materials. In the present study the authors aimed at developing a much simple radio-labeling procedure of pepstatin to obtain a stable derivative. For this purposes, 1-ethyl-3-(3-dimethyl- aminopropyl) carbodiimide hydrochloride (EDC) [42] and N-hydroxy-sulfosuccinimide (Sulfo NHS) [2, 34] were selected as activating agents since they are easy to handle because of their hydrophilicity. The activated pepstatin was coupled with L-[?°S]-methionine which has longer half life of decay than ‘I. In the amphibian metamorphosis, the larval tail tissue is subject to the histolysis by several proteinases such as collage- nase [11, 25] and cathepsin D [18, 29, 38]. This phe- nomenon was utilized to see validity and usefulness of the radio-labeled pepstatin derivative for detecting proteinases in animal tissues. The present study succeeded in demonstrat- ing that the radio-labeled pepstatin is a useful probe for quick and quantitative detection of aspartic proteinases in animal tissues. 222 M. Mukai, T. Konpo AND K. YOSHIZATO MATERIALS AND METHODS Materials Pepstatin A was purchased from Peptide Institute, Inc. (Osa- ka, Japan). EDC and Sulfo NHS were obtained from Pierce (Rockford, IL). L-[*°S]-methionine was from New England Nuc- lear (Boston, MA). SEP-PAK Cig was from Waters (Milford, MASS). Pepsin was a product of Wako Pure Chemicals (Osaka, Japan). All other reagents were of analytical grade. Tadpoles of bullfrog, Rana catesbeiana, were purchased from a local animal supplier. They were staged as described by Taylor and Kollros (TK stage) [36]. Coupling reaction Pepstatin A (0.7 mg, 1 wmoles) was dissolved in 200 pl of dimethyl sulfoxide. Sulfo NHS (10 ~moles) and EDC (100 pmoles) were dissolved in 800 1 of 16 mM sodium phosphate buffer, pH 7.4, containing 20% dimethyl sulfoxide. These were mixed together and were incubated at room temperature (ca. 20°C) for 1hr. The product of this reaction, Sulfo NHS-pepstatin, was diluted with 4 ml of 20 mM sodium phosphate buffer, pH 7.4, and was applied slowly to a SEP-PAK Cg column (1 drop/sec) using a syringe. The column was washed twice with 5 ml of 20mM sodium phosphate buffer, pH 7.4, for removing excess amounts of activating agents (Sulfo NHS and EDC). Sulfo NHS-pepstatin was eluted from the column twice with 2 ml of dimethyl sulfoxide, mixed with 4 ml of 10 mM sodium phosphate buffer, pH 7.4, containing L-[*°S]-methionine (100 pmoles, 4.1 MBq, 1.610° cpm) and was incubated at room temperature (ca. 20°C) for 2hr. The reaction mixture in which [°°S]-methionyl-pepstatin was produced was diluted with deionized water and loaded on a SEP-PAK Cig column in the same way as described above. The column was washed twice with 10 ml of deionized water. This step separated the labeled pepstatin from uncoupled L-[*°S]-methionine and other by-products which did not covalently bind to pepstatin A. The eluate containing [*°S]- methionyl-pepstatin (0.72 MBq, 2.8 10’ cpm/4 ml dimethyl sulfox- ide) was kept at —20°C before use. Separation of {*°S]-methionyl pepstatin A from free pepstatin A [?°S]-Methionyl pepstatin prepared as described above still con- tained excess amounts of pepstatin A. Free pepstatin A was re- moved mainly by HPLC as follows. The preparation was 4-fold diluted with deionized water and loaded on a SEP-PAK Cj, column which was then eluted with acetonitrile. The eluate containing pepstatin A and its [*°S]-methionyl derivative was evaporated to dryness at 40°C in vacuo and was dissolved in an appropriate amount of dimethyl sulfoxide. The concentrated radio-labeled pepstatin A was subjected to reversed phase high-performance liquid chroma- tography (HPLC, Inertsil 300-C,, 4.6x 100mm, Gasukuro Kogyo Inc. Tokyo, Japan) and eluted at 50°C with a linear gradient of acetonitrile (from 25 to 55%) containing 0.1% trifluoroacetic acid (TFA). The flow rate was 0.4 ml/min. The eluate was monitored by 220 nm for pepstatin A and by the radioactivity for the labeled one. Radioactive fractions were collected and concentrated by evaporation. Binding assays Appropriate amounts of pepsin (10-440 «g/20 ul H,O0/mem- brane) were spotted on nitrocellulose membranes (1.5 X1.5 cm’). Membranes were dried and incubated in 20mM sodium acetate buffer, pH 5.0, containing 1% bovine serum albumin (BSA) for 10 min on ice for protecting nonspecific binding of pepstatin A to the membranes. Then, the blocking buffer was replaced with 2 ml of 20 mM sodium acetate buffer, pH 5.0, which contained radio-labeled pepstatin A ([°°S]-methionyl-pepstatin A, 10*-10° cpm/assay). The membranes were incubated for additional 30 min on ice. Mem- branes were then washed 4 times on ice with 2 ml of 20 mM sodium acetate buffer, pH 5.0, to remove radio-labeled pepstatin A that had not bound to pepsin, and were dried and placed in 5ml of a scintillation cocktail (Atomlight, New England Nuclear, Boston, MA) to count radioactivities using a liquid scintillation counter (LSC-3000, Aloka). Nonspecific binding of radio-labeled pepstatin A to the nitrocellulose membrane was estimated through the identi- cal procedure described above with an exception that membranes were washed with neutral buffer (20 mM sodium phosphate buffer, pH 7.4). Binding of [°°S]-methionyl-pepstatin to the tail tissue of bullfrog tadpole The tail of tadpole of bullfrog, Rana catesbeiana, was used as a source of aspartic proteinases for the binding assay of [*S]- methionyl-pepstatin A. Tails were dissected from bullfrog tadpoles at the premetamorphic stage (TK stage XVII) or the climax stage of metamorphosis (TK stage XXIII) and kept at —20°C until use. The frozen tail tissues were defrosted on ice and finely minced to small pieces (about 3X33 mm?) at 0-4°C with scissors in the solution containing 0.1 M NaCl and 20 mM sodium acetate buffer, pH 5.0. Tissue pieces were collected in microfuge tubes (Ca. 100 mg wet weight tissue/tube) and suspended in 1 ml of 20 mM sodium acetate buffer containing 0.1 M NaCl at0-4°C. The radio-labeled pepstatin A (2.0X10* cpm/assay) was added to the tissue suspension and allowed to stand for binding at 0-4°C for 30 min. For removal of unbound pepstatin probe, tubes were then centrifuged at 1,700Xg for 5 min at 4°C and the supernatants were discarded. The precipi- tates were washed 3 times with 1 ml of 20 mM sodium acetate buffer containing 0.1 M NaCl, pH5.0. Then, for recovering radio-labeled pepstatin A bound to tissues, precipitates were lysed in 600 ul of 0.1 N NaOH at 37°C for 2 hr and were centrifuged at 7,000 g for 10 min at room temperature. Radioactivity in the supernatant was counted. The protein concentration of lysates was determined by the procedure described by Lowry et al. using BSA as a standard [20]. RESULTS The carboxyl group of pepstatin A was activated by EDC to form o-acylurea which was then coupled with Sulfo NHS. Sulfo NHS-pepstatin thus obtained was more resistant to hydrolysis than o-acylurea and was hydrophobic. Excess amounts of hydrophilic activating reagents were easily re- moved from the reaction mixture using a reversed phase column of SEP-PAK C;g. Coupling of the activated pepsta- tin A with the amino group of L-[*°S]-methionine produced [*°S]-methionyl-pepstatin A with specific activity of 0.72 MBq/ mole (2.8 10’ cpm/mole). Specificity of the radio-labeled pepstatin A was charac- terized by analyzing its binding properties to pepsin that had been immobilized on a nitrocellulose membrane. The radio- labeled pepstatin was bound to the enzyme at pH 5.0, but not at pH 7.4, same as pepstatin A (Fig. 1). The binding of Radio-labeled Pepstatin A 223 pH 5.0 Bound (X10°cpm) 0 3 6 9 12 15 Pepsin (nmoles/assay) Fic. 1. pH-Dependent binding of [*°S]-methionyl-pepstatin A to pepsin. The radio-labeled derivatives (3.6 10* cpm/assay) were incubated with various amounts of pepsin (0.7-14.6 nmoles) that had been dotted on nitrocellulose membranes. For removing unbound pepstatin A probes, membranes were washed with 20 mM sodium acetate buffer, pH 5.0, (circle) or with 20 mM sodium phosphate buffer, pH 7.4, (square). Each point represents the mean of triplicate assays and bars indicate standard errors of the mean. ) Bound (X10 cpm 0 0 5 10 Concentration of labeled pepstatin A (X10 cpm) Fic. 2. Binding of [*°S]-methionyl-pepstatin A to the fixed amount of pepsin. Pepsin (5.7 nmoles) was immobilized on a nitro- cellulose membrane (1.5 X 1.5 cm?) and incubated with indicated amounts of labeled pepstatin A (4.4 x 10°-8.8 x 10* cpm/assay). Radioactivity bound to membranes was counted as amounts of labeled pepstatin A bound to pepsin. The radioactivity of 10° cpm corresponds to 35 pmoles of the pepstatin A probe. Each point represents the mean of triplicate assays. Bars indicate standard errors of the mean. [°°S]-methionyl-pepstatin A to pepsin is proportional to pep- sin concentrations in the range of 1-12nmoles. Radio- labeled pepstatin A bound increasingly to the fixed amount of pepsin as the amount of labeled pepstatin increased up to around the dose of 2<10*cpm and then the binding was leveled off (Fig. 2), suggesting that the labeled pepstatin A binds to limited and specific sites of pepsin. Binding of [*°S]-methionyl-pepstatin A to pepsin was competitively suppressed by pepstatin A (Fig. 3), indicating that the radio-labeled derivative shows the same binding specificity as pepstatin A. 2.0 Bound (X10°cpm) 5 0 10 20 30 40 50 Concentration of pepstatin A (1M) Fic. 3. Suppression of the binding of the radio-labeled pepstatin A to pepsin by pepstatin A. Pepsin (5.7 nmoles/assay) spotted on membranes was incubated with the fixed amount of the pepstatin derivatives (8.8 x 10° cpm/assay) and varied amounts of pepsta- tin A (0-48 uM) as a competitor. The radioactivity bound to pepsin was counted. Each point represents the mean of tripli- cate assays with its standard error indicated by a bar. Figure 1 shows that the [*°S]-methionyl-pepstatin A pre- pared as above can detect pepsin when its amount is more than 1.0nmoles. To get the probe for aspartic proteinases showing a higher specific radioactivity, the labeled pepstatin A was further purified by subjecting it to reversed phase HPLC (Fig. 4). Pepstatin A was eluted at 38% acetonitrile as a single peak. The radio-labeled pepstatin A was eluted at 44% acetonitrile also as a single peak. The pepstatin derivative thus purified had a specific activity of 3.6 MBq/ pmole and could detect as low as 0.1 nmoles of pepsin. Its binding to pepsin was proportional to the concentration of the enzyme in the range of 0.1—0.3 nmoles (Fig. 5). The radio-labeled pepstatin A was tried to use as a probe to detect aspartic proteinases in the animal tissue. The metamorphosing tadpole tail was quantitated for the pepsta- tin-reactive enzymes (Fig. 6). The specific binding of the probe to the tail of a metamorphosing tadpole was 18-fold higher than the binding to the tail of a premetamorphic animal. If we postulate that all the bound probe recovered 224 Fic. Bound (X10°cpm) M. Mukxal, T. KoNDO AND K. YOSHIZATO ( 220 nm Absorbance at 0 10 20 Retention time (min) - - f - Be - Pied ae labeled pepstatin A i walt oo ES > 2 = = @ 2 2 ® 3 ° Po og 40 50 Fic. 4. Separation of [°°S]-methionyl-pepstatin A and pepstatin A by reversed phase HPLC. The radio-labeled pepstatin A fraction obtained at the second reaction was applied to a Inertsil 300-C, (4.6 x 100 mm) and eluted at 50°C with a linear gradient of 25-55% of acetonitrile (dotted line) containing 0.1% TFA. The flow rate was 0.4 ml/min. The solid line indicates the absorbance at 220nm. The radioactivity of each fraction (0.4 ml) was counted (circle). The first peak of 220 nm contains dimethyl! sulfoxide which was used as the solvent of pepstatin A. The closed arrow at 38% acetonitrile shows the peak of pepstatin A and the open arrow at 44% acetonitrile the peak of [>°S]-methionyl-pepstatin. The shaded fractions were collected as [*°S]-methionyl pepstatin and used as a probe for aspartic proteinases (3.6 MBq/ mole). 25 2.0 1.0 0 Caan 0.3 0.6 Pepsin (nmoles/assay) 5. Binding of the [°°S]-methionyl-pepstatin A to pepsin. HPLC-purified radio-labeled pepstatin A (10* cpm/assay) was incubated for 30 min at 0-4°C with nitrocellulose membranes containing varied amounts of pepsin. Membranes were washed with acidic buffer and counted for radioactivity (circle). Non- specific binding to nitrocellulose membranes was obtained as the radioactivity remaining on membranes when they were washed with neutral pH (square). The radioactivity of 10° cpm is equivalent to 7.1 pmoles of the pepstatin A probe. Each point represents the mean of triplicate assays and bars indicate stan- dard errors of the mean. 4 ~_~ £3 @ 3° oot Cc > oD £eE, ao (Onre os 3 & rood De x. Xvi XXill TK stage of tadpole Fic. 6. Detection of aspartic proteinase in the tail of tadpole by the [>°S]-methionyl-pepstatin A. The specific binding of the probe was calculated by subtracting the value of radioactivity obtained in the presence of excess amounts of pepstatin A (30 «M) from that without it. Tail tissues of bullfrog tadpole (about 100 mg wet weight/assay) at premetamorphosis (TK stage XVII) or the climax of metamorphosis (TK stage XXIII) were incubated with the radio-labeled pepstatin A (HPLC-purified, 2.0 10* cpm/ assay) in solution of 20mM sodium acetate buffer and 0.1M NaCl, pH 5.0, with or without a competitor, pepstatin A (30 uM). The radioactivity of 10° cpm is equivalent to 7.1 pmoles of the pepstatin probe as in Fig. 5. Each value represents the mean of triplicate assays with its standard error indicated by a bar. Radio-labeled Pepstatin A 225 from tissues, the tail contained binding sites of 0.7 nmoles/10 mg protein at the climax stage and 0.04 nmoles/10 mg protein at the premetamorphic stage. These values of binding sites are expressed as those for “pepsin equivalent”. DISCUSSION Pepstatin A is a proteinase inhibitor that is not specific to one species of enzyme but shows the activity toward several species of enzymes grouped as the aspartic proteinase family including pepsin, cathepsin D, cathepsin E, renin, chymosin and gastricsin. Therefore, [*°S-]-methionyl-pepstatin A de- veloped in the present study can be applied to animal tissue for the first screening of proteinases which is grouped into this family. The radio-labeled pepstatin A keeps the same activity as pepstatin A. We used the carboxyl group of pepstatin A as the site of modification, as pepstatin-conjugated resins have been usually prepared by modifying this residue and utilized for purification of aspartic proteinases such as cathepsin D [1]. The present study also confirms that the carboxyl group of pepstatin can be modified as a connecting site for reporter groups without destroying properties which are required for the inhibitor to bind aspartic proteinases. Radioactive methionine can be successfully introduced by this modifica- tion into pepstatin A. Since this additional chain of methionine is neutral in charge, it does not disturb the ionic condition of the binding site of aspartic proteinases. The [*°S]-methionyl-pepstatin A we prepared is more hydrophobic than pepstatin A, which may help the derivative to approach more easily for the binding site of the enzymes than pepstatin A because the site is thought as a hydrophobic pocket in pepsin [27]. Pepstatin A binds to aspartic proteinases in a stoichiometric manner. The present study shows that the binding of the radio-labeled pepstatin A to pepsin is pro- portional to concentrations of pepsin, indicating that it can be used as a sensitive probe for the quantitative analysis of aspartic proteinases. Utilizing this probe, we could detect 0.1 nmoles of pepsin. It seems that other methods such as radio-immuno assay are more sensitive than the method described in the present study. However, the radio-labeled pepstatin A is much useful as compared to the antibody against aspartic proteinases. We have to prepare the specific antibody in each case of the study against the enzyme of target. In contrast, the radio-labeled pepstatin A can be used for detecting the enzyme belonging to the family of aspartic proteinases. We made some alterations in the assay of binding of the pepstatin A derivative when we tried to detect aspartic proteinases in tadpole tissues. Binding was assayed on nitrocellulose membranes for the test enzyme (pepsin), while the assay for the tissue sample was done in tissue pieces suspended in acid solution. The pepstatin A derivative bound to enzymes in the tissue was then removed by centri- fugation in alkaline condition. The reason for this alteration was the relatively low capacity of nitrocellulose membranes to hold tissue proteins. We first tried to assay enzymes in the tissue samples using nitrocellulose membranes as we did for the test enzyme. However, reliable and reproducible values of binding of the radio-labeled pepstatin A could not be obtained in the membrane assay. This probe is shown to detect pepstatin-sensitive pro- teinase(s) in the tail of bullfrog tadpole. The amount of specific binding to the tail tissue much increases at the climax stage of metamorphosis (TK stage XXIII) as compared to the premetamorphic one. Several reports demonstrated that cathepsin D like proteinase activity in the tail increases at the climax of metamorphosis [9, 18, 29, 38, 39, 43] and pepstatin sensitive proteinase plays important roles in the regression of tail during metamorphosis [31]. Pepstatin A binding sites at the climax stage of metamorphosis increases 18-fold as com- pared to those at the premetamorphic stage. The ratio of cathepsin D activity of the tail at the climax stage to that at the premetamorphic stage is 18 in Rana catesbeiana [29] and 16 in Xenopus laevis [39]. These values are very close to that obtained by our method. Pepstatin-sensitive proteinases are found also in a wide variety of life including vertebrates such as human [5], bovine [30], porcine [14], rabbit [6], rat [10], chicken [5, 26], frog [23] and fish [7], and invertebrates such as marine mussel [24] and hemipteran insect [13]. Plant [28], bacteria [17] and retro- virus [40] have also been reported to contain pepstatin- sensitive proteinases. Therefore, it is considered that aspar- tic proteinases might play fundamental roles in metabolic processes of varieties of life. However, the information on the enzyme other than mammalian origin has been poor. There is a possibility that the binding site recognized by pepstatin A is highly conserved in the molecular evolution of aspartic proteinases. The pepstatin derivative reported here is expected to be useful in detecting unknown aspartic proteinases in the wide range of life and studying them from a comparative point of view. It has been shown that aspartic proteinase of HIV virus can be converted by the gene technology into a mutant enzyme that has no enzymatic activity but can bind to pepstatin A [40]. This indicates that the catalytic site of the enzyme is different from the pepstatin binding site, although both sites are in the active center. The radio-active pepsta- tin developed in the present study is expected to be useful in detecting enzymes of the aspartic proteinase family that lose their catalytic activity. We could not develop pepstatin A derivative which shows higher sensivity of detection than the method to directly measure the enzyme activity or conventional im- munological detection method using specific antibody. It remains as a future study to develop a method to prepare [°°S]-methionyl-pepstatin A with much more higher specific activity. 226 M. Mukal, T. KoNDO AND K. YOSHIZATO ACKNOWLEDGMENTS The authors would like to express their thanks to Drs. S. Yasugi, T. Ohoka and S. Tomino for their kind discussions on this study. 10 11 12 REFERENCES Afting EG, Becker ML _ (1981) Two-step affinity- chromatographic purification of cathepsin D from pig myomet- rium with high yield. Biochem J 197: 519-522 Anjaneyulu PSR, Staros JV (1987) Reactions of N- hydroxysulfosuccinimido active esters. Int J Pep Pro Res 30: 117-124 Aoyagi T, Morishima H, Nishikawa R, Kunimoto S, Takeuchi T, Umezawa H, (1972) Biological activity of pepstatins, pepstanone A and partial peptide on pepsin, cathepsin D and renin. J Antibiot 25: 689-694 Azuma T, Liu W, Laan DJV, Bowcock AM, Taggart RT (1992) Human gastric cathepsin E gene. J Biol Chem 267: 1609-1614 Barrett AJ (1970) Cathepsin D, purification of isoenzymes from human and chicken liver. Biochem J 117: 601-607 Barrett AJ (1971) Purification and properties of cathepsin D from liver of chicken, rabbit and man. In “Tissue proteinases” Ed by Barrett AJ, Dingle JT North-Holland Publishing Co., Amsterdam, pp 109-133 Bonete MJ, Manjon A, Llorca F, Iborra JL (1984) Acid proteinase activity in fish II. Purification and characterization of cathepsins B and D from Mujil auratus muscle. Comp Biochem Physiol 78B: 207-213 Faust PL, Kornfeld S, Chirgwin JM (1985) Cloning and sequ- ence analysis of cDNA for human cathepsin D. Proc Natl Acad Sci USA 82: 4910-4914 Fujii Y, Taguchi M, Kobayashi K, Horiuchi S (1991) Im- munochemical studies on cathepsin D like enzyme in the tadpole tail of Rana catesbeiana. Zool Sci 8: 511-520 Fujita H, Tanaka Y, Noguchi Y, Kona A, Himeno M, Kato K (1991) Isolation and sequencing of a cDNA clone encoding rat liver lysosomal cathepsin D and the structure of three forms of mature enzymes. Biochem Biophys Res Commun 179: 190- 196 Gross J, Lapiere CM (1962) Collagenolytic activity in amphib- ian tissues : A tissue culture assay. Proc Natl Acad Sci USA 48: 1014-1022 Hobart PM, Fogliano M, O’Connor BA, Schaefer IM, Chirgwin JM (1984) Human renin gene : Structure and sequence analy- sis. Proc Natl Acad Sci USA 81: 5026-5030 Houseman JG, Downe AER (1983) Cathepsin D-like activity in the posterior midgut of hemipteran insect. Comp Biochem Physiol 75B: 509-512 Huang JS, Huang SS, Tang J (1979) Cathepsin D isozymes from porcine spleens. J Biol Chem 254: 11405-11417 Knight CG, Barrett AJ (1976) Interaction of human cathepsin D with the inhibitor pepstatin. Biochem J 155: 117-125 Knight CG, Hornebeck W, Matthews ITW, Hembry RM, Dingle JT (1980) Interaction of dinitrophenyl-pepstatins with human cathepsin D and with anti-dinitrophenyl antibody. Biochem J 191: 835-843 Kobayashi H, Kusakabe I, Murakami K (1982) Rapid isolation of microbiral milk-clotting enzymes by N-acetyl(or N- isobutyryl)-pepstatin-aminohexyl agarose. Anal Biochem 122: 308-312 Kobayashi K, Hara M, Horiuchi S (1985) Isolation of lyso- somes from the tail of metamorphosing bullfrog tadpole. Comp Biochem Physiol 81B: 603-607 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35) 36 37 38 39 40 41 Kunimoto S, Aoyagi T, Nisizawa R, Komai T, Takeuchi T, Umezawa H (1974) Mechanism of inhibition of pepsin by pepstatin II. J Antibiot 27: 413-418 Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the folin phenol reagent. J Biol Chem 193: 265-275 Matthews ITW, Decker RS, Knight C G (1981) Bimane- labelled pepstatin, a fluorescent probe for the subcellular loca- tion of cathepsin D. Biochem J 199: 611-617 Matthews ITW, Decker RS, Hornebeck W, Knight CG (1983) Dinitrophenyl-pepstatins as active site directed localization re- agents for cathepsin D. Biochem J 211: 139-147 Nanbu M, Kobayashi K, Horiuchi S (1988) Purification and characterization of cathepsin D-like proteinase from the tadpole tail of bullfrog, Rana catesbeiana. Comp Biochem Physiol 89B: 569-575 Okada S, Aikawa T (1986) Cathepsin D-like acid proteinase in the mantle of the marine mussel, Mytilus edulis. Comp Biochem Physiol 84B: 333-341 Oofusa K, Yoshizato K (1991) Biochemical and immunological characterization of collagenase in tissues of metamorphosing bullfrog tadpoles. Develop Growth and Differ 33: 329-339 Retzek H, Steyrer E, Sanders EJ, Nimpf J, Schneider WJ (1992) Molecular cloning and functional characterization of chicken cathepsin D, a key enzyme for yolk formation. DNA Cell Biol 11: 661-672 Rich DH (1986) Inhibitors of aspartic proteinase. In “Pro- tease inhibitors’ Ed by Barret AJ, Salvesen G, Elsevier Science Publishers BV., Amsterdam, pp 179=217 Runeberg-Roos P, Térmakangas K, Ostman A (1991) Primary structure of a barley-grain aspartic proteinase. Eur J Biochem 202: 1021-1027 Sakai J, Horiuchi S (1979) Characterization of cathepsin D in the regressing tadpole tail of bullfrog, Rana catesbeiana. Comp Biochem Physiol 62B: 269-273 Sapolsky AI, Woessner FJ (1972) Multiple forms of cathepsin D from bovine uterus. J Biol Chem 247: 2069-2076 Seshimo H, Ryuzaki M, Yoshizato K (1977) Specific inhibition of triiodothyonine-induced tadpole tail-fin regression by cathep- sin D-inhibitor pepstatin. Develop Biol 59: 96-100 Shewale JG, Tang J (1984) Amino acid sequence of porcine spleen cathepsin D. Proc Natl Acad Sci USA 81: 3703-3707 Sogawa K, Fujii-Kuriyama Y, Mizukami Y, Ichihara Y, Taka- hashi K (1983) Primary structure of human pepsinogen gene. J Biol Chem 258: 5306-5311 Staros JV, Wright RW, Swingle DM (1986) Enhancement by N-hydroxysulfosuccinimido of water soluble carbodiimide- mediated coupling reactions. Anal Biochem 156: 220-222 Takahashi T, Tang J (1983) Amino acid sequence of porcine spleen cathepsin D light chain. J Biol Chem 258: 6435-6443 Taylor AC, Kollros JJ (1946) Stages in the normal develop- ment of Rana pipiens larvae. Anal Rec 94: 7-23 Umezawa H, Aoyagi T, Morishima H, Matsuzaki M, Hamada M, Takeuchi T (1970) Pepstatin, a new pepsin inhibitor pro- duced by actinomycetes. J Antibiot 23: 259-262 Weber R (1967) Biochemical and cellular aspects of tissue involution in development. Exp Biol Med 1: 63-76 Weber R (1977) Biochemical characteristics of tail atrophy during anuran metamorphosis. Colloq Int C N R S 226: 137- 146 Wondrak EM, Louis JM, Mora PT, Oroszlan S$ (1991) Purifica- tion of HIV-1 wild type proteinase and characterization of proteolytically inactive HIV-1 proteinase mutants by pepstatin A affinity Chromatography. FEBS Lett 280: 347-350 Workman RJ, Burkitt DW (1979) Pepsin inhibition by a high 42 Radio-labeled Pepstatin A 227 specific activity radioiodinated derivative of pepstatin. Arc Biochem Biophys 194: 157-164 Yamada H, Imoto T, Fujita K, Motomura M (1981) Selective modification of aspartic acid-101 in lysozyme by carbodiimide reaction. Biochemistry 20: 4836-4842 43 Yoshizato K (1989) Biochemistry and cell biology of amphib- ian metamorphosis with a special emphasis on the mechanism of removal of larval organs. In “International Review of Cytolo- gy” Academic Press, New York, 119: 97-149 Jiriquae te ygoloid iiss bee viieeaioaktt Maer: x obi: i merle ait) ag sade al ret iden a lal fon We od inedteearel” ap auagro' terial \“ vam . oth | Po Tale ey, 2079 hv blaien As, ‘ppom , ee ; 1 II _Peiiekin seat apr Cissn 190 36-2 oe ‘a Ta we ay Nlatshnew? ‘Tw, ‘Decker, » oo , Hebelled! pepstatin é. pana — ih aby, ; Ue. Tol Manan bos \ eek ee ee Maru és TTIW,’ bales ge ost ae Pwevuptel: eeane cuted we ; a Ba ai = WONT Sone Ye Bee he hit § Tee Bye Men We Lie Nitin, ? : Kobayaite” ted ctv ene iienttiss of calbnapeain LIN oF baal Hey Raine cteaaley u : } _ ab wr * } A wy, ? re ie divas 7 ed 7 2 Ty ta SAikewen ¥ (Gag: Wo fey Waar. yt Cie layuarnnee ro p Team Peace May intl au 1 _ cia ts erie / a + d been v4 . ; hailey dar ch thepwlin O, a ney ehatynne top ypaiie iiam ; 1) al oe 7 : j a Be i}? Pay ¢ mi ahh t jtona’” eT My Vibes Wiper Hae * it eres: . +H Pune Mi onpatg: § * Saku Ff Homuct) Yow es Caen 2 prose & : ‘ ta GLE eT Webi, Mis » PA ey | a) rh 4 eyed 418 * : I es! he nh i Rous Mi, Voelker ll 1, Yael if stad f it ichyinlé A a 5 = ee ri titer pean! Ddegiop ‘tes ; t t). Tone) ae) Rin ; fs my i ; Ww 7) 44 a ri - Bee . = ‘= . . > } Ay Ae Ones , sak + VAS I 4) joi Mum iW 2am A c Wee Sry 1 gers War yi 7. i Pahl ind uy at ‘wdcdcvrintee) so yonamllgeallita Hi mh wvnl@ “aN ¥ wiry a ee oh ty sm ; . te ee 5 Sly alee i¥q t an we erieoliety: ut iitiec lay i ‘ fete. queen, 4 : i ar Pp i ewtdde "Wy 2a ZOOLOGICAL SCIENCE 11: 253-260 (1994) Induction of Metamorphosis in the Sea Urchin, Pseudocentrotus depressus, using L-Glutamine Ikuko YAZAKI and Hiroki HARASHIMA Department of Biology, Faculty of Science, Tokyo Metropolitan University Minamiohsawa 1-1, Hachiohji Tokyo 192-03, Japan ABSTRACT—Larvae of the sea urchin, Pseudocentrotus depressus, were induced to metamorphose using 10-°-10-* M L-glutamine. After being subjected to 610° ° M glutamine, the larvae ceased swimming, began to retract their arms, extruded primary podia through an opening of the larva at a vestibule of echinus rudiment (ER) and then everted the ER, requiring 1-9 hr, 7-11 hr, 7-38 hr and 12-48 hr (from the time of the recognition of the first larva to the time of the maximum larvae changes), respectively. The effective times needed in the glutamine-treatment were 2 hr for induction of the cessation of larval swimming and the start of arm-retraction, and 4—8 hr to induce the ER-eversion. There was no correlation between the extruding of primary podia and the amount of time needed for the glutamine-treatment. From the difference seen in the effective time of glutamine-treatment, the early changes in metamorphosis; cessation of swimming and start of arm-retraction, are not autonomously linked to the later change; ER-eversion. L-glutamic acid and y-aminobutyric acid (GABA) did not induce metamorphosis. 10 ’M GABA provoked a rapid cramp in the arms and a spreading of spines, and the larva seemed to undergo ER-eversion, but within a few hours they returned to their original larval shape. Thus eversion-like changes induced by GABA may not reach a true ER-eversion due to an absence of process or processes which were induced by the 4—8 hr treatment of glutamine over a long period of time (12-48 hr). © 1994 Zoological Society of Japan INTRODUCTION Larvae of marine invertebrates undergo metamorphosis in response to environmental cues [6]. Sands and sediments in the adult habitat cause metamorphosis in larvae of the sand dollar [8] and Nassarium [18]. Bacteria and algae substrata are also effective for inducing metamorphosis. In aplysiids and spirorbis (serpulidae), the larvae of each aplysiid or spirorbis species selects a substratum containing their favorite algae in which to settle [7, 20]. In echinoderm, bacterial and algal films, which are prepared by soaking the glass or plastic plates in natural sea water for several days, are effective in inducing metamorphosis of sea star larvae [12, 19] and sea urchin larvae [9, 15, 23]. However, the active factor of these natural cues has been rarely identified. For example, free fatty acid components were extracted from red algae as a chemical inducer of larval settlement and metamorphosis of the sea urchin, Pseudocentrotus depressus and Anthocidaris crassispina [11], and a pheromonal substance, 980 Da- peptide, was isolated as an active factor in the metamor- phosis in the sands in the adult habitat of Dendraster excentri- cus [3]. y-Aminobutyric acid (GABA), which is known as a neurotransmitter in many species, induced abalone larvae to settle and to undergo metamorphosis [13]. In sea urchins, Burke [2] proposed a neural control of metamorphosis according to his electrical and obsolatory experiments that the larvae of D. excentricus were induced to metamorphose by electrical stimulation of their two nerve centers, apical neuropile and oral ganglions, although he noted that GABA Accepted February 15, 1994 Received October 29, 1994 was ineffective [2]. In Strongylocentrotus intermedius and S.milabiris, L-glutamine, which is known as a precursor of glutamic acid and GABA in the central nervous system of mammals, induced metamorphosis in 100% and 50% of the larvae, respectively [16]. In the present study, the effects of L-glutamine on the metamorphosis of P. depressus larvae were examined, and the mechanisms of the glutamine-inducing metamorphosis are discussed. MATERIALS AND METHODS Larvae which derived from one pair of a male and a female of Pseudocentrotus depressus were reared by the method of Noguchi [17] with slight modifications. The cultures in five 3-1 beakers were maintained at 19° +2°C which were stirred with a paddle attached to a 60 rpm motor. At the onset of feeding (3 days after fertilization), there were 25,000 plutei in 2.5 litres of paper-filtered sea water. Two or three ml of Chaetoceros gracilis (3—5 x 10°/ml) were added to each culture every 1 or 2 days [10], and two-thirds of the culture water was changed twice aweek. The larval density was initially 10/ ml but was diluted to 1/ml at the stage of an 8-armed pluteus. Various developmental stages of larvae were found in culture beak- ers. Competent larvae to metamorphosis were obtained over a period of one month, beginning from the 40th day after insemination. Larvae with fully developed ER were selected and transferred into 0.22 4m Millipore-filtered seawater (MFSW) in a 100-ml beaker. To remove microorganisms taken into the larvae as food, the larvae were then kept for 1 day in MFSW. After rinsing twice in MFSW, every 5-10 larvae were randomly apportioned into each well of 12-well plastic plates, which were filled with either 2.5 ml MFSW (control) or with the same volume of various concentrations of L-glutamine- or other substances dissolved in MFSW (adjusted with NaOH at pH 7.8-8.0). Larvae in the wells were retransferred to a well filled with the same medium to avoid diluting the experimental 254 I. YAZAKI AND H. HarAsHIMA medium. Usually, L-glutamine treatment did not exceed 24 hr. If the observations were prolonged, the culture medium (MFSW) was changed every 2 days. RESULTS Effective concentrations of L-glutamine needed to induce metamorphosis Metamorphosis of the echinopluteus occurs after settle- ment, and includes an eversion of the adult rudiment (echinus rudiment: ER) which develop within the larva, resorption of the larval body into the juvenile, and the subsequent differ- entiation of the adult form [5]. Larvae used for the present experiments were swimming and had the fully grown adult rudiment whose primary podia were moving within the larva but not extruding out the larval body (Fig. 1A). When the spines and the primary podia of ER were exposed outside the larva by eversion of the ER, the larva was considered to be metamorphosed (Fig. 1B). Larval arms began to retract prior to the eversion of the ER, but had not completely retracted by the time of ER eversion. Parts of the larval arms and epaulets were retained on the newly metamorph- osed juvenile (Fig. 1B), and remained for an additional 1 or 2 days after eversion of the ER. Larvae were treated with various concentrations of L- glutamine solutions from 3x10~* to 6x10-°M for 24 hr, washed with MFSW and then maintained in MFSW (Fig. 2). Out of ten larvae in each concentration, larvae rarely meta- morphosed after just 24 hr (the end of glutamine-treatment). Most larvae which were treated with glutamine at concentra- tions of 3X 10~* to 3x 10-° M were found to metamorphose after 48 hr. At a concentration of 6x 10° M L-glutamine, no metamorphosis had occurred by 48 hr, and only one larva had metamorphosed at 72 hr (Fig. 2). From these results, 6 <10~° M L-glutamine was used as a conventional concentra- A pe——™ 10 Number of metamorphosed larvae 24 48 72 Time (hr) Fic. 2. Effects of L-glutamine at various concentrations. Groups of ten larvae were treated with L-glutamine in MFSW for 24 hr with concentrations of either 6X10~° (@), 3x10~° (4), 6x 10~° (x), 1.2x10~* (G) or 3X10~*M (0). At the beginning of the experiments, all larvae had not metamorphosed. 1.2X and 3X10~*M glutamine induced metamorphosis in all ten larvae after 48hr. 3X and 6X10-°M glutamine induced metamorphosis in 8 and 9 larvae, but 6X10 * M glutamine did not induce any metamorphosis after 48 hr, with only one larva undergoing metamorphosis after 72 hr. tion for induction of metamorphosis of P. depressus with the resulting metamorphosis usually observed after 48 hr in the following experiments. Changes in larvae induced using L-glutamine treatment To analyse the process of metamorphosis, the behaviour and form of larvae were repeatedly checked from the begin- ning of the glutamine treatment (Fig. 3). Fifty-one larvae were transferred into five wells filled with 6X 107° M gluta- Fic. 1. A competent larva and a juvenile of P. depressus which was induced to metamorphose using L-glutamine. A: An 8-armed pluteus larva competent to metamorphose. The larva is swimming around. Neither the spines nor the tube feet (the primary podia) of the echinus rudiment were extruded. B: A juvenile which was metamorphosed after treatment with 6X 10~° M L-glutamine for 24 hr, and photographed from beneath with an inverted microscope. la, larval arms; ae, anterior epaulettes; pe, posterior epaulettes; ap, apical pedicellaria; as, adult spine; Is, larval spine; p, primary podia; ra, retracting arm (out of focus); ER, echinus rudiment; M, mouth. Bar=100 «m. % of the changed larvae Metamorphosis of Sea Urchin { 3 6 11 16 21 38 Time (hr) Fic. 3. Time course of larval changes induced by treatment with L-glutamine. Fifty-one competent larvae were kept in 6X Number of swimming stopped larvae Fic. 10-°M L-glutamine containing MFSW for 21 hr and then tranferred into MFSW. Observations were carried out repeatedly on four changes which occurred in each larvae; 1) cessation of swimming (0), 2) retraction of larval arms (4), 3) extrusion of primary podia through an opening of larva at the vestibule of ER ((), and 4) eversion of ER (™). The echinus rudiments began to evert since each larva ceased swimming and the retraction of arms had occurred. The primary podia seem to be extruded several hours before the ER-eversion. . 2,4,6 10 6 >| noo 10 Number of arm-retracted larvae Number of ER-everted larvae Number of larvae extruding primary podia 4 0 44 4-4 4 yeti. ithe - ae 2 24 2 24 | 1 | ° to} OE 0 -———————_———4 0 ae 0) J 2 4 6 8 9 Qi 2 4 6 f} 24 48 72 24 48 72 Time (hr) Time (hr) Time (hr) Time (hr) 4. Larval changes induced by various durations of treatment with L-glutamine. Six 10-larvae groups (10 larvae/well) were treated with 10° M L-glutamine in MFSW and then transferred into glutamine-free MFSW after 1 hr (@), 2 hr (A), 4 hr (GQ), 6hr (©), 8hr (4), and9 hr(™). Control larvae (x) were reared in glutamine-free MFSW. The larval changes of each group were followed for up to 72 hr and are shown in a) cessation of swimming, b) retraction of arms, c) extrusion of primary podia and d) eversion of the echinus rudiment (ER). a) and b) are shown from zero to nine hours, and c) and d) are from 24 to 72 hr. Numerals to the right of figures indicate the elapsed time of treatment with glutamine. Cessation of larval swimming (a) and the start of arm retraction (b) were induced by 2-hr treatment, although arm retraction progressed with longer durations of treatment. Extrusion of the primary podia (c) seemed to be accelerated by treatment with L-glutamine relative to the controls, but does not have any correlation with the treatment time. To evert the ER (d), larvae must be treated with glutamine for 4-8 hr. 256 I. YAZAKI AND H. HARASHIMA mine-MFSW and kept there for 21 hr. All of the larvae examined were swimming around after the first 30 min. After one hour, 20% of the larvae had stopped swimming and almost all larvae became immobile between 9-10 hr after immersion in glutamine-MFSW. The number of larvae which retracted larval arms and extruded the primary podia began to increase after 7hr. The arm-retractions were observed in all larvae after 11 hr, while the larvae extruding the primary podia slowly increased. versions of ER began after 12 hr. By the end of the treatment (21 hr), ER everted in 50% of larvae examined, and after 38 hr (17 hr after the 21 hr-glutamine treatment), the ER had everted in 88% of the larvae. Times required for induction of metamorphsis using L- glutamine Seven 10-larvae groups (10 larvae/well) were prepared. Six groups were placed in wells with 610° M glutamine- MFSW while one group was maintained in MFSW alone as a control. Each group of larvae in glutamine-MFSW was transferred into MFSW at an interval of 1 or 2 hr from the beginning up to 9 hours later (Fig. 4). The movement of larval swimming was affected by very short exposure to glutamine. After a 6-hr treatment, all larvae had ceased swimming (Fig. 4a). Only a 2-hr glutamine treatment made all the larvae completely stop swimming after 6hr. The effect of a 1-hr treatment was a little unstable (Fig. 4a). Arms had also begun to retract in all larvae examined after a 6-hr treatment. The start of arm-retraction was much more delayed when the duration of glutamine treatment was shorter, and the progression of arm-retraction was prop- ortional to the duration of glutamine treatment (Fig. 4b). One-hour treatment could induce the start of arm-retraction in all larvae after 24hr (data not shown), and the arms remained a little shortened. As the extrusion of the primary podia and the ER eversion were not recognized up to 9hours, the results achieved after 24hr are given in Figs.4c and 4d. The extrusion of podia was also accelerated by glutamine treat- ment. At 24 hr, in each of the experimental groups (even in the larval group given only a 1-hr treatment), more podia were extruded than in the control group (Fig. 4c). Howev- er, there seemed to be no close relationship between the number of larvae whose primary podia had extruded and the duration of glutamine treatment (Fig. 4c). Eversion of ER at a greater rate than controls could not be induced by either a 1- or 2-hr glutamine treatment. At 24 hr, out of 10 larvae which were treated with glutamine for 9 hr, seven had everted ERs, but in the larvae treated for 8 hr or less, the number of ER everted remained low. At 48 hr, larvae treated for more than 4hr induced ER-eversion in over 60%, and at 72 hr, 100% ER eversion was recorded in larvae given 8 or 9 hr of treatment (Fig. 4d). Effects of L-glutamic acid and GABA on metamorphosis L-glutamic acid, which was deaminated from L- glutamine, was examined at the same concentration (6x 10> M) as L-glutamine for 18-22 hr (Fig. 5). In experiments involving four trials, 69 larvae were treated with glutamic acid, 108 with glutamine and 78 with neither (controls). Observations were carried out for 7 days, although on the 6th and 7th days about half the larvae examined were observed as follows. Among the glutamic acid-treated larvae, only one larva was found to have metamorphosed by the 4th day, after which no more larvae underwent metamorphosis. By con- trast, most of the 108 larvae treated with glutamine had metamorphosed by the 2nd day (Fig.5). Seven larvae, which had metamorphosed by the 4th day, had retracted almost all of their arms to look like black balls and all died. Glutamic acid induced the arm-retractions in 80% of larvae within 2 days (Fig. 6b), but their arms were shortened only slightly. Glutamic acid did not interfere with the larval movement of swimming (Fig. 6a) and the extruding of the primary podia (Fig. 6c). 100 50 % of metamorphosis Days Fic. 5. Effect of L-glutamic acid on metamorphosis compared with that of L-glutamine. Four experiments with different treatment times (18 hr, 20 hr and two experiments with 22-hr treatment) are combined. Observations were continued for 4 days (18-, 20- and 22-hr treatments) and for 7 days (22-hr treatment). The number of experiments totalled 67 for glutamic acid (A), 108 for glutamine (OQ) and 78 larvae for MFSW (controls) (x). In contrast to the high percentage of metamorphosis-induction using glutamine, glutamic acid induce almost no metamorphosis (there was only one metamorphosed larva among those treated with glutamic acid after 4 days, after which no further meta- morphosis occurred). GABA (j-aminobutyric acid), a substance which is formed by the decarboxylation of glutamic acid, was applied on each group of 10 competent larvae in concentrations of 5X and 1x10~°M for 24hr (Exp. 1 in Table 1) or for 8.5 hr (Exp. 2), and at lower concentrations of 1x10-°M or X 10-7 M for 18 hr (Exp. 3) compared with glutamine. When larvae were treated with higher concentrations of GABA for Metamorphosis of Sea Urchin a b 100 100 80 80 60 60 40 40 20 20 % of swiming stopped larvae % of arm- retracted larvae 0 10 20 30 36 0 10 Time (hr) Zi 80 60 40 20 % of larvae extruding primary podia 20 30 36 0 10 20 30 36 Time (hr) Time (br) Fic. 6. Changes in larvae treated with L-glutamic acid. Eighteen larvae were treated with 6x10 ° M L-glutamic acid (4) for 18 hr. for the same duration. To confirm the larval competency to metamorphose, the same concentration of L-glutamine (©) was applied 16 control larvae (x) were kept in MFSW. L-glutamic acid did not effect cessation of swimming and the extrusion of primary podia, although a slight retraction of arms was observed in 80% of the larvae after 36 hr. TABLE 1. Every 10-larvae Effect of various concentration of GABA and various durations of treatment with GABA Number of metamorphosed larvae treated Concentrations with (M) at 24 hr 48 hr 72 hr 1) 24-hr treatment None = 0 0 — L-glutamine 3x10~* 7 8 _ 6x10-° 1 0 — GABA SiO 5 4* — 1 <1O- 6 5° — 2) 8.5-hr treatment None = 0 0 1 L-glutamine 3x10~* 6 10 10 6x10-° 3 8 9 GABA 510° 0 0 2 IS<1o-? 1 1 2 3) 18-hr treatment None = 0 0 0 L-glutamine 6x 10~° 1 6 7 GABA Ixilo-* 1 3 0 AS-moist air, without any supplementation of test agents or serum. Colony- forming effieciency (number of resultant colonies per incubated cell number x 100) was calculated after 20 days of culture. The number of F-type colonies [20] was counted; only those F-type colonies larger than 504m in diameter were scored. In C-type colonies [20], colonies containing more than 6 cells were counted. 3. Assays for cell differentiation in soft-agar culture: Alcian blue staining and immunohistochemical staining The culture dishes were dried at 55°C for 60 min and fixed with 10% formalin for 3hr. Cells were stained with 0.1% Alcian blue dissolved in 0.1 N HCl for 3 hr at room temperature. Anti-PG-H rabbit serum (provided by Dr. Koji Kimata, Aichi Medical University) was used for detecting cartilage-specific proteo- glycan, PG-H [15]. Indirect immunohistochemical staining with biotinylated secondary antibody and /-galactosidase-bound strept- avidin was performed as described previously [20]. 4. Preparation of conditioned medium from protein-free monolayer culture of the scleral fibroblasts and ELISA of the conditioned medium material for anti-TGF-2 immunoreactivity Methods for preparing protein-free monolayer culture and har- vesting conditioned medium were given elsewhere [19]. The har- vested conditioned medium was concentrated from 50 ml to 4 ml by centrifugal concentrator (VC-360, Taitec Co., Japan). The concen- trate and precipitate were mixed, transferred into membrane tubing (Dialysis Membrane, Size 20; cut off 10kDa, Waco Ltd.) and dialysed against distilled water for 5 days. After centrifugation at 12,000 rpm for 20 min, the supernatant was further 4-fold concen- trated and used as the test sample. Aliquots of 50 yl of test samples or control TGF-; (serial dilution from 100 ng/ml) were diluted 2-fold with 0.1 M carbonate- bicarbonate buffer (pH 9.5) and immobilized onto a 96-well micro- plate for ELISA (Japan Intermed Co., Japan) by incubation at 4°C overnight. The plate was washed 3 times with 0.05% Tween-20 (Bio-Rad Lab.) in PBS (TPBS), blocked with 200 ul of 1% gelatin- containing PBS at 37°C for 1 hr, and wahed with TPBS 3 times. A 50-1 aliquot of 150-fold diluted rabbit anti-human TGF-f; antibody (IgG; King Brewing Co., Japan) was added to each well. After 30 min, the plate was washed with TPBS 5 times, and peroxidase- labelled goat anti-rabbit IgG (H+L) (Kirkeguard and Perry Lab. Co.) was added to each well. After 1 hr, the plate was again washed with TPBS 5 times and 100 «1 of peroxidase substrate solution was added to each well. After 30 min, the reaction was stopped by adding 50 ul of 2 M H,SO,. Absorbance at 490 nm was measured with a microplate reader (Bio-Tek, EL-309). As a control for specific antibody, rabbit IgG (Zymed Lab.) was used. 5. Growth factors Basic FGF (bFGF, purified from bovine brain) was purchased from R & D systems, U.S.A. TGF-f (recombinant human TGF-£;) was from king Brewing Co., Japan. Insulin (purified from bovine pancreas) was from Sigma Chemical Co., U.S.A. Platelet-derived growth factor (purified from human leucocytes) was from Collabora- tive Research Inc., U.S.A. RESULTS 1. Effects of serum deprivation of chondrogenic fibroblasts in soft-agar culture Chondrogenic fibroblasts in soft-agar culture in 10% FBS gives rise to two types of clonal colonies [20]. The F-type colony is round and consists of mutually adhering flattened fibroblasts (Fig. 1A), and the C-type colony consists of scat- ered large chondrocytes (Fig. 1B) with a halo of cartilage matrix, which is positive to staining with Alcian blue (Fig. 1C) or anti-PG-H antibody [20]. The colony-forming efficiency decreases by lowering serum concentration (Fig. 1D). Because no colonies were observed in serum-free medium, even in the presence of various growth factors, the analysis of the effects of different factors were performed in medium containing 2% FBS. 2. Basic FGF induces colony formation without promotion of differentiation, and FBS promotes chondrogenic differentia- tion As shown in Table 1, 10 ng/ml bFGF induces many large F-type colonies in chondrogenic fibroblasts, which suggests that, relative to TGF-, PDGF and insulin, bFGF does not promote cartilage differentiation by the proliferating fibro- blasts. The number of F-type colonies increases and that of C-type colonies decreases as a function of bFGF concentra- tion (Table 1). FBS at 10% is effective at promoting C-type colonies, which suggests that FBS contains a strong activity for induc- tion of cartilage differentiation from the fibroblasts (Table 1). The bFGF-induced large F-type colonies can differentiate into C-type colonies (C-type conversion) [20], if the culture medium changed on day 20 to fresh medium containing 2% FBS without bFGF and the cultures are maintained for further 14 days (Fig. 2). This suggests that cells proliferating in the presence of bFGF retain their differentiative capacity for at least 20 days. With differentiated chondrocytes from the scleral carti- lage layer, bFGF also induces rapid proliferation to form many colonies (data not shown). In this case, cell size remains smaller (mean diameter 10 ~m) than in the absence of bFGF (mean diameter 30 «m); this suggests that there is no hypertrophic differentiation, which takes place in the Growth Regulation in Chondrogenic Fibroblasts 263 15 D (%) 10 iciency Colony-forming eff 0 0 5 10 Concentration of FBS(%) Fic. 1. Colony formation of chondrogenic fibroblasts (scleral fibroblasts) in soft-agar culture and the effects of serum concentration. A: F-type colony. B: C-type colony. Both photographs are the same magnification, suggesting large cell size in C-type colony. A and B were observed on the day 20 of culture by phase-contrast microscope. C: The same colony as B, stained with Alcian blue, showing a halo of cartilage matrix surrounding individual chondrocytes, as described previously [20]. Bar, 100 4m. D: Effects of serum (FBS) concentration on colony formation. Each point represents the average of results from three dishes with standard deviation. Fic. 2. Chondrocyte differentiation (C-type conversion) from bFGF-dependent large F-type colony. After culturing for 20 days with 10 ng/ml bFGF, the medium was changed to fresh medium containing 2% FBS andno bFGF. The culture was further incubated for 14 days. The same colony was photographed at the 7th day (A) and 14th day (B) after medium change. Rounded cells, which migrate out from the periphery of the F-type colony, are differentiating chondrocytes (C-type conversion, as described previously) [20]. About 5% of the bFGF-dependent F-type colonies manifested the C-type conversion. Bar, 100 um. Y. OuyYA AND K. WATANABE TaBLE 1. Effects of various growth-promoting substances on colony formation by chondrogenic fibroblasts in soft-agar culture Number of C-type F-type LF-type colonies obtained* colony colony colony” (Experiment A)° None (2% FBS only) 80 ( 4) 13 67 0 bFGF (10 ng/ml) 780 (39) 0 780 585 TGF-f (1 ng/ml) 140 ( 7) 10 130 67 PDGF (3 U/ml) 100 ( 5) 3 97 30 Insulin (10 «g/ml) 140 ( 7) 36 104 7B 10% FBS 420 (21) 311 109 0 10% FBS with bFGF? 1960 (98) 510 1450 1058 (Experiment B) None (2% FBS only) 60 ( 3) 30 30 0 bFGF (0.01 ng/ml) 200 (10) 30 170 68 (0.1 ng/ml) 220 (11) 33 187 103 (1 ng/ml) 180 ( 9) 18 162 97 (10 ng/ml) 280 (14) 0 280 126 a: Dissociated cells were inoculated at 2000 cells per dish. Mean values cbtained from three dishes are shown. Parenthesis means colony-forming efficiency. LF-type (large F-type) colony means F-type colonies with diameters more than 81 um. c: Experiment A and Experiment B were different sets of cell culture. Growth factors were supplemented together with 2% FBS unless otherwise indicated. The variation in the results between these two experiments is as expected for different sets of cultures. d: bFGF was added at 10 ng/ml. ‘with bFGF fe without bFGF a Fic. 3. Two different types of C-type colonies derived from differentiated chondrocytes. A: In the presence of 1 ng/ml bFGF. B: In the absence of bFGF. Both photographs are the same magnification. Both cultures contain 10% FBS. Photographed on day 20 of culture. Bar, 100 ~m. Growth Regulation in Chondrogenic Fibroblasts 265 absence of bFGF (Fig. 3) [7]. 3. TGF-8 modulates the bFGF-dependent proliferation de- pending on the state of differentiation Supplementation with bFGF alone raises colony-forming efficiency of both chondrogenic fibroblasts (@; Fig. 4A) and differentiated chondrocytes (@; Fig. 4C). However, supple- mentation with TGF-f alone does not increase it for either cell type (4; Fig. 4B, D). Chondrogenic fibroblasts Colony-forming efficiency (%) bFGF (ng/ml) When TGF-f is supplemented with 1 ng/ml bFGF, TGF- 8 inhibits bFGF-dependent colony formation by chondro- genic fibroblasts in a dose-dependent manner at a range of 0.001-1 ng/ml (A; Fig. 4B). In contrast, in the presence of bFGF, TGF-f increases cloning efficiency by differentiated chondrocyte at the same range of concentrations (A; Fig. 4D). TGF-f at these concentrations does not affect cellular differentiation in soft-agar culture (data not shown). Colony-forming efficiency (%) 0 0.001 0.01 0.1 1 TGFB (ng/ml) Differentiated chondrocytes 100 80 60 40 20 Colony-forming efficiency (%) 0 0.01 O.1 1 10 bFGF (ng/ml) Fic. 4. Effects of bFGF and TGF-f on colony formation. the case of differentiated chondrocytes. ml). Solid triangles (4): TGF-f alone. three dishes are plotted. A and B: In the case of chondrogenic fibroblasts. Solid circles (@): bFGF alone. Open triangles (4): TGF-@ with bFGF (1 ng/ml). 100 80 60 40 20 Colony-forming efficiency (%) 0 0.001 0.01 0.1 1 TGFB (ng/ml) C and D: In Open circles (O): bFGF with TGF-f (0.1 ng/ Mean values obtained from 266 Y. OHYA AND K. WATANABE 4. Presence of anti-TGF-2 immunoreactivity as autocrine/ paracrine regulators Assay of concentrated conditioned medium by ELISA with anti-TGF-f; antibody indicates that conditioned medium contains a TGF-f-like molecule (Fig.5). The amount in conditioned medium seems to be higher at the early days of culture. 0.16 0.1 Absorbance at 490 nm 0.08 0.06 2 4 6 8 Days of culture Fic. 5. Detection of TGF--like molecule in conditioned medium of chondrogenic fibroblasts in monolayer culture by ELISA with anti-TGF-f, antibody. Concentrated conditioned medium was immobilized onto wells (see Materials and Methods). Absorb- ance of the negative control (non-immune rabbit IgG) was 0.023, and that of the positive control (25 ng/ml TGF-f;) was 0.127. Data given are the average values from 3 replicate wells with standard deviation. DISCUSSION With chondrogenic fibroblasts (scleral fibroblasts), bFGF [6] induced rapid proliferation to produce many large F-type colonies (Table 1). Since many of the chondrogenic fibro- blasts possess the capacity to proliferate and differentiate into C-type colonies when 10% FBS is present [20] (Table 1), bFGF was considered to have a capacity to induce these cells to proliferate without promotion of differentiation. The differentiative potentiality of the proliferating cells was main- tained even up to 20 days after initial treatment with 10 ng/ml bFGF (Fig. 2). With differentiated chondrocytes (scleral chondrocytes), bFGF also induced rapid proliferation with suppression of hypertrophic differentiation (Fig. 3), similar to the case for rabbit chondrocytes [7, 8]. The suppression of differ- entiation by bFGF does not imply inhibition of phenotypic expression. In fact, Kato [8] pointed out that bFGF stabi- lized phenotypic expression and prevented dedifferentiation of proliferating chondrocytes, and, as a result, promoted proteoglycan synthesis. Taken together, the function of bFGF is to induce proliferation of both undifferentiated and differentiated chondrogenic cells without promotion of differentiation, i.e., to expand a cell population as it exists. We emphasize that an initial treatment with bFGF is sufficient for a persistent effect for at least 20 days. This may suggest the existence of a mechanism to hold a cellular differentiative state steady during persistent proliferation, similar to the auto-induction in steroid-hormonal regulation [18]. As to the differentiation-promoting activity, which at least FBS possessed, bone morphogenetic proteins (BMPs) [14] have been reported to be differentiation factors in limb bud mesoderm of the chick [3, 4]. In fact, recombinant BMP-4 has been found to induce C-type conversion (Wata- nabe, Hayashibe and Takaoka, unpublished data). It would be of interest to determine, whether FBS contains a member of the BMP family or stimulates autocrine production of one. As shown in Figure 4, TGF-£ [10, 13] displays an inverse effect on the bFGF-dependent proliferation of chondrogenic fibroblasts and differentiated chondrocytes, which are located adjacent to each other in vivo. Since our three-dimensional cultures should reflect the in vivo function of chondrogenic cells [1, 16], it may be that increased endogenous TGF-/-like molecule gives rise to overall growth of cartilage tissue, together with growth suppression of the adjacent perichon- drium, in the presence of endogenous bFGF-like growth- promoting activity [19]. It was found that the chondrogenic fibroblasts secreted TGF-f-like molecule into their conditioned medium (Fig. 5). High activity at the early days in culture might be explained as an induction by primary monolayer cultivation. It is plausi- ble that these factors play regulatory roles in the differential growth and differentiation of chondrogenic cells in in vivo scleral chondrogenesis. ‘ACKNOWLEDGMENTS We thank Prof. Minoru Amano for his continuous encourage- ment and Dr. Akira Kawahara for his kind advice. Thanks are also due to Dr. David A. Carrino for reviewing manuscript. This work was supported in part by Grants-in-Aid for Scientific Research (No. 03833023 and No. 03304009) to K. W. from the Ministry of Educa- tion, Science and Culture of Japan. REFERENCES 1 Benya PD, Shaffer JD (1982) Dedifferentiated chondrocytes reexpress the differentiated collagen phenotype when cultured in agarose gels. Cell 30: 215-224 Carrington JL, Reddi AH (1990) Temporal changes in the response of chick limb bud mesodermal cells to transforming growth factor f£-type 1. Exp Cell Res 186: 368-373 3 Carrington JL, Chen P, Yanagishita M, Reddi AH (1991) Osteogenin (bone morphogenetic protein-3) stimulates cartilage i) 10 11 Growth Regulation in Chondrogenic Fibroblasts 267 formation by chick limb bud cells in vitro. Develop Biol 146: 406-415 Chen P, Carrington JL, Hammonds RG, Reddi AH (1991) Stimulation of chondrogenesis in limb bud mesoderm cells by recombinant human bone morphogenetic protein 2B (BMP-2B) and modulation by transforming growth factor B,; and #2. Exp Cell Res 195: 509-515 Fujioka M, Shimamoto N, Kawahara A, Amano M, Watanabe K (1989) Purification of an autocrine growth factor in con- ditioned medium obtained from primary cultures of scleral fibroblasts in the chick embryo. Exp Cell Res 181: 400-408 Gospodarowicz D, Neufeld G, Schweigerer L (1986) Molecu- lar and biological characterization of fibroblast growth factor, an angiogenic factor which also controls the proliferation and differentiation of mesoderm and neuroectoderm derived cells. Cell Differ 19: 1-17 Kato Y, Iwamoto M (1990) Fibroblast growth factor is an inhibitor of chondrocyte terminal differentiation. J Biol Chem 265: 5903-5909 Kato Y (1992) Roles of fibroblast growth factor and transform- ing growth factor-f families in cartilage formation. In “Biologi- cal Regulation of the Chondrocytes” Ed by M Adolphe, CRC Press, Florida, pp 241-279 Leonard CM, Fuld HM, Frenz DA, Downie SA, Massague J, Newman SA (1991) Role of transforming growth factor-8 in chondrogenic pattern formation in the embryonic limb: Stimula- tion of mesenchymal condensation and fibronectin gene expres- sion by exogenous TGF-f and evidence for endogenous TGF-- like activity. Develop Biol 145: 99-109 Moses HL, Bramm EB, Proper JA, Robinson RA (1981) Transforming growth factor produced by chemically- transformed cells. Cancer Res 41: 2842-2848 Ohya Y, Watanabe K, Shimamoto N, Amano M (1992) Scle- 15 16 17 18 20 ral fibroblasts of the chick embryo can proliferate without transferrin in protein-free culture. Zool Sci 9: 749-755 Richman JM, Crosby Z (1990) Differential growth of facial primordia in chick embryos: Responses of facial mesenchyme to basic fibroblast growth factor (bFGF) and serum in micromass culture. Development 109: 341-348 Roberts AB, Anzano MA, Lamb LC, Smith JM, Sporn MB (1981) New class of transforming growth factors potentiated by epidermal growth factor: Isolation from non-neoplastic tissues. Proc Natl Acad Sci USA 78: 5339-5343 Rosen V, Thies RS (1992) The BMP proteins in bone forma- tion and repair. Trends Genet 8: 97-102 Shinomura T, Kimata K, Oike Y, Maeda N, Yano S, Suzuki S (1984) Appearance of distinct types of proteoglycan in a well- defined temporal and spatial pattern during early cartilage formation in the chick limb. Develop Biol 103: 211-220 Solursh M, Linsenmayer TM, Jensen KL (1982) Chondrogene- sis from single limb mesenchymal cells. Develop Biol 94: 259- 264 Stewart PA, McCallion DJ (1975) Establishment of scleral cartilage in the chick. Develop Biol 46: 383-389 Tata JR, Baker BS, McChuca I, Rabelo EML, Yamauchi K (1993) Autoinduction of nuclear receptor genes and its signi- ficance. J Steriod Biochem Molec Biol 46: 105-119 Watanabe K, Fujioka M, Takeshita T, Tsuda T, Kawahara A, Amano M (1989) Scleral fibroblasts of the chick embryo pro- liferate by an autocrine mechanism in protein-free primary cultures: Differential secretion of growth factors depending on the growth state. Exp Cell Res 182: 321-329 Watanabe K, Yagi K, Ohya Y, Kimata K (1992) Scleral fibroblasts of the chick embryo differentiate into chondrocytes in soft-agar culture. In Vitro Cell Dev Biol 28A: 603-608 reer! seas Ayiths a Ge “Ni c- veer 2 Bat Sine ats Wot Re tap Meal Wher ira bf oO th csenn7zo bias me eta ' > vas} ft Comes s Hila A Anaiod nioierey 76 mnt) Now (tft) iindael sitast aiw wy neagaliigs i ) fae. DRA iinh KB ae | Kee ory Ear aaa ‘ eed? Perid -fee 4 AT ase? ored ! rr AL 1 Rat : le vorgereae beget 2 } a3) NOI sta Wed : air), ee eri * t usu Ms a AG Wa iat to svt ee se eH Mt ein 4 ios ty ¢. : nse y; AG a 2 “J {Oe if l 3 t £575 en 15 aT ta is »* = E 20) a sf i ly. ne ee lal sont -nisitrny as i - Atl ealileraon wt bith ami ator Wi pbians #0 Decl ie Vibe! on Bee a ang Pays yey EF E a ATED nibhiee Oe Sheba bys ne tsa Mew rg aewiehbat® toad ys sone aia: oa icone Aso hepinshicoa: i ner natoe Unavua YARD 4 ‘ ‘ednes arte cal aio ; 8: i perl Mik (Pen ROSES nie DEM oa was > X Csupeaeet. ts geteuye Ay See BPs es 4 T fi A Areot aa: eC Settler Manned 9h pisciee | eben eyes ; sani on sp cicada dere ath fie do aio caamten a ett eg hiest ai Me tie utes yey NG Tx Hai rary ¥ naitien aa é tad ri EAE D, , Rie iyhtifad SS eee a z pw) Evid Vs “As Me » , i Ayeewtl ter (oe . er iit he ivf ‘ > CGP SF ’ a Aoi T # ie 2 (eet ras i larer = lVistety a Ub Tae Nt et aaltee dt > By ih WiPiteeits “bh pliespepye ay) rer Téye7eical ‘ot Ppcelormpeh Cole ¢ 23) Ped a, le Foimiyetite Of : = oy prac renine y oteice A yihg ih a ZOOLOGICAL SCIENCE 11: 269-274 (1994) © 1994 Zoological Society of Japan Membrane-bound Inclusions in the Leydig Cell Cytoplasm of the Broad-headed Skink, Eumeces laticeps (Lacertilia: Scincidae) NATHAN O. OKIA Department of Biology, Auburn University at Montgomery, 7300 University Drive, Montgomery AL 36117-3596, USA ABSTRACT—The structure of the Leydig cell was studied in laboratory maintained skinks Ewmeces laticeps (Schneider). Overall, skink Leydig cell cytology was similar to that of mammalian steroidogenic cells. Under both the light and electron microscope, for example, skink Leydig cells contained large lipid droplets in their cytoplasm. However, skink Leydig cells also contained elongated membrane bound rods and other rounded structures which were also surrounded by tightly packed concentric layers of membranes. The rounded structures were of two types, oval figures with empty interiors and fibrous structures with ribosome-like elements in their interiors. The elongatd, rod-like structures resembled crystalloids found in aging human corpora lutea. The oval structures resembled myelin sheaths and the round but wooly-looking bodies resembled compact whorls of membranes found in mice Leydig cells. INTRODUCTION A mammalian Leydig cell is often described as having extensive arrays of smooth endoplasmic reticulum appearing as long connected tubulues [3]. In the cytoplasm may be large lipid inclusions, which in some instances is evidence of cellular regression [1]. In addition, human and other mammalian Leydig cells, contain proteinaceous crystals of Reinke of unknown function or origin [8, 9]. Similarly, active Leydig cells in Anolis lizards, have extensive arrays of agranular endoplasmic reticulum, darkly-staining mitochon- dria and lipid granules without crystalline inclusions [13]. In the current study on the ultrastructure of skink Leydig cells, testes of laboratory maintained skinks Eumeces laticeps (Schneider) which were obtained during the periods of tes- ticular recrudescence (January) and breeding (May) in the wild [7] were found to contain membrane-bound inclusions some of which looked crystal-like. MATERIALS AND METHODS Four animals captured in late winter around the campus were kept for 6 or more months in glass tanks, provided with water ad libitum and fed crickets daily. They were sacrificed either in Janu- ary or May by decapitation with one stroke of scissors. Small pieces of testes were immersed in Karnovsky’s fixative, pH 7.4 for 2 hr, postfixed in 1% OsO, in 0.1M sodium cacodylate buffer at room temperature for 2hr. After dehydration through alcohols and ace- tone they were embedded in Spurr’s Epoxy mixture. Semi-thin sections were stained with toluidine blue and basic fuchsin (Electron Microscope Science) and ultrathin sections were stained with uranyl acetate and lead citrate and examined with a Philips 200 electron- microscope at an accelerating voltage of 60 kV. Accepted February 3, 1994 Received November 30, 1993 RESULTS Under the light microscope, Leydig cells which were close to the basement membrane contained large lipid dro- plets in their cytoplasm (Fig. 1). Under the electron micro- scope, some of the Leydig cells similarly showed cytoplasms that were full of lipid granules with few other organelles visible (Fig. 2). The cytoplasm of what were considered Leydig cells was in all cases very granular and in some, contained inclusions consisting of rods which were visible at the level of the light microscope (Fig.3). At the ultra- structural level, Leydig cell cytoplasm also contained other rounded structures that resembled myelin sheaths (Fig. 4). Between the rod-like structures and lipid granules were several free ribosomes and rough endoplasmic reticulum. Except for the solid dark outlines of the rods, their cores were not stained. The lipid granules on the other hand, had unstained cores and edges which were stained to varying degrees. Confluence between the lipid granules were some- times observed (Fig. 5). Other than the simple rectangular pattern, some rods showed pointed ends and their darker outlines consisted of multiple layers of membranes (Fig. 6). Some rods contained fiber-like structures inside their cores (Fig. 7). At higher resolution, the dark walls of myelin-like structures and crystalloids consisted of multiple layers of membranes with alternating dark and light bands which resembled unit membranes (Figs. 8 and 9). In some sec- tions, the light inner band between the 2 dark bands was bisected by a longitudinally positioned dark line (Fig. 10). The myelin structures sometimes enclosed portions of cyto- plasm (Fig. 8) or empty spaces (Fig. 11). Also found was a wooly structure with whorls of concentric fiber enclosing ribosome-like elements (Fig. 12). All the observed struc- tures were found in testes undergoing spermatogenesis and spermiogenesis and their shape or distribution was invariable in the sections examined. 1 27 Skink Leydig Cells in . 10NS o NS, bound Inclus Membrane 22 N. O. Oxia Membrane-bound Inclusions in Skink Leydig Cells 273 DISCUSSION A steroid producing cell should increase its membrane stores during active steroidogenesis and the increase in smooth endoplasmic reticulum in lipid producing cells is evidence of this. The ordered arrangement of smooth endo- plasmic reticulum around liposomes in the armadillo Leydig cell is said to be one means of increasing surface area for cholesterol and steroid production [14]. Based on the appearance of their walls, it is possible that the inclusions found in skink Leydig cell cytoplasm are some special type of smooth endoplasmic reticulum like the kind described as tubular in the opossum Leydig cells [4]. However, only the rod-like inclusions in the skink Leydig cell cytoplasm can be considered tubular. The skink ovoid structures with lamin- ated walls resemble lipid vacuoles found in the armadillo [14] and dog Leydig cells [6], which are surrounded by whorls of smooth endoplasmic reticulum. The skink membranes were however, more tightly packed and uniformly spaced like myelin figures in residual bodies [5]. Similarly, the inclu- sions with concentric whorls of membranes (Fig. 12) resemble those inclusions that have been described as “compact whorls of membranes” [8, 12] of unknown function rather than smooth endoplasmic reticulum. Therefore, these structures are not specialized smooth endoplasmic reticulum. Crystals have been found in mammalian but not reptilian Leydig cells and so the data reported here are probably the first indication of crystal-like structures in reptilian Leydig cells. These rod-like structures in the skink Leydig cell cytoplasm resemble rod-like structures in aging human lutein cells [2, 5] described as lipid-soluble and possibly made of cholesterol. The characteristic of having laminated walls links them to cylindrical bodies that were described in rat Leydig cells [11] as having walls composed of a helical array of tubular units possibly derived from the endoplasmic reticu- lum [11]. But in the skink Leydig cell bodies, there was no connection between the membranes of the laminated walls and the endoplasmic reticulum. Though found in the Leydig cells, these skink rods do not resemble Reinke’s crystals which occur in human and other mammalian Leydig cells. These bodies appear to be unique to the skink and their true nature has to await further investigation. Like crystals, lipid droplets have also been found to increase in regressing Leydig cells [1, 10] and their occurrence in skink Leydig cells could be interpreted as portending their regression, but no evidence of such regression was noted. ACKNOWLEDGMENTS I would like to thank Dr. Emilio Mora for use of the electron microscope facility, Dr. William Cooper, for securing and identifying the skinks, and Dr. Tom Denton for financial support from the Department. Partial support for this work was provided by the NSF’s Instrumentation & Instrument Development Program Grant + DIR-8907860 for the procurement of the ultramicrotome and the Auburn University at Montgomery Grant-in-Aid Program. Fic. 1. The skink testis showing a cluster of Leydig cells (L) containing several lipid vesicles in the cytoplasm next to the basement membrane (b) of the seminiferous tubule, which shows four Sertoli cells (S) resting on the basement membrane. Bar=8 pm. Fic. 2. A Leydig cell with large lipid granules occupying most of the cytoplasm. A few mitochondria (m), lysosome (I), and smooth endoplasmic reticulum (e) are also discernable. Bar=1 ym. Fic. 3. A segment of the interstitial space showing Leydig cells located further from the basal lamina. The granularity of the cytoplasms and the presence of non-stained rods is evident (arrows). Bar=8 yum. Fic. 4. This section through a Leydig cell shows a number of cytoplasmic inclusions including a myelin-like body (M), lysosome (I), non-stained rods (r). Also shown is the nucleus (N) with a ring of heterochromatin all along the inside of the nuclear envelope. Bar=0.6 um. Fic.5. A section through a Leydig cell showing lipid granules (I), tubules of rough endoplasmic reticulum (e) and a non-stained rod or cylinder, here called a crystal (c). Bar: 0.3 ~m. Fic. 6. A Leydig cell cytoplasm showing two crystals (c) with their fibrous wall. An enlarged mitochondrion (m) with degenerating cristae is also shown. e, rough endoplasmic reticulum; n, nucleus. Bar=0.2 um. Fic. Fic. Fic. Fic. Fic. Fic. 7. The cytoplasm of a Leydig cell showing 4 crystalloids, one with a fibrous core (c); liposome (1), lysosome (s) and mitochondrion with laminated cristae (m). Bar=0.2 um. 8. A segment of a Leydig cell cytoplasm showing a number of membrane-bound inclusions all of which show evenly spaced parallel or concentric membranes. The figure is an enlargement of (Fig. 4). c, a crystal; m, myelin-like inclusion; v, vacuole. Bar=0.1 um. 9. The cytoplasm of a Leydig cells showing detail of the crystal (c) wall which consists of regularly spaced alternating light and dark bands. Bar=0.1 um. 10. A tangential section through a small crystal (c) in the cytoplasm of a Leydig cell, showing detail of the membranes making up the crystal wall. It shows that the light longitudinal band is bisected by another dark line (arrow). Bar=0.1 ym. 11. A section through an oval shaped laminated structure within the cytoplasm of a Leydig cell. The wall of this structure consists of concentrically arranged membranes like those found around the crystalloids. Inside the structure are what appear to be cytoplasmic debris. Arrow points at the cell membrane for comparison with crystal wall membranes. Bar=0.1 «m. 12. Another type of membrane-bound structure found in the cytoplasm of Leydig cells which resembles those structures that have been described in literature as “compact whorls of membranes”. The wall of this structure encloses what looks like free ribosomes within the cytoplasm. Bar=0.2 um. N. O. OKIA REFERENCES Andersen K (1978) Seasonal change in fine structure and function of Leydig cells in the blue fox, Alopex lagopus. Int J Androl 1: 424-439 Carsten PM, Merker HJ (1965) Die Darstellung von Cholesterinkristallen im elektronenmikroskopischen. Bild Frankfurt Z Path 74: 539-543 Christensen AK (1965) The fine structure of testicular intersti- tial cells in guinea pigs. J Cell Biol 26: 911-935 Christensen AK, Fawcett DW (1961) The normal fine struc- ture of opossum testicular interstitial cells. J Biophys Biochem Cytol 9: 653-670 Christensen AK, Gillim SW (1969) The correlation of fine structure and function in steroid-secreting cells, with emphasis on those of the gonads. In “The Gonads” Ed. by KW McKerns, Appleton-Century-Crofts, New York pp. 415-488 Connell CJ, Christensen AK (1975) The ultrastructure of the canine interstitial tissue. Biol Reprod 12: 368-382 Cooper WE, Vitt LJ (1987) Intraspecific and interspecific aggression in lizards of the scincid genus Eumeces: chemical detection of conspecific sexual competitors. Herpertologica 43: 10 11 12 13 14 7-14 De Kretser DM, Kerr JB (1988) The cytology of the testis. In “The Physiology of Reproduction” Ed. by KE Neill et al., Raven Press, New York, pp 837-932 Fawcett DW, Burgos MH (1960) Studies on the fine structure of the mammalian testis. The human interstitial tissue. Am J Anat 107: 245-269 Gustafson AW (1987) Changes in Leydig cell activity during the annual testicular cycle of the bat Myotis licifugus lucifugus: Histology and lipid histochemistry. Am J Anat 178: 312-325 Murakami M, Kitahara Y (1971) Cylindrical bodies derived from endoplasmic reticulum in Leydig cells of the rat testis. J Electr Microsc 20: 318-323 Ohata M (1979) Electron microscope study on the testicular interstitial cells in the mouse. Arch Histol Jap 42: 51-79 Pearson AK, Tsui H, Licht P (1976) Effect of temperature on spermatogenesis, on the production and action of androgens and on the ultrastructure of gonadotropic cells in the lizard Anolis carolensis. J Exp Zool 195: 291-303 Weaker FJ (1977) The fine structure of the interstitial tissue of the testis of the nine-banded armadillo. Anat Rec 187: 11-28 ZOOLOGICAL SCIENCE 11: 275-284 (1994) © 1994 Zoological Society of Japan Identification and Localization of a Ligand Molecule of Xenopus Cortical Granule Lectins Norio YOSHIZAKI Department of Biology, Faculty of General Education, Gifu University, Gifu 501-11, Japan ABSTRACT—This study aimed to identify the ligand molecule of the cortical granule lectins which participate in the formation of the fertilization (F) layer in Xenopus laevis. Comparison of fertilization envelopes (FEs) with and without the F layer by SDS-PAGE showed a 105-kDa glycoprotein (gp 105) only in the former FEs. This glycoprotein was isolated by differential centrifugation and electrophoretical extraction from an F layer extract. Immunoblotting with an antiserum against gp 105 produced staining on the gp 105 in the FEs; the relative amount of gp 105 increased during the hatching period due to the digestion of vitelline envelope components. Staining on westernblots with HRP-conjugated peanut agglutinin suggested that gp 105 contains a small amount of galactosides. With immunoelectron microscopy, gold particles indicating the location of gp 105 were visible on the pre-fertilization (PF) layer of eggs obtained from the pars recta 2 (PR2) and the uterus of the oviduct, but they were few on the F layer of activated eggs. With the PA-CrA-Silver method for detecting carbohydrates, silver particles appeared on the PF layer but not on the main body of the F layer. The binding of gold-conjugated lectins to gp 105 on westernblots showed that gp 105 interacts with the cortical granule lectins. It was concluded from these results that gp 105 is a natural ligand of the lectins, and that it resides in the PF layer and is supplied to eggs at the PR2. INTRODUCTION Fertilization in Xenopus laevis induces the formation of a fertilization (F) layer in the egg envelopes as well as a hydrolytic event which changes the vitelline envelope (VE) of unfertilized eggs to that (VE*) of fertilized eggs [7, 12, 21]. Subsequently, the fertilization envelope (FE) consists of the VE* and the F layer and it acts as a block to polyspermy [5, 6]. The F layer is formed by the interaction of cortical granule lectins with ligand molecules. The nature of the lectins has been well documented [3, 16, 28, 29] but that of the ligand molecules not. Two sources for the ligand mole- cules have been proposed, the innermost jelly layer [25] and the pre-fertilization (PF) layer [26, 31]. The jelly layers of Xenopus eggs are composed of four morphologically distinct layers [27]. Wyrick et al. [25] first demonstrated a precipitation reaction on agar plates between the solubilized jelly of the innermost layer and the lectins. Ligand molecules of the jelly were recently identified im- munoelectrophoretically by Birr and Hedrick [2] but have not yet been characterized. Yoshizaki and Katagiri [31] reported a PF layer lying between the VE and the innermost jelly layer of uterine eggs; it is produced by epithelial cells at the pars recta 2 (PR2) of the oviduct [26]. Its honeycomb structure was shown through quick-freeze, deep-etch electronmicroscopy by Larabell and Chandler [11]. When eggs obtained from the PR2 were activated or treated with lectins, their PF layer became similar to the F layer morphologically and biochemi- cally [31]. Furthermore, the F layer could not be produced in activated eggs when they were deprived of the PF layer Accepted February 16, 1994 Receired December 9, 1993 [31]. The PF layer, then, seems essential for F layer forma- tion. However, the identity of the ligand molecule in the PF layer was not known. How could this ligand molecule be found? The strategy was to compare FE with and without the F layer electrophoretically and then isolate the molecule from an F layer extract. The result was a 105-kDa glycopro- tein (gp 105) which was present in the PF layer and interacted with cortical granule lectins. MATERIALS AND METHODS Collection of eggs South African clawed frogs, Xenopus laevis, were purchased from a dealer in Hamamatsu, Japan and reared at 22-24°C. Ovula- tion was induced by injection of 1,000 IU gonadotropin (Gonatropin, Teikoku Zoki Co.) and a sufficient number of oviductal eggs were obtained 7-10 hr after hormone injection. Uterine eggs were obtained by squeezing females every hour after the start of oviposi- tion. Artificial insemination was performed by the method of Moriya [14]. Developmental stages were determined according to the normal table of Nieuwkoop and Faber [15]. Procedures for isolating gp 105 Uterine eggs obtained from 3 females were placed in 0.05 De Boer solution (DB: 110 mM NaCl, 1.3 mM KCl, 1.3 mM CaCl, 10 mM Tris-HCl, pH 7.4), artificially activated with a 100 V AC current for 10sec, and left for 30 min in the solution. Eggs were then dejellied by brief treatment with 20mM dithiothreitol (DTT) in Ca-free 0.05 DB (adjusted to pH9.0 with NaOH) and washed extensively with 0.05 DB. The F layer was extracted from these dejellied eggs by treatment with 5 mM EDTA in Ca-free 0.05 DB for 10min. The F layer extract (ca. 100ml) was dialyzed against distilled water overnight and concentrated to 5 ml by ultrafiltration. A precipitate appeared during the procedure of concentration; this was subsequently centrifuged at 7,000g for 30min at 4°C. The resulting pellet was suspended in a solution of 20 mM DTT and 5 mM 276 N. YOSHIZAKI EDTA in Ca-free 0.05 DB (pH 9.0), freeze-thawed and centrifuged at 1,000xg for 15min. The precipitate was subjected to 7.5% SDS-PAGE. Protein bands were visualized by treatment with 4 M sodium acetate [8]. Sections of 3 mm width at Rf=0.3 were sliced out and cut into several pieces. The gp 105 was extracted elec- trophoretically from the pieces of gel into 2.5 mM Tris-HCl buffer, pH 8.3, in a Max-Yield Protein Concentrator (Atto Co.) at 5 W for 2.5 hr. Preparation of egg envelopes and cortical granule lectins A jelly solution was obtained from activated uterine eggs by DTT-treatment and dialyzed against distilled water. It was then lyophilyzed and dissolved in DB when used. VE*s were isolated by the method of Wolf et al. [24] from eggs whose F layer had been removed. The eggs were crushed by passage through a hypodermic syringe with an 18 G-gauge needle and the VE*s were sieved out through a nylon mesh (82 um). After thorough washing in distilled water, the VE*s were collected by centrifugation. Before being used, they were suspended in DB with an ultrasonic vibrator. Egg FEs were isolated from dejellied eggs by the same method as mentioned above. Embryonic FEs were prepared as follows. Embryos were dejellied at stages 17-19 and cultured in 0.05 DB. The FEs were isolated from the embryos manually with watchmak- er’s forceps. Since dejellied embryos would hatch precociously at stage 28, FEs of later embryos were obtained by incubating the FEs and embryos at stage 28 for 15 hr from the time of reaching stage 28 [32]. Cortical granule lectins were obtained from activated coelomic eggs and purified with an affinity column [29]. The lectins were labeled with colloidal gold prepared by the tannic acid procedure of Slot and Geuze [18]. Lectin-gold complexes were made at 4°C by mixing twice the minimal stabilization amount of lectins with the colloidal gold solution. The mixtures were then centrifuged at 100,000 x g for 30 min and the precipitates suspended at an approxi- mate concentration of 20 g/ml lectin in DB solution containing 1% BSA. Electrophoresis Slab SDS-PAGE in 7.5% gel was carried out as described by Laemmli [10]. The gels were stained with Coomassie blue or PAS. The molecular weights of the protein were estimated from a calibra- tion curve obtained with the standard proteins in the MW-SDS-200 kit (Sigma). Protein was determined by the method of Smith et al. [20] using bicinchoninic acid with BSA as a standard. Detection of carbohydrates on westernblots Electrophoresed samples were electroblotted to nitrocellulose membranes. Galactosides were detected using the method de- veloped by Kitagaki-Ogawa et al. [9]. The membranes were washed with TBS-Tween solution (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.05% Tween 20). Then they were blocked with 0.3% BSA in TBS-Tween for 1 hr and treated with 10 “g/ml horseradish perox- idase-conjugated peanut agglutinin (HRP-PNA) for 1 hr. Control runs were made either by omitting the treatment with HRP-PNA or by treating the membranes with a mixture of HRP-PNA and 2.5 M galactose. The method of Adams [1] was used to detect HRP. Preparation of antiserum Antiserum was raised against isolated gp 105 by injection into rabbits. It was absorbed by glutaraldehyde-fixed pars convoluta (PC) tissue from the oviduct. Immunoblot Westernblotted membranes were blocked overnight in a phos- phate-buffered saline solution (PBS, pH 7.2) containing 10% normal sheep serum and 4% BSA, treated for 1 hr with the antiserum diluted to 1/50 in PBS containing 4% BSA, and immersed for 1 hr with HRP-conjugated sheep antirabbit IgG according to the method of Smith [19]. A control was established by treating the membranes with non-immune rabbit serum in place of the antiserum. HRP was again detected by the Adams method. Electron microscopy Eggs were fixed overnight at 4°C in 2.5% glutaraldehyde in 100 mM cacodylate buffer (pH 7.4), rinsed in buffer, and postfixed for 3 hr in similarly buffered 1% OsO,. The specimens were dehydrated in acetone and embedded in Quetol 812 (Nisshin EM Co.). Ultrathin sections were stained with uranyl acetate and lead citrate and viewed with a JEOL JEM-100SX electron microscope. Immunoelectron microscopy Glutaraldehyde-fixed eggs were dehydrated in ethanol and embedded in Lowicryl K4M (Sigma) at —20°C according to the manufacturer’s instructions. Thin sections were mounted on collo- dion-coated nickel grids. The sections were blocked for 10 min with 0.5% BSA in PBS and washed with PBS. They were then incubated for 1 hr at room temperature with a 1/2,000 solution of antiserum. The sections were subsequently washed with PBS and treated for 1 hr with a gold-conjugated goat antiserum against rabbit IgG (E-Y Lab.). Then they were washed with PBS and with distilled water and stained with uranyl acetate and lead citrate. After staining, the sections were dried and coated with carbon vapor. Control sections were treated with a rabbit non-immune serum and then gold- conjugated goat antiserum. There was no significant labeling on the control sections. Electron microscopic detection of carbohydrates Ultrastructural localization of carbohydrates was detected by the periodic acid-chromic acid-silver methenamine (PA-CrA-Silver) method according to Rambourg er al. [17]. Glutaraldehyde-fixed specimens were dehydrated and embedded in Quetol 812. Thin sections were placed on uncoated stainless-steel grids for staining with PA-CrA-Silver. Control runs were made by treating the sections with silver methenamine without previous oxidation. Assay of interaction of cortical granule lectins with gp 105 Westernblotted membranes were blocked overnight in DB solu- tion containing 1% BSA and treated for 3 hr with gold-conjugated cortical granule lectins in DB solution containing 1% BSA. Con- trols were established by treating membranes with the gold- conjugated lectins plus either galactose (1M) or EDTA (1 mM). The membranes were then processed for silver enhancement with a silver enhancing kit (Bio-Rad). RESULTS Isolation of gp 105 from the fertilization layer The fertilization envelope (FE) consists of a vitelline envelope (VE*) and a fertilization (F) layer in Xenopus (Fig. 1A). The F layer was exposed by removal of the jelly with A Ligand to Cortical Granule Lectin 277 0.5HmM Fic. 1. removed with 20 mM DTT (B), and of an envelope whose fertilization (F) layer was removed with SmM EDTA (C). Electron micrographs showing cross sections of egg envelopes of activated eggs (A), of envelopes whose jelly was FE, fertilization envelope; VE*, modified vitelline envelope in the FE. DTT (Fig. 1B) and was collected after solubilization with EDTA (Fig. 1C). Although DTT may have affected the morphology of the envelopes, as the F layer was loosened from the VE* because of DTT’s low salinity, at this stage of investigation, clearer separation of the jelly and the F layer was desirable. The VE* appeared to decrease in width after jelly removal. Electrophoretic profiles of FEs were compared with those of VE*s to find the constituents of the F layer (Fig. 2). A glycoprotein of 105 kDa (gp 105) appeared with Coomassie blue staining in the profile of the FE but not of the VE* (Fig. 2A), whereas no difference was observed with PAS staining (Fig. 2B). It was also confirmed that gp 105 is not a consti- tuent of the jelly layers (Fig. 2, lane3). Thus gp 105 is a constituent of the F layer. Although it was expected that 39-46-kDa protein bands of cortical granule lectins [29] would be present in the profiles of FE, the present study failed to show any lectin bands. The amount of the lectins contained in the FE loaded may have been subminimal for detection and/or they may have been hidden by relatively large amounts of gp 41. Fig. 3 shows SDS-PAGE profiles of preparations at each step of the isolation procedure. Crude F layer extract (lane 1) was centrifuged at 7,000 g for 30 min; the pellet (lane 2) was suspended once in a 5mM EDTA-20 mM DTT solution, freeze-thawed and reprecipitated by centrifugation at 1,000 x g for 15min. This 1,000Xg precipitate (lane 3) contained glycoproteins of 105, 100, 30 and 29kDa. The 105-kDa band in lane 3 was sliced out and the glycoprotein extracted electrophoretically. The extract (lane 4) gave a single band of gp 105 after reelectrophoresis. 72.3 A ! ae) B 112 —> ss feos | 105— ee 66 61 | 41> eh onan oe 37> a em a Fic. 2. SDS-PAGE of FE (5 yg, lane 1), VE* (5 yg, lane 2) and jelly (5 yg, lane 3) stained with Coomassie blue (A) or PAS (B). A 105-kDa glycoprotein (gp 105) can be seen only in the FE stained with Coomassie blue. In this and subsequent figures the molecular weights are indicated on the left in kilodaltons. Immunoblotting demonstration of gp 105 in F layer Figure 4A shows immunoblots of the 1,000 g precipi- tate (lane 1), FE (lane 2), VE* (lane 3) and jelly (lane 4). The membranes were treated with an antiserum raised against gp 105. There was significant staining on gp 105 and gp 100 in the 1,000Xg precipitate of the F layer extract and on gp 105 in the FE but no specific staining on the VE* and the jelly. The degree of stainability on gp 37 of the FE and VE* was the same as that on the control (Fig. 4B); thus the staining on gp 37 was not specific. 278 N. YOSHIZAKI 105— 1007 NGO oo Ww Fic. 3. SDS-PAGE of F layer extract at each step of the isolation procedure. Lane 1, crude extract (20 ug); lane 2, 7,000xg precipitate (15 yg) of the crude extract; lane 3, 1,000 x g precipi- tate (15 ug) of the 7,000 g precipitates after suspension in a 5 mM EDTA-20 mM DTT solution and centrifugation; lane 4, gp 105 (5 4g) extracted from the lane 3 precipitate. BA 2s Bi AB lee Se vf if 3] | 7 e | Fic. 4. Immunoblot analysis of a 1,000 g precipitate of F layer extract (10 vg, lane 1), FE (20 yg, lane 2), VE* (20 yg, lane 3) and jelly (10 ug, lane 4). Membrane A was treated with an antiserum to gp 105. There was staining on gp 105 and gp 100 in the F layer extract and gp 105 in the FE. Membrane B was treated with non-immune serum and shows faint staining on gp 105, gp 100 and gp 37 (to left of dots). The experiments just discussed showed that gp 105 is a component of the FE but not its location. To localize gp 105 in the FE, the same immunoblotting procedure as above was performed on FEs obtained from embryos at stages 22, 26, 28 and from the 28+15 hr culture. gp 37 staining was used as a The stainability on gp 105 in- creased during the hatching period (Fig. 5), which indicates that the relative amount of gp 105 gradually increased in the control for protein loading. SV hae Fic. 5. Immunoblot analysis of FEs from embryos at various stages showing increased amounts of gp 105 as development proceeds. The amounts of FEs loaded in each lane were adjusted such that the amounts of gp 37 appeared approximately the same, in order to provide a control. Lane 1, stage 22; lane 2, stage 26; lane 3, stage 28; lane 4, stage 28+ 15 hr culture, as described in MATE- RIALS AND METHODS. FEs during the hatching process. Since previous ultra- structural observations showed that the F layer remained substantially unaffected even after the VE* portion of the FE was completely digested by the hatching enzyme [32], it seems that gp 105 resides in the F layer. Immunoelectron microscopical localization of gp 105 in eggs Sections of eggs obtained from various parts of oviducts were treated with antiserum against gp 105 and then with gold-conjugated goat antiserum, followed by electron micro- scope observation. Gold particles were absent in eggs obtained from the pars recta 1 (PR1)(Fig. 6A). In PR2 eggs they were present on the layer covering the outer surface of the vitelline envelope (VE), that is, the pre-fertilization (PF) layer (Fig. 6B). A gold-labeled PF layer lay between the VE and the innermost jelly layer in eggs obtained from the pars convoluta 1 (PC1; Fig. 6C), and it occasionally invagin- ated deeply into the VE in uterine eggs (Fig. 6D). The PF layer appeared to be compressed by the jelly layer. The jelly layer was substantially free of the gold labeling. These observations indicate that gp 105 resides in the PF layer of oviductal eggs which have passed through the PR2. Figure 7 shows a section of activated eggs treated with the same two antisera as above. The F layer could be divided into two areas, a condensed region immediately adjacent to the outer margin of the VE* and a dispersed region peripheral to the condensed one [5, 23]. The gold particles were limited to the dispersed region. Biochemical and ultrastructural demonstration of carbohy- drates Since carbohydrate residues are always found with ligand molecules of lectins, they were tested for in the gp 105. PAS A Ligand to Cortical Granule Lectin 279 Fic. 6. Immunoelectron micrographs of sections of eggs from the pars recta 1 (A), pars recta 2 (B), pars convoluta 1 (C), and the uterus (D) of the oviduct. Gold particles indicating the location of gp 105 reside in the pre-fertilization (PF) layer adhering to the outer surface of the vitelline envelope (VE) in pars recta 2 eggs and between the VE and the jelly (J) layer in pars convoluta 1 eggs. Occasionally they invaginate into the VE in uterine eggs. CG, cortical granule; P, pigment geranule. staining on gels of SDS-PAGE did not mark gp 105. (See significantly decreased the staining (Fig. 8B). These results Fig. 2B again). However, HRP-PNA stained gp 105 and gp suggest that gp 105 possesses, although in very small 100 in the 1,000 x g precipitate of F layer extract and gp 105, amounts, carbohydrates of a galactoside nature. gp 66/61 and gp37 in the FE (Fig.8A, C). Galactose Sections of oviductal or activated eggs were also treated 280 N. YOSHIZAKI Fic. 8. Detection of galactosides on westernblots of a 1,000xg precipitate of the F layer extract (10 wg, lane 1) and of FE (30 pg, lane 2) by treatment with HRP-PNA (A), HRP-PNA plus galactose (B) or no treatment (C). The gp 105 and gp 100 in cy : ; ‘ the F layer extract and the gp 105, gp 66/61 and gp 37 in the FE Fic.7. Immunoelectron micrograph of a section of activated egg. are stained (A). Galactose significantly decreases the staining Gold particles are present on the dispersed region of the (B). fertilization (Fd) layer but absent from its condensed regions (arrows). J, jelly layer; VE*, modified vitelline envelope. . gn Ae RO 1 ei Ne A Ligand to Cortical Granule Lectin 281 Roe 12 30> Fic. 10. Interaction of gp 105 with cortical granule lectins. Cote Westernblots of isolated gp 105 (10 yg, lane 1), of a 1,000xg precipitate of the F layer extract (10 ug, lane 2) and FE (30 yg, lane 3) were treated with gold-conjugated lectins alone (approximately 20 g/ml lectins; A) or with gold-conjugated lectins plus either galactose (1 M; B) or EDTA (1 mM; C). Blotting membranes were further treated with a silver enhancing kit. according to the PA-CrA-Silver method for detecting carbo- hydrates (Fig. 9). In PR1 eggs (Fig. 9A), silver precipitates indicating the location of carbohydrates were present on the cell membrane, cortical granules and glycogen granules. The VEs of these eggs were not stained by the silver, but the PF layer in PR2 eggs (Fig. 9B) and uterine eggs (Fig. 9C) were. The jelly layer in uterine eggs was also stained, but a difference in stainability between the PF layer and the jelly layer is evident, since the silver precipitates on the PF layer are fine whereas those on the jelly layer are coarse and more tightly packed (Fig.9C). In activated eggs (Fig. 9D), the dispersed region of the F layer was stained by the silver but its condensed region not. Interaction of gp 105 with cortical granule lectins Figure 10 shows westernblots of the isolated gp 105, a 1,000Xg precipitate of F layer extract and FE. Gold- conjugated cortical granule lectins stained gp 105 in the first two samples (Fig. 10A). The staining on gp 105 in FE was weak, which may be due to a relatively low amount of gp 105 in FE. gp 100 and gp 30 of the 1,000 xg precipitate were also stained. Galactose and EDTA, which are inhibitors of lectins [16, 28, 29], decreased the staining (Fig. 10B, C). Such results constitute evidence of the occurrence of a lectin-ligand reaction between gp 105 and the cortical granule lectins, thus indicating gp 105 is a ligand molecule of the lectins. D, membrane stained with Coomassie blue. DISCUSSION gp 105, a natural ligand to cortical granule lectins In Xenopus the F layer becomes apparent after fertiliza- tion as an electron-dense layer between the VE* and the innermost jelly layer [25, 31]. Involvement of cortical gra- nule lectins in the reaction forming the F layer was suggested by Wyrick et al. [25] and the localization of the lectins in the F layer was demonstrated by immunoelectron microscopy [30]. However, the identity of the ligand molecule binding to the lectins has been disputed [25, 31]. In the present study gp 105 was isolated from an extract of the F layer. On immunoblot and immunohistochemical test evidence, this glycoprotein is apparently not a component of the jelly. That gp 105 is a constituent of the F layer was suggested by the increase in the relative amount of gp 105 in FEs during the hatching process, an increase which coincides well with the previous ultrastructural finding that the F layer remained substantially unaffected even after the VE* portion of the FE was completely digested [32]. Gold particles indicating the location of gp 105 by immunoelectron micro- scopy were present on the PF layer of uterine eggs, distri- buted in the same area as that comprising the F layer of fertilized eggs according to Grey et al.[5]. Furthermore, binding of gold-conjugated cortical granule lectins to gp 105 on westernblotted membranes demonstrated that lectin- ligand reactions occur. The PF layer is the first extracellular matrix outside the VE met by cortical granule lectins as they emanate from the egg surface and extrude through the VE Fic. 9. Ultrastructural localization of carbohydrates by application of the PA-CrA-Silver method to sections of egg from the pars recta 1 (A), pars recta 2 (B), uterus (C) and uterine egg after activation (D). Silver particles appear on both the pre-fertilization (PF) layer (B and C) and the jelly (J) layer (C and D), but those on the PF layer are fine whereas those on the J layer are coarse and more tightly packed. In activated egg (D), the dispersed region of the fertilization (F) layer is stained by the silver (arrowheads) but the condensed region of the F layer (which is visible as a line marked by the arrows) is not. VE of activated egg. CG, cortical granule; CM, cell membrane; G, glycogen granule; P, pigment granule; VE, vitelline envelope; VE", 282 N. YOSHIZAKI [22]. Then the natural ligand of the lectins seems to be the gp 105 in the PF layer. Resistance to treatment by reducing reagents such as DTT is one of the characteristics of the F layer [23]. The PF layer, however, was dissolved in conjunction with the jelly when treated with mercaptoethanol (see, Wolf et al. [24]). A DTT-stable layer could be produced by activating PR2 eggs whose external surface was covered by the PF layer [31] or by treating them with lectins [28]. The present study demons- trated diminution in both the antigenicity of gp 105 and its stainability by the PA-CrA-Silver method at the time of change from PF to F layer. This diminution is understand- able if the respective reactive sites on gp 105 are being hidden by added cortical granule lectins, as discussed later. Dif- ferential response to DIT between the PF layer and the F layer can be understood along the same lines. The present study suggested that gp 105 is supplied to eggs at the PR2 of the oviduct. Previous studies have shown that an antiserum against secretory granules of the PR2 stained the PF layer [26, 31], suggesting a PR2 origin for PF layer substances. There are two types of secretory cells in the PR2 and PC1, respectively [27]. An intact antiserum raised against gp 105 in the present study stained not only both types of secretory cells in the PR2 but also one type of secretory cell at the ridge in the PC1 (unpublished). Howev- er, when absorbed by the PC1, the antiserum completely lost its ability to stain the cells of either the PC1l or PR2 (unpublished) but retained the ability to stain gp 105. The antiserum may recognize an epitope which is newly produced by an interaction between substances secreted by cells in the PR2. The gp 100 in F layer extract exhibited the same antigeni- city, HRP-PNA stainability and gold-conjugated cortical gra- nule lectin binding as gp 105. It did not appear in the FEs but only in the extract. Since an antiserum raised against gp 105 and absorbed by both the jelly and ovarian homogenates still retained the ability to bind with gp 100 (unpublished), gp 100 is not a contaminant from the jelly, VE, or cellular components of the eggs. gp 100 may be a degraded form of gp 105, since it did not appear when the entire isolation proce- dure was performed in the presence of the protease inhibitors PMSF (1mM) and aprotinin (1 g/ml) (unpublished). Possible agents of degradation are trypsin-like and chymot- rypsin-like enzymes [12, 13], both of which are released extracellularly from their binding sites after activation. The gp 30 in the F-layer extract also exhibited gold-conjugated cortical granule lectin binding but did not show HRP-PNA stainability nor antigenicity like gp 105. Thus gp 30 is diffe- rent from gp 105. The SDS-PAGE profiles of FE in Figure 2 do not show all of the substances seen in those of Gerton and Hedrick [4]. Proteins of 120 kDa and 57 kDa do not appear in this study. The absence of 120-kDa protein is probably due to the difference in the amount of FE loaded, 5 yg in the present study but 25 ~g in Gerton and Hedrick’s. gp 105 migrates on SDS-PAGE very close to gp 112 and they fuse with each other when there are large amounts of FE, so that in this study smaller amounts of FE than usual were chosen in order to separate the two glycoproteins. The 57-kDa protein also failed to appear in this study even though 30 wg FEs were loaded, as seen in Figure 10D. The reason for its absence is not clear yet. It is possible that the 57-kDa protein may have been removed from the FEs during the dejellying process in this study. To clearly separate the jelly from the FEs, this study exploited a Ca-free, 0.05 DB solution (a medium containing reducing agents), whereas Gerton and Hedrick adopted a full-strength DB solution. Further stu- dies are needed to explain why the 57-kDa protein did not appear. Carbohydrates in gp 105 Carbohydrate residues must be demonstrated in order to claim that a particular substance is in fact a ligand. Car- bohydrates were detected histochemically in the PF layer and biochemically by the HRP-PNA method in gp105. The PNA shares carbohydrate-specificity for galactosides with the cortical granule lectins [16, 28, 29]. Apparently, however, there is very little carbohydrate in gp 105, since it was not stained with the PAS method. There seems to be a discrepancy with respect to histo- chemical results: PA-CrA-Silver staining, which detects car- bohydrates ultrastructurally, stained cortical granules and the PF layer in unfertilized eggs but did not stain the F layer in activated eggs. gp 105 is glycosylated, as mentioned above, and so are cortical granule lectins [16]. With present know- ledge, it is difficult to explain the failure in histochemical detection of F-layer carbohydrates, since they are produced by interactions of two biochemically defined glycoproteins. It seems possible that carbohydrate moieties of gp 105 are being bound by lectins in such a way or at such binding sites that the carbohydrate is unavailable for staining in the condensed region of the F layer (see Fig. 9). This region probably corresponds to the area of increased electron densi- ty under conventional heavy metal staining [5, 25, 31], which suggests the presence of some additional substance in the F layer, namely, lectins. Taken together, the increase in elec- tron density and the absence of PA-CrA-Silver staining indicates that some reaction has occurred during fertili- zation; the next studies need to examine exactly how carbohy- drate moieties of the cortical granule lectins behave at the time of binding with gp 105. Ligand molecules in the jelly layer Since cortical granule lectins can react with the jelly, the jelly layer has been proposed as the site of ligand molecules [25]. Using immunoelectrophoretic analyses, Birr and Hed- rick [2] observed three jelly coat ligands bound by cortical granule lectins; two of the three (L-1 and L-2) were sulfated and one (L-3) not. Since L-3 was present in low concentra- tion relative to other components in total jelly solutions and it cross-reacted with anti-envelope sera but not with anti-total jelly sera, they speculated that L-3 is a component of the PF ee —“—“‘“‘(C A Ligand to Cortical Granule Lectin 283 layer. However, neither association of L-3 with the PF layer nor its molecular weight has been established yet. It has already been shown that in PF layer-depleted eggs, secreted lectins produced an electron-dense layer in the space between the outer surface of the VE* and the inner surface of the jelly layer but were dissolved together with the jelly when treated with DTT [31], suggesting that secreted lectins reacted with the jelly at the innermost surface of its layer but did not penetrate deeply into the jelly layer, and that most of the lectins accumulated in the space. Thus L-1 and L-2 (and perhaps L-3), associated with the innermost jelly coat layer, may be ligands to the cortical granule lectins and seem certain to have a role in natural fertilization. The relationship of these substances to gp 105 might warrant investigation. The jelly layer may act as a block to outward diffusion of cortical granule lectins. Limited distribution of secreted lectins was demonstrated in activated eggs immunoelectron microscopically [30]: gold particles indicating the locations of the lectins were present in the perivitelline space, on the VE*, and on the F layer, but not on the jelly. Previous observations on morphological and biochemical changes in the FEs suggested that the F layer is resistant to the hatching enzyme secreted by hatching embryos [32]. A thick F layer might be disadvantageous for embryos to break through. As shown in the present study, the PF layer is compressed by externally loaded jelly layers during jelly deposition around the eggs. The jelly layer may minimize the space in which the F layer will be formed but, by blocking dispersing lectins at its innermost surface, guarantee sufficient lectin-gp 105 interactions for production of a polyspermy block. The present study claims that the F layer is formed by the interaction of gp 105 with cortical granule lectins, but it does not preclude other as yet unrecognized elements of the F layer from candidacy for ligand molecule to the lectins. In particular, immunoelectrophoretically identified ligands of the jelly noted by Birr and Hedrick [2] may also be found in the F layer, for the jelly seems to participate in F layer formation, as discussed above, and the boundary between the jelly layer and F layer is not absolute. That multiple types of secretory cells are involved in producing the extracellular matrix of eggs [27] also suggests the existence of multiple ligand molecules in the F layer. Thus further study is needed to explore other ligand molecules than gp 105 in this layer. ACKNOWLEDGMENTS I wish to thank Dr. H. Kubota of Kyoto University for instruc- tion in immunoelectron microscopy and Ms. M. Lynne Roecklein for reading the manuscript. REFERENCES 1 Adams JC (1981) Heavy metal intensification of DAB-based HRP reaction product. J Histochem Cytochem 29: 775 2 Birr CA, Hedrick JL (1992) Immunoelectophoretic identi- 10 11 12 13 14 15 16 17 18 19 20 21 22 23 fication of jelly coat ligands bound by the cortical granule lectin from Xenopus laevis eggs. Dev Growth Differ 34: 91-98 Chamow SM, Hedrick JL (1986) Subunit structure of a cortical granule lectin involved in the block to polyspermy in Xenopus laevis. FEBS Letters 206: 353-357 Gerton GL, Hedrick JL (1986) The vitelline envelope to fertilization envelope conversion in eggs of Xenopus laevis. Dev Biol 116: 1-7 Grey RD, Wolf DP, Hedrick JL (1974) Formation and struc- ture of the fertilization envelope in Xenopus laevis. Dev Biol 36: 44-61 Grey RD, Working PK, Hedrick JL (1976) Evidence that the fertilization envelope blocks sperm entry in eggs of Xenopus laevis: Interaction of sperm with isolated envelopes. Dev Biol 54: 52-60 Hedrick JL, Nishihara T (1991) Structure and function of the extracellular matrix of anuran eggs. J Electron Microsc Tech 17: 319-335 Higgins RC, Dahmus ME (1979) Rapid visualization of protein bands in preparative SDS-polyacrylamide gels. Anal Biochem 93: 257-260 Kitagaki-Ogawa H, Matsumoto I, Seno N, Takahashi N, Endo S, Arata Y (1986) Characterization of the carbohydrate moiety of Clerodendron trichotomum \ectins. Eur J Biochem 161: 779-7185 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227: 680-685 Larabell C, Chandler DE (1991) Fertilization-induced changes in the vitelline envelope of echinoderm and amphibian eggs: self-assembly of an extracellular matrix. J Electron Microsc Tech 17: 294-318 Lindsay LL, Hedrick JL (1989) Proteases released from Xeno- pus laevis eggs at activation and their role in envelope conver- sion. Dev Biol 135: 202-211 Lindsay LL, Larabell CA, Hedrick JL (1992) Localization of a chymotrypsin-like protease to the perivitelline space of Xenopus laevis eggs. Dev Biol 154: 433-436 Moriya M (1976) Required salt concentration for successful fertilization of Xenopus laevis. J Fac Sci Hokkaido Univ Ser VI 20: 272-276 Nieuwkoop PD, Faber J (1967) Normal Table of Xenopus laevis (Daudin). North-Holland Publ, Amsterdam Nishihara T, Wyrick RE, Working PK, Chen YH, Hedrick JL (1986) Isolation and characterization of a lectin from the cortical granules of Xenopus laevis eggs. Biochemistry 25: 6013-6020 Rambourg A, Hernandez W, Leblond CD (1969) Detection of complex carbohydrates in the Golgi apparatus of rat cells. J Cell Biol 40: 395-414 Slot JW, Geuze HJ (1985) A new method of preparing gold probes for multiple-labeling cytochemistry. Eur J Cell Biol 38: 87-93 Smith JC (1987) A mesoderm-inducing factor is produced by a Xenopus cell line. Development 99: 3-14 Smith PK, Krohn RI, Hermanson GY, Mallia AK, Gartner FH, Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC (1975) Measurement of protein using bicinchoninic acid. Anal Biochem 150: 76-85 Urch UA, Hedrick JL (1981) The hatching enzyme from Xenopus laevis: Limited proteolysis of the fertilization envelope. J Supramol Str Cell Biochem 15: 111-117 Wolf DP (1974) The cortical granule reaction in living eggs of the toad, Xenopus laevis. Dev Biol 36: 62-71 Wolf DP (1974) On the contents of the cortical granules from 284 24 25 26 27 N. YOSHIZAKI Xenopus laevis. Dev Biol 38: 14-29 Wolf DP, Nishihara T, West DM, Wyrick RE, Hedrick JL (1976) Isolation, physicochemical properties, and the macro- molecular composition of the vitelline and fertilization envelope from Xenopus laevis eggs. Biochemistry 15: 3671-3678 Wyrick RE, Nishihara T, Hedrick JL (1974) Agglutination of jelly coat and cortical granule components and the block to polyspermy in the amphibian Xenopus laevis. Proc Natl Acad Sci USA 71: 2067-2071 Yoshizaki N (1984) stration of the pre-fertilization layer in Xenopus eggs. Growth Differ 26: 191-195 Yoshizaki N (1985) Fine structure of oviducal epithelium of Xenopus laevis in relation to its role in secreting egg envelopes. J Morphol 184: 155-169 Immunoelectron microscopic demon- Dev 28 29 30 31 32 Yoshizaki N (1986) Properties of the cortical granule lectin isolated from Xenopus eggs. Dev Growth Differ 28: 275-283 Yoshizaki N (1989) Comparison of two lectins isolated from Xenopus cortical granules. Zool Sci 6: 507-514 Yoshizaki N (1989) Immunoelectron microscopic demon- stration of cortical granule lectins in coelomic, unfertilized and fertilized eggs of Xenopus laevis. Dev Growth Differ 31: 325— 330 Yoshizaki N, Katagiri C (1984) Necessity of oviducal pars recta secretions for the formation of the fertilization layer in Xenopus laevis. Zool Sci 1: 255-264 Yoshizaki N, Yamasaki H (1991) Morphological and bioche- mical changes in the fertilization coat of Xenopus laevis during the hatching process. Zool Sci 8: 303-308 ZOOLOGICAL SCIENCE 11: 285-290 (1994) Localization and Purification of Serum Albumin in the Testis of Xenopus laevis Masauisa Nakamura’, Tomoyo YAMANOBE* and MINORU TAKASE Laboratory for Amphibian Biology, Faculty of Science, Hiroshima University, 1-3-1 Kagamiyama, Higashi-Hiroshima, Hiroshima 724, and *Central Laboratory of Analytical Biochemistry, School of Medicine, Teikyo University, 2-11-1 Kaga, Itabashi-ku, Tokyo 173, Japan ABSTRACT—The distribution of serum albumin is of interest in the Xenopus (X.) laevis testis, since albumin is probably a major protein that binds testosterone (T) in the plasma and interstitial fluid. This study was undertaken to determine the localization and purification of serum albumin in the X. Jaevis testis. The interstitial tissue and spermatogonia immunoreacted strongly with a sheep antiserum raised against X. /aevis albumin. A weak staining was also seen in spermatocytes and early spermatids, but there was no staining in Sertoli cells. In order to clarify whether serum albumin was really localized on the surface of testicular cells in the X. /aevis testis, a membrane-rich fraction was prepared from testes and extracted with 0.6M KCl. The KCl extract was then subjected to gel filtration, ammonium sulfate precipitation and high-performance liquid chromatography (HPLC). A protein with Mr=74 kD was obtained by this procedure and its NH>-terminal amino acid sequence was determined. The sequence of the first 19 amino acids was DTDADXXKXIADVYTALTE, suggesting that this protein was identical to serum albumin (Mr=74kD). When the membrane fraction of blood cells in this animal was handled in the same manner, no appreciable amount of albumin was detected. These results suggest that the 74 kD serum albumin, possibly associated with bound T, may play an important © 1994 Zoological Society of Japan role in the differentiation of germ cells during spermatogenesis of X. laevis testis. INTRODUCTION Sperm formation, spermatogenesis, is the result of a complex process of biochemical and morphological dif- ferentiation of germ cells. Pituitary gonadotropins and ster- oid hormones control spermatogenesis [20-22]. As yet, the stage-specificity of hormonal control of spermatogenesis re- mains unclear. In order to clarify the stage-specificity of steroid hormonal control of this process, immunohistoche- mical studies have been performed in the mammalian testis using antibodies raised against serum albumin, because this protein has a high capacity to bind T [5] and serves as the major protein transporting T in the plasma and interstitial fluids in adult rats [4]. It is probable that albumin acts on Leydig cells and stimulates steroidogenesis of these cells [5]. In fact, Christensen et al. [3] showed under electron- microscopic immunocytochemistry that albumin was localized on the surface of Leydig cells in rat testis, and that im- munoreactivity extended between Sertoli cells as well as around spermatogonia and early spermatocytes, but albumin was not present beyond Sertoli cell junctions. In human testis, albumin was observed in Sertoli cells, secondary sper- matocytes and early spermatids [6, 15]. The precise localiza- tion of albumin within the testis is still controversial. In amphibians, the regulation of spermatogenesis by steroid hormones is not clear except that T may be required for spermatid formation [17]. This study was undertaken to determine the localization of serum albumin in the X. /aevis Accepted February 26, 1994 Received January 12, 1994 " To whom request of reprint should be addressed testis, and also to confirm by purifying this protein from the membrane fraction that albumin is really localized on the surface of testicular cells. MATERIALS AND METHODS Experimental animals Adult male X. laevis (50-70 gm) were used for all the experi- ments. Immunohistochemistry Testes were fixed in Bouin’s solution and further treated accord- ing to conventional histological technique. Sections (approximately 5 ym thick) were cut on a microtome (Yamato), placed on alcohol- washed slides, and warmed on a hot plate for 3hr and then rehydrated in phosphate buffered saline (PBS; pH 7.4) for 10 min. The avidin-biotin-peroxidase complex (ABC) method [8] was used for immunohistochemical stainings using sheep antisera raised against X. laevis albumin (a gift of Dr. D. R. Schéenberg) at a 1: 15000 dilution in PBS. Purification of a 74 kD protein (albumin) To confirm whether albumin was really localized on the surface of testicular cells, albumin was purified from the membrane-rich fraction of X. Jaevis testes. The membrane-rich fraction of testes was prepared by the method of Millette et al. [12]. Testes were removed, wiped with Kimwipes around the tissue, frozen immediate- ly in liquid nitrogen and stored at —80°C until use. Frozen testes were then thawed and homogenized with a glass-Teflon homogenizer in 40 ml of TBS buffer containing 0.16 M NaCl, 3 mM MgCl, 5 mM KCl in 10 mM Tris-HCl (pH 7.4 at 4°C) [12]. The homogenate was centrifuged at 1000 g for 10 min at 4°C to remove large aggregates and debris. The supernatant was used for preparation of plasma membranes by centrifugation on discontinuous sucrose gradients in TBS. Exactly 2.5 ml of the supernatant was mixed with 2.5 ml of 286 M. NaAKAmurRA, T. YAMANOBE AND M.TAKASE 80% sucrose (w/v) to yield 5 ml! of 40% sucrose containing mem- branes. All the 40% sucrose material (5 ml) was layered on top of 2 ml of 45% sucrose (w/v) in TBS in a cellulose nitrate centrifuge tube (Hitachi RPS 40T). Two ml of 30% sucrose (w/v) in TBS were then layered above the 40% sucrose, followed by 1 ml of TBS to fill the tube. Gradients were centrifuged at 125,000 g for 2 hr at 4°C ina Hitachi SCP 85H2 ultracentrifuge equipped with an RPS 40T rotor. Fractionated material (the interface between 30% and 40% sucrose) was collected, diluted~1: 10 in TBS and pelleted at 125,000 g for 40 min at 4°C. To prepare the membrane-rich fraction of X. laevis blood cells the same protocol was used. The membrane-rich frac- tion of testes or blood cells was suspended in 10 ml of 0.6 M KCl, and stirred for 48h at 4°C. Then, insoluble materials were removed by centrifugation at 105,000 g for 1 hat 4°C. The resultant KCl extract was fractionated through Sephadex G-200 (Pharmacia) gel filtration. Proteins were eluted with 0.6 M KCi at a flow rate of 10 ml/h. The effluent was collected in 1.8-ml fractions, and the protein content of each fraction was monitored by absorbance at 280nm. After this, the effluent from the membrane-rich fraction of either testes or blood cells was divided into three fractions. The fraction (designated F2 or F3’, respectively) was dialyzed for 12h at 4°C against 1 liter of saturated ammonium sulfate solution. Precipitates were collected by centrifugation at 105,000 g for 30 min at 4°C and dissolved in 2 ml of 50 mM Tris-HCl (pH 7.4). The sample was applied to a column of Mono Q Sepharose (HR 5/5; Pharmacia) equilibrated with 50 mM Tris-HCl (pH 7.4). Protein concentrations were determined by the method of Peterson [16] using bovine serum albumin as the standard. SDS-PAGE and immunoblot analysis Proteins were added to the SDS sample buffer, heat denatured, and electrophoresed on a 12% acrylamide gel [10]. For immunoblot analysis, nitrocellulose membranes were stained after transfer [23] with a sheep anti-X. /aevis albumin serum at a 15,000 dilution in PBS [13]. NH,-terminal sequence analysis An NH,-terminal amino acid sequence analysis was performed using the 74 kD protein obtained from X. laevis testes. An auto- mated protein sequence analysis was performed on an Applied Biosystems Model 470A gas-liquid phase protein sequencer con- nected on-line to an Applied Biosystems Model 120A HPLC [14]. RESULTS Immunohistochemical studies for localization of albumin The immunohistochemical localization of serum albumin was examined by use of a highly diluted specific antiserum. None of cells was stained when non-immune serum was used (Fig. la). However, a strong staining was observed in the interstitial tissue and spermatogonia, when the sheep anti- serum raised against X. laevis albumin was used (Fig. 1b). A weak staining was also seen in spermatocytes and early spermatids, but not in Sertoli cells (Fig. 1b). Purification of a 74 kD protein (albumin) The membrane-rich fraction from testes was extracted with 0.6 M KCl and then the 0.6 M KCl extract was applied to a Sephadex G-200 column. The effluent was divided into three fractions (Fig. 2a). The last peak was not saved be- Fic..1. Localization of albumin in the X. laevis testis by indirect ABC analysis with sheep antiserum raised against X. laevis albumin. Immunostaining with sheep non-immune serum (a) and with sheep antiserum raised against X. laevis albumin (b). Arrow and arrowheads indicate Sertoli cell and spermatogonia, respec- tively. Sg, spermatogonia; sc, spermatocytes; st, spermatids; SZ, spermatozoa; S, Sertoli cells; it, the interstitial tissue. cause no detectable amounts of proteins was obtained, although it had an absorbance at 280nm. This may be due to free amino acids and/or small peptides since all substances in this peak were dialyzable. When the membrane-rich fraction from blood cells was used instead of that from testes, three peaks appeared in the elution profile from the gel filtration (Fig. 2b). The first three fractions were designated F1, F2 and F3 for the testes, or Fl’, F2’ and F3’ for the blood cells, respectively (see Figs. 2a and 2b). The F2 fraction for the testes was dialyzed against a saturated ammonium sulfate solution and then the precipi- tates were obtained, followed by HPLC. As shown in Figure 3a, a major peak was obtained by the first HPLC. This peak with a dotted area was pooled and dialyzed for 1 hr against 1 liter of 50 mM Tris-HCl (pH 7.4). After dialyzed, the sample was subjected to the second HPLC. When the second HPLC was done, the symmetrical peak with a dotted area containing a 74 kD protein was eluted with 0.25 to 0.35 M NaCl (Fig. 3b). When this peak was analyzed for the Albumin in the Testis of X. laevis al im he il 0.3 h aa BSA Cyt C 4 v me) D2 oe | oe o 4 =! Absorbance at 280nm Fraction Number Absorbance at 280 nm Fraction Number Fic. 2. Gel filtration chromatography on Sephadex G-200 of the 0.6 M KCI extract of X. laevis testes (a) and blood cells (b). A column (1.5 120 cm) was calibrated with standard molecular weight proteins [bovine serum albumin (BSA; Sigma, Mr=68 kD)] and cytochrome c (Cyt C; Miles, Mr=14kD). The void volume (Vo) is indicated with an arrow. heterogeneity of proteins by SDS-PAGE, a 74kD protein was not a major protein in the KCI extract (Fig. 4A; lane c). However, the 74kD protein was a major protein in the precipitate of the fraction F2 obtained by a dialysis against a (M) NaCl Absorbance at 280nm (x10) 0 10 20 30 40 Fraction Number 287 Absorbance at 280 nm (x10) ro) oo NaCl (M) 0) 10 20 Fraction Number 30 Fic. 3. Elution profiles from the first (a) and second (b) HPLC. Proteins in the F2 fraction obtained from the Sephadex G-200 gel filtration chromatography were eluted with 20 ml of a linear gradient of NaCl (0.0—1.0 M) in 50 mM Tris-HCI (pH 7.4) at a flow rate of 2 ml/min. saturated ammonium sulfate solution. After the second HPLC, a very strong band with Mr=74kD and a much weaker band with Mr=68 kD were observed (Fig. 4A; lane g). Based on densitometric tracings of stained gels on SDS-PAGE (Joyce-Loebel Chromatoscan 3), the total amount of the 68 kD protein was <5% of that of the 74 kD protein. In contrast, a 74 kD protein could not be detected 288 M. NAKAMURA, T. YAMANOBE AND M.TAKASE B a b in the precipitate of the F3’ faction from blood cells after dialysis against a saturated ammonium sulfate solution (Fig. 4B;laneb). Yields of each step in the process of purification of the 74 kD protein are summarized in Table 1. The 74kD protein was obtained to a final yield of 0.1%. 74kD> eer Identification of the 74 kD protein In order to identify the 74 kD protein, an NH>-terminal amino acid sequence analysis of this protein was performed. As seen in Fig. 5, the sequence of the first 19 NH>-terminal amino acids of the 74 kD protein was identical, except for 3 unidentified amino acids, to that of the 74 kD X. laevis serum albumin published by Maskaitis et al. [11] and Schorpp et al. [18]. «27kD DISCUSSION This study has clearly shown that serum albumin is present in the interstitial tissue of the X. laevis testis. According to Christensen et al. [3], immunoreactivity of albumin was detected on the surface, but not in the cytoplasm of Leydig cells. It is not clear in this study whether both the cell surface and cytoplasm of cells in the interstitial tissue of the X. laevis testis contain albumin. We need further inves- tigation at an ultrastructural immunocytochemical level to answer this question. However, it seems probable that the Fic. 4. Profiles of protei SDS-PAGE as observed during the : 3 at ; } g Hee: SUA sR B surface of cells in the interstitial tissue [probably steroid purification procedure (A) and the gel filtration (B). (A): Lane a, the homogenate (50 yg); lane b, the membrane hormone (SH)-secreting cells] is associated with albumin, fraction (50 yg); lane c, the KCI extract (5 ug); lanes d-f, the F1 since albumin was purified from the membrane-rich fraction (10 ug), F2 (12 ug), F3 (8 ug) fractions from gel filtration, of X. laevis testes, but not from membranes of blood cells. respectively; lane g, the sample from the second HPLC (2 zg). An arrow and an arrowhead indicate 74 kD and 68 kD proteins, respectively. Spermatogonia also had a strong response to the albumin antibody, and spermatocytes and early spermatids had a weak (B): Lane a, the precipitate in the F2 fraction from testes after response. A question also arises as to whether albumin is dialysis against a saturated ammonium sulfate solution (10 4g/ localized on the surface of these germ cells. Presently, we lane); lane b, the precipitate in the F3’ fraction from blood cells have no direct evidence for this. Immunocytochemical stu- after dialysis a saturated ammonium sulfate solution (8 g/lane). TaBLE 1. Yields of the 74kD protein from X. laevis testes. Total protein Yield (mg) (%) exp. 1 exp. 2 exp. 3 Membrane fraction 23.0 28.0 37.8 100 KCI extract 3.63 4.59 1.23 17.4 Sephadex G-200 0.240 0.320 0.425 1.11 gel filtration HPLC 0.025 0.024 0.038 0.10 Fic. 5. The N-terminal sequence of the 74 kD protein from the X. /aevis testis. Prepeptide Propeptide Mature Protein | | | | 74kD Albumin® MKWITLICLLISSSFIES RILFKR DTDADHHKHIADVYTALTERTFKG:::: This work bitte SEY Guid ChE De Stee BL hs “Data from Schorpp et al. [18]. The leader peptide of the 74kD X. /aevis albumin consists of a hydrophobic sequence of 24 amino acids [11,18]. Note. Regions of identity are noted by an asterisk. X, not determined. Albumin in the Testis of X. laevis 289 dies will answer this question. It is of great interest to note that Sertoli cells did not respond to the antibody. Several investigators have local- ized albumin in Sertoli cells of mammalian testes such as human [6, 15], hamster [9] and rat [3]. It is not clear presently why the immunoreactivity of albumin was not observed in Sertoli cells of X. laevis testis. In the semini- ferous tubules of mammalian testis, Sertoli cells form a barrier, so-called the blood testis-barrier, to retard or exclude many substances in the blood plasma from entrance into the lumen [19, 25]. Most germ cells, except for spermatogonia, reside within the barrier or the adluminal compartment. In anurans, on the other hand, spermatogenesis takes its course in the cysts of the testes. Germ cells develop within groups of ’follicle” cells which are thought to be comparable to the Sertoli cells in the mammalian testis. According to Berg- mann et al. [1], substances like nutrients and hormones in the blood in this species is probably accessible to most developing germ cells. Taking all these findings into consideration, it is not surprising that Sertoli cells had no response to the antibody of albumin. Perhaps, albumin is not associated with Sertoli cells. T may be transported to germ cells from the interstitial space without going via Sertoli cells. Finally, the HPLC sample consisted of two proteins, as judged from the result of SDS-PAGE analysis. One with Mr=74 kD was a very strong band and the other with Mr=68 kD was a very faint band (see Fig. 4A; lane g). Both bands immunoreacted with the antibody of albumin (data not shown). This is not unusual. The frog, X. laevis, has two albumin genes that code for a 74kD and a 68kD serum albumin [11, 18]. In addition, two molecular forms of proteasome [7], calreticulin (a Ca**-binding protein) [24] and prolactin [26] have also been reported in this animal. Two forms of these proteins may have occurred from a duplication of the entire genome in the genus Xenopus [2]. In view of these findings, we must have purified two albumins together, but could not separate one from another by the methods used in this study. One explanation for this may be as follows; albumin exists in two forms that migrate on SDS-PAGE with relative molecular weights of 74 kD and 68 kD, respectively. As the number of amino acids of the two albumins is equivalent (608 residues), the anomalous be- haviour on SDS-PAGE may be due to the glycosylation, which is specific for the 74kD albumin [18]. It might be possible to separate one from another by changing the range of NaCl concentrations on HPLC. As to which albumins are more closely associated with immunoreacted cells remains unclear at the present time. In the serum of X. laevis, the 74kD albumin exists to a much greater extent than the 68 kD albumin (data not shown). It seems, therefore, very likely that the former is more closely associated with the surface of testicular cells. We do not know yet how spermatogenesis in X. /aevis is controlled by T. Nevertheless, it is extremely interesting to note that the developing germ cells and the interstitial tissue (probably SH-secreting cells) are associated with albumin. Consider- ing that serum albumin can bind T, spermatogenesis may be influenced under T with the aid of serum albumin in this species as well as in others. ACKNOWLEDGMENTS We are indebted to Dr. D. R. Schéenberg, Uniformed Services University of the Health Science, for the generous gift of sheep antisera raised against X. /aevis serum albumin. We gratefully acknowledge Dr. S. Tanaka, Gunma University, for helpful advice for the identification of specific cell types in the X. laevis testis. We wish to thank Dr. E. P. Widmaier, Boston University, for his stimulating discussions and criticisms. REFERENCES 1 Bergmann M, Schindelmeiser J, Greven H (1984) The blood- testis barrier in vertebrates having different testicular organiza- tion. Cell Tiss Res 238: 145-150 2 Bisbee CA, Baker MA, Wilson AC (1977) Albumin Physiolo- gy for clawed frogs (Xenopus). Science 195: 785-787 3 Christensen AK, Komorowski TE, Wilson B, Ma S-F, Stevens III. RW (1985) The distribution of serum albumin in rat testis, studied by electron microscope immunocytochemistry on ultrathin frozen sections. Endocrinology 116: 1983-1996 4 Corvol P, Bardin CW (1973) Species distribution of testoster- one binding globulin. Biol Reprod 8: 277-282 5 Ewing LL, Chubb CE, Robaire BR (1976) Macromolecules, steroid binding and testosterone secretion by rabbit testis. Nature 264: 84-86 6 Forti G, Barni T, Vanelli G, Balboni GC, Orlando C, Serio M (1989) Sertoli cell proteins in the human seminiferous tubule. J Steriod Biochem 32: 135-144 7 Fujii G, Tashiro K, Emori Y, Saigo K, Shiokawa K (1993) Molecular cloning of cDNA for two Xenopus proteasome subun- its and their expression in adult tissues. Biochim Biophys Acta 1216: 65-72 8 Hsu S-M, Soban E (1982) Color modification of diaminobenzi- dine (DAB) precipitation by metalic ions and its application for double immunohistochemistry. J Histochem Cytochem 30: 1079-1982 9 Krishna A, Spanel-Borowski K (1990) Albumin localization in the testis of adult golden hamsters by use of immunohistochemis- try. Andrologia 22: 122-128 10 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T,. Nature 227: 680- 685 11 Maskaitis JE, Sargent TD, Smith Jr LH, Pastori RL, Schéenberg DR (1989) Xenopus laevis serum albumin: sequence of the complementary deoxyribonucleic acids encoding the 68- and 74-kilodalton peptides and the regulation of albumin gene expression by thyroid hormone during development. Mol En- docrinol 3: 464-473 12 Millette CF, O’Brien DA, Moulding CT (1980) Isolation of plasma membranes from purified mouse spermatogenic cells. J Cell Sci 43: 279-299 13. Nakamura M, Michikawa Y, Baba T, Okinaga S, Arai K (1992) Calreticulin is present in the acrosome of spermatids of rat testis. Biochem Biophys Res Commun 186: 668-673 14 Nakamura M, Moriya M, Baba T, Michikawa Y, Yamanobe T, Arai K, Okinaga S, Kobayashi T (1993) An endoplasmic reticulum protein, calreticulin, is transported into the acrosome of rat sperm. Exp Cell Res 205: 101-110 15 Orlando C, Casano R, Forti G, Barni T, Vanelli GB, Balboni 290 20 21 M. NAKAMURA, T. YAMANOBE AND M.TAKASE GC, Serio M (1988) protein in human testis and seminal plasma. 83: 687-692 Peterson GL (1977) A simplification of the protein assay method of Lowry et al. which is more generally applicable. Anal Biochem 83: 346-356 Rastogi RK, Iela L, Saxena PK, Chieffi G (1976) The control of spermatogenesis in the green frog, Rana esculenta. J Exp Zool 196: 151-166 Schorpp M, Dobbeling U, Wagner U, Ryffel M (1988) 5’- Flanking and 5’-proximal exon regions of the two Xenopus albumin genes. Deletion analysis of constitutive promoter function. J Mol Biol 199: 83-93 Setchell BP (1967) The blood testicular barrier in sheep. J Physiol (Lond) 189: 63-65 Steinberger E (1971) Hormonal control of mammalian sperma- togenesis. Physiol Rev 51: 1—22 Steinberger E, Duckett GE (1967) Hormonal control of sper- Immunologically reactive albumin-like J Reprod Fertil 22 23 24 De 26 matogenesis. J Reprod Fertil Suppl 2: 75-87 Steinberger E, Steinberger A, Ficher M (1970) Study of sper- matogenesis and steroid metabolism in cultures of mammalian testes. Rec Prog Hormonal Res 26: 547-588 Towbin H, Staehelin T, Gordon J (1979) Electrophoretic transfer of protein from polyacrylamide gels to nitrocellulose sheet: procedure and applications. Proc Natl Acad Sci USA 76: 4350-4354 Treves S, Zorzato F, Pozzan T (1992) Identification of calreti- culin isoforms in the central nervous system. Biochem J 287: 579-581 Waites GMH, Setchell BP (1969) Physiology of the testis, epididymis and scrotum. Adv Reprod Physiol 4: 1-63 Yamashita K, Matsuda K, Hayashi H, Hanaoka Y, Tanaka S, Yamamoto K, Kikuyama S (1993) Isolation and characteriza- tion of two forms of Xenopus prolactin. Gen Comp Endocrinol 9: 307-317 ZOOLOGICAL SCIENCE 11: 291-297 (1994) Spatio-Temporal Pattern of DNA Synthesis Detected by Bromodeoxyuridine Labeling in the Mouse Endometrial Stroma during Decidualization Naosui Onta!?, Takao Mort’, Sencuiro KAWASHIMA!, SHINOBU SAKAMOTO’, HIDESHI KoBAYASHI” ‘Zoological Institute, School of Science, University of Tokyo, Bunkyo-ku, Tokyo 113, Research Laboratory, Zenyaku Kogyo Co., Ltd. Nerima-ku, Tokyo 178, 3Department of Endocrinology, Medical Research Institute, Tokyo Medical and Dental University, Bunkyo-ku, Tokyo 113, Japan ABSTRACT—In order to examine the patterns of proliferation and differentiation of endometrial stromal cells before and during decidualization in pseudopregnant mice, the rate of DNA synthesis was immunocytochemically determined by means of bromodeoxyuridine (BrdU) labeling. On day 4 of pseudopregnancy induced by mating with vasectomized male, both uterine horns were traumatized to induce deciduoma. On day 5 of pseudopregnancy (one day after traumatization), BrdU-labeling index was markedly increased, and the labeled cells were found in almost all parts of endometrial stroma. From day 6 to day 8 of pseudopregnancy (2-4 days after traumatization), the labeling index remained high in the stromal cells of all parts except for the periluminal region. In the endometrial stromal cells in the peripheral region of myometrium, however, the labeling index was maximum on day 8 and decreased remarkably on day 9. In the stromal cells in the periluminal region where deciduomal cells developed, the labeling index was high on day 5 and low on day 6, no labeled cells being found on day 8. There results clearly show that each region of uterine endometrial stroma has a different responsiveness to traumatization, and each region plays a different role in the formation of © 1994 Zoological Society of Japan deciduoma. INTRODUCTION Immunohistochemical detection of bromodeoxyuridine (BrdU), which is a uridine analogue and incorporated selec- tively into the cellular DNA at S-phase of the cell cycle, has been proven useful for the analysis of cell proliferation in place of 3H-thymidine incorporation into replicating cells [5]. Differentiation of the endometrial stromal cells into decidual cells occurs soon after the implantation of blasto- cysts. In mice and rats, however, decidual reaction can be induced artificially in the uteri without blastocysts by mecha- nically scratching the luminal surface [17]. Changes in the structure and function of uterine tissue during decidualization are mainly controlled by the ovarian estrogen and progester- one [2, 11, 13, 14]. Decidualization is a highly regulated process characterized by a variety of events including increase in DNA synthesis [1, 8, 15], changes in vascular permeability [7], and polyploidization and hypertrophy of stromal cells [12, 16]. Therefore, formation of deciduoma has been widely applied as a useful experimental model for the study not only of implantation but also of the mechanisms of cell prolifera- tion and differentiation. BrdU labeling patterns of the cells in uterine tissue were reported in normal cycling and pre- pubertal mice [6]. Because the changes of cell proliferation during de- cidualization as a function of time have not been reported, the present study was designed to examine the spatial and Accepted February 3, 1994 Received November 11, 1993 temporal patterns of DNA synthesis in the mouse uterine stromal cells during decidualization after traumatization. MATERIALS AND METHODS Animals Female mice of the ICR strain purchased from Japan CLEA Inc. (Tokyo, Japan) were used in the present study. They were housed in plastic cages (3-7 mice per cage) under controlled lighting (12—hr light and 12—-hr darkness; lights on at 06:00) and temperature (25+ 0.5°C), and were provided with a commercial diet (CE-7: Japan CLEA) and tap water ad libitum. Induction of deciduoma Virgin female mice at 50-60 days of age were mated with vasectomized males to induce pseudopregnancy. The day when a vaginal plug was found was designated as day 1 of pseudopregnancy. On day 4 of pseudopregnancy, the anti-mesometrial luminal surface in both uterine horns was traumatized by a single scratch with a bent needle. The needle was inserted into the uterine lumen from a small incision made with a small scissors at the posteriar end of uterine horn, adjacent to the uterine cervix, under light nembutal anesthesia [17]. The pseudopregnant mice were killed by cervical dislocation on various days after traumatization (Fig. 1). In order to check the effect of trauma, some pseudopregnant mice without traumatization were killed as controls between 2 and 10 days after mating with vasectomized males. Immediately after autopsy, the uterine horns were removed and weighed. The uterine weight was used as a parameter of decidual reaction. In addition, virgin cycling mice at 50-60 days of age were also killed at varius phases of the estrous cycle and the uterine weights were recorded. ESTROUS CYCLE N. Onta, T. Mort et al. DAYS OF PSEUDOPREGNANCY Ml I iP E 2ueeS 4.45 6 wi 8 9 20 Pa Le IL Ul A A A A A A mating trauma- tization Fic. 1. Experimental schedule. Pseudopregnant mice received a deciduogenic stimulus (traumatization) on day 4 of pseudopreg- nancy and were given a single injection of BrdU on various days (4) after traumatization. M: metestrus, D: diestrus, P: proes- trus, E: estrus. BrdU labeling and immunocytochemistry The pseudopregnant mice received a single intravenous injection of bromodeoxyuridine (BrdU, 30 mg/kg body weight: Amersham, UK) at 24-hr intervals after traumatization. Four hr after BrdU injection, the uteri were fixed in ice cold 10% phosphate-buffered neutral formalin for 5hr at room temperature. The uteri were dehydrated, embedded in paraffin, and the sections were cut at 44m thickness. After deparaffinization, the sections were washed in 0.01M phosphate-buffered saline (PBS, pH7.4) three times and digested with 0.1% trypsin (Sigma) in 0.1% CaCl, (pH 7.8) for 20 min at 37°C. After washing in PBS (15 min, 3 times), endogenous perox- idase activity was blocked by immersing the sections in 0.3% H,O; in methanol for 20 min, followed by washing in PBS (15 min, 3 times). Thereafter, the sections were incubated with monoclonal anti-BrdU antibody containing 10 units/ml nuclease (Amersham) for 1 hr at room temperature and then rinsed in PBS (15 min, 3 times). Final- ly, the sections were incubated with peroxidase-conjugated rabbit anti-mouse IgG for 30 min at room temperature. After washing in PBS (15 min, 3 times), the antibody binding sites were visualized by 0.05% 3, 3’-diaminobenzidine tetrahydrochloride solution. Each incubation was conducted in a moist chamber at room temperature. After immunostaining, the sections were counterstained with 0.1% Kernechtrot in 5% Al,(SO,)3, dehydrated through an ethanol series, cleared in xylene, and mounted. The immunocytochemistry was controlled by sections overlaid with PBS instead of anti-BrdU antibody, which showed no immunoreactivity. Measurement of labeling index In order to examine the BrdU labeling index, two sections which were separated by at least 404m apart were randomly chosen from the middle part of the uterine horn or the middle part of deciduoma in each mouse. Cell counting was carried out in the four regions; anti-mesometrial side of the endometrial stroma (AME), periluminal endometrial stroma (PLE), peripheral endometrium adjacent to the myometrium (PPE), and mesometrial side of the endometrial stroma (MME) (Fig. 2). Total number of BrdU labeled cells was counted out of 1,000 cells each in two sections from the four regions by using Fic. 2. Four regions of uterus for examining BrdU labeling index. L: lumen, M: myometrium, AME: antimesometrial side of endometrial stroma, PLE: periluminal endometrial stroma, PPE: peripheral endometrial stroma adjacent to myometrium, MME: mesometrial side of endometrial stroma. 1200 — e£ D ® 1000 4 : 800 4 5 600 —) Y 400 4 : 200 + nae ee o——__e——e-___,. . 0 facies lel Ina eS Sa re Mi Lipttgiprieiie: Loknit sre 8 9 10 11 12 19 14 45° 46 17 48-19) 20 ESTROUS CYCLE DAY OF PSEUDOPREGNANCY Fic. 3. traumatization. point depicts the mean and SEM of 3-5 mice. Changes in uterine weights in mice during estrous cycle and pseudopregnancy with (-O-) or without (-@-) Deciduoma were induced on the next day of traumatization given on day 4 of pseudopregnancy. M: metestrus, D: diestrus, P: proestrus, E: estrus. Each Decidualization in Mouse Uterus 293 an image processor-analyzer (LUZEX; NIRECO Co. Ltd, Tokyo). The labeled indices were expressed as percentages of labeled cells per 1,000 cells. Statistical analysis The statistical significance of the difference between groups were evaluated by one-way analysis of variance (ANOVA), followed by Duncan’s multiple range test for the uterine weights and labeled indices. RESULTS Changes in uterine weight during decidualization Changes in uterine weights during the estrous cycle and pseudopregnancy before and after traumatization are shown in Figure 3. In normal cycling mice, the uterine weight was lowest at metestrus, followed by an increase during diestrus (P<0.01). The weight reached maximum at proestrus, the value being significantly higher than those in the other phases of estrous cycle (P<0.01). The uterine weight of pseudopregnant mice without traumatization between 2 and 10 days after the mating was almost the same as that of diestrous mice. Traumatization of the uteri on day 4 of pseudopregnancy resulted in the development of deciduoma in the next day. The weight increased rapidly after traumatization, reaching maximum on day 8 of pseudopregnancy, and decreased on day 12 and onward. The weights were significantly higher between days 6 and 12 than the other stages of pseudopregnancy with traumatization (P<0.01, in all comparisons). BrdU labeling index On day 2 of pseudopregnancy before traumatization, BrdU labeled cells were observed in the luminal and glandu- lar epithelia but not in the endometrial stroma. On days 3 and 5 of pseudopregnancy before and one day after trauma- tization, a large number of labeled cells were observed in the endometiral stroma but very rarely in the luminal epithelium (Fig. 4a). On days 7 and 8 of pseudopregnancy given no traumatization, labeled cells were not found in the stroma and appeared again in the luminal epithelium. The BrdU-positive cells on day 5 of pseudopregnancy in mice given traumatization (Fig. 4b) were more numerous compared to those on day 3 of pseudopregnancy (Fig. 4a) and on day 5 of pseudopregnancy without traumatization (data not shown, but regardless of the regions, the indices were less than 1% in mice given no traumatization). Detailed spatio-temporal patterns of BrdU labeling in- dex during decidualization are shown in Figure 5. On day 5 through 8 of pseudopregnancy with traumatization, the in- dices were significantly higher in the endometrial stroma than Fic. 4. Uteri of mice on day 3 (a) and day 5 (b) of pseudopregnancy, one day before and after traumatization, respectively. BrdU-labeled cells (black dots) were visible in the endometrial stroma. Bar: 5004m 294 N. Onta, T. Mori et al. AME 50 (0 Sct te ace u ae T a PPE DS < LW Q 2 50 oO Zz =] a PLE <= — Os 5 S| MME 50 | OQ) a Se | T oo aro 3 5 6 7 8 9 DAY OF PSEUDOPREGNANCY Fic. 5. Changes of BrdU lebeling index in the endometrial stromal cells during decidualization. Each point depicts the mean and SEM of 3-5 mice. AME: antimesometrial side of endometrial stroma, PPE: peripheral endometrial stroma adjacent to myometrium, PLE: periluminal endometrial stroma, MME: mesometrial side of endometrial stroma. On days 3 and 5, as PPE and PLE could not be counted separately, pooled data are shown. those on day 3 at 0.01 level, except for PLE (Figs. 5-7). On day 6 of pseudopregnancy with traumatization, the plump cells with large nuclei over 25m in diameter appeared in the PLE and AME (Fig. 8). In PLE, the percentage of labeled stromal cells on day 5 was significantly higher than that in the other stages, respectively (P<0.01). The percentage re- duced rapidly on day 6 and BrdU labeling was no more detected on day 8 (Fig.5). On day 8, labeling indices tended to be lower in the regions of the endometrium except for PPE. On day 9, all labeling indices were almost the same as those on day 3 except for MME (Figs. 5 and 9). Many degenerating cells with pyknotic nuclei appeared on days 8 and 9 of psudopregnancy with traumatization. DISCUSSION Cell proliferation and differentiation during decidualiza- tion in pregnant [3] or pseudopregnant rodents [4, 9, 17] have extensively been studied. It is well known that the dif- ferentiation of endometrial stromal cells to decidual cells occurs in response to the implantation of blastocysts or traumatization of artificial stimuli. In mice, the sensitivity of the uterus to a deciduogenic stimulus is known to be the Fic. 6. Uterus of a mouse on day 6 of pseudopregnancy 2 days after traumatization. BrdU-labled cells were present in almost all regions of the endometrial stroma except for PLE. Bar: 500um highest on day 4 of pseudopregnancy [2, 10]. In the present study, the uterine weight increased immediately after trauma- tization on day 4 and reached maximum on day 10 of pseudopregnancy. The weight markedly decreased on day 12 and returned to the normal diestrous level on day 18. These findings accord well with the previous results in rats [17]. In the present study, DNA synthesis was detectable by the presence of BrdU-labeled cells in the luminal epithelial cells on day 2 of pseudopregnancy. On day 3 of pseudopreg- nancy, some stromal cells in the endometrium began to show DNA synhesis. If deciduogenic stimuli were not given to the uterus, the activity of DNA synthesis in the stromal cells decreased within a few days. These findings may reflect that the stromal cells are ready to respond to deciduogenic stimuli on day 3 of pseudopregnancy. On day 5 of pseudopregnan- cy with traumatization, the labeled cells were present exten- sively and evenly in all the four regions. of endometrial stroma. Thereafter, BrdU-labeled cells greatly decreased in the PLE on days 6 and 7 of pseudopregnancy. Ledford et al. [8] have stated that in mice rapid cell proliferation begins approximately 30 hr after deciduogenic stimulation and con- tinued for 72 hr in the endometrial stroma. After the initia- tion of decidualization, however, a population of stromal cells Decidualization in Mouse Uterus 295 Fic. 7. Uterus of a mouse on day 7 of pseudopregnancy 3 days after traumatization. visible in PLE. Bar: 500um Fic. 8. Antimesometrial side of endometrial stroma in a mouse on day 6 of pseudopregnancy 2 days after traumatization. Plump cells with large nuclei (arrowheads) appeared. Bar: 1004m Only a few BrdU-labeled cells were is known to synthesize DNA and differentiate into polyploid decidual cells without cell division [1]. Deciduomal cells called plump cells in the present study are distributed exclu- sively in the periluminal part of endometrial stroma where the implantation normally ccurs. Thus, it seems likely that the proliferation of deciduomal cells ceases and differentiation begins on day 6 or 7 of pseudopregnancy (2-3 days after traumatization/implantation). During decidualization, a remarkable rise of DNA synth- esis in peripheral endometrial stroma adjacent to the myometrium (PPE) occurred between day 6 and day 8 of pseudopregnancy. Proliferated stromal cells in this region may contribute to the reconstruction of endometrial tissue after the regression of preformed deciduomal tissue. It is known that the life span of the rat deciduoma is limited and frequent cell death occurs on day 9 of pseudopregnancy [14]. The present findings clearly show that regression of de- ciduoma begins in most parts of the endometrium from day 8 of pseudopregnancy, because many degenerated cells were encountered on days 8 and 9 of pseudopregnancy. 296 N. Ounta, T. Mort et al. faut atlas i ee Fic. 9. Uterus of a mouse on day 9 of pseudopregnancy 5 days after traumatization. The number of BrdU-labeled cells became decreased. Bar: 500um The present findings demonstrated that activity of DNA synthesis in the endometrium varies during decidualization in several regions of the endometrial stroma. This implies that each region has a different responsiveness to traumatization and that it plays a different role during the development of deciduoma. ACKNOWLEDGMENTS The authors would like to express our cordial thanks to Dr. M. K. Park for his advice and encouragement during this study. This study was supported by Grants-in-Aid for Scientific Research from the inistry of Education, Science and Culture, Japan to T. M., S. K. and S. S., respectively and a Research Grant from Zenyaku Kogyo, Ltd. to S. K. REFERENCES 1 Das RM, Martin L (1978) Uterine DNA synthesis and cell proliferation during early decidualization induced by oil in mice. J Reprod Fert 53: 125-128 2 Finn CA (1966) Endocrine control of endometrial sensitivity during the induction of the decidual cell reaction in the mouse. J Endocr 36: 239-248 3 Finn CA, Martin L (1967) Patterns of cell division in the mouse uterus during early pregnancy. J Endocr 39: 593-597 4 Galassi L (1968) Autoradiographic study of the decidual cell reaction in the rat. Dev Biol 17: 75-84 5 Gratzner HG (1982) Monoclonal antibody to 5-bromo- and 10 11 Decidualization in Mouse Uterus 297 5-iododeoxyuridine: a new reagent for detection of DNA re- plication. Science 218: 474-475 Hanazono M, Yoshiki A, Ota K, Kitoh J, Kusakabe M (1990) DNA replication in uterine cells of adult and prepubertal mice under normal and hormonally stimulated conditions detected by bromodeoxyuridine labeling method. Endocrinol Japon 37: 183-191 Kennedy TG (1979) Prostaglandins and increased endometiral vascular permeability resulting from the application of an arti- ficial stimulus to the uterus of the rat sensitized for the decidual cell reaction. Biol Reprod 20: 560-566 Ledford BE, Rankin JC, Froble VL, Serra MJ, Markwald RR, Baggett B (1978) The decidual cell reaction in the mouse uterus : DNA synthesis and autoradiographic analysis of respon- sive cells. Biol Reprod 18: 506-509 Marcus GJ (1974) Mitosis in the rat uterus during the estrous cycle, early pregnancy, and early pseudopregnancy. Biol Re- prod 10: 447-452 Martin L, Finn CA (1968) Hormonal regulation of cell division in epithelial and connective tissues of the mouse uterus. J Endocr 41: 363-371 Moulton BC, Blaha GC (1978) Separation of deciduomal cells 12 13 14 15 16 17 by velocity sedimentation at unit gravity. 147 Moulton BC, Koening BB (1984) Uterine deoxyribonucleic acid synthesis during preimplantation in precursors of stromal cell differentiation during decidualization. Endocrinology 115: 1302-1307 Ohta Y (1991) Deciduoma formation in pseudopregnant rats bearing pituitary grafts. Zool Sci 8: 75-80 O’shea JD, Kleinfeld RG, Morrow HA (1983) Urtrastructure of decidualization in the pseudopregnant rat. Am J Anat 166: 271-298 Tachi C, Tachi S, Lindner HR (1972) Modification by prog- esterone of oestradiol-induced cell proliferation, RNA synthesis and oestradiol distribution in the rat uterus. J Reprod Fertil 31: 59-76 Takewaki K (1969) Formation of deciduomata in immature rats with luteinized ovaries. Annot Zool Japon 42: 126-132 Velardo JT, Dawson AB, Olsen AG, Hisaw FL (1953) Sequ- ence of histological changes in the uterus and vagina of the rat during prolongationof pseudopregnancy associated with the pre- sence of deciduomata. Am J Anat 93: 273-305 Biol Reprod 18: 141- ZOOLOGICAL SCIENCE 11: 299-303 (1994) Endocrine Control of Cartilage Growth in Coho Salmon: GH Influence in Vivo on the Response to IGF-I in Vitro PETER I. TSAI, STEFFEN S. MADSEN!, STEPHEN D. McCormick and Howarp A. BERN® Department of Integrative Biology, Bodega Marine Laboratory and Cancer Research Laboratory, University of California, Berkeley, California 94720, USA ABSTRACT—Ceratobranchial cartilages from coho salmon (Oncorhynchus kisutch) parr, injected with growth hormone (GH) at 4ug/g body weight or with saline, were sampled monthly from February to July. Thymidine and sulfate uptakes by cartilages were determined as measures of DNA and chondroitin sulfate synthesis, respectively. Cartilages were incubated with IGF-I at 0.01, 0.1 and 1ug/ml to examine the in vitro response to this hormone. GH injection increased cartilage thymidine and sulfate uptakes at least four-fold in all experiments. IGF-I treatment in vitro further increased sulfate but not thymidine uptake in cartilages from GH-injected coho and increased uptake of both in cartilages from © 1994 Zoological Society of Japan saline-treated coho. However, the IGF-stimulated uptakes were still significantly below the uptakes in cartilages from GH-injected coho. The dual effector hypothesis of GH action [12] in mammals is supported at least in part in teleost fishes by the observation that addition of IGF-I in vitro was not equivalent to injection of GH in vivo. INTRODUCTION The endocrine control of cartilage growth has only recently been examined in teleost fish [see 1, 2, 20]. Studies on the Japanese eel, Anguilla japonica, by Duan and Inui [8, 9] have shown that the stimulatory action of GH on sulfate uptake by cartilage is indirect. Duan and Hirano [6, 7] later showed that sulfate uptake by eel cartilage is stimulated by mammalian IGF-I and raised the possibility of regulation by a similar principle in teleosts. McCormick et al. [14] and Gray and Kelley [10] have subsequently shown that mammalian IGF-I stimulated sulfate incorporation in vitro in cartilages from coho salmon (O. kisutch) and goby (Gillichthys mirabi- lis), respectively. These observations are consistent with the somatomedin hypothesis [4]. In anadromous salmonids, smoltification is a period during which the fish undergoes various physiological changes, many of which are cued by the endocrine system. Endogenous levels of growth hormone, prolactin, thyroid hormones and cortisol change in a distinctive pattern in coho salmon (Oncorhynchus kisutch) undergoing smoltification [19]. In the period when GH levels in the coho are increas- ing [16, 19], cartilage growth rate would be expected to increase due to increased liver-derived IGF-I in the circula- tion [3, 18] and to sensitization of the cartilage to IGF-I by GH [12]. Although injection of GH leads to transient Accepted February 3, 1994 Received October 12, 1993 ' Present address: Institute of Biology, Odense University, Campus- vej 55, DK-5230 Odense M, Denmark 2 Present address: Anadromous Fish Research Center, National Biological Survey, 1 Migratory Way, P. O. Box 796, Turners Falls, Massachusetts 01376, USA 3 To whom correspondence should be directed. down-regulation of liver GH receptors [10, 11, 15, 17] in several teleost species, increased expression of IGF-I mRNA in the liver was observed in coho salmon [5]. Injection of GH, comparable to natural increases in endogenous GH, may thus stimulate cartilage growth. The purpose of this study is to examine further the effect of GH on cartilage growth and its potential interaction with IGF-I. Jn vivo GH action and in vitro IGF-I action on ceratobranchial cartilage in coho salmon were judged by determining thymidine and sulfate incorporation. Experi- ments were done repeatedly during the period of parr-smolt transformation to detect possible developmental or seasonal changes in response to GH in vivo and to IGF-I in vitro. MATERIALS AND METHODS Animals Coho salmon (Oncorhynchus kisutch) parr (10-20 g) were obtained from Iron Gate Hatchery, California Department of Fish and Game, in December 1991. They were maintained outdoors at Bodega Marine Laboratory at 12-14°C in a 2000-liter concrete raceway supplied with filtered fresh water and were fed twice daily with Oregon Moist Pellets (Moore-Clarke, LaConner, WA) at a ration of 2% body wt/day. Injections The fish for each monthly experiment were randomly separated into two groups: GH-injected (NIADDK-oGH-15 at 4ug/g body wt; n=10) and saline-injected (n=10). The oGH was dissolved in 0.01 N NaOH followed by saline solution to yield a final concentration of 2ug/ 1 solution (with pH less than 9); the same volume of 0.01 N NaOH was added to the saline used for injecting controls. Fish were injected with 21 solution/g body wt on alternate days (total of 4 injections). 300 P. I. Tsar, S. S. MADSEN et al. Sampling Cartilage samples were taken monthly from the above groups from February to July. Fish remained unfed for 7 days before sampling in an attempt to increase the sensitivity of their cartilage to IGF-I [13]. Fish were killed by a blow to the head 24 hr after the last injection. Ceratobranchial cartilages were dissected from the bone of the first three pairs of gill arches of each fish under a dissecting microscope and placed in a pre-culture medium: Minimum Essential Medium (MEM) with Hanks’ salts, penicillin (100 U/ml) and streptomycin (100ug/ml). Randomly-selected cartilages (with an average dry wt of 53+15.6yg in February to 101 +27.2g in July) from each fish were then placed into wells (24-well plate, Falcon 3047) for different treatments (n=7-10 for each treatment): basal (untreated); non-specific (cartilages frozen at —80°C to measure non-specific thymidine and sulfate uptake); recombinant bovine IGF-I (rbIGF-I; a gift from Monsanto, St. Louis, MO, U.S.A.) at 0.01, 0.1 and 1yg/ml. Each well contained 1 ml culture medium: MEM with Earle’s salts, bovine serum albumin (BSA; 25yg/ml), penicillin (50 U/ml), streptomycin (50ug/ml), SO, (1Ci/ml) and 3H-thymidine (24Ci/ml). The cartilages were then incubated in a chamber filled with 95% O,/5% CO, at 14°C for 48 hr. During this period, the ceratobranchial cartilage incorporated radioactive sulfate into chondroitin sulfate and radioactive thymidine into DNA. The experiment was terminated by freezing at —80°C. Cartilages were then soaked in cold Na,SO, twice and rinsed with distilled water 3 times in order to eliminate residual unincorporated radioactive sulfate and thymidine. The cartilages were dried in an oven at 60°C and weighed to the nearest wg. Each cartilage was then placed in a scintillation vial containing 0.5 ml 99% formic acid which dissolved the cartilage, thereby releasing the radioactivity (*°SO,+7°H- thymidine) into the acid. Liquid scintillation fluid (4.5 ml) was added to each vial. *°S and °H radioactivities in dpm were deter- mined by a dual-label (7H and *°S) program in a Beckman 5000 scintillation counter. The dpm count was normalized for each cartilage weight to yield dpm/ yg. Statistical analysis Two-way analysis of variance (ANOVA) was used to test for significance of GH injection over time. All other statistical com- parisons were done by one-way ANOVA followed by Newman-Keuls analysis for post-hoc comparisons of factor means. Regression analyses and ANOVA were conducted using the Crisp statistical program (CRUNCH, Berkeley, CA). All groups comprised 7-10 fish, and P<0.05 was considered significant. RESULTS Body weight and smoltification Mean body weight increased linearly (with a slight decrease in June) from 18 g in February to 39 g in July (data not shown). Signs of smoltification (loss of parr marks, silvering of scales, darkening at edge of fins, and increased condition factor) were most evident in May. Thymidine and sulfate uptakes Thymidine uptake (Fig. 1) by the cartilages in GH- injected coho was 4 to 16 times higher than in saline-injected coho throughout the study period (two-way ANOVA, P< 0.0001). From February to June, levels of thymidine uptake d thymidine uptake (dpm/yg) IIS NO nN ~ nN ao w ~ wo > ~ —_ p-S on ~ J o ~ = o “I ~ ow ui date Fic. 1. Basal thymidine uptake (dpm/g cartilage) by ceratobran- chial cartilages from GH- and saline-injected coho salmon (Oncorhynchus kisutch) sampled monthly from February to July. Fish were given injections on alternate days (total of 4 injections) and sampied 1 day after the last injection. Hatched and clear boxes represent GH- and saline-injected fish, respec- tively. Values are mean+SEM (n=7-10). Values with shared letters are not significantly different (P >0.05). by cartilage from GH-injected coho showed a decreasing trend without statistical significance, averaging 10 dpm/ jg from February to May and dropping to 4.4 dpm/ zg in June. Thymidine uptake increased to 15.7 dpm/g in July (P<0.05 compared to the uptake in March-June). Thymidine uptake in saline-injected coho averaged 1.6 dpm/g from February to May, and dropped (P<0.05) to 0.8 dpm/g in June and July. Sulfate uptake (Fig. 2) by the cartilages in GH-injected 10 sulfate uptake (dpm/pg) 2/25 3/9 4/14 S/7 6/16 7/15 date Fic. 2. Basal sulfate uptake (dpm/g cartilage) by ceratobranchial cartilages from GH- and_ saline-injected coho salmon (Oncorhynchus kisutch) sampled monthly from February to July. Fish were given injections on alternate days (total of 4 injections) and sampled 1 day after the last injection. Hatched and clear boxes represent GH- and saline-injected fish, respec- tively. Values are meant+SEM (n=7-10). Values with shared letters are not significantly different (P >0.05). GH/IGF-I Influence on Coho Cartilage TABLE 1. Monthly measurements from February-July of thymidine uptake (dpm/sg) by cartilages from GH- and saline-injected (SAL) coho in response to IGF-I in vitro IGF-I (ug/ml) 0 0.01 0.1 1 Feb. GH 11.5+2.2 (8) 11.1+1.6 (8) 17.2+2.9 (9) 18.7+2.0 (9) SAL 1.5+0.4 (8) 2.1+0.4 (9) 2.6+0.3 (9) 3.2+0.4 (10)* Mar GH 8.141.3 (9) 6.7+1.8 (8) 8.5+1.8 (9) 7.3+1.0 (9) SAL 1.7+0.2 (10) 1.5+0.2 (8) 1.9+0.2 (10) 2.6+0.2 (10)* Apr. GH 8.2+1.9 (9) 6.6+1.1 (8) 11.3+1.8 Q) 11.3+1.8 Q) SAL 1.8+0.2 (8) 2.1+0.2 (9) 2.9+0.3 (7)* 3.0+0.2 (8)* May GH 9.3+2.2 (9) 11.8+2.2 (10) 22.7+4.7 (10)* 16.9+3.4 (10) SAL 1.3+0.2 (9) 2.4+0.3 (10)* 3.0+0.2 (10)* 2.9+0.3 (10)* June GH 4.4+0.5 (7) 4.8+0.5 (8) 6.2+0.5 (8) 6.2+0.6 (9) SAL 0.7+0.1 (8) 1.9+0.3 (8)* 1.9+0.3 (8)* 2.1+0.2 (8)* July GH 15.8+2.2 (7) 21.8+2.8 (7) 21.2+4.3 (7) 18.3+2.3 (7) SAL 0.9+0.1 (7) 1.9+0.4 (7)* 2.3+0.4 (7)* 2.2+0.3 (7)* data expressed as *H-thymidine dpm/yg+SEM (N) * P<0.05 over basal (0) uptake TasBLeE2. Monthly measurements from February-July of sulfate uptake (dpm/g) by cartilages from GH- and saline-injected (SAL) coho in response to IGF-I in vitro IGF-I (ug/ml) 0 0.01 0.1 1 Feb. GH 9.2+1.4 (8) 9.5+0.9 (9) 14.4+1.8 (9)* 14.9+1.0 (9)* SAL 0.8+0.2 (8) 1.8+0.3 (9)* 2.7+0.3 (9)* 3.1+0.2 (40)* Mar GH 7.5+1.4 (9) 6.3+1.2 (8) 8.2+1.4 (9) 8.2+1.3 (9) SAL 1.5+0.2 (10) 1.4+0.2 (10) 1.9+0.2 (10) 2.4+0.3 (10)* Apr GH 7.0+1.3 (9) 6.30.8 (8) 10.8+1.2 (9)* 10.5+1.2 (9)* SAL 1.4+0.1 (8) 1.5+0.1 (9) 2.7£0.2 (7)* 2.6+0.2 (8)* May GH 7.2+1.4 (9) 11.1+1.6 (10) 17.5+2.8 (10)* 14.1+2.0 (10)* SAL 1.4+0.2 (9) 2.8+0.3 (10)* 3.4+0.2 (10)* 3.3+0.3 (10)* June GH 5.4+0.7 (7) 6.10.5 (8) 7.1+0.5 (8) 7.7£0.9 (9)* SAL 0.9+0.1 (8) 2.4+0.4 (8)* 2.2+0.3 (8)* 2.2+0.2 (8)* July GH 4.9+0.5 (7) 7.8+0.7 (7)* 6.4+0.8 (7) 6.7+0.7 (7) SAL 0.4+0.1 (7) 1.0+0.2 (7)* 1.2+0.2 (7)* 1.1+0.2 (7)* data expressed as SO, dpm/yzg+SEM (N) * P<0.05 over basal (0) uptake 302 P. I. Tsar, S. S. MADSEN et al. coho was 5 to 11 times higher than in saline-injected coho throughout the study period (two-way ANOVA, P<0.0001). Sulfate uptake by cartilages in GH-injected coho showed a decreasing trend without statistical significance, averaging 8 dpm/ yg from February to May and dropping to 5 dpm/g in June and July. Uptake in saline-injected coho was 0.8 dpm/ yg in February, then increased (P<0.05) to an average of 1.4 dpm/g from March to May, and dropped (P<0.05) to 0.9 and 0.4 dpm/g in June and July, respectively. Thymidine and sulfate uptakes were positively correlated in both GH- treated fish (r=0.65, P<0.001) and in saline-treated fish (r= 0.68, P< 0.001). Cartilage response to IGF-I in vitro Cartilages from GH- and saline-treated fish in each month were tested for their response to IGF-I in vitro; the results are presented in Table 1 (thymidine uptake) and Table 2 (sulfate uptake). In thymidine uptake (see Table 1), cartilages from GH-treated fish did not respond significantly to further stimulation by IGF-I in vitro at 0.01, 0.1 and 1yg/ml; an exception was in May, when cartilage treated with IGF-I in vitro at 0.14g/ml showed an increase over the basal uptake (P <0.05). Cartilages from the saline-treated group responded to in vitro IGF-I at 1yg/ml in February and March (P<0.05), then to both 0.1 and 1yg/ml IGF-I in April (P<0.05), and to all IGF-I doses from May to July (P<0.05). However, cartilages did not show a dose-dependent response to IGF-I over the doses tested: uptake generally plateaued with in- creasing doses of IGF-I after the initial or smallest dose that elicited a response. In sulfate uptake (see Table 2), cartilages from GH- treated fish responded to IGF-I in vitro with increases in all months except March. Cartilages sampled in February, April and May responded to IGF-I in vitro at 0.1 and 1yg/ml (P<0.05). A dose-dependent response to IGF-I was not found, as stimulated sulfate uptake plateaued after 0.1g¢/ml. Cartilages in June and July only responded to 1yg/ml and 0.014g/ml IGF-I, respectively. Cartilages from the saline- treated group responded to in vitro IGF-I in all months. Cartilages sampled in February, May, June and July all responded to in vitro IGF-I at 0.01, 0.1 and 1ug/ml. Carti- lages in March only showed stimulated sulfate uptake at 1ug/ ml IGF-I, whereas cartilages in April responded to IGF-I at 0.1 and lvg/ml. A dose-dependent response to IGF-I again was not found, as stimulated sulfate uptake usually plateaued after the initial or smallest dose that elicited a response. DISCUSSION As seen in Figures 1 and 2, GH injection markedly increased thymidine and sulfate uptakes, and the two para- meters are strongly correlated (r=0.65, P<0.001). This indicates that thymidine and sulfate uptakes are generally coupled, even after GH stimulation. Cartilages from saline-injected fish showed decreased thymidine uptake (Fig. 1) and decreased sulfate uptake (Fig. 2) in June and July. A marked decline in Na* , K*-ATPase activity (an indicator of hypoosmoregulatory activity) also occurred in these fish in June and July [13], which may indicate the end of the smoltification period. Cartilages from GH-injected fish also showed a non-significant trend of decreasing thymidine (Fig. 1) and sulfate uptakes (Fig. 2) from February to June and from February to July, respective- ly. Injected GH might have compensated for the expected decrease in endogenous GH (plasma GH levels were not measured), resulting in a lack of significant decrease in both thymidine and sulfate uptakes. A significant increase in GH-stimulated thymidine uptake (P<0.05) was, however, observed in July. This is contrary to the sulfate uptake which stayed low. One explanation may be that the carti- lages undergo a new cycle of chondrocyte proliferation at this time. A major increase in thymidine uptake indicating mitotic activity in prechondrocytes/chondrocytes resulted from GH injection in July; this may have led to increased sulfate uptake by maturing chondrocytes at a later date, but this was not examined. Although GH injection consistently increased thymidine and sulfate uptakes, it did not result in a consistent sensitiza- tion to IGF-I in vitro as judged by thymidine uptake. The dual effector hypothesis [12] predicts that GH would increase serum levels of IGF-I and increase cartilage sensitivity to IGF-I. Thus, injected GH may have maximally stimulated the mitotic activity of chondrocytes in vivo so that further stimulation by IGF-I in vitro was not seen. This is in contrast to the findings which showed that priming of gill Na*, K*t-ATPase resulted from either endogenous or exoge- nous GH, so that further stimulation by IGF-I in vitro was possible [13]. As chondrocytes can also respond to IGF-I by synthesizing chondroitin sulfate, further stimulation of sulfate uptake by IGF-I in vitro was still possible. Thus, consistent IGF-I stimulation of sulfate uptake was observed in cartilages from GH-injected coho (March was the only exception). Although cartilages from the saline-injected group con- sistently responded to IGF-I in vitro with stimulated thymi- dine and sulfate uptakes, the stimulated uptakes did not approach the basal uptake seen in the GH-injected group (Tables 1 and 2). The observation that GH in vitro at lug/ ml did not increase thymidine or sulfate uptake [14; Tsai, unpublished] suggested that GH has no direct effect on cartilage growth. GH injection may thus act by increasing endogenous IGF-I levels and/or by sensitizing the cartilage cells to hepatic and/or local IGF-I. The dual effector hypothesis of GH action [12] is supported by the observation that no dose of IGF-I alone in vitro was able to parallel the effects of GH injection in vivo. However, the organ-culture system used did not allow testing of IGF-I in vitro in the presence of other serum factors, including IGF-binding pro- teins. Such factors may modify the responsiveness of carti- lage to stimulation by IGF-I. Furthermore, the exposure time of cartilages to IGF-I in vitro for 48 hr (maximal sulfate incorporation by eel and salmon cartilage occurs between 24 GH/IGF-I Influence on Coho Cartilage 303 and 48 hr [8, 9, 14]) was significantly less than the exposure in vivo to GH, which was given as 4 injections during a 9-day period (these fish were the same as those used by Madsen and Bern [12]). As the cartilages seemed to have responded maximally to GH injection and also to IGF-I addition in vitro, no signif- icant seasonal change could be discerned. These studies support the relevance of the dual effector theory of GH action [12] to teleost cartilage growth: injection of GH in vivo induced consistently higher thymidine and sulfate uptakes by cartilage than were seen in the control cartilages exposed to IGF-I in vitro. ACKNOWLEDGMENTS P. I. Tsai was a California Sea Grant trainee in 1991-2. S.S. Madsen was a postdoctoral fellow of the Carlsberg Foundation (Denmark). We would like to thank Prof. Charles S. Nicoll for his critical comments on this work, Dr. Richard S. Nishioka for his review of the manuscript, Dr. Elisabeth S. Gray and Richard J. Lin for aid in the statistical analysis, and Kimmakone Siharath, Robert Tsai and Jeanette Endersen for assist in the experiments. This research was funded by a grant from the National Sea Grant College Program, National Oceanic and Atmospheric Administration, U. S. Department of Commerce, under grant number NA89AA-D-SG 138, project R/F-145, through the California Sea Grant College, and in part by the California State Resources Agency. The views expressed herein are those of the authors and do not necessarily reflect the views of NOAA or any of its sub-agencies. The U. S. Government is authorized to reproduce ‘and distribute for gov- ernmental purposes. We are grateful to Zenyaku Kogyo Co. of Tokyo for additional support, to NIH and the National Pituitary Program (Baltimore, MD) for the ovine growth hormone and to Monsanto Co. (St. Louis, MO) for the recombinant bovine IGF-I. REFERENCES 1 Bern HA, McCormick SD, Kelly KM, Gray ES, Nishioka RS, Madsen SS, Tsai PI (1991) Insulin-like growth factors “under water”: Role in growth and function of fish and other poiki- lothermic vertebrates. In “Modern Concepts of Insulin-like Growth Factors” Ed by EM Spencer, Elsevier, New York, pp 85-96 2 Bern HA, Nishioka, RS (1993) Aspects of salmonid endocri- nology: the known and the unknown. Bull Fac Fish Hokkaido Univ 44: 55-67 3 Cao O-P, Duguay S, Plisetskaya E, Steiner DF, Chan SJ (1990) Nucleotide sequence and growth hormone-regulated expression of salmon insulin-like growth factor! mRNA. Mol Endocrinol 3: 2005-2010 4 Daughaday WH, Phillips LS, Herington AC (1975) Measure- ment of somatomedin by cartilage in-vitro. Methods in En- zymology 37: 93-109 5 Duan C, Duguay SJ, Plisetskaya EM (1993) Insulin-like growth factor I (IGF-I) mRNA expression in coho salmon, Oncorhynchus kisutch: Tissue distribution and effects of growth 13 14 15 16 17 18 19 20 hormone/prolactin family proteins. Fish Physiol Biochem 11: 371-379 Duan C, Hirano T (1990) Stimulation of *°S-sulfate uptake by mammalian insulin-like growth factor I and II in cultured cartilages of the Japanese eel, Anguilla japonica. J Exp Zool 256: 347-350 Duan C, Hirano T (1992) Effects of insulin-like growth factor-I and insulin on the in vitro uptake of sulphate by eel branchial cartilage: Evidence for the presence of independent hepatic and pancreatic sulphation factors. J Endocrinol 133: 211-219 Duan C, Inui Y (1990) Effects of recombinant eel growth hormone on the uptake of [*°S]sulfate by ceratobranchial carti- lages of the Japanese eel, Anguilla japonica. Gen Comp Endo- crinol 79: 320-325 Duan C, Inui Y (1990) Evidences for the presence of a somatomedin-like plasma factor(s) in the Japanese eel, Anguilla Japonica. Gen Comp Endocrinol 79: 326-331 Gray E, Kelley KM (1991) Growth regulation in the gobiid teleost, Gillichthys mirabilis: Roles of growth hormone, hepatic growth hormone receptors and insulin-like growth factor I. J Endocrinol 130: 57-66 Gray E, Kelley KM, Law S, Tsai R, Young G, Bern HA (1992) Regulation of hepatic growth hormone receptors in coho salmon (Oncorhynchus kisutch). Gen Comp Endocrinol 88: 243-252 Green H, Morikawa M, Nixon T (1985) A dual effector theory of growth hormone action. Differentiation 29: 195-198 Madsen SS, Bern HA (1993) In-vitro effects of insulin-like growth factor-I on gill Na*, K*-ATPase in coho salmon, Oncorhynchus kisutch. J Endocinol 138: 23-30 McCormick SD, Tsai PI, Kelley KM, Nishioka RS, Bern HA (1992) Hormonal control of sulfate incorporation by branchial cartilage of coho salmon: role of IGF-I. J Exp Zool 262: 166- 172 Mori I, Sakamoto T, Hirano T (1992) Growth hormone (GH)- dependent hepatic GH receptors in the Japanese eel, Anguilla japonica: effects of hypophysectomy and GH injection. Gen Comp Endocrinol 85: 385-391 Prunet P, Boeuf E, Bolton JP, Young G (1989) Smoltification and seawater adaptation in Atlantic salmon (Salmon salar): Plasma prolactin, growth hormone, and thyroid hormones. Gen Comp Endocrinol 74: 355-364 Sakamoto T, Hirano T (1991) Growth hormone receptors in the liver and osmoregulatory organs of rainbow trout: Charac- terization and dynamics during adaptation to seawater. J En- docrinol 130: 425-433 Sakamoto T, Hirano T (1992) Mode of action of growth hormone in salmonid osmoregulation: Expression of insulin-like growth factor I gene during seawater adaptation. Proc Natl Acad Sci USA 90: 1912-1916 Young G, Bjornsson BTh, Prunet P, Lin RJ, Bern HA (1989) Smoltification and seawater adaptation in coho salmon (Oncorhynchus kisutch): Plasma prolactin, growth hormone, thyroid hormones, and cortisol. Gen Comp Endocrinol 74: 335-345 Siharath K, Bern HA (1993) The physiology of insulin-like growth factor (Bengal) (IGF) and its binding proteins in teleost fishes. Proc Zool Soc, Haldane Comm Vol 113-124 ZOOLOGICAL SCIENCE 11: 305-311 (1994) © 1994 Zoological Society of Japan Mesostigmatic Mites (Acari) Associated with Ground, Burying, Roving Carrion and Dung Beetles (Coleoptera) in Sapporo and Tomakomai, Hokkaido, Northern Japan GEN TakaAku!, Haruo Katakura'! and Nospuyo YOSHIDA” ‘Division of Biological Sciences, Graduate School of Science, Hokkaido University, Sapporo, Hokkaido 060, and *Tohoku Agricultural Experiment Station, Morioka, Iwate 020-01, Japan ABSTRACT—A total of 19 species belonging to 5 families of mesostigmatic mites were collected in Sapporo and Tomakomai, northern Japan, on four groups of beetles, i.e., ground beetles (Carabinae, Carabidae), burying beetles (Nicrophorini, Silphinae, Silphidae), roving carrion beetles (Silphini, Silphinae, Silphidae) and dung beetles (Scarabaeidae and Geotrupidae), all of which mainly forage on the ground surface. No mite species was found on more than one group of beetles except for Poecilochirus carabi, which was found almost exclusively on burying beetles and rarely on ground beetles. Mites also seemed to be specific to particular beetle group(s) at the family level. Thus, the “phoretic” mite faunas were distinctly different between the beetle groups: ground beetles (seven species) were characterized by carrying only one mite species, Iphidosoma fimetarium, burying beetles (two species) by two mite species (Alliphis necrophilus, P. carabi), roving carrion beetles (three species) by one mite species, Rhodacaridae sp., and dung beetles (11 species) by 15 mite species that included 8 species of Macrocheles. Mites associated with dung beetles included specialist species such as Holostaspella sp. 1 which was specific to subsocial Copris ochus, and generalist species like Macrocheles sp. aff. glaber that was found on nine dung beetle species. INTRODUCTION Various species of mites are found clinging to or moving on the body surface of other organisms, particularly insects. Although such an association between mites and insects sometimes involves complex relationships such as mutualism [37, 41, 42], the association is usually called phoretic on the assumption that the mites use the beetles as vehicles. The majority of larger species of phoretic mites belong to the suborder Mesostigmata [8, 13, 15, 16]. There are a lot of carrier records of these mesostigmatic mites, and carrier specificity of some species has been investigated [5, 9, 10, 12, 27, 28, 30, 34, 40]. Phoretic mesostigmates are also known as an important control agent of house flies [1, 7, 29]. In Japan, too, phoretic mesostigmatic mites are com- mon, and so far 26 species have been recorded on various imsect species [21-25]. However, these studies dealt with the mites in the middle and the southern parts of Japan, and no comprehensive study has been undertaken on the mites of the northern part. Furthermore, we know very little on the biology of the Japanese species, except for the carrier records of some species. Since 1989, we have studied relationships between mesostigmatic mites and various groups of beetles in Hok- kaido, northern Japan. As an outcome of this study, we will present below a list of mesostigmatic mites so far confirmed by us on four taxonomically and ecologically different groups of beetles, i.e., ground beetles (Carabidae), burying beetles (Nicrophorini, Silphinae, Silphidae), roving carrion beetles Accepted January 28, 1994 Received December 24, 1993 (Silphini, Silphinae, Silphidae), and dung beetles (Geotrupi- dae and Scarabaeidae). We will also summarize suggested relationships between mesostigmatic mites and carrier bee- tles. MATERIALS AND METHODS We collected beetles with the aid of pitfall traps set at forest margins and in the forests in the vicinity of Sapporo, and Hokkaido University Tomakomai Experiment Forest, Tomakomai, Hokkaido, northern Japan. The pitfall traps were baited with meat or without any bait. We also collected beetles from cattle dung by hand in pastures of Hokkaido Agricultural Experiment Station in Sapporo. Beetles collected on other occasions were also examined for their phoretic mites. No quantitative sampling of beetles was under- taken. Beetles were anesthetized with chloroform or ethyl-ethel. Mites detached from the anesthetized beetles were fixed in 70% ethanol, softened and cleared in lactophenol and mounted with a gum-chroral medium on a glass slide [27, 36]. Observation and identification of mites were made on these mounted specimens under a phase-contrast microscope. RESULTS Beetles examined and those bearing mesostigmatic mites We examined 33 species of beetles during the course of this study. Mesostigmatic mites were obtained from 23 beetle species, which are asterisked in the following list (scientific names of the beetles followed Hirashima [17], except those of Aphodius dung beetles for which we follow Masumoto ef al. [33]): Ground beetles (Carabinae, Carabidae, 12 species): Cychrus morawitzi Géhin, 1885, Calosoma maximowiczi 306 G. TAKAKU, H. KATAKURA AND N. YOSHIDA (Morawitz, 1863), Cl. inquisitor cyanescens Motschulsky, 1858, Campalita chinense (Kirby, 1818), *Carabus granulatus yezoensis Bates, 1883, *C. conciliator hokkaidensis Lapouge, 1924, *C. albrechti albrechti Morawitz, 1862, *Leptocarabus arboreus arboreus (Lewis, 1882), *L. opaculus opaculus (Putzeys, 1875), Procrustes kolbei aino (Rost, 1908), *Damaster gehinii gehinii (Fairmaire, 1876), *D. blaptoides rugipennis (Motschulsky, 1861). Burying beetles (Nicrophorini, Silphinae, Silphidae, 5 species): *Nicrophorus maculifrons Kraatz, 1877, *N. quadri- punctatus Kraatz, 1897, N. investigator investigator Zetter- stedt, 1824, N. vespilloides (Herbst, 1784), Ptomascopus morio (Kraatz, 1877). Roving carrion beetles (Silphini, Silphinae, Silphidae, 4 species): *Silpha perforata venatoria Harold, 1877, *Eusilpha japonica (Motschulsky, 1860), *Phosphuga atrata (Linnaeus, 1758), Dendroxena sexcarinata (Motschulsky, 1866). Dung beetles (Geotrupidae and Scarabaeidae, 12 spe- cies): *Geotrupes auratus Motschulsky, 1857, *G. laevistriatus Motschulsky, 1857, *Copris ochus Motschulsky, 1860, *Liatongus phanaeoides (Westwood, 1840), *Caccobius jes- soensis Harold, 1867, *Onthophagus ater Waterhouse, 1875, *O. atripennis atripennis Waterhouse, 1875, *Aphodius ele- gans Allibert, 1847, *A. haemorrhoidalis (Linnaeus, 1758), *A. quadratus Reiche, 1847, *A. pusillus (Herbst, 1789), A. rectus (Motschulsky, 1866). List of mesostigmatic mites collected A total of nineteen species of mites were collected from 23 species of beetles. They are enumerated below with some notes. Full taxonomic accounts of these mites will be published elsewhere by the first author ((38]; in preparation). Mites referred to by the combinations of the generic name and the species code number are undescribed species. Asterisked species are those recorded from Japan for the first time. For each species, a) stages of mites collected; b) carrier beetles confirmed by us; c) attaching site and d) known geographic distribution, are given. Superfamily Eviphidoidea Family Eviphididae 1) Eviphis cultratellus (Berlese, 1910)*: a) female, male, deutonymph; b) Copris ochus; c) ventral surface of body, mainly intersegmental membrane between prothorax and mesothorax; d) Japan (Hokkaido), Java, Egypt, India, South Africa. This species has been collected on the dung beetles Onitis spp. [35], Copris sp. [2], and from cattle dung [3]. 2) Eviphis sp. 1*: a) female; b) Copris ochus; c) ventral surface of body, mainly intersegmental membrane between prothorax and mesothorax; d) Japan (Hokkaido). 3) Alliphis halleri (G. & R. Canestrini, 1881): a) female, male, deutonymph; b) Copris ochus, Caccobius jes- soensis, Aphodius elegans; c) mainly around the mouthparts; d) Japan (Hokkaido, Shikoku), Europe, Israel. This species has so far been collected on the dung beetle Geotrupes laevistriatus and on the burying beetle Nicrophorus quadripunctatus in southern Japan [21], and on five species of dung beetles (Copris [8], Geotrupes [6, 18, 26]) in Europe and Israel. 4) Alliphis necrophilus Christie, 1983*: a) female, male, deutonymph; b) Nicrophorus maculifrons, N. quadri- punctatus; c) specifically found on the ventral membranous portion connecting prothorax and mesothorax; d) Japan (Hokkaido), UK. Unlike A. halleri, the present species was exclusively found on burying beetles of the genus Nicrophorus. A. necrophilus also has been collected on several species of Nicrophorus beetles in the UK [6]. 5) Scarabaspis spinosus Ishikawa, 1968: a) female, male, deutonymph; b) Geotrupes laevistriatus; c) ventral surface of body, mainly prothorax; d) Japan (Hokkaido, Shikoku). This species seems to be specific to the dung beetle G. laevistriatus [21, present study]. 6) Pelethiphis hogai Ishikawa, 1984: a) female, male; b) Geotrupes auratus; c) specifically found on the ventral side of the prothorax; d) Japan (Hokkaido, Honshu). This species is thus far known only on G. auratus ([{25]; present study). Family Macrochelidae 7) Macrocheles insignitus Berlese, 1918: a) female; b) Liatongus phanaeoides, Aphodius quadratus; c) mainly found on the ventral surface of body, in particular the membranous portion between prothorax and mesothorax; d) Japan (Hok- kaido, Honshu, Shikoku), Europe, USA. This species has been found in a variety of habitats including beetles, a rodent, compost, cattle dung, grassland soil, wet mosses, and nests of bumble bees (Bombus sp.) [20, 24, 31]. 8) Macrocheles serratus Ishikawa, 1968: a) female, male, deutonymph; b) Geotrupes laevistriatus, Aphodius quadratus; c) mainly found on the ventral surface of body, in particular the membranous portion between prothorax and mesothorax; d) Japan (Hokkaido, Shikoku). 9) Macrocheles sp. aff. glaber (Miller, 1860)*: a) female; b) Geotrupes auratus, G. laevistriatus, Copris ochus, Liatongus phanaeoides, Caccobius jessoensis, Aphodius quadratus, A. haemorrhoidalis, A. pusillus, A. elegans; c) mainly found on the ventral surface of body, in particular the membranous portion between prothorax and mesothorax; d) Japan (Hokkaido); M. glaber is nearly cosmopolitan, being distributed in Europe, Meditteranian areas, Russia, USA, Australia and New Zealand. Although this species can be identified with M. glaber we could not obtain deutonymphs which are indispensable for exact identification. M. glaber has so far been collected on various groups of coprophagous scarabaeid beetles (Geot- rupes, Typhaeus, Aphodius, Onthophagus, Copris) in Europe, Australia and Japan. Also collected on burying beetles (Nicrophorus), bumble bees (Bombus), small mam- mals and their nests, and in a variety of manure and rotting vegetation habitats [11, 20]. 10) Macrocheles sp. aff. monchadskii Bregetova & Koroleva, 1960*: a) female; b) Geotrupes laevistriatus, Onthophagus ater; c) mainly found on the ventral surface of body, in particular the membranous portion between prothorax and mesothorax; d) Japan (Hokkaido); M. mon- chadskii was described from Adzhar (Russia). Macrocheles monchadskii was recorded from leaf litter [4]. 11) Macrocheles sp. aff. hallidayi Walter & Krantz, 1986*: a) female; b) Geotrupes auratus, Onthophagus ater, O. atripennis atripennis, Copris ochus, Liatongus phanaeoides, Aphodius quadratus; c) mainly found on the ventral surface of body, in particular the membranous portion between prothorax and mesothorax; d) Japan (Hokkaido); M. halli- dayi is widespread in SE Asia, covering India, Thailand, Cambodia, Java and Sarawak. Macrocheles hallidayi has been found on dung beetles of the genera Onitis, Heliocopris, and Catharsius [39]. 12) Macrocheles sp. aff. moneronicus Bregetova & Koroleva, 1960*: a) female; b) Geotrupes laevistriatus, Liatongus phanaeoides; c) mainly found on the ventral sur- face of body, in particular the membranous portion between prothorax and mesothorax; d) Japan (Hokkaido); M. moner- onicus was described from Moneron Island (Russia) [4]. 13) Macrocheles sp. 1*: a) female; b) Geotrupes laevis- triatus, Onthophagus ater, O. atripennis atripennis; c) mainly found on the ventral surface of body, in particular the membranous portion between prothorax and mesothorax; d) Japan (Hokkaido). 14) Macrocheles sp. 2*: a) female; b) Copris ochus, Liatongus phanaeoides, Onthophagus ater, Aphodius quadra- tus, A. elegans; c) mainly found on the ventral surface of body, in particular the membranous portion between prothorax and mesothorax; d) Japan (Hokkaido). 15) Holostaspella sp. 1*: a) female, male, deutonymph, protonymph; b) Copris ochus; c) mainly found on the ventral surface of body, in particular the membranous portion be- tween prothorax and mesothorax; d) Japan (Hokkaido). This species was found exclusively on the subsocial dung beetle C. ochus. Not only adults, but also deutonymphs and protonymphs were collected on beetle body surface and from dung balls which were made by the parental beetles as the larval food [38]. Family Pachylaelapidae 16) Pachylaelaps copris Ishikawa, 1984: a) female, male; b) Copris ochus; c) mainly found on the ventral surface of body, in particular the membranous portion between prothorax and mesothorax; d) Japan (Hokkaido, Kyushu). Thus far collected only on C. ochus ((25]; present study). Superfamily Parasitoidea Family Parasitidae 17) Poecilochirus carabi G. & R. Canestrini, 1882*: a) Mesostigmatic Mites on Beetles 307 deutonymph; b) Nicrophorus maculifrons, N. quadripunc- tatus, Carabus albrechti albrechti, Damaster gehinii gehinii; c) mainly found on the ventral surface of body, in particular the membranous portion between prothorax and mesothorax; d) Japan (Hokkaido), Europe, Russia, China, USA. P. carabi has been collected on various species of burying beetles (Nicrophorus) and carcasses of birds and small mam- mals, and has been said to be mutualistic with Nicrophorus beetles [5, 37, 41, 42]. In the present study, too, this species was mainly collected on Nicrophorus beetles. In addition, P. carabi has been collected on dung beetles (Aphodius) and ground beetles (Carabus, Pterostichus) ({19, 34, 37, 40]; present study). The nature of association between P. carabi and beetles other than burying beetles is yet unknown. Superfamily Rhodacaroidea Family Rhodacaridae 18) Iphidosoma fimetarium (Miller, 1859)*: a) female; b) Carabus granulatus yezoensis, C. conciliator hokkaidensis, C. albrechti albrechti, Leptocarabus arboreus arboreus, L. opaculus opaculus, Damaster gehinii gehinii, D. blaptoides rugipennis; c) mainly attached to the dorsal side of mesothor- ax, metathorax and abdomen covered by elytra; d) Japan (Hokkaido), Russia, Europe. We collected I. fimetarium on seven species of carabid beetles. In Russia and Europe, too, this species has been collected on ground beetles of the genera Carabus, Pteros- tichus and Nebria [14, 32]. 19) Rhodacaridae sp. a) deutonymph; b) Silpha perfor- ata venatoria, Eusilpha japonica, Phosphuga atrata; c) mainly attached to the dorsal side of mesothorax, metathorax and abdomen covered by elytra. d) Japan (Hokkaido). The genus, to which this species belongs, is difficult to determine since only deutonymphs are available. REMARKS Summarizing the above findings, we prepared a synoptic list of mesostigmatic mites and carrier beetles (Table 1). Some ecological properties of examined beetles were summa- rized in Table2, together with associated mite families. Recent studies have shown that carrier specificity may be different between populations of a single mite species [5, 34, 40]. Furthermore, there may be seasonal difference in mite frequencies on beetles. Since we did not make any quantita- tive sampling of beetles, our data is not appropriate to analyze such spatio-temporal variation of beetle-mite interac- tion. However, our results suggest certain noteworthy aspects of mite-beetle associations as shown below. First, mesostigmatic mites were never collected on the beetles foraging on tree foliage; all the beetles that carried mesostigmatic mites forage on the ground (Table 2), or forage on the foliage of undergrowth plants (Phosphuga atrata). Secondly, except for P. carabi, which was almost exclusively found on burying beetles but rarely on ground beetles, no mite species were found on more than one beetle 308 G. TakaAku, H. KATAKURA AND N. YOSHIDA TaBLE 1. Synopsis of carrier beetles and associated mesostigmatic mites examined in the present study Eviphis cultratellus E. sp. 1 A. necrophilus Alliphis hallert Groups and species of carrier beetles Eviphididae Scarabaspis spinosus Pelethiphis hogat Macrochelidae Families and species of mesostigmatic mites | Macrocheles insignitus M. serratus M. sp. aff. glaber M. sp. aff. monchadskit M. sp. aff. hallidayi M. sp. aff. moneronicus M. sp. 1 M. sp. 2 Holostaspella sp. 1 Pachylaelapidae Pachylaelaps copris Parasitidae Poecilochirus carabi Rhodacaridae Iphidosoma fimetarium Rhodacaridae sp. Ground beetles Carabus glanulatus yesoensis C. conciliator hokkaidensis C. albrechti albrechti Leptocarabus arboreus arboreus L. opaculus opaculus Damaster gehinii gehinii D. blaptoides rugipennis t++++4+4++4 Burying beetles Nicrophorus maculifrons + N. quadripunctaus + Roving carrion beetles Silpha perforata venatoria Eusilpha japonica Phosphuga atrata Dung beetles Geotrupes auratus G. laevistriatus Copris ochus tej atencete Liatongus phanaeoides Caccobius jessoensis sf Onthophagus ater O. atripennis atripennis Aphodius quadratus A. haemorrhoidalis A. pusillus A. elegans aF + +++ 44 + + + + +++ + group (Table 1). Such strong associations with particular groups of beetles are also noticed at the family level of mites (Table 2). We briefly examine below these mite-beetle asso- ciations for each mite family. Eviphididae: Five eviphidid species were collected on five species of dung beetles, and one eviphidid species on two species of burying beetles. Since all these beetle species are either presocial or subsocial and bury food under the ground for their larvae, eviphidid mites might be specific to beetles that share this particular behavioral characteristic. Carrier specificity of eviphidid mites seems to be intense. Two closely related Alliphis species differ in their carriers, A. halleri on dung beetles and A. necrophilus on burying beetles (Nicrophorus). Furthermore, Scarabaspis spinosus and Pelethiphis hogai were found to be specific to different species of Geotrupes dung beetles, the former to G. Jaevistriatus and the latter to G. auratus. Macrochelidae: About half of the mite species reported in the present paper are macrochelids, and they belong to a single genus Macrocheles, except for Holostaspella sp. 1. All the species were collected on dung beetles. Some Mac- rocheles species, in particular M. sp. aff. glaber, are general- ists and were found on various species of dung beetles, whereas Holostaspella sp. 1 was a specialist and found only on the subsocial Copris ochus and in the dung balls prepared by the beetles. Pachylaelapidae: Only one species, Pachylaelaps copris, was collected. Like Holostaspella sp. 1, P. copris was spe- Mesostigmatic Mites on Beetles 309 TABLE 2. Some ecological properties of beetles examined and associated mite families. Associated mite family*** Beetle group and genus Food type Foraging at Flight ability Sociality Senn ae acres aia Ground beetles Cychrus* Live invertebrates Ground — None Calosoma* Live invertebrates Tree foliage + None Campalita* Live invertebrates Tree foliage dE None Carabus Live invertebrates Ground = None (+) + Leptocarabus Live invertebrates Ground = None + Procrustes Live invertebrates Ground None Damaster Live invertebrates Ground = None (+) + Burying beetles Nicrophorus Small vertebrate Ground AF Subsocial + aE carcasses Ptomascops* Vertebrate carcasses Ground + None (?) Roving carrion beetles Silpha Live and dead organic Ground = None + matter Eusilpha Live and dead organic Ground — None + matter Phosphuga Live invertebrates Foliage of — None + undergrowth Dendroxena* Live invertedrates Tree foliage + None Dung beetles Geotrupes auratus Vertebrate dung Ground aE Presocial ++ G. laevistriatus Vertebrate dung, dead Ground +(?) Presocial + + organic matter Copris Vertebrate dung Ground 4 Subsocial + + + Liatongus Vertebrate dung Ground ae Presocial + Caccobius Vertebrate dung Ground + Presocial + + Onthophagus Vertebrate dung Ground + Presocial =F Aphodius \** Vertebrate dung Ground 4p Presocial a Aphodius I1** Vertebrate dung Ground + None + * No mesostigmatic mites were found on these beetles ** Aphodius 1: A. quadratus, A. elegans; Aphodius I: A. haemorrhoidalis, A. rectus, A. pusillus. *** EV, Eviphididae; MA, Macrochelidae; PC, Pachylaelapidae; PR, Parasitidae; RD, Rhodacaridae. cific to Copris ochus. Parasitidae: Only one species, Poecilochirus carabi, was collected. ‘This is the only species that was collected on two different groups of beetles, the ground beetles and burying beetles. In USA and Europe, this species is known to be mutualistic with burying beetles [5, 37, 41, 42], but nothing is known about the association with ground beetles as men- tioned before. Rhodacaridae: Rhodacarid mites in the surveyed habi- tats may be specific to beetles that forage exclusively on the ground. All the beetle species bearing rhodacarids are either ground beetles or roving carrion beetles that cannot fly, whereas burying beetles and dung beetles, which were rhoda- carid-free, can fly well (with the possible exception of G. laevistriatus which rarely flies as far as we know). Iphidoso- ma fimetarium was specific to ground beetles but did not show preference to particular species of ground beetles. Like- wise, Rhodacaridae sp. was specific to roving carrion beetles but did not show species specificity. Reflecting these carrier specificities of mesostigmatic mites, the “phoretic mite” faunas of the four beetle groups were distinctly different (Tables1, 2). Ground beetles (seven species) were characterized by only one mite species (Iphidosoma fimetarium), burying beetles (two species) by two species (Alliphis necrophilus, Poecilochirus carabi), rov- ing carrion beetles (three species) by one species (Rhodacar- idae sp.), and dung beetles (11 species) by 15 species, including 8 species of Macrocheles. Thus, most mite species, or mite families, showed de- finite preference for a particular group of beetles. The reason is not yet clear. Since the four beetle groups treated in the present study differ phylogenetically and ecologically, both phylogenetic constraints and ecological factors could affect the carrier specificity. Future studies will clarify which factor is most important in shaping the carrier preference of these mites. 310 G. TaKAKuU, H. KATAKURA AND N. YOSHIDA ACKNOWLEDGMENTS We express our thanks to Dr. G. W. Krantz for his review of the manuscript, and two anonymous reviewers for their helpful com- ments. This study was supported by a Grant in Aid (No. 04264203) from the Ministry of Education, Science and Culture of Japan. REFERENCES 1 Axtell RC (1964) Phoretic relationship of some common manure-inhabiting Macrochelidae (Acarina: Mesostigmata) to the house fly. Ann Entomol Soc America 57: 584-587 2 Berlese A (1910) Lista di nuove specie e nuovi generi di Acari. Redia 6: 242-271 3. Bhattacharyya SK (1971) The genus Eviphis in India (Acarina: Mesostigmata: Eviphididae). Acarologia 13: 266-271 4 Bregetova NG, Koroleva EV (1960) The macrochelid mites (Gamasoidea, Macrochelidae) in the USSR. Parazit Sb 19: 32- 154 5 Brown JM, Wilson DS (1994) Poecilochirus carabi: Behavioral and life-history adaptations to different hosts and the consequ- ences of geographical shifts in host communities. In “Mites: Ecological and Evolutionary Analyses of Life-History Patterns” Ed by MA Houck, Chapman & Hall, pp 1-22 6 Christie JE (1983) A new species of Alliphis (Mesostigmata: Eviphididae) from Britain. Acarologia 24: 231-242 7 Cicolani B (1992) Macrochelid mites (Acari: Mesostigmata) occurring in animal droppings in the pasture ecosystem in central Italy. Agri Ecosys Environ 40: 47-60 8 Costa M (1963) The mesostigmatic mites associated with Cop- ris hispanus (L.) (Coleoptera, Scarabaeidae) in Israel. J Linn Soc Zool 45: 25-45 9 Costa M (1967) Notes on macrochelids associated with manure and coprid beetles in Israel. II. Three new species of the Macrocheles pisentii complex, with notes on their biology. Acarologia 9: 304-329 10 Costa M (1969) The association between mesostigmatic mites and coprid beetles. Acarologia 11: 411-426 11 Evans GO, Browning E (1956) British mites of the subfamily Macrochelinae Tragardh (Gamasina - Macrochelidae). Bull Br Mus nat Hist (Zool) 4: 1-55. 12 Evans GO, Hyatt KH (1963) Mites of the genus Macrocheles Latr (Mesostigmata) associated with coprid beetles in the collec- tions of the British Museum (Natural History). Bull Br Mus nat Hist (Zool) 9: 327-401 13 Evans GO, Sheals JG, Macfarlane D (1961) The Terrestrial Acari of the British Isles. An Introduction to their Morpholo- gy, Biology and Classification. 1. Introduction and Biology. British Museum, London 14 Ghilarov MS, Bregetova NG (1977) A Key to the Soil- Inhabiting Mites. Mesostigmata. Zool Inst Akad Sci USSR, Leningrad, 718 pp 15 Halffter G., Matthews GE (1971) The natural history of dung beetles. A supplement on associated biota. Rev lat amer Microbiol 13: 147-164 16 Halliday RB (1986) Mites of the Macrocheles glaber group in Australia (Acarina: Macrochelidae). Aust J Zool 34: 733-752 17 Hirashima Y (Ed) (1989) A Check List of Japanese Insects I. xi +540 pp. Published by Entomological Laboratory, Faculty of Agriculture, Kyushu University, Fukuoka. 18 Hyatt KH (1959) Mesostigmatic mites associated with Geot- rupes stercorarius (L.) (Col., Scarabaeidae). Entomologist’s Mon Mag 95: 22-23 19 Hyatt KH (1980) Mites of the subfamily Parasitinae (Mesostig- 20 21 22 23 24 25 26 27 28 29 30 31 32 34 35 36 37 39 mata: Parasitidae) in the British Isles. Bull Br Mus nat Hist (Zool) 38: 237-378 Hyatt KH, Emberson RM (1988) A review of the Macrocheli- dae (Acari: Mesostigmata) of the British Isles. Bull Br Mus nat Hist (Zool) 54: 63-125 Ishikawa K (1968) Studies on the mesostigmatid mites associ- ated with the insects in Japan (I). Rep Res Matsuyama Shino- nome Jr Coll 3: 197-218 Ishikawa K (1977) On the mesostigmatid mites associated with the cerambycid beetle, Monochamus alternatus Hope (I). Annot Zool Japon 50: 99-104 Ishikawa K (1977) On the mesostigmatid mites associated with the cerambycid beetle, Monochamus alternatus Hope (II). Annot Zool Japon 50: 182-186 Ishikawa K (1980) Macrocheles insignitus. In “Illustrations of the Mites and Ticks of Japan” Ed by S Ehara, Zenkoku Noson Kyoiku Kyokai, Tokyo, pp 90-91 (In Japanese) Ishikawa K (1984) Studies on the mesostigmatid mites associ- ated with the insects in Japan (II). Rep Res Matsuyama Shinonome Jr Coll 15: 89-102 Koyumdjieva MI (1981) Gamasid mites (Gamasoidea, Para- sitiformes) associated with scarabaeid beetles (Coleoptera, Scar- abaeidae) in Bulgaria. Acta Zool Bul 17: 17-26 Krantz GW (1978) A Manual of Acarology, Second Edition. Oregon State Univ. Book Stores, Inc., Corvallis, 509 pp Krantz GW (1981) Two new glaber group species of Mac- rocheles (Acari: Macrochelidae) from southern Africa. Int J Acarol 7: 3-16 Krantz GW (1983) Mites and biological control agents of dung-breeding flies, with special reference to the Macrochelidae. In “Biological Control of Pests by Mites” Ed by MA Hoy, GL Cunningham, L Knutson, University of California, Special Pub- lication 3304, Berkley, pp 91-98 Krantz GW, Mellott JL (1972) Studies on phoretic specificity in Macrocheles mycotrupetes and M. peltotrupetes Krantz and Mellott (Acari: Macrochelidae), associates of geotrupine Scar- abaeidae. Acarologia 14: 317-344 Krantz GW, Whittaker JO Jr (1988) Mites of the genus Macrocheles (Acari: Macrochelidae) associated with small mam- mals in North America. Acarologia 29: 225-259 Lundqvist L (1991) Rearing deutonymphs of Iphidosoma fimetarium (J. Miller), a mesostigmatic mite associated with carabid beetles. In “The Acari: Reproduction, Development and Life-history Strategies” Ed by R Schuster, PW Murphy, pp 445-452 Masumoto K., Dellacasa G, Kiuchi M (1990) On the Aphodius species of Japan. Ent Rev Japan 45: 145-156 Miller JK, Schwarz HH (1990) Differences in carrier prefer- ence and evidence of reproductive isolation between mites of Poecilochirus carabi (Acari, Parasitidae) living phoretically on two sympatric Necrophorus species. Zool Jb Syst 117: 23-30 Ryke PAJ, Meyer MKP (1957) Eviphidinae Berlese 1913 (Mesostigmata: Acarina) associated with South African beetles. Ann Mag Nat Hist (Ser 12) 10: 593-618 Shirasaka A, Ito Y (1980) Technique for the collecting and preparation of mites. In “Illustrations of the Mites and Ticks of Japan” Ed by S Ehara, Zenkoku Néson Kyéiku Kyékai, Tokyo, pp 511-520 (In Japanese) Springett BP (1968) Aspects of the relationship between burying beetles, Necrophorus spp. and the mite, Poecilochirus necrophori Vitz. J Anim Ecol 37: 417-427 Takaku G (1994) A new species of the genus Holostaspella (Acari; Macrochelidae) from northern Japan. Acarologia 35 (In press) Walter DE, Krantz GW (1986) Description of the Macrocheles 40 41 311 Mesostigmatic Mites on Beetles kraepelini species complex (Acari: Macrochelidae) with two new species. Can J Zool 64: 212-217 Wilson DS (1982) Genetic polymorphism for carrier prefer- ence in a phoretic mite. Ann Entomol Soc America 75: 293- 296 Wilson DS (1983) The effect of population structure on the 42 evolution of mutualism: a field test involving burying beetles and their phoretic mites. Amer Nat 121: 851-870 Wilson DS, Knollenberg WG (1987) Adaptive indirect effects: the fitness of burying beetles with and without their phoretic mites. Evol Ecol 1: 139-159 ZOOLOGICAL SCIENCE 11: 313-317 (1994) © 1994 Zoological Society of Japan First Fossil Record of the Family Phormosomatidae (Echinothurioida: Echinoidea) from the Early Miocene Morozaki Group, Central Japan SHONAN AmemiyA!, YOSHIAKI MIzUNO? and SuGuRU OHTA? ' Misaki Marine Biological Station, University of Tokyo, Miura-shi, Kanagawa 238-02, *Tokai Fossil Society, 9-21, Sawashita-cho, Atsuta-ku, Nagoya, Aichi 456, 3Ocean Research Institute, University of Tokyo, 1-15-1, Minamidai, Nakano-ku, Tokyo 164, Japan ABSTRACT—A fossil echinothurioid echinoid is described from the Early Miocene Morozaki Group in the Chita Peninsula of Aichi Prefecture, central Japan. Based on geological observations, the fossil species is supposed to be the inhabitant of the bathyal zone. The diagnosis of the species is as follows: A small body size for an echinothurioid, large and deep areoles in the oral side, slender teeth with a sharp point, and a large peristome area. From the supposed habitat and the diagnosis, this species is considered to be identical with or a direct ancestor of Phormosoma bursarium, a common extant species in the bathyal zone of the Indo-Pacific region including the southern coasts of the Japanese main islands. This is not only the first fossil record in the world of the family Phormosomatidae, but also the third fossil species for the order Echinothurioida. INTRODUCTION Fossil echinothurioid echinoids have scarcely been found because the echinothurioids have flexible tests with rich connective tissue and poorly calcified ossicles [19]. Extant echinothurioids are divided into two families, Echinothur- iidae and Phormosomatidae [7, 18, 19]. As the firm records, only two fossile species of echinothurioids have been reported [19]. One species is Echinothuria floris from the Upper Cretaceous of the British Chalk [20], and another is Araeoso- ma thetidis from the Pliocene of New Zealand [4, 19]. No fossil phormosomatid has been reported so far [19]. The sea along the Pacific coast of the Japanese main island is so extensively exploited that some groups of organ- isms known elsewhere only from the deep sea seems to occur there in relatively shallower waters. Echinothurioids are one of such groups, and a considerable number of species of the order can be found. The extant species of echinothur- ioids found on the bottom of or shallower than the upper _bathyal zone in the Japanese coast include 5 species of the Echinothuriidae: Asthenosoma ijimai from the infralittoral zone, Araeosoma owstoni and Hapalosoma gemmiferum from the circalittoral (lower sublittoral covering the edge of con- tinental shelf) zone, and Calveriosoma gracile and Hygroso- ma hoplacantha from the upper to middle bathyal zone [1, 2, 8, 10-13, 16, 21], and one species of the Phormosomatidae, that is Phormosoma bursarium from the upper bathyal zone [8, 10-13, 16]. However, fossil specimens of echinothur- ioids have never been reported from Japan. Miocene deposits known as the Morozaki Group are distributed widely in the southern area of Aichi Prefecture. The Group is characterized by extremely rich marine fossil Accepted January 7, 1994 Received April 17, 1992 fauna consisting of bathyal and mesopelagic assemblage [14]. Recently, we obtained a large amount of deep-sea fossils comprising fishes, echinoderms, crustaceans and molluscs, from sandstone and mudstone beds in the Morozaki Group. Among those, some specimens of the Phormosomatidae were found. In the present paper, a fossil phormosomatid species is described on the basis of seven specimens, and its affinity with P. bursarium is discussed. MATERIALS The Morozaki Group is distributed in Saku-shima and Himaka- shima Islands in Mikawa Bay, and the southern edge of the Chita Peninsula in Aichi Prefecture situated in central Japan (Fig. 1). The group is subdivided into four formations which are the Himaka, Toyohama, Yamami, and Utsumi Formations in ascending order [14]. The fossil phormosomatid specimens examined were collected with many other echinoderm species (ca. 17 species) from the Toyohama and Yamami Formations (Fig. 1: Locs. 1-3) and were found in the tuffaceous sandstone beds deposited in the bathyal zone between around 500 m to 1,000 m depth or more [5]. The geologic age of the Toyohama and Yamami Formations was estimated to be Middle Miocene, 15.5 Ma to 16.5 Ma [14, 15]. But recent biostratigraphic studies based on planktonic foraminifers [3] and paleomagnetic study [6] showed that the formations are Early Miocene in age. These formations contain various kinds of fossil species including phormosomatid echinoids. They are composed of benthic, nektonic and planktonic organisms such as fishes, crusta- ceans, molluscs and protozoans having inhabited various depth zones from littoral to bathyal [17]. Echinoids were abundant and accounted for more than a half of the echinoderm specimens col- lected from the formations. More than ten specimens of fossil phormosomatids were col- lected. Six of them (MFM 38057-38060, MI 01, SA 01) are almost intact, and others are fragmental. The fossil specimens examined in this study are deposited in the Mizunami Fossil Museum (MFM 314 S. Amemiya, Y. MizuNo AND S. OHTA 38057-38060), the Misaki Marine Biological Station (MMBS A1), and in the private collections of Y. Mizuno (MI 01) and F. Sakakura (SA 01). MFM 38057, 38060, and MI01 were found in a tuffaceous sandstone bed of the lower part of the Yamami Formation (Fig. 1: Loc. 2), together with crinoids, ophiuroids, crustaceans and mol- luscs. MEM 38057 was collected from a sandstone bed of the upper- most part of the Toyohama Formation (Fig. 1: Loc. 1), together with asteroids and ophiuroids. MEM 38058 and SA 01 was collected from a tuffaceous mud- stone bed of the middle part of the Yamami Formation (Fig. 1: Loc. 3), together with asteroids, ophiuroids, crustaceans, molluscs and fishes. The living specimens of Phormosoma bursarium (MMBS P1— P7) for comparison were collected from Suruga Bay at 550m in depth. SYSTEMATIC DESCRIPTION Order Echinothurioida Claus, 1880 Family Phormosomatidae Mortensen, 1934 Subfamily Phormosomatinae Mortensen, 1934 Genus Phormosoma Thomson, 1872 Phormosoma sp. cf. P. bursarium A. Agassiz, 1881 Diagnosis: 1, a small body size for an echinothurioid; 2, large and deep areoles in the oral side; 3, slender teeth with a CHITA PENINSULA sharp point; 4, a large peristome area. Description: This description is based on six intact (MFM 38057-38060, MI01, SA01) and one fragmental (MMBS A1) specimens. The test is circular and rather small in size for an echinothurioid with the diameters ranging from 48 mm to 118mm (Table 1). The size is comparable to that of Phormosoma bursarium, an extant species, but somewhat larger than the other extant species, P. placenta, P. rigidum, and P. verticillatum, none of which reaches a larger size than 90 mm in diameter [8]. Specimens are squashed along the oral-aboral axis. In most specimens, the oral side is well preserved (Fig. 2), but the aboral side is damaged heavily, or lost. The oral side of the test looks like a honey-comb (Figs. 2, 3), because the areoles of the primary tubercles in the oral side are very large and deep. The five pieces of teeth in the lantern are slender with sharp points (Figs.2, 4). The areoles in the aboral side, in contrast to the oral side, are small and not so deep (Fig. 4). The peristome area is large (Figs. 2, 4), although not so large as in P. bursarium (Table 1). Interambulacra on the oral side are almost 2.5 times as broad as ambulacra, somewhat larger (ca. twice) than in P. bursarium. Plates are large, not numerous. An ambulac- ral plate is composed of a large primary plate and two small demiplates which are situated at the side of the lower edge of the primary plate (Fig. 3). Three pairs of pores are found on these plates. A pair of pores is situated in the margin of the primary plate. Another pair of pores is on each demiplate. UTUMI N \ YAMAN 1 o~ ISE BAY —~ MORQ’ 2 Fic. 1. Map showing the Toyohama and Yamami Formations and Localities 1-3 (Locs. 1-3) in the Morozaki Group in the Chita Peninsula, from where the Miocene phormosomatid specimens were collected. Curved arrow shows the location of the Chita Peninsula in the Japanese main islands. Fossil Echinothurioid 315 TABLE 1. Size distribution of tests and peristomes in fossil Phormosoma sp. and living Phormosoma bursarium Phormosoma sp. P. bursarium test peristome test peristome Cat. No. diameter diameter B/A Cat. No. diameter diameter BA’ (A) (mm) —— (B) (mm) (A’) (mm) — (B’) (mm) MEM 38060 48 15 0.31 MMBS P1 78 30 0.38 MEM 38058 63 17 0.27 MMBS P2 79 24 0.30 MI 01 74 23 0.31 MMBS P3 79 28 0.35 MFM 38057 82 21 0.26 MMBS P4 89 33 0.37 MFM 38059 100 25 0.25 MMBS P5 92 28 0.30 SA 01 118 31 0.26 MMBS P6 94 36 0.38 MMBS P7 109 36 0.33 Mean 0.277 0.344 +S.D. + 0.027 +0.035 There are five to six pairs of primary plates in each ambulac- rum on the oral surface, and an ambulacral primary tubercle is on each primary plate. The size of the primary tubercles is almost equal except in the ones situated along the peristome which are somewhat smaller than others. There are four to six interambulacral plates in each column on the oral surface. The number of primary tuber- cles on an interambulacral plate decreases adapically from 3 (for the outermost 1-2 plates), 2 (for the inner 2 plates) to 1 (for the 1—2 plates adjacent to the peristome). Secondary and miliary tubercles are not preserved. The spines pre- served on the marginal fringe of the test of MFM 38058 are short and slender (Fig.5). The apical system is not pre- served well. Remarks: The specimens from the Morozaki Group have large and deep areoles in the oral side, slender teeth with a sharp point, a large peristome area, and a comparatively small body size. These characteristics clearly indicate that they belong to the genus Phormosoma Thomson, 1872. Four extant species belonging to the genus Phormosoma have so far been reported [8]. Their distributional ranges are as follows: P. placenta Thomson, 1872: Northern Atlantic from Iceland and the Davis Strait down to the Azores and the Gulf of Guinea, and to the West Indies, 215-2500 m depths. P. verticillatum Mortensen, 1904: Indian Ocean, from the Bay of Bengal to the Arabian Sea, 1165-1925 m depths. P. rigidum A. Agassiz, 1881: Off New Zealand, 1260 m depth. P. bursarium A. Agassiz, 1881: Indo-Pacific, from the Natal Coast and the Arabian Sea to Australia, New Caledonia, Japan, and the Hawaiian Islands, overall depth range in records 170-2340 m, and dominate between depths of 500- 1700 m on the continental slope of the Pacific coast of Japan. The Morozaki specimens have affinity with P. bursarium in some morphological characters such as the test size, or the arrangement of ambulacral plates and pore pairs on the plates, although some differences are also found between them in the peristome area and in the relative width of the ambulacra and interambulacra. Their habitats seem also to be similar, and a great number of living specimens of P. bursarium have been trawled and sometimes gregarious paches of them were photographed on the silty mud bottoms of the bathyal zone along the southeastern coast of Japanese main islands includ- ing the area off the Chita Peninsula [12]. The depth range of Calveriosoma gracile and Hygrosoma hoplacantha overlaps with that of P. bursarium in the above locations. However, the usual habitats of them are segregated. The former two species prefer sandy mud bottom on the topographic highs swept by strong bottom currents expecting the occasional drifting sea algae as facultative vegetarians [9, 12], whereas the latter species predominates on the silty mud floor where the regime of bottom water movement is relatively calm. If the habitat of the fossil species is the same as that of the extant species, the taphonomy of the soft-shelled sea urchin in the taffaceous sandstone in a formation of Morozaki Group suggests instantaneous burial beneath turbidity currents. The present Phormosoma specimens were collected together with four other echinoid species, 13 species of other echi- noderms and with some fishes, crustaceans and molluscs from tuffaceous sandstone beds deposited in the bathyal zone. The four echinoid species are Temnopleurus sp. (upper bathyal member), Brissopsis sp. (upper to middle bathyal member), a diadematid and a scutellid (littoral members). Their habitats today range from the littoral to bathyal zones, suggesting that some of the fossils have been brought from shallower bottoms into the bathyal zone by turbidity flows and/or episodic disasters involving the littoral, midwater and upper continental slope realms. The fossil specimens found in the Morozaki Group can be identical with or a direct ancestor of P. bursarium, although the information obtained from the present fossil specimens is still insufficient to identify the species exactly. This is the first fossil record in the world of the family Phormosomatidae, and is also the third unequivocal fossil species for the order Echinothurioida [19]. 316 Fic. Fic. Fic. S. AMEMIYA, Y. MIZUNO AND S. OHTA 2. Oral view of Phormosoma sp. (MFM 38057) showing a honey-comb-like structure with large and deep areoles and a large peristome area. A sharp and slender tooth (arrow-head) in the mouth is also seen. The smallest graduation of the measure corresponds to 1 mm. 3. Ambulacral and interambulacral plates on the oral-side of Phormosoma sp. (MMBS A1) viewed under a higher magnification than in Fig. 2. The areoles (arrows) of the primary tubercles are very large and deep. Large and small arrow-heads show the primary ambulacral plates and the demiplates, respectively. Bar represents 1 cm. 4. Oral view of Phormosoma sp. (SA 01). Arrow-heads indicate the five slender teeth in the mouth. Part of the test has been removed and the coelom-side of the aboral test is seen in the left-side of the field. The areoles (arrows) on the aboral test are small. The smallest graduation of the measure corresponds to 1 mm. Fossil Echinothurioid 317 4 Fell HB (1966) Diadematacea. In “Treatise on Invertebrate Paleontology Part U, Echinodermata 3, Vol. 1” Ed. by R. C. Moore, The Geological Society of America Inc, New York pp U340-U366 5 Hachiya K, Yamaoka M, Mizuno Y (1988) Deep sea fauna from the Middle Miocene Morozaki Group in the Chita Peninsu- la, Aichi Prefecture, central Japan. J Growth 27: 119-139 6 Hayashida A (1985) Paleomagnetic study of Miocene strata in the Chita Peninsula, central Japan. Sci Engineer Rev Doshisha Univ 26: 180-185 7 Jensen M (1981) Morphology and classification of Euechi- noidea Bronn, 1860 - a cladistic analysis. Vid Meddr Dansk Naturh Foren 143: 7-99 8 Mortensen Th (1935) A Monograph of the Echinoidea II. C.A. Reitzel Publisher, Copenhagen, pp 1-647; pls 1-89 9 Mortensen T (1938) On the vegetarian diet of some deep-sea echinoids. Annot Zool Japan 17: 225-228 10 Nishiyama S (1968) The echinoid fauna from Japan and adja- cent regions. In “Part II, Special papers of Palaeontological Society of Japan No. 13” Ed by Palaeontological Society of Japan, University of Tokyo Press, Tokyo p 491 11 Ohta S (1980) Photographic census of the larger-sized epibenthos on the pacific coast of central Japan. In “The Kuroshio IV, Proc 4th CSK Symp Tokyo 1979” pp 602-620 12 Ohta S (1983) Photographic census of large-sized benthic organisms in the bathyal zone of Suruga Bay, central Japan. Bulletin of the Ocean Research Institute, University of Tokyo, No. 15, 1-244 13 Okutani T (1969) Synopsis of bathyal and abyssal megalo- invertebrates from Sagami Bay and the south off Boso Peninsula ‘ : ; trawled by the R/V Soyo-Maru. Bull Tokai Reg Fish Res Lab Fic.5. The short and slender spines (arrows) on the marginal fringe 57: 1-62 of the test of Phormosoma sp. (MFM 38058). Bar represents 1 14 Shibata H (1977) Miocene mollusks from the southern part of cm. Chita Peninsula, central Honshu. Bull Mizunami Fossil Mus 4: 45-53 15 Shibata H, Ishigaki T (1981) Heteropodous and pteropodous biostratigraphy of Cenozoic strata of Chubu Province, Japan. ACKNOWLEDGMENTS Ibid 8: 55-70, pls. 12, 13 , ; : ‘ 16 Shigei M (1986) The sea urchins of Sagami Bay, Ed. by We wish to express our cordial thanks to Mr. K. Hachiya for his Biological Laboratory Imperial Household, Maruzen, Tokyo technical assistance to take photographs for Figures 2, 4 and S. 204+ 126 pp pincer’ thanks gieralsordue ie Dest Salo sng O it ro thet 17 Shikama T, Kase T (1976) Molluscan fauna of the Miocene careful reading of the manuscript and advice. We are also indebted Morozaki Group in the southern part of Chita Peninsula, Aichi Sere a ecu omnis waluablecadvice: Prefecture, Japan. Sci Rep Yokohama Nat Univ Sec 2, 23: 1- 25 REFERENCES 18 Smith AB (1984) Echinoid Palaeobiology. George Allen & Unwin, London. 1 Amemiya S, Tsuchiya T (1979) Development of the echi- 19 Smith AB, Wright CW (1990) British Cretaceous Echinoids. nothurid sea urchin Asthenosoma ijimai. Marine Biol 52: 93- The Palaeontographical Society, London. pp 101-198; pls 33-72 96 20 Woodward SP (1863) On Echinothuria floris, a new and ano- 2 Amemiya S, Suyemitsu T, Uemura I (1980) Morphological malous echinoderm from the Chalk of Kent. The Geologist 6: observations on the spermatozoa of echinothurid sea urchins. 327-330 Develop. Growth Differ 22: 327-335 21 Yoshiwara S (1897) On two species of Asthenosoma from the 3 Doi K (1983) On stratigraphy and age of the Miocene Utsumi sea of Sagami. Annot Zool Jap 1: 5-12 Formation, Morozaki Group, Southwest Japan. News Osaka Micropaleont. 10: 14-21 ZOOLOGICAL SCIENCE 11: 319-335 (1994) Phylogeny, Classification, and Biogeography of Goniurosaurus kuroiwae (Squamata: Eublepharidae) from the Ryukyu Archipelago, Japan, with Description of a New Subspecies L. Lee Grismer!, Hipetosui Ora’ and SatosH! TANAKA? ‘Department of Biology, San Diego State University, San Diego, CA 92182, U.S.A., Department of Biology, University of the Ryukyus, Nishihara, Okinawa 903-01, and *Motobu Senior High School, Tokuchi 337 Motobu, Okinawa 905-02, Japan ABSTRACT—The phylogenetic relationships of populations of Goniurosaurus kuroiwae from the Ryukyu Archipelago, Japan, are resolved using a cladistic analysis and their classification modified accordingly. The resultant phylogeny indicates that five subspecies should be recognized: G. k. yamashinae from Kumejima, G. k. orientalis from Tonakijima, Tokashikijima, Akajima, and lejima, G. k. kuroiwae from Okinawajima, Sesokojima, and Kourijima, G. k. toyamai subsp. nov. from Iheyajima, and G. k. splendens from Tokunoshima. Goniurosaurus k. yamashinae is the sister taxon of the remainder. Goniurosaurus k. kuroiwae and G. k. orientalis are sister taxa which collectively form the sister group to the lineage composed of G. k. toyamai and G. k. splendens. Goniurosaurus k. yamashinae and G. k. orientalis are designated as metataxa because they are not demonstrably monophyletic or paraphyletic. A discriminant function analysis of the five subspecies shows G. k. yamashinae and G. k. splendens are separated from each other and from the other subspecies by a wide morphological gap. A\ll these lineages were recognized as subspecies because this increases the phylogenetic content of the classification. The ancestor of the G. kuroiwae group may have dispersed into the Ryukyu Archipelago from continental China by way of a late Miocene to early Pliocene landbridge. The paleogeography of the Ryukyu Archipelago suggests that differentiation within G. kuroiwae resulted from the formation of the Ryukyu’s contemporary configuration caused by rising Pleistocene sea levels. © 1994 Zoological Society of Japan INTRODUCTION In a revision of the gecko family Eublepharidae, the genus Goniurosaurus was resurrected to contain the insular populations of eublepharids from the Gulf of Tonkin, China and the Ryukyu Archipelago, Japan [14, 15]. Thus, as it is currently constituted, Goniurosaurus contains at least two allopatric insular species; G. lichtenfelderi (Mocquard, 1897) [38] from the Island of Hainan and Iles de Norway in the Gulf of Tonkin, China [14] and G. kuroiwae (Namiye, 1912) [42] from 10 islands within the Ryukyu Archipelago, Japan [47]. There may also exist an undescribed species of Goniuro- saurus from the Guizhou province of mainland China [33]. Grismer [15] revised the taxonomy of G. kuroiwae and recognized three subspecies; G. k. yamashinae (Okada, 1936) [44] of Kumejima Island, G. k. kuroiwae of Okinawajima Island and satellite islands to the west, and G. k. splendens (Nakamura et Uéno, 1959) [40] of Tokunoshima Island. Grismer’s [15] classification resulted in the synonymy of G. k. orientalis (Maki, 1930) [35] with G. k. kuroiwae and placed the Kumejima Island population, considered at the time to be G. k. orientalis [41, 51], under the resurrected name of G. k. yamashinae. However, Grismer’s [15] conclusions were provisional because he was only able to examine small series of specimens from Okinawajima and Tokunoshima Islands and relied on literature descriptions for some of the other Accepted January 14, 1994 Received October 28, 1993 insular populations. In a more recent revision, Ota [47] referred all the insular populations west of Okinawajima Island (including that of Kumejima Island) to G. k. orientalis. The Okinawa- jima, Sesokojima, and Kourijima island populations re- mained under G. k. kuroiwae and the Tokunoshima popula- tion remained G. k. splendens. Both classifications [15, 47] suffer because they are based on overall similarity rather than derived similarity and not all insular populations were ex- amined. In this paper we readdress the classification of G. kuroiwae from a phylogenetic standpoint based on a cladistic analysis of all known insular populations. MATERIALS AND METHODS Data were obtained from preserved (Appendix I) and living specimens. Grismer [15] demonstrated that Goniurosaurus kuroiwae is monophyletic based on its possession of the derived character states of tuberculate gular scales, unsheathed claws, and the absence of preanal pores. All specimens from the same island were treated as a single operational taxonomic unit and, as such, ana priori assumption of insular monophyly was adopted. Character states were polarized [34] based on the relationships suggested by Grismer [14-17] where Goniurosaurus lichtenfelderi served as the first outgroup and Eublepharis, Hemitheconyx, and Holodactylus as the second (Fig. 1). Maddison et al. [34] stressed the importance of two sequentially aligned outgroups to ensure that polarity assignments are maximally parsimonious. It is likely, however, that the Eublepharis- Hemitheconyx-Holodactylus clade does not actually comprise the 320 L. L. Grismer, H. Ora AND S. TANAKA Holodactylus Hemitheconyx Eublepharis Goniurosaurus cornii africanus caudicinctus taylori hardwickii angramainyu macularius turcmenicus lichtenfelderi kuroiwae Second outgroup First outgroup Fic. 1. Outgroup taxa and relationships of the first and second outgroups to Goniurosaurus kuroiwae after Grismer [14-17]. immediate second outgroup owing to the possible existence of an undescribed form of Goniurosaurus from mainland China [33]. This population is known from only a single specimen which was unavail- able to us. Thus, the Eublepharis-Hemitheconyx-Holodactylus clade serves, at this point, as the best approximation of an immediate second outgroup. Because of a lack of homoplasy in the data set, computer algorithms were not necessary to aid in tree construction. Following the cladistic analysis, scale counts from the Kume- jima, Tokashikijima, Okinawajima, Iheyajima, and Tokunoshima populations were subjected to a multigroup discriminant function analysis (DA) with the MacIntosh version of Bio” tatII [50], in which each insular population was used as a predefined group. The sample sizes from the remaining islands (Table 1) were too small to yield statistically reliable results. Results from the DA were not used to construct phylogenetic relationships among the populations but only to aid in the morphological characterization and distinction of the terminal taxa suggested by the results of the cladistic analysis. Terminology follows Grismer [15] and scale counts were taken as follows. Supralabials—the series posterior to the rostral and termi- nating with a scale at least twice the size of the surrounding granular scales. Infralabials—the series posterior to the mental and termi- nating with a scale at least twice the size of the surrounding granular scales. Postmentals—all scales except the first infralabials which contact the mental. Preoculars—the linear arrangement of granu- lar scales between the anterior corner of the eye and the posterior margin of the external nares. Eyelid fringe scales—the lateralmost enlarged triangular scales encircling the eye. Paravertebral tuber- cles—the number of paravertebral tubercles between the limb inser- tions. Midbody scales—number of granular scales surrounding the body midway between limb insertions. Fourth toe lamellae— counting from the union of the third and fourth toes and terminating with the distal penultimate scale. Scales surrounding claw on fourth toe—all scales contacting the claw. The following counts were not used in the DA because of their incomplete representation in some or all of the populations. Caudal scales—the number of granular scales in the transverse caudal whorl of a non-regenerated tail at a point midway between the ankle and the knee when the hindlimb is adpressed against the tail. Body bands—number of transverse body bands between the nuchal loop and caudal constriction. PHYLOGENETIC ANALYSIS OF CHARACTERS 1. Dorsal tubercles In the Tokunoshima population, the dorsal body tubercles between the limb insertions are triangular to elliptical in cross-section and sharply keeled anteriorly. The tubercle keel is most pronounced dorsally, becoming less pronounced towards the base. Also, the degree of tubercle keeling increases posteriorly on the body with those tubercles be- tween hind limb insertions being the most strongly keeled. Keeled tubercles do not occur on the head, nape of the neck, or tail. In the Kumejima population, the dorsal body tuber- cles are smooth and conical. In all other populations, tuber- cle-keeling only rarely occurs. When present, it is usually very weak and occurs only in a few tubercles between the hind limb insertions. In Goniurosaurus lichtenfelderi, Eub- lepharis, and Holodactylus, the tubercles are smooth and conical. Only in Hemitheconyx are the tubercles keeled. Therefore, the condition of sharply keeled tubercles in the Tokunoshima population is considered derived. 2. Ventral scales The ventral scales of eublepharid geckos are usually hexagon- al, flat, subimbricate to imbricate, and grade laterally into the granular scales of the dorsum. In the Tokunoshima popula- tion, the ventrals are juxtaposed and sharply raised, giving them a pointed or weakly tuberculate appearance. This condition is most evident in the pectoral region and fades posteriorly, grading into flat hexagonal imbricate interfemor- al scales. Anteriorly, the pectoral scales grade into even more sharply pointed and raised gular scales. In all other populations of Goniurosaurus kuroiwae, as well as G. lichten- felderi, Eublepharis and Hemitheconyx, the ventral scales are flat, wide, and subimbricate to imbricate. Anderson and Leviton [1] stated that the ventral scales of E. angramainyu are juxtaposed but we find them to be subimbricate. In Holodactylus, the ventral scales are juxtaposed but they are Systematics of Goniurosaurus kuroiwae 321 TaBLE 1. Meristic differences between the insular populations of Goniurosaurus kuroiwae. SL=supralabials; IL= infralabials; PM=postmentals; PO=preoculars; EF=eyelid fringe scales; TU=paravertbral tubercles; BO= midbody scales; 4T=fourth toe lamellae; CL=scales surrounding claw on fourth toe; CA=caudal scales; BB= body bands; I=incomplete; and A=absent. Character SL IL PM PO EF TU BO 4T CL CA BB yamashinae Kumejima (n=14) x 8.0 8.4 46 21.0 53.6 28.8 148.6 18.4 6.0 47.7 4 range 7-9 7-9 4-5 19-23 44-65 21-33 132-156 17-20 6 43-53 4 SE +02 360.35 220.2 se0.0 sEx3 aE sea se0.3 0 aE 2,9) 0 (n=6) orientalis Akajima (n=1) 9 10 5 22 63 35 144 19 6 48 4 Iejima (n=1) 9 8 5 22 63 29 145 20 6 50 4 Tokashikijima (n=50) i 9.4 8.6 38 Zils 59.8 35.1 147.2 17.3 6.0 475 4or!I range 8-11 7-11 3-5 17-24 53-66 31-40 135-159 16-19 5-7 42-52 SE 220.0 S202 SEO 203 sb0./ sed, abil se02 asac@hil sei (n=16) Tonakijima (n=2) # 9 10 3.5 20.0 52.00 30.5 139.5 20.5 6 A 4 range 9 10 3-4 19-21 48-56 29-32 137-142 20-21 6 A 4 SE 0 0 ae. EI) an4i(0) seis) SEAS) a E()-5) 0 A 0 kuroiwae Okinawajima (n=221) x 9.4 8.7 5.0 21.4 60.2 33.4 150.9 17.1 6.0 49.8 A range 7-11 7-11 4-6 19-23 54-70 29-43 139-162 15-20 6-7 40-S7 A SE 202 02 sO, 60,2 3607 320.0 Geld s202 =O, 12 A (n=124) Kourijima (n=1) 9 Oe ec 23 51 34 140 15 6 51 A Sesokojima (n=1) 8 8 5 21 57 31 137 17 6 A A toyamat Iheyajima (n=14) x 9.3 8.3 46 21.4 55.6 37.2 149.8 17.3 5.9 47.5 3.8 range 8-10 7-10 3-5 20-23 53-59 34-42 140-158 16-20 5-6 40-53 3-4 SE £02 403 202 2203 s200.5 220.0 cael 220.3 s20)3 si +0.1 (n=6) splendens Tokunoshima (n=27) x 9.0 8.6 2.9 20.5 54.1 20.0 132.4 16.2 7.4 50.4 3 range 8-10 7-10 2-4 18-24 46-59 22-29 121-146 15-18 6-9 46-61 3 SE +0.1 +01 +01 +403 +405 +404 41.4 +402 +02 es 0 n= flat and not sharply raised. Therefore, juxtaposed and sharply raised ventrals in the Tokunoshima population is considered derived. 3. Scales at base of digits In all Goniurosaurus kuroiwae except those from Kumejima Island, there are one to three (usually two) enlarged scales at the base of each digit on the hand and foot. These scales are two to three times the size of the scales of the adjacent palmar and plantar regions. This condition shows a slight indication of ontogenetic variation, being somewhat less pronounced in hatchlings and juveniles. In the Kumejima population, there is a single scale at the base of each digit which is occasionally slightly enlarged but rarely reaching twice the size of the surrounding scales. This is similar to the condi- tion found only in the manus of G. lichtenfelderi. Enlarged scales at the base of the digits do not occur in Eublepharis, Hemitheconyx, and Holodactylus. Therefore, this condition in the manus of all populations of G. kuroiwae, except that from Kumejima Island, is considered derived. 4. Lineate middorsal pattern Goniurosaurus kuroiwae from Akajima, Tokashikijima, Tonakijima, lejima, Okinawajima, Kourijima, and Sesoko- 322 L. L. Grismer, H. Ora AND S. TANAKA jima Islands have lineate tendencies in their dorsal banding patterns (Figs. 2 and 3). In the Akajima, Tokashikijima, Tonakijima, and Iejima populations, there is a middorsal stripe in the nape which usually extends far enough posterior- xe wen ly to contact the first transverse body band near the forelimb insertions. Occasionally this stripe will continue far enough to contact the second transverse body band, nearly midway between the forelimb and hindlimb insertions, but rarely any Fic. 2. Photographs of Goniurosaurus kuroiwae taken at the site of collection. Upper: juvenile G. k. yamashinae from Kumejima Island. Middle: G. k. kuroiwae from southern Okinawajima Island. Lower: adult G. k. splendens from Tokunoshima Island. Photographs by L. Lee Grismer. Systematics of Goniurosaurus kuroiwae 323 z 52 4 ee Can iS Te ae ae = pete Ay Fic. 3. Photographs of Goniurosaurus kuroiwae taken at the site of collection. Upper: adult G. k. toyamai from Iheyajima Island. Middle: G. k. orientalis from Tokashikijima Island. Lower: juvenile G. k. orientalis from Akajima Island. Photographs by Masanao Toyama. further except in the Okinawajima, Kourijima and Sesoko- stripe extending nearly one-half the way down the body. jima populations. The Tonakijima population is known The juvenile (OPM 489) is banded and shows only weak from one adult and one juvenile specimen. The adult evidence of striping on the nape of the neck. Middorsal (NSMT 02522: holotype of G. k. orientalis) has a middorsal lineation is most pronounced in the Okinawajima, Kourijima, 324 L. L. Grismer, H. Ora AND S. TANAKA and Sesokojima populations where a stripe usually extends the entire length of the body and terminates at the caudal constriction (Fig. 2). On Okinawajima Island, there is some variation in that striping in the northernmost populations is not as well defined and a weak, faded banding pattern is sometimes present. Populations from the southern Okina- wajima, however, have a bold, well-defined stripe and seldom show evidence of banding. In the remaining populations of G. kuroiwae, there is no evidence of middorsal striping (Figs. 2 and 3). In large specimens from Tokunoshima Island, there is occasionally a lack of dark pigmentation in the vertebral region that appears in preserved specimens to be a stripe. However, in living specimens this is merely a light- ened area that develops with ontogeny. This region does not contain the same color pigments as the dorsal bands and as such does not constitute a middorsal stripe homologous with that described above. Such pigment loss in the verteb- ral region is common in large individuals of other eublephar- ids [18]. Tendencies toward dorsal pattern lineation do not occur in G. lichtenfelderi. Lineate tendencies are absent in Eubelpharis hardwickii but variable in E. macularius, E. turcmenicus, E. angramainyu, Hemitheconyx caudicinctus, and Holodactylus. Striping is absent in Hemitheconyx taylori. Therefore, the most parsimonious assumption is that lineate tendencies in the dorsal pattern of the Akajima, Tokashikijima, Tonakijima, lejima, Okinawajima, Kouri- jima, and Sesokojima populations are derived. 5. Banding pattern In all populations of Goniurosaurus kuroiwae except those from Okinawajima, Kourijima, and Sesokojima, there is a prominent dorsal pattern consisting of three to four trans- verse bands between the nape of the neck and the caudal constriction. Banding is bold and prominent in the Kume- jima, Iheyajima, and Tokunoshima populations (Figs. 2 and 3). Inthe Akajima, Tokashikijima, Tonakijima, and Iejima populations, banding is prominent but the first band in the vicinity of the forelimb insertion may be incomplete (rarely absent). In the Okinawajima, Kourijima, and Sesokojima populations, banding is absent or incomplete. This condi- tion is most obvious in populations from the southern portion of Okinawajima where there is usually no trace of banding (Fig. 2). In populations from northern Okinawajima Island as well as Kourijima and Sesokojima Islands, some specimens may retain some portions of the more posterior body bands. However, the bands are usually very irregular in shape and suffused with dark pigments from the surrounding ground color. This condition shows little or no ontogenetic varia- tion. All outgroup taxa except Holodactylus cornii have a prominent banding pattern. Therefore, the incomplete to absent transverse dorsal banding pattern of the Okinawajima, Kourijima, and Sesokojima populations is considered de- rived. 6. Hind limb banding In the Tokunoshima population, the posteriormost transverse body band extends laterally onto the dorsal surface of the thigh. In many specimens, the band runs parallel to the long axis of the hind limb uninterrupted to the knee. In others, it is interrupted medially and exists as a lineate blotch on the thigh and knee (Fig.2). In all other populations of Goniurosaurus kuroiwae, the posteriormost body band does not extend along the long axis of the hind limbs. There may be light-colored elongate blotches on the dorsomedial surface of thigh of some specimens, but close examination reveals that they are oriented perpendicular rather than parallel to the long axis of the hind limb and in many cases, are discontinuous with the posteriormost transverse body band. In the outgroup taxa, the posteriormost transverse body band does not extend onto the hind limb. Therefore, the condi- tion for the Tokunoshima population is considered derived. 7. Interspace mottling The common condition in eublepharine geckos sensu Grismer [15] is for hatchlings to have dark, unicolored interspaces (or ground color) between the body bands, which become in- creasingly suffused and/or mottled with lighter coloration with age. This is the case in all Goniurosaurus kuroiwae except the Iheyajima and Tokunoshima populations where the adults retain unmottled dark interspaces (Figs. 2 and 3). Although the specimen of the Akajima population (OPM 341) is a juvenile, mottling in the interspaces is still observ- able (Fig. 3) and as with other eublepharines, assumed to be characteristic of adults. In the specimen from Iejima Island (KUZ 9991), the interspaces appear superficially like those in the Iheyajima and Tokunoshima populations. On close examination, however, it is clear that they are considerably lightened and not uniformly dark throughout. Adults of G. lichtenfelderi have both mottled and unicolored interspaces. The interspaces of adult Eublepharis macularius, E. turcmeni- cus, and E. angramainyu are mottled whereas those of E. hardwickii are unicolored. Holodactylus and Hemitheconyx taylori have mottled interspaces but they are generally unico- lored in Hemitheconyx caudicinctus. Therefore, dark unico- lored interspaces of the adults of the Iheyajima and Toku- noshima populations are considered derived. 8. Juvenile coloration In all populations of Goniurosaurus kuroiwae except that of Kumeyjima Island, the color of the dorsal pattern (striped or banded) overlying the dark-brown ground color is always bright-orange to pink in hatchlings and juveniles. This color intensity usually remains into adulthood but sometimes fades into a cream-yellow hue (Figs. 2 and 3). The juvenile col- oration of the dorsal pattern in specimens from the Kumejima population is whitish. The dorsal color pattern in the juve- niles of both outgroups consists of yellow to whitish hues and is never bright-orange to pink at any stage of life. There- fore, the latter condition is considered derived for all popula- tions of G. kuroiwae except that of Kumejima Island. Systematics of Goniurosaurus kuroiwae 32 9. Eye color The color of the iris in all Goniurosaurus kuroiwae except the Kumejima population is blood-red (Figs. 2 and 3). In the Kumejima population, G. lichtenfelderi, and Eublepharis (E. hardwickii and E. angramainyu not available for examina- tion), the iris is yellow-brown to gold in color. In Hemithe- conyx caudicinctus and Holodactylus africanus, the iris is usually a very dark brown. Living Hemitheconyx taylori and Holodactylus cornii were unavailable for examination. Therefore, based on the relationships of the outgroup taxa observed (Fig. 1), a blood-red iris is considered to be derived for the ingroup. If, however, a blood-red iris is present in the outgroup taxa that were not examined, the polarity assignment would be equivocal. 10. Body stature The overall body stature in the Iheyajima population is robust, whereas that in the other populations of G. kuroiwae, as well as G. lichtenfelderi, more slender (Fig. 4). Thus, although the state of this character is not defined in the Nn members of the second outgroup due to their much divergent body proportion, the robust body stature in the Iheyajima population is considered derived within the genus Goniuro- saurus by assuming the Kumejima population as a first functional outgroup and G. lichtenfelderi as a second out- group [60] (see below). RESULTS There is only one single most parsimonious tree which has a consistency index value of 1.0 (Fig. 5). The distribu- tion of the derived character states (Table 2) suggests that the Kumejima population is the sister taxon of the remaining nine insular populations of Goniurosaurus kuroiwae (Fig. 5). Ota [47] placed the Kumejima population (G. k. yamashinae: sensu Grismer [15]) in G. k. orientalis. However, it is apparent here that such a classification would result in the demonstrative paraphyly of the latter (Fig. 5). Although the Kumejima population is discretely diagnosable from all other G. kuroiwae, it lacks character state support for its monophy- TABLE 2. Distribution of derived (1) and primitive (0) character states among the island populations of Goniurosaurus kuroiwae and the outgroup taxa. examined because live animals were not available to us. ?=Character was not —=Character state was not defined due to the great divergence in related body portions. Characters 1 2 3 4 5 6 7 8 9 10 Taxa Ingroup G. k. yamashinae Kumejima 0 0 G. k. orientalis Akajima Tonakishima Tokashikijima So oS 2S © Se 2 Se ©& Iejima . kuroiwae Okinawajima 0 0 Kourijima 0 0 Sesokojima 0 0 G. k. toyamai Iheyajima 0 0 . splendens Tokunoshima 1 1 Outgroups G. lichtenfelderi lichtenfelderi G. I. hainanensis E. angramanyu 0 0 Eublepharis hardwickii 0 0 E. turcmenicus 0 0 E. macularius Holodactylus africanus H. cornii Hemitheconyx caudicinctus So oo qe0eeoe 2 © (—< js ©& H. taylori = = — So eo eaqQoo eg © 2S © a So Se © SoS eo eo © So 2 2 © aS ey Ss Pp Re Re Re Se 2 So ©& — aN i=) = — i=) S — Meo qaqa eo oagoe So oeeeaeeeed OSoeqa aq qq eee HSE WPOooonwy co © | 326 L. L. GrisMER, H. OTA AND S. TANAKA Fic. 4. Body stature of the five subspecies of Goniurosaurus kuroiwae. From left to right G. k. yamashinae, TPN 78050402 (SVL=87.5) from Kumejima Island; G. k. orientalis, TPN 78052101 (SVL=83.8) from Tokashikijima Island; G. k. toyamai, KUZ 9983 (SVL=84.6) from Iheyajima Island; G. k. splendens, TPN 76102202 (SVL=77.2) from Tokunoshima Island; and G. k. kuroitwae TPN 76102111 (SVL=79.9) from Okinawajima Island. yamashinae_ orientalis Akajima Iejima Tokashikijima Kumejima Tonakijima kurotwae toyamai splendens Okinawajima Kourijima Sesokojima Theyajima Tokunoshima Fic. 5. Cladistic relationships of the insular populations of Goniurosaurus kuroiwae and the resultant classification. Numbered horizontal bars represent the presence of the following derived character states: 1=tubercles sharply keeled; 2 =ventrals juxtaposed; 3=enlarged scale(s) at base of digits; 4=lineate tendencies in middorsal pattern; 5=dorsal banding absent; 6=posteriormost body band extending onto hind limb; 7=interspace mottling absent; 8=orange-pink dorsal pattern in juveniles; 9=iris red; 10=body robust. ly or demonstrable paraphyly, and thus, is given a metataxon designation [12] and recognized here as G. k. yamashinae (see [6, 8, 10, 25] for differing viewpoints on usage of this designation). The DA shows that G. k. yamashinae is well isolated from all other populations examined along the second axis except for a very slight degree of overlap with that of Okinawajima (Fig. 6). Standardized canonical coef- ficients for the first two variates presented in Table 3 account for 92.26% of the observed variation. The remaining nine insular populations form a well- corroborated monophyletic group diagnosed by the derived character states of an enlarged scale at the base of each digit, a blood-red iris, and a bright orange to pink hatchling and juvenile color pattern (Fig. 5). Within this clade, there are two major monophyletic lineages. The first lineage consists of the Okinawajima, Kourijima, Sesokojima, Tonakijima, Tokashikijima, Iejima, and Akajima populations. This group is diagnosed by the derived acquisition of lineate Systematics of Goniurosaurus kuroiwae 327 o—o yamashinae m—as orientalis 4 o—o kuroiwae O—i toyamai e—e splendens 1 t ‘is -1 oO Fic. 6. First two canonical variates of the DA (C.V.1=78.78% and C.V.2=13.48% of the total variation). Goniwrosaurus kuroiwae splendens from Tokunoshima Island (n=27); G. k. kuroiwae from Okinawajima Island (n=221); G. k. orientalis from Tokashikijima Island (n=50); G. k. yamashinae from Kumejima Island (n=14), G. k. toyamai from Iheyajima Island (n=14). Polygons were constructed by connecting the most peripherally located points of each plot. TABLE 3. Standardized canonical coefficients for the first two variates (C.V.1 and C.V.2) from the multigroup discriminant function analysis (DA) of the nine characters from the Kumejima, Tokashikijima, Okinawajima, Iheyajima, and Tokunoshima populations. Symbols follow those of Table 1. ‘Character C.V.1 C.V.2 SL —0.10 0.70 IL —0.01 0.05 PM 0.70 —0.78 PO 0.02 0.11 EF 0.05 0.16 TU 0.29 0.15 BO 0.03 —0.02 4T 0.25 —0.38 CL —0.92 0.31 tendencies in the dorsal pattern (Fig. 5). Within this group there are two recognizable subgroups. The first is a monophyletic lineage containing the Okinawajima, Kouri- jima, and Sesokojima populations which is diagnosed by the derived acquisition of a dorsal pattern consisting of bands that are incomplete to absent. This group was previously refer- red to as Goniurosaurus kuroiwae kuroiwae (15 (in part), 41, 47| and its monophyly supports its continued recognition. The DA shows that G. k. kuroiwae is nearly completely separated from G. k. yamashinae along the second axis and completely separated from the Tokunoshima population along the first axis (Fig. 6). However, it greatly overlaps the Tokashikijima and Iheyajima populations along both axes (Fig. 6). The second subgroup within this lineage is composed of the Tonakijima, Tokashikijima, Iejima, and Akajima (see below) populations (Fig. 5). Although this group is not demonstrably monophyletic or paraphyletic, it is diagnosable from Goniurosaurus kuroiwae kuroiwae as well as all the other populations (Fig.5). Because Tonakijima Island is the type locality of G. k. orientalis [35], that name has priority for this group and it is given a metataxon designation (G. k. orientalis). The DA shows that G. k. orientalis is completely separated from G. k. yamashinae along the second axis and from the Tokunoshima population along the first axis but that it greatly overlaps the Okinawajima and Iheyajima popula- tions along both axes (Fig. 6). The second major monophyletic lineage also comprises two subgroups and is diagnosed by its derived lack of interspace mottling (Fig.5). The first subgroup is a monophyletic lineage composed of the Tokunoshima popula- tion and diagnosed by the derived acquisition of keeled dorsal tubercles, juxtaposed and sharply raised ventral scales, and the posteriormost transverse body hand extending onto the hind limbs. This population was first described as Eub- lepharis splendens by Nakamura and Uéno [40], and has subsequently been recognized as Goniurosaurus kuroiwae splendens [15, 41, 47, 51]. The evidence presented here for its monophyly supports its continued recognition. The DA shows that G. k. splendens is well separated from G. k. yamashinae along the second axis and from all other popula- tions along the first axis (Fig. 6). The second subgroup of this lineage is composed of the Iheyajima population (Fig.5). This population was placed within G. k. orientalis (sensu Ota [47]) based on its overall similarity in color pattern and scale meristics to the popula- tions of Tonakijima, Tokashikijima, Iejima, and Akajima [47, 52, 56]. It is shown here, however, that the Iheyajima population differs from G. k. orientalis in that it lacks the derived state of lineate tendencies in the dorsal banding pattern that unite G. k. orientalis with G. k. kuroiwae. Furthermore, it has the derived state of a lack of interspace mottling which groups it with G. k. splendens. Thus, con- tinued placement of this population in G. k. orientalis would make the latter demonstrably paraphyletic. Additionally, it is well-separated from G. k. splendens by its lacking the derived character states of keeled dorsal tubercles, juxta- posed and sharply raised ventral scales, and posteriormost body band extending onto the hind limbs. The DA shows that it is completely separated from G. k. yamashinae, along the second axis and from G. k. splendens along the first axis but greatly overlaps G. k. orientalis and G. k. kuroiwae along 328 L. L. Grismer, H. OTA AND S. TANAKA both axes (Fig. 6). It is clear that the Iheyajima population forms a separate monophyletic lineage differing from splendens by its lack of the derived states of characters 1, 2, and 6, and the possession of derived state in character 10. Additionally, its continued recognition as Goniurosaurus kuroiwae orientalis would re- sult in the demonstrative paraphyly of the latter. Thus, separate subspecific recognition is warranted and we consider this population to be: Goniurosaurus kuroiwae toyamai subsp. nov. (Fig. 7) Suggested English name: Iheyajima Leopard Gecko Suggested Japanese name: Iheya-Tokagemodoki Eublepharis kuroiwae orientalis: Toyama, 1984, p. 270 [56] (part). Goniurosaurus kuroiwae orientalis: Ota, 1989, p. 230 [47] (part). Fic. 7. Holotype of Goniurosaurus kuroiwae toyamai KUZ 9983 from Iheyajima Island, Okinawa Prefecture, Japan. Holotype. KUZ 9983, collected by S. Tanaka on Iheya- jima Island, Okinawa Prefecture, Japan, on 4 July 1977. Paratypes. ‘Thirteen paratypes from the same locality as the holotype: KUZ 9978-9982, 9985-9988; TPN 77032201, yamashinae ao _Iheyajima cryin kuroiwae p 9° lejima Tonakijima@— orientalis aly “ee fe) Sesokojima of §2_/ Tokashikijima Akajima o a Okinawajima toyamal ® Kourijima 77070301, 7707401-7707402. Diagnosis. Goniurosaurus kuroiwae toyamai differs from all other subspecies of G. kuroiwae in its overall robust body stature and greater mean number of paravertebral tubercles (37.2:34-42). It differs further from G. k. yamashinae, G. k. orientalis, and G. k. kuroiwae in that adults lack interspace mottling; from G. k. orientalis and G. k. kuroiwae in lacking lineate tendencies in its dorsal pattern and having a lower mean number of eyelid fringe scales (55.6 :53-59); from G. k. orientalis by having a slightly higher mean number of postmental scales (4.6:3-5); from G. k. kuroiwae in having a complete dorsal banding pattern; from G. k. yamashinae by having enlarged scales at the base of its digits, orange-pink juvenile color pattern, blood-red iris, a higher mean number of supralabial scales (9.3 :8-10), and a lower mean number of fourth toe lamellae (17.3 : 16-20); and from G. k. splendens by lacking sharply keeled dorsal tuber- cles, juxtaposed and sharply raised ventral scales, the lateral extension of the posteriormost body bar onto the hind limb, and by having a greater mean number of postmental scales (4.6:3-5), midbody scales (149.8:140-158), fourth toe lamellae (17.3: 16-20), and a lower mean number of scales surrounding the claw on the fourth toe (5.9 :5—6). Distribution. Goniurosaurus kuroiwae toyamai is known only from Iheyajima Island of the Okinawa Group, Ryukyu Archipelago, Japan (Fig. 8). Description of holotype. Adult male; SVL 83.8 mm; head triangular, wider than neck, covered with uniform granular scales interspersed with enlarged tubercles increas- ing in size posteriorly; tubercles absent from rostrum, group of enlarged tubercles immediately anterior to orbit; rostral convex and rectangular, twice as wide as high, middorsal portion partially sutured dorsomedially, bordered laterally by first supralabial and prenasal, dorsolaterally by supraprenasal on left and intercalary scale on right, and dorsally by four enlarged granular scales; external nares subelliptical with splendens Tokunoshima oA Okinoerabujima ess 0 Yoronjima é | Taiwan Fic. 8. Distribution of the subspecies of Goniurosaurus kuroiwae in the Ryukyu Archipelago, Japan. Systematics of Goniurosaurus kuroiwae 329 long axis sloping forward, bordered anteriorly by prenasal and supraprenasal, dorsally and posteriorly by 8(R)-7(L) granular scales, and ventrally by one (R and L) granular scale and prenasals; prenasals with long recurved ventral portion; supraprenasal square, separated medially by four granular scales; supralabials 9(R)-10(L), first two of series square, the remaining rectangular, decreasing in size posteriorly, grading into granular scales along ventral margin of posterior section of upper jaw, posteriormost raised centrally; rostral granules equal in size; preoculars 21(R)-20(L); eyes relatively large, pupils vertical and with convex slightly serrate margins (vis!- ble in life); eyelid fringe scales 59 (R and L), triangular, those of upper eyelid slightly enlarged and conical; outer surface of upper eyelid consisting of small uniform granular scales equal in size to those on top of head, base of upper eyelid bordered by row of enlarged tuberculate scales; 63 scales across top of head between posterior corners of eyes; a fold of skin consisting of granular scales originating in suborbital region extends posteroventrally across angle of jaw; external audi- tory meatus elliptical with long axis directed dorsoventrally, single elongate tubercle bordering anterior margin; tympa- num deeply recessed; mental triangular acutely tapering but rounded at posterior tip, bordered laterally by first infrala- bials and posteriorly by four slightly enlarged postmentals, 10(R)-9(L) infralabials, anteriormost square grading post- eriorly into smaller rectangularly shaped infralabials that grade posteriorly into granular scales bordering dorsal margin of upper jaw; ventral margin of posteriormost infralabials well elevated from surrounding gulars; gular region covered with juxtaposed conical scales interspersed with enlarged tubercles; 56 rows of gulars between postmentals and an imaginary line between posterior margins of auditory meati; gulars grading posteriorly into flat hexagonal subimbricate pectoral scales and larger hexagonal imbricate ventral and interfemoral scales. Neck narrower than body, covered with uniform granu- lar scales, interspersed with several large sharply pointed conical tubercles on nape; tubercles on body conical and prominent, long axes directed posteriorly; body tubercles numerous, distributed evenly on dorsum and increasing in size posteriorly from nape of neck to caudal constriction, grading into distinct repeating caudal whorls; tubercles at caudal constriction twice the size of those on nape, tubercles surrounded by 9-12 granular scales; 36 paravertebral tuber- cles between limb insertions, strict vertebral row absent. Limbs robust, covered with uniform granular scales interspersed with tubercles roughly one-half the size of those on body; granular scales grading distally into slightly flattened, weakly subimbricate scales of dorsal surface of manus and pes; 43(R)-45(L) granular scales around humeral region and 39(R)-38(L) around forearm; hindlimbs roughly twice as thick as forelimbs, covered with uniform granular scales interspersed with enlarged tubercles, those of post- erofemoral region equal in size to those on body; 51(R)-53(L) granular scales around femoral region and 47(R)-48(L) around forelegs; pes covered ventrally with juxtaposed scales, those at heel enlarged; 1-2 enlarged scale(s) at base of each digit; subdigital lamellae narrow, nearly equal in size to slightly smaller lateral digital scales; 18(R)-17(L) subdigital lamellae on fourth toe; digits conical, increasing in length from first to fourth, fifth shorter than fourth. Body robust, covered with granular scales grading ven- trally into flattened, subimbricate ventral scales; 141 granular scales around midbody; ventral interfemoral scales large, flat, imbricate, grading posteriorly into small granular scales anterior to vent; region immediately posterior to vent co- vered with large flat imbricate scales, greatly swollen, with two upward-curving bony spurs arising from lateral margins. Tail conical, thickest at base, covered with small rec- tangular imbricate scales arranged in transverse caudal whorls and repeated series of greatly enlarged and sharply pointed tubercles occurring in caudal whorls; caudal tubercles absent ventrally and decreasing in size laterally and posteriorly; ventral caudals larger and more nearly square than dorsal caudals; posterior one-quarter of tail regenerated and co- vered with slightly raised granular scales. Coloration in life. Ground color of dorsum (including limbs and tail) uniform dark brown; top of head mottled in dark brown and cream; wide dark brown pre- and postorbital stripes; eyelid fringe scales cream-colored giving slight appearance of eye-ring; labial region light-brown to cream- colored; four wide, immaculate, transverse, cream-colored body bands between nape of neck and caudal constriction; bands blend ventrolaterally into light-colored ventrum; back of head bordered by incomplete nuchal band; forelimbs with large cream-colored symmetrical blotches in brachium, elbow, and antebrachium; hind limbs with large cream- colored symmetrical blotches in groin, knee, and dorsopos- terior crus regions; two bands on original portion of tail suffused with reticulate pattern of dark-brown ground color; ventral surface of head, limbs, and body, immaculate light- brown; iris blood-red. Variation. Paratypes closely approximate the holotype in morphology, meristics, and coloration. The most signi- ficant variation occurs in banding pattern and coloration. In three specimens (KUZ 9981-82, 9988), the third transverse body band is incomplete and interrupted medially. Three other (KUZ 9986-87; TPN 77070301) have three instead of four bands between the nuchal band and caudal constriction. All other populations of Goniurosaurus kuroiwae have four bands except its sister taxon G. k. splendens, which always has three. The coloration of the bands in hatchling and juvenile G. k. toyamai is bright orange-pink. This colora- tion may fade to a yellowish to cream-colored banding pattern in adulthood, although some adults maintain the juvenile coloration to some extent. Regenerated tails lack all aspects of banding observed in original tails. Regener- ated tails are mostly dark-brown like the body ground color but become overlain with a reticulum of uneven light-purple uneven blotches. Etymology. This poulation is named in honor of Mr. Masanao Toyama in recognition of his vast number of 330 L. L. Grismer, H. Ora AND S. TANAKA contributions to the herpetology of the Ryukyu Archipelago including the first distributional record of Goniurosaurus kuroiwae from Theyajima Island [56]. DISCUSSION Phylogeny and similarity The classification of Goniurosaurus kuroiwae presented here does not differ greatly from those of Grismer [15] and Ota [47]. It differs from Grismer [15] only in that it consid- ers the populations of Tonakijima and Tokashikijima Islands to be G. k. orientalis rather than G. k. kuroiwae. It differs from Ota [47] in that the population from Kumejima Island is considered to be G. k. yamashinae rather than G. k. orienta- lis, and that the population from Iheyajima Island is consi- dered to be a distinct subspecies G. k. toyamai, rather than G. k. orientalis. Owing to the diagnosability of Goniurosaurus kuroiwae yamashinae (Figs. 5 and 6), it is clearly a lineage in the sense of Frost and Hillis [10]; that is, it is a sexual plexus viewed through time. We are less convinced, however, about the lineage status of G. k. orientalis primarily because of the inclusion of the Iejima population and the distance of that island from Tonakijima, Tokashikijima, and Akajima Islands (Figs. 8 and 9). The latter three islands are in close geog- raphic proximity and situated within a cluster of islands approximately 35 km west of the southwestern tip of Okina- wajima between it and Kumejima Island (Fig. 8). Iejima Island on the other hand, lies only 15 km off the western tip of the Motobu Peninsula of Okinawajima. Populations of G. kuroiwae from Kourijima and Sesokojima Islands, which are also geologically associated with the Motobu Peninsula, are not distinguishable from those of Okinawajima and are considered to be G. k. kuroiwae. The placement of the Iejima Island population in G. k. orientalis is based on the complete dorsal banding pattern of the single known speci- men. More specimens from this island would certainly help to clarify its relationships but unfortunately this population may be greatly reduced or extinct due to habitat alteration. The DA demonstrates the great morphological similarity among Goniurosaurus kuroiwae orientalis, G. k. kuroiwae, and G. k. toyamai. If we were to construct our classification based on the phenetic relationships of the DA, we would recognize G. k. yamashinae and G. k. splendens but would consider G. k. orientalis, G. k. toyamai, and G. k. kuroiwae to all be G. k. kuroiwae. However, it is probable that the apparent meristic similarity among the Tokashikijima, Oki- nawajima, and Iheyajima populations has resulted from independent microevolution in these allopatric populations. Therefore, we prefer to base our classification on the hypoth- esis of phylogenetic relationships. We believe this results in a more meaningful classification because it is based on common ancestry rather than overall similarity, and thus reflects similarities due to the evolution of shared novelties rather than parallel evolution. Species vs. subspecies There recently has been renewed controversy over the use of the subspecies category [2—5, 9-11, 19, 20, 30, 39, 53, 58] as it specifically pertains to herpetology. We believe this issue can be broken down into two general themes. The first concerns the ontology of species and determining wether or not a population represents an individual that should be named and if so, at what taxonomic level. The second is a more general issue concerning the types of information to be retrieved from classifications. In regard to the second question, phylogenetic systemat- ists want a classification that is consistent with recoverable phylogenetic history [10]. Therefore, they are opposed to classifications that recognize demonstrably paraphyletic taxa including subspecific entities that subjectively divide up va- rious sections of a continuous populational cline. Evolution- ary systematists prefer classifications that represent the his- torical extent of phenetic divergence [36]. Thus, they are willing to recognize demonstrably paraphyletic taxa and consider them one of the consequences of evolution. There- fore, dividing up a continuously breeding population into definable subspecies is acceptable because this may represent the early stages of evolutionary divergence. Neither system is right or wrong, they are just different. And each is attempting to construct a classification from which different types of information can be extracted. What needs to be decided is which kind of information has more utility to science in general. It is our opinion that a phy- logenetic classification has a broader utility because it forms the phylogenetic foundation upon which both systematic and non-systematic evolutionary biologists can formulate hypoth- eses as to how and why character states and other biological systems have evolved [e.g., 43, 55, 59]. This does not mean that we are advocating abandoning the subspecies concept, however. In fact, at this point in time, we believe that the subspecies categoly may be able in some situations, to provide a classification with additional phylogenetic informa- tion. However, the advantages and disadvantages of each situation need to be carefully evaluated on a case by case basis. We do believe that demonstrably paraphyletic taxa should not be used in a classification because they obscure more information than they reveal as well as misrepresent history. In regard to the first question, the subspecies category was originally intended to be used for populations of geog- raphic variants of continously interbreeding populations [37, 62] and has subsequently enjoyed a broad use of this type of pattern class designation. The problem, however, is that this type of category in a classification, is often inconsistent with the recoverable phylogenetic history. Because these categorical assignments are not lineage (historically) based, they usually do not represent individuals (sensu [13, 21, 22]) but classes, and can actually distort history. Additionally, such categories often offer little information because the populations being recognized are usually only weakly di- agnosable. It recently has been argued, that if adjacent Systematics of Goniurosaurus kuroiwae populations are interbreeding with one another (e.g. they are not on separate phylogenetic trajectories) they cannot be considered independent evolutionary units [10, 26]. And in such situations, a phylogenetic classification would argue for a single taxon. This criterion is often correct and will go a long way in resolving problems surrounding pattern class taxonomy but it is far too broad. Because of the degrees of differences between varying zones of contact between many diagnosable entities, some of these entities may not be pattern classes or simple geographic variants. They may actually be monophyletic lineages undergoing secondary con- tact in which case their lineage identity should be retained. Because phylogenetic systematics is a retrodictive discipline and not a predictive discipline, there is no way of knowing from just a slice of time (the present) whether or not these populations will subsume one another. Therefore, it is more conservative to base our decisions on what can be inferred from the past. Before any type of an “all or nothing” criterion can be convincingly argued (or if it can be argued at all) concerning the importance of reproductive compatibility in contact zones, many more case studies will need to be examined. We believe that there may very well be discretely diagnosable widely distributed populations (lineages) that narrowly intergrade on the fringes of their ranges with other such adjacent populations and that this intergradation does not preclude them from being on separate phylogenetic trajectories. By abandoning the traditional concept of what a subspe- cies is supposed to represent (geographic variants or pattern classes) and applying the criteria of historical individuality (sensu Kluge [26]) to these smallest evolving lineages, the phylogenetic content of classifications may be augmented. This situation is actually less problematic when dealing with yamashinae orientalis 331 allopatric populations and Goniurosaurus is a good example with which to work. The evolutionary species concept (sen- su Frost and Hillis [10]) would recognize all the different taxa of Goniurosaurus as distinct species because each is diagnos- able and allopatric with respect to one another. This is true, and it clearly indicates that these taxa are on their own separate phylogenetic trajectories. However, this classifica- tion would provide less recoverable phylogenetic information (Fig. 10). All that would exist is a list of seven species of Goniurosaurus with no indication of how any of them are related. However, if these taxa are considered as subspecies of G. kuroiwae and G. lichtenfelderi, then more phylogenetic knowledge is imparted into the classification (Fig. 11). In other words, we know that there are at least two monophyle- tic lineages within Goniurosaurus and that within each of those lineages, there has been additional evolution. And if we want our classifications to be information retrieval systems consistent with recoverable phylogenetic history [10], then this system is clearly providing more information. All the populations of Goniurosaurus could be consi- dered as species and the classification indented accordingly [61] which would provide the same amount of information. However, this classification only would be useful when it is at hand for referral. For example, if a paper was written describing various physiological differences between the insu- lar popoulations of Goniurosaurus under an indented phy- logenetic classification system with no subspecies, we would not know how the taxa were related or the evolutionary implications of the results simply by reading the paper. If, however, the subspecific classification proposed above was used, an evolutionary interpretation of the data becomes more readily apparent because we are able to infer something about relationships from the nomenclature. Now obviously, splendens toyamat oD (eS Fic. 9. Area cladogram of the subspecies of Goniurosaurus kuroiwae. 332 L. L. Grismer, H. OTA AND S. TANAKA Genus Goniurosaurus species lichtenfelderi hainanensis kuroiwae yamashinae orientalis toyamai splendens Fic. 10. Classification of Goniurosaurus on the left recognizing all populations as species and the amount of recoverable phylogenetic information from that classification on the right. Genus Goniurosaurus species lichtenfelderi subspecies lichtenfelderi hainanensis species kuroiwae subspecies kuroiwae yamashinae orientalis toyamai splendens Fic. 11. Classification of Goniurosaurus on the left using subspecies and the amount of recoverable phylogenetic information from that classification on the right. this information could be obtained by reviewing the original phylogeny but unfortunately this is far too often neglected by non-systematic evolutionary biologists (see Dial and Grismer [7)). The disadvantage in recognizing these taxa as subspecies lies primarily in the original connotation of the subspecies category: that of intergrading geographic variants or pattern classes, which is something these populations are not. In fact, it is the historical inertia of this connotation that is the major disadvantage to considering the name “subspecies” to replace the name Kluge [26] uses for lineage individuality (“species”). However, we believe at this point it is still best to emphasize the phylogeny of this group by way of its nomenclature in light of the disadvantages that the subspecies category carries. It may be that in the future, this disadvan- tage will be too much to overcome by just a few workers and these populations will best be recognized as different species. Biogeography The distribution of Goniurosaurus kuroiwae throughout Systematics of Goniurosaurus kuroiwae 333 the Ryukyu Archipelago seems to be rather incomplete. There are large gaps between insular localities within which there are islands with well suited habitat where G. kuroiwae is apparently absent (Fig. 8). This absence appears not to be an artifact of collecting because the herpetology of these islands is known quite well [46, 57]. The best example of this is the presence of G. k. splendens and G. k. toyamai on Tokunoshima and Theyajima Islands, respectively, and the absence of G. kuroiwae from the geographically intermediate Okinoerabujima Island (Fig. 8). The geological and tectonic history of the Ryukyu Archipelago indicates that the backbone of the Ryukyus was uplifted during the late Miocene as a result of oceanic crust from the Philippine Plate being thrust below the continental crust of Asia [27, 29, 31]. Subsequently, the Ryukyus have had two separate landbridge connections to Taiwan and continental China [23, 24]. The first of these lasted from the late Miocene to the early Pliocene and the second occurred in the early Pleistocene. The majority of the terrestrial reptiles from the Ryukyu Archipelago have their closest relatives occurring in Taiwan and Fukien, the nearest coast on the Chinese continent. Thus, it is assumed here that their ancestors dispersed through the landbridge from the latter regions during the presence of the second landbridge in the early Pleistocene [24, 48, 54]. On the other hand, a few Ryukyu species do not have sister taxa in Taiwan and Fukien and such a relictual distribution pattern is interpreted here as being indicative of a more ancient entry into the Ryukyu Archipelago [23, 24]; perhaps during the first landbridge connection in the late Miocene. The absence of eublepharid geckos from Taiwan and other surrounding regions and the presence of other Goniurosaurus from the Gulf of Tonkin (G. lichtenfelderi) and southern China (Goniurosaurus sp. [14, 15, 33]) may indicate that the ancestral form of G. kuroiwae entered the Ryukyus from continental China during the first landbridge formation. With the submergence of the first landbridge in the middle Pliocene, a large island was formed connecting Kume- jima Island of the south to Amamioshima Island of the extreme north [23, 24]. Such a large island would have effectively isolated this ancestral Goniurosaurus and presum- ably promoted the evolution of the ancestor of G. kuroiwae. Despite the reunion of this island with Taiwan and continen- tal China during the early Pleistocene, the ancestor of G. kuroiwae presumably remained in the Ryukyus probably due to various ecological factors and did not reinvade the conti- nent. Rising sea levels in the middle Pleistocene resulted in the submergence of the second landbridge. This also began the formation of the contemporary configuration of the Ryukyu Archipelago by creating primary divisions within its central portion. Initially five island groups were formed: Kumejima Island; an island which composed of Tonakijima, Akajima, Tokashikijima, and the other geographically proximate is- lands; an island composed of Okinawajima, Kourijima, Seso- kojima, and lejima Islands; Iheyajima Island; and Toku- noshima Island [23, 24, 28] (Fig. 8). Unfortunately, there is very little geological data regarding the relative sequence of the isolation of these islands although the biological data presented here (i.e. the phylogenetic relationships) would suggest that it occurred from south to north (Fig. 9). It is highly probable that all the above islands except Iheyajima and Tokunoshima Islands were re-connected with each other by land bridges during the most recent continental glaciation (ca 15000-18000 yr ago) when the sea level drop- ped by no less than 120m [49]. This, however, does not seem to have provided good opportunities for gene flow or range extension for most forest-dwelling reptiles especially on Kumejima Island. The reason is likely due to the very short duration of the connection or the absence of appropriate habitat which would have offered favorable environments on the bridges. Presence of the endemic montane snake, Opis- thotropis kikuzatoi, on Kumejima [56, 57] and the large genetic divergences between Kumejima, Tokashikijima, and Okinawajima populations of several forest-dwelling species of reptiles revealed by biochemical studies (Ota in prep.), seem to support for this hypothesis. Subsequent rising of the sea level near the end of the Pleistocene would have divided this land mass into the current islands of Kumejima, Tonakijima, Akajima, Tokashikijima, and their surrounding islands, as well as Okinawajima, Sesokojima, Kourijima, and Iejima Islands. The paleogeographical scenario seems to largely coincide with the evolution of various lineages within Goniurosaurus kuroiwae (Fig. 5), although it does not offer any corrobora- tive data as to the chronological sequence in which these lineages evolved. The classification presented above gener- ally reflects the paleogeography of the islands except for the assignment of the Iejima population to G. k. orientalis. No evidence suggests that this island was connected to Tonaki- jima, Akajima, and Tokashikijima Islands to the exclusion of Okinawajima Island [23, 24, 49]. Moreover, similarities in both extant and fossil faunas imply that Iejima Island is historically related to Okinawajima [45]. Thus, it is likely that the apparent similarity of the Iejima population to the Tonakishima, Akajima, and Tokashikijima populations is the result of parallel evolution and not dispersal (contra Shimo- jana [52]). ACKNOWLEDGMENTS We wish to express our gratitude to Y. Chigara of the Okinawa Prefectural Museum (OPM), S.-I. Uéno of the National Science Museum (Nat. Hist), Tokyo (NSMT), J. Vindum of the California Academy of Sciences (CAS), C. Myers of the American Museum of Natural History (AMNH), E. N. Arnold of the British Museum (BM), H. Marx of the Field Museum of Natural History (FMNH), J. Wright of the Natural History Museum of Los Angeles County (LACM), I. Darevsky of the Zoological Institute, Academy of Sciences, Leningrad (ZIL), and P. Alberch of the Museum of Comparative Zoology (MCZ) for the loan of specimens. For field assistance we wish to thank H. Hoshikawa, Y. Kane, K. Komemoto, and H. Tomiyama. Special thanks are due to M. Toyama for providing specimens and color photographs of Goniurosaurus 334 L. L. Grismer, H. Ora AND S. TANAKA kuroiwae from various lacalities, and for the permission to name the subspecies in his honor. For helpful comments on the manuscript we wish to thank R. Etheridge, J. McGuire, B. Hollingsworth, D. Archibald, M. Simpson, and two anonymous reviewers. The work of H. Ota was partially supported by Grants-in-Aid from the Japan Ministry of Education, Science, and Culture (A-63790257, 05740523; to Ota), and the U.S. National Geographic Society Grant (No. 4505-91; to M. Matsui). Goniurosaurus kuroiwae has been designated as a natural monu- ment of Okinawa Prefecture and the handling of this species is strictly regulated by law. This research was carried out under the permis- sion from the Section of Culture, Okinawa Prefectural Government. 12 REFERENCES Anderson SC, Leviton AE (1966) A new species of Eublephar- is from southwestern Iran (Reptilia: Gekkonidae). Occ Pap California Acad Sci 53: 1-5 Cole CJ (1990) When is an individual not a species? Herpeto- logica 46: 104-108 Collins JT (1991) Viewpoint: a new taxonomic arrangement for some North American amphibians and reptiles. Herpetol Rev 22: 42-43 Collins JT (1992a) Reply to Grobman on variation in Opheod- rys aestivus. Herpetol Rev 23: 15-16 Collins JT (1992b) The evolutionary species concept: a reply to Van Devender et al. and Montanucci. Herpetol Rev 23: 43- 46 de Queitoz K, Donoghue MJ (1988) Phylogenetic systematics and the species problem. Cladistics 4: 317-338 Dial BH, Grismer LL (1992) A phylogenetic analysis of phy- siological-ecological character evolution in the lizard genus Coleonyx and its implications for historical biogeographic recon- struction. Syst Biol 41: 178-195 Donoghue MJ (1985) A critique of the biological species concept and recommendations for a phylogenetic alternative. Bryologist 88: 172-181 Echelle AA (1990) In defense of the phylogenetic species concept and the ontological status of hybridogenetic taxa. Her- petologica 46: 109-113 Frost DR, Hillis DM (1990) Species in concepts and practice: herpetological applications. Herpetologica 46: 87-104 Frost DR, Kluge AG, Hillis DM (1992) Species in contempor- ary herpetology: comments on phylogenetic inference and tax- onomy. Herpetol Rev 23: 46-54 Gauthier J, Estes R, de Queiroz K (1988) A Phylogenetic analysis of Lepidosauromorpha. In “Phylogenetic Rela- tionships of the Lizard Families” Ed by R Estes, G Pregill, Stanford University Press, Stanford, California, pp 15—98 Ghiselin MT (1987) Species concepts, individuality, and objec- tivity. Biol Phylos 2: 127-143 Grismer LL (1987) Evidence for the resurrection of Goniuro- saurus Barbour (Reptilia: Eublepharidae) with a discussion of geographic variation in Goniurosaurus lichtenfelderi. Acta Herpetologica Sinica 6: 43-47 Grismer LL (1988) The phylogeny, taxonomy, classification, and biogeography of eublepharid geckos (Reptilia: Squamata). In “Phylogenetic Relationships of the Lizard Families” Ed by R Estes, G Pregill, Stanford University Press, Stanford, California, pp 369-469 Grismer LL (1989) Eublepharis ensafi Baloutch and Thireau, 1986: A junior synonym of E. angramainyu Anderson and Leviton, 1966. J Herpetol 23: 94-95 Grismer LL (1991) Cladistic relationships of the lizard Eub- lepharis turcmenicus (Squamata: Eublepharidae). J Herpetol 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 25: 251-253 Grismer LL, Ottley JR (1988) A preliminary analysis of geog- raphic variation in Coleonyx switaki (Squamata: Eublepharidae) with a description of an insular subspecies. Herpetologica 44: 143-154 Grobman A (1992) On races, clines, and common names in Opheodrys. Herpetol Rev 23: 14-15 Highton R (1990) Taxonomic treatment of genetically diffe- rentiated populations. Herpetologica 46: 114-121 Hull DL (1976) Are species really individuals ? Syst Zool 25: 174-191 Hull DL (1980) Individuality and selection. Syst 11: 311-332 Kizaki K, Oshiro I (1977) Paleogeography of the Ryukyu Islands. Marine Sciences Monthly 9: 542-549 (in Japanese) Kizaki K, Oshiro I (1980) The origin of the Ryukyu Islands. In “Natural History of the Ryukyus” Ed by K Kizaki, Tsukiji— Shokan, Tokyo, pp 8-37 (in Japanese) Kluge AG (1989) Metacladistics. Cladistics 5: 291-294 Kluge AG (1990) Species as historical individuals. Biol Phi- los 5: 417-431 Koba M (1980) Distribution and age of the marine terraces and their deposits in the reef-capped Ryukyu Islands, Japan. Quart Res 18: 189-208 (in Japanese with English abstract) Konishi K (1965) Geotectonic framework of the Ryukyu Is- Ann Rev Ecol lands. J Geol Soc Japan 71: 437-457 (in Jupanese with English abstract) Konishi K, Sudo K (1972) From Ryukyus to Taiwan. Kagaku 42: 221-230 (in Japanese) Lazell J (1992) Taxonomic tyranny and the esoteric. Herpe- tol Rev 23: 14 Lee C, Shor GC, Bibee LD, Lu RS, Hilde TWC (1980) Okinawa trough: origin of a back-arc basin. Marine Geol 35: 219-241 Leviton AE, Gibbs RH Jr., Heal E, Dawson CE (1985) Stan- dards in herpetology and ichthyology: Part I. Standard symbo- lic codes for institutional resource collections in herpetology and ichthyology. Copeia 1985: 802-832 Li D, Wang J, Zhao Z (1985) Reptilian survey of karst in Libo, Guizhou. Acta Herpetologica Sinica 4: 140-143 (in Chinese with English abstract) Maddison WP, Donoghue MJ, Maddison DR (1984) Outgroup analysis and parsimony. Syst Zool 33: 83-103 Maki M (1930) A new banded gecko, Eublepharis orientalis, sp. nov. from Ryu Kyu. Annot Zool Japon 13: 9-11 Mayr E (1969) Principles of Systematic Zoology. McGraw Hill, New York Mayr E (1982) Of what use are subspecies ? Mocquard F (1897) Notes herpetologiques. Nat Paris 5: 211-217 Montanucci RR (1992) Commentary on a proposed taxonomic arrangement for some North American amphibians and reptiles. Herpetol Rev 23: 9-10 Nakamura K, Uéno S-I (1959) The geckos found in the limes- tone caves of the Ryu Kyu islands. Mem Coll Sci, Kyoto Univ (ser B) 26: 42-52 Nakamura K, Uéno S-I (1963) Japanese reptiles and amphi- bians in colour. Hoikusha, Osaka (in Japanese) Namiye M (1912) On the gekkonid lizards of Okinawa. Mag (Tokyo) 24: 442-445 (in Japanese) O’Hara RJ (1988) Homage to Clio, or, toward historical philosophy for evolutionary biology. Syst Zool 37: 142-155 Okada Y (1936) A new cave gecko, Gymnodactylus yamashi- nae from Kumejima, Okinawa Group. Proc Imp Acad, Japan 6: 71-73 Auk 99: 593-595 Bull Mus Hist Zool 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 Systematics of Goniurosaurus kuroiwae 335 Oshiro I, Nohara T (1990) Vertebrate fossils in Okinawa. In “Nature of Okinawa-Geomorphology and Geological Features” Ed by H Ujiie, Hirugi-sha, Naha, pp 213-229 (in Japanese) Ota H (1986) A review of reptiles and amphibians of the Amami Group, Ryukyu Archipelago. Mem Fac Sci, Kyoto Univ (Ser Biol) 11: 57-71 Ota H (1989) A review the geckos (Lacertilia: Reptilia) of the Ryukyu Archipelago and Taiwan. In “Current Herpetolo- gy in East Asia” Ed by M Matsui, T Hikida RC Goris, Herpetological Society of Japan, Kyoto, pp 222-261 Ota H (1991) Systematics and biogeography of terrestrial reptiles of Taiwan. In “Proceedings of the International Sym- posium of Wildlife Conservation” Ed by Y-S Lin, K-H Chang, Council of Agriculture, Taipei, pp 47-112 Ota H, Sakaguchi N, Ikehara S, Hikida T (1993) The herpeto- fauna of the Senkaku Group, Ryukyu Archipelago. Pacific Sci 47: 248-255 Pinmentel RA, Smith JD (1986) Bio tat II. A multivariate statistical toolbox. Sigma Soft, Placentia, California Sengoku S (1979) Amphibians and Reptiles in Color. Hikari Kyokai, Tokyo (in Japanese) Shimojana M (1979) Cave animals of Okinawa and adjacent islands. In “Survey Report of the Caves of Okinawa Prefecture Il” Ed by Educational Committee of Okinawa Prefecture, Educational Committee of Okinawa Prefecture, Naha, pp 97- 153 (in Japanese) Smith HM (1990) The universal species concept. gica 46: 122-124 Takara T (1962) Studies on the terrestrial snakes of the Ryukyu Archipelago. Sci Bull Agr Home Econ Div, Univ Ryukyus 9: 1-202 (in Japanese with English summary) Thope RS (1987) Geographic variation: a synthesis of cause, data, pattern and congruence in relation to subspecies, multi- variate analysis and phylogenesis. Boll Zool 54: 3-11 Toyama M (1984) Amphibians and reptiles. In “A Guide to Animals of Okinawa, Vol. 1: Terrestrial Vertebrates” Ed by S Ikehara, Y Yonashiro, K Miyagi, M. Toyama, Shinsei-tosho, Naha, pp 209-323 (in Japanese) Toyama M, Ota H (1991) Amphibians and reptiles of the Ryukyu Islands. In “Study of Essential Factors for Preserva- tion of Wildlife in Nansei Islands” Ed by World Wildlife Fund Japan, Japan Agency of Environment (Nature Conservation Department), Tokyo, pp 233-254 (in Japanese) Van Devender TR, Lowe CH, McCrystal H, Lawler H (1992) Viewpoint: reconsider suggested systematic arrangements for some North American amphibians and reptiles. Herpetol Rev 23: 10-14 Wantop H-E (1983) Historical constraints in adaptation theory: traits and non-traits. Oikos 41: 157-160 Watrous LE, Wheeler OD (1981) The out-group comparison method of character analysis. Syst Zool 30: 1-11 Wiley EO (1981) Phylogenetics: the theory and practice of phylogenetic systematics. John Wiley and Sons, New York Wilson EO, Brown WL Jr (1953) The subspecies concept. Syst Zool 2: 97-111 Te no Herpetolo- APPENDIX I Specimens Examined All abbreviations follow Leviton et al. [32] except those as follows: KUZ-Department of Zoology, Faculty of Sciences, Kyoto Universi- ty, Japan; LLG-L. Lee Grismer, Department of Biology, San Diego State University, San Diego, CA 92182, USA; and TPN-Satoshi Tanaka, Motobu Senior High School, Tokuchi 377, Motobu, Okina- wa 905-02, Japan. Goniurosaurus kuroiwae kuroiwae: JAPAN-RYUKYU AR- CHIPELAGO: OKINAWAJIMA ISLAND; Naha (KUZ 10005-06, TPN 76042508, 76042802, 77041005, 7711060-611, 78011101, 78011401, 78020201, 78031801-05, 78041401-09, 78051601-09, 78060502, 78061501-09, 78061801, 78062501, 78063001, 78070301, 78070401, 78071401-10, 78071701, 78071902, 78072001-02, 78072101, 78072201, 78081401-11, 78081601, 78091601, 78091701- 07, 78100601, 78091701-07, 78100601, 78102301-09, 79061001, 79070201, 79070301-02, 79072801, 79082601), Seihua (KUZ 7955, 7957-58, 8002, 8004-05, 8051-53, TPN 6091507-13, 76102111-13, 71040802-05, 77040807-—42, 78051701-02, 78052202, 78052701-02, 78052901-04, 78062201-02, 78062401, 7806201-03), Chinen (KUZ 10007), Yakena (TPN 77041701), Oyakebaru (TPN 77121001), Yonaha (TPN 76052401, 76052406, 78032701-03), Yona (TPN 77041401, 77050201-04, 77052101, 77060901, 77081804—12, 77081911-12, 77082102-04, 78042302), Hentona (TPN 78042305- 08), Haneji (TPN 79081101-03). KOURIJIMA ISLAND; OPM 424. SESOKOJIMA ISLAND; OPM 369. Goniurosaurus kuroiwae orientalis: JARAN-RYUKYU ARCHIPELAGO: AKA- JIMA ISLAND; OPM 341. IEJIMA ISLAND; KUZ 9991. TOKASHIKIJIMA ISLAND; KUZ 7976, OPM 325, NSMT 2523- 24, TPN 77101501, 77111701, 77120701-05, 78030101—-03, 78052001- 03, 78052101-09, 78052201, 78060501, 78061101, 78061401, 78061601, 78071101, 78071201, 78071901, 78072003, 78080501-02, 78080508-10, 78080512—15, 78090202—05, 79082602-03. TONAKI- JIMA ISLAND; OPM 489, NSMT 02522. Goniurosaurus kuroiwae splendens: JAPAN-RYUKYU ARCHIPELAGO: TOKUNOSHI- MA ISLAND; Kedoku (KUZ 8408-11, OPM 10, LLG 1180, 1251- 52, NSMT 2514-20, 3211-12, TPN 76102201-04, 77041001-04, 77120901). Goniurosaurus kuroiwae toyamai: JAPAN-RYUKYU ARCHIPELAGO: THEYAJIMA ISLAND; KUZ 9978-83, 9985- 88, TPN 77032201, 77070301, 77070401-02. Goniurosaurus kuroiwae yamashinae: JAPAN-RYUKYU ARCHIPELAGO: KUMEJIMA ISLAND; KUZ 9989, 9998, 13249, OPM 3, 8, 9, TPN 78050401-02, six uncatalogued specimens. Eublepharis angra- mainyu: CAS 86333, 86337, 86361—-62, 86366, 86381, 86383, 86385, 86396-98, 86416, 157129. Eublepharis harwickii: AMNH 57593, BM 1927.8.9.1, 1962.23, 1962.38. Eublepharis macularius: AMNH 57594, CAS 96212, 96245, 101440-41, 104361-63, 133826, FMNH 161142, LACM 109902-03. LLG 1393-94, 131-38. Eublepharis turcmenicus: ZIL 10103, 15380, 19414. Hemitheconyx caudicinctus: AMNH 104409, CAS 55114, 154299-302, LLG 2119-96. Hemithe- conyx taylori: BM 1946.8.26.57-59, 1946.8.26.72, 12.5.372-74. Holodactylus africanus: CAS 125431, MCZ 21663, 38693, 53593, 77365-67, 77369-70, 96928, 160707. Holodactylus cornii: BM 1931.7.20.269. ZOOLOGICAL SCIENCE 11: 337-342 (1994) © 1994 Zoological Society of Japan Polymorphism of Lampbrush Chromosomes in Japanese Populations of Rana nigromaculata Hiromi OHTANI Laboratory for Amphibian Biology, Faculty of Science, Hiroshima University, Higashi-hiroshima 724, Japan ABSTRACT— The 13 pairs of the lampbrush chromosomes of R. nigromaculata are characterized by one to five landmarks situated at specific positions on the axis of each chromosome. In R. nigromaculata collected from 28 sites in Japan, eight chromosome pairs did not show any variations, whereas the remaining five pairs consisted of two forms of chromosome which differed in the number or type of the landmarks. The variations in chromosomal polymorphism indicated that Japanese R. nigromaculata has genetically differentiated into four groups which were formed by a break in the migration due perhaps to geographic obstacles and the expansion of new genetic materials provided from continental R. nigromaculata in the Wurm glacial stage. By comparing the distribution of the variations and the lampbrush chromosomes of continental R. nigromaculata, the history of the migration of R. nigromaculata into Japan is discussed. INTRODUCTION Rana nigromaculata is distributed over the north-eastern area of China, the whole of the Korean Peninsula, and Kyushu, Shikoku and Honshu (except the Kanto and Sendai plains) in Japan [5]. This species is supposed to have evolved from an ancestral species population common to Rana plancyi in the continent, to adapt to the arid and cold climates [6]. After its own evolution was sufficiently com- pleted, R. nigromaculata came to Japan through the Korean Peninsula when Japan was still a part of the continent [6]. Recently, Nishioka et al. [12] found that R. nigromaculata collected from 45 sites in Japan have differentiated into four groups on the basis of Nei’s genetic distances obtained from electrophoretic analyses. The genetic differentiation was reflected in the character- istics of its lampbrush chromosomes which are composed of 13 pairs of bivalent chromosomes at the diplotene stage of oogenetic meiosis. The lampbrush chromosomes of R. nig- romaculata are characterized by one to five landmarks, consisting of simple-type giant loops, compound-type giant loops, spheres, and an oval-like structure, at specific positions of each chromosome axis [10], though the landmarks grow conspicuous by accumulating their own gene product around their axes [1]. However, some lampbrush chromosomes showed variations in the landmarks among populations of R. nigromaculata. These variations are presumed to relate to the processes of the habitat expansion of R. nigromaculata in Japan in the same way as the mitotic chromosomal mutations of Rattus species [15-17]. Herein, the geographical distribution and the history of the polymorphism of the lampbrush chromosomes of Japanese R. nigromaculata in comparison with those of continental species are described. Accepted January 19, 1994 Received October 26, 1993 MATERIALS AND METHODS Lampbrush chromosomes were removed from the ova of 199 female Rana nigromaculata Hallowell collected from 28 sites in Japan. Figure 1 shows the collection sites and the number of females. In addition, lampbrush chromosomes were removed from three females from Beijing, China, and six females from Suwon, Korea, which had been bred in the Laboratory for Amphibian Biology, Hiroshima University. Lampbrush chromosomes were prepared by means of a slight modification of Gall’s method [2, 13]. Because the size of the landmarks varies with the stages of oogenesis, ten or more lampbrush chromosome preparations per female were examined under a phase- contrast microscope to avoid misjudgements on the presence or absence of landmarks. RESULTS Previously, Nishioka et al. [10] described a map of the lampbrush chromosomes of R. nigromaculata, which repre- sented the positions and types of landmarks on each chromo- somal axis. The map was constructed from the lampbrush chromosomes of three female offspring of a pair collected from Hiroshima. According to that map, 13 lampbrush chromosomes possess one to five landmarks each, or a total of 35 landmarks, and they are all distinguishable by position and type. In this study, eight of the 13 lampbrush chromosomes of all the populations of Japanese R. nigromaculata examined showed the same characteristics (2-5, 8-10, and 12). By contrast, the remaining five, 1, 6, 7, 11, and 13, showed two forms which differed in the number or type of the landmarks they carried. In chromosome 1, one form possessed two simple-type giant loops, one on the short arm and the other on the long arm; the other form possessed one compound- type giant loop on the long arm in addition to these (Fig. 2a). In chromosome 6, one form had one compound-type and one simple-type giant loop on the long arm; the other had another H. OHTANI Fukui(10) Kanazawa(10) Toyama(9) Joetsu(15) Y amaguchi(10) Shibata(7) 2%: Hiroshima(17) Sakata(5) Bee “ Gotsu(5) Akiva(3)/ /: soccer ins _Konko(4) Ss : oP : 2 [.Tottori(10) . ws piano B - Himeji(5) ope Cita) Shingu(5) Izu ‘/ Munakata(9) Takamatsu(5) | Sule Kumamoto(10) Nangoku(2) de is peated Kagoshima(10) Matsuyama(3) aya(5) Sasebo(10) ee He Maibara(4) Fic. 1. Map showing the collection sites of Rana nigromaculata and the number of females studied. Bold lines mark the mountain ranges and a dotted line marks the location of the presumptive coastline about 2 10* years ago [3]. Fic. 2. Variations in lampbrush chromosomes 1(a), 6(b), 7(c), 11(d), and 13(e) of R. nigromaculata. In diagrammatic representation, simple- and compound-type giant loops are represented by solid and dotted lines, respectively, and spheres A segment between two short parallel lines shows a range consisting of larger normal loops are shown with black circles. In than any other parts and including a centromere. A, B, and C indicate A-, B-, and C-form, respectively. photographs, simple-type giant loops appear as a stiff loop which becomes thick by covering its axis with a large quantity of matrixes and it has a smooth outline. Compound-type giant loops consist of two or more simple-type giant loops which are more slender than the independent simple-type giant loops and have a notched outline. Scale bar represents 100 ~m. Polymorphism of Lampbrush Chromosomes 339 simple-type giant loop on the long arm in addition to these (Fig. 2b). In chromosome 7, one form had two simple-type and one compound-type giant loop on the long arm; the other had another compound-type giant loop on the short arm in addition to these (Fig. 2c). In chromosome 11, one form had one compound-type giant loop on the long arm and two spheres on both the arms; the other had three simple-type giant loops in place of the compound-type giant loop of the former (Fig. 2d). In chromosome 13, one form had one compound-type giant loop on the short arm and one simple- type giant loop on the long arm; the other had another simple-type giant loop on the short arm in addition to these (Fig. 2e). In each of these five chromosomes, the former was named the A-form, and the latter was the B-form. Based upon the frequencies of the A- and B-forms in each of the 28 collection sites, the R. nigromaculata popula- tions were divided into four groups, named the Eastern Honshu, Chubu, Western Honshu-Shikoku, and Kyushu groups, respectively (Fig.3; Table 1). In chromosome 1, the A-form was predominant in the Chubu group, while the B-form was predominant in the Eastern Honshu and Western Honshu-Shikoku groups. Inthe Kyushu group, the frequen- cy of the B-form was a little higher than that of the A-form. It was remarkable, however, that nine of the 10 females from Kagoshima had the homologous A-form. Western Honshu-Shikoku group >» a ce @) wey Yamaguchi. R, Hiroshima. S,Konko. T, Himeji. Y, Sasebo. Z, Kumamoto. a, Oita. b, Kagoshima. In chromosome 6, the two forms were somewhat similar in distributional pattern to those of chromosome 1. The A-form was found in the Chubu group at a fairly high frequency, not being so predominant as the A-form of chromosome 1. The B-form was predominant in the East- ern Honshu and Western Honshu-Shikoku groups, like chromosome 1. The Kyushu group also involved the B-form at high frequency. It was remarkable, however, that five females from Kagoshima were homozygous for the A-form and four were heterozygous for the A- and B-forms. In chromosome 7, all the females of the Eastern Honshu, Chubu, and Western Honshu-Shikoku groups had the only A-form. On the other hand, the Kyushu group had the B-form at a nearly equal frequency to that of the A-form. It was remarkable, however, that all females from Kagoshima had the homologous A-form. In chromosome 11, the B-form was distributed in a similar manner to the A-form of chromosome 7. The B- form was found in all the females of the Eastern Honshu, Chubu, and Western Honshu-Shikoku groups in the homolo- gous condition. The Kyushu group involved both the A- and B-forms at nearly equal frequencies. In chromosome 13, all the females of the Eastern Hon- shu and Chubu groups were homologous for the A-form. On the other hand, most of the females of the Western Eastern Honshu group Chromosome no. as Form Fic. 3. Frequencies of the A- and B-form of five lampbrush chromosomes, 1, 6, 7, 11, and 13, in each collecting site. Areas partitioned with broken lines represent groups of R. nigromaculata females with about the same characteristics in their lampbrush chromosomes. A, Akita. B, Sakata. C, Shibata. D, Joetsu. E, Toyama. F, Kanazawa. G, Fukui. H, Okaya. I, Sutama. J, Mishima. K, lida. L, Nagoya. M, Maibara. N, Shingu. O, Tottori. P, Gotsu. Q, U, Takamatsu. V,Nangoku. W, Matsuyama. X, Munakata. 340 H. OnTANI TABLE 1. Frequencies of bivalent chromosomal types in four groups of R. nigromaculata Figures in parentheses show expected values based on Hardy-Weinberg hypothesis. Type Eaten Chubu Wes Gat Wousbe Kyushu Total Chromosome 1 AA 0 31 1 19 51 ( 0.2) (31.0) ( 0.4) ( 9.7) AB 6 1 8 7 22 @7) ( 1.0) ( 9.1) (25.5) BB 53 0 47 26 126 (53.2) ( 0.0) (46.4) (16.7) Total 59 32 56 52 199 Chromosome 6 AA 0 12 0 5 17 (10.1) ( 0.0) ( 1.7) AB 0 12 2 9 23 (15.8) ( 2.0) (15.5) BB 59 8 54 38 159 ( 6.1) (54.0) (34.7) Total 59 32 56 52 199 Chromosome 7 AA 59 32 56 18 165 (14.0) AB 0 0 0 18 18 (26.0) BB 0 0 0 16 16 (12.0) Total 59 32 56 52 199 Chromosome 11 AA 0 0 0 17 17 (11.5) AB 0 0 0 15 15 (25.9) BB 59 32 56 20 167 (14.5) Total 59 32 56 52 199 Chromosome 13 AA 59 32 4 4 99 (@320) ( 1.9) AB 0 0 18 12 30 (20.0) (16.2) BB 0 0 34 36 70 (33.0) (33.9) Total 59 32 56 52 199 Honshu-Shikoku and Kyushu groups were homozygous for the B-form. The lampbrush chromosomes of continental R. nigroma- culata were identical in characteristics with those of Japanese R. nigromaculata, except for chromosome 1. The character- istics of chromosomes 6, 7, 11, and 13, in which Japanese R. nigromaculata had variations, were of the A-form. Howev- er, of the six females from Suwon, two had the homozygous and heterozygous B-form in chromosome 6 and one was homozygous for the B-form in chromosome 11. Chromo- some 1 of all the females from Beijing and Suwon had one simple-type giant loop on the short arm and one compound- type giant loop on the long arm (Fig. 2a). This was named the C-form. Of the two landmarks, the former was also found in the A- and B-forms, and the latter in the B-form. In chromosome 13 of three females from Beijing, there was a variation which was not found in Japanese R. nigromaculata. This variation had two compound-type giant loops on the short and long arms and was named the C-form (Fig. 2e). Two females were homologous for the C-form and one was heterologous for the A- and C-forms. DISCUSSION Nishioka et al. [12] found that R. nigromaculata collected from all over Japan can be divided into four groups by a cluster analysis of the genetic distances between local popula- tions. The geographic areas of the four groups are as Polymorphism of Lampbrush Chromosomes 34] follows: 1) the Tohoku area comprising the prefectures of Aomori, Akita, and Niigata; 2) the Chubu area comprising the prefectures of Shizuoka, Aichi, Nagano, and Yamanashi; 3) the Hokuriku, Kinki, Sanyo, Shikoku, and Kagoshima areas comprising the prefectures of Toyama, Ishikawa, Fukui, Shiga, Mie, Wakayama, Osaka, Hyogo, Okayama, Hiroshima, Kagawa, Kochi, Ehime, and Kagoshima; and 4) the San-in and Kyushu areas comprising the prefectures of Tottori, Shimane, Yamaguchi, Fukuoka, Nagasaki, Kuma- moto, Oita, and Miyazaki. In this study, it was found that the spread of variations in five lampbrush chromosomes, 1, 6, 7, 11, and 13, is also coincident with this grouping, though there were slight discrepancies in dividing lines. In the four groups based on the chromosomal variations, the geographic- al area of the Eastern Honshu group comprises the Tohoku and Hokuriku areas, that of the Chubu group comprises the Chubu area, that of the Western Honshu-Shikoku group comprises the Kinki, Sanyo, San-in (except Yamaguchi Pre- fecture) and Shikoku areas, and that of the Kyushu group comprises the Kyushu area and Yamaguchi and Kagoshima Prefectures. In each group, the frequencies of the forms of lampbrush chromosomes agreed well with expected values based on the Hardy-Weinberg hypothesis (P>0.05), except for chromo- somes 1, 6, and 11 of the Kyushu group. This seems to show that three groups other than the Kyushu group are geographi- cally isolated from one another by major mountain ranges. The Kyushu group, however, is not partitioned from the adjoining group by a clear topographical obstacle, since a part of this group encroaches on the western Honshu area. Moreover, this group seems to include a heterogeneous subgroup, because in chromosomes 1 and 6 many R. nigroma- culata from Kagoshima had the A-forms which were also found in the Chubu group. The Kyushu group may be divided into two subgroups by increasing the sites of examina- tion further. It seems that the estimation of the period when the ancestral population of R. nigromaculata invaded Japan gives an important clue to the process of the formation of the four groups. Japan was a part of the continent up to the end of the Riss glacial stage [4]. It is reasonably certain that the complete separation of Japan occurred in the Riss-Wiirm interglacial stage (about 0.12 million years ago) [8]. After entering Japan, R. nigromaculata is supposed to have moved along the coastal plains without passing over the highly upheaved mountain ranges; the coastal plains which had widened owing to marine regression in the glacial periods were more favorable. The population which had reached the Chubu area, however, could not go further than this area. The reason is that a collision between the central Honshu region and Izu island (presently a peninsula) broke down the coastal plain (about 0.5 million years ago) [7]. The difficulty of migration in this region is evidenced by the cluster analyses of genetic distances in the populations of Rana rugosa, Rana Japonica, and Rana brevipoda [9, 11, 12]. The populations of these species in the Kanto and Sendai plains, which are situated in the Pacific side of eastern Honshu, are divided in the dendrograms from those in the other areas. Conse- quently, the absence of R. nigromaculata in the Kanto and Sendai plains implies that its ancestral population reached Japan through the Korean Peninsula between about 0.5 and 0.12 million years ago. The area of the north end of the eastern Honshu region could not provide another entrance because the mountain ranges virtually reached the shore. The formation of the Eastern Honshu, Chubu, and Western Honshu-Shikoku groups is attributable to the release of migration due to marine regression during the glacial stage and the geographical isolation due to marine transgression during the interglacial stages. The formation of the Kyushu group is attributed to the introgression of new genetic mate- rials provided from continental R. nigromaculata which in- vaded again during the Wurm glacial stage when the Japanese Islands were temporarily reconnected with the Korean Penin- sula (between about 20,000 and 18,000 years ago) [14]. This introgression is typified by the presence of the A-form of chromosome 11 which is found in continental R. nigromacu- lata. In each of the lampbrush chromosomes 6, 7, 11 and 13, it is quite certain that the A-form is older than the B-form, that is, the B-form was derived from the A-form by mutations. The reasons are, firstly, that continental R. nigromaculata is of the A-form in these lampbrush chromosomes and the Chubu group, of which the R. nigromaculata migrates to the furthest area in Japan, also has the A-form in the lampbrush chromosomes 6, 7, and 13. Secondly, the lampbrush chromosomes 6, 11 and 13 of R. brevipoda, which diverged from an ancestral species common to R. nigromaculata [6], have the same characteristics as the A-form [10, 13]. In the two forms of lampbrush chromosome 1, an age comparison is difficult because continental R. nigromaculata and R. brevi- poda have forms that differ from these. However, both forms of chromosome 1 somewhat resembled those of chromosome 6 in their distribution pattern. Therefore, the A-form is also probably older than the B-form in chromo- some 1. In view of the probable migration course of R. nigroma- culata and the distributional extent of the B-forms, the B-form of each lampbrush chromosome is presumed to have occurred in the order of chromosomes 11, 1 and 6, 13, and lastly 7. The B-form of chromosome 7 seems to have occurred toward the Wtrm glacial stage, because it is very similar in distribution pattern to the A-form of chromosome 11. The B-form of chromosome 13 probably occurred be- fore the Wurm glacial stage and spread in this stage, judging from the absence in the Eastern Honshu and Chubu groups. This is because the western Honshu, Shikoku, and Kyushu areas united during the Wiirm glacial stage, since the Inland Sea dried up owing to marine regression [4]. The B-forms of chromosomes 1 and 6 probably spread during the glacial stages before the Witrm glacial stage. The B-form of chromosome 11 spread at the beginning of the migration of R. nigromaculata. 342 H. OHTANI REFERENCES Callan HG (1986) Lampbrush Chromosomes. Verlag, Berlin, pp 50-105 Gall JG (1966) Techniques for the study of lampbrush chromo- somes. In “Methods in Cell Physiology 2” Ed by DM Prescott, Academic Press, New York, pp 37-60 Japan Association for Quaternary Research (1987) Explana- tory Text for Quaternary Maps of Japan. (In Japanese) Uni- versity of Tokyo Press, Tokyo, pp 119 Kaizuka S, Naruse Y (1977) Paleogeographic changes. (In Japanese) In “The Quaternary Period: Recent Studies in Japan” Ed by Japan Association for Quaternary Research, University of Tokyo Press, Tokyo, pp 333-351 Kawamura T, Nishioka M (1977) Aspects of the reproductive biology of Japanese anurans. In “The Reproductive Biology of Amphibians” Ed by DH Taylor, SI Guttman, Plenum Press, New York, pp 103-139 Kawamura T, Nishioka M (1979) Isolating mechanisms among the water frog species distributed in the Palearctic region. Mitt Zool Mus Berlin 55: 171-185 Matsuda T (1978) Collision of the Izu-Bonin Arc with central Honshu: Cenozoic Tectonics of the Fossa Magma, Japan. J Phys Earth 26 Suppl: S 409-421 Matsuura N (1977) Molluscan fossils from the late Pleistocene marine deposits of Hokuriku region, Japan Sea side of central Japan. Sci Rep Kanazawa Univ 22: 117-162 Nishioka M, Kodama Y, Sumida M, Ryuzaki M (1993) Sys- tematic evolution of 40 populations of Rana rugosa distributed in Japan elucidated by electrophoresis. Sci Rep Lab Amphibian Springer- 10 11 12 13 14 15 16 17 Biol Hiroshima Univ 12: 83-131 Nishioka M, Ohtani H, Sumida M (1980) Detection of chromosomes bearing the loci for seven kinds of proteins in Japanese pond frogs. Sci Rep Lab Amphibian Biol Hiroshima Univ 4: 127-184 Nishioka M, Sumida M, Borkin LJ, Wu Z (1992) Genetic differentiation of 30 populations of 12 brown frog species distributed in the Palearctic region elucidated by the elec- trophoretic method. Sci Rep Lab Amphibian Biol Hiroshima Univ 11: 109-160 Nishioka M, Sumida M, Ohtani H (1992) Differentiation of 70 populations in the Rana nigromaculata group by the method of electrophoretic analyses. Sci Rep Lab Amphibian Biol Hiroshima Univ 11: 1-70 Ohtani H (1990) Lampbrush chromosomes of Rana nigroma- culata, R. brevipoda, R. plancyi chosenica, R. p. fukienensis and their reciprocal hybrids. Sci Rep Lab Amphibian Biol Hiroshi- ma Univ 10: 165-221 Ono Y (1984) Late glacial paleoclimate reconstructed from glacial and periglacial landforms in Japan. Geographical Re- view of Japan 57: 87-100 Yosida TH, Tsuchiya K, Moriwaki K (1971) Frequency of chromosome polymorphism in Rattus rattus collected in Japan. Chromosoma 33: 30-40 Yosida TH, Tsuchiya K, Moriwaki K (1971) Karyotypic differ- ences of black rats, Rattus rattus, collected in various localities of East and Southeast Asia and Oceania. Chromosoma 33: 252- 267 Yosida TH (1973) Evolution of karyotypes and differentiation in 13 Rattus species. Chromosoma 40: 285-297 ZOOLOGICAL SCIENCE 11: 343-349 (1994) Biochemical Systematics of Five Asteroids of the Family Asteriidae Based on Allozyme Variation Normasa Matsuoxa!, Krvoko FuKuDA, Kyoko YOSHIDA, Mino SUGAWARA and MEGUMI INAMORI Department of Biology, Faculty of Science, Hirosaki University, Hirosaki 036, Japan ABSTRACT— The family Asteriidae of the order Forcipulatida from Japanese waters includes the five common starfish species belonging to the five different genera. They are Asterias amurensis, Aphelasterias japonica, Distolasterias nipon, Coscinasterias acutispina and Plazaster borealis. he phylogenetic relationship of these five members were investigated by electrophoretic analyses of 15 different enzymes. From the allozyme variation observed in 31 genetic loci, the Nei’s genetic distances between species were calculated and the molecular phylogenetic tree for the five species was constructed. The phylogenetic tree indicated the following: (1) The five species are phylogenetically divided into three clusters: (i) A. amurensis and P. borealis; (11) A. japonica and D. nipon; and (ii) C. acutispina. (2) A. amurensis and P. borealis are the most closely related to each other and more recent species which evolved later. (3) A. japonica is more closely related to D. nipon than to other species. (4) C. acutispina is the most distant species of the five members. These electrophoretic results were discussed through the detailed comparison with molecular and non-molecular data, and the differentiation © 1994 Zoological Society of Japan process of five species was speculated. INTRODUCTION During the last 10-15 years, the taxonomic, phylogenetic and evolutionary studies have been revitalized by the applica- tion of techniques from biochemistry or molecular biology. Protein sequencing, immunological methods, protein elec- trophoresis, DNA hybridization test and sequence analysis of mitochondrial DNA or ribosomal RNA (DNA) are among the molecular techniques used in evolutionary studies. Of these, enzyme electrophoresis has been most widely used in the field of biochemical systematics [5]. Such molecular studies have made it possible for us to estimate the phy- logenetic relationships among taxa and their evolutionary processes quantitatively with common parameters such as enzymes or DNA, and they have been providing much relevant, and in some cases critical, information about phy- logenetic relationships in various groups of organisms [10]. One of the present authors (N.M.) has been investigating the phylogeny, taxonomy and evolution within the class Echinoidea (sea-urchins), which is one of the major groups of the phylum Echinodermata, by using the electrophoretic and immunological techniques [13-17, 19, 20]. Another large group of Echinodermata is the class Asteroidea (starfish) and we are also interested in the evolutionary aspect of starfish. The taxonomy and phylogeny of the starfish have been extensively studied by many workers from the morphological and/or paleontological standpoint [1, 2, 4, 9, 28]. However, there are disagreements between asteroid taxonomists, and many unresolved problems concerning the phylogenetic and evolutionary relationships among starfish still remain. For an elucidation of these problems, it would be desirable to Accepted January 3, 1994 Received October 12, 1993 " To whom reprint requests should be addressed. actively introduce the molecular approaches which are more analytic and quantitative than the traditional and usual mor- phological methods into the field of asteroid taxonomy. As already mentioned above, we have been investigating biochemically the phylogenetic relationships among sea- urchins, and found that enzyme electrophoresis is one of the reliable methods in the field of echinoid phylogeny and taxonomy. Therefore, we have an advantage in the biochemical systematic studies of the starfish belonging to echinoderms by enzyme electrophoresis. In the present study, with the background noted above, we have attempted to investigate the phylogeny within the family Asteriidae from the order Forcipulatida by using enzyme elec- trophoresis. Five common species were adopted in the present study. They are Asterias amurensis, Aphelasterias japonica, Distolas- terias nipon, Coscinasterias acutispina, and Plazaster borealis. As evident from Figure 1, the former three species have standard five-armed forms, while the latter multi-armed forms. Particularly, P. borealis has many arms and shows the clear differentiation between arms and disk. They are common starfish to many zoologists, because A. amurensis which is a representative species of the family has been widely used in the embryological, physiological or biochemical study. C. acutispina is widely distributed from central Hon- shu to the Ryukyus, while the other four species are arctic starfish which are commonly found in the seas of northern Japan. Although each of these five species of the family has the characteristic external morphology, it seems to be difficult to establish their phylogenetic relationship and the sequence of the evolutionary divergence by the morphological criteria. Further, there is a little quantitative information available concerning the phylogenetic relationship among these mem- bers of the family. In fact, as far as we are aware, there are 344 N. Matsuoka, K. FukupDa et al. only a few reports: the immunological studies by Kubo [12] and Mochizuki and Hori [22]. Under such situation, further biochemical systematic studies of these species would be desirable and informative. In this paper, we report on the results of an electrophore- tic study designed to clarify the phylogenetic relationship of the five common starfish species of the family Asteriidae from Japanese waters. MATERIALS AND METHODS Starfish The starfish examined in this study were five species from the family Asteriidae of the order Forcipulatida: Asterias amurensis Liitken, Aphelasterias japonica (Bell), Distolasterias nipon (Déder- lein), Coscinasterias acutispina (Stimpson), and Plazaster borealis (Uchida) (Fig. 1). A. amurensis and A. japonica were collected from the coast near the Asamushi Marine Biological Station, Tohoku University, facing Mutsu Bay, Aomori Pref., by snorkelling. D. nipon and P. borealis were provided by the Fishermen’s Cooperative Association of Yokohama-machi, Kamikita-gun, Aomori Pref. They were collected in the breeding ground of scallops in Mutsu Bay by fishermen. C. acutispina was provided by the Misaki Marine Biological Station, University of Tokyo. It was collected from the rocky shore near the Station facing Sagami Bay, Kanagawa Pref. The number of individuals examined was 20 for A. amurensis, 39 for A. japonica, 20 for D. nipon, 21 for C. acutispina, and 9 for P. borealis. After collection, the pyloric ceaca were cut off from these specimens and stored at —80°C until being analyzed. Electrophoresis Electrophoresis was performed on 7.5% polyacrylamide gels by the method of Davis [3] as described previously [14]: About 1 g of pyloric caecum was individually homogenized in 2 vols of 20mM phosphate buffer, pH 7.0, containing 0.1 M KCl and 1 mM EDTA by using a glass homogenizer of the Potter-Elvehjem type in an ice water bath. After centrifugation at 108,800 g for 20 min at 4°C, 0.05- 0.10 ml of clear supernatant was used for electrophoretic analyses of enzymes. Electrode buffer was 0.38 M glycine-tris buffer, pH 8.3. After electrophoresis, the gels were stained for the following 15 different enzymes: hexose-6-phosphate dehydrogenase (H6PD), ma- late dehydrogenase (MDH), malic enzyme (ME), nothing dehy- drogenase (NDH), octanol dehydrogenase (ODH), sorbitol dehy- drogenase (SDH), xanthine dehydrogenase (XDH), glucose-6- phosphate isomerase (GPI), hexokinase (HK), superoxide dismutase (SOD), aspartate aminotransferase (AAT), alkaline phosphatase (ALK), peroxidase (PO), esterase (EST), and leucine amino pepti- dase (LAP). Stain recipes for these enzymes have been described previously [21]. RESULTS From the allozyme variation observed in 15 different enzymes, 31 genetic loci were inferred. Figure 2 shows diagramatically allozyme patterns of four enzymes presenting the typical electrophoretic band patterns, which were chosen from among 15 enzymes analyzed in this study. The major features of variation in these enzymes are summarized as follows: ME exhibited a single band of the same mobility Fic. 1. Five starfish species of the family Asteriidae from Japanese waters. 1=Asterias amurensis, 2=Aphelasterias japonica, 3=Distolasterias nipon, 4=Coscinasterias acutispina, 5= Plazaster borealis. Biochemical Systematics of Asteroids 345 ME HK 12345 1 2 3 4 5 ODH #1 }2 <3 t4 65 LAP 12345 1 2 3 4 5 Fic. 2. Electrophoretic band patterns of four typical enzymes in five starfish species of the family Asteriidae. For each enzyme the origin is at the top and the direction of mobility toward the bottom. The number of 1-5 described in the right side of the zymogram of LAP shows the genetic loci of LAP-1 to LAP-5. Genetic loci are numbered downwards from 1, starting with that nearest the origin (i.e., of lowest electrophoretic mobility). The number of 1—5 described in the bottom of each enzyme shows the five starfish species: 1=Asterias amurensis, 2=Aphe- lasterias japonica, 3=Coscinasterias acutispina, 4= Distolasterias nipon, 5=Plazaster borealis. between five species and was monomorphic as well as NDH. ODH also showed a single active band, but varied interspeci- fically. The similar band pattern was also observed in H6PD and GPI. HK showed extensive polymorphism and exhib- ited single- and double-banded phenotypes. This variation was interpreted as a diallelic system at a single locus coding for a monomeric protein, with single-banded pattern corres- ponding to the homozygous state, and double-banded pattern to the heterozygous state. The similar variation was also observed in the following 12 loci: XDH, AAT, MDH-1, PO-2, SOD-3, SOD-4, ALK-1, ALK-4, EST-1, EST-4, LAP- 1, and LAP-2. LAP of digestive enzyme was detected as several bands which were grouped into five zones (LAP-1 to LAP-5). LAP-1 and LAP-2 showed single- and double- banded phenotypes as well as HK. The single band of LAP-3 was not scored in two species. Each of LAP-4 and LAP-S5 exhibited a single monomorphic band of high enzy- matic activity. The multi-banded patterns such as LAP were also observed in EST, ALK and SOD. The LAP activity of the starfish was much stronger than that of various sea-urchin species reported previously at the electrophoretic level [20]. On the other hand, the AMY activity which was scored easily in sea-urchins could not be detected in these starfish species. The allele frequencies for all loci in the five species are given in Table 1. With respect to the degree of enzyme variation within populations, Table 1 shows that enzymes involved in glucose metabolism (catalysing steps in, or adja- cent to, the glycolytic pathway and tricarboxylic acid cycle) were on average less variable than those (e. g., EST or SOD) involved in other reactions, which contain many that are relatively nonspecific with respect to substrate. Table 2 summarizes the extent of genetic variation in five species. The number of aileles per locus was in the range of 1.15-1.38, with a mean of 1.22, the proportion of polymorphic loci (P), in the range of 14.3-38.5%, with a mean of 21.4%, and the expected average heterozygosity per locus (H), in the range of 5.9-16.7%, with a mean of 9.0%. As evident from this table, D. nipon showed considerably higher genetic variabil- ity than the other four species. In order to quantify the degree of genetic differentiation among five species, the genetic identity (I) and genetic distance (D) between each species were calculated by the method of Nei [23] from the allele frequencies data in Table 1. Table 3 shows the matrices of I and D values between all pairs of species examined. The highest I value (0.598) was found between A. amurensis and P. borealis. Figure 3 shows the molecular phylogenetic tree for the five species which was constructed from the Nei’s genetic distance matrix of Table 3 by using the unweighted pair-group arithmetic average (UPGMA) clustering method of Sneath and Sokal [27]. The molecular phylogenetic tree indicated the fol- lowing: (1) The five species are phylogenetically divided into three large clusters: (i) A. amurensis and P. borealis; (ii) A. japonica and D. nipon; and (iii) C. acutispina. (2) Of the five species, A. amurensis and P. borealis are the most closely related to each other. (3) A. japonica is more closely related to D. nipon than to the other three species. (4) C. acutispina is the most distinct species of the five members. The divergence time (T) of the five species estimated from the genetic distance by the Nei’s equation [24] is also given in the phylogenetic tree. The molecular phylogenetic tree with the divergence time provides valuable information with respect to the evolutionary divergence of the five species of the family Asteriidae. DISCUSSION Enzyme variation within populations With respect to the relationship between enzyme func- tion and heterozygosity, Yamazaki [30] showed, using data from various Drosophila species, that the substrate-specific enzymes have lower heterozygosity than the nonspecific enzymes. A similar analysis was carried out by Gojobori [7] using data on 20 different proteins (mostly enzymes) from 14 Drosophila species, 14 Anolis species and 31 other species. As a result, he found that enzymes with various functional constraints tend to have low heterozygosity. These findings are well consistent with our serial electrophoretic studies of echinoderm enzymes. In this study, the glucose metaboliz- ing enzymes (the mean H=6.0%) with functional constraints were less variable than the non-glucose metabolizing enzymes (the mean H=9.9%), and also nonspecific enzymes such as SOD or EST were more highly polymorphic. The similar results have also been obtained in many other echinoderm 346 N. Matsuoka, K. FUKUDA et al. TaBLE1. Allele frequencies at various enzyme loci in the five species of the family Asteriidae Locus Aa Aj Ca Dn Pb H6PD b b a b b MDH-1 b b — a (0.45) = b (0.55) MDH-2 a Cc b = a ME a a a a a NDH a a a a a ODH b c c a b SDH b b b b a XDH a (0.53) b b c b b (0.47) HK b a (0.07) a (0.04) a (0.47) a (0.20) b (0.76) b (0.96) b (0.53) b (0.80) c (0.17) PGI b b a c — AAT b b b a (0.39) a b (0.61) SOD-1 a c b — a SOD-2 — a b b a SOD-3 b a — a (0.73) = c (0.27) SOD-4 a a c a (0.75) a b (0.25) PO-1 b a c — b PO-2 b a © c (0.75) b d (0.25) ALK-1 b c a (0.50) c a (0.25) ce (0.50) c (0.75) ALK-2 a b a b a ALK-3 a = a _ a ALK-4 b b b a (0.35) b b (0.65) ALK-5 — a b c a EST-1 b (0.70) a (0.40) a (0.78) a (0.42) b (0.44) ce (0.30) b (0.60) b (0.22) b (0.58) c (0.56) EST-2 b a a b — EST-3 a a a ‘a b EST-4 a (0.81) a (0.71) b (0.38) a (0.53) a (0.78) c (0.19) c (0.29) c (0.62) b (0.47) c (0.22) LAP-1 a (0.37) - a (0.40) - - b (0.63) b (0.60) LAP-2 a (0.50) a (0.47) a (0.50) a (0.83) b c (0.50) ¢ (0.53) b (0.50) c (0.17) LAP-3 a _ a a LAP-4 b a b b b LAP-5 a a a a a Alleles are correspondingly lettered from “a”. frequency of each allele in population. Aa=Asterias amurensis, Aj=Aphelasterias japonica, Ca=Coscinasterias acutispina, Dn= Distolasterias nipon, Pb=Plazaster borealis. The value in parenthesis represents the Biochemical Systematics of Asteroids 347 TaBLe 2. Genetic variation in the five species of the family Asteriidae Parameter Aa Aj Ca Dn Pb No. of alleles per locus 1.17 1.18 1.21 1.38 11,1) Proportion of polymorphic loci:P(%) 17.2 14.3 21.4 38.5 15.4 Expected average heterozygosity per locus:H(%) 7.6 6.4 8.5 16.7 5.9 Aa=Asterias amurensis, Aj=Aphelasterias japonica, Ca=Coscinasterias acutispina, Dn=Distolaster- las nipon, Pb=Plazaster borealis. TABLE 3. Genetic identities (above diagonal) and genetic distances (below diagonal) between five species of the family Asteriidae Species 1 2 3 4 5 1. Asterias amurensis — 0.475 0.434 0.484 0.598 2. Aphelasterias japonica 0.744 — 0.433 0.506 0.397 3. Coscinasterias acutispina 0.835 0.837 = 0.401 0.370 4. Distolasterias nipon 0.726 0.681 0.914 — 0.360 5. Plazaster borealis 0.514 0.924 0.994 1.022 ~ DIVERGENCE TIME(T) T:5x10°D (YEARS) 5.0 2.5 0 Asterias amurensis Plazaster borealis Aphelasterias japonica Distolasterias nipon Coscinasterias acutispina —_— 1.0 0.5 GENETIC DISTANCE (D) Fic. 3. A molecular phylogenetic tree for the five starfish species of the family Astertidae. It was constructed from Nei’s genetic distances by using the UPGMA clustering method of Sneath and Sokal [27]. The divergence time estimated from the Nei’s equation [24] using the genetic distance is given in the phy- logenetic tree. species [14, 15, 19-21]. These results can be explained by the neutral mutation theory of Kimura [11]: The more strictly functional constraint would decrease the neutral regions of the molecules and the probability of a mutation change not being harmful (i.e.,selective neutral) is smaller for the sub- strate-specific enzymes than for nonspecific enzymes. I have previously reported on the amount of genetic variation within populations of various echinoderm species [18]. It is interesting to compare the extent of genetic variation (the average heterozygosity per locus: the H value) in the five starfish species studied here with H values observed in other echinoderm populations. A. amurensis, A. japoni- ca, C. acutispina and P. borealis showed low genetic variabili- ties (H=7.6, 6.4, 8.5, 5.9%) and these H values were comparable to those (H=0-8.7%) of many other shallow water echinoderms reported previously [18]. On the other hand, D. nipon showed much higher genetic variability (H= 16.7%) than other shallow water echinoderms and the value was comparable to the H values of deep-sea echinoderms. Nei [25] and Nei and Graur [26] examined the relationship between average heterozygosity and population size for 77 different species. As a result, they found a significant cor- relation between them. From this evidence, it may be considered that the difference in the extent of genetic varia- tion among five species is related to their population sizes. Namely, it may be expected that D. nipon showing the higher genetic variability has larger population size than the other four species. Further extensive population surveys in var- 10uS marine invertebrates would be required for establishing the validity of this prediction. Phylogenetic relationship of five species of the family Aster- lidae The molecular phylogenetic tree shown in Figure 3 clear- ly indicated that the five species of the family Asteriidae are phylogenetically divided into three large clusters: (1) A. amurensis and P. borealis; (ii) A. japonica and D. nipon; and (iii) C. acutispina. Fisher [6] suggested from the morpholog- ical standpoint that the family Astertidae may be a large and polyphyletic aggregation of genera, and he proposed the subfamily system within the family Asteriidae. The heter- ogeneity of the family suggested by Fisher seems not contra- dictory to the present results, excluding the problems what species belongs to each subfamily. The electrophoretic results showed that A. amurensis and P. borealis are the most closely related to each other among five species. The genetic distance (D=0.514) be- 348 N. Matsuoka, K. FukuDaA et al. tween them was comparable to the D values reported be- tween congeneric species in other animals [29]. The close affinity between them was also suggested by the immunolog- ical study of Mochizuki and Hori [22]. They examined the phylogenetic relationships among various starfish species by using the enzyme inhibition method with the specific antibody against purified hexokinase (HK) from the pyloric ceaca of A. amurensis. The immunological data indicated that P. borealis has the highest immunological similarity to A. amurensis among seven species of the family Asteriidae examined. In addition, Fisher [6] stated from the morpho- log-ical standpoint that A. amurensis and A. japonica may be closely related to P. borealis. His view is partially consistent with these biochemical results, excepting the phylogenetic position of A. japonica. The allozymic study also showed the close affinity be- tween A. japonica and D. nipon. The genetic distance (D= 0.681) between them was comparable to the D values observed between congeneric species or closely related con- familial genera in other animals [29]. Interestingly, there are two conflicting views on the systematic position of A. japonica from the morphological standpoint: Fisher [6] and Hayashi [8] proposed the close affinity between A. japonica and A. amurensis, and included these two species into the subfamily Asteriinae. On the other hand, Shigei and Saba (personal communication) suggested that A. japonica may be rather closely related to D. nipon. The present results are in favor of the view of Shigei and Saba. In contrast, the view of Fisher [6] and Hayashi [8] is inconsistent with not only the present electrophoretic study but also the immunological studies by other workers: Mochizuki and Hori [22] showed by the enzyme inhibition method that A. japonica is distantly related to A. amurensis. Prior to their study, Kubo [12] examined the phylogenetic relationships among various starfish species by the following immunological method: He prepared rabbit antisera against extracts of tube feet of ambulacral zones taken from several starfish species and measured the cross-reactivity of the antisera to antigens from various species by the quantitative precipitin technique. As a result, he obtained the similar results to those of Mochizuki and Hori [22]. However, there are some differences be- tween the present electrophoretic data and Kubo’s immuno- logical results. Namely, Kubo [12] showed that A. amurensis was more distantly related to A. japonica than to D. nipon and C. acutispina. On the other hand, the present electrophoretic results (Fig. 3) indicated that A. amurensis and P. borealis are more closely related to the cluster of A. japonica and D. nipon than to C. acutispina. In spite of such differences, these biochemical studies did not support the close affinity between A. japonica and A. amurensis which was suggested by the morphological studies [6,8]. The molecular phylogenetic tree (Fig. 3) also indicated that C. acutispina is the most distant species of the five members. The result seems to be consistent with the zoogeographical evidence: Of the five species, C. acutispina is not commonly found in the cold seas of northern Japan and distributes widely in the more southern regions from central Honsyu to the Ryukyu Islands. On the other hand, the main distributional region of the other four species is the cold seas of northern Japan. From the morphological studies, Fisher [6] and Hayashi [8] proposed the close affinity between C. acutispina and D. nipon, and included these two species into the subfamily Coscinasterinae. However, their taxonomic system is inconsistent with the present electrophoretic results. The molecular phylogenetic tree (Fig. 3) shows not only their genetic relationships, but also the sequence of their evolutionary divergence. According to Nei [24], genetic distance (D) corresponds well with the divergence time (T) from the common ancestor, and T of two taxa can be estimated by T=5 x 10° D (years). Application to this equa- tion to the molecular dendrogram constructed from the genetic distances leads to the following speculation of evolu- tionary process of the five species: Firstly, the common ancestor of the five species diverged into two lineages (one is Coscinasterias lineage and the other the common ancestor of the other four genera) 4.5 million years (MY) ago. Then, the latter ancestor diverged into two lineages (one is the Asterias-Plazaster lineage and the other the Aphelasterias- Distolasterias lineage) after a short time (4.3MY ago). Finally, these four genera differentiated from one another 2.6-3.4 MY ago. The phylogenetic tree suggests that Aster- ias and Plazaster are more recent genera which evolved later. The biochemical systematic studies of sea-urchins re- ported previously suggested that the more recent species which evolved later tend to become predominant species [14— 17, 20]. Among the five species of the family Asteriidae, A. amurensis which evolved later seems to be more predominant species than others. The species is more frequently found in Japanese waters than the other four species and shows the extensive morphological variations between local Japanese populations in some morphological characters such as body color, body size, spine and so on. This may suggest the speciation within A. amurensis. At present, we have been investigating the genetic differentiation between local popula- tions of A. amurensis by using enzyme electrophoresis and attempting the molecular approach concerning the speciation and evolution of the species. As evident from Figure 1, P. borealis is considerably specialized at the morphological level, and shows the clear differentiation between disk and arm, in contrast with the other four species with standard morphology. The molecu- lar phylogenetic tree (Fig. 3) implies that P. borealis of such specialized morphology might have differentiated from the Asterias-like starfish with standard morphology, since the cluster consisted of P. borealis and A. amurensis is also closely related to the cluster of A. japonica and D. nipon with standard morphology. If it is true, the evolutionary rate at the morphological level in the Plazaster lineage might have been much accelerated. In future, further detailed inves- tigation on the close genetic relationship between P. borealis and A. amurensis which highly differentiated with each other at the morphological level would produce some useful and Biochemical Systematics valuable information on the morphological evolution in starfish. ACKNOWLEDGMENTS We are grateful to the Fishermen’s Cooperative Association of Yokohama-machi, Aomori Prefecture, the Misaki Marine Biological Station, University of Tokyo, and the Asamushi Marine Biological Station, Tohoku University, for their kind help in collecting the starfish specimens. We also thank Dr. M. Shigei, Kyoto Institute of Technology, and Dr. M. Saba, Mie Prefectural Ise High School, for their valuable advice. This study was supported by a Grant-in-Aid (Grant No. 05640779) for scientific research from the Ministry of Education, Science and Culture of Japan to Norimasa Matsuoka. REFERENCES 1 Blake DB (1987) A classification and phylogeny of post- Palaeozoic sea stars (Asteroidea: Echinodermata). J Nat Hist 21: 481-528 2 Clark AC, Downey ME (1992) Starfishes of the Atlantic. Chapman and Hall, London 3 Davis BJ (1964) Disc electrophoresis—II. Method and ap- plication to human serum proteins. Ann NY Acad Sci 121: 404-427 4 Downey ME (1973) Starfishes from the Caribbean and the Gulf of Mexico. Smithsonian Contr Zool 126: 1-158 5 Ferguson A (1980) Biochemical Systematics and Evolution. Blackie, Glasgow 6 Fisher WK (1928) Asteroidea of the North Pacific and adjacent waters. Part 2. Forcipulata (part). US Nat Mus Bull 76 7 Gojobori T (1982) Means and variances of heterozygosity and protein function. In “Molecular Evolution, Protein Poly- morphism and the Neutral Theory” Ed by M Kimura, Japan Scientific Societies Press Berlin, Springer-Verlag pp 137-148 8 Hayashi R (1943) Contributions to the classification of the sea-stars of Japan. II. Forcipulata, with the note on the rela- tionships between the skeletal structure and respiratory organs of the sea-stars. J Fac Sci Hokkaido Univ Ser VI 8: 133-281 9 Hayashi R (1974) Asteroids. In “Systematic Zoology” Vol. 8b Ed by T Uchida, Nakayama, Tokyo (In Japanese) pp 82-141 10 Hills DM, Moritz C (1990) Molecular Systematics. Sinauer, MA 11 Kimura M (1983) The Neutral Theory of Molecular Evolution. Cambridge University Press, Cambridge 12 Kubo K (1961) Studies on the systematic serology of sea-stars. V. Jpn J Zool 13: 15-37 13 Matsuoka N (1980) Immunological relatedness of sea-urchin glucose-6-phosphate dehydrogenases: Phylogenetic implication. 14 15 16 17 18 19 20 21 22 23 24 25 26 Dil, 28 29 30 of Asteroids 349 Comp Biochem Physiol 66B: 605-607 Matsuoka N (1985) Biochemical phylogeny of the sea-urchins of the family Toxopneustidae. Comp Biochem Physiol 80B: 767-771 Matsuoka N (1987) Biochemical study on the taxonomic situa- tion of the sea-urchin, Pseudocentrotus depressus. Zool Sci 4: 339-347 Matsuoka N (1989) Biochemical systematics of four sea-urchin species of the family Diadematidae from Japanese waters. Biochem Syst Ecol 17: 423-429 Matsuoka N (1990) Evolutionary relationships of sea-urchins at the molecular level. Comp Biochem Physiol 97B: 31-36 Matsuoka N (1991) Maintenance mechanism of enzyme poly- morphism in echinoderms. Sci Rep Hirosaki Univ 38: 38-45 Matsuoka N, Hatanaka T (1991) Molecular evidence for the existence of four sibling species within the sea-urchin, Echi- nometra mathaei in Japanese waters and their evolutionary relationships. Zool Sci 8: 121-133 Matsuoka N, Suzuki H (1989) Electrophoretic study on the phylogenetic relationships among six species of sea-urchins of the family Echinometridae found in the Japanese waters. Zool Sci 6: 589-598 Matsuoka N, Yoshida K, Fukuda K, Shigei M (1991) Genetic variation in the starfish Coscinasterias acutispina. Comp Biochem Physiol 99B: 893-898 Mochizuki Y, Hori SH (1980) Immunological relationships of starfish hexokinases: Phylogenetic implication. Comp Biochem Physiol 65B: 119-125 Nei M (1972) Genetic distance between populations. Am Nat 106: 283-292 Nei M (1975) Molecular Population Genetics and Evolution. North-Holland, Amsterdam Nei M (1983) Genetic polymorphism and the role of mutation in evolution. In “Evolution of Genes and Protein” Ed by M Nei, R Koehn, Sinauer,MA pp 165-190 Nei M, Graur D (1984) Extent of protein polymorphism and the neutral mutation theory. Evol Biol 17: 73-118 Sneath PHA, Sokal PR (1973) Numerical Taxonomy. Free- man, San Francisco, CA Spencer WK, Wright CW (1966) Asterozoans. In “Treatise on Invertebrate Paleontology” Part U Ed by RC Moore, Geol Soc Am Univ Kansas Press, pp 4-107 Thorpe JP (1982) The molecular clock hypothesis: Biochemi- cal evolution, genetic differentiation, and systematics. Ann Rev Ecol Syst 13: 139-168 Yamazaki T (1977) Enzyme polymorphism and functional dif- ference: mean, variance, and distribution of heterozygosity. In “Molecular Evolution and Polymorphism” Ed by M Kimura, Mishima: National Institute of Genetics pp 127-147 Development Published Bimonthly by the Japanese Society of Developmental Biologists 7 mtr Distributed by Business Center for Academic Growth & Differentiation Societies Japan, Academic Press, Inc. Papers in Vol. 36, No. 2. (April 1994) 13. REVIEW: S. V. Cooke and B. D. Shur: Cell Surface £1,4-Galactosyltransferase: Expression and Function 14. C. C. Lambert, G. P. Gonzales and K. M. Miller: Independent Initiation of Calctum Dependent Glycosidase Release and Cortical Contractions during the Activation of Ascidian Eggs 15. Y. Hagiwara and E. Ozawa: A New Method for Fibroblast-less Primary Skeletal Muscle Cell Culture by the Use of Hydroxyurea 16. K. Yamazaki, R. Suzuki, E. Hojo, S. Kondo, Y. Kato, K. Kamioka, M. Hoshi and H. Sawada: Trypsin-like Hatching Enzyme of Mouse Blastocysts: Evidence for Its Participation in Hatching Process before Zona Shedding of Embryos 17. A. lio, M. Mochii, K. Agata, R. Kodama and G. Eguchi: Expression of the Retinal Pigmented Epithelial Cell-Specific pP344 Gene during Development of the Chicken Eye and Identification of Its Product 18. S. Kuno, T. Nagura and I. Yasumasu: Insulin-Induced Outgrowth of Pseudopodial Cables from Cultured Micromere-Derived Cells Isolated from Sea Urchin Embryos at the 16 Cell Stage, with Special Reference to the Insulin-Receptor 19. T.Iwamatsu: Medaka Oocytes Rotate within the Ovarian Follicles during Oogenesis 20. L-N. Wei and Y-C. Hsu: Identification of a New Gene Expressed Specifically in Early Mouse Embryos 21. A.M. Fausto, M. Carcupino, M. Mazzini and F. Giorgi: An Ultrastructural Investigation on Vitellophage Invasion of the Yolk Mass during and after Germ Band Formation in Embryos of the Stick Insect Carausius morosus Br. 22. T. Shimizu, K. Yoshino and N. Suzuki: Identification and Characterization of Putative Receptors for Sperm-Activating Peptide I (SAP-I) in Spermatozoa of the Sea Urchin Hemicentrotus pulcherrimus 23. M. Kosaka, M. Takeda, K. Matsumoto and Y. Nishimune: F9 Cells Can be Differentiated toward Two Distinct, Mutually Exclusive Pathways by Retinoic Acid and Sodium Butyrate 24. T. Kaname, S. Matsubara, F. Murata, K. Yamamura, K. Miyata, and T. Muramatsu: The Upstream Sequence of a New Growth/Differentiation Factor, Midkine (MK), Mediates Developmentally Regulated lac Z Gene Expression in Transgenic Mice Development, Growth and Differentiation (ISSN 0012-1592) is published bimonthly by The Japanese Society of Developmental Biologists. Annual subscription for Vol. 35 1993 U.S. $ 191,00, U.S. and Canada; U.S. $ 211,00, all other countries except Japan. ill prices include postage, handling and air speed delivery except Japan. Second class postage paid at Jamaica, N.Y. 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 32887-4900, 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 32887-4900, U.S. A. In Japan: Send nonmember subscription orders and notices of change of address to Business Center for Academic Societies Japan, 16-9, Honkomagome 5-chome, Bunkyo-ku, Tokyo 113, Japan. Send inquiries about membership to Business Center for Academic Societies Japan, 16-9, Honkomagome 5-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. BRANCGHES NOW OPEN IN (USA VAIN ID SE UsROrRE Narishige. The complete range for micromanipulation or over 30 years Narishige have been developing their This large range includes micro- manipulators, microelectrode pullers and micro- S55 forges, micro- CF injectors, microgrinders Shown here is a small and stereotaxic selection of instruments from instruments. the extensive Narishige range. MF-83 Patch Clamp Forge SN-3 Stereotaxic Instrument extensive range of precision instruments for Physiology, — See US-I Miniature Pharmacology, Zoology and IM4 Microinjector Stand Psychology research. Narishige are pleased to announee that repair facilities have been opened in Europe and the USA. EG-6 Microgrinder This enables us to provide an improved service to the many users of our quality products throughout the world, including the upgrading of previous types of water MX-1 3 Axis filled hydraulic Please contact us for PN-3 Microelectrode Puller iim micromanipulators. more information. x | Ss 5) ee — om ee PE-2 Microelectrode Puller REPAIRS, AFTER SALES SERVICE AND TECHNICAL SUPPORT Narishige Scientific Instrument Laboratory 9-28 Kasuya 4-Chrome, Setagaya-Ku, Tokyo 157, Japan. Tel: +81 (0) 3 3308 8233 Fax: +81 (0) 3 3308 2005 Narishige International London Branch Unit 7 Willow Business Park, Willow Way, London SE26 4QP U.K. Tel: +44 (0) 81 699 9696 Fax: +44 (0) 81 291 9678 U.S. Narishige International Inc 104 Glen Cove Avenue, Sea Cliff, New York 11579 U.S.A. Tel: +1 (516) 759 6167 Fax: +1 (516) 759 6138 ZOOLOGICAL SCIENCE VOLUME 11 NUMBER 2 APRIL 1994 CONTENTS REVIEWS Ward, A., P. Bierke, E. Pettersson and W. Engstrém: Insulin-like growth factors: Growth, transgenes and WM PHNUNG Sase eye aegast neces 03s See eee eRe eer 167 Mizunami, M: Processing of contrast signals in the insect ocellar'syst€m: s..cos. Aesecned Mra enipmiear sscsee ees 175 ORIGINAL PAPERS Physiology Aonuma, H., T. Nagayama, M. Hisada: Output effect of identified interneurons upon the abdominal postural system in the crayfish Procambarus clarkii (Gerard) Immunology Hirose, E., T. Ishii, Y. Saito, Y. Taneda: Phagocytic activity of tunic cells in the colonial ascidian Aplidum yamazii (Polyclinidae, Aplousobranchia) ........... 303 Biochemistry Harumi, T., K. Hoshino, N. Suzuki: Jn vitro autophos- phorylation and cyclic nucleotide-dependent dephos- phorylation of sea urchin sperm histone kinase .... 209 Mukai, M., T. Kondo, K. Yoshizato: Rapid and quan- titative detection of aspartic proteinase in animal tissues by radio-labeled pepstatin A ....................... 221 Furukohri, T., S. Okamoto, T. Suzuki: Evolution of phosphagen kinase (III). Amino acid sequence of arginine kinase from the shrimp Penaeus japonicus i rapide. anabapebee ee mttaer Skee aetna 229 Developmental Biology Furuya, H., K. Tsuneki, Y. Koshida: The development of the vermiform embryos of two mesozoans, Dicyema acuticephalum and Dicyema japonicum ............ 235 Kimura, K., K. Usui, T. Tanimura: Female myoblasts can participate in the formation of a male-specific muscles rosopiila wae ere eee eee eee eee 247 Yazaki, I., H. Harashima: Induction of metamorphosis in the sea urchin, Pseudocentrotus depressus, using E-plutamine’ 3... sacettioe aang ciesrseaiiacere sete deat 253 Ohya, Y., K. Watanabe: Control of growth and dif- ferentiation of chondrogenic fibroblasts in soft-agar Role of basic fibroblast growth factor and sip Hae eta abd nitrates loud asettons 261 culture: transforming growth factor-2 INDEXED IN: Current Contents/LS and AB & ES, Science Citation Index, ISI Online Database, CABS Database, INFOBIB Reproductive Biology Okia, N. O.: Membrane-bound inclusions in the Leydig cell cytoplasm of the broad-headed skink, Eumeces PEAS 5 Sc ss So 25 269 Identification and localization of a ligand molecule of Xenopus cortical granule lectins ....... 275 Nakamura, M., T. Yamanobe, M. Takase: Localization and purification of serum albumin in the testis of Aenopus laevis’ 2... 0500s sees vee 285 laticeps (Lacertilia: Yoshizaki, N.: Scincidae) Endocrinology Ohta, N., T. Mori, S. Kawashima, S. Sakamoto, H. Kobayashi: Spatiotemporal pattern of DNA synthesis detected by bromodeoxyuridine labeling in the mouse endometrial stroma during decidualization ......... 291 Tsai, P. I., S. S. Madsen, S. D. McCormick, H. A. Bern: Endocrine control of cartilage growth in coho salmon: GH influence in vivo on the response to IGF-I in vitro ve eedeteeteue des deste beboeee tet 299 Environmental Biology Takaku, G., H. Katakura, N. Yoshida: Mesostigmatic mites (Acari) associated with ground, burying, roving carrion and dung beetles (Coleoptera) in Sapporo and Tomakomai, Hokkaido, northern Japan ........... 305 Systematics and Taxonomy Amemiya, S., Y. Mizuno, S. Ohta: First fossil record of the family Phormosomatidae (Echinothurioida: Echi- noidea) from the Early Miocene Morozaki Group, céntral Japan on coset see ciccicienie + coe eee 313 /Grismer, L. L., H. Ota, S. Tanaka: Phylogeny, clas- sification, and biogeography of Goniurosaurus kuroiwae (Squamata: Eublepharidae) from _ the Ryukyu Archipelago, Japan, with description of a new subspecies Ohtani, H.: Polymorphism of lampbrush chromosomes in Japanese populations of Rana nigromaculata .... 337 Matsuoka, N., K. Fukuda, K. Yoshida, M. Sugawara, M. Biochemical systematics of five asteroids of the family Asteriidae based on allozyme variation Inamori: Issued on April 15 Front cover designed by Saori Yasutomi Printed by Daigaku Letterpress Co., Ltd., Hiroshima, Japan ye ISSN 0289-0003 OOLOGICAL = SCIENCE ZOOLOGICAL SCIENCE The Official Journal of the Zoological Society of Japan Editors-in-Chief: ; The Zoological Society of Japan: Seiichiro Kawashima (Tokyo) Toshin-building, Hongo 2—27-2, Bunkyo-ku, dee Ve ECS CLES) Tokyo 113, Japan. Phone 03-3814-5461 Division Editors: Fax (3-3814-5352 Shunsuke Mawatari (Sapporo) Officers: Yoshitaka Nagahama (Okazaki) Takashi Obinata (Chiba) Suguru Ohta (Tokyo) Noriyuki Satoh (Kyoto) Assistant Editors: Akiyoshi Niida (Okayama) Masaki Sakai (Okayama) Sumio Takahashi (Okayama) President: Hideo Mohri (Chiba) Secretary: Takao Mori (Tokyo) Treasurer: Makoto Okuno (Tokyo) Librarian: Masatsune Takeda (Tokyo) Auditors: Hideshi Kobayashi (Tokyo) Hiromichi Morita (Fukuoka) Editorial Board: Kiyoshi Aoki (Tokyo) Makoto Asashima (Tokyo) Howard A. Bern (Berkeley) Walter Bock (New York) Yoshihiko Chiba (Yamaguchi) Aubrey Gorbman (Seattle) Horst Gurnz (Essen) Robert B. Hill (Kingston) Yukio Hiramoto (Chiba) Tetsuya Hirano (Tokyo) Motonori Hoshi (Tokyo) Susumu Ishii (Tokyo) Hajime Ishikawa (Tokyo) Sakae Kikuyama (Tokyo) Makoto Kobayashi (Higashi-Hiroshima) Kiyoaki Kuwasawa (Tokyo) John M. Lawrence (Tampa) Koscak Maruyama (Chiba) Roger Milkman (Iowa) Kazuo Moriwaki (Mishima) Richard S. Nishioka (Berkeley) Chitaru Oguro (Toyama) Masukichi Okada (Tsukuba) Andreas Oksche (Giesen) Hiraku Shimada (Higashi-Hiroshima) Yoshihisa Shirayama (Tokyo) Takuji Takeuchi (Sendai) Ryuzo Yamagimachi (Honolulu) ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and rapid 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 Jasponenses (1897-1983), the former official journals of the Zoological Soceity of Japan. An annual volume consists of six regular numbers and one supplement (abstracts of papers presented at the annual meeting of the Zoological Society of Japan) of more than 850 pages. The regular numbers appear bimonthly. MANUSCRIPTS OFFERED FOR CONSIDERATION AND CORRESPONDENCE CONCERNING EDITORIAL MATTERS should be sent to: Dr. Tsuneo Yamaguchi, Editor-in-Chief, Zoological Science, Department of Biology, Faculty of Science, Okayama University, Okayama 700, 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, 5—16—9 Honkomagome, Bunkyo-ku, Tokyo 113. Subscriptions outside Japan should be ordered from the sole agent, VSP, Godfried van Seystlaan 47, 3703 BR 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. () Copyright 1994 by the Zoological Society of Japan. In the U.S.A., authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by [copyright owner’s name], provided that designated fees are paid directory to Copyright Clearance Center. For those organizations that have been granted a photocopy license by CCC, a separate system of payment has been arranged. Copyright Clearance Center, Inc. 27 Congress St., Salem, MA, U.S.A. (Phone 508—744— 3350; Fax 508-741-2318). Publication of Zoological Science has been supported in part by a Grant-in-Aid for Publication of Scientific Research Results from the Ministry of Education, Science and Culture, Japan. ZOOLOGICAL SCIENCE 11: 351-361 (1994) © 1994 Zoological Society of Japan REVIEW Regulation of Gonadotropin Receptors and Its Physiological Significance in Higher Vertebrates Kazuyosui Tsutsur! and SENCHIRO KAWASHIMA~ ‘Faculty of Integrated Arts and Science, Hiroshima University, Higashi-Hiroshima 724 and *Zoological Institute, School of Science, University of Tokyo, Tokyo 113, Japan INTRODUCTION Gonadotropins, follicle-stimulating hormone (FSH) and luteinizing hormone (LH), are secreted by the anterior pituitary in response to gonadotropin releasing-hormone (GnRH) of the hypothalamus. It is well established that in mammals and birds both FSH and LH are indispensable for the biological function of the gonad. In the testis, FSH first acts on Sertoli cells [3, 27, 56, 74], and Sertoli cells are involved in the initiation and maintenance of spermatogenesis [39, 87]. LH acts on Leydig cells to stimulate the androgen production in the testis [78]. The male sex characters and spermatogenesis are under androgenic regulation. In the ovary, LH together with FSH interacts with the theca interna and membrana granulosa, and regulates the follicular matura- tion, ovulation and steroidogenesis (for review, see [23]). Binding of gonadotropins to specific receptor sites on the plasma membrane is accepted as the first indispensable step for their action. Therefore, the activity of target cells is dependent not only on the circulating gonadotropin level but also on the number of gonadotropin receptor sites. Thus, to understand the cellular function of the gonad, the studies concerning receptor changes and the regulatory mechanism are essential. Over the past 20 years, radioimmunoassay has been the main tool for the determination of circulating gonadotropin levels. As for gonadotropin receptor assay, the pioneering studies by Means and Vaitukaitis [57] and Bhalla and Reichert [11] on the radioligand receptor assay in rats opened the door to a wide range of research application. This paper summarizes the advances made in our under- standing of physiological changes in gonadotropin receptors and the regulatory mechanism in higher vertebrates. Although both studies on the ovarian and testicular gonado- tropin receptors will be dealt with, more emphasis is placed on the latter. For detailed reviews on the ovarian gonado- tropin receptors the reader is referred to References [1, 15, 23, 40]. Received April 7, 1994 RECEPTORS FOR GONADOTROPINS Radioligand receptor assay For the identification of target cells for gonadotropins in the gonad, the development of techniques for radiolabeling protein hormones was essential. Precaution is needed in that radiolabeled FSH and LH retain biological activities. In several species of domestic and laboratory mammals, specific receptor sites in the testis for FSH and LH have been found exclusively in Sertoli cells and Leydig cells, as ascer- tained by autoradiography and radioligand binding assays [14, 55, 66, 82, 88]. Similar results have been obtained in a domestic avian species [47]. It is also well established in the ovary that FSH receptors are located in the granulosa cells and LH receptors are located in the cells of the theca interna, membrana granulosa and corpora lutea. Binding of FSH or LH to target cell receptors shows common physicochemical properties, such as, hormone spe- cificity, high affinity, etc. With these discoveries, several investigators developed specific and sensitive radioligand receptor assays based on the ability of radioiodinated FSH or LH to bind specific receptor sites in the gonadal homogenates of mammals [14, 18, 58, 69, 103] and birds [12, 29, 44, 45, 52, 95]. Specificity for hormones Hormone specificity of the receptor can be examined by means of competitive inhibition of the binding of radioiodin- ated hormone by various unlabeled hormone preparations. For example, testicular FSH receptors of rodents bind spe- cifically mammalian FSHs but not LHs or TSH (Fig. 1) [98, 99, 103]. Similarly, FSH receptors in the avian gonad speci- fically recognize mammalian and avian FSHs [12, 44, 45, 101]. Hormone specificity of LH receptors has also been demon- strated [22]. Affinity To estimate the affinity of binding, equilibrium analysis of the binding should be performed. Scatchard plot analysis of the FSH binding data unanimously reveals that the equilib- wy) n i) rat LH-I-4 A) MOUSE —s ts, ete PN rat SHR ©. *, a, DN z \ o% Ze 7. Binding (cpm x 10°) eo B) RAT rat LH-1-4 — — Bees po OB ] e e a Ea Fe NS Me “Oy te, e 73) Binding (cpm x 10 Competitor in log (ng ) Fic. 1. Competition of specific binding of labeled rat FSH to the particulate fraction of testicular homogenates of mice (A) and rats (B) by various gonadotropin preparations. Incubation for 120 min at 37°C. rium dissociation constant (Kd) ranges between 10’ to 107 '" M in the testis of mammals [11, 14, 18, 57, 98, 99, 103] and of birds [12, 29, 44, 45, 75]. A similar affinity value (Kd) for LH has been obtained with the mammalian [100] and avian testis [52]. CHANGES IN GONADOTROPIN RECEPTORS AND GONADOTROPIN RESPONSIVENESS During 1970s and 1980s, several investigators asked the following questions. First, when during development do gonadotropin receptors appear in the gonad? Second, what are the changes of gonadotropin receptors in the maturing and matured gonad? Many studies have been carried out to determine the number of gonadotropin receptor sites and binding affinity during fetal and postnatal life. Fetal life The ontogeny of testicular FSH receptors during fetal life was first presented by Warren eral. [108]. In fetal rats, FSH binding was detectable at 17.5 days of gestation and the binding level significantly increased at 20.5 days [108]. The kinetics of basic properties of FSH binding were conducted by Tsutsui and Kawashima [96]. The FSH binding to the testis of rat fetuses at 17.5 days of gestation showed a saturable process with respect to the concentration of FSH (Fig. 2). K. Tsutsul AND S. KAWASHIMA 17.5 days of gestation 400 300 0 ae ° 200 — e E — a oO Fy v to) 1.0 2.0 30 D £ to 1 postnatal day e c= 2 20) y = P <0,005 = r :-0,88 o o a x= un 7) rig © o ) e quail. o:sK xyes : mh Until the recent studies by Closset and Hennen [21] and 2 a Tsutsui [92], the up-regulation of testicular FSH receptors had not been established in mammalian species. Tsutsui [92] used hypophysectomized immature (25 days of age) rats. The absence of pituitary and/or gonadal hormones was not followed by an increase in FSH-binding sites per two testes (Fig. 4), suggesting that the induction of receptors depends on these hormones. The results of replacement therapy supported this hypothesis. FSH administration to hypophysectomized immature rats induced a dose-dependent increase in the total number of FSH-binding sites in the testis and testicular weight. The stimulatory effect of FSH on the testicular FSH receptors has already been reported by Closset and Hennen [21]. The study of Tsutsui [92] further indi- cated that testosterone induced an increase in FSH-binding sites with no influence on the testicular weight. There is evidence indicating that Sertoli cells of rats possess a cytosol receptor-nuclear acceptor system for testosterone [60, 61, 76]. Thus, direct effect of testosterone on Sertoli cells is probable. From the results in quails and rats, it may be generally stated that Sertoli cells require at least FSH and testosterone to induce FSH receptors in the developing testis (Fig. 5). As for the synergistic effects of FSH and testosterone on Sertoli cell function, secretion of androgen-binding protein (ABP) has been reported in immature rats [37]. Synergism of FSH and testosterone on FSH receptors may be present in imma- ture male rats [92]. The up-regulation of FSH receptors by FSH and estradiol was also observed in the ovarian granulosa cells of female rodents [71, 72]. It has been shown in the rat that FSH activates mem- brane bound adenylate cyclase of Sertoli cells and stimulates various biological processes, such as production of ABP, transferrin and insulin-like growth factor (IGF), activation of protein kinase and RNA polymerase, steroid metabolism and morphological changes including hypertrophy [3, 27, 56, 74, 76]. On the other hand, testosterone enhances the action of FSH to increase ABP production [36] and potentiates the fo} J H Se e #K AK oo 20 ——+ 9 Ego ePe E BE a 10 ott Control Sd eee u OOO CARES 2.0 per testes >) = Specific FSH binding (cpm x 10 Age in days Fic. 4. Long term effects of hypophysectomy (Hypox) on specific FSH binding per 4 mg testicular tissue and per two testes in rats. Rats were hypophysectomized at 25 days of age, and binding was measured 10 and 16 days after surgery. Intact rats served as controls. Incubation was performed for 150 min at 35°C. Sig- nificant differences from matched controls: ***P<0.001. See [92] for details. effect of FSH on Sertoli cells to increase incorporation of *’P [46]. The increase in FSH receptors is the only reported FSH-independent effect of testosterone on Sertoli cells. Accordingly, it is highly probable that testosterone, acting on chromatin of Sertoli cells, specifically induces synthesis of FSH receptors (Fig.5). In contrast, FSH may activate 356 K. Tsutsul AND S. KAWASHIMA Sertoli cell R FSH synthesis transducing FSHD>RFsii| systems some paracrine factors Leydig cell testosterone qe_ttansducing LH systems RLH transducin synthesis ¢ Systems | RPRL[PRL Fic. 5. A model for up-regulation of testicular gonadotropin receptors. FSH stimulates, by acting with testosterone secreted by Leydig cells, the synthesis of its receptors (R¢sy) in Sertoli cells. Sertoli cells possess specific receptor sites for not only FSH but also testosterone. LH receptor (Ry) synthesis in Leydig cells is controlled by PRL and FSH (indirectly by some paracrine factors). There are no receptors for FSH on Leydig cells, although Leydig cells possess PRL receptors (Rprr)- general protein synthesis nonspecifically acting at some level after transcription (Fig. 5). Interestingly, other factors that are independent of pituitary and sex hormones may also contribute to the testicular FSH receptor induction. According to Tsutsui [92], hypophysectomy in neonatal (9 days of age) rats was followed by an increase of FSH-binding sites in the testis after the operation, although the rate of increase in receptors in hypophysectomized rats was less than that in intact control rats. Such an increase in SH _ receptors after hypophysectomy suggests not only pituitary and gonadal hormones but also other factors contribute to the receptor induction. Recent studies with mammals have revealed that not only hormones but also paracrine factors modulate Sertoli cell function [3, 28, 62, 83]. According to Skinner [83], P-Mod-S was effective in stimulating Sertoli cell function. However, it is considered that the production of P-Mod-S in the peritubular myoid cell is androgen regulated [83]. Accordingly, some unknown paracrine modulators that are not under direct influence of pituitary and gonadal hormones may take part in the induction of FSH receptors during fetal and neonatal periods. In contrast to FSH receptors, the induction of LH receptors is more closely linked with the pituitary hormones, since hypophysectomy was followed by a marked loss of LH-binding sites in both immature and adult rats [92]. In terms of the LH receptor induction, the most striking observation is the heterologous up-regulation by prolactin (PRL) and FSH (Fig.5). It has been reported that PRL treatment enhances both LH-binding sites and LH-stimulated steroidogenesis in hamsters exposed to short day (SD) photo- periods [9], and in immature and adult hypophysectomized rats [68, 110] and mice [89, 90]. In addition, knowledge has been accumulated in rats and mice that FSH augments the number of LH receptors and Leydig cells [17, 21, 50, 89, 90]. Prepubertal rise in testicular LH receptors coincides with the increase in plasma level of FSH in rats [51] and mice [59]. However, the mechanism of action of FSH on LH receptor induction is not known. Although Leydig cells possess PRL receptors, there are no receptors for FSH on Leydig cells. Presumably FSH effect on Leydig cells is mediated by Sertoli cells. Some paracrine factors secreted by Sertoli cells may have a positive effect on LH receptor induction in Leydig cells (Fig. 5) [79]. Another possibility is that the described effect of FSH was due to the small amounts of LH contami- nating the FSH preparations [68]. However, treatment with the estimated quantity of LH alone had no effect on the in vitro Leydig cell response [105]. Down-regulation In addition to up-regulatory effect of FSH on FSH receptors, FSH reduces its own receptors in the testis. This phenomenon is called down-regulation. Tsutsui et al. [103] reported that in mice persistently high concentrations of plasma FSH levels since the onset of puberty act to reduce the number of FSH receptors in the testis and through this mechanism FSH receptors are maintained at a low level throughout the adult life (Fig. 6). To confirm this hypoth- esis, hypophysectomy and hormonal replacement therapy were conducted by Tsutsui ef a/. [103]. If down-regulation is crutial for the regulation of testicular FSH receptors in adult mice, the number of FSH receptors should increase after hypophysectomy. Hypophysectomy at adulthood (90 days of age) induced a decrease in the testicular weight, but the concentration and total number of FSH-binding sites were increased after the operation (Fig. 7). In contrast, the administration of FSH to hypophysectomized adult mice reduced the number of FSH-binding sites (Fig. 7). O’Shaughnessy and Brown [66] also reported in intact adult rats that direct injection of a high dose of FSH into the testis induced a decrease in the FSH binding to about 50% within 24hr. Recently, down-regulation of testicular FSH recep- tors has been found in adult hamsters [53]. Thus, it appears that FSH contributes as an inhibitory factor to the mainte- nance of adult level of FSH receptors in these rodents. The higher relative plasma FSH concentrations in mice after puberty compared to rats seems to be the cause for the observed difference in the adult level of FSH receptors Regulation of Gonadotropin Receptors 357 Plasma FSH a) ie E “” E au Bp ee = 600 “a, 2 2 in Me A aunesnendueseste be =. 2 FSH receptor Be o g 400 ae = = = & 200 1 uv 0 0 20 0-60 80 100 Age in day a a Z = cigs [—**) SUL = == = — E 2 = S = om 5 = = 3 mt S 40 «G0 80 100 Age in days Fic. 6. Changes with age in plasma FSH level and specific binding of labeled rat FSH per two testes in mice (A) and rats (B). Concentrations of plasma FSH are expressed as nanograms of NIADDK-rat FSH-RP-1 per ml. See [103] for details. between rats and mice (Fig. 6). In rats and mice several biochemical responses to FSH decrease after puberty [3, 25-27, 56, 76]. Since these re- sponses are mediated by cAMP, the decrease may be due to the decrease in the capacity of FSH receptors in Sertoli cells. Tsutsui et al. [104] found that the accumulation of cAMP by FSH treatment was lower in adult mice than prepubertal mice, and the low responsiveness of adult mice was modified by hypophysectomy. These changes in FSH responsiveness were in parallel with those in FSH-binding sites per Sertoli cell [103]. Therefore, it is considered that the down- regulation of FSH receptors brings forth a lowered respon- siveness to FSH at adulthood. The desensitization of adenylate cyclase system associated with G-proteins may also account for the lowered responsiveness based on the studies with rats [106]. With the use of light and electron microscopic auto- radiography, Shimizu et al. [82] found that '°"I-FSH after binding to the cell-surface receptors translocates into the intracellular organelles by receptor-mediated endocytosis. Shimizu and Kawashima [80] further investigated the be- havior of '°'I-FSH after binding to cell-surface receptors in 0075 Hypox Kd=2.39x107'2 y Hypox + FSH Kd=2.66X107!0m 0 10 20 B (10—'2m) Fic. 7. Scatchard plots of the binding of rat FSH to the particulate fraction of testicular homogenates of intact control mice, hypophysectomized mice and FSH-treated hypophysectomized mice. Hypophysectomy was performed at 90 days of age. Hypophysectomized mice were given injections of 50ug NIH- FSH-P-2 twice daily for 10 days. B, Concentration of bound hormone at apparent equilibrium; F, concentration of free hormone at apparent equilibrium. See [103] for details. purified Sertoli cells of mice in culture, and proposed a kinetic model for the intracellular processes. The model clearly shows that the internalization of ligand-receptor complexes and the degradation of the complexes in lysosomes are important processes for down-regulation [80, 81]. In LH receptors, a number of investigators have also demonstrated that systemic injections of LH or human chorionic gonadotro- pin decrease the apparent number of LH-binding sites in the testis of rats [16, 22, 38, 41, 68, 77]. These findings altogether indicate that FSH or LH interacts with its receptor in Sertoli or Leydig cells and the hormone-receptor com- plexes are internalized and then degraded in lysosomes. Species difference in receptor regulation The effect of hypophysectomy on the capacity of FSH- binding sites was different among species of animals. A marked decrease in the capacity of FSH-binding sites in hypophysectomized quails suggests that up-regulation is actually manifested in the testis of quails [94, 95]. However, a reduction in the number of FSH-binding sites was less 358 K. Tsutsul AND S. KAWASHIMA obvious in hypophysectomized adult rats [92]. In contrast, both the concentration and content of FSH-binding sites increased after hypophysectomy in adult mice [103]. Tosum up, down-regulation is effectively operative in mice especially after maturation and the function of up-regulation is apparent in quails. It is well known that photosensitive birds show a rapid testicular growth when exposed to long day (LD) photoperiods [31]. Tsutsui et al. [102] stated that the testicu- lar growth during sexual maturation was much more rapid in quails and fowls than in mice and rats. The active up- regulation as a consequence of the elevation in gonadotropin level may induce rapid testicular growth in photostimulated birds [94-96]. ENVIRONMENTAL CUES AND RECEPTOR REGULATION In the majority of animals, reproductive functions do not continue throughout the year. Although the timing of the onset and termination of reproductive function depends on a complex interaction of internal biological rhythms with a variety of environmental cues, including day length, ambient temperature, and the availability of food, seasonal changes in the time of sunrise and sunset (photoperiod) appear to be most important in many species. For investigations on the role of environmental cues in the regulation of gonadotropin receptors in mammals, the hamster has served as an excellent model. Tsutsui et al. [99] observed that photoperiod is a more important environmental factor than ambient temperature for the regulation of FSH receptors in the Djungarian (Sibe- rian) hamster, Phodopus sungorus (Fig. 8). They studied the effects of artificial photoperiod and ambient temperature on testicular FSH receptor numbers and plasma FSH levels in adult males. In their experiments, hamsters were transfer- red to LD photoperiods (16-hr light, 8-hr dark) after adapta- tion in SD photoperiods (8-hr light, 16-hr dark), but the ambient temperature was maintained at 25°C. An increase in the content of FSH-binding sites and testicular growth occurred after transfer to LD, concomitant with the increase in circulating FSH levels. When hamsters reared under LD were transferred to SD, the content of FSH-binding sites and the plasma FSH level decreased. In contrast to the photo- periodic influence, they could not detect any clear-cut in- fluence of different ambient temperatures on the capacity of FSH binding in adult Djungarian hamsters (Fig. 8) [99]. An inhibitory effect of SD on the content of FSH-binding sites in the golden (Syrian) hamster, Mesocricetus auratus was also reported by Amador et al. [2]. Photoperiod-related changes in the testicular FSH receptor levels and circulating FSH levels suggest that up-regulatory mechanism of FSH on FSH receptors does exist in the hamster. Information on the effect of artificial photoperiod on the number of testicular LH receptors in the hamster has also been accumulated. A number of studies demonstrated that testicular levels of LH-binding sites declined after transfer to Testes weight r=) 0.5 (x) o fo) Specific FSH binding per 5 mg tissue (cpmx10°) rs) i (=) testes (cpm x10”) = (=) Specific FSH binding per I) oO RK LD LD SD SD NC: 25C iy BSC aie Fic. 8. Effects of photoperiod and ambient temperature on the testicular weight, specific binding of labeled rat FSH per 5 mg tissue, and specific binding of labeled rat FSH per two testes in hamsters. Incubation for 120 min at 35°C. Significant differ- ences from matched LD groups: **P<0.01. See [99] for de- tails. SD in the golden hamster [6, 7, 9, 10, 85] and Djungarian hamster [100]. Tsutsui et al. [100] further demonstrated in the Djungarian hamster that LD exposure after adaptation to SD induced an increase in the capacity of LH binding. Experimental evidence suggests that PRL is a major stimula- tor of the number of LH-binding sites in the hamster [7, 8]. Yellon and Goldman [109] reported in the hamster that circulating PRL concentration was higher after LD exposure than after SD exposure. Similar photoinduced changes in the testicular gonado- tropin receptor number have been extensively reported in several species of temperate zone birds (white-crowned spar- row [45]; domestic quail [94]; domestic fowls [47]). Recent- ly, Kawashima et al. [48] have also demonstrated that photo- period regulates the number of FSH receptors in the testis of a subtropical bird (Indian weaver bird, Ploceus philippinus). In addition, Ishii [43] reported that testicular responsiveness to gonadotropins was elevated in photostimulated white- crowned sparrows with growing testes. As indices of gonad- otropin sensitivity of the testis, rates of changes in testoster- one release and cAMP accumulation in excised testes incu- Regulation of Gonadotropin Receptors 359 bated with graded doses of FSH were employed in his study. Furthermore, the parallel increase in the circulating gonado- tropin level and the number of FSH receptors in male Indian weaver birds subjected to LD [48] agrees well with the findings of Tsutsui and Ishii [94—96] that in the quail FSH and testosterone act synergistically on Sertoli cells to increase FSH receptor numbers and to elevate the testicular respon- siveness to FSH. CONCLUSIONS Gonadotropins exert their action after binding to specific membrane receptors in target cells of the gonad. Testicular Sertoli and Leydig cells are target cells for FSH and LH, respectively. From the studies with laboratory or domestic animals, FSH receptors begin to appear in Sertoli cell during fetal life. Steady increase in the number of FSH receptors takes place during testicular growth of postnatal life. After the puberty the adult FSH receptor levels are maintained. Ontogenetic and developmental changes in LH receptors are almost similar to those in FSH receptors. Sertoli cell- to -Leydig cell cooperation contributes to the gonadotropin receptor induction. A homologous hormone, FSH, and a heterologous hormone, testosterone secreted by Leydig cells, act to Sertoli cells to induce FSH receptors in the developing testis (up-regulation). Testosterone acting on chromatin, may specifically synthesize or activate FSH receptors. FSH may activate general protein synthesis nonspecifically acting at some level after transcription. The LH receptor synthesis or activation can be controlled by several heterologous hor- mones including PRL and FSH. PRL can act directly on Leydig cells and FSH action on LH receptor induction may be mediated by Sertoli cells. In addition to up-regulation of gonadotropin receptors by homologous and heterologous hormones, down-regulation of FSH and LH on their own receptors becomes apparent especially after maturation. The circulating gonadotropin level may be a main determinant of the adult level of gonadotropin receptors. The acceleration of internalization of hormone-receptor complexes seems to be an important cause for down-regulation. Unlike laboratory or domestic animals, the reproductive function in the majority of wild animals does not continue throughout the year but is confined to a fixed short breeding period in the majority of animals. Seasonal changes in endocrine and gametogenic function of the testis in wild animals are associated with the changes in gonadotropin levels and the availability of gonadotropin receptors in the testis. Photoperiod is a more important environmental fac- tor than temperature for the regulation of gonadotropin receptors. To conclude, the testicular responsiveness to gonadotro- pins in both domestic and wild animals is generally correlated with the level of gonadotropin receptors. Thus, not only circulating levels of gonadotropins but also capacity of their receptors must be taken into account for the elucidation of testicular function. ACKNOWLEDGMENTS This study was supported in part by the Japan Society for the Promotion of Science as the International Joint Research Project, and Grants-in-Aid from the Ministry of Education, Science, and Culture, Japan. We are grateful to Prof. S. Ishii (Department of Biology, Waseda University) for his constant collaboration during this study and valuable advice in preparing the manuscript. Cordial thanks are also due to Dr. S. Raiti, National Institute of Arthritis, Diabetes, and Digestive and Kidney Diseases (NIADDK) and Dr. A. F. Parlow, Pituitary Hormones/ Antisera Center, Harbor-UCLA Medical Center for the supply of pituitary hormones. We are grateful to Prof. R. N. Saxena (Department of Zoology, University of Delhi, India) and to Prof. T. Oishi (Department of Biology, Nara Women’s University) for their collaborations. REFERENCES 1 Adashi EY, Hsueh AJW (1984) In“The Receptors” Ed by PJ Conn, Academic Press, London, Vol 1, pp 587-636 2 Amador A, Bartke A, Klemcke HG, Siler-Khodr TM, Stallings MH (1985) J Reprod Fertil 74: 693-701 3 Armstrong DT, Moon YS, Fritz IB, Dorrington JH (1975) In“Current Topics in Molecular Endocrinology Vol 2” Ed by FS French, V Hansson, EM Ritzen, SN Nayfeh, Plenum Press, New York, pp 117-191 4 Barenton B, Hochereau-de Reviers MT, Perreau C, Saumande J (1983) Endocrinology 112: 1447-1453 5 Barenton B, Pelletier J (1983) Endocrinology 112: 1441-1446 6 Bartke A, Goldman BD, Bex FJ, Kelch RP, Smith MS, Delterio S, Doherty PC (1980a) Endocrinology 106: 167-172 7 Bartke A, Goldman BD, Klemcke HG, Bex FJ, Amador AG (1980b) In“Functional Correlates of Hormone Receptors in Reproduction” Ed by VB Mahesh, TG Muldoon, BB Saxena, WA Sadler, Elsevier North Holland, Inc., New York, pp 171- 185 8 Bartke A, Klemcke HG, Amador A, van Sickle A (1982) Ann NY Acad Sci 383: 122-134 9 Bex FJ, Bartke A (1977) Endocrinology 100: 1223-1226 10 Bex FJ, Bartke A, Goldman BD, Dalterio S (1978) Endocri- nology 103: 2069-2080 11 Bhalla VK, Reichert LE Jr. (1974) J Biol Chem 249: 43-51 12 Bona Gallo A, Licht P (1979) Gen Comp Endocrinol 37: 521- 532 13 Catt KJ, Dufau ML, Neaves WB, Walsh PC, Wilson JD (1975) Endocrinology 97: 1157-1165 14 Catt KJ, Dufau ML, Tsuruhara T (1972) J Clin Endocrinol 34: 123-132 15 Catt KJ, Harwood JP, Clayton RN, Davies TF, Chan V, Katikineni M, Nozu K, Dufau ML (1980) Rec Progr Horm Res 36: 557-622 16 Chen Y-D, Payne AH (1977) Biochem Biophys Res Commun 74: 1589-1596 17 Chen Y-D, Payne AH, Kelch RP (1976) Proc Soc Exp Biol Med 153: 473-475 18 Cheng KW (1975) J Clin Endocrinol Metab 41: 481-491 19 Chowdhury M, Steinberger E (1976) J Endocrinol 69: 381- 384 20 Clausen OPF, Purvis K, Hansson V (1981) Acta Endocrinol 96: 569-576 21 Closset J, Hennen G (1989) J Endocrinol 120: 89-96 360 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 535) 54 K. TsutsuI AND S. KAWASHIMA Cooke BA, Rommerts FFG (1988) In “Hormones and Their Actions, Part II” Ed by BA Cooke, RJB King, HJ van der Molen, Elsevier, Amsterdam, pp 155-180 Dahl KD, Hsueh AJW (1988) In “Hormones and Their Actions, Part II” Ed by BA Cooke, RJB King, HJ van der Molen, Elsevier, Amsterdam, pp 181-192 Desjardins C, Zeleznik AJ, Midgley AR Jr. (1974) In“Cur- rent Topics in Molecular Endocrinology Vol1” Ed by ML Dufau, AR Means, Plenum Press, New York, pp 221-235 Dorrington JH, Armstrong DT (1975) Proc Natl Acad Sci USA 72: 2677-2681 Dorrington JH, Fritz IB, Armstrong DT (1976) Mol Cell Endocrinol 6: 117-128 Dorrington JH, Roller NF, Fritz IB (1975) Mol Cell Endocri- nol 3: 57-70 Eskeland NL, Lugo DI, Pintar JE, Schachter BS (1989) En- docrinology 124: 2914-2919 Etches RJ, Cheng KW (1981) J Endocrinol 91: 11-22 Fabbri A, Tsai-Morris CH, Luna S, Fraioli F, Dufau ML (1985) Endocrinology 117: 2544-2546 Follett BK, Farner DS (1966) Gen Comp Endocrinol 7: 111- 124 Frowein J, Engel W (1974) Nature 249: 377-379 Gariyali V, Saxena RN (1975) Proc All India Rodent Semin 2: 133-139 George FW, Catt KJ, Neaves WB, Wilson JD (1978) Endo- crinology 102: 665-673 Grover JL (1985) Ph D thesis, Univ of Delhi, India Hansson V, French FS, Weddington SC, Nayfeh SN, Ritzen EM (1974) In“Current Topics in Molecular Endocrinology Vol1” Ed by ML Dufau, AR Means, Plenum Press, New York, pp 287-290 Hansson V, Weddington SC, McLean WS, Smith AA, Nayfeh SN, French FS, Ritzen EM (1975) J Reprod Fertil 44: 363- 375 Haour F, Saez JM (1977) Mol Cell Endocrinol 7: 17-24 Hodgson Y, Robertson DM, de Kretser DM (1983) Int Rev Physiol 27: 275-327 Hsueh AJW, Adashi EY, Jones PBC, Welsh TH (1984) Endocr Rev 5: 76-127 Hsueh AJW, Dufau ML, Catt KJ (1976) Biochem Biophys Res Commun 72: 1145-1153 Huhtaniemi IT, Korenbrot CC, Jaffe RB (1977) J Clin En- docrinol Metab 44: 963-967 Ishii S (1980) In“Avian Endocrinology Vol4” Ed by DS Farner, JR King, KC Parkes, Academic Press, pp 1-15 Ishii S, Adachi T (1977) Gen Comp Endocrinol 31: 287-294 Ishii S, Farner DS (1976) Gen Comp Endocrinol 30: 443-450 Ishii S, Hirano S (1979) Gen Comp Endocrinol 37: 137-140 Ishii S, Tsutsui K, Adachi T (1978) In“Comparative Endocri- nology” Ed by PJ Gaillard, HH Boer, Elsevier, Amsterdam, pp 73-76 Kawashima S, Tsutsui K, Saxena RN, Ishii S (1993) Gen Comp Endocrinol 92: 250-259 Kawashima S, Yamada Y, Tsutsui K (1986) J Sci Hiroshima Univ, Ser B Div 1, 32: 271-282 Kerr JB, Sharpe RM (1985) Endocrinology 116: 2592-2604 Ketelslegers J-M, Hetzel WD, Sherins RJ, Catt KJ (1978) Endocrinology 103: 212-222 Kikuchi M, Ishii S (1989) Biol Reprod 41: 1047-1054 Klemcke HG, van Sickle M, Bartke A, Amador A, Chan- drashekar V (1987) Biol Reprod 37: 356-370 Magre S, Jost A (1980) Arch Anat Micros Morphol Exp 69: 297-318 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 iS) 76 77 78 79 80 81 82 83 84 85 86 87 Means AR, Dedman JR, Tash JS, Tindall DJ, van Sickle M, Welsh MJ (1980) Ann Rev Physiol 42: 59-70 Means AR, Tindal DJ (1975) In“Current Topics in Molecular Endocrinology Vol2” Ed by FS French, V Hansson, EM Ritzen, SN Nayfeh, Plenum Press, New York, pp 383-398 Means AR, Vaitukaitis J (1972) Endocrinology 90: 39-46 Mendelson C, Dufau M, Catt K (1975) J Biol Chem 250: 8818-8823 Mori H, Tsutsui K, Kawashima S (1985) J Sci Hiroshima Univ, Ser B, Div 1 32: 143-156 Mulder E, Peters MJ, van der Molen HJ (1987) In “Current Topics in Molecular Endocrinology Vol 2” Ed by FS French, V Hansson, EM Ritzen, SN Nayfeh, Plenum Press, New York, pp 287-291 Nakhla AM, Mather JP, Janne OA, Bardin CW (1984) En- docrinology 115: 121-128 Norton JN, Skinner MK (1989) Endocrinology 124: 2711- 2719 Orth J (1982) Anat Rec 203: 485-492 Orth J, Christensen AK (1977) Endocrinology 101: 262-278 Orth JM (1984) Endocrinology 115: 1248-1255 O’Shaughnessy PJ, Brown PS (1978) Mol Cell Endocrinology 12: 9-15 Pahnke VG, Leinberger FA, Kunzig HJ (1975) Acta Endocri- nol 79: 610-618 Purvis K, Clausen OPF, Olsen A, Haug E, Hansson V (1979) Arch Androl 3: 219-230 Reichert LE Jr., Bhalla VK (1974) Endocrinology 94: 483- 491 Reichet LE Jr., Dattatreyamurty B (1989) Biol Reprod 40:13-26 Richards JS (1980) Physiol Rev 60: 51-83 Richards JS, Irland JJ, Rao MC, Bernath GA, Midgley AR, Reichert LE Jr. (1976) Endocrinology 99: 1562-1570 Richards JS, Kersey KA (1979) Biol Reprod 21: 1185-1201 Risbridger GP, Hodgson YM, de Kretser DM (1981) In“The Testis” Ed by H Burger, DM de Kretser, Raven Press, New York, pp 195-212 Ritzhaupt LK, Bahr JM (1987) J Endocrinol 115: 303-310 Sanborn BM, Elkington JSH, Steinberger A, Steinberger E (1975) In“Current Topics in Molecular Endocrinology Vol 2” Ed by FS French, V Hansson, EM Ritzen, SN Nayfeh, Plenum Press, New York, pp 293-309 Sharpe RM (1976) Nature 264: 644-646 Sharpe RM (1982) In “Oxford Reviews in Reproductive Biology Vol4” Ed by CA Finn, Oxford University Press, Oxford, pp 241-317 Sharpe RM (1986) Clin Endocrinol Metab 15: 185-207 Shimizu A, Kawashima S (1989) J Biol Chem 264: 13632- 13639 Shimizu A, Kawashima S (1989) J Biol Chem 264: 13639- 13641 Shimizu, A, Tsutsui K, Kawashima S (1987) Endocrinol Jpn 34: 431-442 Skinner MK (1988) Internat Congr of Endocrinol, Kyoto, Japan, pp 51 (Abstract) Sprengel R, Braun T, Nikolics K, Segaloff DL, Seeburg PH (1990) Mol Endocrinology 4: 494-499 Stallings MH, Matt KS, Amador A, Bartke A, Soares MJ, Talamantes F (1985) J Reprod Fertil 75: 663-670 Steinberger A, Steinberger E (1971) Biol Reprod 4: 84-87 Steinberger A, Steinberger E (1972) In“Reproductive Biolo- gy” Ed by H Balin, S Glasser, Excerpta Medica, Amsterdam, pp 144-267 88 89 90 91 92 93 94 95 96 97 98 99 100 Regulation of Gonadotropin Receptors 361 Steinberger A, Thanki KH, Siegal B (1974) In“Current Topics in Molecular Endocrinology Vol 1” Ed by ML Dufau, AR Means, Plenum Press, New York, pp 177-192 Takase M, Tsutsui K, Kawashima S (1990a) J Exp Zool 256: 200-209 Takase M, Tsutsui K, Kawashima S (1990b) Endocrinol Jpn 37: 193-203 Thanki KH, Steinberger A (1978) Andrologia 10: 195-202 Tsutsui K (1991) Endocrinology 128: 477-487 Tsutsui K (1992) J Exp Zool 264: 167-176 Tsutsui K, Ishii S (1978) Gen Comp Endocrinol 36: 297-305 Tsutsui K, Ishii S (1980) J Endocrinol 85: 511-518 Tsutsui K, Ishii S (1985) In‘“Current Trends in Comparative Endocrinology” Ed by B Lofts, WN Holmes, Hong Kong Univ Press, Hong Kong, pp 761-762 Tsutsui K, Kawashima S (1987) Endocrinol Jpn 34: 717-725 Tsutsui K, Kawashima S, Kumar V, Kapania R, Saxena RN (1989a) Gen Comp Endocrinol 73: 442-451 Tsutsui K, Kawashima S, Masuda A, Oishi T (1988) Endocri- nology 122: 1094-1102 Tsutsui K, Kawashima S, Masuda A, Oishi T (1989b) J Exp Zool 251: 91-10 101 102 103 104 105 106 107 108 109 110 Tsutsui K, Kawashima S, Saxena RN, Ishii S (1992) Gen Comp Endocrinol 88: 444-453 Tsutsui K, Shimizu A, Kawamoto K, Kawashima S (1983) In“Recent Trends in Life Sciences” Ed by A Gopalakrishna, SB Singh, AK Saxena, Manu Publications, Kanpur, pp 93-102 Tsutsui K, Shimizu A, Kawamoto K, Kawashima S (1985) Endocrinology 117: 2534-2543 Tsutsui K, Shimizu A, Kawashima S, Ishii S (1986) 3: 497-501 van Beurden, Roodnat WMO, van der Molen HJ (1978) Int J Androl (Suppl) 2: 374-382 van Sickle M, Oberwetter JM, Birnbaumer L, Means AR (1981) Endocrinology 109: 1270-1280 Warren DW, Dufau ML, Catt KJ (1982) Science 218: 375- 377 Warren DW, Huhtaniemi IT, Tapanainen J, Dufau ML, Catt KJ (1984) Endocrinology 114: 470-476 Yellon SM, Goldman BD (1984) Endocrinology 114: 664-670 Zipf WB, Payne AH, Kelch RP (1978) Endocrinology 103: 595-600 Zool Sci ZOOLOGICAL SCIENCE 11: 363-373 (1994) © 1994 Zoological Society of Japan REVIEW Gonadotropin-Releasing Hormone: Present Concepts, Future Directions* RAKESH K. RAstToci and LuIsA IELA Department of Zoology, University of Naples “Federico II’, via Mezzocannone 8, 80134 Napoli, Italy ABSTRACT— There is an overwhelming amount of evidence to indicate that gonadotropin-releasing hormone (GnRH), a decapeptide, is found in multiple molecular forms, and is vital for the functional integration of brain-pituitary-gonadal axis in vertebrates. In simple terms, there is an overall agreement that GnRH acts as a neuroendocrine regulator of pituitary gonadotropin secretion, gonadal steroid secretion, sexual behavior, and reproduction. GnRHs are distributed widely within the vertebrate body, particularly in the brain. The brain GnRH neuronal system(s) varies in its morphology, ontogenesis and function across vertebrates. It is a highly dynamic structure which does not function at the same level throughout life. A large framework of studies completed to date attests to the emerging concept that GnRH neuronal system is regulated by a complex neural circuitry, comprised of diverse neurochemical signals, which may provide excitatory or inhibitory input to GnRH neurons. While general considerations on GnRH systems may be similar among vertebrates, it must not seduce us to generalize the more specific details. In fact, there may occur ontogenesis and reproductive status-related changes and a timetable of complex neuroendocrine events that are probably (certainly) species-specific. INTRODUCTION Gonadotropin-releasing hormone (GnRH=LHRH), a 10-amino acid bioregulatory neuropeptide, has a widespread occurrence within the living kingdom, and is not restricted just to vertebrates. In fact, from the evolutionary viewpoint one can go as far back as the yeast cells in which Loumaye et al. [55] demonstrated that a mating factor, called a-factor, necessary for sexual reproduction, has a strong similarity to GnRH, as evaluated by its ability to influence pituitary gonadotropin (GtH) secretion. Brain is certainly the most complex organ of the verte- brate body. Within this highly specialized tissue there are cells characterized by their ability to synthesize GnRH. Originally isolated as a brain peptide, GnRH is well known for its regulatory role in the release of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) from the ante- rior pituitary (pars distalis, adenohypophysis, ventral lobe). Since the isolation and structural elucidation of the hypotha- lamic GnRH in the early-1970s there have been numerous investigations to unveil the distribution pattern of GnRH-like peptide(s) in the brain of all vertebrate classes, from mam- mals down to cyclostomes [4, 6, 10, 22, 23, 44, 64, 77, 87]. During evolution this neuropeptide has undergone gene duplication with consequent structural diversification which is evident in the presence of multiple molecular forms across vertebrate species [see 45]. So far eight GnRH molecular forms have been described in the vertebrate brain, character- ized and named for the vertebrate species in which they were Received May 9, 1994 * This paper is dedicated to our mentor Professor Giovanni Chieffi. first identified: mammalian (mGnRH), chicken I (CGnRH-I), chicken II (cGnRH-II), salmon (sGnRH), lamprey I and III (IGnRH), catfish (cfGnRH) and dogfish (dfGnRH) [10, 44-46, 64, 87, 86]. From a phylogenetic viewpoint cGnRH-II is considered to be the most conserved GnRH across vertebrates and residues 5, 7 and 8 vary among different forms. So far, only cGnRH-II has been described in all jawed vertebrates. LLamprey GnRH, however, varies from cGnRH-II in positions 3, 5, 6 and 8, and its presence has been demonstrated in cyclostomes only. An endogenous posttransational product of the GnRH precursor has been characterized in mammalian and frog hypothalamus: (Hydroxyproline’)GnRH [29]. More recently, in Xenopus laevis, this form was shown to be distributed in the forebrain, midbrain and hypothalamus [46]. This peptide, as well as the C-terminal fragments of GnRH, are supposed to enhance sexual behavior as GnRH itself [30]. Novel GnRH forms described in oviparous mammals, reptiles, amphibians, bony and cartilaginous fishes and cyclostomes are yet to be char- acterized and still other GnRH forms are expected to be discovered [45—47, 64, 75]. Furthermore, over a couple of thousand analogs of GnRH (GnRHa) have been synthesized whose availability becomes a powerful tool in understanding the regulatory roles and mechanisms of action of GnRH [42, 83]. The molecular heterogeneity is apparently the basis for a variety of regulatory functions of GnRH: as the stimulator of the reproductive system and sexual behavior, as a neuro- modulator and/or neurotransmitter in the central and sym- pathetic nervous system, and as a paracrine /autocrine regula- tor in the pituitary, gonads, placenta, and in tumor cells. This article is intended to give a “bird’s eye view” of the morpho-functional features of GnRH _ neuronal systems 364 R. K. RastoGi AND L. IELA across vertebrates, and to spotlight the new knowledge that has emerged from latest studies. DISTRIBUTION Evidences based upon radioimmunoassay (RIA), chro- matography (HPLC), in situ hybridization histochemistry (ISHH) and immunohistochemistry (ICC) techniques indi- cate the presence of GnRH-like peptide(s) in several brain as well as outside brain areas of a variety of vertebrate species. Localization of GnRH-producing neurons and their projec- tions has been studied in the brain of mammals, birds, reptiles, amphibians, bony and cartilaginous fishes and cy- clostomes [4, 6, 10, 19, 21-23, 40, 50, 56, 57, 64, 68, 77, 78, 87]. In several vertebrates HPLC/RIA analyses have con- firmed ICC data as far as the distribution of GnRH-like material in the brain is concerned. However, in some species of birds, reptiles and amphibians ICC and HPLC/ RIA data have shown discrepancy regarding the presence and distribution of different molecular forms of GnRH [19, 24, 28, 57, 61, 67, 75, 78, 101, unpublished data]. Modern ICC techniques are considered highly efficient, and yet we can not exclude the possibility that there may be GnRH neurons which fall below the level of detectability of the ICC procedures. Taking into account this reservation, at the present time GnRH-containing neurons and their projections have been described in the rhinencephalon, tel- encephalon, diencephalon, mesencephalon, metencephalon and myelencephalon [6, 10, 19, 21, 23, 57, 64, 87]. Although there are remarkable interspecies differences, a large hypothalamic population constitutes the biggest group of GnRH neurons in the brain of most vertebrates. In the median eminence (ME) no GnRH cell bodies are found. Among mammals, only musk shrew brain contains a cluster of GnRH neurons in the mesencephalon [21], and it is not known whether this feature is common to other primitive mammals. No GnRH neurons have been described pos- terior to the midbrain in any mammalian species. GnRH neurons in the forebrain, however, send fiber projections to the midbrain in all mammals. In nonmammalian verte- brates, except urodele amphibians [19] and cyclostomes [48], a mesencephalic group of GnRH neurons has been described [50, 57, 61, 64, 65, 101, 106, unpublished data]. Projections of midbrain GnRH neurons in bony fishes may innervate the caudal neurosecretory system (urophysis) [23]. Conspic- uous differences in the distribution pattern have been de- scribed within each vertebrate subgroup. For example, among anuran amphibians, Rana esculenta brain shows perhaps the most extensive network of GnRH neurons and fibers [24, 78], whereas Pachymedusa dacnicolor brain con- tains only a small number of GnRH neurons all located exclusively in the anterior preoptic area (POA) [40] from where axonal projections reach the ME. Incyclostomes, the POA-located GnRH neurons project to the neurohypophysis (PN; pars nervosa) [48]. In most vertebrates GnRH fiber endings in the ME are derived from the forebrain-located neurons, mainly in the hypothalamus and septum, and only in bony fishes do the POA-located GnRH neuron projections directly contact pituitary gonadotropes [23]. Several studies have established that the populations of GnRH cells in the forebrain that project to the pituitary, PN and/or ME are often completely separate from those GnRH neurons that give rise to brain stem and/or spinal cord projections [65]. In fact, a growing body of evidence indicates that there are subpopulations of anatomically distinct GnRH-neuronal sys- tems in the vertebrate brain, and that GnRH neurons may be unipolar, bipolar or multipolar. At a subcellular level GnRH immunoreactivity is distributed around the outside of the nuclear envelope, associated particularly with the rough endoplasmic reticulum, secretory vesicles and Golgi appa- ratus, and in the neuronal projections [54, 59, 110]. Sex- related differences in subcellular organelles have been de- scribed in mares and stallions [59]. To the best of our knowledge, no ultrastructural study of GnRH neurons is available in nonmammalian vertebrates. Numerous studies have demonstrated that more than one GnRH molecular form may be present in the brain, and there are several reports on a differential distribution of GnRH variants in distinct brain areas [10, 13, 44, 46, 61, 65, 96, 109]. Moreover, in the lizard, Podarcis sicula, more than one GnRH form were colocalized in the same neuron [57]. In the musk shrew, mGnRH is distributed all through the forebrain and diencephalon, whereas midbrain-located neurons contain only cGnRH-II [21]. In the midbrain of adult chicken only cGnRH-II is present, while cGnRH-I is present chiefly in the forebrain, diencephalon and ME [43]. However, in a later study, van Gils et al. [101] detected cGnRH-II in the ME of chicken as well as Japanese quail. Similarly, Millam ef al. [61] did not find cGnRH-II in the ME of turkey by ICC, while in a later study RIA analysis showed the presence of cGnRH-II in the posterior pituitary [El Halawani, unpublished, see 60]. Further, in a turtle species, cGnRH-I and II are differentially distributed, the former being most concentrated in the ME in a ratio of 8:1 against the latter which is more abundant in the caudal regions [96]. Likewise, in _X. /Jaevis mGnRH is distributed throughout the brain, whereas cGnRH-II is more concentrated in the mid- brain and hindbrain [46]. In the frog, R. ridibunda, cGnRH-II neurons, but not mGnRH neurons innervate the neurointermediate lobe of the pituitary; similarly, in R. esculenta mGnRH neurons project to ME and PN [78], and using a highly specific cGnRH-II antiserum GnRH neurons were revealed in the midbrain tegmentum and fiber endings in the pars intermedia [unpublished]. Differential distribu- tion of GnRHs has been described in the brain of bony fishes as well [see 45, 47]. Recently, ring dove brain has been shown to contain GnRH-like material-containing nonneuron- al cells [88]. Characterized as mast cells, component of the immune system, they are distributed mainly in the medial habenula. This interesting finding could be leading to in- vestigate other vertebrate groups as well. How does brain GnRH reach the target organs? In GnRH in Vertebrates 365 tetrapods GnRH is conveyed to the pituitary via the hypotha- lamo-hypophyseal portal vessels; however, there are evi- dences to indicate that GnRH can also be secreted into the cerebral ventricles as well as in brain vessels. In bony fishes, which lack the ME, GnRH neurons are found to directly innervate pituitary gonadotropes. In elasmobranchs, the ventral lobe of the pituitary does not receive a portal supply, and GnRH is evidently released into the general circulation, where, in fact, GnRH and GnRH-binding protein molecules have been determined [83]. In cyclostomes, King et al. [48] have suggested that GnRH can be released into the third ventricle and transported by tanycytes to the pituitary; GnRH may also reach pituitary by simple diffusion from neuronal projections in the PN. In elasmobranchs, GnRH may reach pituitary and gonads via general circulation, and its presence has been ascertained in the cerebrospinal fluid [86, 107]. In tetrapods and bony fishes, GnRH can act on the gonads as a paracrine regulatory factor, and the idea is reinforced by the presence of GnRH-like molecules as well as of GnRH- binding sites in the gonads [27, 32, 35]. GnRH neurons have been described in the terminal nerve of mammals, birds, amphibians, and bony and cartila- ginous fishes [19, 23, 64, 67, 103]. Fibre projections from these neurons may reach areas as far ahead as nasal capsule and as far behind as rhombencephalon, except the pituitary [69, 103]. In the dwarf gourami, a bony fish, using whole brain in vitro, it was shown that terminal nerve GnRH neurons may be the most extensively projecting GnRH cells in the brain [69]. Reptiles and cyclostomes remain the only vertebrate taxa in which GnRH neurons have not been described yet in the terminal nerve. In cyclostomes, howev- er, the presence of a terminal nerve-like structure is still a question of debate. Besides the brain and terminal nerve, GnRH-like mate- rial has been detected in a variety of other structures: gonads, mammary gland, tumor cells, human placenta and pancreas, follicular fluid, milk, olfactory epithelium, sympathetic gang- lia, adrenal gland, liver, intestine and retina [10, 12, 35, 44]. GnRH-like material outside the brain and terminal nerve may be similar to or different than one of the GnRH variants characterized in the brain. ONTOGENESIS The embryonic origin(s) of GnRH neurons has received considerable interest in recent years. Several reports have described that these neurons, unique among brain cells, originate not in the brain but in the olfactory placode and migrate into the forebrain/diencephalon along the olfactory/ terminal nerve. The extracranial origin, time course and route of GnRH neuronal migration has been clarified in some species of mammals, birds and amphibians [19, 64, 65, 84, 92]. In birds and urodele amphibians, this line of evidence has been confirmed by surgical removal of the olfactory placode which results in the elimination of GnRH neurons in the forebrain and diencephalon [1, 63, 64]. In anuran amphibians, however, the midbrain-located group of GnRH neurons is supposed to have an intracranial origin [18, 65, unpublished data]. Among tetrapods, the reptilian GnRH neuronal system is morphologically “atypical” in that GnRH neurons are mainly located in the midbrain and infundibulum [57], and they are not considered to have an extracranial origin [unpublished data]. Interestingly, the midbrain clus- ter of GnRH neurons is the first to appear during lizard ontogenesis, followed by their appearance in the infundibu- lum; we argue that these neurons may take origin in the nearby neuroepithelium [unpublished data]. However, the ontogenesis of GnRH neuronal system in reptiles need be investigated in more species to unequivocally clarify the intracranial origin of GnRH neurons in this group. In a bony fish, Pterophyllus scalare, the first ontogenetic appear- ance of GnRH-immunoreactive cells in the pituitary precedes that in the POA and this may be indicative of still another site of origin for GnRH-producing cells [16]. No information on the extracranial origin of GnRH neurons in the brain is available for cyclostomes; in this group Muske [64] suggests another line of origin of the POA-located GnRH neurons, and that is from the ventricular ependyma because of the periventricular localization of GnRH cell bodies. Needless to say, the phylogenetic picture of the embryonic origin of GnRH neuronal systems in the vertebrate brain is far from complete. Biochemical and ultrastructural differentiation of GnRH neurons, during their olfactory placode-forebrain migration, has been analyzed in the mouse [54], in which GnRH gene is expressed early in ontogenesis [107] and the differentiation of GnRH neurons continues and is coordinated during migra- tion. In this mammal, however, axonal projections are formed only upon entering the forebrain. In contrast, in the rhesus monkey axonal projections are elaborated while GnRH neurons are still in the nasal septum [82]. Similarly, in the chick [92] and newt [19] GnRH axons are seen in the terminal/olfactory nerve during early migratory stage. Prior to migration GnRH neurons are scattered as individual cell bodies in the placode, while during migration they come to lie in close apposition, as confirmed by electron microscopy [92, 110]. Cell-to-cell contact appears to be established when these neurons are located at the adult site. The nature of these contacts must be investigated at the ultrastructural level. During ontogenesis, besides the morphological and biochemical changes, a remarkable change may also occur in the number of GnRH neurons. In the mouse, in fact, GnRH cell number decreases drastically in the last stages of migration [110], and it is suggested that programmed cell death may be one reason, or that after migration some neurons stop synthesizing GnRH and become undetectable byICC. There is the need for further investigation as well as the necessity to draw more vertebrate groups in this reper- toire. 366 R. K. RAstoci AND L. IELA CORRELATES OF BIOSYNTHETIC AND SECRETORY ACTIVITY, AND FUNCTIONS Morphological and biochemical correlates Indeed, the basic theme of investigations related to the morphological, histochemical and biochemical correlates of GnRH secretion and functions has almost always been around the age, sex and reproductive status-related changes in the hypothalamo-hypophyseal-gonadal system. The va- rious steps are, in sequence, comprised of the identification of GnRH-like material in the brain, morphology, distribution and ontogenesis of the GnRH neuronal system, anatomical connections with the pituitary, and an analysis of the GnRH content in the brain, ME, portal blood and systemic circula- tion under different conditions. Among the central regula- tory mechanisms involved in the control of the development of the hypothalamo-hypophyseal-gonadal axis until sexual maturity, the importance of GnRH has been emphasized in all vertebrates [see 31, 98, 104]. In the adult, besides the pulsatile pattern of GnRH secretion, abundantly referred to in mammals, insight into the functional relationships of the hypothalamo-hypophyseal-gonadal system is also gained through evaluating seasonal pattern of GnRH secretion [see 3, 26, 28, 80]. Among tetrapods GnRH content has been evaluated in tissue extract as well as in portal blood, and only in bony and cartilaginous fishes has the radioimmunoassayble GnRH been determined in general circulation [47, 73]. It has hitherto been demonstrated that GnRH neurons undergo morphological, distributional and numerical changes associated with the developmental stage and reproductive status of the animal. Prominent seasonal cycles in GnRH cell morphology are observed in a variety of vertebrates [see 78, 98, 108]. It is evident that a particular morphological feature of the GnRH neuron may be correlated with a particular aspect of reproduction. In hibernating mammals, degranulation of GnRH cells occurs during hibernation, and increased storage after arousal. In the male Djungarian hamster, there occurs an increase in the number of unipolar neurons correlated with the onset of puberty, whereas bipolar GnRH neurons increase only in postpubertal period [108]. An ultrastructural analysis of GnRH neurons in the pony brain has, furthermore, indicated that irregularly-contoured cells in the POA/organum vasculosum lamina terminalis may have higher synthetic activity as compared to most other neurons, and it is suggested that GnRH neurons may utilize ultrashort feedback [59]. In the Syrian hamster, it was ascertained that the inhibitory effects of short days on the reproductive axis are mediated through a suppression of GnRH neurons which, in turn, is reflected as an increase in the net content of GnRH within the brain [97]. In the cow, a quantitative light microscopical study describing morpholog- ical changes in GnRH neurons supports the hypothesis of reduced activity of GnRH neurons during early to middle stages of the puerperium [51] which is in line with the concept of postpartum infertility due to the suckling stimulus- mediated suppression of GaRH, and consequently of LH, secretion. Functions In all major vertebrate groups, the effects of exogenous GnRH/GnRH, have been investigated in vivo and in vitro. These comprise stimulation of pituitary GtH and, in some cases, growth hormone secretion, stimulation of pituitary- thyroid axis, gonadal steroidogenesis, gametogenesis, sper- miation, ovulation, and sexual behavior. However, ex- tremely varied experimental protocols, as well as different GnRH forms and/or GnRH, have been used, and thus it becomes rather arduous to unify these data in order to draw generalized considerations. Nevertheless, all native GnRH variants appear to have at least one characteristic in common, and that is the stimulation of GtH release from the pituitary. A series of studies have also provided evidences that naturally occurring GnRH variants may exhibit different biological potencies in different species in terms of their ability to enhance pituitary GtH secretion [see 10, 53, 73]. In a variety of mammals, including man, and in the domestic fowl, reproductive function can be inhibited by a prolonged treatment with GnRH/GnRH, [93, 94]. In the adult, initially, GnRH administration enhances pituitary GtH secretion but a continuous exposure to GnRH eventually leads to desensitization of the pituitary to GnRH with the consequent suppression of gonadal function [102]. It is also known that GnRH/GnRHz, exert a differential control over FSH and LH. This may be credited to the fact that GnRH is released in pulses, inducing a pulsatile pattern of LH release, but not of FSH. Nonetheless, chronic GnRH treatment suppresses both LH and FSH. In contrast, in amphibians, it was demonstrated that pituitary is relatively resistant to desensitization due to chronic in vivo GnRH exposure which enhances GtH biosynthesis and secretion [53, 89]. Howev- er, in the goldfish GnRH desensitization has been reported [34]. Further, in some mammals and birds evidences are that, during ontogenesis, GnRH plays a role in the develop- ment of pituitary gonadotropes, and that GnRH, at least in the rat, is important for maintaining FSH synthesis [see 9, 104]. Differential distribution, distinct roles Do naturally occurring GnRH variants play distinct roles? The answer is yet far from clear. However, it is becoming increasingly evident that native GnRHs are dif- ferentially distributed within the brain of jawed vertebrates [13, 21, 43, 46, 61, 96]. In the musk shrew, and some species of birds, reptiles, amphibians and bony fishes a quantitative predominance of cGnRH-II is found not in the forebrain, but in the midbrain and/or hindbrain, and this has led to the assertion that cGnRH-II may have an extrapituitary role. However, a specific role for this form has not yet been established. In birds, in which cGnRH-I seems to pre- dominate in the forebrain/hypothalamus, it is argued that the potential function of cGnRH-I and II may diverge early in development [43, 60]. Among amphibians, based upon the GnRH in Vertebrates 367 predominance of mGnRH in the hypothalamus in X. /aevis it was suggested that this variant may be the prime regulator of GtH release, whereas cGnRH-II, abundant in the hindbrain, may have an extrapituitary role [45]. In R. esculenta, however, it was suggested that cGnRH-II may have a hypophysiotropic activity [28]. Further, in this species it was also seen that cGnRH-I and II and mGnRH all enhanced androgen production in intact males, whereas in hypophysectomized males only cGnRH-II enhanced testicu- lar androgen production, suggesting that a cGnRH-II-like molecule, produced in the testis, may be the local paracrine regulator of testicular activity [20, 79]. None of the tetrapod species studied is responsive to |GnRH [see 53], in terms of pituitary GtH secretion. However, |GnRH enhances plas- ma androgen levels in intact male frog [20]. In the forebrain/hypothalamus of fishes the predominant form varies from mGnRH, sGnRH to catfish or dogfish GnRH, making it evident that any of these variants may have a hypophysiotropic role. However, cGnRH-II neuronal en- dings may terminate in the bony fish pituitary [see 45]. Perhaps, in lower vertebrates the specificity of GnRH variants is very low. Lovejoy et al. [56] have proposed the division of known native GnRHs in two groups: mGnRH, ceGnRH-I and catfish GnRH with hydrophilic residues, and cGnRH-II, sGnRH and dogfish GnRH with hydrophobic residues. ‘They suppose that this may lead to clarify, on a structural basis, their distinct functional roles. Naturally, 1GnRH is not included in this classification, being this variant present exclusively in cyclostomes. Interaction with sex steroids Among vertebrates, the repertoire of GnRH functions appears to be remarkably complex and involves interactions with several hormonal and other factors. GnRH acts upon the pituitary gonadotropes through interaction with mem- brane-associated high affinity receptors [14]. GnRH- modulated pituitary GtH secretion regulates gonadal activity and reproduction. Sex steroids in turn influence GnRH- regulated GtH biosynthesis and secretion. For this the sex steroid signals from the gonads must be correctly interpreted within the brain. Further, it is necessary to mention that GnRH regulation of pituitary GtH secretion can be mod- ulated by sex steroids at the level of the pituitary as well, positively or negatively [see 71]. Although GnRH neurons do not appear to have sex steroid receptors, their distribution pattern has a remarkable overlapping with that of sex steroid- concentrating neurons, and the latter may even project to other brain areas in order to transmit steroid-influenced signals [10, 22, 25, 62, 74]. Most likely, the GnRH neurons which send their axon terminal in the ME are affected by estrogen-sensitive afferent neuronal systems. Indeed, GnRH neurons in such areas may represent targets for the feedback of steroids, and this is substantiated by a recent study in the rat, in which the rostral medial POA contains estrogen receptors and the GnRH neurons situated in this area are sensitive to estrogen treatment. Changes in GnRH neuronal mRNA levels in this area of estrogen-supplemented ovariectomized rats are taken as the cellular correlates of the positive feedback effects of estrogen on GnRH neurons [74]. This experimental model may be used to examine temporal changes in GnRH gene transcription, GnRH mRNA stability and GnRH translation all contributing to the understanding as to how estrogen triggers the preovulatory hypersecretion of GnRH which leads to LH surge followed by ovulation. However, in the meanwhile pituitary sensitivity to GnRH is enhanced by preovulatory progesterone surge. Similarly, in R. esculenta, in which sex steroid-concentrating neurons abound in the anterior POA, gonadectomy in both sexes caused a severe quantitative depletion of GnRH neurons in this area, as evaluated by ICC [41]. Sex steroid-replacement therapy enhanced somal accumulation of immunoreactive material, indicative probably of an increased synthesis and storage. In R. catesbeiana, it is suggested that GnRH can enhance pituitary GtH secretion in juveniles, but biosynthesis of GtH is enhanced only at a later time, coincident with gonadal steroid production [90]. These authors suggest that androgens may augment GnRH receptor molecules in the pituitary as well as endogenous secretion of GnRH in the hypothalamus. It thus appears that GnRH-stimulated GtH release is influenced not only by age, sex, season or morpho- functional heterogeneity of pituitary gonadotropes, but also by sex steroids [see 89, 90, 96]. External factors Of the behavioral cues, courtship enhances GnRH con- centration in the terminal nerve in a urodele amphibian [76], whereas in the ring dove, it determines the appearance of nonneuronal GnRH-containing cells in the habenula [88], thus making it obvious that GnRH-producing brain areas may be differentially correlated with diverse reproductive aspects in different species. An upsurge in interest in the implication of odours in animal reproduction dates long-long back, and it is widely contended that reproductive activity is dependent upon a fully functional olfactory system, and that pheromone signals can be transduced by terminal nerve GnRH system to in- fluence reproductive activity through the modulation of GnRH secretion in the hypothalamus [22, 23, 91]. This concept is strengthened by the fact that limbic (hypothala- mus, amygdala, hippocampus, POA) and olfactory systems indeed control reproductive behavior, and extensive GnRH projections between the terminal nerve and limbic systems are well placed to create a chain through the olfactory mucosa, brain, pituitary, and on to the gonad. In a recent study on the electrical activity and morphology of terminal nerve-GnRH neurons in the dwarf gourami, it was shown that these neurons display an endogenous rhythmic discharge pattern, a feature common to all peptidergic and monoaminergic modulator neurons, but are not projected to the pituitary, and thus it was assumed that they do not function as a hypophysiotropic GnRH system, but rather as neuromodulator [69]. 368 R. K. RastoGi AND L. IELA Neural modulation of GnRH release and function Besides the fact that GnRH neurons display neuroendo- crine as well as neuromodulator function, their own secretory activity is influenced by multiple neurotransmitters and/or neuromodulators. Only recently is the pivotal importance of such relationhips becoming fully appreciated. Indeed, there are evidences to indicate that several neural circuits are involved in controlling the GnRH neuronal systems. Trans- mitters and modulators of importance include GABA (7- amino butyric acid), neuropeptide-Y (NPY), neurotensin, catecholamines, FMRFamide, endogenous opioid peptides (EOP), and others. All these circuits may provide impor- tant inputs to the GnRH system and some of them might contain excitatory as well as inhibitory input to GnRH system. It is also conceivable that one or the other compo- nent may, in turn, be influenced pre-synaptically by other neural circuits or factors, like sex steroids, cytokines etc. In mammals, GnRH-induced LH release in vivo is potentiated by NPY, either probably as a paracrine regulator (NPY neurons may synapse on GnRH neurons in the POA), or through enhancing the binding of GnRH to pituitary GnRH receptors [8, 70]; NPY secretion is enhanced by the preovula- tory surge of sex steroids, and in turn it stimulates GnRH secretion. GnRH enhances intracellular calcium levels through openining voltage-sensitive calcium channels and this post-receptor GnRH action is also potentiated by NPY [15]. A number of investigations has suggested that EOPs can act as powerful inhibitors of GtH release acting by primarily decreasing the amplitude of GnRH pulses, and are in turn influenced by gonadal steroids [5, 8]. EOP-containing neurons may communicate with GnRH neural circuitry by adjusting locally the influx of excitatory adrenergic signals along the hypothalamo-hypophyseal axis. There is evidence that corticotropin-releasing hormone (CRF) may deplete GnRH neuronal activity through central opioidergic path- ways, although direct effects are also likely to occur [2]. In the rat, dopamine can inhibit calcium ionophore-induced GnRH release [49]. In the goldfish, GnRH release from the POA is inhibited by dopamine and enhnaced by noradrena- line [73]. Noradrenaline, however, may be excitatory to GnRH in the presence of estrogens and inhibitory in their absence [33]. GnRH release and GnRH gene expression can be markedly inhibited by cytokines and the effects may be direct or mediated by opioids and prostaglandins [81]. GABAergic fibers are reported to directly innervate GnRH neuronal cells and until recently this provided morphological basis for a role of the GABAergic system in the regulation of GnRH secretion. Recently, however, Li and Pelletier [52] have shown that the use of GABA, receptor agonists inhibit not only the release of GnRH but also GnRH gene express- ion evaluated by ISHH. Simultaneous localization of multi- ple neuropeptides in the same brain section may yield useful information in relation to the modulation of GnRH neuronal activity on part of the complex neuronal circuitry in verte- brates. Neurons and fiber projections containing GnRH, FMRFamide and EOPs are interspersed in same brain areas and this may be the morphological substrate of physiological interactions between these systems [17, 38, 99, 100]. The coexistence of FMRFamide-like peptide and GnRH within the same neuron in a fish has provided morphological basis to suggest that FMRFamide may have an autocrine action on GnRH secretion [7]. The neural mechanisms involved in the stimulation or inhibition of GnRH release need further study. Secretory mechanism One exciting line of investigations is related to the study of neurosecretory mechanisms of GnRH neurons. Based on the conservation of ultrastructural features of differentiated neuroendocrine cells, the pulsatile manner of GnRH secre- tion and the presence of functional gap junctions in the GnRH neuronal cell line GT1-7 it has recently been sug- gested that gap junctional coupling between GnRH neurons may be the signalling machinery underlying periodic (pulsatile/circadian/seasonal) secretion of hypothalamic GnRH neuronal system [57]. GnRH-GnRH cell contacts, like synaptic organizations, have been described in mammals [58, 105], and are supposed to be involved in the synchronous activity of an entire population of GnRH neurons. Conse- quently, innervation of only a few GnRH neurons on part of another category of neuron may account for a cascade excitatory or inhibitory paracrine role of a neurotransmitter/ neuromodulator/neuropeptide. At the same time, GnRH- GnRH contact will suggest an autocrine regulation within a GnRH-neuronal population. Taken together, the current data would suggest that a full understanding of the anatomical substrates of GnRH/other neuropeptides interactions may provide the key to the correct interpretation of functional studies. Binding sites, receptors and mechanism of action High affinity binding sites for GnRH/GnRH, have been demonstrated in the pituitary and/or gonads of several verte- brate groups, as well as in the mammalian placenta, adrenal gland and mammary carcinoma [see 10, 11, 27, 44, 45]. GnRH receptor in the mammalian pituitary cells has been characterized and cloned [72]. It is composed of seven transmembrane segments, a feature of G-protein-coupled receptors. Studies are needed to understand the role of G-proteins in mediating the effects of GnRH as well as the amino acid sequence of the GnRH receptor of nonmamma- lian vertebrates. In studying the mechanism of action of GnRH on its targets, extracellular and intracellular Ca?*, protein kinase C, inositol phosphates, calmodulin, leukotrienes and arachi- donic acid have been implicated in the mediation of GnRH- induced GtH secretion [see 66]. GnRH acts upon the pituitary gonadotropes through specific high affinity mem- brane-bound receptors [14], by modifying the frequency and amplitude of action potentials generated spontaneously in gonadotropes [37]. In mammals, chronic GnRH/GnRH, treatment causes desensitization of the pituitary by down- GnRH in Vertebrates 369 regulating pituitary GnRH receptors and simultaneously by uncoupling GnRH signal transduction system and inactivat- ing voltage-sensitive Ca** channels [11, 36, 37, 64]. Among bony fishes, there are two lines of evidences, one supporting the participation of protein kinase C pathway in the media- tion of GnRH-stimulated GtH release, while the other sup- ports the involvement of cAMP pathway [see 41]. FINAL COMMENTS 1. The most logical inference that can be drawn from the rather bulky and somewhat bizzare repertoire of data is that a uniform methodological approach should be used throughout vertebrates. However, because of the existence of amazingly diverse situations related to reproduction, the morpho-functional charatecristics of the GnRH neuronal system of a vertebrate group can not necessarily be general- ized and extrapolated to other vertebrates. The discrepan- cies over the identification and distribution of GnRH-like material in the brain and other tissues may depend upon methodological variations, and despite the recognition of inherent difficulties, isolation and sequence analysis of GnRH-like material from more vertebrate species are needed. 2. A relevant amount of studies has provided the basis for additional investigations to fully resolve the questions of how GnRH regulates pituitary-gonadal activity, and repro- duction as whole, involving direct GnRH effects on these organs as well as factors (hormonal and not) which in turn influence GnRH system. In relation to the last point it is imperative to clarify whether the stimulatory/inhibitory effects are exerted both on GnRH secretion and biosynthesis, and if there exists a direct relationship between alteration in GnRH release and biosynthesis. Moreover, in ICC analyses of reproductive status-related changes, corresponding ultra- structural studies are required to define whether diminished immunoreactivity is due to decreased biosynthesis or en- hanced secretion and, vice versa, increased immunoreactivity is Owing to an increased biosynthesis or a decline in release. 3. Although there have been some investigations, the potential importance of GnRH on the ontogenesis and post- natal development of pituitary gonadotropes is far from understood. Nevertheless, it is clear that an analysis of the number and distribution pattern of GnRH neuronal subtypes may turn out to be useful in understanding the ontogenetic mechanisms and the potential (differential) role of regional subpopulations of GnRH neurons in the process of sexual maturation. It is also conceivable that embryonic migration of GnRH neurons into the forebrain from the nasal area continues in the postnatal/posthatching/pro and postmeta- morphosis period, and it is possible that precocious migration includes undifferentiated cells from the olfactory placode that are capable of division and differentiation at the time of sexual maturation/puberty. Future studies may involve more vertebrate species and, hopefully, be aimed to provide details not only about the extra- and intracranial migratory course of GnRH neurons and the developmental stage at which GnRH neurons reach their adult position, but also on the differential localization of GnRH neurons and processes showing immunoreactivity for different GnRH forms. 4. Morphological, distributional and molecular heter- ogeneity of GnRH render the variety of roles played by this neuropeptide plausible. There are GnRH receptor sybtypes in the pituitary which has made it possible to ascertain that C-terminal fragments of GnRH, or GAP, or GnRH, may have similar effects as GnRH itself. Much remains to be done before more concrete inferences may be drawn. 5. Besides electrophysiological methods to study the activity of individual neurons, it is argued that changes in mRNA levels reflect changes in neuronal activity. The availability of quantitative ISHH technique allows for measuring changes in mRNA levels in individual neurons. However, the current methods used to measure GnRH content and GnRH mRNA may not be sensitive enough to detect small changes in GnRH secretion. Hopefully, in the forthcoming years more sophisticated technologies will emerge which will allow us to reveal, record and interpret the electro-chemical and molecular signals released by individual GnRH neurons. There is a total dearth of information on the role of the genome in modulating GnRH neuronal activity. To date only one putative gene has been cloned in all species examined, and it is of comfort that ISHH unveils GnRH gene transcription and expression in neurons where GnRH itself is revealed by ICC. 6. It is hoped that latest studies will stimulate col- laborative research among physiologists, biochemists, molecular biologists, genetists, neurobiologists and compara- tive endocrinologists to unveil the yet “secret” world of GnRH. ACKNOWLEDGMENTS The authors’ personal research cited here has been supported by grants (40% and 60%) from M.U.R.S.T. and National Research Council (grant n. 93.00392.CT04) of Italy. A debt of gratitude is extended to Professor Joseph T. Bagnara for his ongoing support and advice. Thanks are also due to Dr Judy A. King, Professor V. Botte and Dr B. D’Aniello for useful suggestions. REFERENCES 1 Akutsu S, Takada M, Ohki-Hamazaki H, Murakami S, Arai Y (1992) Origin of luteinizing hormone-releasing hormone neurons in the chick embryo: effect of the olfactory placode ablation. Neurosci Lett 142: 241-244 2 Almeida OFX, Nikolarakis KE, Sirinathsinghji DJS, Hertz A (1989) Opioid-mediated inhibition of sexual behavior and luteinizing hormone secretion by corticotropin-releasing hor- mone. In “Brain Opioid Systems in Reproduction” Ed by RG Dyer, RJ Bicknell, Oxford University Press, Oxford, pp 149- 164 3 Amano M, Aida K, Okumoto N, Hasegawa Y (1992) Changes in salmon GnRH and chicken GnRH-II contents in the brain and pituitary, and GtH contents in the pituitary in female Masu salmon, Oncorhynchus masou, from hatching 370 10 11 12 13 19 R. K. RaAstoGi AND L. IELA through ovulation. Zool Sci 9: 375-386 Anderson AC, Tonon M-C, Pelletier G, Conlon JM, Fasolo A, Vaudry H (1992) Neuropeptides in the amphibian brain. Int Rev Cytol 138: 89-210 Angioni S, Nappi RE, Genazzani AD, Carunchio P, Criscuolo M, Guo A-L, Bidzinska B, Petraglia F (1993) Brain opioids and reproductive function. In “Cellular Communication in Reproduction” Ed by F Facchinetti, [W Henderson, R Pieran- toni, A Polzonetti-Magni, J Endocrinol Ltd, Bristol, pp 33-40 Barry J, Hoffman GE, Wray S (1985) LHRH-containing systems. In “Handbook of Chemical Neuroanatomy Vol 4” Ed by A Bjorklund, T Hokfelt, Elsevier, Amsterdam, pp 166- 215 Batten TFC, Cambre ML, Moons L, Vandesande F (1990) Comparative distribution of neuropeptide-immunoreactive sys- tems in the brain of the green molly, Poecilia latipinna. J Comp Neurol 302: 893-919 Bauer-Dantoin AC, Knox KL, Schwartz NB, Levine JE (1993) Estrous cycle stage-dependent effects of neuropeptide Y on luteinizing hormone (LH)-releasing hormone-stimulated LH and follicle-stimulating hormone secretion from anterior pituit- ary fragments in vitro. Endocrinology 133: 2413-2417 Blake CA, Gary TC, Mascagni F, Culler MD, Negro-Vilar A (1993) Effects of injection of anti-luteinizing hormone (LH)- releasing hormone serum and anti-gonadotropin-releasing hor- mone-associated peptide serum into neonatal rats on LH and follicle-stimulating hormone cells. Biol Reprod 49: 965-971 Chieffi G, Pierantoni R, Fasano S (1991) Immunoreactive GnRH in hypothalamic and extrahypothalamic areas. Int Rev Cytol 127: 1-55 Clayton RN (1989) Gonadotropin-releasing hormone: its ac- tions and receptors. J Endocrinol 120: 11-19 Clayton RN, Eccleston L, Gossard F, Thalbard J-C, Morel G (1992) Rat granulosa cell expression of the gonadotropin- releasing hormone gene: evidence from in situ hybridization histochemistry. J Mol Endocrinol 9: 189-195 Conlon JM, Collin F, Chiang YC, Sower SA, Vaudry H (1993) Two molecular forms of gonadotropin-releasing hormone from the brain of the frog, Rana ridibunda: purification, character- ization, and distribution. Endocrinology 132: 2117-2123 Conn PM, Huckle WR, Andrews WV, McArdle GA (1987) The molecular mechanisms of action of gonadotropin-releasing hormone (GnRH) in the pituitary. Rec Prog Horm Res 43: 29-61 Crowley WR, Shah GV, Carroll BL, Kennedy D, Dokher ME, Kalra SP (1990) Neuropeptide Y enhances luteinizing hor- mone (LH)-releasing hormone-induced LH release and eleva- tion in cytosolic Ca** in rat anterior pituitary cells: evidence for involvement of extracelular Ca** influx through voltage- sensitive channels. Endocrinology 127: 1487-1494 D’Aniello B (1994) Il GnRH nell’ontogenesi dei vertebrati: analisi immunoistochimica. Doctorate Thesis, University of Naples “Federico IT” D’Aniello B, Imperatore C, Fiorentino M, Vallarino M, Rastogi RK (1994) Immunocytochemical localization of POMC-derived peptides (adrenocorticotropic hormone, a- melanocyte stimulating hormone and /-endorphin) in the pituitary, brain and olfactory epithelium of the frog, Rana esculenta, during development. Cell Tissue Res, (in press) D’Aniello B, Masucci M, di Meglio M, Ciarcia G, Rastogi RK (1991) Distribution of gonadotropin-releasing hormone-like peptides in the brain during development of juvenile male Rana esculenta. Cell Tissue Res 265: 51-55 D’Aniello B, Masucci M, di Meglio M, lela L, Rastogi RK (1994) Immunohistochemical localization of GnRH in the 20 21 22 23 24 25 26 Pi 28 29 30 31 32 34 35 crested newt (Triturus carnifex) brain and terminal nerve: a developmental study. J Neuroendocrinol 6: 167-172 D’Antonio M, Fasano S, de Leeuw R, Pierantoni R (1992) Effects of gonadotropin-releasing hormone variants on plasma and testicular androgen levels in intact and hypophysectomized male frogs, Rana esculenta. J Exp Zool 261: 34-39 Dellovade TL, King JA, Millar RP (1993) Presence and differential distribution of distinct forms of immunoreactive gonadotropin-releasing hormone in the musk shrew brain. Neuroendocrinology 58: 166-177 Demski LS (1984) The evolution of neuroanatomical sub- strates of reproductive behavior: sex steroid and LHRH- specific pathways including the terminal nerve. Am Zool 24: 809-830 Demski LS (1987) Phylogeny of luteinizing hormone- releasing hormone system in protochordates and vertebrates. In “The Terminal Nerve (Nervus Terminalis) Structure, Func- tion, and Evolution Vol 519” Ed by LS Demski, M Schwanzel- Fukuda, Ann NY Acad Sci, pp 1-14. di Meglio M, Masucci M, D’Anillo B, Iela L, Rastogi RK (1991) Immunohistochemical localization of multiple forms of gonadotropin-releasing hormone in the brain of the adult frog. J Neuroendocrinol 3: 363-368 di Meglio M, Morrell JI, Pfaff DW (1987) Localization of steroid-concentrating cells in the central nervous system of the frog, Rana esculenta. Gen Comp Endocrinol 67: 149-154 Fahien CM, Sower SA (1990) Relationship between brain gonadotropin-releasing hormone and final reproductive period of the adult male sea lamprey, Petromyzon marinus. Gen Comp Endocrinol 80: 427-437 Fasano S, de Leeuw R, Pierantoni R, Chieffi G, van Oordt PGWJ (1990) Characterization of gonadotropin-releasing hormone (GnRH) binding sites in the pituitary and testis of the frog, Rana esculenta. Biochem Biophys Res Commun 168: 923-932 Fasano S, Goos HJTh, Janssen C, Pierantoni R (1993) Two GnRHs fluctuate in correlation with androgen levels in the male frog Rana esculenta. J Exp Zool 266: 277-283 Gautron JP, Patton E, Bauer K, Kordon C (1991) (Hydroxyproline’) luteinizing hormone-releasing hormone: a novel peptide in mammalian and frog hypothalamus. Neurochem Inter 18: 221-235 Gautron JP, Patton E, Leblanc P, L’HSitier A, Kordon C (1993) Preferential distribution of C-terminal fragments of (Hydroxyproline’)LHRH in the rat hippocampus and olfactory bulb. Neuroendocrinology 58: 240-250 Goos HJTh (1993) Pubertal development: big questions, small answers. In “Cellular Communication in Reproduction” Ed by F Facchinetti, IW Henderson, R Pierantoni, A Pol- zonetti-Magni, J Endocrinol Ltd, Bristol, pp 11-20 Goubau S, Bond CT, Adelman JP, Misra V, Haynes MF, Murohy BD (1992) Partial characterization of the gonadotro- pin-releasing hormone (GnRH) gene transcript in the rat ovary. Endocrinology 130: 3098-4000 Grossmann R, Dyer RG (1989) Opioid inhibition of brain- stem projections to the medial preoptic area in female rats. In “Brain Opioid Systems in Reproduction” Ed by RG Dyer, RJ Bicknell, Oxford University Press, Oxford, pp 112-124 Habibi HR (1991) Homologous desensitization of gonadotro- pin-releasing hormone (GnRH) receptors in the goldfish pituit- ary: effects of native GnRH peptides and a synthetic GnRH antagonist. Biol Reprod 44: 275-283 Habibi HR, Prati D (1993) Endocrine and paracrine control of ovarian function: role of compounds with GnRH-like activity in goldfish. In “Cellular Communication in Reproduction” Ed 36 37 38 39 40 41 42 43 45 46 47 48 49 50 51 GnRH in Vertebrates 371 by F Facchinetti, [W Hendersen, R Pierantoni, A Polzoneti- Magni, J Endocrinol Ltd, Bristol, pp 59-70 Hazum E, Conn PM (1988) Molecular mechanisms of gona- dotropin-releasing hormone (GnRH) action. I. The GnRH receptor. Endoc Rev 9: 379-386 Heyward PM, Chen C, Clarke IJ (1993) Gonadotropin- releasing hormone modifies action potential generation in sheep pars distalis gonadotropes. Neuroendocrinology 58: 646-654 Hoffman GE, Fitzsimmons MD, Watson Jr RE (1989) Rela- tionship of endogenous opioid peptide axons to GnRH neurones in the rat. In “Brain Opioid Systems in Reproduc- tion” Ed by RG Dyer, RJ Bicknell, Oxford University Press, Oxford, pp 125-134 Iela L, D’Aniello B, di Meglio M, Rastogi RK (1994) In- fluence of gonadectomy and steroid hormone-replacement therapy on the gonadotropin-releasing hormone neuronal sys- tem in the anterior preoptic area of the frog (Rana esculenta) brain. Gen Comp Endocrinol (in press) Iela L, Rastogi RK, Bagnara JT (1991) Reproduction in the Mexican leaf frog, Pachymedusa dacnicolor. V. Ymmunohis- tochemical localization of gonadotropin-releasing hormone in the brain. Gen Comp Endocrinol 84: 129-134 Jobin RM, Ginsberg J, Matowe WC, Chang JP (1993) Down- regulation of protein kinase C levels leads to inhibition of GnRH-stimulated gonadortopin secretion from dispersed pituitary cells of goldfish. Neuroendocrinology 58: 2-10 Karten MI, Rivier JE (1986) Gonadotropin-releasing hor- mone analog design. Structure function studies toward the development of agonists and antagonists. Rationale and pers- pective. Endoc Rev 7: 44-66 Katz IA, Millar RP, King JA (1990) Differential regional distribution and release of two forms of gonadotropin-releasing hormone in the chicken brain. Peptides 11: 443-450. King JA, Millar RP (1991) Gonadotropin-releasing hor- mones. In “Vertebrate Endocrinology: Fundamentals and Biomedical Applications VolIV” Ed by PKT Pang, MP Schreibman, Academic Press, New York, pp 1-31 King JA, Millar RP (1992) Evolution of gonadotropin- releasing hormones. Trends Endocrinol Metab 3: 339-346 King JA, Steneveld AA, Millar RP (1993) Differential dis- tribution of gonadotropin-releasing hormones in amphibian (clawed toad, Xenopus laevis) brain. Regulatory Peptides 50: 277-289 King JA, Steneveld AA, Millar RP, Fasano S, Romano G, Spagnuolo A, Zanetti L, Pierantoni R (1992) Gonadotropin- releasing hormone in elasmobranch (electric ray, Torpedo marmorata) brain and plasma: chromatographic and immuno- logical evidence for chicken GnRH II and novel molecular forms. Peptides 13: 27-35 King JC, Sower SA, Anthony ELP (1988) Neuronal systems immunoreactive with antiserum to lamprey gonadotropin- releasing hormone in the brain of Petromyzon marinus. Cell Tissue Res 253: 1-8 Koike K, Kadowaki K, Hirota K, Ohmichi M, Ikegami H, Sawada T, Miyake A, Tanizawa O (1993) Prolactin stimulates (7H) dopamine release from dispersed rat tubero-infundibular dopaminergic neurons and dopamine decreases gonadotropin- releasing hormone release induced by calcium ionophore. Acta Endocrinol 129: 548-553 Kuenzel WJ, Blahser S (1991) The distribution of gonadotro- pin-releasing hormone neurons and fibers throughout the chick brain (Gallus domesticus). Cell Tissue Res 264: 481-495 Leshin LS, Rund LA, Kraeling RR, Crim JW, Kiser TE (1992) Morphological differences among luteinizing hormone- 52 53 54 35) 56 57 58 59 60 61 62 63 64 65 66 67 releasing hormone neurons from postpartum and estrous cyclic cows. Neuroendocrinology 55: 380-389 Li S, Pelletier G (1993) Chronic administration of muscimol and pentobarbital decreases gonadotropin-releasing hormone mRNA levels in the male rat hypothalamus determined by quantitative in situ hybridization. Neuroendocrinology 58: 136-139 Licht P, Porter DA (1987) Role of gonadotropin-releasing hormone in regulation of gonadotropin secretion from amphi- bian and reptilian pituitaries. In “Hormones and Reproduc- tion in Fishes, Amphibians, and Reptiles”, Ed by DO Norris, RE Jones, Plenum, New York, pp 61-85 Livne I, Gibson MJ, Silverman A-J (1993) Biochemical dif- ferentiation and intercellular interactions of migratory gonadot- ropin-releasing hormone (GnRH) cells in the mouse. Dev Biol 159: 43-656 Loumaye E, Thorner J, Catt KJ (1982) Yeast mating pher- omone activates mammalian gonadotrophs: evoltionary con- servation of a reproductive hormone? Science 218: 1323-1325 Lovejoy DA, Fischer WH, Ngamvongchon S, Craig AG, Nahorniak CS, Peter RE, Rivier JE, Sherwood NM (1992) Distinct sequence of gonadotropin-releasing hormone (GnRH) in dogfish brain provides insight into GnRH evolution. Proc Natl Acad Sci USA 89: 6373-6377 Masucci M, D’Aniello B, Iela L, Ciarcia G, Rastogi RK (1992) Immunohistochemical demonstration of the presence and loca- lization of diverse molecular forms of gonadotropin-releasing hormone in the lizard (Podarcis s. sicula) brain. Gen Comp Endocrinol 86: 81-89 Matesic DF, Germak JA, Dupont E, Madhukar BV (1993) Immortalized hypothalamic luteinizing hormone-releasing hor- mone neurons express a connexin 26-like protein and display functional gap junction coupling assayed by fluorescence recov- ery after photobleaching. Neuroendocrinology 58: 485-492 Melrose PA, Littlefiedl-Chabaud MA, Pickel C, French DD (1993) Regional differences in the sunbcellular characteristics and synaptic connections with gonadotropin-releasing hormone (GnRH) neurons in the basal forebrain of adult ponies: sex- dependent differences found at the beginning of the anovula- tory season. Proc 13th Int Symp Equine Physiol Nutrition, Gainsville, Fl, pp 344-345 Millam JR, Craig-Veit CB, Petitte JN (1993) Brain content of cGnRH I and II during embryonic development in chickens. Gen Comp Endocrinol 92: 311-317 Millam JR, Faris PL, Yougren OM, El Halawani ME, Hartman BK (1993) Immunohistochemical localization of chicken gonadotropin-releasing hormones I and II (CGnRH I and II) in turkey brain. J Comp Neurol 333: 63-82. Morrell JI, Pfaff DW (1981) Characterization of estrogen- concentrating hypothalamic neurons by their axonal projec- tions. Science 217: 1273-1276 Murakami S, Kikuyama S, Arai Y (1992) The origin of the luteinizing hormone-releasing hormone (LHRH) neurons in newt (Cynops pyrrhogaster): the effect of olfactory placode ablation. Cell Tissue Res 269: 21-27 Muske LE (1993) Evolution of gonadotropin-releasing hor- mone (GnRH) neuronal systems. Brain Behav Evol 42: 215— 230 Muske LE, Moore FL (1990) Ontogeny of immunoreactive gonadotropin-releasing hormone neuronal system in amphib- ians. Brain Res 534: 177-187 Naor Z (1990) Signal transduction mechanisms of Ca** mobi- lizing hormones: the case of gonadotropin-releasing hormone. Endocr Rev 11: 326-353 Norgren RB, Lippert J, Lehman MN (1992) Luteinizing 372 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 R. K. Rastoci AND L. IELA hormone-releasing hormone in the pigeon terminal nerve and olfactory bulb. Neurosci Lett 135: 201-204 Nozaki M, Tsukahara T, Kobayashi H (1984) Neuronal sys- tems producing LHRH in vertebrates. In “Endocrine Corre- lates of Reproduction” Ed by K Ochiai, Y Arai, T Ohioda, M Takahashi, Springer, Berlin, pp 3-27 Oka Y, Matsushima T (1993) Gonadotropin-releasing hor- mone (GnRH)-immunoreactive terminal nerve cells have in- trinsic rhythmicity and project widely in the brain. J Neurosci 13: 2161-2176 Parker SL, Kalra SP, Crowley WR (1991) Neuropeptide Y modulates the binding of a gonadotropin-releasing hormone (GnRH) analog to anterior pituitary GnRH receptor sites. Endocrinology 128: 2309-2316 Pavgi S, Licht P (1993) Inhibition of in vitro pituitary gona- dotropin secretion by 17f-estradiol in the frog Rana pipiens. Gen Comp Endocrinol 89: 132-137 Perrin MH, Bilezikjian LM, Hoeger C, Donaldson CJ, Rivier J, Haas Y, Vale WW (1993) Molecular and functional charac- terization of GnRH receptors cloned from rat pituitary and mouse pituitary tumor cell line. Biochem Biophys Res Com- mun 191: 1139-1144 Peter RE, Habibi HR, Chang JP, Nahorniak CS, Yu KL, Huang YP, Marchant TA (1990) Actions of gonadotropin- releasing hormmone (GnRH) in goldfish. In “Progress in Clinical and Biological Researc. Progress in Comparative Endocrinology” Ed by A Epple, CG Scanes, MH Stetson, Wiley-Liss, New York, pp 393-398 Petersen SL, McCrone S, Shores S (1993) Localized changes in LHRH mRNA levels as cellular correlates of the positive feedback effects of estrogen on LHRH neurons. Am Zool 33: 255-265 Powell RC, Ciarcia G, Lance V, Millar RP, King JA (1986) Identification of diverse molecular forms of GnRH in reptile brain. Peptides 7: 1101-1108 Propper CR, Moore FL (1991) Effects of courtship on brain gonadotropin hormone-releasing hormone and plasma steroid concentrations in a female amphibian (Taricha granulosa). Gen Comp Endocrinol 81: 304-312 Rance NE, Young WS, McMullen NT (1994) Topography of neurons expressing luteinizing hormone-releasing hormone gene transcripts in the human hypothalamus and basal fore- brain. J Comp Neurol 339: 573-586 Rastogi RK, di Meglio M, Iela L (1990) Immunoreactive luteinizing hormone-releasing hormone in the frog (Rana esculenta) brain: distribution pattern in the adult, seasonal changes, castration effects, and developmental aspects. Gen Comp Endocrinol 78: 444-458 Rastogi RK, Iela L (1992) Spermatogenesis in amphibia: dynamics and regulation. In “Sex Origin and Evolution” Ed by R Dallai, Mucchi Editore, Modena, Italy, pp. 231-249 Rhim T-J, Kuehl D, Jackson GL (1993) Seasonal changes in the relationships between secretion of gonadotropin-releasing hormone, luteinizing hormone, and testosterone in the ram. Biol. Reprod. 48: 197-204 Rivest S, Rivier C (1993) Interleukin-1/ inhibits the endoge- nous expression of the early gene c-fos located within the nucleus of LHRH neurons and interferes with hypothalamus LHRH release during proestrous in the rat. Brain Res 613: 132-142 Ronnekleiv OK, Resko JA (1990) Ontogeny of gonadotro- pin-releasing hormone-containing neurons in early fetal de- velopment of rhesus macaques. Endocrinology 126: 498-511 Schally AV, Bajusz S, Redding TW, Zalatnai A, Comaru- Schally AM (1989) Analogs of LHRH: the present and the 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 future. In “GnRH Analogues in Cancer and in Human Re- production: Basic Aspects Vol1” Ed by BH Vickery, V Lunenfeld, Kluwer, Dordrecht, pp 5-31 Schwanzel-Fukuda M, Pfaff DW (1989) Origin of luteinizing hormone-releasing hormone neurons. Nature 338: 161-164 Sherwood NM (1986) Evolution of a neuropeptide family: gonadotropin-releasing hormone. Am Zool 26: 1041-1054 Sherwood NM, Lovejoy DA (1993) Gonadotropin-releasing hormone in cartilaginous fishes: structure, location, and trans- port. Env Biol Fish 38: 197-208 Silverman A-J (1988) The gonadotropin-releasing hormone (GnRH) neuronal systems: immunocytochemistry. In “The Physiology of Reproduction Vol 2” Ed by E Knobil, J Neil, Raven Press, New York, pp 1283-1304 Silverman A-J, Millar RP, King JA, Zhuang X, Silver R (1994) Mast cells with gonadotropin-releasing hormone-like im- munoreactivity in the brain of doves. Proc Natl Acad Sci USA 91 (in press) Stamper DL, Licht P (1993) Further studies on the influence of GnRH on the biosynthesis of gonadotropins in female frogs (Rana pipiens) Gen Comp Endocrinol 92: 104-112 Stamper DL, Licht P (1994) Influence of androgen on the GnRH-stimulated secretion and biosynthesis of gonadotropins in the pituitary of juvenile female bullfrog, Rana catesbeiana. Gen Comp Endocrinol 93: 93-102 Stoddart DM (1990) The Scented Ape: The Biology and Culture of Human Odour. Cambridge University Press, pp 79-119 Sullivan KA, Silverman A-J (1993) The ontogeny of gonado- tropin-releasing hormone neurons in the chick. Neuroendoc- rinology 58: 597-608 Tilbrook AJ, Galloway DB, Williams AH, Clarke IJ (1993) Treatment of young rams with an agonist of GnRH delays reproductive development. Horm Behav 27: 5-28 Tilbrook AJ, Johnson RJ, Eason PJ, Walsh DJ, Trigg TE, Clarke IJ (1992) Short-term reduction in egg production in laying hens treated with an agonist of GnRH. Br Poult Sc 33: 611-628 Trudau VL, Murthy CK, Habibi HR, Sloley BD, Peter RE (1993) Effects of sex steroid treatments on gonadotropin- releasing hormone-stimulated gonadotropin secretion from the goldfish pituitary. Biol Reprod 48: 300-307 Tsai PS, Licht P (1993) Differential distribution of chicken-I and chicken-II in the turtle brain. Peptides 14: 221-226 Urbanski HF, Doan A, Pierce M (1991) Immunocytochemi- cal investigation of luteinizing hormone-releasing hormone neurons in Syrian hamster maintained under long or short days. Biol Reprod 44: 687-692 Urbanski HF, Doan A, Pierce M, Fahrenbach H, Collins PM (1992) Maturation of the hypothalamo-pituitary-gonadal axis of the male Syrian hamster. Biol Reprod 46: 991-996 Vallarino M, Feuilloley M, D’Aniello B, Rastogi RK, Vaudry H (1994) Distribution of FMRFamide-like immunoreactivity in the brain of the lizard Podarcis sicula. Peptides (in press) Vallarino M, Salsotto-Cattaneo M, Vaudry H (1993) Im- munohistochemical localization of FMRFamide-like peptides in the brain of the frog, Rana esculenta. In “Cellular Com- munication in Reproduction” Ed by F Facchinetti, [W Hender- son, R Pierantoni, A Polzonetti-Magni, J Endocrinol Ltd, Bristol, pp 45-48 Van Gils J, Absil P, Grauwels L, Moons L, Vandesande F, Balthazart J (1994) Distribution of luteinizing hormone- releasing hormone (LHRH) (LHRH-I and -II) in the quail and chicken brain as demosntrated with antibodies directed against synthetic peptides. J Comp Neurol (in press) 102 103 104 105 106 107 108 109 GnRH in Vertebrates 373 Vickery BH (1986) Comparisons of the potential for ther- apeutic utilies with gonadotropin-releasing hormone agonists and antagonists. Endoc Rev 7: 115-124 Wirsig-Wiechmann CR (1993) Pripheral projections of nervus terminalis LHRH-containing neurons in the tiger salamander, Ambystoma tigrinum. Cell Tissue Res 273: 31-40 Woods JG, Honan MP, Thomas RC (1989) Hypothalamic regulation of the adenohypophyseal-testicular axis in the male chick embryo. Gen Comp Endocrinol 74: 167-172 Wray S, Grant P, Gainer H (1989) Evidence that cells expres- ing luteinizing hormone-releasing hormone mRNA in the mouse are derived from progenitor cells in the olfactory placode. Proc Natl Acad Sci USA 86: 8132-8136 Wright DE, Demski LS (1991) Gonadotropin hormone- releasing hormone immunoreactivity in the mesencephalon of sharks and rays. J Comp Neurol 307: 49-56 Wright DE, Demski LS (1993) Gonadotropin-releasing hor- mone (GnRH) pathways and reproductive control in elasmob- ranchs. Env Biol Fish 38: 209-218 Yellon SM, Newman SW (1991) A developmental study of the gonadotropin-releasing hormone neuronal system during sexual maturation in the male Djungarian hamster. Biol Reprod 45: 440-446 Yu KL, Sherwood NM, Peter RE (1988) Differential distribu- tion of two moelcular forms of gonadotropin-releasing hor- mone in discrete brain areas of goldfish (Carassius auratus). Peptides 9: 625-630 110 Zheng L-H, Pfaff DW, Scwanzek-Fukuda M (1992) Electron microscopic identification of luteinizing hormone-releasing hor- mone-immunoreactive neurons in the medial olfactory placode and basal forebrain of embryonic mice. Neuroscience 46: 407-418 Note added in proof In a redent study, Licht et al. [111] identified only cGnRH-II and mGnRH in the brain of three species of Rana (pipiens, ridibunda and esculenta). While both GnRHs were distributed in the telencephalon and diencephalon, only cGnRH-II was found in the cerebellum and medulla. In the platyfish, IGnRH, together with mGnRH and sGnRH was revealed by ICC in some cells of the pituitary gland of animals of all ages [112]. 111 112 Licht P, Tsai P-S, Sotowska-Brochocka J (1994) The nature and distribution of gonadotropin-releasing hormones in brains and plasma of ranid frogs. Gen Comp Endocrinol 94: 186-198 Magliulo-Cepriano L, Schreibman MP, Blum V (1994) Dis- tribution of variant forms of immunoreactive gonadotropin- releasing hormone and f-gonadotropins I and II in the platyfish, Xiphophorus maculatus, from birth to sexual matur- ity. Gen Comp Endocrinol 94: 135-150. ZOOLOGICAL SCIENCE 11: 375-379 (1994) Effects of Rapid Cooling on Heart Rate of the Japanese Lobster in vivo MotTonori NAKAMURA, MASAKI TANI and TAKETERU KURAMOTO Shimoda Marine Research Center, Laboratory of Physiology University of Tsukuba, Shimoda, Shizuoka 415, Japan ABSTRACT — Seasonal changes of the effects of rapid cooling on the in vivo heart rate of the lobster, Panulirus japonicus, were studied. Cardiac activity was monitored with chronically implanted electrodes. The mean heart rate was 45.5+6.0 bts/min during winter and 99.5+7.5 bts/min during the other seasons at the same acclimation temperature (20+1°C). When the ambient temperature was lowered (0.1-1.0°C/min, dT=5-6'C), the heart rate decreased along a linear line in spring, summer and fall. The correlation ratio of the heart rate and temperature was 0.98 with the minimum rate of 67.0 +12.0 bts/min. The Qo value was 2.2. Incontrast, the heart rate in winter decreased only for the initial few minutes of cooling and little for the later phase with the minimum rate of 28.8+9.5 bts/min. Qo was2.6. The correlation plots for the heart rate and temperature appeared to regress on two linear lines. The ratio was 0.88 in the range of 19-21°C and 0.41 in the range of 15-19°C. These data may suggest that the lobsters in winter have some compensatory mechanisms for a heart rate drop by cooling as well as for the mean heart rate, which are different from those in the other seasons. © 1994 Zoological Society of Japan implanted electrodes. The recording methods were similar to those INTRODUCTION The heart rate of crustaceans changes with variations in ambient temperature within the normal environmental range. Temperature coefficient (Q10) for the range of 10-20°C of the crustacean heart rate is around 2, measured in vivo for slow temperature changes ([24] for review, [10]). Even though marine lobsters live under mild temperature environment, they meet daily with a warm or cold water brought about by tidal cycle. Moreover, they suffer seasonal changes. Our preliminary experiments show that temperature of the lobster pericardium will follow the change in ambient temperature within 30 sec. Electromechanical coupling of muscle fibers becomes less efficient with decreasing temperature [8]. A compensation mechanism has been described for leg muscle activity of crabs [9]. It may be that the lobsters are equipped with compensation mechanisms in the heart activity for a rapid drop in temperature. Few studies on the heart re- sponse to rapid temperature changes in crustaceans have been reported [28]. However, the presence of compensa- tion in heart rate has been still unclear. In the present study, the influence of rapid cooling on heart rate in vivo is examined over all seasons using Japanese spiny lobsters. The behavior of heart rate in winter was different from that in the other seasons. MATERIALS AND METHODS Japanese spiny lobsters (Panulirus japonicus, around 200g in body weight, both sexes, n=8) were used. They were captured around Nabeta Bay (Izu Peninsula, Japan) in April and October and were reared in an indoor aquarium (1.54 1m*) where the fresh natural sea water was continuously supplied. Natural light (around 20k Ix at noon) was provided via windows. Electrocardiogram (ECG) was recorded from the unrestrained animals with chronically Accepted April 22, 1994 Received February 15, 1994 described previously [13, 14]. The electrodes consisted of silver wires, which were introduced through small holes drilled in the carapace of the dorsal thorax and fixed in position with epoxy resin. Then each lobster with the electrodes mounted was kept in a tank (25 x 40x30 cm?) for a month before being subjected to cooling experi- ment. They were fed on a diet of live Mytilus. Coral sands were laid on the bottom of the tank as suggested by Florey and Kriebel [10]. The sea water in the tank was aerated and stirred with air and water pumps, respectively. The acclimation temperature was 20+ 1°C because seasonal changes of the sea water temperature in the aquarium have ranged from 15 to 25°C. To cool the animal, the sea water cooled at 5°C was poured into the tank and mixed by stirring. Water temperature was monitored with an electronic thermometer. The cooling rates (0.1-1.3°C/min) and magnitudes (dT=5-6°C) were determined by pouring rate and volumes of the cold sea water. During experiment, major part of the tank was covered with black screen to minimize disturbance. Electrical signals amplified were recorded on a video tape (DC-2K Hz). These measurements were performed once aweek. The electrical recordings were played back to a pen recorder (100 Hz). The heart rate was counted by hand and/or with an electronic counter using pulses of ECG. Statistical analysis of the data was performed with a personal computer software (Stat View II: Abacus Concepts, Inc.). RESULTS Beating patterns of the heart became constant at 2 weeks after the lobster was transferred into the experimental tank. Thus the acclimation period of a month was long enough for the lobster to adapt to the experimental conditions. Large variations of heart rate accompanying with the body move- ment and spontaneous bradycardia often occurred. Howev- er, it was possible to measure the rate in a stable state since the animal often kept quiescent for 10 or more minutes in daytime. Complete recordings of ECG in vivo were obtained frequently during cooling. However, only few recordings were reliable for both cooling and rewarming processes because the unrestrained lobsters moved vigorously aoe (eo) (es) @ HEART RATE (bts/min) w [o0) (oe) | oO TIME (min) Fic. 1. Effects of cooling on the in vivo M. NaKAmurA, M. TANI AND T. KuRAMOTO 15°C 50 bts/min A TEMPERATURE (‘C) heart rate of the lobster in summer. A, a simultaneous recording of electrocardiogram (ECG, upper), ambient temperature (T, middle) and heart rate (HR, lower) displayed with fast and slow speeds. B, a graphic presentation of the heart rate change with decreasing tempera- ture. Means of ten beats are successively plotted with the temperature. The data are from the same videotape recording shown in A. during rewarming. Cooling rates ranged usually from 0.1 to 1.0°C/min and in few instances, to 1.3°C/min. Table 1 shows distributions of the heart rate measured from the 8 animals at 20+1°C and during cooling. At the acclimation temperature, the heart rate was 99.5+7.5 bts/min in the period from spring to fall while it was 45.5+6.0 bts/min in winter (December to February). In spring, summer and fall, the heart rate decreased with the lowering of the ambient temperature (dT=5-6C). Sig- nificant decreases in heart rate were observed by the lowering of temperature of more than 3°C and the minimum rate was 67.0+12.0 bts/min (Table 1). The heart rate recovered with rewarming. Moreover, a little higher heart rates (105- 110 bts/min) than before cooling were often observed for a few minutes after the water temperature was resumed to 20+ 1°C (not shown in figure). Figure 1A shows an example of the ECG recordings obtained in June, where the ambient temperature was lowered from 21 to 15°C at the rate of 0.7°C/min. The mean values of ten heartbeats were plotted with the change in temperature in Figure 1B. The heart rate decreased with a rate of 4.5 bts/min, and a decrement of 6.4 bts/°C. Qjo for the range of 15-20°C of the heart rate was 2.2. These parameters did not change remarkably by the cooling rates within a range of 0.1-1.3°C/min (not shown in figure). In winter, the heart rate decreased for the initial few minutes of cooling but did not for the later phase. Signi- ficant decreases in heart rate were observed by the lowering of temperature of more than 4°C and the minimum rate was 28.0+9.5 bts/min (Table 1). One of the ECG recordings obtained in February is shown in Figure2A. The mean heart rate of each ten beats was plotted with the temperature in Figure 2B. In this case, the ambient temperature de- creased from 21 to 16°C at the rate of 1.0°C/min but the decremental change of the heart rate against the temperature was small (3.2 bts/"C). Moreover, the heart rate did not TABLE 1. Effects of cooling on heart rate of P. japonicus Degree of cooling Heart rate in seasons from spring to fall Heart rate in winter dT C) Mean+SE bts/min) (Mean+SE bts/min) 0.0 WNae 7/5 45.5+ 6.0 0.5 96.0+ 12.0 41.0+ 5.0 1.0 97.0+ 7.0 40.0+ 5.0 125) 92.0+ 7.5 38.5+ 9.0 2.0 89.0+ 9.0 29.0+ 5.0 2.5 84.0+11.5 31.0+ 9.0 3.0 81.0+ 10.0 33.0+ 10.5 3h) 78.0+ 9.0 29.0+ 10.0 4.0 76.0+ 8.5 29.0+ 8.5 4.5 73.0+ 8.0 30.04 11.5 5.0 67.0+ 12.0 28.0+ 9.5 Acclimation temperature was 20+1°C. Ambient temperature of the lobster was lowered to 16 or 15°C (dT= to May. Summer: June to August. Fall: September to November. Winter: The cooling experiment was repeated four times per month from March, 1991 to 5°C). Spring: March December to February. February, 1993. Cooling Effect on Lobster Heart Rate 377 A ECG | aoe cr a T 21°C 16°C HR 50 a ee | Mian aioe ear ft mine le Ah bts/min 1s. 1 min 1s. B ES 50 pin & 2 2) = Ba 5 a oo (ae S on z mh Ee 25 S ‘ > | ) 10 “a! 16 0 5 TIME (min) Fic. 2. Effects of cooling on the in vivo heart rate of the lobster in winter (14-2-1992). A, a simultaneous recording of ECG, ambient temperature (T) and heart rate (HR) displayed with fast and slow speeds. B, a graphic presentation of the heart rate change with decreasing temperature. Means of ten beats are successively plotted with the temperature. The data are from the same videotape recording shown in A. always show a linear change with the decreasing temperature from 20 to 16°C. With rewarming, the heart rate returned and reached to a level of 5—10 bts/min higher than the initial rate before cooling (not shown in figure). Correlation between the heart rate and temperature was analyzed statistically using the datashownin Table 1. Inthe period from spring to fall, the mean heart rates (68—99.5 bts/ min) regressed on a positive linear line (Y =6.43X-27.66, r= 0.99) within the temperature range of 15—20°C (Fig. 3a). In contrast, the mean heart rates in winter showed a lower value (28—45.5 bts/min) and regressed on another positive line (Y =3.26X-26.38, r=0.88) against the experimental tempera- ture decrease (Fig. 3b). However, some plots digressed at around 19°C; the plots appeared to regress on two lines (Fig. 4A and B). The correlation ratio was 0.94 in the partial range of 19-21°C while it was 0.41 in the range of 15-19°C. 105 100 95 90 85 80 75 70 HEART RATE (bts/min) Sm e416 17 1 10 ao a TEMPERATURE (°C) Fic. 3. Relationships between the heart rate and lowered tempera- ture in the Japanese lobsters. The plots a, correlation between the mean heart rate and temperature measured during the period from spring to fall (Y =6.43X-27.66, r=0.99). The plots b, correlation between the mean heart rate and temperature measured during winter (Y =3.26X-26.38, r=0.88). A linear line and paired curves indicate the regression and 95% confide- nce bands for the true mean of the heart rate, respectively. The all data shown in Table 1 are used. > Y = 4.40X - 47.85, 1 = 0.94 46 45 44 43 42 At 40 39 38 HEART RATE (bts/min) 20 24 TEMPERATURE (‘C) Y = 0.64X - 18.61, r= 0.41 34 33 32 31 30 29 28 HEART RATE (bts/min) fee 16 165 17 175 18 185 19 195 TEMPERATURE (C) Fic. 4. The heart rate drop with cooling of the winter lobsters. A, the correlation between the mean heart rate and temperature in the range of 19.5-21°C. B, the correlation between the mean heart rate and temperature in the range of 15.5-19.5°C. A straight line and paired curves indicate the regression and 95% confidence bands for the true mean of the heart rate, respective- ly. The same data as shown in Fig. 3b are used. 378 M. NAKAmuRA, M. TANI AND T. KurAMoTO DISCUSSION The mean heart rate (45.5 bts/min) in winter was as low as half of that (99.5 bts/min) in the other seasons at the same acclimation temperature (Table 1). In P. japonicus, the heart rate around 60 bts/min is unstable both in vivo and in vitro [22]. In both the shrimp and the crayfish, the mean heart rate is lowered when reared under dark conditions [11, 23]. Since the lobsters used had been under natural light, they might perceive the decreases in both light dose and period toward winter. Therefore, the conditions of illu- mination may cause to lower the heart rate to 45.5 bts/min in winter lobsters. However, this idea is not acceptable at present because the experiments to prove it will be necessary. The lobster heart in the pericardium is suspended by the ligaments and the arteries. The ligaments are associated with the pericardial alary muscles while the arteries with the pericardial septum, the circumference of which contains muscle fibers ({24] for review, [1]). These pericardial mus- cles are innervated by the segmental nerves coming from the thoracic ganglia [1]. Thus, the central nervous system (CNS) probably controls the heart rate via the suspensory elements mechanically. The importance of the suspensory elements for heart pumping activity has been noticed by several authors ({24, 25] for reviews, [13]). Meanwhile, we have observed that the beat rate of the isolated lobster heart is determined by pulling of the suspensory elements as well as filling pressure ([{15], unpublished data). Therefore, the seasonal difference in heart rate levels in vivo might be determined by the degree of tension of the suspensory elements controlled by the pericardial muscles and the CNS. The lobster heart in vivo is also controlled centrally via the dorsal nerves and the pericardial neurohemal organs [1- 3, 26,27]. The nervous control might contribute to the rapid change in heart rate observed frequently as fluctuations ([13, 14], Figs.1A and 2A). The pericardial organs of lobsters contain monoamines (serotonin, octopamine and dopamine) and peptides (proctolin and FMRFamide-related peptides) ({6] for review, [12, 29, 30]). The pericardial hormones slowly augment the beat of isolated lobster heart for minute- long periods [5, 17, 18]. By internal cooling of the body, the pericardial organs of P. japonicus are activated and probably release the pericardial hormones [21]. Even under the low temperature, octopamine markedly enhanced the beat rate and amplitude of the isolated heart [20]. Therefore, octopa- mine released from the pericardial organs could activate the in vivo heart under the low temperature. The pericardial peptides also may activate the in vivo heart [31, 32] but their actions under low temperature have not yet been proved. The cooling rate of 1°C/min is faster than that applied to decapod crustaceans by others [7, 10, 28]. Despite of the rapid cooling, the Qo value for the lobster heart rate was around 2, similar to that reported in the crab [10]. Thus we failed to find the effect of rapid cooling on the Qjo value. When the water temperature was resumed, however, the heart rate was often higher than that before cooling. This may suggest that the heart activation occurs during cooling and still continues during rewarming. The Qo between 10 and 25°C in the crustacean heart rate has been constant with an average value of about 2 while the Qo of 3-5 has been often observed between 5 and 10°C ([24] for review). This indicates that the chemical pacemaking processes in the heart are enhanced under the cold condi- tions. The Qo for the range of 15—-20°C of the lobster heart rate was 2.6 in winter while it was 2.2 in summer. Moreoy- er, the correlation between the heart rate and the lowered temperature was weaker in winter than in the other seasons (Fig. 3). Even if the starting heart rate was different, the decrease in heart rate against the same drop in temperature is much slow in winter (Fig. 4). It is therefore suggested that the heart activation for the cooling may occur markedly in winter. The lobster cardiac muscle develops tension by excita- tory junction potentials generated by motor neurons in the cardiac ganglion [4, 15, 19]. Tension produced in the myocardium is fed back to the cardiac ganglion since the cardiac neurons are sensitive for filling pressure ({24] for review, [15-18]). Thus, the responses of the heart to cool- ing should be explained by activities of the cardiac neurons and muscle fibers. The studies using the isolated hearts and their nerve-muscle preparations are in progress. ACKNOWLEDGMENTS The authors thank the staff of Shimoda Marine Research Center for supporting the present study. Contribution No. 569 from Shi- moda Marine Research Center. REFERENCES 1 Alexandrowicz JS (1932) The innervation of the heart of the crustacea. I. Decapoda. Q J Microsc Sci 75:181—249 2 Alexandrowicz JS (1953) Nervous organs in the pericardial cavity of the decapod Crustacea. J Mar Biol Ass UK 31: 563- 580 3 Alexandrowicz JS, Carlisle DB (1953) Some experiments on the function of the pericardial organs in Crustacea. J Mar Biol Ass UK 32: 175-192 4 Anderson M, Cooke IM (1971) Neural activation of the heart of the lobster Homarus americanus. J Exp Biol 55: 449-468 5 Beltz BS, Kravitz EA (1986) Aminergic and peptidergic neuro- modulation in crustacea. J Exp Biol 124: 115-141 6 Cooke IM, Sullivan RE (1982) Hormones and neurosecretion. In “THE BIOLOGY OF CRUSTACEA Vol. 3” Ed by HL Atwood, DC Sandeman, Academic Press, New York, pp 205- 391 7 Dauscher H, Flindt R (1969) Vergleichende Untersuchungen zur Herztaigkeit bei freibeweglichen dekapoden Krebsen (Asta- cus fluviatilis Fab., Astacus leptodactylus Escholz und Cambarus affinis Say.). Z vergl 62: 291-300 8 Dudel J, Ruedel R (1968) Temperature dependency of elec- tromechanical coupling in crayfish muscle fibres. Pfltier Arch 301: 16-30 9 Fisher L, Florey E (1981) Temperature effects on neuromuscu- lar transmission (opener muscle of crayfish, Astacus leptodacty- lus). J Exp Biol 94: 251-268 10 11 12 14 15 16 ity) 18 19 20 21 Cooling Effect on Lobster Heart Rate 379 Florey E, Kriebel ME (1974) The effects of temperature, anoxia and sensory stimulation on the heart rate of unrestrained crabs. Comp Biochem Physiol 48A: 285-300 Hara J (1952) On the hormones regulating the frequency of the heart beat in the shrimp, Paratya compressa. Annot Zool Jap 25: 162-171 Kobierski LA, Beltz BS, Trimmer BA, Kravitz EA (1987) FMRFamidelike peptides of Homarus americanus: distribution, immunocytochemical mapping, and ultrastructural localization in terminal varicosities. J Comp Neurol 266: 1-15 Kuramoto T (1990) Cardiac activity and pressure change in the lateral pericardium of the unrestrained lobster, Panulirus japont- cus. Physiol Zool 63: 182-190 Kuramoto T (1993) Cardiac activation and inhibition involved in molting behavior of a spiny lobster. Experientia 49: 682-685 Kuramoto T, Ebara A (1984) Effects of perfusion pressure on the isolated heart of the lobster, Panulirus japonicus. J Exp Biol 109: 121-140 Kuramoto T, Ebara A (1985) Effects of perfusion pressure on the bursting neurones in the intact or segmented cardiac gang- lion of the lobster, Panulirus japonicus. J Neurosci Res 13: 569-580 Kuramoto T, Ebara A (1988) Combined effects of 5- hydroxytryptamine and filling pressure on the isolated heart of the lobster, Panulirus japonicus. J Comp Physiol B 158: 403- 412 Kuramoto T, Ebara A (1991) Combined effects of octopamine and filling pressure on the isolated heart of the lobster, Panulirus japonicus. J Comp Physiol B 161: 339-347 Kuramoto T, Kuwasawa K (1980) Ganglionic activation of the myocardium of the lobster, Panulirus japonicus. J Comp Physiol 139: 67-76 Kuramoto T, Nakamura M (1992) Effects of cooling on the heart beat of the Japanese lobster in vitro. Zool Sci 9: 1217 Kuramoto T, Tani M (1994) Cooling-induced activation of the pericardial organs of the spiny lobster, Panulirus japonicus. 22 23 24 25 26 27 28 29 30 31 32 Biol Bull Mar Biol Lab Woods Hole 186: 319-327 Kuramoto T, Yamagishi H (1990) Physiological anatomy, burst formation, and burst frequency of the cardiac ganglion of crustaceans. Physiol. Zool, 63: 102-116 Larimer JL, Tindel JR (1966) Sensory modifications of heart rate in crayfish. Anim Behav 14: 239-245 Maynard DM (1960) Circulation and heart function. In “The Physiology of CRUSTACEA Vol. 1” Ed by TH Waterman, Academic Press, New York, pp 16l—214 McMahon BR, Wilkens JL (1983) Ventilation, perfusion, and oxygen uptake. In “THE BIOLOGY OF CRUSTACEA Vol 5” Ed by LH Mantel, Academic Press, New York, pp 290-372 Shimahara T (1969a) The inhibitory synaptic potential in the cardiac ganglion cell of the lobster, Panulirus japonicus. Sci Rpt Tokyo Kyoiku Daigaku B14: 9-26 Shimahara T (1969a) The effect of the acceleratory nerve on the electrical activity of the lobster cardiac ganglion. Zool Mag 78: 351-355 Spaargaren DH, Achituv Y (1977) On the heart rate response to rapid temperature changes in various marine and brackish water crustaceans. Neth J Sea Res 11: 107-117 Sullivan RE, Friend BJ, Barker DL (1977) Structure and function of spiny lobster ligamental nerve plexuses: evidence for synthesis, storage and secretion of biogenic amines. J Neuro- biol 8: 581-605 Timmer BA, Kobierski LA, Kravitz EA (1987) Purification and characterization of FMRFamidelike immunoreactive subst- ances from the lobster nervous system: isolation and sequence analysis of two closely related peptides. J Comp Neurol 266: 16-26 Wilkens JL, McMahon BR (1992) Intrinsic properties and extrinsic neurohormonal control of crab cardiac hemodynamics. Experientia 48: 827-833 Yazawa T, Kuwasawa K (1992) Intrinsic and extrinsic neural and neurohumoral control of the decapod heart. Experientia 48: 834-840 ZOOLOGICAL SCIENCE 11: 381-384 (1994) Effectiveness of Metoclopramide, Domperidone and Ondansetron as Anti-emetics in the Amphibian, Xenopus laevis Tomio NairoH, Moroko MATuUURA and RICHARD J. WASSERSUG* Department of Biology, Shimane University, Matsue 690, Japan ABSTRACT—We examined the effectiveness of three anti-emetic agents—metoclopramide, domperidone (both dopamine antagonists) and ondansetron (a SHT; receptor antagonist)—in the frog Xenopus laevis. Apomorphine and cisplatin were used to induce emesis. All three anti-emetics significantly retarded emesis when induced by apomorphine. The drugs, however, were not effective against cisplatin induced emesis in the limited dosage range that we examined. Paradoxically, both metoclopramide and domperidone themselves induced vomiting in some frog specimens at dosages where they retarded apomorphine-induced emesis in others. Xenopus laevis appears to be particularly sensitive to a variety of emetic challenges. Our results suggest that neural mechanisms involved in the control of emesis in amphibians and mammals are similar, although there are differences in the sensitivity of frogs and mammals to emetic and anti-emetic agents. Since Xenopus is an easy frog to maintain and breed in captivity, it may be valuable as an alternative model to mammals in the pharmaceutical search for effective anti-emetic © 1994 Zoological Society of Japan agents. INTRODUCTION We report here on the anti-emetic properties of three drugs in the African clawed frog, Xenopus laevis. Two are well established, anti-emetic, dopamine antagonists: metoc- lopramide and domperidone. The third, ondansetron, is a newer 5-HT3 receptor antagonist, also an effective anti- emetic in man and other mammals (see “Proceedings of the Ondansetron Symposium” In European Journal of Cancer & Clinical Oncology 25 (Suppl. 1), 1989. Pergamon Press, Oxford). There are two reasons for specifically exploring their effects in a lower vertebrate. The first is to understand the evolution and development of emesis in vertebrates in gener- al. Some animals (e.g. dogs, cats, certain primates) have a well developed emetic response whereas others, such as rats, are incapable of vomiting [1], even when exposed to a strong emetic challenge. Similarly, in some organisms such as anurans (frogs and toads), the ability to vomit is linked to developmental stage [5]. The neural bases for these taxono- mic and developmental differences are not known. We reasoned that similarity in the response to both emetic and anti-emetic agents among diverse vertebrates would provide indirect evidence of similar underlying neurochemical mechanisms. There is a second, more applied reason for undertaking this study. If it can be shown that the neural mechanisms controlling emesis are conserved in vertebrate evolution, then lower vertebrates may serve as a convenient alternative to Accepted May 10, 1994 Received February 15, 1994 * Present address: Department of Anatomy and Neurobiology, Sir Charles Tupper Medical Building, Dalhousie University, Halifax, Nova Scotia, B3H 4H7, Canada carnivorous mammals or primates in pharmaceutical research directed at improving anti-emetic therapies. We have used the African clawed frog Xenopus laevis for our investigation because it is easy to breed and maintain in captivity. It is the most common amphibian used in research around the world. MATERIALS AND METHODS Xenopus laevis adults were initially purchased from a commer- cial dealer. They were maintained in the laboratory on fish food pellets, then bred. The animals used in our experiments were either from the parental stock or from subsequent generations. All of the frogs used were either adults or young that had fed freely and exhibited normal growth since metamorphosis. Both males and females were used. The frogs ranged in wet mass from 3.9—-12.7 g as juveniles, 39.7-103.0 g as adult females, and 29.0-58.8 g as adult males (plus two sub-adult males of 18.9 and 25.1 g); however, only animals of similar size were used in any one experiment. An hour or less before each experiment, the frogs were fed pieces of beef liver that weighed approximately 1.0-2.6% of the frog’s wet mass. Ejection of this bolus of food was used to indicate emesis. No frogs were used more than once. Nausea was promoted by either apomorphine hydrochloride (Sigma, St. Louis) or cisplatin (Bristol-Myers Squibb, Tokyo). Both are well known emetics. The former stimulates dopamine receptors and its effectiveness in mammals is inhibited by dopamine antagonists, like domperidone [7]. The latter is commonly used in chemotherapy for cancer patients. The apomorphine was dissolved in 0.65% NaCl and injected into the dorsal lymph sac at a dose of 20ug per g wet body weight, following methods outlined in Naitoh et al. [4]. The cisplatin, made up for injection, was administered by similar injection at dosages of 0.5 to 100ug per g body weight. The absolute volume of any drug injection did not exceed 200y1/g. In all cases, where the frogs were subjected to these emetic agents, concurrent controls were injected with 0.65% NaCl, distilled water, or lactic acid solution (as appropri- 382 ate) and similarly observed (Fig. 1). Metoclopramide monohydrochloride (Sigma, St. Louis), dom- peridone (Sigma, St. Louis) and ondansetron, in the form of ondan- setron hydrochloride dihydrate (Glaxo Group Research Ltd., Green- ford) were compared for their anti-emetic properties. Metoclopra- mide monohydrochloride dissolved in 0.65% NaCl, which had pre- viously been shown to delay apomorphine-induced emesis in X. laevis [4], was administered at dosages of 10, 30, and 60ug per g body The domperidone was administered at dosages of 1.0, 10 and 20g per g body weight. Because domperidone solubility is pH sensitive, it was dissolved in 2.5% lactic acid solution. This was further diluted to equal 10ug/1 and the final dosages (as shown in Fig. 1) were made from that solution by additional dilution with The ondansetron, which was also dissolved in weight. distilled water. -180 -90 -30 0 60 f--*>- +! 0.65 PojNacl Apo. 20 a a See eS SS 1 T. NatroH, M. MATuuRA AND R. J. WASSERSUG 0.65% NaCl, was administered at dosages of 0.01, 0.1 and 1.0 per g body weight. The specific dosages used in each experiment are noted in the figures and Results. Anti-emetic agents were adminis- tered by injection into the dorsal lymph sac anywhere from 3 hr to 30 min before or after the emetic administration. The order and times of all drug administrations are similarly noted in Figures 1-2 and the Results. Some frogs were injected with ondansetron without apomor- phine or cisplatin to determine whether this putative anti-emetic agent had any intrinsic emetic effects in and of itself. For the first 4 hr after apomorphine administration and 10 hr after cisplatin administration, the frogs were observed continuously After that they were observed at least once an If the frogs had not vomited by for ejected vomitus. hour for an additional two hours. 120 180 fsesossssoS= @0@0- @0 -@ ------------------------— ------- ------~ ----~ ----~ n=8 Metoc. 10 Apo. 20 aera bo gg 25 ites) es eeeti sss: ee ee ee ee eee n=5 * Metoc. 30 Apo. 20 Ge IES ig Ihe. ~@ ----@- ----@@ ------------------ ------- --- (Ss 22555 222-5 n= 10 %* Metoc.60 Apo. 20 ven Oe eee @--@ ----------e-- ------~ ------- --9-— ----— =10 0.5 % Lactic acid Apo. 20 02S) ee Sees + 906 --005-=65.---— a2 Sees Se isee a ree he ieee n=7 DW Apo. 20 Wigih ung Ses 2eeeL ee e--- 6 @--- Cpe eee Cae eee eee see a eae n=7 Domp. 10 Apo. 20 eee eee + -=-@--@ 00 © ---==~== === = St ee ees ee eee iy 1212) anew Domp. 10 Apo. 20 bens ee a aes See see eeneses __——_ | _ - ee eee n=7 Domp. 20 Apo. 20 CED S + ge ----- en Pt sas ese ee ee eee n=7 Domp. 20 Apo. 20 = Sf} --@ --@------ se @ve - - - ----------------- @-- @----- ------- ----~ ----- n= 10 x Domp. 20 Apo. 20 45-99 4\@@---------- Wide erereescsore @e----------------------. __- @-- ----.-- 9. -9 = a 0.65% NaC! Apo. 20 cssoase=4 Bape 6: oe--en eee on ee ee Sree tS Sces=) ===: = Ste Ond. 0.01 Apo. 20 We pacer Sees ~@-@e--@---------~--~------------. ------~ ------~ ----~ ----- n=5 Ond.0.1 Apo. 20 octsseoss + -- 9.0 0m ~ 002 =~ = =< = = 55 = See ty os ed eee ee ee * Ond. 1.0 Apo. 20 esecsesee eons @---@---e-e------ @---@----------- ------~ ------- -.--- ~ = Ond. 1.0 beens nee eee ene nent ee ne nee Le Te Ugggo NaF Fic. 1. Apomorphine-induced emesis in the frog Xenopus laevis and its inhibition by metoclopramide monohydrochloride (top), domperidone (middle) and ondansetron hydrochloride dihydrate (bottom). “Apo. 20” indicates administration of 204g/g apomorphine hydrochloride. “DW” is distilled water. Each circle equals one frog; a solid circle indicates that an individual vomited at that time, open circles indicate individuals that did not vomit. Asterisks indicate statistically significant inhibition (P<0.05) of emesis in frogs which received both an anti-emetic and apomorphine compared to controls, shown at the top of each section. Top: “Metoc. 10”, “Metoc. 30” etc. indicate times for the administration of 10, 30 and 60ug/g metoclopramide monohydrochloride, respectively. Middle: “Domp. 1.0”, “Domp. 10” etc. indicate times for the administration of 1.0, 10 and 20ug/g domperidone, respectively. Bottom: “Ond. 0.01”, “ Ond. 0.1” ete. indicate times for the administration of 0.01, 0.1 and 1.0ug/g ondansetron hydrochloride dihydrate, respectively. The number of frogs used in each test is given to the right. The control tests in the top and bottom sections, labeled “0.65% NaCl Apo. 20”, are one and the same. The data indicate that all three anti-emetics can significantly retard the onset of apomorphine-induced emesis at selected dosages, but metoclopramide and domperidone themselves induced emesis in some of the specimens. The time of vomiting for those particular specimens is indicated on the time lines to the left of the time when apomorphine was injected into the remaining specimens. Specimens that reacted to the anti-emetics alone were removed from the study once they vomited and were not further treated with apomorphine. Effectiveness of Anti-emetics in Frogs 383 -1 (0) 1 2 3 4 a es 0.65 96 Ne Lye, eke Metoc. 30 Cisp. 100 4 4 Je Sono Meas Geseh es oo OS SEae So a= aS — Domp. 20 Cisp.100 4 4 Ond. 1.0 Cisp.100 4 + Jesas ctheoosecessoscSs cose Scosesssse= e Gispald 0.65 96 NaCl 4 Cisp.100 Metoc. 30 4 4 = -- -S- - - - - - -- - -- — - - - = -- ~~ - - = - - - = == 2 eispice Dome: 20 wm .__-- _-------------- e@--~2------- Cisp.100 Ond. 1.0 4 4 6 7 8 10 14 ----- -@ ----@-------- ---B-- ------ ------ ----- n=4 ----@----@--------- ------ -@@-- ------ ----- n=5 Fic. 2. Cisplatin-induced emesis in the frog Xenopus laevis and the effort to inhibit it with metoclopramide monohydrochloride, domperidone and ondansetron hydrochloride dihydrate. “Cisp. 100” indicates administration of 100ug/g cisplatin. The symbols are the same as in Fig. 1 except that squares indicate juvenile frogs and a solid square indicates that an individual vomited at that time, whereas an open square indicated no emesis. Again, the number of frogs used in each test is given to the right. “Metoc. 30”, “Domp. 20” and “Ond. 1.0” indicate, respectively, the times when 304g/g metoclopramide monohydrochloride, 204g/g domperidone and 1.0ug/g of ondansetron hydrochloride dihydrate were administered. Although these dosages were most effective against apomorphine-induced emesis (Fig. 1), they were not effective against cisplatin. then, they were intermittently checked for an additional day, at which point the experiment was terminated. All experiments were performed between 18-26.5°C. Statistical significance in the re- sponses of the frogs was assessed with the non-parametric, Mann- Whitney U test. Except for some preliminary tests, all experiments concerning apomorphine-induced vomiting were performed in December through March and the studies with cisplatin ran from November until March. A total of 148 frogs were used in these experiments. As indicated by the circles and squares in Figures 1 and 2, sample sizes for each individual anti-emetic test ranged from five to eleven for apomorphine-induced vomiting and from two to six for cisplatin- induced vomiting at any single dosage. Certain frogs, noted on the left side of Figure 1, vomited after injection with either metoclopra- mide or domperidone, but before exposure to apomorphine. These frogs were not treated with apomorphine. The time that they vomited was recorded in Figure 1 and then they were removed from the study. RESULTS Figure 1 (top) confirms that apomorphine is an effective emetic in X. /aevis, as has been reported previously by Naitoh et al. [5]. Metoclopramide did not block this drug-induced emesis at the lowest dosages tested (=10ug/g), but it did significantly (P<.05) delay onset of apomorphine-induced emesis at the next two higher dosages (i.e. 30 and 60u¢/g). A disturbing side effect of metoclopramide was that at all dosages for which it was effective against apomorphine- induced emesis, it was itself emetic in certain individual frogs. The mean latency period for metoclopramide-induced eme- sis, at dosages of 30ug/g or higher, was 9.1 min (see left side of Fig. 1, top section). This contrasts with a mean latency period of 28.8 min for apomorphine at 204g/g, or more than three times as long as with metoclopramide. A similar pattern is seen for domperidone (Fig. 1, middle section). This drug significantly (P<.05) delayed apomor- phine-induced emesis, but, once again, only at dosages where it was itself an emetic (i.e. 20ug/g). Domperidone differed from metoclopramide in the speed with which it induced emesis in Xenopus. Although our sample of domperidone- induced emesis (n=2+4; left side of Fig. 1, lines 10 and 11, respectively) was smaller than for metoclopramide-induced emesis (n=3-+6; left side of Fig. 1, lines 3 and 4, respective- ly) and the dosages were not exactly the same, the data suggest that domperidone takes from one half to two hours longer to induce emesis (Fig. 1). Ondansetron, in contrast to the previous two anti- emetics, did not induce emesis itself at pharmacologically effective dosages. With dosages at or below 0.1ug/g ondan- setron did not significantly inhibit or delay apomorphine induced emesis (Fig. 1, bottom section). At the next highest dosage tested, 1.0u%g/g, ondansetron did significantly retard apomorphine-induced emesis (P< .05). Cisplatin, at dosages of 30ug/g and higher, induced emesis in the frog as it does in mammals. Response times ranged from 3.5 hr up to 23 hr after the cisplatin administra- tion. This anti-cancer agent did not provoke vomiting at dosages of 15yg/¢g or lower (in 15 frogs, not shown in Fig. 2). The response of Xenopus to cisplatin was more delayed and more variable than its response to apomorphine (cf. top of Fig. 1 with top of Fig. 2). None of the three anti-emetic agents proved effective against emesis induced by 100ug/g dosages of cisplatin; i.e. the dosage that was most effective against apomorphine (Fig. 2). No differences in the sensitivity to or effectiveness of these drugs was noted depending on whether small (juvenile) 384 T. Narrou, M. Matuura AND R. J. WASSERSUG or large (adult) frogs were used. Two protocols were tried —one where the anti-emetic agent was presented one hour before the cisplatin and the other where the order of pre- sentation was reversed. No differences between the two administrative regimes were observed. DISCUSSION Our data suggest that both the dopamine antagonists, metoclopramide and domperidone, and the 5-HT3 antago- nist, ondansetron, are anti-emetic agents in amphibians. This result is consistent with those of mammals where metoclopra- mide and domperidone [11] and the 5-HT3 antagonist ICS 205-930 [2] are effective in inhibiting apomorphine-induced vomiting. As a5-HT3 receptor antagonist, ondansetron is a fundamentally different type of anti-emetic than either dom- peridone or metoclopramide [9]. The fact that it works as an anti-emetic in Xenopus suggests that the same or similar serotonin-based, as well as dopamine-based, neural path- ways, associated with emesis in mammals, occur in frogs. Recently we have demonstrated that frogs, like mammals, are susceptible to motion-induced emesis [10]. Collectively these observations raise the prospect that frogs may serve as an alternative model to mammals in emesis and anti-emesis research, where financial or societal considerations restrain mammalian use. The dosage ranges over which Xenopus responds to emetic and anti-emetic medications are not the same as for mammals, but it is difficult to directly compare dosages between homeothermic and poikilothermic vertebrates. The fact is that the Xenopus showed an emetic response to virtually all of the drugs that we experimented with, excluding the controls and the ondansetron. It is noteworthy that metoclopramide and domperidone, in particular, could them- selves induce emesis in some frogs (although when they did not, they remained effective anti-emetics against apomor- phine). This suggests that these frogs are particularly sensi- tive to certain emetic challenges, though the paradoxical emetic response induced by anti-emetics is not restricted to amphibians. Dogs may vomit in response to elevated dos- ages of domperidone [6] and ondansetron [8], and ferrets in response to the 5-HT3 antagonist, zacopride [3]. The drug sensitivity of anurans may be viewed as a positive attribute in an animal being considered as a model species in emetic and anti-emetic research. In the same vein, it must be realized that cisplatin, as a cytotoxic agent, is a far more potent drug than apomorphine. The fact that none of the anti-emetics were effective in preventing or retarding emesis induced by cisplatin testifies to the toxicity of this agent in amphibians. against cisplatin-induced emesis in anurans, but more re- search would be necessary to establish the most effective time Anti-emetics may still be effective course for presenting the anti-emetics in relation to the emetic administration. This work was conducted following the “Guiding Princi- ples for the Care and Use of Animals in the Field of Physiological Sciences” set by the Physiological Society of Japan. ACKNOWLEDGMENTS This research was supported by funds from the Ministry of Education, Science and Culture (Japan), the Natural Science and Engineering Council (Canada) and the Japan Science and Technolo- gy Fund (Canada). We thank Glaxo Group Research Limited (Greenford) and Bristol-Myers Squibb KK (Tokyo) for supplying us with ondansetron and cisplatin, respectively. The draft manuscript profited from critical comments by B. M. Bain, L. Bourque, M. Fejtek and S. Pronych. REFERENCES 1 Daunton NG (1990) Animal models in motion sickness re- search. In “Motion and Space Sickness” Ed by GH Crampton, CRC Press, Boca Raton, FL, pp 87-104 2 Costall B, Naylor RJ, Owera-Atepo JB, Tattersall FD (1989) The responsiveness of the ferret to apomorphine induced eme- sis. Brit J Pharmacol 96: Suppl 329P 3 King G L (1990) Emesis and defecations induced by the 5-hydroxytryptamine (S-HT3) receptor antagonist zacopride in the ferret. J Pharmacol Exp Therap 253:1034-1041 4 Naitoh T, Imamura M, Wassersug RJ (1991) Interspecific variation in the emetic response of anurans. Comp Biochem Physiol 100C: 353-359 5 Naitoh T, Wassersug RJ, Leslie RA (1989) The physiology, morphology, and ontogeny of emetic behavior in anuran amphi- bians. Physiol Zool 62: 819-843 6 Niemegeers CJE, Schellekens KHL, Janssen PAJ (1980) The antiemetic effects of domperidone, a novel potent gastrokinetic. Archiv intern Pharmacodyn Therap 244:130-140 7 Schwartz J-C, Agid Y, Bouthenet M-L, Javoy-Agid F, Llorens- Cortes C, Martres M-P, Pollard H, Sales N, Taquet H (1986) Neurochemical investigations into the human area postrema. In “Nausea and Vomiting: Mechanisms and Treatment” Ed by CJ Davis, GV Lake-Bakaar, DG Grahame-Smith. Advances in Applied Neurological Sciences 3, Springer-Verlag, Berlin, pp 18-30 8 Tucker ML, Jackson MR, Scales MDC, Spurling NW, Tweats DJ, Capel-Edwards K (1989) Ondansetron: Pre-clinical safety evaluation. Eur J Cancer Clin Oncol 25: Suppl 1, S79-S93 9 Tyers MB, Bunce KT, Humphrey PPA (1989) Pharmacologic- al and anti-emetic properties of ondansetron. Eur J Cancer Clin Oncol 25: Suppl 1, $15-S19 10 Wassersug RJ, Izumi-Kurotani A, Yamashita M, Naitoh T (1993) Motion sickness in amphibians. Behav Neural Biol 60: 42-51 11 Wauquier A, Niemegeers CJE, Janssen PAJ (1981) Neuropharmacological comparison between domperidone and metoclopramide. Jpn J Pharmacol 31: 305-314 ZOOLOGICAL SCIENCE 11: 385-390 (1994) Electrical Responses of Non-Taste Cells in Frog Tongue and Palate to Chemical Stimuli OsAMU SATA* and TOSHIHIDE SATO** Department of Physiology, Nagasaki University School of Dentistry, 1-7-1 Sakamoto, Nagasaki 852, Japan ABSTRACT—The response characteristics of non-taste epithelial cells on the dorsal and ventral surface of the tongue and the palate of the bullfrogs were investigated with microelectrodes. The resting potential of the non-taste cells ranged from —11.7 to —20.1 mV, which was smaller than that in taste cells of the tongue. The mean amplitudes of depolarizing and hyperpolarizing responses of non-taste cells on the dorsal surface of the tongue for 0.5 mM acetic acid, 0.5 M NaCl, 10 mM quinine-HCl! (Q-HCl) and water except for 1 M sucrose were mostly similar to those of the responses in taste cells. The time to peak of depolarizing responses in non-taste cells was the shortest (1-5 sec) with acetic acid, middle (7-21 sec) with sucrose and Q-HCI and the longest (26-34 sec) with NaCl. These values were almost the same as those in taste cells. It is probable that a depolarization of non-taste epithelial cells in response to taste stimuli is initiated by the generative mechanisms similar to those of a depolarization of taste cells. © 1994 Zoological Society of Japan INTRODUCTION A taste cell responds to gustatory stimulation with a depolarizing or a hyperpolarizing response [20]. The re- sponse behaviors of taste cells have been reported in various species such as frog [1, 19, 20], mudpuppy [15, 24], tiger salamander [21], rat [5, 14], mouse [23] and hamster [5]. On the other hand, supporting non-taste cells in the taste bud or taste disk are known to respond to taste stimuli with a depolarization or a hyperpolarization [17, 18]. The response characteristics of supporting cells in the frog taske disk such as the dose-response curves, response profiles and changes in cell membrane resistance for four basic taste stimuli are comparable with those of taste cells [17]. Various types of non-taste cells, such as neuroblastoma cells [6], sciatic nerve fibers [3], Tetrahymena [2], free nerve endings [3] and Nitella cells [2] have been reported to respond to tastants. Since it is supposed that non-gustatory epithelial cells in various areas of the mouth also respond to a variety of chemicals, the present study was undertaken to examine characteristics of non-taste cell responses in frog oral epithe- lia and to compare them with taste cell responses. An abstract of this study has appeared elsewhere [16]. MATERIALS AND METHODS Preparation The experiments were conducted with adult bullfrogs (Rana catesbeiana) weighing 250-520 g at a room temperature of 17—24°C. * Present address: Department of Endodontics and Operative Dentistry, Nagasaki University School of Dentistry, Nagasaki 852, Japan Accepted June 10, 1994 Received May 25, 1994 ** To whom all correspondence should be addressed. The animal was anesthetized by i.p. injection of a 50% urethane- Ringer solution (3 g/kg body wt.). To prevent the tongue from the spontaneous twitches the hyoglossal and geniohyoid muscles and the hypoglossal nerves were cut bilaterally. Gustatory stimulation The four basic taste stimuli (0.5 M NaCl, 0.5 mM acetic acid, 1 M sucrose and 10 mM quinine-HC! (Q-HCl) and deionized water were used for gustatory stimulation. Acetic acid, Q-HCl and suc- rose were dissolved in 0.1 M NaCl to remove the hyperpolarizing shift of the membrane potential by solvent water of a stimulus solution. As shown in Fig. 1, non-gustatory epithelial cells on the dorsal and ventral tongue surface and the palate surface were used. These epithelial surfaces were adapted to a normal Ringer solution (mM) (115 NaCl, 2.5 KCl, 1.8 CaCh, 5 HEPES; pH 7.2), which was flowed at a rate of 0.08 ml/sec with a solution deliverer [10, 19]. The nozzle of deliverer was put on the tongue or palate about 2 mm away from the microelectrode inserted into a cell. After gustatory stimulation of the tongue and the palate, these surfaces were rinsed with Ringer. The time interval between each stimulation was more than 3 min. Tongue Palate dorsal vomerine tooth filiform p. fungiform p. ventral Fic. 1. Schematic illustration of the tongue and the palate in bullfrog. The intracellular responses of non-taste epithelial cells were recorded from the filiform papillae, the fungiform papillae, the ventral surface of the tongue and the palate near the vomerine teeth (arrows and dotted circle). 386 O. SATA AND T. SATO Recording The apical and middle area of the tongue was used for record- ings. Intracellular recordings were made from non-gustatory epithelial cells of the filiform papillae, the side wall of the fungiform papillae, the ventral side of the tongue and the vomerine teeth area of the palate (Fig. 1). Intracellular responses were also recorded from taste cells within the taste disks of the fungiform papillae. Criteria for the taste cell identification have already been mentioned [1, 10, 17]. Glass capillary microelectrodes were filled with 3M KCI and had a resistance of 20-50 M2. An indifferent electrode of glass capillary (tip outer diameter, 100m) filled with 3% agar-3 M KCl was placed on the tongue or palate surface. The membrane poten- tials were amplified with a microelectrode amplifier (DPZ-16, Dia Medical System, Tokyo), monitored on an oscilloscope and recorded on a pen-recorder. RESULTS Resting potential The resting potentials were recorded from taste cells of the taste disks in the fungiform papillae and from non-taste Resting potentials of epithelial cells : E 2 (e) (e) = = % 2 © EE 0 a -_ O) F 5 zeEeEda — 18) To) £ — r= = © o wo = LL ~ ~ toy) (2) P=] rs] o Cc — Cc = © =) = cD © e ire re > a 0 -10 > E -20 13 P<0.01 - 30 27 0.02 2 9 19 4 10 8 6 24°" 15° 12°14 Hyperpolarization 0 0 0 2 8 0 5 l 3 8 0 0 0 1 No response 0 0 0 2 5 1 l 0 0 0 0 0 0 0 I Total 23 ; 17 [26 22 20) LOS ede 24°15)" 12" 1s ee A; taste cells in the fungiform papillae. B-E; non-taste cells in the fungioform papillae, the filiform papillae, the Non-Taste Cell Responses in Frog mouth are shown in Fig. 3. The responses were depolarizing or hyperpolarizing. The mean amplitudes of acid responses in taste cells, non-taste cells in the fungiform papillae and Responses of epithelial cells to 1 M sucrose A Taste cell in fungiform p. E Palate Ee ree aes Poa Parone Foace Pe ioumy B Fungiform p. 110 mV 35 sec Sa ee eI ow oa | 10 mV C Filiform p. ann 0.05 +20 [= 15 —— be +10 —__J10 mV oo 20 sec 5 a 0 Fa ABCODE Fic. 5. Electrical responses of taste cells and non-taste cells to 0.5M NaCl. (A) Taste cell response. (B)-(E) Non-taste epithelial cell responses at various portions described. The inset shows the mean amplitudes of the NaCl responses. non-taste cells in the filiform papillae were 27.0, 15.9 and 20.3 mV, respectively, which did not show any statistical differ- ence. These values were much larger than those in non-taste cells and non-taste cells in different epithelial areas calls a tine venta Savas o une toneue Ane Une (palette. : ee time to peak of a depolarization induced by 0.5 mM acetic 10mM Q-HCl water Axi By GC nieDsor® Bye Berl nu Bae E These values were the smallest compared with the peak times M 12 iD i iO qe CoD FO qi his of depolarizations induced by the other basic taste and water AED SHO): PEA ED 14 i@ il WM 15 stimuli. Some non-taste cells in the ventral surface of the OURS *2CUCTORee 30S 00) O10 0 tongue and the palate were hyperpolarized or did not respond Die ie TAs Me 1 Th SG 2 Oa vie |): ventral surface of the tongue, and the palate, respectively. acid was 1-—5sec in taste cells as well as non-taste cells. Sucrose responses. Electrical responses to 1 M sucrose of taste cells and non-taste cells are shown in Fig.4. The 388 response magnitude in taste cells was 8.0+2.0 mV, which was significantly larger than those in non-taste cells in four different areas. As shown in Table 1, the hyperpolarized responses appeared in non-taste cells in a high percentage. The time to peak of a depolarization in response to the sucrose was a range of 11-21 sec in both types of cells. NaCl responses. Figure 5 shows an example of re- sponses to 0.5M NaCl in taste cells and non-taste cells. Both types of cells investigated were mostly depolarized by 0.5M NaCl (Table 1). The mean amplitudes of the re- sponses were a range of 10.2-14.6 mV, where no significant difference was found. The peak time of a depolarization in both cells was from 26 to 34 sec, which was the longest of all taste stimuli used. Q-HCI response. Figure 6 shows an example of Q-HCI responses in taste cells and non-taste cells. The amplitude of the Q-HCI responses in 22 taste cells was 3.2+0.5 mV, which was the smallest value of responses evoked by four basic taste stimuli. The Q-HCI responses in the other non-taste cells were almost the same values. Excepting non-taste cells in the filiform papillae, the hyperpolarized and no responses appeared in 20-30% of the non-taste cells in the other epithelia (Table 1). The peak time of a depolarization in both cells evoked by Q-HCI was a range of 7-21 sec. Water responses. As shown in Table 1, many taste cells and non-taste cells responded to deionized water with hyper- polarizing responses. These responses are due to removal of the adapting Ringer solution covering each epithelium by deionized water. Some taste and non-taste cells were de- polarized by water. As shown in the inset of Fig. 7, the mean magnitudes of responses in taste and non-taste cells for water were all negative (Fig. 7). A hyperpolarization range was from —2.7 to —14.7mV. The hyperpolarizing re- sponse of the filiform papilla cells was generally larger than that of the other non-taste cells. O. SATA AND T. SATO Responses of epithelial cells to 10 mM quinine -HCl A Taste cell in fungiform p. E Palate ee oO aoa B Fungiform p. —~ ~— 1 10 mV = P>0.05 +10 12 uw 5 22 16 17 15 C Filiform p. fo 9 & AB. Coby E a IO UY, D Ventral surface of tongue [10m 15 sec Fic. 6. Electrical responses of taste cells and non-taste cells to 10 mM Q-HCI. (A) Taste cell response. (B)-(E) Non-taste epithelial cell responses at various portions described. The inset shows the mean amplitudes of the Q-HCI responses. DISCUSSION It has been reported that supporting cells in the taste bud and the taste disk are depolarized by various chemical stimuli [17, 18, 24]. Depolarizing responses in non-taste cells be- sides the taste organ by various chemical stimuli are reported in mudpuppy epithelial cells [24], neuroblastoma cells [6], Tetrahymena and Nittela cells [2]. In the time course of electrical responses in taste and non-taste cells, the peak time of depolarization evoked by acid stimulus (Fig. 3) was much shorter than that by the other taste stimuli (Fig. 4-7). This result is consistent with the Responses of epithelial cells to deionized water A Taste cell in fungiform p. D Ventral surface of tongue E pO ee a ALB) cep => E 21 £F elma ye aa E Palate ah at 12 20 6 15 a. ee af i — 2 - B Fungiform p. ee eee C_ Filiform p. | 25 sec WOT ee 1 10 mV Fic. 7. Electrical responses of taste cells and non-taste cells to deionized water. (A) taste cell response. (B)-(E) non-taste epithelial cell responses at various portions described. The inset indicates the mean amplitudes of the water responses. Non-Taste Cell Responses in Frog 389 previous study with frog supporting cells in taste disk [17]. The amplitude of resting potentials of non-taste cells investi- gated in the present study was, on the average, 55% of that of the taste cells. The response amplitudes of taste cells and non-taste cells were almost the same when NaCl and Q-HCl were used. However, the response magnitude of taste cells for acetic acid was the same as those of non-taste cells in the dorsal surface of the tongue, but was much larger than those of non-taste cells in the ventral surface of the tongue and the palate. The sucrose responses in taste cells were much larger than those in non-taste cells at every region, suggesting that snear-hinding receptors are formed mostly in the taste cells, but hardly in non-taste cells (Fig. 4). The response characteristics such as the amplitude of response and the peak time of response were, on the whole, very similar between taste cells and non-taste cells. We have been studying ionic mechanisms of receptor potentials in frog taste cells induced by four taste stimuli and deionized water. We have proposed the following mechan- isms: (1) In case of NaCl stimulation, the receptor poten- tials are generated by functions of cationic and anionic channels at the receptive membrane and second messenger- dependent cation channels at the basolateral membrane of the taste cells [8, 9]. (2) In case of acid stimulation, Ca** channels and H* transporters such as H* pump at the receptive membrane play an important role in generating acid-induced receptor potentials [7, 13]. (3) In case of bitter stimulation, the depolarization is produced by a secretion of intracellularly accumulated Cl~ through the apical receptive membrane [10]. (4) In case of sugar stimulation, the recep- tor potential is generated by an entry of extracellular H* through the apical receptive membrane [11]. (5) In case of water stimulation, the receptor potential is generated by a secretion of Cl~ through the apical membrane and by a blockage of K* outflow through the basolateral membrane [12]. It has been reported that a taste cell responds to odorants with a depolarization [4], while an olfactory cell responds to tastants with a depolarization [22]. However, the mechan- isms underlying these responses have not yet been under- stood. Although some non-taste cells in lingual epithelia and other tissues respond to chemical stimuli of very low concen- trations [2, 17, 24], other non-taste cells slightly respond to chemical stimuli of very high concentrations alone (for exam- ple: frog striated muscle fibers, Drosophila salivary gland cells and frog stomach epithelial cells, unpublished data by Sato T). Since the response characteristics of non-taste cells in the frog mouth examined in the present experiments are, on the whole, very similar to those of taste cells, it is probable that tastant-induced responses in both taste cells and non-taste cells are induced by some common chemo-electrical transduc- tion mechanisms, which involve receptor sites and ionic channels of voltage-sensitive and ligand-sensitive types. Molecular transduction mechanisms in non-taste cells have to be clarified in the next step. ACKNOWLEDGMENTS This study was supported in part by Grants-in-Aid (Nos. 03304042, 05404063) for Scientific Research from the Ministry of Education, Science and Culture of Japan and by a Research Grant from Human Frontier Science Program Organization. REFERENCES 1 Akaike N, Noma A, Sato M (1976) Electrical responses of frog taste cells to chemical stimuli. J Physiol 254: 87-107 2 Ataka M, Tsuchii A, Ueda T, Kurihara K, Kobatake Y (1978) Comparative studies on the perception of bitter stimuli in the frog, Tetrahymena, slime mold and Nittela.. Comp Biochem Physiol 61A: 109-115 3 Beidler L M (1965) Comparison of gustatory receptors, olfac- tory receptors, and free nerve endings. Cold Spring Harbor Symp Quant Biol 30: 191-200 4 Kashiwagura T, Kamo N, Kurihara K, Kobatake Y (1977) Responses of the frog gustatory receptors to various odorants. Comp Biochem Physiol 56C: 105-108 5 Kimura K, Beidler L M (1961) Microelectrode study of taste receptors of rat and hamster. J Cell Comp Physiol 58: 131-140 6 Kumazawa T, Kashiwayanagi M, Kurihara K (1985) Neurob- lastoma cell as a model for a taste cell: mechanism of depolariza- tion in response to various bitter substances. Brain Res 333: 27-33 7 Miyamoto T, Okada Y, Sato T (1988) Ionic basis of receptor potential of frog taste cells induced by acid stimuli. J Physiol 405: 699-711 8 Miyamoto T, Okada Y, Sato T (1989) Ionic basis of salt- induced receptor potential in frog taste cells. Comp Biochem Physiol 94A: 591-595 9 Miyamoto T, Okada Y, Sato T (1993) Cationic and anionic channels of apical receptive membrane in a taste cell contribute to generation of salt-induced receptor potential. Comp Biochem Physiol 106A: 489-493 10 Okada Y, Miyamoto T, Sato T (1988) Ionic mechanism of generation of receptor potential in response to quinine in frog taste cell. Brain Res 450: 295-302 11 Okada Y, Miyamoto T, Sato T (1992) The ionic basis of the receptor potential of frog taste cells induced by sugar stimuli. J Exp Biol 162: 23-36 12 Okada Y, Miyamoto T, Sato T (1993) The ionic basis of the receptor potential of frog taste cells induced by water stimuli. J Exp Biol 174: 1-17 13 Okada Y, Miyamoto T, Sato T (1993) Contribution of proton transporter to acid-induced receptor potential in frog taste cells. Comp Biochem Physiol 105A: 725-728 14 Ozeki M, Sato M (1972) Responses of gustatory cells in the tongue of rat to stimuli representing four taste qualities. Comp Biochem Physiol 41A: 391-407 15 Roper S D, McBride D W (1989) Distribution of ion channels on taste cells and its relationship to chemosensory transduction. J Membr Biol 109: 29-39 16 Sata O, Sato T (1988) Electrical responses of oral epithelial cells to taste stimuli in bullfrogs. J Physiol Soc Jpn 50: 507 17 Sata O, Sato T (1990) Electrical responses of supporting cells in the frog taste organ to chemical stimuli. Comp Biochem Physiol 95A: 115-120 18 Sata O, Okada Y, Miyamoto T, Sato T (1992) Dye-coupling among frog (Rana catesbeiana) taste disk cells. Comp Biochem 390 19 20 21 O. SATA AND T. Sato Physiol 103A: 99-103 Sato T, Beidler L M (1975) Membrane resistance change of the frog taste cells in response to water and NaCl. J Gen Physiol 66: 735-763 Sato T (1980) Recent advances in the physiology of taste cells. Prog Neurobiol 14: 25-67 Sugimoto K, Teeter J H (1991) Stimulus-induced currents in isolated taste receptor cells of the larval tiger salamander. Chem Senses 16: 109-122 we 23 24 Takagi S F, lino M, Yarita H (1978) Effects of gustatory stimulants upon the olfactory epithelium of the bullfrog and the carp. Jpn J Physiol 28: 109-128 Tonosaki K, Funakoshi M (1984) The mouse taste cell re- sponse to five sugar stimuli. Comp Biochem Physiol 79A: 625— 630 West C H K, Bernard R A (1978) Intracellular characteristics and responses of taste bud and lingual cells of the mudpuppy. J Gen Physiol 72: 305-326 ZOOLOGICAL SCIENCE 11: 391-398 (1994) Immunogold Colocalization of Opsin and Actin in Drosophila Photoreceptors That Undergo Active Rhabdomere Morphogenesis KENTARO ARIKAWA and ATSUKO MATSUSHITA Department of Biology, Yokohama City University, 22-2 Seto, Kanazawa-ku, Yokohama 236, Japan ABSTRACT—This paper describes the localization of visual pigment opsin and its association with actin in the photoreceptors of newly emerged (within 12 hr after emergence) Drosophila melanogaster. The photoreceptor of newly emerged flies was characterized by the rich content of rough-surfaced endoplasmic reticulum (rER) and the small rhabdomere: the photoreceptor is actively constructing rhabdomere, and therefore suitable to study the mechanism of thabdomere morphogenesis. The photoreceptor specifically contained opsin-bearing structures some of which were enclosed by several layers of membranes. The structure became sparse in 10 d old flies. Opsins in the structure may be incorporated into the new rhabdomere. The antiopsin also labeled the plasma membrane facing to the intraommatidial space and the endomembranes in the cell body. Both regions were furnished by uniformly oriented actin filaments with the plus ends towards the rhabdomere. Such orientation makes the actin filaments possible to be involved in the vectorial transport of materials towards the rhabdomere by a presumptive interaction with the myosin-like ninaC proteins identified © 1994 Zoological Society of Japan in Drosophila photoreceptors. INTRODUCTION Photoreceptor function is maintained throughout the life by continuous turnover of the photoreceptive membrane both in vertebrates [6, 9, 18] and invertebrates [7, 21]. In the arthropod compound eye, old membranes are removed from the phototransductive rhabdomere and digested by the lyso- somal system in the photoreceptor itself [7, 25]. The remov- al of photoreceptive membranes is accompanied by the reciprocal addition of new membranes to the rhabdomeral microvilli. The new membranes should contain visual pig- ment opsin as an integral membrane protein, and should be transported from the cell body towards the base of the rhabdomeral microvilli where the membrane addition is taking place [8, 21-23]. The transport requires force. The force in this case must be able to transport materials towards the rhabdomere. A possible candidate of such a force-producing system is the actin-myosin interaction, because the photoreceptor cell body is furnished by actin filaments [2] and the myosin-like ninaC proteins (NINAC) [13, 17]. If the conventional actin- myosin interaction occurs between the actin and NINACs, the produced force could transport materials along the actin filaments towards their plus ends, which attach to the tip of the rhabdomeral microvilli in arthropod photoreceptors [2, 5, 12]. In fact, mutations in the myosin domain of a NINAC isoform disrupt the accumulation of calmodulin in the rhabdomere [19], suggesting that the actin-NINAC interac- tion is involved in the calmodulin transport into the rhabdo- mere. The interaction may also transport other rhabdo- meral proteins such as opsin. Accepted April 5, 1994 Received November 8, 1993 If the transport of opsin is mediated by actin, the moving opsins are expected to be found close to the actin filaments. Detection of the situation must be easier in the photorecep- tors that are actively constructing the rhabdomere than in the mature photoreceptors. The first aim of this paper is to demonstrate that the newly emerged flies undergo active rhabdomere morphogenesis. Furthermore we present the distribution pattern of opsin and its association with actin filaments in the photoreceptors of the newly emerged flies revealed by the electron microscopic histochemistry. MATERIALS AND METHODS Animals Newly emerged (within 12 hr after adult eclosion) and 10 d old flies of wild-type Drosophila melanogaster (Canton S strain) were obtained from a laboratory stock culture kept under a 12 hr light/ 12 hr dark cycle at 25°C. Conventional electron microscopy Light-adapted compound eyes were fixed with 2% glutaral- dehyde plus 2% paraformaldehyde in 0.1M sodium cacodylate buffer at pH 7.4 (CB) overnight at 4°C. After a brief wash with CB, the tissues were post fixed with 2% OsO, in CB for 2 hr at room temperature. The tissues were then dehydrated through a graded series of ethanol and embedded in Epon. UJtrathin sections, cut at the level of photoreceptor nuclei, were double stained with 4% uranyl acetate in 50% ethanol and Reynolds’ lead citrate solution. The electron micrographs were taken with a JEOL 1200EX electron microscope. We measured the size of the rhabdomeres and other structures on electron micrographs using a digitizer tablet connected to a computer. Electron-microscopic immunogold labeling Light-adapted compound eyes were fixed with 2% glutaral- dehyde plus 2% paraformaldehyde in 0.1 M sodium phosphate buffer 392 at pH 7.4 (PB) for 1 hr at room temperature. The tissues were then dehydrated through a graded series of methanol and embedded in L. R. White resin. Ultrathin sections were cut with a diamond knife and collected on nickel grids. We used a monoclonal mouse IgG against Drosophila Rh1 (anti R1-6 opsin, provided by Dr. T. Tanimura) [10] and a monoclonal mouse IgM against chicken-gizzard actin (Amersham, code N.350). The antiactin detects a single band with apparent molecular weight of 42kD on a Western blot of Drosophila head homogenate: the antibody labels both G- and F-actins [2]. The labeling was done by the following two methods. Mixed-antibody method: Each step of the labeling was done by floating the grid on 5~50yl drop of solution. The sections were first etched with saturated sodium metaperiodate in distilled water for 1 hr, and then blocked with 4% bovine serum albumin (BSA) in PBSG (0.1 M sodium phosphate buffer at pH 7.4 plus 0.5 M NaCl, 0.25% gelatin) for 30min. The blocking was followed by incubation with the mixture of antiopsin (final conc. 1:200~1:400 of the original) and antiactin (final conc. 10~25ug/ml) in 1% BSA in PBSG overnight at 4°C. After washing with PBSG the primary antibodies were detected by the mixture of goat-anti-mouse (GAM) IgG- conjugated 15nm-gold (Janssen, final conc. 1:50) and GAM IgM- conjugated 5 nm-gold (Janssen, final conc. 1:50) in PBSG. Thus the GAM IgG-15 nm gold detects antiopsin (mouse IgG) whereas the GAM IgM-5 nm gold detects antiactin (mouse IgM). Control label- a) Fic. 1. Transverse section of Drosophila photoreceptor. (arrows) in the cell body. (b) 10 d old fly. body, m; mitochondria, R; rhabdomere. Scale bar=1um. SF ae (a) Newly emerged fly (within 12 h after emergence). I; intraommatidial space, LB; lysosomal body, LCB; large complex body, M; multivesicular K. ARIKAWA AND A. MATSUSHITA ing was done by removing either the antiactin or antiopsin from the primary antibody-mixture. Two-surface method: Each step of the labeling was done by floating the grid with the appropriate side down on 10~50yl drop of solution. Etched and blocked surface was incubated with antiactin in PBSG plus 1% BSA (10~25yg/ml) overnight at 4°C. The antiactin was detected by GAM IgM-S nm gold in PBSG (1:50). After washing with distilled water and air-drying, the other surface was etched, blocked, and incubated with antiopsin in PBSG plus 1% BSA (1:200 ~1:400 of the original) overnight at 4°C. The antiopsin was detected by GAM IgG-conjugated 15 nm-gold in PBSG (1:50). Control labeling was done by replacing either antiopsin or antiactin with 1% BSA in PBSG. The sections were then stained and observed as described above. Decoration of actin filaments with myosin subfragment-1 This procedure follows that of Arikawa and Williams [4]. Briefly, isolated light-adapted compound eyes were first incubated with 1.0% Triton X-100 in a buffer solution (150 mM KCl, 2mM DTT, 20 mM Tris-HCl, pH 7.4) at room temperature for 40 min with gentle agitation. After a wash with the buffer for 30-40 min, the eyes were incubated with myosin subfragment-1 (S1, 10~15 mg/ml in the buffer) for 2 hr at room temperature. The tissues were then similarly processed as for the conventional electron microscopy. Note the rich content of rER Opsin and Actin in Fly Photoreceptors 393 RESULTS Anatomy of the photoreceptors of the newly emerged flies A Drosophila ommatidium contains eight photoreceptor cells (R1-8) each bearing a rhabdomere. R1-6 provide the six peripheral rhabdomeres, and R7 and R8 form a tiered thabdomere in the center of the ommatidium. Thus, only seven rhabdomeres are observed in any given transverse section of an ommatidium. The photoreceptor of the newly emerged flies (within 12 hr after emergence) has a well-organized rhabdomere (Fig. 1). The rhabdomere is, however, significantly smaller than the fully developed rhabdomeres of 10 d old flies (Fig. 2): the rhabdomeres are still developing in this stage. The periphery of the photoreceptor cell body is characterized by the rich content of rER (Fig. 1). The amount of rER was quantified as the length appeared in transverse sections. The newly emerged flies contained significantly more rER compared to 10 d old flies (Fig. 2, P<0.01, Student’s ¢-test). Multivesicular bodies (MVBs, Figs. 1 and 3a) and lyso- 1 newly emerged 10d old rER 0 20 40 60 Length / R1-6 (um) Fic. 2. Cross sectional area (mean + se) of the rhabdomeres and the total length of the rER in R1-6 measured in transverse sections at the level of the photoreceptor nuclei. Any peripheral photo- receptor of the newly emerged flies has smaller rhabdomere (P< 0.05, Student’s f-test) and contains more rER (**; P<0.01, Student’s t-test). Measurement was done on 20 ommatidia from 4 individuals for each age (n=4). somal bodies (LBs, Figs. 1 and 3c) were commonly found. Also common was the large structure of irregular shape containing vesicles, ribosomes, and/or rER (Fig. 3e, g). The structures themselves were embedded in the rER mass. Several layers of membranes enclosed the structure in some cases (Fig. 3e). We hereafter refer the structure as the large complex body (LCB). We measured the areas occupied by the MVBs, LBs, and LCBs in R1-—6 in transverse sections at the level of nuclei of the photoreceptors (Fig. 4). The LCB occupied about 1.6% (4.77+1.61ym’) of the total area of R1-6 in newly emerged flies. The area significantly de- creased in 10 d old flies to about 0.1% (0.23+0.084m’, P< 0.05, Student’s t-test), in which the LB reciprocally increased. The area occupied by the MVB remained constant (Fig. 4). Distribution of opsin and actin Figure 5 shows the results of control labeling for the mixed antibody method. Each section was first incubated with either antiactin (Fig. 5a) or antiopsin (Fig. 5b), and both were then reacted with the mixture of GAM IgG-15 nm gold and GAM IgM-5nm gold. Since one of the primary anti- bodies was removed from the initial incubation, gold particles of only one size bound on each section, indicating that the detection system functioned properly. However, the density of antiactin labeling on the rhabdomeres was not consistent. Both in control and experimental double labeling, the density varied between rhabdomeres even in a single section (data not shown). The antiactin labeling in the cell body region and the antiopsin labeling are rather constant. Two surface method gave virtually the same result. The antiactin recognizes actin in all photoreceptors [2], whereas the antiopsin specifically binds to Rh1, the opsin of R1-6 photoreceptors [10]. The following observations were therefore made on R1-6 photoreceptors. Apparent colocalization of antiopsin aad antiactin label- ing was observed on the rhabdomere (e.g., Fig. 6). The antiopsin labeled the MVBs (Fig. 3b), LBs (Fig. 3d), and LCBs (Fig. 3f, h) in the cell body. The vesicles and the lamellated membranes contained in the LCBs were densely labeled with antiopsin, while the labeling was hardly detected on the associated ribosomes and rER. Other regions labeled with antiopsin were the plasma membrane facing to the intraommatidial space (Fig. 6a, b) and the endomem- branes in the cell body (Fig. 6g, h). Opsin-bearing vesicles were found close to the plasma membrane (Fig. 6e). Although rarely, patchy labeling was detected on the rER (Fig. 6b, f). Table 1 summarizes the distribution of anti- opsin labeling in a single photoreceptor. Nearly 45% of labeling was found outside the rhabdomere. Note the signi- ficant decrease in the labeling on the plasma membrane facing to the intraommatidial space in 10 d old flies (Table 1, P< 0.05, Student’s t-test). The difference in the particle num- bers is directly attributed to the difference in the labeling density on the membrane, because unlike the LBs and LCBs, the length of the plasma membranes in the sections does not change between newly emerged and 10 d old flies. 394 K. ARIKAWA AND A. MAatTSusHITA Opsin and Actin in Fly Photoreceptors 895 qin i De aes] C1 newly emerged MVB 10d old LCB Area / R1-6 (um’) Fic. 4. Areas occupied by multivesicular body (MVB), large com- plex body (LCB), and lysosomal body (LB) in R1-6. Measure- ment was done on 23 ommatidia from 5 individuals for each age (n=5). **; P<0.01, *; P<0.05 (Student’s ¢-test). inde . ee “f. : Fic. 5. Control labeling. The L. R. White sections were first labeled either with antiactin (a) or antiopsin (b), and were further treated with the mixture of 15 nm gold-conjugated GAM IgG and 5 nm gold-conjugated GAM IgM. Scale bar=0.2um. Figures 6c and d show S1-decorated actin filaments lining the plasma membrane. The actin filaments point away from the rhabdomere (arrowhead triplets): we could not detect any actin filaments with the opposite polarity in this region. The antiactin labeling around the plasma membrane most likely corresponds to these actin filaments (arrows in Fig. 6g). Antiactin also labeled the filaments in the space between the thabdomere and the cell body (Fig. 6g, h, see also [2]). DISCUSSION The results reported here are the following. First, the thabdomeres of newly emerged flies are developing. These flies are therefore suitable to study the cellular events associ- ated with the rhabdomere morphogenesis. Second, we de- scribed opsin-containing large complex body (LCB) in the rER mass. The opsins in the LCB may be incorporated into the rhabdomeres. Finally, electron microscopic histochem- istry of opsin and actin suggested that the actin filaments serve as a route for opsin transport towards the rhabdomere. Double labeling Two double labeling methods employed here both func- tioned properly, and gave similar results. The labeling density was consistent except for the antiactin labeling on the thabdomere in both methods. The observed variation in the density of antiactin labeling is probably attributed to the arrangement of actin in the rhabdomere: actin contributes to the slender core of the microvilli [2, 20]. We cut the hexagonally packed microvilli, each of which is about 100 nm in diameter, roughly along their longitudinal axes. In the ultrathin sections, about 70 nm in thickness, the number of cores contained in a section would vary depending on the cutting angle and/or the degree of distortion of the microvilli arrangement. The number of cores in the section affects the frequency of epitopes exposed on the section surface, which is a determinant of the labeling density. Opsin synthesis The photoreceptors of newly emerged flies actively synthesize proteins, which is indicated by the rich content of rER in the cell body (Figs. 1 and 2). The proteins should include opsin, for the rER was occasionally labeled with antiopsin (Fig. 6b, f, Table 1). The newly emerged flies appeared to have large opsin- bearing structures of irregular profile, which contain vesicles, ribosomes and/or rER, making the cross sectional appear- ance complex (Figs. 1 and 3). We therefore termed the structure as large complex body (LCB). Some LCBs were enclosed by several layers of membranes densely labeled with antiopsin (Fig. 3f, h). Although some of these resemble autophagic vacuoles, there is also a possibility that the opsins in the LCBs, in the lamellated membranes in particular, are newly synthesized rather than removed from the rhabdo- mere. Our preliminary observation indicates that the LCBs are abundant even in the late pupal stage and then mostly disappear within two days after emergence: the appearance of the LCBs coincides with the activity of the rhabdomere morphogenesis [15]. The functional significance of the LCBs is remained for further investigation. MVBs and LBs are both involved in the membrane degradation [8, 11]. Although not frequent, the MVBs and LBs, both contain opsin, were also found in the newly Fic. 3. Opsin-bearing structures embedded in the rER (arrows) in the photoreceptor cell body. The left column (a,c,e,g) represents the images of conventional EM, whereas the right column (b,d,f,h) represents the L. R. White sections double labeled with antiopsin and antiactin. (a) (b) Multivesicular body. (c) (d) Lysosomal body. (e) (f) Large complex body enclosed by the lamellated membranes (arrowheads). The membranes enclose some rER (arrows). (g) (h) Large complex body without the lamellated membranes. In (h), an LCB enclosed by the lamellated membranes (arrowheads) is also seen. I; intraommatidial space, LB; lysosomal body, M; multivesicular body, R; rhabdomere, r; ribosome, v; vesicles. Scale bar=0.5um. 396 4, A tale! we 2h > ae S Seo NES Fic. 6. Histochemistry of opsin and actin in the photoreceptors of newly emerged flies. Arrows in (b) indicate antiopsin labeling on the rER. The polarity of the filaments was indicated on the left side of the filaments by arrowhead triplets. membrane facing to the intraommatidial space. lining the plasma membrane (arrows). K. ARIKAWA AND A. . MATSUSHITA i a rs Set } owheads) on the plasma (c) (d) S1-decorated actin filaments (e) Opsin-bearing vesicles (arrowheads) found close to the plasma membrane (arrowheads), which was partially lined with antiactin labeling (arrows). mere. Scale bar=0.2um. emerged flies (Fig. 4), suggesting that the rhabdomere de- gradation was not inhibited during the stage where the rhabdomere morphogenesis is actively taking place. Actin and transport of opsin Transverse sections of the retina clearly show that the photoreceptors have two distinct domains: the rhabdomere and the cell body (Fig. 1). The opsins synthesized in the cell body should be transported to the site of function, the trhabdomere, through the gap between the domains. (f) Some portions of rER mass were labeled with antiopsin (arrowheads). appeared to be associated with some endomembranes (arrowheads). (g) (h) Antiopsin labeling scattered in the cell body arrows; antiactin labeling. I; intraommatidial space, R; rhabdo- There are two possible routes for new opsins to reach the rhabdomere: via the plasma membrane facing to the intraom- matidial space [21], and through the subrhabdomeral cister- nae (SRC) [14, 24]. Here the antiopsin clearly labeled the plasma membrane (Fig. 6a, b, Table 1). The opsins in the plasma membrane is probably being transported towards the rhabdomere rather than removed from there. One reason for this is that the rhabdomere volume is increasing in this stage (Fig. 2): the flies retain high activity of rhabdomere morphogenesis that Opsin and Actin in Fly Photoreceptors TaBLE 1. Distribution of 15nm gold particles (antiopsin labeling) on the photoreceptors of newly emerged and 10d old flies newly emerged, n=4 mean particle number+se (%) 10 d old, n=4 thabdomere plasma membrane MVB LB LCB other regions 116.7+18.1 (54.7) 90.14 8.9 (61.3) 1.3+ 0.1 (0.8)** Wilae i. 7/ (7.8) 33.1+ 2.3 (22.4) LCB not found 11.2+ 1.3 (7.7) 49+ 1.0 (2.5) 5.6+ 2.5 (2.5) 84+ 41 (3.6) 65.9+ 5.4 (33.1) 64+ 12 (3.5) 397 Total 2.7.9+21.4 (100.0) 152.1+11.3 (100.0) “Other regions” include endomembranes and rER. LCBs were not found in the measured micrographs of 10 d old flies. The particle numbers were compared with the corresponding regions in the different developmental stage. Particle counting was done on 57 photoreceptors (ranging 2-14 photoreceptors per individual) from 4 individuals for each age (n=4). **- P<0.05, Student’s f-test should require opsin incorporation into the rhabdomere. Secondly, the labeling density on the plasma membrane is significantly higher in the newly emerged than in 10 d old flies (Table 1). Thirdly, the plasma membrane is lined with the uniformly oriented actin filaments with the plus ends towards the rhabdomere (Fig. 6c, d) and anti-NINAC labeling [13]. As suggested by Adams and Pollard [1] in Acanthamaeba myosin I, the myosin-like NINAC could move the mem- brane-embedded opsin towards the plus end of the actin filaments, i.e., towards the rhabdomere. In fact, actin is probably involved in the rhabdomere morphogenesis. In a crab, Hemigrapsus sanguineus, the rhabdom volume increases at dusk and decreases around dawn [3]. The volume increase was inhibited by treating the isolated eye at dusk with cytochalasin D, which disrupts actin filaments. As the cytochalasin D treatment had no effect at night on the enlarged rhabdom, the inhibition of the volume increase should be attributed not to the disintegration of once established rhabdom but to the disruption of some proces- s(es) in the rhabdom morphogenesis [16]. The most plausi- ble process mediated by the presumptive actin-myosin interaction is the transport of opsin. When applied also at dusk, colchicine, a microtubule inhibitor, failed to stop the volume increase, suggesting that the actin was more directly involved in the rhabdomere morphogenesis than the micro- tubules [16]. Further analyses using other inhibitory drugs are in progress. The actin-NINAC interaction also explains another route of opsin transport through the SRC, which is an opsin-containing network that wraps the entire base of the rhabdomere [14, 24]. The SRC is connected to the rER with membrane tubules forming an extensive endomembrane sys- tem. The opsins in the SRC are probably originated from the rER, the site of opsin synthesis, through the membrane tubules, although no opsin had been so far localized in the tubules. Here we detected antiopsin labeling in the region between the rhabdomere and the cell body. The labeling appeared to be associated with the vesicles or pieces of membranes (Fig. 6g, h). These endomembranes are most likely parts of the tubules, for the tubules appear in transverse sections as elongated or swollen vesicles [14], or even as fragmented membranes if one side of the tubule was tangen- tially sectioned. The antiopsin labeling on the rER probably represents new opsins that will be transferred into the tubules (Fig. 6b, f). The region between the rhabdomere and the cell body is furnished also with the NINAC [13]. By the presumptive interaction between actin and NINAC, the opsins embedded in the endomembranes can be transported towards the rhabdomerte. ACKNOWLEDGMENTS We thank Dr. T. Tanimura for providing the monoclonal anti-Drosophila Rh1 opsin antibody. Drs. T. Tanimura, S. Stowe, and E. Eguchi provided helpful comments in the initial stages of the work. We also thank two anonymous referees for many valuable suggestions on the manuscript. The work was supported by the Grants to K. Arikawa from Whitehall Foundation (Florida, USA), Kihara Foundation for Life Sciences (Yokohama), and the Ministry of Education, Science, and Culture of Japan. REFERENCES 1 Adams RJ, Pollard TD (1989) Binding of myosin I to mem- brane lipids. Nature 340: 565-568 2 Arikawa K, Hicks JL, Williams DS (1990) Identification of actin filaments in the rhabdomeral microvilli of Drosophila photoreceptors. J Cell Biol 110: 1993-1998 3 Arikawa K, Kawamata K, Suzuki T, Eguchi E (1987) Daily changes of structure, function and rhodopsin content in the compound eye of the crab Hemigrapsus sanguineus. J Comp Physiol A 161: 161-174 4 Arikawa K, Williams DS (1989) Organization of actin fila- ments and immunocolocalization of alpha-actinin in the connect- 398 10 11 12 13 14 15 K. ARIKAWA AND A. ing cilium of rat photoreceptors. J Comp Neurol 288: 640-646 Baumann O (1992) Structural interactions of actin filaments and endoplasmic reticulum in honeybee photoreceptor cells. Cell Tissue Res 268: 71-79 Besharse JC, Horst CJ (1990) The photoreceptor connecting cilium. A model for the transition zone. In “Ciliary and flagellar membranes” Ed by R. A. Bloodgood, Plenum, New York, pp 389-416 Blest AD (1980). Photoreceptor membrane turnover in arthro- pods: Comparative studies of breakdown processes and their implications. In “The effects of constant light on visual proces- ses” Ed by T. P. Williams and B. N. Baker, Plenum, New York London, pp 217-245 Blest AD (1988) The turnover of phototransductive membrane in compound eyes and ocelli. Adv Insect Physiol 20: 1-53 Bok D (1985) Retinal photoreceptor-pigment epithelium interactions. Invest Ophthalmol Vis Sci 26: 1659-1694 de Couet HG, Tanimura T (1987) Monoclonal antibodies provide evidence that rhodopsin in the outer rhabdomeres of Drosophila melanogaster is not glycosylated. Eur J Cell Biol 44: 50-56 Eguchi E, Waterman TH (1976) Freeze-etch and histochemical evidence for cycling in crayfish photoreceptor membranes. Cell Tissue Res 169: 419-434 Hafner GS, Tokarski TR, Kipp J (1992) Localization of actin in the retina of the crayfish Procambarus clarkii. J Neurocytol 21: 94-104 Hicks JL, Williams DS (1992) Distribution of myosin I-like ninaC proteins in the Drosophila retina and ultrastructural analysis of mutant phenotypes. J Cell Sci 101: 247-254 Matsumoto-Suzuki E, Hirosawa K, Hotta Y (1989) Structure of the subrhabdomeric cisternae in the photoreceptor cells of Drosophila melanogaster. J Neurocytol 18: 87-93 Matsushita A, Arikawa K, Eguchi E (1992) Distribution of opsin in the photoreceptors in the late pupal and newly emerged 16 17 18 19 20 21 22 23 24 25 MATSUSHITA Drosophila. Zool Sci 9: 1232 Matsushita A, Arikawa K, Eguchi E (1993) Disruption of actin filament organization in the photoreceptors inhibits the rhabdom morphogenesis in a crab Hemigrapsus. Zool Sci 10: 86 Montell C, Rubin GM (1988) The Drosophila ninaC locus encodes two photoreceptor cell specific proteins with domains homologous to protein kinases and the myosin heavy chain head. Cell 52: 755-772 Papermaster DS, Schneider BG, Besharse JC (1985) Vesicular transport of newly synthesized opsin from the Golgi apparatus toward the rod outer segment: Ultrastructural immunocytoche- mical and autoradiographic evidence in Xenopus retinas. In- vest Ophthalmol Vis Sci 26: 1386-1404 Porter JA, Yu MJ, Doberstein SK, Pollard TD, Montell C (1993) Dependence of calmodulin localization in the retina on the NINAC unconventional myosin. Science 262: 1038-1042 Saibil HR (1982) An ordered membrane-cytoskeleton network in squid photoreceptor microvilli. J Mol Biol 158: 435—456 Schwemer J (1986) Turnover of photoreceptor membrane and visual pigment in invertebrates. In “The molecular mechanism of photoreception” Ed by H Stieve, Springer-Verlag, Berlin Heidelberg New York Tokyo, pp 303-326 Stark WS, Sapp R, Schilly D (1988) Rhabdomere turnover and rhodopsin cycle: Maintenance of retinula cells in Drosophila melanogaster. J Neurocytol 17: 499-509 Stowe S (1980) Rapid synthesis of photoreceptor membrane and assembly of new microvilli in a crab at dusk. Cell Tissue Res 211: 419-440 Suzuki E, Hirosawa K (1991) Immunoelectron microscopic study of the opsin distribution in the photoreceptor cells of Drosophila melanogaster. J Electron Microsc 40: 187-192 Williams DS, Blest AD (1980) Extracellular shedding of photoreceptor membrane in the open rhabdom of a tipulid fly. Cell Tissue Res 205: 423-438 ZOOLOGICAL SCIENCE 11: 399-406 (1994) © 1994 Zoological Society of Japan Experimental Perturbations of the Litonotus-Euplotes Predator-Prey System NicoLa Ricci and FRANCO VERNI Dipartimento di Scienze dell’Ambiente e Territorio, via A. Volta 6 56100 Pisa, Italy ABSTRACT—A model previously proposed to demonstrate the interactions between Litonotus (predator) and Euplotes (prey), led to a new round of experiments. The different experimental approaches used to solve these questions (starved cells; killed cells; enzymes; lectins; ions; inhibitors) resulted in quite a new model of the cell interactions which accounts for the different steps of the phenomenon: the main point demonstrated by these experiments is that the cellular cortex of both predator and prey is involved in many of the successive steps of the cascade reactions enabling Litonotus to prey upon Euplotes INTRODUCTION Efforts spent in attempting to deepen our understanding of predation among protozoa are completely justified by the basic importance of the process. Protozoa, indeed, were not only the first primary consumers in the primeval Oceans, but the first predators as well [12]. Such a new trophic niche is quite an important one, due to the two consequences it leads to: (a) it creates new empty spaces for new organisms to settle in, (b) it triggers a sort of evolutionary competition between preys and predators (to escape and to strike each other, respectively) as to their morpho-functional acquisitions. Many examples have been already studied and the knowledge of the Didinium-Paramecium [1], Dileptus-Colpidium [34], Enchelys-Tetrahymena [5], Chaenea-Uronema [5]; Homalo- zoon-Paramecium [4] predator-prey systems, cannot but help us to complete our overall picture of this phenomenon considered from a more specific sinecological point of view: in this perspective, indeed, the study of predator-prey interac- tions also lends itself to be used in an attempt to penetrate the adaptive strategies, conditioning the reciprocal (co)evolution of predators and preys. Let us recall the example of the bat-moth relationships, as a truly paradigmatic one, to clarify our idea [28]. In this context we studied another predatory model, namely that of Litonotus lamella-Euplotes crassus, focusing our attention successively on (a) the ultrastructure of the toxicysts of the predators [11], (b) the ultrastructural analysis of the consequences to Euplotes of toxicyst discharge by Litonotus [32]; (c) the peculiar digestion process [33] and, finally, (d) the behavioural patterns following each other along a path characterised by the succession of several basic steps, namely casual encounter (CE); toxicyst discharge (TD); research (R); engulfment of the prey (PE)[8]. The predator-prey interaction model proposed in the previous paper focused our attention on several closely related prob- lems, which represented the targets of the next round of Accepted May 31, 1994 Received December 16, 1993 experiments: (a) how is the toxicyst-discharging system trig- gered and controlled? (b) which are the spatio-temporal sequences of a toxicyst-discharge phenomenon? The unique nature of our pet-organism protozoa (i.e. indeed perfect eukaryotic cells and complete organisms, at the same time) offers a double advantage: (a) it enables us to use all those techniques typical of experimental studies on cell interactions to investigate also the relationships between entire organisms and (b) it allows us to transfer any results obtained for these truly sophisticate organisms to the general field of cell biology. MATERIALS AND METHODS Both L. lamella and E. crassus were grown, collected and used as already described by Ricci and Verni [24]. The observations were made with a Wild M5 (20-60X) stereomicroscope, and a Leitz Orthoplan (400X) microscope (together with its Nomarsky inter- ferential contrast), coupled to a Panasonic TVC camera and a VHS videorecorder. Unless otherwise indicated, the prey organism was E. crassus. he following specific procedures were followed for the different kinds of experiments: Expt. I The effects of starvation on preys and on predators were studied using normal and starved Litonotus, exposed to both normal and starved Euplotes. Normal populations of predators were used 4 days after the last feeding, while the starved ones were tested after 11 days. Normal Euplotes were not fed for 24 hr, while the starved ones were used 7 days after the last feeding. Expt. 2 To study the role played by the body itself, of both the predator and the prey, in the specific toxicyst discharge (TD) processes at the very moment when the two organisms come into direct contact with each other and the TD itself is triggered and actually occurs, both Litonotus and E. crassus were frozen, and then thawed at room temperature (the experimental populations were immersed in liquid nitrogen for 2 min): in this way the structurally and chemically preserved, but physically inert bodies of both Litono- tus and Euplotes were tested with living prey or predator respectively. In some of these experiments homogenized Euplotes were also used. Expt. 3 Whenever a predator contacts a prey TD occurs: does TD affect the microenviroment where it occurred? How far for TD area is such an effect perceived? How long does it persist? How is the behaviour of Euplotes affected? To solve these problems, many TD 400 N. Ricci AND F. VERNI events were videorecorded and the videotapes scored frame by frame according to the standard technique for behavioural studies reported elsewere by Ricci [20]. In this way we quantified: a) the subcircular area where the TD effects are perceived by Euplotes; b) their duration; c) the changes in behaviour of the prey. Expt. 4 To assess the role possibly played by calcium concentration in the sea water, 3 different standard set ups (cf. Expt. 5) were prepared: the first contained standard marine water (control), the second 15 mM calcium chloride, the third the same concentration of CaCl, plus 0.1mM EDTA, to inhibit the effects of the calcium. Previous experiments, carried out with 5, 10, 15, 20, and 25 mM Catt and with 0.01, 0.1 and 10 mM EDTA had shown that the best results were obtained with 15 mM Ca‘* and with 0.1 mM EDTA. In other words these two concentrations were the lowest capable of inducing clearcut results. Five experiments were then carried out and microvideorecorded, to measure: (a) the time lag between the introduction of Litonotus and the first instance of TD (this period of time will be referred to as TD At throughout this paper, it somehow measures the efficiency to Litonotus in intercepting the prey); (b) the time lag between the introduction of Litonotus and the actual engulfment of the killed prey (this period of time will be referred to as I At in this paper, it somehow measures the efficiency of Litonotus feeding on the prey); (c) the length of the backward motion of Litonotus following TD; (d) the number of TD per Litonotus. Expt. 5 The effects of trypsin (Sigma, T8253; concentration 2.5, 2, 1.5, 1 and 0.5%) were also studied. About 50 Litonotus were incubated in the different concentration for various time periods (15, 30, 60, 90, 120, 180, 210 and 240 min); they were washed 3 times and then used in a 50 yl droplet with concentrated Euplotes, to study the TD At, I At and the percentage of inhibited (namely not-toxicyst discharging) Litonotus. When 100% TD inhibition was induced, the Litonotus still incubated by trypsin were washed free of the enzyme fresh water and then used in Euplotes populations to monitor their recovery period. Expt. 6-A The effects of concanavalin-A (Con-A, Sigma C2010; concentration 2, 1.5, 1, 0.5 and 0.25%; treatment time 15, 30, 60, 90, 120, 150, 180 and 210 min) on Litonotus were studied by washing the treated cells after different periods of time and measuring TD 4t, I At and the percentage of inhibited cells, when these experimental Litonotus were transferred into a 50 ul droplet of concentrated preys. The same treatments were also carried out on incubating Litonotus in the same dosages of Con-A and for the same times as before, in the presence of 40mM of a-methyl-D-mannoside, well-known as a specific competitor of Con-A. The same parameters were meas- ured. Expt. 6-B Concentrations of 2 and 1% of Con-A were also used to incubate Euplotes for 15, 30, 60, 90, 120, 150, 180, 210, 270, 330, and 390 min: these populations were washed three times and then ex- posed to Litonotus, to measure the TD 4t, the I At, the percentage of non-discharging Litonotus. Expt. 7 In a final experimental approach, Litonotus was treated with different concentrations (0.5, 0.25, 0.125, 0.06, 0.03 and 0.015%) of cycloheximide (Chx)(Sigma C-6255) for different times (1 to 9 hr); they were used singly with populations of Euplotes to measure TD At, 1 At and the percentage of TD inhibition. When 100% TD inhibition was obtained the still incubated cells were washed free of Chx and then used with Euplotes to study the kinetics of their recovery in terms of the percentage of TD inhibited cells. RESULTS Expt. I The effects of different starvation of Litonotus and Euplotes. Previous microscope observations (Verni, unpublished results) had shown that the longer the starvation, the more caudal the distribution of toxicysts: the effects were statisti- cally significant. On the basis of these findings, the conse- quences of starvation were studied more specifically. The results obtained in this round of experiments de- monstrated that (a) severe starvation affects the efficiency of predator’s TD: Table 1, the III and the IV columns vs the I and the II; (b) the starvation of the preys affects, to a limited extent, the ingestion capability of Litonotus (Table 1, I4t, the II vs the I column and the IV column vs the III), while it does not affect the corresponding TD At. Expt. 2 The predatory interactions between Litonotus and frozen-thawed Euplotes. Litonotus cannot be frozen and then thawed, without being disrupted: no result could be obtained, except that the area where a disrupted Litonotus lies is avoided by the preys. Normal Litonotus exposed to a population of frozen-thawed Euplotes demonstrated that: (a) Litonotus can ingest them without discharging any toxicysts (Fig. 1B); (b)I4t is longer than twice as much as that of the control (Fig. 1: B vs A); (c) when freshly prepared homogenate of Euplotes is added to the system, ID At is strikingly reduced (Fig. 1C). Expt. 3 The TD-affected area. time (sec) 400 300 200 100 Lat UY NO TD ff NOT? ff TDAt L+ftE+heE A B (G Fic. 1. The TD4t (white bars) and the [4t (shadowed bars) of Litonotus (L) in presence of [A] normal Euplotes (E), [B] frozen-thawed Euplotes (ftE) and [C] ftE and homogenated Euplotes (hE). On the ordinates the times in seconds are reported. L+E L+ftE Litonotus-Euplotes Predator-Prey System TABLE 1. 401] The effects of starvation, on the predation of Litonotus lamella on Euplotes crassus, measured by the durations of the TDAt and of the I4t (described in Materials and Methods section). normal Litonotus starved Litonotus + normal + starved + normal + starved Euplotes Euplotes Euplotes Euplotes x 1’25”’ 1’20”’ PALA) APU”? TDAt s 35) 50” 1712 1718” n 35 26 29 82, X 2°30”’ 3’ 3’42”’ 4’9”’ IA4t S V IPS" IPI? 1718” n 35 26 29 32 % of E avoiding a struck E % of E avoiding frozen-thawed L i 90 120 150 180 210 times (sec) Fic. 2. The temporal trend of avoidance of the “toxic” area by Euplotes, expressed in terms of individuals (ordinate: %) creep- ing away from it, in the time (abscissa: sec). The shadowed bars show the same trend of Euplotes avoiding frozen-thawed Litonotus The results we obtained are the following: (a) the effects of a TD event extend around the TD point over a sub-circular area of about 300um in diameter, (b) the same effects increase up to their maximum for 90 sec after the TD event and last for about 3 min; (c) the behaviour of the preys is clearly affected by the TD event as demonstrated by the high frequency of avoidances induced in Euplotes by TD im- mediately afterwards (Fig. 2, shadowed areas); the avoi- dances correspond to the behavioural pattern called Side Stepping Reaction [19] and indicated as SSR; (d) these SSR occur according to a temporal pattern parallel to that of Euplotes avoiding the area where a frozen-thawed disrupted Litonotus is placed (Fig. 2, white areas). TaBLe2. The effects of Ca** (15 mM) on the 4 parameters indicated on the left, BM=the length of the backward motion of a Litonotus lamella after discharging its toxicysts: it is measured in Relative Units (RU): 1 RU=1 species-specific length of Litonotus lamella =250um. The number of observations for each parameter was of 50 populations 15mM Catt & Control 15mM Catt Pan 0.1mM EDTA TDAt (sec) 43+17 41.5+18 42+26 BM (RU) 1.140.7 2.5+0.7 1.2+0.6 IAt (sec) 132+30 186+80 139+33 0 n- of TD per iene P 1.2+0.4 2.440.5 1.2+0.5 Expt. 4 The role of Ca*~ in the environment. The results given in Table 2 clearly show that calcium ions affect the general physiology of the Litonotus-Euplotes system at least at three different levels: (a) TD At is not affected; (b) the length of the backward motion is doubled; (c) I4t increases by about 50%; (d) the number of TD per Litonotus is doubled. EDTA inhibits calcium effects, as expected. As considered in the Materials and Methods section, the differential effects of calcium ions on TD At (no effect) and on [At (strong inhibition), actually demonstrate the importance of both parameters in interpreting correctly the effects of calcium ions on the Litonotus-Euplotes system. Expt. 5 The effect of trypsin on Litonotus The enzyme was already known to affect the physiology of Litonotus; progressively higher concentrations and longer treatments, indeed, induce (a) a progressive darkening of the cytoplasm, (b) a rounding of the body shape, (c) reduced locomotion, (d) immobilization and (e) lysis of the cell (Ricci and Verni, unpublished results). It was found (Fig. 3) that the different concentrations used for the different times indicated in Material and Methods, clearly affect TD At (the higher the concentration, the longer TD At) but not I4t (the difference between TD At and [At is indeed rather constant). For a certain concentration a longer treatment affects the percentage of cells not discharging their toxicysts (Fig. 4), while TD At tends to be more or less constant, the different values depending solely upon trypsin concentration. Once 402 N. Ricct AND F. VERNI controls: TD At=1"30” | % of discharging Litonotus 0 30 60 90 120 150 180 210 240 270 treatment (min) Fic. 4. The progressive action of the trypsin in the time: for a certain concentration the longer the treatment (abscissa) the smaller the percentage of discharging Litonotus. The figure close to each of the different points along the curves, represents the average TDAt observed for that point; the figures in the small panels represent the trypsin concentration for the different curves and the relative mean values of the TDAt. 0 0.5 1.0 1.5 within 60-120 min. Similar treatments with trypsin were also conducted on Euplotes; the enzyme either has no effect al all on both TD At or IAt, and it kills the preys. Expt. 6A The effects of Con-A on Litonotus. trypsin (%) Fic. 3. The average TDAt and IAt induced by the concentrations of trypsin, given on the abscissa; 0% of trypsin=controls 100% inhibition is induced (namely, when no Litonotus can The results of the experiments are the following: (a) the discharge its toxicysts) they still behave quite normally and higher the Con-A concentration, the stronger the effect, on their physiology seems to be unaffected: after washing, these TD At, but not on I4t (Table 3) (b) for the same concentra- completely inhibited populations recover 80% of their TD tion, the longer the treatment, the stronger the effect on TD activity within 30-60 min, and 100% of their potentialities At but not on [dt (Table 3) (c) as the Con-A concentration TABLE 3. The TDAt and the [4t of different populations of Litonotus treated by different concentrations of Con A (shown in the first line) for different times, given in the left column ; Cond 2% 1.5% 1% 0.5% 0.25% control | TD4 I4t TDA at TDA ‘It TDA ‘It TDA I4t TDA It | n 15 15 10 10 10 10 10 10 8 8 15a ESB 0 Ie SSeS? 4 ? 27200 3° 220" = SND=) INDEeee? 3° see uty 1720” 1°46” 50” 07 =e ibid? 1705” © 1725” n 14 14 14 14 12 12 11 11 12 12 8 8 30° x 330°" 35° :10°50” 15°30” «6°23 =——s730 = ss 210”,Ss«330” )=—s 21S?) 3724 «1°20 soe geen = 219703? B55 3°27 tame LoTeest 21 150° n 16 16 9 9 10 10 1 1 8 8 60k + j) wne=30 $357 1807". 11507520700 7 227° 3402) 1230) D558 5 Ld a 5715) 5? 220? 2? 12”) 225? P15? 220% n 10 10 7 7 8 8 | 97 =x t t 11°15”. 13°30 10°35” 12°15. ND». SND © /2710") 737152 s 518? TAT», 22024 23302 050” 1°35” n 18 18 13 13 8 8 1200 x >30? >35%17 12°35? 1343? | ND). <=ND., 1°20%., 27207 5 wh = 24" 1°21? 1705”, 1°50% | n ; 6 6 10 10 8 8 | 150° x t + 12°40” 13°26” 5’ 642”. 1°50”. 3? | = i) n an — S 1367, 40 a On Se oOr | Litonotus-Euplotes Predator-Prey System 403 0.25% & control % of discharging Litonotus 0 20 40 60 80 100 120 140 160 180 200 treatments (min) Fic.5. The effects of Con-A (at the concentrations indicated for the different curves) are expressed as percentages of Litonotus discharging their toxicysts (ordinate), after being treated for different times (abscissa). The arrowheads indicate the onset of the inhibitory effects for the different concentrations. % of inhibition of discharging Litonotus an lasearsay Lary 1 0 60 120 180 240 300 360 420 480 540 treatment (min) Fic. 6. The percentages of inhibited Litonotus (ordinate) given as a function of the different concentrations (indicated in the figure itself) and of the length of treatment (abscissa). decreases (1.5; 1.0; 0.5%), the time which elapses between the onset of the treatment and the beginning of the inhibitory effects of the lectin itself increases (0; 50; 110 min respective- ly, as shown by the arrowheads in Fig. 5); (d) for the same concentration, the longer the treatment, the lower the per- centage of Litonotus discharging their toxicysts (Fig. 5); (e) when Litonotus treated with Con-A are washed free from the lectin, the inhibitory effects are removed after a period whose length is roughly proportional to the Con-A concentration: while, for instance, 1% Con-A inhibits completely the TD after 90 min and the recovery period, after washing, is about 4 hr, 0.5% Con-A inhibits TD completely after 3.5 hr, but the populations start discharging their toxicysts again only 40 min after washing; (f) the effects of Con-A are specific, being completely inhibited by a -methyl-D-mannoside, the specific competitor of Con-A. Expt. 6B The effects of Con-A on Euplotes. The results showed that this lectin affects the prey in the same way as it proved to do with the predator: (a) Con-A treated Euplotes induce a reduction of TD At and the percentage of TD predators, according to a clearcut dose- dpendence; (b) the longer the treatment period of Euplotes with a certain concentration of Con-A, the longer the TD At of Litonotus and the higher the percentage of non TD predators; (c) when Con-A treated Euplotes do not induce any TD at all, they are washed: the recovery periods required for Litonotus to discharge its toxicysts again are proportional to the Con-A concentration used on Euplotes; (d) the effects of Con-A on Euplotes are specific, being completely inhibited by a-methyl-D-mannoside. Expt. 7 The effects of cycloheximide on Litonotus. When Litonotus is treated with cycloheximide TD is severely affected: (a) the higher the Chx concentrations, the longer TD At (Table 4) and the larger the amount of non TD Litonotus (Fig. 6). Also the interference effects of Chx are easily reversed upon washing and the recovery times are proportional to the concentrations used for a certain treat- ment. DISCUSSION The experiments reported here led to a widening and deepening of the general knowledge of the predator-prey relationships between Litonotus and Euplotes [24]. First of all, it was found that starvation acts on both elements of the system, although differentially: (a) in Litono- tus it affects both the cortical areas where TD occurs (the more severe starvation, the more caudal the TD) and TD 4t, a parameter somehow measuring the “hunting efficiency” of Litonotus, (the longer the starvation, the more inefficient it becomes): (b) starved Euplotes are ingested in longer times. TaBLE 4. The inhibitory effects of Chx on TD of Litonotus: the TDAt (min) on organisms treated by different concentrations (%, shown in the left column) for different hours (indicated in the first line) % 1 2 3 4 5 6 7 8 9 hr 0.015 VY 1730” 3°30” 5°00” 5°00” 5°00” 6700” 8700” >30’ 0.03 136” 154” 400” 6°00” 6700” >30’ >30’ >30’ >30’ 0.06 De 154” 5°00” 6°00” >30’ >30’ >30’ >30’ >30’ 0.125 3°50” 3°24” >30 >30’ >30’ >30’ >30’ >30’ >30’ 0.25 4°24” 3°48” >30° >30’ >30’ >30’ >30° > 30’ >30’ 0.5 >30’ >30 >30’ >30’ >30’ >30’ >30’ >30’ >30’ 404 N. Ricci AND F. VERNI In general it seems possible to conclude that a basic physiolo- gical “health” of both predator and prey represents a sort of prerequisite for predation to occur efficiently. Freezing-thawing experiments provided several clues to a deeper understanding of our predation model. Frozen- thawed Euplotes (a) maintain their shape, (b) cannot induce TD in Litonotus, and (c) still suitable preys for Litonotus to eat. These results seem to suggest that some “activation energy” (related to the kinetic energy of a normally swim- ming Euplotes) might be required for TD. Similar results on the other hand, have also been found by Ricci et al. [26] regarding the mechanisms which trigger specific cell dif- ferentiation (giantic cell) in Oxytricha bifaria. Experiments carried out with paralyzed mutants of Euplotes [23] are expected to give more precise answers to the question: the fairly small amount of energy involved in this process, however, should not to be considered exceptional since the world of ciliates has already proved to be ruled by unex- pectedly small forces, as demonstrated by Machemer [16] and by Ricci et al. [27]. According to these results, moreover, TD does not represent a step necessarily occurring before ingestion, although TD itself seems to facilitate somehow the actual ingestion, as demonstrated by I4t of Litonotus en- gulfing frozen-thawed Euplotes: the values of these [At are, on average, more than twice than those of the controls. A further clue to the comprehension of the factors involved in the Litonotus-Euplotes interactions is given by the use of the homogenate of Euplotes, which strikingly reduces Idt of Litonotus engulfing frozen-thawed Euplotes. These evi- dences seem to suggest that a cytoplasmatic factor (or com- plex of factors) of Euplotes (called «) is capable of “activact- ing” Litonotus. The nature of this “activation” is at present being studied through a behavioural approach, namely inves- tigation of the specific locomotory pattern of Litonotus ex- posed to the homogenate of Euplotes, and related to the normal searching activity of the predator already reported by Ricci and Verni [24]. The findings of experiment 3 show that a “toxic” (in its broadest sense) area persists, where a Euplotes has been killed by a Litonotus: its spatial extension and temporal duration recalls quite closely that of the area where a frozen-thawed Litonotus is placed. This finding strongly suggests that Litonotus releases a factor (or complex of factors), whenever it “breaks down” or discharges its toxi- cysts. According to our present understanding, a Litonotus must depolarize its membrane to discharge its toxicysts and this, in turn, cannot but lead it to creep backwards for a while. To “buffer”, somehow, this unavoidable physiologic- al handicap, Litonotus may be supposed to release a subst- ance (called 4) at the TD point, capable of guiding it towards the prey: on the other hand the same A might represent the “stay away” signal for Euplotes, thus producing what we called the “toxic area”. Similar “repulsive” effects have been described also for Colpidium avoiding killed Dileptus [6]. The use of 15 mM Ca** produced clearcut results, in terms of a larger number of TD per Litonotus and longer backward locomotion following the TD: these observations strengthened our hypothesis of electrically controlled TD for Litonotus, according to both what is already known for Dileptus [6] and to the more specific report of Hara et al. [13] for Didinium. The dramatic effects of proteolytic enzymes on a wide range of biological phenomena of ciliates are already known: binary fission [36], conjugation [15, 22, 35] and feeding processes [6, 7, 30, 37]. The data reported here show that while no effect can be detected when trypsin acts on Euplotes, (Expt.5) the same enzyme affects (a) the physiology of Litonotus (cytoplasm, body shape, and locomotion) in much the same ways as in Dileptus [6], (b) the average TD At and (c) the recovery periods of washed populations, with a generally clearcut dose dependence. An exception is repre- sented by the difference between I4t and TD At, which is fairly constant at the different trypsin concentrations: this observation indicates that TD and ingestion differ from each other from a physiological point of view, as also suggested by the results of the experiments carried out with frozen-thawed Euplotes. The most surprising effect of trypsin, however, is the finding that induced TD At depends only on the concen- tration itself and not on the length of the treatment, which is on the contrary capable of inhibiting (completely and revers- ibly) the TD of the Litonotus population. Such a puzzling result (strong numerical inhibition of Litonotus vs constant TD At at a certain concentration of trypsin, upon longer treatment times) might be accounted for by a working hypothesis, based on some kind of accumulation of proteins, possibly involved in the TD processes (called TDP). Similar accumulations have already been hypothesized by Beisson and Rossignol [2] for the trichocyst discharging of Para- mecium, by Heckmann and Siegel [14] for the preconjugant cell interactions of Euplotes crassus, and by Ricci et al. [25] for the differentiation of carnivorous giants in Oxytricha bifaria. According to our hypothesis, the higher the trypsin concentration (i.e. the larger the number of TDP molecules) the larger also the number of digested cortical TDP. If Litonotus can substitute them at a certain velocity, recruiting them from a cytoplasmic pool, the effect of trypsin can somehow be counterbalanced, at least for a certain time: the average number of newly-exposed and ready-to-work TDP can be expected to be inversely proportional to the number of trypsin molecules in the environment (and not related to the length of the treatment!). According to this way of thinking the reason why the trypsin concentration is directly prop- ortional to the time required to reach the physiological threshold of TDP per surface unit requested for the TD itself can be easily explained. Only when the entire cytoplasmic pool of presynthetized TDP is exhausted is a Litonotus inhibited from discharging its toxicysts. The temporal trend of the progressive inhibition of the entire population, on the contrary, would depend upon the relative differences occur- ring among different Litonotus with regard to the pool of presynthetized TDP and to the TD threshold. This general working hypothesis seems to be supported and strengthened Litonotus-Euplotes Predator-Prey System 405 The PREDATOR © discharges its toxicysts © releases Species- Predator 98 -Prey Specific : DISET RECOGNITION CONTACT The PREY-FULLY STRUCK - moves searches backwards for the prey (upon successful search) lies immobile and releases ©) Fic. 7. The schematic drawing of the successive steps occuring when a Litonotus (predator) comes in direct contact with an E. crassus (prey): the upper, vertical arrows show the parameters affecting the different steps. by the observed effects of cycloheximide, which induces increasingly long recovery periods for the synthesizing machinery, when increasingly strong treatments are used. A final consideration must be made about the data obtained with Con-A, in the context of what is already known about its effects on different aspects of the biology of ciliates: feeding [10, 29], growth [3, 29, 35], conjugation [10, 17, 18, 21, 31], locomotion [9], cell differentiation [26]. The effects of Con-A show a clear dose dependence for both the inhibi- tory effect of TD and TD 4t itself. Con-A seems to affect the Litonotus-Euplotes system by reacting with cortical gly- coproteins of both Litonotus and Euplotes and not through an aspecific toxicity, as demonstrated by the use of both the washing procedure and a@-methyl-D-mannoside. What still remains to be ascertained is the degree of its specificity for TDP: does Con-A react specifically with TDP or, rather, with aspecific cortical glycoproteins, thus simply masking the TDP themselves? Ad hoc experiments presently being conducted are expected to solve the problem. Regardless of its way of acting on the Litonotus-Euplotes system, Con-A works dif- ferently from trypsin: (a) stronger dosages of this lectin, in fact, induce longer TD At; (b) Con-A inhibits TD in a population of Litonotus with a time delay roughly proportion- al to its dilution, while trypsin acts immediately. The differentiated approches used to study the Litono- tus-Euplotes biological system led us to draw a complex picture of the different steps following each other during the time elapsing between the initial contact between predator and prey and the actual ingestion (Fig. 7). Direct cell con- tacts are required for the phenomenon to occur, a species- specific recognition represents the first step of sort of double scale: the Predator (a) discharges its toxicysts and releases some chemical, A; (b) it moves backwards, (c) it searches for the prey and if everything works out properly, (d) it ingests the prey. The Prey, on the other hand, may be either fully struck (it quits moving and releases a particular chemical, «) or only partially struck (its locomotion is severely affected, but it slowly recovers its normal physiology). As indicated on the upper part of the figure itself, many factors, all contributing to the clarification of our understanding of the phenomenon, may interfere either positively or negatively with the physiology of the predator. Although quite com- plex, this synoptic scheme represents an interesting achieve- ment, being not only an exhaustive summary of what is so far known, but rather a polyhedric working hypothesis, extreme- ly useful for proceeding towards the next phases of investiga- tion of the Litonotus-Euplotes story. ACKNOWLEDGMENTS The authors are truly indebted to Dr. Fabrizio Erra, for both the scientific discussions and the capable assistance in handling the graphic problems. REFERENCES 1 Balbiani EG (1873) Observations sur le Didinium nasutum. Arch Zool Exp 2: 363-394 2 Beisson J, Rossignol M (1975) Movements and positioning of organelles in Paramecium aurelia. In “Molecular biology of nucleocytoplasmic relationships” Ed by S. Puiseux-Dao, Else- vier, Amsterdam, pp 291-294 3 Bessler WG, Lipps HJ (1976) Concanavalin A-induced nuclear division in Stylonychia mytilus. J Gen Microbiol 92: 221-223 4 Diller WF (1964) Observation on the morphology and life history of Homalozoon vermiculare. Arch Protistol 107: 351- 362 5 Dragesco J (1962) Capture et ingestion des proies chez les infusories ciliés. Bull Biol Fr Belg 46: 123-167 6 Esteve JC (1981) Perturbation du comportement alimentaire par atteinte du revétement cellulaire chez le cilié Dileptus. Protistologica 17: 479-488 7 Esteve JC (1984) Calcium, calmoduline, sites glucidiques et 406 10 11 12 13 14 15 16 17 18 19 20 21 22 N. Ricci AND F. VERNI ingestion chez le cilié gymnostome Dileptus sp. Protistologica 20: 15-25 Fenchel T (1987) Ecology of Protozoa. Berlin Frisch A, Bessler W, Lipps HJ, Ammermann D (1976) Im- mobilization of ciliates by concanavalin-A. J Protozool 23: 427- 430 Frisch A, Loyter A (1977) Inhibition of conjugation in Tet- rahymena pyriformis by Con A. Localization of Con A binding sites. Exp Cell Res 23: 337-346 Giambruni A, Verni F (1986) Studio ultrastrutturale delle toxicisti del ciliato predatore Litonotus lamella Boll Zool 53: 20 Gould SJ, Raup DM, Sepkoski JJ Jr, Schopf TJM, Simberloff DS (1977) The shape of evolution. A comparison of real and random clades. Paleobiology 3: 173-181 Hara R, Asai H, Naitoh Y (1985) Electrical responses of the carnivorous ciliate Didinium nasutum in relation to discharge of the extrusive organelles. J Exp Biol 119: 211-224 Heckmann K, Siegel RW (1964) Evidence for the induction of mating-type substances by cell to cell contacts. Exp Cell Biol 36: 688-691 Kitamura A, Hiwatashi K (1978) Are sugar residues involved in the specific cell recognition of mating in Paramecium? J Exp Biol 203: 99-108 Machemer H (1993) Ciliate gravireception and graviresponses. IX Inter Congress Protozool July 25-31 Berlin, Germany Offer L, Levkovitz H, Loyter A (1976) Conjugation in Tet- Springer Verlag, rahymena pyriformis. The effect of polylysine, concanavalin A and bivalent metals on the conjugation process. J Cell Biol 70: 287-293 Revoltella R, Ricci N, Esposito F, Nobili R (1976) Cell surface control in Blepharisma intermedium Bhandary (Protozoa Ciliata). 1. Distinctive membrane receptors-sites in mating type 1 cells for various interacting ligands. Monitore Zool Ital (N.S.) 10: 279-292 Ricci N (1981) The ethogram of Oxytricha bifaria (Ciliata, Hypotrichida). II. The motile behaviour. Acta Protozool 20: 393-410 Ricci N (1992) Qualitative study and quantitative analysis of behavior of ciliated protozoa: principles, tecniques, tricks. In “Protocols in Protozoology” Ed by JJ Lee, AT Soldo, Soc Protozol, Allen Press, Lawrence, Kansas, pp B-14.1-14.16 Ricci N, Esposito F, Nobili R, Revoltella R (1976) Cell surface control in Blepharisma intermedium (Protozoa Ciliata). Block- ing of gamone IJ-induced omotypic pairing by different mem- brane interacting ligands. J Cell Physiol 88: 363-370 Ricci N, Cetera R, Banchetti R (1980) Cell to cell contacts mediating mating type dependent recognition(s) during the 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 preconjugant cell interactions of Oxytricha bifaria (Ciliata, Hypotrichida). J Exp Zool 211: 171-183 Ricci N, Miceli C, Giannetti R (1987) The ethogram of Eu- plotes crassus (Ciliata, Hypotrichida). II. The paralyzed mutant. Acta Protozool 26: 295-307 Ricci N, Verni F (1988) Motor and predatory behavior of Litonotus lamella (Protozoa, Ciliata). Can J Zool 66: 1973- 1981 Ricci N, Grandini G, Bravi A, Banchetti R (1991) The giant of Oxytricha bifaria: a peculiar cell differentiation triggered and controlled by cell to cell contacts. Eur J Protistol 27: 127-133 Ricci N, Bravi A, Grandini G, Cifarelli D, Gualtieri P, Coltelli P, Banchetti R (1991) The formation of giants in Oxytricha bifaria: a peculiar multi-step cell differentiation. Eur J Protis- tol 27: 264-268 Ricci N, Erra F, Russo A, Banchetti R (1992) The interface between air and water: a perturbation source adaptive behaviour in Ciliates. J Protozool 39: 521-525 Roeder KD (1968) Interactions of bats and moths. tik 13: 75-87 Sharabi Y, Gilboa-Garber N (1980) Interactions of Pseudomo- nas aeruginosa hemagglutinins with Euglena gracilis, Chlamydo- monas reinhardi and Tetrahymena pyriformis. J Protozool 27: 80-83 Tolloczko B (1975) Endocytosis in Paramecium. 1. trypsine and pronase. Acta Protozool 14: 313-320 Tsukii Y, Hiwatashi K (1978) Inhibition of early events of sexual process in Paramecium by concanavalin A. J Exp Zool 205: 439-446 Verni F (1985) Litonotus-Euplotes (predator-prey) interaction: ciliary structure modification of the prey caused by toxicysts of the predator (Protozoa, Ciliata). Zoomorphology 105: 333- 335 Verni F, Rosati G (1992) A comparative study of digestion in a raptorial ciliate and in the facultative carnivorous form of a filter feeding ciliate. Tissue and Cell 24: 443-453 Visscher JP (1923) Feeding reactions in the Ciliate Dileptus gigas with special reference to the function of the trichocysts. Biol Bull 45: 113-143 Watanabe T (1977) Chemical properties of mating substances in Paramecium caudatum: effect of various agents on mating reactivity of detached cilia. Cell Struct Funct 2: 241-247 Wyroba E (1978) Cell multiplication following partial enzyma- tic removal of surface coat. Cytobiologie 17: 412-420 Wyroba E, Brutkowska M (1978) Restoration of phagocytic activity of Paramecium tetraurelia suppressed by previous diges- tion of surface coat. Acta Protozool 17: 515-524 Kyberne- Effects of ZOOLOGICAL SCIENCE 11: 407-412 (1994) © 1994 Zoological Society of Japan Immunochemical Studies of an Actin-binding Protein in Ascidian Body Wall Smooth Muscle Yukio OntsuKka!, Hiroki NAKAE*, HirosHI ABE and TAKASHI Osinata!> ‘Department of Biology, Faculty of Science, Chiba University, Yayoi-cho, Inage-ku, Chiba, Chiba 263, and *Advanced Research Laboratory, Research and Development Center, Toshiba Corporation 1, Komukai Toshiba-cho, Saiwai-ku, Kawasaki 210, Japan ABSTRACT—A monoclonal antibody (McAb, AS23) was prepared to the proteins extracted from ascidian body wall smooth muscle by a dilute alkaline salt solution. The McAb recognized a protein of about 80 kDa (termed 80K protein) and the protein was detected most abundantly in the body wall smooth muscle by both immunoblotting and immunocytochemical methods. The 80K protein was enriched in isolated thin actin filaments, and in addition, it was observed that AS23 binds to one end of thin filaments by immunoelectron microscopy. The results indicate that the 80K protein is an actin-binding protein which can associate with filamentous actin in ascidian body wall smooth muscle. INTRODUCTION The body wall, or mantle, of ascidian is constituted of two layers of smooth muscle which resembles smooth muscles of vertebrates morphologically, especially in the pattern of myofilament arrangement [17]. This muscle has attracted attention since it is unique multinucleated [16, 18] and troponin-containing smooth muscle [22]; it is the only known example of a troponin/tropomyosin-regulated muscle which contains no organized sarcomeric structures. Thus, the asci- dian smooth muscle is not obviously homologous with either vertebrate smooth or striated muscle, but appears intermedi- ate between smooth and striated muscles of vertebrates. Contractile proteins, actin and myosin, of this muscle have been investigated [14]. In addition, three troponin components, troponin T, I and C, [5] and tropomyosin [11] have been isolated and characterized. However, nothing is known as to the actin-binding proteins which may be involved in the regulation of actin assembly and/or filament organiza- tion in this muscle. According to electron microscopic observations [17-19], the contractile apparatus consists of many irregularly arranged filament bundles which are sepa- rated from each other by a network of intermediate filaments. Actin-containing thin filaments appear considerably longer than those in vertebrate smooth or striated muscles [17]. During the contraction of the body wall muscle, the arrange- ment of myofilaments is significantly altered [20]. There- fore, it is of interest how the arrangement of myofilaments is controlled in the cytoplasm. In order to understand the ascidian muscle cells from functional and structural aspects, it Accepted May 16, 1994 Received April 14, 1994 > To whom correspondence and reprint requests should be addressed. seems to be indispensable to characterize actin-binding pro- teins which may be involved in assembly, organization and re-distribution of actin filaments. In this investigation, as the first step to clarify actin- binding proteins in ascidians, we prepared monoclonal anti- bodies to ascidian muscle protein extracts which we assumed to contain actin-binding proteins as judged by the knowledge on vertebrate smooth muscle [7]. One of the antibodies recognized an 80 kDa protein which is enriched in ascidian body wall smooth muscle and is associated with its thin actin filaments. MATERIALS AND METHODS Muscle source Ascidians, Halocynthia roretzi, were obtained from a local seafood market. The fresh body wall muscle was isolated, frozen immediately in liquid nitrogen, and stored at —80°C until use. Antibodies The proteins as an immunogen were prepared as described previously [12]; Briefly, the isolated body wall muscle was homoge- nized in ice-cold distilled water containing 0.5mM phenylmethyl sulfonyl fluoride (PMSF) and washed extensively in the same solu- tion. The final residue was extracted with 5 volumes of 4mM Tris-HCl, pH 8.5, containing 1 mM EDTA and 0.5 mM PMSF for 30 min at room temperature. The pH of the extract was adjusted to 7.2, and then 1 M MgCl, was added gradually to give 10 mM at final concentration. The solution was stirred on ice for 15min, and centrifuged at 7,000g for 10min. The supernatant was subjected to ammonium sulfate fractionations; The proteins salted out at 28- 78% (0.15-0.55 g/ml) ammonium sulfate were collected and they were dialyzed against 20 mM Tris-HCl, pH 7.6, 20mM NaCl, 0.1 mM EDTA, 15 mM 2-mercaptoethanol, and 0.5mM PMSF. The proteins of about 70-150 kDa were then prepared by preparative SDS-PAGE and used for an immunogen. The monoclonal antibody, AS23, was produced through hybri- 408 Y. Ontsuka, H. NAKAE et al. doma formation between spleen cells of BALB/c mice immunized with the immunogen and nonsereting myeloma cell line P3x63.Ag8U1 cells using the technique of Galfre et al. [8] as modified by Gefter et al. [9]. Hybridoma cells producing antibody were subcloned twice by a limiting dilution method. Supernatants from subcloned cultures were used as the sources of antibodies. The monoclonal antibody to troponin T (NT-302) was prepared as described previously [1]. _Horseradish-peroxidase (HRP)-labeled goat anti-mouse IgG (GAM) was purchased from Bio-Rad (Rich- mond, California) and fluorescein (FITC)-labeled GAM was from Tago (Burlingame, California). Gel electrophoresis and immunoblotting Isoelectric focusing gel electrophoresis (IEF) was carried out according to O’Farrell [15]. SDS-PAGE was performed using 10% polyacrylamide gel in a discontinuous Tris-glycine buffer system as described by Laemmli [10]. Proteins were electrophoretically trans- ferred from SDS-polyacrylamide gel to nitrocellulose filter by the method of Towbin etal. [21]. The filter was treated with 3% gelatin in TBS, 0.5 M NaCl containing 10 mM Tris-HCl! (pH 7.5), and then incubated with AS23 for lhr, followed by the treatment with HRP- GAM for 1 hr. After immunoreaction, the filter was washed with TBS. HRP-GAM bound to the filter was detected as the product of diaminobenzidine reaction with nickel and cobalt ions according to De Blas et al. [3]. Preparation of ascidian native thin filaments Native thin filaments were isolated from ascidian body wall muscle as described by Toyota er al. [22]; briefly, the body wall muscle was homogenized in a KMP buffer (0.05 M KCl-1mM MgClL-10 mM K-phosphate buffer, pH 7.0) and centrifuged at 15,000 xg. Thin filaments in contractile structures were released from the precipitate by a relaxing buffer containing 5mM ATP and 0.1 mM EGTA and collected by ultracentrifugation at 100,000 x g. Indirect immunofluorescence microscopy Pieces of freshly dissected ascidian tissues were immersed in liquid nitrogen-cooled isopentane and transverse cryosections were cut at 10m. The sections were fixed with 4% paraformaldehyde containing phosphate-buffered saline (PBS) and 10mM glycine, washed with PBS containing 10 mM glycine without air-drying, and treated with PBS/2% bovine serum albumin (BSA). They were then exposed to AS23, followed by staining with FITC-labeled GAM. The antibody reaction was performed for 60 min at room temperature. The specimens were mounted in 10 mg/ml p- phenylenediamine-50% _ glycerol-50% PBS, pH8.0, and their fluorescence was observed and photographed under a Zeiss fluorescence/phase contrast microscope. Electron microscopy Isolated thin filaments suspended in a KMP buffer were placed on carbon-coated Formbar grids, treated with 1% BSA in the KMP-buffer and then reacted with AS23, followed by the treatment with 10 nm-gold conjugated GAM in 1% BSA/KMP buffer. The antibody reaction was performed for 30 min at room temperature. The specimens were washed thoroughly with the KMP-buffer after each antibody reaction and stained negatively with 1.5 % uranylace- tate. The specimens were observed under a JEOL JEM 100CX electron microscope at an accelerating voltage of 80 kV. RESULTS We extracted proteins from the smooth muscle of asci- dian body wall by a dilute alkaline salt solution according to the procedure which was originally devised for extracting the proteins from adhesion plaques [6, 7]. Several proteins in the molecular weight range of 70-150 kDa were detected as major components in the extract [12]. In this study, we isolated the proteins around 70-150kDa by preparative SDS-PAGE and used as immunogens to prepare monoclonal antibodies. The culture medium of a hybridoma clone named AS23 recognized a single protein band of about 80 kDa, when the whole lysate of the ascidian body wall muscle was examined by immunoblotting combined with SDS-PAGE (Fig. 1, lanes b&i). The lysates of the other ascidian tissues scarcely reacted with the antibody (Fig. 1, lanes j-n). These results indicate that the antigen recognized by AS23, tenta- tively we call 80K protein, is enriched in the body wall Avthue Dvinin Gom HEL freeones koe lelceat mpd Fic. 1. Immunoblot analysis of the AS23 antigen. Proteins were extracted with 8 M guanidine-HCl from the tissue pieces of ascidian (b-g and i-n) or chicken (h, 0), dialyzed against an SDS-buffer, and then applied to SDS-PAGE. Sample amount was adjusted so that the same wet weight of the tissue was in each lane. The left panel (a-h) denotes the electrophoresis patterns visualized by staining with Coomassie Brilliant Blue. The right panel (i-o) denotes the patterns of immunoblotting stained with AS23. a, molecular weight markers; b and i, body wall muscle; c and j, gut; d and k, stomach; e and I, liver; f and m, branchial sac; g and n, gonad; h and 0, chicken gizzard. An Actin-binding Protein in Ascidian Smooth Muscle 409 muscle. We also examined the extracts of chicken gizzard (Fig. 1, h&o) and mouse smooth muscles with AS23, but none of the protein bands from these tissues were recognized by AS23. To further clarify the protein recognized by AS23, the lysate of body wall muscle was displayed on a two- dimensional gel by a combination of isoelectric focusing and SDS-PAGE, and the protein blotted were reacted with AS23 by an immunoblotting method. AS23 recognized three ma- jor spots of about 80 kDa which differ slightly in pI (Fig. 2). Although it remains to be examined empirically, it seems likely that the spots recognized by AS23 were generated by post-translational modification of the protein, for example phosphorylation or acetylation etc. SDS-PAGE Fic. 2. Immunoblotting combined with two-dimensional PAGE. Proteins were extracted with a lysis buffer for IEF from ascidian body wall muscle and applied for two-dimensional PAGE (IEF/ SDS-PAGE). Left, electrophoresis pattern stained by Coomassie Brilliant Blue; right, the result of immunostaining with AS23. The major spot of the 80K protein is marked by arrowheads. Tissue distribution of the 80K protein was examined by immunocytochemical methods. Ascidian muscle and non- muscle tissues were cryosectioned, treated with AS23, and examined by indirect immunofluorescence methods. The cells in the body wall but not the extracellular matrix were positively stained with AS23 (Fig. 3, b). Almost all of the cells in the body wall were positively stained with AS23 as well as a monoclonal antibody (NT-302) to troponin T, a specific marker for muscle cells (Fig. 3, b&c). The AS23- positive cells were regarded as muscle cells. These cells were brightly stained by rhodamine-phalloidin (Fig. 3, d), indicating enrichment of actin filaments in the cytoplasm. In the immunoblot assay, the 80K protein was scarcely detected in non-muscle tissues, but by the immunocytochemical method, we detected AS23-positive cells among non-muscle tissues. As shown in Figure 3-f, we detected AS23-positive regions in the cryosection of hepatopancreas and associated tissues, although we have not yet been determined what type of tissues or cells they were. They were not stained by NT-302 (Fig. 3, f). Therefore, it is concluded that the 80K protein exists predominantly in muscle cells but its existence is not restricted to muscle cells. The amount in non-muscle tissues appears small. In order to clarify intracellular localization of the 80K protein in ascidian body wall smooth muscle, the muscle was homogenized in a KMP buffer and the homogenate was subjected to subsequent fractionation, and the distribution of the 80K protein in various fractions was examined by im- munoblotting. As shown in Figure 4, we found that the protein was concentrated in the fractions containing thin filaments during the procedure of thin filament isolation; the 80K protein appeared in the supernant at 15,000 g (Fig. 4, b&d), and it was detected more clearly in the isolated thin filaments not only with AS23 but also by staining with Coomassie Brilliant Blue (Fig. 4, c, e&f). These results suggest that the 80K protein may be an actin-binding protein, but its amount in thin filaments was much smaller than the amounts of actin, tropomyosin, and troponin. The 80K protein was detected in the solbule cytoplasmic fraction although small in amount, suggesting that the 80K protein can be free from actin filaments in the cells (data not shown). Since the 80K protein was present in the thin filament fraction, it was matter of interest how the protein is associ- ated with actin filaments. In order to visualize binding of the 80K protein to thin filaments, we applied an immunoelectron microscopic method: isolated thin filaments were treated with AS23 and subsequently with gold-labeled GAM. Gold particles certainly associated with one end of some thin filaments. The thin filaments without the particles were also observed (Fig. 5). DISCUSSION In this investigation, we succeeded in preparing a mono- clonal antibody which recognizes a specific protein of about 80kDa. The antigen, 80K protein, was detected predomi- nantly in body wall smooth muscle cells. Therefore, it is likely that this protein plays some functional role in the ascidian body wall smooth muscle. For further characteriza- tion of the 80K protein, attempts to isolate this protein and its cDNA are now in progress. The monoclonal antibody, AS23, could be an excellent probe for such purpose. Our present data suggest that the 80K protein is an actin-associated protein. Since nothing is known as to reg- ulatory proteins for actin assembly in ascidian body wall smooth muscle, characterization of the 80K protein may provide to the clue to clarify regulation of actin dynamics in the ascidian body wall smooth muscle. When we examined binding of AS23 to isolated thin filaments under an electron microscope, the antibody bound to the end of the filaments. However, many thin filaments without immuno-labeling were also observed. We assume the reasons as follow: 1) Affinity of the 80K protein to actin filaments would be rather weak and the protein might be released from the filaments during preparation of the speci- men, 2) during the preparation of thin filaments, filaments were fragmented and then filament ends without the 80K protein were generated, and 3) procedure for immunostain- 410 Y. Ontsuka, H. NAKAE et al. @ oO 6,7? ., os ¥ ser, ‘gen of Fic. 3. Immunocytochemical localization of the AS23 antigen on sections of ascidian tissues. Cryosections (8 «m) of ascidian body wall muscle (a-d) and hepatopancreas and associated tissues (e-g) were examined. Phase-contrast micrographs (a, e) and corresponding im- munofluorescence micrographs (b-d and f-g) are shown. The sections were stained with AS-23 (b, f), with anti-troponin T antibody (NT-302) (c, g) or with rhodamine-labeled phalloidin (d), respectively, Bar, 100 pm. An Actin-binding Protein in Ascidian Smooth Muscle 411 eaigtly. eCxnG wo ae Fic. 4. Immunoblot analysis of the AS23 antigen in isolated thin filaments. Actin-containing fractions, the supernatant at 15,000 xg (b&d) and the isolated thin filaments (c&e) (see Materials and Methods), were dissolved in an SDS-solution, and applied to SDS-PAGE. a, molecular weight markers; b and c, the electrophoresis patterns visualized by staining with Coomassie Brilliant Blue; d and e, the patterns of immunoblotting stained with AS23. f, an electron micrograph of the isolated thin filaments. Fic. 5. Electron micrographs of ascidian thin filaments which were stained by AS23 and 10 nm-gold conjugated GAM. Location of gold particles were marked by arrowheads. Bar, 0.2 um. ing were not satisfactory. Nevertheless, the results suggest that the 80K protein can bind to the end of ascidian thin filaments, in other words, that the 80K protein might be a capping protein for actin filaments. Although it is too early to discuss about the nature of the 80K protein, as judged by the size, binding to the filament end, and the tissue distribution, the 80K protein might be something like gelsolin in vertebrates. According to a pre- vious report [2], vertebrate gelsolin gives several spots on two dimentional gels which focus in pH range 6.0-6.5, just as the 80K protein. Gelsolin is known to exist in vertebrate smooth muscle [4, 13]. AS23, however, did not recognize any proteins in chicken and mammalian smooth muscles. ACKNOWLEDGMENTS This work was supported by research grants from Ministry of Education, Science and Culture, Yamada Science Foundation, and Futaba Denshi Memorial Foundation. 11 12 13 14 Bar, 0.2 um. REFERENCES Abe H, Komiya T, Obinata T (1986) Expression of multiple troponin T variants in neonatal chicken breast muscle. Dev Biol 118: 42-51 Bader M-F, Trifaro J-M, Langley OK, Thierse D, Aunis D (1986) Secretory cell actin-binding proteins: identification of a gelsolin-like protein in chromaffin cells. J Cell Biol 102: 636- 646 De Blas AL, Chervinski HM (1983) Detection of antigens on nitrocellulose paper immunoblots with monoclonal antibodies. Anal Biochem 133: 214-219 Ebisawa K, Maruyama K, Nonomura Y (1984) Ca** regula- tion of vertebrate smooth muscle thin filaments mediated by an 84K Mr actin-binding protein: purification and characterization of the protein. Biomed Res 6: 161-173 Endo T, Obinata T (1981) Troponin and its components from ascidian smooth muscle. J Biochem. (Tokyo) 89: 1599-1608 Evans RR, Robson RM, Stromer MH (1984) Properties of smooth muscle vinculin. J Biol Chem 259: 3916-3924 Feramisco JR, Burridge K (1980) A rapid purification of @ -actinin, filamin, and a 130,000-dalton protein from smooth muscle. J Biol Chem 255: 1194-1199 Galfre G, Howe SC, Milstein C, Butcher GW, Howard JC (1977) Antibodies to major histocompatibility antigens pro- duced by hybrid cell lines. Nature 266: 550-552 Gefter ML, Margulies DH, Scharff MD (1977) A simple method for polyethylene glycol-promoted fusion of mouse myeloma cells. Somatic cell Genet 3: 231-236 Laemmli U (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680= 685 Meedel TH, Hastings KEM (1993) Striated muscle-type tro- pomyosin in a chordate smooth muscle, ascidian body-wall muscle. J Biol Chem 268: 6755-6764 Nakae H, Sugano M, Ishimori Y, Endo T, Obinata T (1993) Ascidian entactin/nidogen. Implication of evolution by shuf- fling two kinds of cysteine-rich motifs. Eur J Biochem 213: 11- 19 Nodes BR, Shackelford JE, Lebherz HG (1987) Synthesis and secretion of serum gelsolin by smooth muscle tissues. J Biol Chem 262: 5422-5427 Obinata T, Ooi A, Takano-Ohmuro H (1983) Myosin and actin from ascidian smooth muscle and their interaction. Comp Biochem Physiol 76B: 437-442 412 15 16 Y. Outsuka, H. NAKAE et al. O'Farrell, PH (1975) High resolution two-dimensional elec- trophoresis of proteins. J Biol Chem 250: 4007—4021 Shinohara Y, Konishi K (1982) Ultrastructure of the body wall muscle of the ascidian, (Halocynthia roretzi): smooth muscle cell with multiple nuclei. J Exp Zool 221: 137-142 Terakado K (1987) Fine structure of ascidian smooth muscle. Zool Sci 4: 751-761 Terakado K, Obinata T (1987) Structure of multinucleated smooth muscle cells of the ascidian Halocynthia roretzi. Cell Tissue Res 247: 85-94 Terakado K (1988) The pattern of organization of intermediate filaments and their asymmetrical association with dense bodies in 20 21 22 smooth muscle of an ascidian Halocynthia roretzi. Cell Tissue Res 252: 23-32 Terakado K (1989) Pattern of filament arrangement that pro- duces easily their ball-like aggregates in smooth muscle fibers. Zool Sci 6: 1206 Towbin H, Stahlin T, Gordon J (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitro-cellulose sheets: Procedure and some applications. Proc Natl Acad Sci USA 76: 4350-4354 Toyota N, Obinata T, Terakado K (1979) Isolation of tropo- nin-tropomyosin-containing thin filaments from ascidian smooth muscle. Comp Biochem Physiol 62B: 433-441 ZOOLOGICAL SCIENCE 11: 413-421 (1994) Developmental Changes in Pteridine Biosynthesis in the Toad, Bufo vulgaris SHIN-ICHIRO TAKIKAWA and Motroxo NAKAGOSHI Biological Laboratory, Kitasato University, Sagamihara City, Kanagawa 228, Japan ABSTRACT—The presence of blue-violet fluorescent pteridines, neopterin, biopterin, pterin and isoxanthopterin, in a toad tadpole in the various development stages was determined by HPLC with fluorimetric detection. Changes in the activities of pteridine- synthesizing enzymes, GTP cyclohydrolase I, 6-pyruvoyl-tetrahydropterin synthase and sepiapterin reductase, were also investigated. In contrast to an older report by Hama, neopterin was already found in the tadpole larval stage, as well as in the young adult stage. Also, biopterin was found not only in the tadpole larval stage but also in the young adult stage. Neopterin constitutes the major portion of the pteridines of the animal and biopterin makes up the minor one throughout the whole development stage. In spite of the high content level of neopterin in the young adult stage, the activity of GIP cyclohydrolase I, which produces dihydroneopterin triphosphate, declined; and the activity of 6-pyruvoyl-tetrahydropterin synthase, which metabolizes dihydroneopterin triphosphate as a substrate, rose in the same © 1994 Zoological Society of Japan Stage. INTRODUCTION Tailless amphibians contain fairly large amounts of pter- idines as fluorescent compounds in their skin, both in the larval and the adult stages. Hama proposed the name “Rana-chromes” for the fluorescent substances isolated from the skin of Rana nigromaculata as the collective name [15]. Afterwards, Rana-chrome 1 was identified as biopterin [12], Rana-chrome 2 as riboflavin [14], Rana-chrome 3 as 6- hydroxymethylpterin [1, 17], Rana-chrome 4 as isoxantho- pterin [29] and Rana-chrome 5 as pterin-6-carboxylic acid [15], respectively. On the other hand, the other fluorescent substances of the pteridine type were isolated from the skin of the adult toad, Bufo vulgaris, and were named “Bufo- chromes” [18]. Bufo-chrome 2 was identified as pterin [18], and Bufo-chrome itself, which was generally considered specific for Bufo, was proved to be neopterin [8]. The kind and quantity of pteridines differ according to the kind of chromatophore, the age, and the species of the amphibians. At the time of metamorphosis in amphibians, a striking change in the pteridine patterns occurs in some species [1, 13, 16]. In the toad (Bufo vulgaris) neopterin appears at the end of metamorphosis (stage 46) and is very abundant in the adult stage, whereas biopterin is abundant in the larval stage and disappears at the end of metamorphosis [16]. The biosynthesis of pteridines in the skin of the tadpole, Rana catesbeiana, has also been investigated, and guanine [20], guanine nucleotide [32] and GTP [7] were proposed as precursors of pteridines. In all of the above-mentioned investigations, however, pteridines were semi-quantitatively determined by fluorimetric estimation on paper chromatograms and the biosynthetic pathway of pteridines was not elucidated yet at that time. Nowadays, high performance liquid chroma- Accepted May 20, 1994 Received February 9, 1994 tography (HPLC) with fluorimetric detection allows for excel- lent separation and accurate quantitative determination of pteridines [31]. Furthermore, in the past few years, the biosynthetic pathway of tetrahydrobiopterin (BH,) from GTP via dihydroneopterin triphosphate and 6-pyruvoyl- tetrahydropterin (PPH,) has been revealed, as shown in Figure 1 [19, 24, 27, 28, 30, 33-35]. In this pathway, GTP cyclohydrolase I (EC 3.5.4.16), PPH,4 synthase and sepiapter- in reductase (EC 1.1.1.153) are shown to be the indispensable enzymes for the biosynthesis of BHy. The purposes of the present study are to demonstrate the developmental change in the activities of these three pteridine-synthesizing enzymes in relation to the change in the pteridine patterns in Bufo vulgaris and to measure the accurate content of various pteridines by HPLC with fluorimetric detection in several stages of the toad’s development. MATERIALS AND METHODS Animals Fertile eggs of Bufo vulgaris, deposited by a single female, were collected in the suburbs of Tokyo and hatched in our laboratory at 23°C. The tadpoles were fed on boiled spinach. The stages of the development from the fertilized egg through metamorphosis were classified according to Limbaugh and Volpe [21]. Whole bodies of embryos at the prefeeding stage (external gill completed; stage 21, each set of sample composed of 20 individuals) were collected and weighed without dissection. Tadpoles at stage 25 (operculum com- pleted, each set 10 individuals), stage 30 (limb-bud bullet-shaped and about 1 mm in length, each set 5 individuals), stage 40 (hind legs completed, each set 2 individuals), stage 42 (both forelimbs protrude, each set 2 individuals) and stage 46 (metamorphosis complete and tail resorbed, each set 2 individuals) were collected and weighed after the removal of stomach and intestine by dissection. They were stored at —30°C until use. Chemicals Sepiapterin, biopterin, neopterin, pterin and isoxanthopterin were purchased from Dr. B. Schircks Laboratories (Jona, Switzer- 414 S. TAKIKAWA AND M. NAKAGOSHI O N =e » q Ms ~ COOH H.N~ ~N N __O_ CH,0 P-O HN S er, OH }3 gh | Te] OH OH HN N a O HN Guanosine 5'-triphosphate Isoxanthopterin Pterin 6-carboxylic acid GTP-CH | Oo OHOH fo) OHOH H P - ae Ge = CH.2OH HN Nw C- c- CH20 2 | H H OH H H.N* SN~ nq] H Dihydroneopterin triphosphate Neopterin Hc 0 m O N CRCHs Nw HN S ill C H H Ble | Fm H2N N N H NH HN N N H 6-Pyruvoyl-tetrahydropterin 7,8-Dihydropterin Pterin | re) H O re) H Ou4H re) H H C-C-cH NH C_¢-cH N. _¢-G-CH HN S HN nome HN Qc’ Fares a Ce eG sl. = S HN“ SN~ Sn HN* SN~ ny HoN~ SN~ Sw H H H 6-(1'-Hydroxy-2'-oxopropy))- 6-Lactoyl-tetrahydropterin Seplapterin tetrahydropterin = = H Hi at N HC-C-CH, om iS -CH3 fe ie -CH; Ik | [ends ry CY HOH i yen H Tetrahydroblopterin Dihydroblopterin Blopterin Fic. 1. The biosynthetic pathway of tetrahydrobiopterin and degradation process of the related pteridines. Neopterin also degraded to pterin-6-carboxylic acid. Enzymes are shown in synthase; SPR, sepiapterin reductase; AR, aldose round boxes: GTP-CH I, GTP cyclohydrolase I; PPH,S, 6-pyruvoyl-tetrahydropterin reductase; PHO, phosphatase; XO, xanthine oxidase. Pteridine Biosynthesis in the Toad 415 land). Sepiapterin, the substrate for sepiapterin reductase, was purified by column chromatography on phospho-Sephadex [11] and by recrystallization. The other pteridines were used without further purification. Dihydroneopterin triphosphate, the substrate for PPH, synthase, was prepared enzymatically from GTP by the method of Yoshioka et al. [37] using GTP cyclohydrolase I purified from chicken liver [6]. GTP lithium salt, alkaline phosphatase, Trizma base and bovine serum albumin were obtained from Sigma Chemical Co. Sephadex G-25 (fine) was purchased from Pharmacia Fine Chemicals. NADPH was bought from Oriental Yeast Co. Other reagents were of an analytical grade from commercial sources. Sample preparation for pteridine analysis Pteridines in tadpoles are present in their reduced and oxidized forms [9]. Acidic oxidation converts all forms of pteridines to the fully oxidized forms [10]; thus, the total amount of pteridines present is measured in this experiment. Preliminary experiments proved that pteridines were found mainly in the skin of tadpoles, but it was very difficult to separate the skin from other tissues; therefore, whole bodies of tadpoles were used through the experiment. Tadpoles were thawed and homogenized with 4 volumes of 50% ethanol in a tapered tissue grinder with Teflon-pestle (Wheaton, NJ USA) for 2 min, and the homogenate was heated at 70°C for 10 min in a boiling water bath. After cooling, the homogenate was centri- fuged at 10,000 x g for 10 min and the supernatant fluid was collected. Pteridines were extracted once more from the precipitate in the same way. The supernatant from the second extraction was combined with that from the first extraction, and the total volume was meas- ured. A portion of the supernatant (2001) was acidified by adding 401 of 20% trichloroacetic acid (TCA) and oxidized with 401 of iodine solution (1% Ib, 2% KI). After standing for 60 min in the dark, excess iodine was reduced by the addition of 40u1 of 2% ascorbic acid. The mixture was then centrifuged at 10,000 g for 7 min. The oxidized pteridines in the supernatant were determined by HPLC. Sample preparation for enzyme activities All subsequent procedures were carried out at 4°C. Tadpoles were thawed and homogenized with 4 volumes of 25 mM triethanol- amine-HCl buffer (pH 7.4) in a tapered tissue grinder with Tefion- pestle for 2 min. The homogenate was centrifuged at 10,000 x g for 20min. The resultant supernatant was applied to a Sephadex G-25 column (fine, 1 cmi.d. x15 cm) equilibrated with the same buffer as extraction buffer, and then eluted with the same buffer. The elution pattern was monitored by mini UV monitor II (Atto Co.) at the wavelength of 280 nm, and the proteins in the main peak were used as the preparation of enzymes. Protein concentration was deter- mined by the method of Lowry et al. [22] using bovine serum albumin as the standard. Assay for GTP cyclohydrolase I The enzyme activity was measured by a modification of the method reported by Duch ef al. [2]. The reaction mixture was composed of 100mM Tris-HCl buffer (pH7.8), 2mM GTP and 2001 of the Sephadex eluate, in a final volume of 250u1. The mixture was incubated for 60 min at 37°C in the dark. The reaction was terminated by the addition of 251 of a 5:1 mixture of 1% Ih, 2% KI:5N HCl. The mixture was allowed to stand for 20 min in the dark and then excess iodine was reduced by the addition of 25, of 2% ascorbic acid. The reaction mixture was neutralized by the addition of 251 of IN NaOH. The neopterin triphosphate pro- duced in the reaction mixture was dephosphorylated by incubation with 2 units of alkaline phosphatase for 60 min at 37°C in the dark. The reaction was terminated by the addition of 5041 of 20% TCA. Following centrifugation, the amount of neopterin in the supernatant was determined by HPLC. One unit of the enzyme activity was defined as the amount that catalyzes the production of 1 nmole of neopterin under the above conditions. Assay for PPH4 synthase The assay of the enzyme activity is based on the fact that PPH, is decomposed to pterin and pyruvic acid under acidic condition [36]. The enzyme activity was measured by pterin production with a slight modification of the methods reported by Masada et al. [25] and Yoshioka et al. [37]. The reaction mixture was composed of 100 mM Tris-HCl buffer (pH 7.4), 8 mM MgCl, 30”M dihydroneopterin triphosphate and 1171 of the Sephadex eluate, in a final volume of 2001. The mixture of the buffer and enzyme was heated at 80°C for 1 min before use to inactivate heat-labile phosphatases. The reac- tion was started by the addition of the substrate and the reaction mixture was flushed with N> gas and sealed. After incubation at 37°C for 60 min in the dark, the reaction was terminated by adding 40ul of 20% TCA and 401 of iodine solution (1% I,, 2% KI). After standing for 10 min in the dark, excess iodine was reduced by the addition of 4041 of 2% ascorbic acid. The mixture was then centrifuged and the supernatant was subjected to HPLC for estima- tion of pterin. One unit of the enzyme activity was defined as the amount that catalyzes the production of 1 nmole of pterin under the above conditions. Assay for sepiapterin reductase The enzyme activity was measured by a modification of the method reported by Ferre and Naylor [5]. The reaction mixture contained 100mM potassium phosphate buffer (pH 6.4), 100~M NADPH, 50M sepiapterin and 1401 of the Sephadex eluate, in a final volume of 20041. The mixture was incubated at 37°C for 60 min in the dark. The reaction was stopped by adding 401 of 20% TCA and 40] of iodine solution (1% In, 2% KI). After standing for 20 min in the dark, excess iodine was reduced by the addition of 401 of 2% ascorbic acid. The mixture was centrifuged and the fluores- cence of the resulting biopterin in the supernatant was measured by HPLC. One unit of the enzyme activity was defined as the amount that catalyzes the production of 1 nmole of biopterin under the above conditions. Analysis of pteridines by HPLC The HPLC system consisted of an 887 PU pump (Japan Spec- troscopic Co.), an FP-210 spectrofluorometer (Japan Spectroscopic Co.), a D-2500 chromato-integrator (Hitachi Co.) and a 655A-40 auto sampler (Hitachi Co.). An Asahipak GS-320H column (7.6 x 250 mm, Asahikasei Co.) was used for the analyses of pteridines in the tadpoles, for the determination of GTP cyclohydrolase I activity with neopterin detection and for PPH, synthase activity with pterin detection, using 5 mM ammonium acetate as the mobile phase. For the assay of sepiapterin reductase activity with biopterin detection, HPLC was performed on a pre-column (4.650 mm) of ODS-80T), reverse phase (104m, Tosoh Co.) in series with an analytical column (4.6 150 mm) of ODS-80T,, (Sum) using 7% aqueous methanol as the elution solvent. Pteridines were detected by fluorimetry using an excitation wavelength of 360 nm and an emission wavelength of 445 nm at a flow rate of 1.0 ml/min. 416 S. TAKIKAWA AND M. NAKAGOSHI RESULTS Pteridines in the toad tadpoles As shown in Figure 2, blue and violet fluorescent pter- idines in the toad tadpoles were well separated by HPLC on Asahipak GS-320H column; therefore, the pteridines levels could be quantitatively determined using the known amount of the authentic samples of pteridines by fluorimetric detec- tion. Neo r= 7 c eo = oO (Ss) [= (<}) (Ss) ” ® = Ss T IXP Pt Bio 0 5 10 15 20 25 30 Retention Time (min) Fic. 2. An HPLC chromatogram of oxidized pteridines extracted from toad tadpoles at stage 40 with fluorimetric detection. The analysis was performed on an Asahipak GS-320H column (7.6 x 250 mm) with 5 mM ammonium acetate as the mobile phase at a flow rate of 1.0 ml/min. Excitation at 360 nm, emission at 445 nm. Neo: neopterin, Bio: biopterin, Pt: pterin, LXP: isoxan- thopterin. Pteridine levels in individual tadpoles in the various development stages are shown in Figure3. The results of the quantification of pteridines by HPLC indicate that neo- pterin constitutes the major portion of the pteridines of this animal and, in contrast with this finding, that the biopterin level is relatively low throughout metamorphosis. The levels of neopterin, pterin and isoxanthopterin continued to rise gradually in the development process (Fig. 3A, 3C, 3D). On the contrary, changes in the biopterin level during de- velopment displayed a particular pattern. The average value of biopterin reached a minimum at stage 25 and increased gradually during development, reaching a max- imum at mid-climax and then declining slightly at the end of metamorphosis (stage 46) (Fig. 3B). As for the content of pteridines per body weight, the changes in the levels of neopterin, pterin and isoxanthopterin showed the same pattern. They continued to rise gradually in the development process (Fig. 4A, 4C, 4D). In contrast, the biopterin content decreased gradually during develop- ment (Fig. 4B). Activities of pteridine-synthesizing enzymes The changes in the activities of the three pteridine- synthesizing enzymes, GTP cyclohydrolase I, PPH, synthase and sepiapterin reductase, during development are repre- sented in three ways: the total activities contained in an individual tadpole (Fig. 5); the specific activities per mg protein (Fig. 6); and the activities per mg body weight (Fig. De Concerning the changes in GTP cyclohydrolase I activ- ity, the total activity increased gradually in the development process, reaching a maximum at stage 42, and then decreasing suddenly at the end of metamorphosis (stage 46) (Fig. SA). The specific activity of the enzyme tends to decline gradually in the development process and reached a minimum at stage 46 (Fig.6A). The activity of the enzyme per body weight also decreased gradually in the development process and reached a minimum at stage 46 (Fig. 7A). As for changes in PPH, synthase activity, the activity levels were initially low, and then increased gradually in the development process, showing the highest values at the end of metamorphosis. This phenomena can be observed in all three changes in the total activity (Fig. 5B), the specific activity (Fig. 6B) and the activity per body weight (Fig. 7B). The levels of the total activity of sepiapterin reductase were initially low and then increased gradually in the develop- ment process, reaching a maximum at stage 42 and declining at the end of metamorphosis (Fig. SC). On the contrary, the specific activity of the enzyme and the activity per body weight were initially high and rapidly decreased in the development process, then they increased gradually, showing a small peak at stage 42 (Fig. 6C, 7C). DISCUSSION Hama [16] reported the changes in pteridine patterns during the metamorphosis of amphibians. According to his data, the toad, Bufo vulgaris, in the tadpole larval stage contains biopterin and no neopterin, whereas an adult toad at the end of metamorphosis contains neopterin and no biopter- in. However, according to our present data, neopterin was already present in the tadpole larval stage, and the levels of total neopterin in individual tadpoles increased extremely in the development process (Fig. 3A). Also, the neopterin level per body weight increased in the development process (Fig. 4A). Furthermore, our data indicate that the total amount of biopterin in individual tadpoles increased gradual- Pteridine Biosynthesis in the Toad 417 100 0.5 5S 80 S$ 04 2 3 = 2 as] i] & & 2 60 S 03 re) ro) E E & 3 £ & m 20 = 0.2 r=] _ 2. 2. 9 £) = e ) s 0.1 Fe = 0 0 21 25 30 40 42 46 21 25 30 40 42 46 Stages Stages 50 5 o 3 = 3 we =e SI 3] ie} = 2 o = To} re 3 30 — 3 ro) E 2 & 2 © 20 E> 2 5 2 a rs = i) : e 10 2 1 & (& [) = 0 = 0 Stages Stages Fic. 3. The changes in levels of total pteridines contained in individual toad tadpoles in the various development stages. Each point and vertical line represent the mean of 5 determinations and standard error of the mean, respectively. A: Change in neopterin content. B: Change in biopterin content. C: Change in pterin content. ly in the development process, although the level decreased at the end of metamorphosis (stage 46) (Fig. 3B). On the contrary, the biopterin content per body weight decreased gradually in the development process (Fig. 4B). However, the level of biopterin did not fall to zero at the end of metamorphosis in our present experiment. Such a discrep- ancy between our present data and previously reported data by Hama [16] can be attributed to the difference in the respective analytical methods employed. Analysis by HPLC with fluorimetric detection is much more sensitive and accu- rate than that by paper chromatography. In any case, neopterin is a major pteridine and biopterin is a minor one in Bufo vulgaris throughout the whole development stage. The improved analytical method using HPLC revealed the complete absence of pterin-6-carboxylic acid in Bufo vulgaris throughout the whole development stage, while Hama found the compound on the paper chromatogram using the same D: Change in isoxanthopterin content. experimental animal [16]. However, he already supposed that pterin-6-carboxylic acid might be an artifact product [17], the likelihood of which was proved by our present work. The developmental change in the total GTP cyclohydro- lase I activity indicates that the pteridine synthesizing activity decreased rapidly at the end of metamorphosis (Fig. 5A). The decrease in the enzyme activity seems to cause the decrease in the level of various pteridines, because the enzyme catalyzes the first step to synthesize pteridine from GTP (Fig. 1). However, the level of various pteridines, except biopterin, increased gradually in the development process and did not decrease at stage 46 in spite of the decrease in the enzyme activity. Also, in the case of Dro- sophila melanogaster, there is no relationship between the drastic change in GTP cyclohydrolase activity and levels of isoxanthopterin, sepiapterin and drosopterin during develop- ment [4]. The change in biopterin levels during the develop- 418 S. TAKIKAWA AND M. NAKAGOSHI 0.6 z ie) oO S 0.5 > ne) fo} Qa D 0.4 £ & 2 iS 0.3 ie © 5 0.2 (8) £ ® a 0.1 [o} oO z 0 21 25 30 40 42 46 Stages 0.3 ie to) J) = > no) S§ 02 oD) £ 2 ie} £ & 6 0.1 c fe) (@) & = g a C 0 Se 21 25 30 40 42 46 Stages Biopterin Content (nmole/mg body welght) Isoxanthopterin Content (nmole/mg body welght) 0.012 0.010 0.008 0.006 0.004 0.002 0.025 0.020 0.015 0.010 0.005 Stages Fic. 4. The changes in pteridines content per body weight of toad tadpoles in the various development stages. Each point and vertical line represent the mean of 5 determinations and standard error of the mean, respectively. A: Change in neopterin content. B: Change in biopterin content. C: Change in pterin content. D: Change in isoxanthopterin content. ment of D. melanogaster also was not affected by the change in biopterin synthase (it must be just the same as sepiapterin reductase) activity. That is, the level of biopterin increased gradually in the development process in spite of the decrease in the enzyme activity [3]. Such contradictions may be explained by the following. In lower vertebrates and in- vertebrates, pteridines are contained in pigment granules such as pterinosomes in chromatophores [26]. Therefore, pteridines may accumulate continuously in the granules dur- ing development, and they may not be lost from the granules when once they have been accumulated in the granules. However, the decrease in the total amount of biopterin at stage 46 (Fig. 3B) cannot be explained by this idea. This decrease may be attributed to the morphological change the tadpole undergoes as it loses its tail at the end of metamor- phosis. The pteridines contained in the tail are lost [23]. Then the decrease in the total amount of biopterin is marked- ly affected because of its small amount, while the decrease in the levels of other pteridines are not so affected because of their relatively large amounts. As for PPH, synthase, the total activity (Fig. 5B), the specific activity (Fig. 6B) and the activity per body weight (Fig. 7B) were initially low and increased gradually in the development process. The three profiles of these changes resemble one another. At the end of metamorphosis the enzyme activity still increased; therefore, much more of the enzymatic product, PPH4, may be produced. Finally, the product will be changed into biopterin, which will accumu- late, if the sepiapterin reductase activity is high. On the contrary, the sepiapterin reductase activity decreased at this stage; therefore, PPH, will decompose to pterin instead of being catalyzed by sepiapterin reductase (see Fig. 1). Then, the accumulation of pterin at stage 46 can be seen, as shown in Figure 3C. The increase in the enzyme activity at stage Pteridine Biosynthesis in the Toad 419 0.4 0.3 0.2 0.1 Total GTP-CH | Activity (unit/individual) 20 15 10 Total PPH,S Activity (unit/individual) 40 30 20 10 Total SPR Activity (unit/individual) Stages Fic.5. The changes in levels of the total activity of pteridine- synthesizing enzymes contained in individual toad tadpoles in the various development stages. Each point and vertical line represent the mean of 5 determinations and standard error of the mean, respectively. A: Change in GTP cyclohydrolase I activ- ity. B: Change in 6-pyruvoyl-tetrahydropterin synthase activ- ity. C: Change in sepiapterin reductase activity. 0.03 = © i) = o 1o)) = 0.02 2 Cc = = 2 ~_ (5) < 0.01 x= 2 a = 5 A 0 21 25 30 40 42 46 Stages 1.2 =e 10 g fe) = Qo mo 0.8 E 2 Cc = 06 is 2 G gt 0.4 7) + = o 0.2 0) 21 #2 30 40 42 46 Stages 4 SPR Activity (unlit/mg protein) 1s) Stages Fic. 6. The changes in levels of the specific activity of pteridine- synthesizing enzymes contained in toad tadpoles in the various development stages. Each point and vertical line represent the mean of 5 determinations and standard error of the mean, respectively. A: Change in GTP cyclohydrolase I activity. B: Change in 6-pyruvoyl-tetrahydropterin synthase activity. C: Change in sepiapterin reductase activity. 420 S. TAKIKAWA AND M. NAKAGOSHI GTP-CH | Activity (unit/mg body weight) 0.0020 0.0015 0.0010 0.0005 A 0.10 0.08 0.06 0.04 0.02 B 0.4 0.3 0.2 0.1 C 40 42 46 Stages PPH,S Activity (unit/mg body welght) SPR Activity (unit/mg body welght) Fic. 7. The changes in levels of the activity of pteridine-synthesizing enzymes per body weight of toad tadpoles in the various development stages. Each point and vertical line represent the mean of 5 determinations and standard error of the mean, respectively. A: Change in GTP cyclohydrolase I activity. B: Change in 6-pyruvoyl-tetrahydropterin synthase activity. C: Change in sepiapterin reductase activity. 46 also seems to cause the decrease in the concentration of the substrate for the enzyme at this stage by being catalyzed with the enzyme. Thus, a decrease in the concentration of dihydroneopterin triphosphate at stage 46 can be expected; consequently, a decrease in neopterin content can be also expected (see Fig. 1). On the contrary, the neopterin con- tent still increased at this stage (Fig. 3A, 4A). We cannot explain the reason for this interesting result at present. The characteristic feature in the activity of sepiapterin reductase could be seen in an early stage. Although the total activity was very low in this stage (Fig. SC), the specific activity and the activity per body weight were markedly high only in this stage (Fig. 6C, 7C). This may be a reason why biopterin content per body weight is very large in this stage (Fig. 4B). In the present work we investigated the changes in activities of only pteridine-synthesizing enzymes and the correlation between these activities and levels of several pteridines during the development of the toad. However, it would be necessary to investigate also the changes in activities of pteridine-decomposing enzymes, phosphatase which pro- duces neopterin via dihydroneopterin from dihydroneopterin triphosphate, xanthine oxidase which produces isoxantho- pterin from pterin and dihydropterin oxidase reported by Fan and Brown [3]. The comparison of K,, values of every enzyme or investigations into mechanism of positive accu- mulation of pteridines in the pterinosomes would give more accurate information about our experimental results. REFERENCES 1 Bagnara JT, Obika M (1965) Comparative aspects of in- tegumental pteridine distribution among amphibians. Comp Biochem Physiol 15: 33-49 2 Duch DS, Bowers SW, Woolf JH, Nichol CA (1984) Biopterin cofactor biosynthesis: GTP cyclohydrolase, neopterin and bio- pterin in tissues and body fluids of mammalian species. Life Sci 35: 1895-1901 3. Fan CL, Brown GM (1979) Partial purification and some properties of biopterin synthase and dihydropterin oxidase from Drosophila melanogaster. Biochem Genet 17: 351-369 4 Fan CL, Hall LM, Skrinska AJ, Brown GM (1976) Correlation of guanosine triphosphate cyclohydrolase activity and the syn- thesis of pterins in Drosophila melanogaster. Biochem Genet 14: 271-280 5 Ferre J, Naylor EW (1988) Sepiapterin reductase in human amniotic and skin fibroblasts, chorionic villi, and various blood fractions. Clin Chim Acta 271: 271-282 6 Fukushima K, Richter WEJp, Shiota T (1977) Partial purifica- tion of 6-(D-erythro-1’,2’,3’-trihydroxypropyl)-7,8-dihydro- pterin triphosphate synthetase from chicken liver. J Biol Chem 252: 5750-5755 7 Fukushima T (1970) Biosynthesis of pteridines in the tadpole of the bullfrog, Rana catesbeiana. Arch Biochem Biophys 139: 361-369 8 Fukushima T, Akino M (1968) Nuclear magnetic resonance studies of some biologically active dihydropterins. Arch Biochem Biophys 128: 1-5 9 Fukushima T, Kobayashi K, Eto I, Shiota T (1979) A differen- tial microdetermination for the various forms of biopterin. 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Pteridine Biosynthesis in the Toad 42] Anal Biochem 89: 71-79 Fukushima T, Nixon JC (1980) Analysis of reduced forms of biopterin in biological tissues and fluids. Anal Biochem 102: 176-188 Fukushima T, Shiota T (1972) Pterins in human urine. Chem 247: 4549-4556 Goto T, Hama T (1958) Uber die fluoreszierenden Stoffe aus der Haut eines Frosches Rana nigromaculata. 1: Isolierung und Eigenschaften des “Rana-chrom 1”. Proc Jap Acad 34: 724— 729 Giinder I (1955) Quantitative Untersuchung tiber die Entwick- lung der Pterine und des Riboflavins in der Haut und im Auge von Bufo bufo. Z Naturf 10b: 173-177 Hama T (1952) Chemical properties of Rana-chromes. Mag 61: 89 (Abstract, in Japanese) Hama T (1953) Substances fluorescentes du type ptérinique dans la peau ou les yeux de la grenouille (Rana nigromaculata) et leurs transformations photochimiques. Experientia 9: 299-300 Hama T (1963) The relation between the chromatophores and pterin compounds. Ann N Y Acad Sci 100: 977-986 Hama T, Fukuda S (1964) The role of pteridines in pigmenta- tion. In “Pteridine Chemistry” Ed by W Pfleiderer, EC Taylor, Pergamon Press, Oxford, pp 495-505 Hama T, Obika M (1958) On the nature of some fluorescent substances of pterin type in the adult skin of toad, Bufo vulgaris formosus. Experientia 14: 182-184 Heintel D, Ghisla S, Curtius H-Ch, Niederwieser A, Levine RA (1984) Biosynthesis of tetrahydrobiopterin: Possible involve- ment of tetrahydropterin intermediates. Neurochem Int 6: 141-155 Levy CC (1964) Pteridine metabolism in the skin of the tadpole, Rana catesbeiana. J Biol Chem 239: 560-566 Limbaugh BA, Volpe EP (1957) Early development of the gulf coast toad, Bufo valliceps Wiegmann. American Museum Novitates No. 1842, 1-32 Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the folin phenol reagent. J Biol Chem 193: 265-275 Manabe T, Kobayashi K, Horiuchi S (1989) Observation on release of pterins during metamorphosis of bullfrog, Rana catesbeiana. Zool Sci 6: 1139 Masada M, Akino M, Sueoka T, Katoh S (1985) Dyspropterin, an intermediate formed from dihydroneopterin triphosphate in the biosynthetic pathway of tetrahydrobiopterin. Biochim Bio- phys Acta 840: 235-244 Masada M, Matsumoto J, Akino M (1990) Biosynthetic path- ways of pteridines and their association with phenotypic expres- J Biol Zool 26 27 28 29 30 31 32 33 34 35 36 37 sion in vitro in normal and neoplastic pigment cells from goldfish. Pigment Cell Res 3: 61-70 Matsumoto J (1965) Studies on fine structure and cytochemical properties of erythrophores in swordtail, Xiphophorus helleri, with special reference to their pigment granules (pterinosomes). J Cell Biol 27: 493-504 Milstien S, Kaufman S (1983) Tetrahydro-sepiapterin is an intermediate in tetrahydrobiopterin biosynthesis. Biochem Biophys Res Commun 115: 888-893 Milstien S, Kaufman S (1985) Biosynthesis of tetrahydrobio- pterin: Conversion of dihydroneopterin triphosphate to tetrahy- dropterin intermediates. Biochem Biophys Res Commun 128: 1099-1107 Nawa S, Goto M, Matsuura S, Kakizawa H, Hirata Y (1954) Studies on pteridines: VI. The pterin from Bombyx mori (silkworm) and Rana nigromaculata (frog). J Biochem (Tokyo) 41: 657-660 Smith GK, Nichol CA (1984) Two new tetrahydropterin in- termediates in the adrenal medullary de novo biosynthesis of tetrahydrobiopterin. Biochem Biophys Res Commun 120: 761-766 Stea B, Halpern RM, Smith RA (1979) Separation of unconju- gated pteridines by high-pressure cation-exchange liquid chro- matography. J Chromatog 168: 385-393 Sugiura K, Goto M (1968) Biosynthesis of pteridines in the skin of the tadpole, Rana catesbeiana. J Biochem (Tokyo) 64: 657-666 Switchenko AC, Brown GM (1985) The enzymatic conversion of dihydroneopterin triphosphate to tripolyphosphate and 6- pyruvoyl-tetrahydropterin, an intermediate in the biosynthesis of other pterins in Drosophila melanogaster. J Biol Chem 260: 2945-2951 Switchenko AC, Primus JP, Brown GM (1984) Intermediates in the enzymic synthesis of tetrahydrobiopterin in Drosophila melanogaster. Biochem Biophys Res Commun 120: 754-760 Takikawa S, Curtius H-Ch, Redweik U, Leimbacher W, Ghisla S (1986) Biosynthesis of tetrahydrobiopterin: Purification and characterization of 6-pyruvoyl-tetrahydropterin synthase from human liver. Eur J Biochem 161: 295-302 Tanaka K, Akino M, Hagi Y, Doi M, Shiota T (1981) The enzymatic synthesis of sepiapterin by chicken kidney prepara- tions. J Biol Chem 256: 2963-2972 Yoshioka S, Masada M, Yoshida T, Inoue K, Mizokami T, Akino M (1983) Synthesis of biopterin from dihydroneopterin triphosphate by rat tissues. Biochim Biophys Acta 756: 279- 285 ZOOLOGICAL SCIENCE 11: 423-431 (1994) © 1994 Zoological Society of Japan Neural Crest Development in Reptilian Embryos, Studied with Monoclonal Antibody, HNK-1 Linc Hou! and Taxusi TAKEUCHI-* Biological Institute, Faculty of Science, Tohoku University, Aoba-yama, Sendai 980, Japan ABSTRACT—We report here early development of cranial and trunk neural crest, and migratory routes of HNK-1- immunoreactive cells in embryos of a primitive amniote, the softshell turtle. The cranial neural crest cells begin to emigrate shortly after the neural tube closure. It seems that the migratory timing of the cranial neural crest cells in the turtle is different from those of birds and mammals. Presence of two major migratory routes were found in the trunk: (1) a dorsolateral pathway between the epidermal ectoderm and somite; (2) a ventral pathway between the anterior portion of dermomyotome and the sclerotome of each somite. Our results suggest that the trunk neural crest cells have characteristic mode of migration in development of amniotes including the turtle unlike that of teleosts and Xenopus. It is likely that ventrally migrating neural crest cells are the source of pigment cells of some extracutaneous tissues in the turtle embryo. INTRODUCTION Cellular morphogenic movement in embryonic develop- ment of most organisms, particularly that of the vertebrates, is a significant process because it is related to the cytodif- ferentiation and the spatial patterning of the body. The neural crest is an extraordinary embryonic tissue that appeared during the evolution of vertebrates and contributes to a variety of neuronal and nonneuronal structures as a result of extensive migration during embryogenesis [14, 23]. In avian embryos, this transient tissue originates along the dorsal midline of the closing neural tube during neurulation and migrates along characteristic pathways throughout cepha- lic and trunk regions of the embryo to reach appropriate destination of the body, where they give rise to numerous cellular phenotypes and structures such as (1) most of the craniofacial cartilage, skeletal and connective tissues; (2) neurons and supporting cells of the peripheral nervous sys- tems; (3) pigment cells; (4) adrenomedullary cells, etc. [23]. Their highly migratory, multipotent features have long attracted attention of many biologists [14, 23, 38]. Thus, neural crest has been widely used in the investigation of cell migration and differentiation [2, 3, 11, 19, 22, 29, 34, 39]. However, our knowledge concerning the migration and dif- ferentiation of neural crest cells is based on limited informa- tion from a few species, which mostly belong to three groups of vertebrates; amphibians, birds and mammals. Our understanding on the neural crest development in reptiles is " Present Address: Department of Microbiology and Immunology, Indiana University School of Medicine, 635 Barnhill Dr., Indiana- polis, IN 46202, USA 2 Present Address: Nihon Gene Research Laboratories, Inc. 3-11- 18 Tsubamesawa-higashi, Miyagino-ku, Sendai 983, Japan Accepted April 14, 1994 Received Feburary 10, 1994 * To whom all correspondence should be addressed. poor [14] though reptiles have been recognized as materials that provide unique biological information such as tempera- ture-dependent sex determination in embryonic development [7], migration mode of primordial germ cells [16], spatial pattern of neural crest-derived pigment cells [10]. It seems, therefore, significant to investigate the development and migration pattern of neural crest of reptiles as ancestors to both birds and mammals in animal phylogeny. On the other hand, the neural crest-derived pigment cells in the turtle embryos are found in the skin as well as in the various extracutaneous tissues such as dorsal aorta, aorta of kidney, lung and skeletal muscle, etc. [17]. It was also found that the melanoblasts were located corresponding to the ventral route of neural crest cell migration in warm-blooded animals. This leads us to a question regarding the enigmatic migration routes of the melanogenic crest cells in the turtle. In order to visualize migratory neural crest cells in situ, various methods including radioactive labeling, differential chromatin marking and vital dye labeling have been utilized in other species of vertebrates [amphibian: 30; birds: 22, 32, 37; mammals: 32]. However, the study on the neural crest cell migration in reptiles has been difficult because of the lack of specific label or marker that recognizes the nascent crest to follow the migratory route in embryos. In recent years, development of a monoclonal antibody, HNK-1 (NC-1) which recognizes an acidic sulfated glycosphingolipid with five sugars, made it possible to identify migrating neural crest cells in chick and rat, as well as teleost, embryos [2, 5, 11, 29, 31, 36]. We have previously studied the differentiation of neural crest cells in vitro in the turtle embryos and found that the antibody specifically recognizes migrating neural crest cells [18]. Judging from the results mentioned above, this antibody provides a useful tool for the study of neural crest cell migrating of reptiles. In the present study, we investigate development of the turtle neural crest using the monoclonal antibody, HNK-1, 424 L. Hou AnD T. TAKEUCHI and examined the source of the neural crest-derived pigment cells in some extracutaneous tissues. These studies will be an important step towards understanding the developmental patterns of reptile neural crest. The information we obtain can also be compared with that of other species for further understanding the mechanism of cell migration and dif- ferentiation in relation to the evolutionary process of the developmental mechanisms. MATERIALS AND METHODS Incubation of eggs The softshell turtle, Trionyx sinensis japonicus was used in our experiments. Fertilized eggs of the turtle were obtained from Hattori-Nakamura Nursery of Shizuoka, Japan. The method of incubation for this species was previously described [15]. Briefly, the eggs were incubated in laboratory incubator at 33°C, with about 90% relative humidity. Isolation of embryos The softshell turtle embryos were staged according to the criteria of authors [15]; its specific feature of the early development is presented in Table 1. The assignment of each embryo to develop- mental stages was made by isolating it from the eggs. Embryos of stages 6 to 14 were used in our experiment. Histological methods For light microscopic studies, embryos were fixed at the indi- cated stages in Bouin’s fixative (75 parts saturated picric acid, 25 parts formalin, 5 parts glacial acetic acid) at room temperature for 24 hr, then transferred to 70% ethanol. The embryos were dedhydrated, embedded in paraffin (Merk, Germany) and sectioned serially at 7-8 ym thickness. Sections were stained with hematoxylin and eosin. Immunohistochemical methods The monoclonal antibody, HNK-1 (Beckton-Dickinson, CD57, IgMx), which ha been shown to recognize the surface epitope of neural crest cells of some vertebrates [31, 36] was used for this immunostaining: The turtle embryos of various stages were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS, pH 7.4) either at 4°C for overnight or at room temperature for 3-6 hr, and then washed in several changes of PBS. The fixed embryos were dehydrated, embedded in paraffin and serially sectioned at 10 «m. The sections were placed on neoprene-coated microslide glasses. They were deparaffinized in xylene, then dehydrated through a graded series of ethanol, and rinsed in PBS. Before staining, specimens were treated with PBS-3% bovine serum albumin (BSA) and were then incubated with the monoclonal antibody at 1:50 dilution in PBS-0.5% BSA for overnight at 4°C. After several washes with PBS, they were incubated for 1 hr with rabbit anti-mouse IgM antibody (RAM), followed by several washes in PBS. They were then stained with FITC-conjugated goat anti-rabbit antibody (FITC-GAR, Cappel) for 1 hr. The sections were washed in PBS and mounted in PBS containing 2.5% diazobicyclooctane (DABCO). In control experiments, mouse IgM protein (Zymed) was used as the first antibody. No specific staining was found in the control series at any embryonic stage. For whole mount staining, embryos of stages 11 and 12 were for the sections, except the time and concentration of antibody reaction. The embryos were incubated with the antibody HNK-1 at 1:25 dilution in PBS-0.5% BSA for 2-3 days at 4°C, then incubated for 2 days with RAM, and stained with FITC-GAR for 2 days. The specimens were examined and photographed with a Olympus epifluorescence microscope, BH2-RFL. RESULTS Early development of neural crest In embryos of stage 6, the edges of the neural plate moved fo form the neural folds, while the neural groove appeared in the anterior region of embryo. At stage 7, the neural groove closed to form the neural tube in the head region; the enclosure progressed toward the anterior region of trunk. The presumptive neural crest was observed on the dorsal wall of the neural tube of head region (Fig. 1A); they are located at junction between the ectoderm and the neural tube in the anterior region of embryo (Fig. 1B). At stage 8, neural crest was clearly segregated from the neural tube and appeared at the cephalic region of embryos. A mass of crest cells covered the neural tube of presumptive mesencephalon and rhombencephalon region (Fig. 1C). At stage 9, neural crest cells detached from neural tube and located at the space TasBLe 1. Developmental features in the early stages of embryos of the turtle, Trionyx sinensis japonicus Stage Age Bee Specific features 5 15-16 hr 0 Notochord and neural plate appear. 6 24 hr 1 Neural groove formation; amnion and blood islands appear. v 33-35 hr 4-5 Neural tube formation; primordium of heart appears. 42-43 hr 6-7 Formation of mesencephalon and rhombencephalon; prim- ary optic vesicles appear. 9 48-56 hr 9-11 Otic pits appear. 10 64-65 hr 13-14 Otic vesicle formation; amnion extends to blastopore region. 11 72-74 hr 16-18 Lens formation; heart beat starts. 12 96 hr 20-21 Amnion closed; nasal placode appears. 13 5 day 27-28 Limb-buds appear. 14 6 day Pigment cells of retina appear. i Reptile Neural Crest Development 42 Wn a NEO a Oe! CL ow reas Fic. 1. Cross sections of the neural tube of the turtle embryos. A, stage 7. Neural crest population located at the dorsal border of the neural tube of head region. B, Stage 7. Presumptive neural crest appears at the junction between the ectoderm and the neural tube. C, Stage 8. A mass of crest cells cover the dorsum of the mesencephalon and rhombencephalon. D, Stage 9. The front line of crest cell population moves the area between the ectoderm and the mesoderm (arrow). E, Stage 9. A cross-section through first somite. Crest cells are found between the ectoderm and the somite. F, Stage 9. A cross section through the 7th somite. Crest cells are located laterally to the neural tube. Ec, ectoderm; En, endoderm; N, notochord; NT, neural tube; Ot, otic placode; S, somite. Arrowhead indicate neural crest cells. Scale bar=100 ~m. between the ectoderm and mesoderm in the cephalic region (Fig. 1D, E), while individual neural crest cells were found on the neural tube of trunk region at this stage; most of them were located between the ectoderm and the neural tube prior to the beginning of their migration (Fig. 1F). The onset of migration of the neural crest follows an anterior-to-posterior order. At these stages, however, cells identified as indi- vidual neural crest cells by morphological observation did not react to the antibody, HNK-1, except a few positive cells. Cephalic crest cell migration The earliest cells to emigrate from the mesencephalon appeared at stage 8 towards stage 9. By advanced stage 10, HNK-1-immunoreactive crest cells have separated from the neural tube and dorsolaterally moved under the epidermis (Fig. 2A), being localized laterally to the optic vesicles. Numerous migrating crest cells were also found ventrally to the space between the optic vesicle and prosencephalon, and the area surrouding the developing eye (Fig. 2B). At more rostral level, some crest cells from the mesencephalon moved cranially over the optic vesicle to the cell free space between the epidermis and prosencephalon (Fig. 2C). The HNK-1- immunoreactive cells can be detected in the vicinity of the nasal placode at stage 12. At later stage, the mesencephalic L. Hou Anpb T. TaKEUCHI Fic. 2. Migration of HNK-1-immunoreactive crest cell in cephalic region. A, Stage 10. Crest cells are seen in the mesencephalon region dorsolaterally migrating along the epidermis. B, Stage 11. Crest cells are ventrally located at the space between optic vesicle and prosencephalon. C, Stage 11. Sole population of crest cells distributes between the epidermis and prosencephalon. D, Stage 10. Crest cells migrating dorsolaterally under the epidermis and ventrally toward the bottom of the anterior rhombencephalon (arrow). _E, Crest cells Reptile Neural Crest Development 427 crest cells will give rise to mesenchymal components to contribute to the maxillary and nasal processes as previously described [13]; some of them also differentiate into melano- cytes at the choroid and nasal connective tissue (Data not shown). At stage 10, crest cells arose out of anterior rhom- bencephalon (anterior to the otic placode) and continued to migrate dorsolaterally along the ectoderm, while some of the cells ventrally along the tube of anterior rhombencephalon (Fig.2D). Two distinct streams of migration of rhom- bencephalic neural crest were observed at this region. First, front line of the crest cell polulation moved under the epidermis and reached the pharynx; they intercede into second visceral (hyoid) arch. Second, other crest cells move toward the bottom region of rhombencephalon and contri- bute to the trigeminal ganglion formation at stage 11 of Fic. 3. Detection of HNK-1-immunoreactive cell in trunk. A, A cross section through the anterior half of the 8th somite at stage 12. Crest cells pass a space between the dermomyotome and the sclerotome and move toward the dorsal aorta. B, A cross section through the posterior half of the somite. Crest cells are located between the neural tube and the somite, but do not migrate into the posterior sclerotome. n, notochord; nt, neural tube; da, dorsal aorta. Scale bar=100 um. embryos. The crest cell migration was seen at the region of posterior rhombencephalon (hind brain after the optic pla- code) at this stage. The foremost cells of posterior rhom- bencephalic crest population moved along the area between the ectoderm and mesoderm, and already reached the pharynx (Fig. 2E, F). The crest cells in cephalic region were stained intensely probably reflecting the fact that the density of the crest cells was high. At the level of the first visceral arch, a population of crest cells have migrated laterally between the ectoderm and mesoderm. They moved into the developing first visceral arch and into the area surrounding the first aortic arch in the embryo of stage 10 (Fig. 2G). They seem to provide the Fic. 4. HNK-1-immunoreactive cells migrating along somites (6th to 9th). Longitudinal sections of a stage 11 embryo are shown from top (a) to bottom (c). a, A section below the dorsal surface of the neural tube and the somite. Crest cells are uniformly distributed as a continuous longitudinal population at the space between the neural tube and the somite. b, A. section 10 ~m ventral to (a). c, 20 um ventarl to (b). Neural crest cells are only observed in the anterior half of each somite. (d) and (e) are enlarged from (a) and (c). A, anterior; NT, neural tube; P, posterior; S, somite. Scale bar=100 um. located under the epidermis at the level of posterior region of the anterior rhombencephalon. F, Crest cells at posterior rhombencephalon. G, A cross-section through 1st visceral arch. Crest cells move into the 1st visceral arch (arrow). H, Stage 12. A cross section through the 3rd and 4th visceral arches, crest cells are located in the 3rd and 45th visceral arch (arrow). I, Stage 12. Crest cells are mostly detected in the 4th visceral arch (arrow). ao, aorta; m, mesencephalon; n, notochord; OpV, optic vesicle; OtV, otic vesicle; p, prosencephalon; ph, pharynx; rh, rhombencephalon. Scale bar=100 um. 428 L. Hou anpb T. TAKEUCHI mesenchymal components to contribute the mandibular pocesses and morphological differentiation of Meckel’s cartil- age in 12-day-old embryos (data not shown). By the stage 12, numerous crest cells from posterior rhombencephalon were found in third and fourth visceral arches (Fig. 2H, I). In cephalic region, the characteristics of neural crest cell are to migrate as a coherent sheet of cells in contrast with their migration in trunk region. Trunk neural crest migration The neural crest cell migration mostly occurred after morphological differentiation of the somite into the der- momyotome and sclerotome in the trunk of embryos. At stage 10, HNK-1-immunoreactive crest cells seem to migrate into the somite; numerous migrating crest cells were observed in the somite at stage 11 and 12. These cells are usually located at the anterior portion of each somite, where they seem to migrate between the anterior dermomyotome and sclerotome (Fig.3A). Very few cells were found in the posterior part of the somites (Fig. 3B). Fig. 4 illustrates the distribution of neural crest cells in longitudinal section through the trunk. The migration of trunk neural crest cells in the somite was always restricted to the anterior half of each sclerotome (Fig. 4C, E) in spite of the fact that there is no morphological difference between the anterior and posterior half of each somite. It seems that their migration is suppres- sed in the posterior half of each sclerotome, but is stimulated in the anterior half. At stage 11, neural crest cells were detected dorsolateral- ly between the ectoderm and the somite in the trunk (Fig. 5A). In general, they seem to migrate as individual cells. HNK-1-immunoreactive crest cells located in the area be- tween the ectoderm and dermomyotome are illustated in Fig. 5B. It is interesting to note that the neural crest cells in the developing somite are often located in the medial space of the dermomyotome of the anterior half of somite and within the myotome (Fig. 5C,D). Insome cases, we observed HNK-1- immunoreactive cells migrating into the dermomyotome. Thus, our results indicate the presence of two major pathways of their migration in this region; (1) a dorsolateral pathway between the epidermal ectoderm and somite; (2) a ventral pathway between the anterior portion of the der- momyotome and sclerotome of each somite. No crest cell was observed migrating around the notochord area. In addition, we found that numerous HNK-1- immunoreactive crest cells passing through the dorsal mesen- Fic. 5. Fluorescence micrographs of neural crest cell migrating in trunk region of a stage 11 embryo. A, A cross section through the 7th somite of trunk. Several crest cells migrate into the area between the ectoderm and the dermomyotome (arrow). Positive cells are also detected in the area between the dermomyotome and the sclerotome. b, A cross section through the 8th somite of trunk. Crest cell is located between the epidermis and the dermomyotome (arrow). Numerous crest cells migrate along ventral route at the same time. C, A cross section through the anterior somite of posterior region of trunk at stage 11. Crest cells intercalate into the medial space of the dermomyotome (arrow). D, Crest cells localized within the myotome at stage 14. d, dermomyotome; ec, ectoderm; m, myotome; n, notochord; nt, neural tube. Scale bar=100 «xm. Reptile Neural Crest Development 429 Fic. 6. HNK-1-immunoreactive creast cells distributed in the de- veloping lung and foregut at stage 14 embryo. A, A cross section through the level of oesophagus (oe) and lung bud (Ig). Crest cells are seen in the mesenchyme of the vicinity of lung bud. B, A longitudinal section through the level of foregut, the crest cells are observed between the endoderm and mesoderm of foregut. en, endoderm of the gut; 1, liver. Scale bar=100 yan. tery at stage 12. They were found at stage 13 migrating into the lung buds. By stage 14, a number of crest cells were observed in the lung buds and foregut (Fig. 6A, B). DISCUSSION To our knowledge, there are only a few reports on the distribution of cranial neural crest cells in reptiles [12, 25]. The principal contribution of the present investigation was to demonstrate precise migratory routes of cranial and trunk neural crest cells in the reptilian embryo by using monoclonal antibody, HNK-1. We found that the mode of neural crest cell migration is different according to various cephalic re- gions and that the cranial neural crest cells seem to follow pathways similar to those of other vertebrate species studied previously [chick: 8, 26; Xenopus: 30; teleost: 31]. Howev- er, it has been noticed that the time of onset of cranial neural crest migration varies in non-reptile vertebrate groups [13, 14]. In mammals, Tan and Morriss-Kay [35] reported that the cranial neural crest migration already began when the neural tube is still open at neural fold stage. In birds, it has been shown that quail neural crest cells start their migration immediately before the complete closure of the neural tube [23]. On the other hand, the neural crest cell migration generally corresponds with the closure of the neural tube in amphibia [14]. In the turtle, no distinguishable neural crest cell was found when the neural tube is closed in the cephalic region. At advanced stage 8, individual neural crest cells segregated from the tube of mesencephalon and rhom- bencephalon, and begin their migration after aggregating once on the neural tube. Our results indicate that the cranial neural crest of the turtle has characteristic migration pattern and timing different from birds and mammals. We do not know the underlying mechanism responsible for the above-mentioned difference. Recently, however, it is in- creasingly accepted that the ECM (extracellular matrix) molecules may play an important role in determing the direction of neural crest cell migration and/or that ECMs provide permissive migratory substrates for their migration [28, 38]. The difference in the migratory timing of cranial neural crest cells among different vertebrate groups is likely due to the difference in environmental factors such as ECM and other unknown factors. The migratory pathway of trunk neural crest cells has not been reported previously for reptiles. So far, the studies of the migration of trunk neural crest cells in non-reptile verte- brates have shown the following two major routes: (1) a dorsolateral pathway between the ectoderm and somite; (2) a ventral pathway between the neural tube and somite. It is noteworthy that a ventral pathway of migrating neural crest cells in somewhat different between anamniotes such as teleosts or Xenopus and aminotes such as birds or mammals. In birds and mammals, neural crest cell along the ventral pathway passes through the area between the anterior der- momyotome and the sclerotome [2, 11, 29]. The trunk neural crest cells in teleosts and Xenopus embryos migrate along the ventral pathway passing near the notochord and do not migrate into the somite [30, 31]. In our study, neural crest cells in the trunk were found to migrate along ventral pathway, which is similar to that of birds and mammals unlike that of teleosts and Xenopus. The difference in the antero- posterior pattern of ventral crest cell migration might be attributable to the somite tissue [4]. The ventral migrating pathway through the somite seems to be evolutionary con- served as principal character of neural crest development of amniotes. The pattern of neural crest migration along the somite is likely to be regulated by the similar nature of developmental and molecular mechanisms among reptiles and other aminotes [11]. It has been assumed that migratory pathways of neural crest cells are related to their differentiated cellular phenoty- pes, for example, a dorsolateral pathway is the migrating route of non-neuronal cell type including pigment cells in other vertebrate groups [fish: 31; amphibia: 27, 30; birds: 2, 20, 32; mammals: 33], whereas neural crest cells migrating along a ventral pathway will contribute to various neuronal phenotypes in the chick and mouse [20, 23, 24, 32]. We have found that, in the turtle embryos, pigment cells are 430 L. Hou anp T. TAKEUCHI located in the skin as well as in the various extracutaneous tissues such as skeletal muscle, dorsal aorta, aorta of kidney and lung [17]. Question arises as to the migration route of these extra- cutaneous pigment cells. The neural crest cells have been found along ventral route migrating into the myotome, on the dorsal aorta and in the lung buds. In addition, presence of some melanogenic neural crest cells located at the sites corresponding to the ventral pathway seems to indicate that pigment cells of some extracutaneous tissues are derived from neural crest cells migrating along the ventral pathway [17]. This assumption was verified by another series of experiments involving the culture of isolated somite from stage 12 embryo; the results suggest that the melanogenic crest cells along the ventral route are associated with the somite (unpublished data). For the migration of neural crest-derived pigment cells in the trunk of the turtle embryos three possible routes can be suggested: a) They migrate along a dorsolateral pathway and give rise to pigment cells of the skin; b) they migrate along ventral pathway, while some crest cells intercalate into the myotome during midway of migration and differentiate into pigment cells of skeletal muscle; c) other precursor cells continue to migrate along ventral pathway toward the dorsal aorta and the developing kidney, where they finally differentiate into pigment cells. It has been known that the pigment cells are not normal- ly localized in the above-mentioned sites in the white leghorn, the quail or the mouse and that the neural crest cell migrating along a ventral pathway contributes to the dorsal root ganglia (DRG) and sympathetic ganglia (SG) [1, 20, 21]. However, the cells of DRG and SG are shown to differentiate in vitro into several cell types including pigment cells [6, 9, 35]. The cells of DRG and SG seem to lose the potency of differentia- tion into pigment cells at later stages of development [6, 9]. It is conceivable that crest cells migrating along ventral pathway are basically pluripotent in amniotes and that the loss of differentiation potency toward melanogenesis is reg- ulated under the tissue environment in the embryo. These observations lead us to an assumption that species-specific environmental factor(s) responsible for the final differentia- tion of neural crest cells is present in this region. ACKNOWLEDGMENTS The authors thank Hattori-Nakamura Nursery of Shizuoka for providing us with the turtle eggs used in our study. This research was supported by Grant-in-aid from the Ministry of Education, Science and Culture to T. T., and Japanese Government Research Scholorship from the Ministry of Education, Science and Culture to Pe REFERENCES 1 Bronner ME, Cohen AM (1979) Migratory patterns of cloned neural crest melanocytes injected into host chicken embryos. Proc Natl Acad Sci USA 76: 1843-1847 a 10 11 12 13 15 16 20 Bronner-Fraser M (1986) Analysis of the early stages of trunk neural crest migration in avian embryos using monoclonal anti- body HNK-1. Dev! Biol 115: 44-55 Bronner-Fraser M, Fraser S (1988) Cell lineage analysis re- veals multipotency of some avian neural crest cells. Nature 335: 161-164 Bronner-Fraser M, Stern C (1991) Effects of mesodermal tissues on avian neural crest cell migration. Devl Biol 143: 213-217 Chou KH, Ilyas AA, Evans JE, Quarles RH, Jungalwala FB (1985) Structure of a glycolipid reacting with monoclonal IgM in neuropathy and with HNK-1. Biochem Biophys. Res Comm 128: 383-388 Ciment G, Glimelius B, Nelson D, Weston JA (1986) Reversal of a developmental restriction in neural crest-derived cells of avian embryos by phorbol ester drug. Devi Biol 118: 392-398 Deeming DC, Ferguson MWJ (1988) Environmental regula- tion of sex determination in reptiles. Phil Trans R Soc Lond B 322: 19-39 Duband JL, Thiery JP (1982) Distribution of fibronectin in the early phase of avian cephalic neural crest cell migration. Devl Biol 93: 308-323 Duff RS, Langtimm CJ, Richardson MK, Sieber-Blum M (1991) In vitro clonal analysis of progenitor cell patterns in dorsal root and sympathetic ganglia of the quail embryo. Devl Biol 147: 451-459 Buncker HR (1985) The neural crest. In “Pigment Cell, Vol 7” Ed by JT Bagnara, S. Klaus, E Paul, M Schartl, Univ Tokyo Pres, Tokyo, pp 255-269 Erickson CA, Loring JF, Lester SM (1989) Migratory path- ways of HNK-1-immunoreactive neural crest cells in the rat embryo. Devl Biol 134: 112-118 Ferguson MWJ (1985) Reproductive biology and embryology of the crocodilians. In “Biology of the Reptiles, Vol. 14” Ed by C Gans, F Billett, PF Maderson, John Wiley & Sons, New York, pp 329-492 Hall BK (1987) Tissue interaction in the development and evolution of the vertebrate head. In “Development and Evolu- tionary Aspects of the Neural Crest” Ed by PFA Maderson, John Wiley & Sons, New York, pp 215-259 Hall BK (1988) THe Neural Crest, Oxford Univ Press, London Hou L (1984) Studies on the embryonic development of the turtle, Trionyx sinensis. Nat Sci J Hunan Normal Univ 2: 59- 70 Hou L (1987) Cytogenesis of the primordial germ cells in the turtle, Trionyx sinensis. Acta Herp Sin 6: 5-10 Hou L, Takeuchi T (1991) Differentiation of extracutaneous melanocytes in embryos of the turtle (Trionyx sinensis japoni- cus). Pigment Cell Res 4: 158-162 Hou L, Takeuchi T (1992) Differentiation of reptilian neural crest cells in vitro. In Vitro Cell Dev Biol 28A: 348-354 Ito K, Takeuchi T (1984) The differentiation in vitro of the neural crest of mouse embryo. J Embryol exp Morphol 84: 49- 62 Kitamura K, Takiguchi-Hayashi K, Sezaki M, Yamamoto H, Takeuchi T (1992) Avian neural crest cells express a mela- nogenic trait during early migration from the neural tube: observation with the new monoclonal antibody, “MEBL-1”. Development 114: 367-378 Klein-Szanto A, Bradl M. Porter S, Mintz B (1991) Melanosis and associated tumors in transgenic mice. Proc Natl Acad Sci USA 88: 169-173 Le Douarin NM (1973) A biological cell labeling technique and its use in experimental embryology. Devl Biol 30: 217-222 Le Douarin NM (1982) The neural crest. Cambridge Univ 24 25 26 27 28 29 30 31 Reptile Neural Crest Development 431 Press, Cambridge Le Douarin NM (1986) Cell line segregation during peripheral nervous system ontogeny. Science 231: 1515-1522 Meier S, Packard DS Jr (1984) Morphogenesis of the cranial segments and distribution of neural crest in the embryos of the snapping turtle, Chelydra serpentina. Devl Biol 102: 309-323 Noden DM (1975) An analysis of the migratory behavior of avian cephalic neural crest cells. Dev! Biol 42: 106-130 Perris R, Von Boxberg Y, Lofberg J (1988) Local embryonic matrices determine region-specific phenotypes in neural crest cells. Science 241: 86-89 Perris R, Bronner-Fraser M (1989) Recent advences in de- fining the role of the extracellular matrix in neural crest develop- ment. Comm Develop Neurobiol 1: 61-83 Rickmann M, Fawcett JW, Keynes RJ (1985) The migration of neural crest cells and the growth of motor axons through the rostral half of the chick somite. J Embryol exp Morphol 90: 437-455 Sadaghiani B, Thiebaud CH (1987) Neural crest development in the Xenopus laevis embryos, studies by interspecific trans- plantation and scanning electron microscopy. Devl Biol 124: 91-110 Sadaghiani B, Vielkind JR (1990) Distribution and migration 32 33 34 35 36 37 38 pathways of HNK-1-immunoreactive neural crest cells in teleost fish embryos. Development 110: 197-209 Serbedzija GN, Bronner-Fraser M, Fraser SE (1989) A Vital dye analysis of the timing and pathways of avian neural crest cell migration. Development 106: 809-819 Serbedzija GN, Fraser SE, Bronner-Fraser M (1990) Pathways of trunk neural crest cell migration in the mouse embryos as revealed by vital dye labeling. Development 108: 605-612 Sieber-Blum M (1989) Commitment of neural crest cells to the sensory neuron. Science 243: 1608-1611 Tan SS, Morriss-Kay GM (1985) The development and dis- tribution of the cranial neural crest in the rat embryo. Cell Tissue Res 240: 403-416 Vincent M, Thiery JP (1984) A cell surface marker for neural crest and placodal cells: Further evolution in peripheral and central nervous system. Devl Biol 103: 468-481 Weston JA (1963) An radioautographic analysis of the migra- tion and localization of trunk neural crest cells in the chick. Devl Biol 6: 279-310 Weston JA (1983) Regulation of neural crest cell migration and differentiation. In “Cell Interaction and Development” Ed by KM Yamada, John Wiley & Sons, New York, pp 153-184 ZOOLOGICAL SCIENCE 11: 433-439 (1994) © 1994 Zoological Society of Japan The Second Maturation Division and Fertilization in the Spider Achaearanea japonica (Bos. et Str.) Hironumi Suzukxt and Axio Konpo Department of Biology, Faculty of Science, Toho University, 2-1, Miyama 2 chome, Funabashi-shi, Chiba 274, Japan ABSTRACT— Formation of the main layer of the vitelline membrane, the second maturation division and fertilization were examined in the spider Achaearanea japonica by light and electron microscopy. Newly laid eggs had already accepted a sperm nucleus near the periphery. The sperm nucleus was observed as a mass of condensed chromatin, which was not enclosed by a nuclear envelope. Bundles of microtubules, which separated the large yolk granules that surrounded the perinuclear cytoplasm into radial arrays immediately after oviposition, seem likely to play an important role in the migration of the sperm nucleus together with perinuclear cytoplasm to the center of the egg. The matrix that was released from the vesicles, by exocytosis, into the space between the cell membrane and the outer layer of the vitelline membrane appeared to form the main layer of the vitelline membrane. The second maturation division of the first polar body progressed to telophase, but daughter nuclei were not formed. The maturation division of the secondary oocyte was strictly synchronized with respect to that of first polar body, and it generated the female pronucleus and the nucleus of the second polar body. The female pronucleus moved toward the egg’s center, in which the male pronucleus was now located, for conjugation. The larger pronucleus was probably the male pronucleus, and the other one was probably the female pronucleus. INTRODUCTION Embryos of many spiders have been examined in detail under the light microscope. However, very little informa- tion is available about the maturation and fertilization of the egg. We have previously examined the fine-structural changes at the egg’s surface during the first maturation division in the spider Achaearanea japonica (unpublished data). The cell membrane begins to invaginate from the egg’s surface, and then the lumen of each invagination becomes wider, forcing the large yolk granules into outer radial columns and an inner spherical mass (see Fig. 5). The matrix that is discharged at the egg’s surface by secretion granules appears to form the outer layer of the vitelline membrane. In this report, we describe the formation of the main layer of the vitelline membrane, the second maturation division and fertilization in A. japonica. MATERIALS AND METHODS Mature females of Achaearanea japonica (Bés. et Str.) lay eggs 5-6 times in mid- and late summer, and about 100 eggs are released at each oviposition. The eggs are spherical and 0.5 mm in diameter. Eggs collected on the campus of Toho University were used for the present study. They were incubated at 25°C. Routine light microscopy was carried out using paraffin- embedded serial sections. Eggs fixed in FAA (formalin, ethyl alcohol and acetic acid, 5:15:1, v/v) were dehydrated in a graded alcohol series, cleared in toluene and embedded in paraffin. The samples were cut off by 30 ~m with a microtome, left in distilled Accepted June 6, 1994 Received March 28, 1994 ' To whom correspondence should be addressed. water for 3 days and then sectioned at 6- to 8- um thickness. For more detailed light-microscopic observations, eggs were fixed for the most part in a mixture of 2.5% glutaraldehyde and 2% paraformal- dehyde in 0.1M phosphate buffer, pH 7.4, that contained 0.2M sucrose, and they were punctured during fixation with a tungsten needle. Dehydrated samples were embedded in methacrylate resin Technovit 7100 (Kulzer). The resin-embedded specimens were sectioned at 1- to 5- wm thickness with a glass knife on a LKB-4800 ultramicrotome. The sections were stained with Mayer’s acid- haemalum and eosin. For fine-structural observations, the eggs were prefixed, at room temperature, for 3 hr in a mixture of 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, that contained 0.2 Msucrose. During fixation, the eggs were cut in half with a tungsten needle. After rinsing for more than one hour with the same buffer plus 0.2 M sucrose, the samples were postfixed, at room temperature, for one hour in 2% osmic acid in 0.1 M phosphate buffer, pH 7.4, without sucrose. After rinsing with the same buffer without su- crose, samples were dehydrated in a graded alcohol series, transfer- red to propylene oxide, and embedded in epoxy resin Quetol 812 (Nisshin EM). Ultrathin sections were cut with a diamond knife on the ultramicrotome, stained with uranyl acetate and lead citrate, and examined under a JEOL JEM-1210 or a Hitachi HU-12A electron microscope. Thick sections were prepared simultaneously, and these sections were stained with toluidine blue for light microscopy. They were reembedded in the same epoxy resin for electron micros- copy as needed. RESULTS Just at the time when oviposition occurred, the eggs had already accepted a sperm nucleus near the periphery, about 50 4m from the egg’s surface (Fig. 1). No spermatozoa were observed at all. The cytoplasm was distributed at the egg’s surface as periplasm, 10-20 um in thickness, and in the H. Suzuki AND A. Konpo Fics. 1-4. Technovit-embedded sections. Scale bar=20 um. Fig. 1. Just at the time of oviposition. The egg has already accepted a sperm nucleus (arrow). The cytoplasm is distributed at the egg’s surface as periplasm (pp) and in the perinuclear zone. Fig. 2. Twenty minutes after oviposition. The large yolk granules surrounding the perinuclear cytoplasm of the sperm nucleus are organized in radial arrays. Fig. 3a. Forty-five minutes after oviposition. The large yolk granules are arranged in outer radial columns and an inner spherical mass (compare with Fig. 5). The maturation division of the secondary oocyte at metaphase (arrowhead) is seen at the surface of the inner spherical yolk mass. The equatorial plane is perpendicular to the egg’s surface. ch, chorion. Fig. 3b. Same egg as in Fig. 3a. The second maturation division of the first polar body at metaphase (arrowhead) is seen in the cytoplasm located among the outer radial columns of yolk granules. The equatorial plane is parallel to the egg’s surface. Fig. 4. A different egg, 45 min after oviposition. The first polar body at metaphase during the second maturation division (arrowhead) is seen in the periplasm. The equatorial plane is perpendicular to the egg’s surface. Meiosis and Fertilization in Spider 435 aon Hf » : ae TRS: Bron,” i PP Ro Oka ———en 34? 5 Gs ee IS. VO SAO, 5 Fic. 8. Electron micrograph showing the sperm nucleus. envelope. A cross section of axial filament of the spermatozoon (arrow) and fragments of the nuclear envelope (arrowheads) are visible in the perinuclear cytoplasm. Scale bar=1 sm. m, mitochondrion with a weakly electron-dense matrix. Fic. 9. The perinuclear cytoplasm of the sperm nucleus. Microtubules are scattered around (arrows). Mitochondrion with a highly electron-dense matrix (m, white lettering) and that with a weakly electron-dense matrix (m, black lettering) are visible. Scale bar=0.5 pm. arrowheads, fragments of the nuclear envelope; fg, fatty granule; sy, small yolk granule. perinuclear zone around the sperm nucleus. The first maturation division of the oocyte, which was in the telophase, was visible in the periplasm (data not shown). The sperm nucleus, about 5 ~m in diameter, migrated toward the center of the egg as development proceeded. It was clear that this nucleus was not cleavage nucleus. Under the electron microscope, the sperm nucleus was observed as a mass of electron-dense chromatin (Fig. 8). A complete nuclear en- velope was not observed but fragments of nuclear envelope were observed in the perinuclear cytoplasm (Fig. 8 and 9). Axial filament of the spermatozoon was observed near the nucleus (Fig. 8). Microtubules were scattered about (Fig. 9). Mitochondria were spherical or oval with faintly de- veloped cristae. Some of them had a highly electron-dense matrix but others had a weakly electron-dense matrix (Fig. 9). The large yolk granules surrounding the perinuclear cytoplasm were organized in radial arrays from 20 min after Oviposition (Fig. 2). Thirty minutes after oviposition, the bursting of vesicles was observed at the surface of the egg, and the matrix of the vesicles was scattered in the space between the cell membrane and the outer layer of the vitelline membrane (Fig. 10). Mucous material and fibrils, which were of the same electron density as the matrix of the vesicles, began to accumulate under the outer layer of the vitelline membrane. Méicrovilli protruded from the cell membrane. The eggs were difficult to dechorion from this stage. Forty-five minutes after oviposition, two chromosome plates at metaphase, one being that of the secondary oocyte and the other being that of the first polar body, were observed. The second maturation division of the secondary oocyte occurred at the surface of the inner spherical yolk mass, and its equatorial plane was perpendicular to the egg’s surface (Fig. 3a). A total of 15 eggs was employed for examination of the second maturation division of the first polar body. The spindle was situated at the periplasm in 9 eggs, while in the other 6 eggs it was observed at the cytoplasm somewhere among the outer radial yolk columns. In the former case, the equatorial plane was located perpen- dicularly to the egg’s surface (Fig. 4), by contrast, in the latter case it was parallel to the egg’s surface (Fig. 3b). Electron micrographs of maturation divisions could not be obtained because of failure in the staining of thick, epoxy-resin- embedded sections. One hour after oviposition, the sperm nucleus arrived at the center of the egg (Fig. 5). Until this stage, in a few eggs, several nuclei resembling the sperm nucleus were observed (Fig. 6). The number of such nuclei per egg was usually two or three, while the maximum number was seven, in the present investigation. ‘They were considered to be the result of polyspermy. Ninety minutes after oviposition, the vitelline membrane Fic.5. One hour after oviposition. Technovit-embedded section. The large yolk granules are arranged in outer radial columns (oc) and an inner spherical mass (im). The sperm nucleus, accompanied by perinuclear cytoplasm, is seen at the center of the egg (arrow). The large yolk granules surrounding the perinuclear cytoplasm are organized as radial arrays. Two sperm nuclei, a result of polyspermy, are visible in the egg (arrows). Fic. 6. One hour after oviposition. Paraffin-embedded section. Scale bar=50 «am. Scale bar=100 zm. ch, chorion. Fic. 7. One hour and thirty minutes after oviposition. Technovit-embedded sections of anegg. Scale bar=20 4m. Fig. 7a. The nucleus of the second polar body (arrow) and one chromosomal mass after the second maturation division of first polar body (arrowhead) are seen in the periplasm. Fig. 7b. Another chromosomal mass, generated by the division of the first polar body, (arrowhead) is seen in the periplasm, and the female pronucleus (arrow) is seen at the surface of the inner spherical yolk mass. 436 H. Suzuki AND A. KonpDo Fic. 10. The surface of the egg, thirty minutes after oviposition. The bursting of vesicles is seen at the egg’s surface (asterisks). The matrix of vesicles is dispersed into the space between the cell membrane and the outer layer (arrowhead) of the vitelline membrane. It forms the mucous material and the fibrils under the outer layer, generating the main layer. Microvilli (arrows) protrude from the cell membrane. Scale bar=0.5 um. ch, chorion; gg, glycogen granules; v, vesicle. Fic. 11. The surface of the egg, one hour and thirty minutes after Oviposition. The vitelline membrane consists of a highly elec- tron-dense outer layer (arrowhead) and a weakly electron-dense main layer (ml). The inner surface of the main layer lacks any limiting structures. A desmosome-like structure (arrow) is faintly visible in the invagination of the cell membrane (ic). Scale bar=0.5 um. fg, fatty granule; v, vesicle. oo consisted of an electron-dense outer layer of about 30 nm in thickness and a weakly electron-dense main layer of 0.1—0.3 ym in thickness (Fig. 11). The main layer increased in thickness as development proceeded. The inner surface of the main layer lacked any limiting structures. Microvilli were faintly visible. Desmosome-like structures were faintly visible between invaginated cell membranes in the surface region (Fig. 11), and they became more apparent as develop- ment proceeded. In the periplasm, which decreased in thickness, two masses of chromatin and a nucleus of about 7 ym in diameter were observed. These structures were all situated close to each other (Fig. 7a, b). The two masses of chromatin were due to the condensation of chromosomes after the second maturation division of first polar body. The nucleus was considered to be the nucleus of the second polar body, generated during the second maturation division of the secondary oocyte. While these profiles of polar bodies were observed until 14 hr after oviposition, no protrusion of polar bodies from the egg was observed. Just under the masses of chromatin of the first polar body and the nucleus of the second polar body, a nucleus, which was the same size as the nucleus of the second polar body, was located at the surface of the inner spherical yolk mass (Fig. 7b). It was considered to be the female pronucleus. At the center of the egg, the male pronucleus, which was about 154m in diameter, was observed. Decondensed chromatin was enclosed by a complete nuclear envelope (Fig. 12). Bundles of microtubules radiating from the perinuclear cytoplasm, were observed among the large yolk granules (Fig. 13). Two hours after oviposition, at the center of the egg, two pronuclei lay closed to one another (Fig. 14). The nuclear envelopes of the two pronuclei ran alongside each other at a distance of about 0.2 wm (Fig. 15). There was always a size difference between the two pronuclei, but no other morpho- logical differences were observed. The larger pronucleus, which was about 15 ~m in diameter, was probably the male pronucleus and the other, which was about 12 ~m in dia- meter, was probably the female one, as judged from the sizes at the previous stage. Much vesicular, smooth-surfaced endoplasmic reticulum appeared in the perinuclear cyto- plasm, but no rough-surfaced endoplasmic reticulum was observed (Fig. 16). Mitochondria with a highly electron- dense matrix increased in number. Three hours after oviposition, two pronuclei of about 22 ym and about 20 4m in diameter, respectively, or a single nucleus of about 25 ~m in diameter were observed at the center of the egg. Figure 17 shows the single nucleus at this stage. This nucleus is probably the zygote nucleus formed by the conjugation of the male pronucleus and the female pronucleus. No mitochondria with a weakly electron-dense matrix were found at the stage at which the putative zygote nucleus was observed. Three hours and thirty minutes after oviposition, the first nuclear division-had occurred at the center of the egg. A chromosome plate at metaphase was observed (Fig. 18). DISCUSSION The vitelline membrane It is generally accepted that the vitelline membrane is located internally with respect to the chorion in spider eggs. There are a few observations relevant to the details of its formation. Rempel [8] studied the embryonic development of Latrodectus mactans and reported that, during the first few hours after oviposition, the vitelline membrane adhered closely to the underlying periplasm and the overlying chorion. Hence, the removal of the chorion was difficult. Soon, however, the chorion and the vitelline membrane separated from each other, and the removal of the chorion became easy. Rempel suggested that the vitelline membrane con- Meiosis and Fertilization in Spider 437 Fic. 12. Electron micrograph showing a male pronucleus and perinuclear cytoplasm. The male pronucleus is enclosed by a complete nuclear envelope (ne) and contains dispersed chromatin. The perinuclear cytoplasm is surrounded by large yolk granules (yg), which are arranged radially (see Fig. 2). Scale bar=5 um. arrows, weakly electron-dense mitochondria; arrowheads, highly electron-dense mitochondria; fg, fatty granules; sy, small yolk granules. Fic. 13. Bundles of microtubules between large yolk granules (yg), which are arranged radially around the male pronucleus. Scale bar=0.5 ym. Fic. 14. Two pronuclei close to one another. The larger pronucleus is probably the male pronucleus (mp) and the other is probably the female pronucleus (fp). Scale bar=5 ~m. m, mitochondria; sy, small yolk granules; yg, large yolk granule. Fic. 15. The nuclear envelopes (ne) of two pronuclei run parallel to each other at a distance of 0.2 ~m. The putative female pronucleus (fp) is on the left and the putative male pronucleus (mp) is on the right. Scale ar=0.2 wm. arrowheads, nuclear pores. Fic. 16. Smooth-surfaced endoplasmic reticulum round the putative female pronucleus (fp). Scale bar=1 “m. arrowhead, nuclear pore; m, highly electron-dense mitochondria; ne, nuclear envelope. trolled the entrance of sperm into the egg while sperm could pass freely through the chorion. According to Kondo [4], in lycosid spiders, only the thin outer layer of a vitelline membrane, which could not be recognized by light micros- copy, was present 30 min after oviposition. The main layer of the vitellime membrane was formed as a very low electron dense layer, containing fibrils, under the outer layer. Kondo suggested that the vitelline membrane bore some resem- blance, in terms of its formation, to the fertilization mem- brane. By contrast, Seitz [9] reported that “the funiculus cells” secreted a precursor component of the vitelline mem- brane during the first vitellogenic phase in Cupiennius salei. 438 H. Suzuki AND A. Konpo ikea Fic. 17. The zygote nucleus after conjugation. Scale bar=5 um. m, mitochondria; ne, nuclear envelope; sy, small yolk granules. The observations of these authors appear to contradict one another in terms of the origin of the vitelline membrane. The results of the present investigation strongly support the observations of Kondo [4]. The matrix scattered from the vesicles, by exocytosis, in the space between the cell membrane and the outer layer appeared to form the mucous material and fibrils under the outer layer, and it probably forms the main layer of the vitelline membrane. The forma- tion of the outer layer and that of the main layer may be successive processes that require different materials. Maturation of the egg The second maturation division of the first polar body progressed until telophase, but no daughter nuclei were formed. Such division seems similar to that in eggs of A. tepidariorum [6]. According to Warren [11], division of the first polar body ceased at metaphase in Palystes natalius. The maturation division of the secondary oocyte was strictly synchronized with respect to that of the first polar body. Montgomery [6] reported that the nucleus of the second polar body was located in the outer radial column of yolk granules. He may have observed the nucleus, as it migrated to the periplasm, immediately after division. War- ren [11] described only anaphase in a discussion of the division of the secondary oocyte. According to Montgomery [6], the chromosomes of the first polar body were not found more than 2 hr after oviposi- tion, and the last time at which the nucleus of the second polar body was observed was 169 min after oviposition. In A. japonica, two masses of chromatin of the first polar body and the nucleus of the second polar body were found until 14 hr after oviposition. It is unclear from the present study whether or not polar bodies are eliminated from the egg via the invaginations of the cell membranes. Fic. 18. The first nuclear division at metaphase. Chromosomes (cr) are aligning at the equatorial plane. Scale bar=S um. fg, fatty granules; m, mitochondria; mt, microtubules. Inset. Centrioles at a spindle pole. Scale bar=1 um. Meiosis and Fertilization in Spider 439 Fertilization The time at which the sperm is incorporated into a spider’s egg remains unknown. Montgomery [6] reported that eggs accepted a sperm nucleus that was located about halfway between the periphery and the center of the egg at oviposition, while Warren [11] and Rempel [8] suggested that fertilization occurs after oviposition. The present investiga- tion supports the evidence reported by Montgomery. From the location of the sperm nucleus in the newly laid egg, the incorporation of sperm in the ovarian cavity can be postulated. The sperm nucleus was observed as a mass of condensed chromatin, and it was not enclosed by a nuclear envelope. The nuclear envelope may be broken down at the time of incorporation of the sperm. Mitochondria that had a weakly electron-dense matrix may have been of paternal origin. The large yolk granules surrounding the perinuclear cyto- plasm were arranged radially from immediately after oviposi- tion. This arrangement of large yolk granules may be caused by the radially distributed cytoplasm that contains numerous bundles of microtubules that radiate from the perinuclear cytoplasm. Holm [2] called the cytoplasm that connects the perinuclear cytoplasm to the periplasm “plasm threads”, and he suggested that they allow cleavage nuclei to migrate toward the periphery of the egg. By contrast, Kondo [4] suggested that the migration of the cleavage nuclei might be related to the decomposition of yolk granules and to motion of the cell membrane, and he also suggested that the plasm threads, which he described as a protoplasmic reticu- lum, only lay between the perinuclear cytoplasm and the periplasm. He would not have detected any bundles of microtubules because he prepared his samples by fixation at 0-4°C. Microtubules tend to be depolymerized at such low temperatures [10]. We suggest that the radial bundles of microtubules are equivalent to the plasm threads and that they play an important role in the migration of the sperm nucleus, accompanied by the perinuclear cytoplasm, to the center of the egg. In previous studies of many spiders [1-3, 5, 7, 12], a nucleus has been noted at the center of the egg immediately after oviposition. This nucleus may have been the male pronucleus. The sperm nucleus arrives at the center of the egg and then develops into the male pronucleus. The chro- matin ceases to be condensed, and fragments of the original nuclear envelope may be reconstructed as the complete nuclear envelope. The female pronucleus, generated by the second maturation division, should then move toward the center of the egg, in which the male pronucleus is now located. However, no migrating female pronucleus was observed in the present study. In A. japonica, the male and female pronuclei lie close to one another for about one hour before conjugation, and each increases in size. Montgomery [6] stated that the larger nucleus was certainly the sperm nucleus because it was similar in volume to the supernumerary sperm nuclei. Most of the mitochondria with a highly electron-dense matrix may have immigrated with the female pronucleus. While rod-shaped mitochondria were observed in the peri- plasm (Suzuki and Kondo, unpublished data), they were hardly ever observed in the perinuclear cytoplasm. REFERENCES 1 Holm A (1952) Experimentelle Untersuchungen iiber die Entwicklung und Entwicklungsphysiologie des Spinnenembryos. Zool Bidr Uppsala 29: 293-424 2 Holm A (1954) Notes on the development of an orthognath spider, Ischnothele karschi Bos. & Lenz. Zool Bidr Uppsala 30: 109-221 3 Kishinouye K (1891) On the development of Araneina. J Coll Sci Imp Uni Tokyo 4: 55-88 4 Kondo A (1969) The fine structures of the early spider embryo. Sci Rep Tokyo kyoiku Daigaku Sec B 14: 47-67 5 Locy WA (1886) Observations on the development of Agelena nevia. Bull Mus Com Zool 12: 63-103 6 Montgomery TH (1908) On the maturation mitoses and ferti- lization of the egg of Theridium. Zool Jb Anat 25: 237-250 7 Morin I (1887) Zur Entwicklungsgeschichte der Spinnen. Biol Zentralbl 6: 658-663 8 Rempel JG (1957) The embryology of the black widow spider, Latrodectus mactans (Fabr.). Canad J Zool 35: 35-74 9 Seitz KA (1971) Licht- und elektronenmikroskopische Unter- suchungen zur Ovarentwicklung und Oogenese bei Cupiennius salei Keys. (Araneae, Ctenidae). Z Morph Tiere 69: 283-317 10 Tilney LG, Porter KR (1967) Studies on the microtubules in Heliozoa, II. The effect of low temperature on the formation and maintenance of the axopodia. J Cell Biol 34: 327 11 Warren E (1926) On the habits, egg-sacs, oogenesis and early development of the spider Palystes natalius (Karsch). Ann Natal Mus 5: 303-349 12 Yoshikura M (1955) Embryological studies on the liphistiid spider, Heptathela kimurai. Part 2. Kumamoto J Sci Ser B 2: 1-86 ZOOLOGICAL SCIENCE 11: 441-444 (1994) © 1994 Zoological Society of Japan Neuron-like morphology expressed by perinatal rat C-cells in vitro Icuiro NisHtyAMa!, TADACHIKA OoTA’ and MANABU OGISO> ‘Biological Laboratory, Komazawa Women’s College, Inagi, Tokyo 206, *Isotope Center, Tokyo University of Agriculture, Setagaya-ku, Tokyo 156, Japan, *Cell and Information, PRESTO, Research Development Corporation of Japan (J.R.D.C.), and Department of Physiology, Toho University School of Medicine, Ohta-ku, Tokyo 143, Japan ABSTRACT— Thyroid C-cells (calcitonin-producing cells) are endocrine derivatives of the neural crest. The morpholo- gical plasticity of rat C-cells was examined under cell culture. In a primary culture of perinatal rat thyroid glands, a small number of C-cells were found to extrude neurite-like processes, some of which reached 350m in length. The processes frequently branched and had varicosity-like structures. The processes were intensely stained with anti-a-tubulin antibody, suggesting that microtubular cytoskeleton participated in their elongation and maintenance. In primary cultures of C-cells derived from postnatal rats at day 2 or later, no neurite-like processes were observed. These findings suggest that at least some C-cells in the perinatal rat thyroid retain the potential to extrude neurite-like processes, as do chromaffin cells in adrenal medulla, another type of crest-derived endocrine cell. INTRODUCTION Although calcitonin-producing cells (C-cells) in the thyr- oid gland are classified as endocrine cells, they have phy- logenetic, embryological and biochemical relationships with enteric serotonergic neurons [1]. As befits their neuroec- todermal origin, C-cells have several neuronal properties. They produce calcitonin gene-related peptide (CGRP), a putative neurotransmitter, as an alternative product of the calcitonin gene [12]. C-cells are capable of synthesizing serotonin from L-tryptophan and accumulate it in their secretory granules with neuron-specific serotonin binding protein [1]. Our recent study [9] revealed that C-cells express neural cell adhesion molecules on their surfaces. Neuronal characteristics of C-cells have also been confirmed by electrophysiological techniques [5, 13, 14]. Although C-cells share some neuronal properties, they do not display neuron-like morphological features in the thyroid gland [4]. In this communication, we report that a small number of C-cells derived from perinatal rats displayed neuronal mor- phology in vitro, suggesting their morphological plasticity. MATERIALS AND METHODS For immunohistochemical study, thyroid glands were excised from 20-day-old fetuses or 9-week-old male rats of Wistar strain. The day on which sperm were observed in vaginal smears was designated as day 0 of pregnancy. The glands were fixed in freshly prepared 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) at 4°C overnight. After being rinsed with phosphate-buffered saline (PBS), the specimens were immersed in 20% sucrose in PBS at 4°C for 24 hr, and subsequently frozen in an OCT compound (Miles). Transverse sections of 6m in thickness were cut on a cryostat and Accepted June 6, 1994 Received March 28, 1994 mounted on glass slides coated with egg white. Calcitonin immunoreactivity in the sections was detected using rabbit anti-human calcitonin antiserum (1:800, ICN ImmunoBiolo- gicals) and rhodamin-labeled goat anti-rabbit IgG antiserum (1 : 100, Cappel). The specimens were observed under a NIKON DI- APHOT-TMD microscope equipped with fluorescence optics. The procedures were detailed in our previous paper [8]. For primary cultures, the thyroid glands were dissected out from perinatal rats ranging in age from embryonic day 16 (E16) to postnatal day 4 (P4), and dispersed with collagenase (Worthington, CLS II) and Dispase (Godo Shusei) [7]. The dispersed cells were cultured on glass coverslips (14mm in diameter) in 24-well multi- dishes (Falcon) with Dulbecco’s modified Eagle’s medium (GIBCO) containing 5% fetal bovine serum (GIBCO). The cell density was controlled to obtain approximately 500 C-cells/well after 48 hr of incubation, as detailed in our previous paper [7]. The cultures were maintained at 37°C in a humidified atmosphere of 5% CO, in air. Primary cultures were also prepared from the thyroid glands of young (4- and 9-week-old) and aged (80-week-old) rats as described above. After being incubated for 48hr, the cells were fixed with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) for 1 hr, and permeabilized with 0.25% Triton X-100 for 10 min. Calcitonin was immunocytochemically detected as described above, and the number of neuron-like C-cells, which were defined as calcitonin- immunoreactive cells with at least one process longer than 150m, were counted. In some experiments, CGRP was detected using rabbit anti-rat CGRP antiserum (1:400, Peninsula). For double- immunostaining of calcitonin and a-tubulin, mouse monoclonal anti- a-tubulin antibody (1:500, BioMakor) and fluorescein-labeled goat anti-mouse IgG antiserum (1:100, Kirkegaard and Perry Lab) were used in combination with the immunostaining of calcitonin described above. The specificity of the immunoreactions was determined by omitting the primary antibody, or by preincubation of the antibody with an excess of antigen. No specific immunoreactivity was found in these controls. RESULTS AND DISCUSSION In the sections of thyroid gland, C-cells were oval or 442 I. NisHtyAMA, T. OoTA AND M. Ociso polygonal in shape, and were located close to the basal portion of the follicular epithelium (Fig. 1). In the E20 fetal thyroid, some of the C-cells were found to extrude short processes (Fig. 1A, arrows). The length of these processes was estimated to be 40um at most. In contrast, no processes were observed in adult thyroid C-cells (Fig. 1B). To determine whether these C-cells extend processes in vitro, the thyroid glands of both fetal and adult rats were dissociated into single cells and grown in culture. In the cultures of 9-week-old rat thyroid, C-cells displayed oval or triangular shapes. Although a few C-cells had short proces- ses of up to 30um, C-cells with long processes were not observed in the cultures of adult thyroid up to day 5 in vitro (data not shown). In the cultures of E20 fetal thyroid glands, most of the C-cells exhibited ovoid or triangular shapes, as in the cultures of adult thyroid (Fig. 2A). However, a small fraction of these C-cells displayed neuron- like features (Fig. 2 B-D). Most of the neuron-like C-cells were monopolar, and the processes frequently bifurcated. Varicosity-like structures were observed along some of these processes (arrows in Fig.2 B and C). The cytoplasm, in- cluding the processes, of the C-cells was filled with calcitonin and CGRP immunoreactant. All the C-cells examined ex- pressed CGRP immunoreactivity irrespective of whether the C-cells possessed the neurite-like processes. The longest process reached 350m in length. Two to seven neuron-like C-cells (0.4-1.4%) were detected among approximately 500 C-cells in each culture well. Prolonged incubation of up to 5 days did not increase the proportion of neuron-like C-cells (data not shown). The neurite-like processes of C-cells were intensely stained with anti-a-tubulin antibody (Fig. 2 E and F), suggest- ing that microtubular cytoskeleton participated in the forma- tion and/or maintenance of these processes. Microtubules in neurites are stabilized by microtubule-associated proteins, such as MAP-2 and tau [15]. Our recent study revealed that both thyroid C-cells and a C-cell line produce tau- immunoreactive protein with an apparent molecular mass of 110,000 (Nishiyama et al., in preparation), probably corres- ponding to the high molecular weight tau found in PC-12 cells [3] and neuroblastoma cells [2]. Therefore, it is plausible that microtubules in the processes of the neuron-like C-cells are associated with the high molecular weight tau protein. The results of this study indicated that a small percentage of fetal rat C-cells, but none of adult origin, had the ability to extend long neurite-like processes. Next, we determined the ratio of neuron-like C-cells in thyroid cell cultures of perinatal rats of varying ages. As shown in Table 1, neuron- like C-cells were also observed at almost the same ratio in cultures of E19 fetal rat thyroid glands as in those of E20 fetuses, and at much lower ratios in cultures of E17, E18, PO and P1 rat thyroid glands. No neuron-like C-cells were detected in the thyroid cell cultures of E16 fetuses, P2 and P3 pups (Table 1), or P28, P63 and P560 rats (data not shown). These results show that neuron-like C-cells appeared in thyroid cell cultures during only a limited perinatal period. A question arises as to why only a subpopulation of C-cells extend processes in thyroid cultures of perinatal rats. This implies a heterogeneity in the microenvironment sur- rounding the C-cells or in the properties of C-cells them- selves. To answer this question, it is important to know whether C-cells without process have the potential to extend processes in response to some factor(s). An attempt to induce neurite-like process outgrowth in the C-cells in vitro is now in progress using various growth factors and hormones. Although the neural crest origin of C-cells has been confirmed by several investigators [6, 10, 11], the develop- mental process of restricting C-cell phenotypic traits in the neural crest lineage remains to be elucidated. Our findings suggested that a small number of C-cells from perinatal rats, especially those obtained one or two days before birth, had the potential to extrude in vitro long neurite-like processes, a hallmark of the neuronal phenotype. Future experiments using this culture system will shed light upon commitment and Fic. 1. Morphology of C-cells in the thyroid glands of fetal (A) and young adult (B) rats. Transverse sections of 20-day-old rat fetuses (A) and 9-week-old rats (B) were immunostained using an anti-calcitonin antiserum. Fetal C-cells appeared to extrude short processes (arrows in A). Scale bar=20um. Neuron-like Morphology in Rat C-Cells 443 Fic. 2. C-cells in primary culture of dissociated thyroid glands of 20-day-old rat fetuses. with antisera to calcitonin (A-C) and calcitonin gene-related peptide (D). The cells were cultured for 48 hr, and immunostained C-cells with neurite-like processes were observed in the cultures at a low ratio (B-D). Their processes frequently branched and had varicosity-like structures (arrows in B and C). Calcitonin (E) and a-tubulin (F) immunoreactivities were detected in the same specimen. anti-a-tubulin antibody (arrows in F). Scale bar=20m. TaBLE1. Frequency of neuron-like C-cells in primary cultures derived from rat thyroid glands of various ages Age Number of Neuron-like C-cells per 1,000 C-cells Embryonic day 16 0 17 0.1+0.07 18 1.2+0.73 19 7.3+1.54 20 8.4+2.16 Postnatal day 0 Upyae N35) 1 0.2+0.09 2 0 3 0 The neurite-like process of the C-cell was intensely reactive with the plasticity of the phenotype in C-cells. REFERENCES Bernd P, Gershon MD, Nunez EA, Tamir H (1979) Localiza- tion of a highly specific neuronal protein, serotonin binding protein, in thyroid parafollicular cells. Anat Rec 193: 257-268 Couchie D, Mavilia C, Georgieff IS, Liem RKH, Shelanski ML, Nunez J (1992) Primary structure of high molecular weight tau present in the peripheral nervous system. Proc Natl Acad Sci USA 89: 4378-4381 Goedert M, Spillantini MG, Crowther RA (1992) Cloning of a Thyroid glands of rat fetuses or pups at the ages indicated were enzymatically dissociated and cultured for 48hr. The cells were immunostained using an anti-calcitonin antiserum, and the ratios of neuron-like C-cells, which were defined as calcitonin- immunoreactive cells bearing at least one process longer than 150m, were determind. At least three samples, each of which contained approximately 500 C-cells, were counted for each experiment. The values are means+SEM of four independent experiments. 444 I. NisHtyAMA, T. Oota AND M. Ociso big tau microtubule-associated protein characteristic of the peripheral nervous system. Proc Natl Acad Sci USA 89: 1983- 1987 Kameda Y (1987) Localization of immunoreactive calcitonin gene-related peptide in thyroid C cells from various mammalian species. Anat Rec 219: 204-212 Kawa K (1988) Voltage-gated sodium and potassium currents and their variation in calcitonin-secreting cells of the chick. J Physiol 399: 93-113 Le Douarin N, Fontaine J, Le Liévre C (1974) New studies on the neural crest origin of the avian ultimobranchial glandular cells—Interspecific combinations and cytochemical character- ization of C cells based on the uptake of biogenic amine precursors. Histochemistry 38: 297-305 Nishiyama I, Fujii T (1989) Somatostatin-immunoreactive C- cells in primary culture of fetal, neonatal and young rat thyroid glands. Biomed Res 10: 353-359 Nishiyama I, Fujii T (1992) Laminin-induced process out- growth from isolated fetal rat C-cells. Exp Cell Res 198: 214- 220 Nishiyama I, Seki T, Oota T, Ohta M, Ogiso M (1993) Ex- pression of highly polysialylated neural cell adhesion molecule in 10 11 12 13 14 15 calcitonin-producing cells. Neuroscience 56: 777-786 Pearse AGE, Polak JM (1971) Cytochemical evidence for the neural crest origin of mammalian ultimobranchial C cells. His- tochemie 27: 96-102 Polak JM, Pearse AGE, Le Liévre C, Fontaine J, Le Douarin NM (1974) Immunocytochemical confirmation of the neural crest origin of avian calcitonin-producing cells. Histochemistry 40: 209-214 Sabate MI, Stolarsky LS, Polak JM, Bloom SR, Varndell IM, Ghatei MA, Evans RM, Rosenfeld MG (1985) Regulation of neuroendocrine gene expression by alternative RNA processing. J Biol Chem 260: 2589-2592 Sand O, Jonsson L, Nielsen M, Holm R, Gautvik KM (1986) Electrophysiological properties of calcitonin-secreting cells de- rived from human medullary thyroid carcinoma. Acta Physiol Scand 126: 173-179 Sand O, Ozawa S, Gautvik KM (1981) Sodium and calcium action potentials in cells derived from a rat medullary thyroid carcinoma. Acta Physiol Scand 112: 287-291 Tucker RP (1990) The roles of microtubule-associated proteins in brain morphogenesis: a review. Brain Res Rev 15: 101-120 ZOOLOGICAL SCIENCE 11: 445-449 (1994) © 1994 Zoological Society of Japan Proliferation of Pituitary Cells in Streptozotocin-induced Diabetic Mice: Effect of Insulin and Estrogen SuMIO TAKAHASHI, SOUICHI OOMIZU and YASuO KOBAYASHI Department of Biology, Faculty of Science, Okayama University, Tsushima, Okayama 700, Japan ABSTRACT— Insulin has growth-stimulatory actions in various tissues. The present study is aimed to clarify whether insulin stimulates the proliferation of anterior pituitary cells. Estrogen is the mitogenic factor in pituitaries and female reproductive tracts. We studied the insulin-estrogen relationship in the pituitary cell proliferation. The proliferation of uterine epithelial cells was also studied, since the uterus was one of the most typical estrogen-responsive organs. Ovariectomized ICR mice were given streptozotocin (STZ, 100 mg/kg) intraperitoneally to make mice insulin-deficient. Estradiol-17 8 (50 4g, E>) and insulin (0.2 or 0.41U per twice a day) were given in STZ-treated mice. Insulin significantly stimulated the mitosis of pituitary cells and PRL cells in both normal control mice and STZ-treated diabetic mice. E> stimulated the mitosis of pituitary cells (10.93+0.73 cells/mm), compared with controls (3.59-+0.67 cells/ mm7) in control mice. In STZ-treated mice E, failed to increase the mitosis (4.91+0.99 cells/mm7), compared with controls (2.47+0.43 cells/mm). Insulin recovered the diminished response to estrogen in pituitary cells of STZ-treated mice to the level comparable to control mice. In the uterus insulin at the high dose used stimulated the proliferation of luminal epithelial cells in normal control mice. However, insulin deficiency did not alter the responsiveness of uterine epithelial cells to estrogen. The present study suggests that insulin is involved in the proliferation of pituitary cells and probably uterine luminal epithelial cells, but the mechanism of insulin action on the cell proliferation may differ between pituitary cells and uterine cells. INTRODUCTION Estrogen stimulates proliferation of various cells includ- ing pituitary cells and uterine cells. Recent reports concern- ing the estrogen-induced cell proliferation suggest that estrogen action is not a direct action and is mediated by autocrine or paracrine growth factors [24, 27, 32]. Trans- forming growth factor a and epidermal growth factor are growth factors for pituitary cells, particularly for prolactin (PRL) cells [3]. It is highly probable that other growth factors are involved in pituitary cell proliferation. Insulin is known to have growth-promoting activity in various tissues [26]. Insulin is indispensable for estrogen-induced pituitary growth [6]. Pituitary tumor cells, GHz cells, require insulin for the optimal growth in serum-free medium [11]. The present study was undertaken to clarify effects of insulin on the estrogen-induced proliferation of pituitary cells in ovar- iectomized mice made diabetic by streptozotocin (STZ) ad- ministration. The proliferation of PRL cells was particularly examined, since insulin stimulates PRL secretion [13, 25]. Uterine growth is regulated by estrogens and progesterone. The proliferation of uterine luminal epithelial cells in STZ- treated diabetic mice was also studied in the present study. MATERIALS AND METHODS Adult female mice (2 month old) of the Jcl:ICR strain (Clea Accepted May 25, 1994 Received March 14, 1994 Japan) were used. They were kept under temperature-controlled conditions, and given food and water ad libitum. All mice were ovariectomized under ether anesthesia. Ten days after ovariec- tomy, STZ (Wako Pure Chemicals) was intraperitoneally given at a dose of 100 mg/kg BW. STZ was dissolved in 0.05 M citric acid (65 mg/ml). Control mice were given the vehicle. All mice had been in advance deprived of food for 16 hr before STZ treatment. Estrogen treatment Estradiol-17 8 (E>, Sigma), dissolved in sesame oil (1 mg/ml), was subcutaneously given 7 days after STZ treatment at a dose of 50 yg. Control mice were given sesame oil. Insulin treatment in STZ-treated mice Seven days after STZ treatment insulin (0.2 or 0.41U for each mouse, Novo Actrapid MC, Novo Nordisk), diluted in saline (1 or 2 IU/ml), was intraperitoneally given twice a day (9.00 AM and 5.00 PM) for 3 days. Control mice were given the vehicle. On the last day insulin was given only in the morning, and 5 hr later pituitaries were collected for the study. The mitotic activity of pituitary cells and uterine epithelial cells Colchicine (Wako Pure Chemicals), dissolved in saline (1 mg/ ml), was subcutaneously given at a dose of 5mg/kg BW. The pituitary gland and uterine horns were removed Shr after the colchicine injection, that is, 48 hr after estrogen treatment, and then fixed in Bouin’s solution. The pituitary glands and uteri were embedded in paraplast, and horizontal pituitary and cross uterine serial sections (5 um thickness) were cut. The colchicine-arrested mitotic cells were observed. For the counting of mitotic pituitary cells, sections near the horizontal medial plane were selected. The number of mitotic pituitary cells was counted. In the sections used for the counting, the area of anterior pituitary glands was measured with a planimeter. The mitotic activity was expressed as the number of mitotic cells per mm?. For the counting of mitotic uterine luminal epithelial cells, sections near the cross medial plane were selected. 446 S. TAKAHASHI, S. OomMIzu AND Y. KoBaAyASHI The number of total luminal epithelial cells and the mitotic luminal epithelial cells was counted. The mitotic activity of uterine cells was expressed as the percentage of mitotic luminal epithelial cells in total luminal epithelial cells. Immunocytochemical identification of PRL cells PRL cells were identified in the ABC method using the rabbit anti-mouse PRL serum (Shikibo). The specificity of anti-mouse PRL serum was checked with the immunoblotting method. The antiserum detected a single band of mouse prolactin, and did not cross-react with mouse growth hormone. Measurement of blood glucose level Mice were lightly anesthetized with ether vapor, and the blood was obtained from the tail vein. The blood glucose level was determined by the glucose oxidase method using the kit (Glucoscan, Eiken), and expressed as mg/dl. Statistics Statistical analysis was carried out by Bonferroni’s multiple t-test. Differences of P<0.05 were considered as statistically significant. RESULTS Mitotic activity of pituitary cells STZ treatment lowered the basal mitotic activity of pituitary cells, although statistical difference was not detected between control mice and STZ-treated mice (Fig.1). Insulin alone (0.2 IU, 0.4 IU) increased the mitotic activity of pitui- tary cells in control and STZ-treated mice. Estrogen admin- istration resulted in a three-fold increase in the mitotic activity of pituitary cells of control mice (control, 3.59+0.67 cells/mm7; E2, 10.93+0.73 cells/mm7; Fig. 1). However, CL OO Me = stradiol Streptozotocin Number of mitotic cells (cells/mm?) Vehicle Insulin Insulin Vehicle Insulin Insulin 0.2IU =-0.4TU 0.2tU = 0.4TU Fic. 1. Influence of estradiol-17 8 and insulin (0.21U, 0.41U) on the number of mitotic pituitary cells (mean+SE of the mean) in control ovariectomized mice and streptozotocin-treated ovar- iectomized mice. The number above each column depicts the number of mice. *, P<0.05; **, P<0.01 vs. respective vehicle group. +, P<0.01 vs. respective corresponding group of con- trol mice. a, P<0.05;b, P<0.01; c, P<0.001 vs. respective oil group. in STZ-treated diabetic mice estrogen administration failed to increase the mitotic activity of pituitary cells (control, 2.47+ 0.43 cells/mm7?; E>, 4.91+0.99 cells/mm7?). The number of mitotic pituitary cells in STZ-treated mice with estrogen treatment was significantly lower than that of corresponding control mice (P<0.001). Insulin administration in associa- tion with estrogen further increased the mitotic activity in control mice. In STZ-treated mice receiving insulin, estrogen was able to increase the mitotic activity of pituitary cells to the level comparable to control mice. Thus, insulin administration stimulated the proliferation of pituitary cells and recovered the diminished response of pituitary cells to estrogen in STZ-treated mice. STZ treatment without estrogen injection did not lower the mitotic activity of PRL cells (Fig. 2). Estrogen adminis- tration resulted in an increase in mitotic activity of PRL cells, but the level in STZ-treated mice was lower than that in control mice (P<0.01). Insulin alone or in association with estrogen increased the mitotic activity of PRL cells in control and STZ-treated mice. However, high dose of insulin (0.4 IU) failed to increase the mitotic activity of PRL cells in control mice. Insulin administration at the high dose may have depleted the storage of PRL, resulting in the decrease in the number of immuno-positive PRL cells. [stom Control GE Estradiol 8 Number of mitotic PRL cells (cells/mm?) Vehicle Insulin Insulin Vehicle Insulin Insulin 0.2IU 0.4TU 0.2IU = 0.41U Fic. 2. Influence of estradiol-17 8 and insulin (0.2 IU, 0.41U) on the number of mitotic PRL cells (mean+SE of the mean) in control ovariectomized mice and streptozotocin-treated ovar- iectomized mice. The number above each column depicts the number of mice. *, P<0.05; **, P<0.01 vs. respective vehicle group. #, P<0.05; ##, P<0.01 vs. respective correspond- ing group of control mice. a, P<0.05; b, P<0.001 vs. respec- tive oil group. Mitotic activity of uterine luminal epithelial cells Insulin administration at the high dose (0.41U) in- creased the mitotic activity of luminal epithelial cells in control mice. E> administration increased the mitotic activ- ity of luminal epithelial cells in control mice and in STZ- Proliferation of Mouse Pituitary Cells 447 ae) Oil GE Estradiol Streptozotocin d Control 14 Percentage of mitotic cells Vehicle Insulin Insulin Vehicle Insulin Insulin 0.2IU 60.41U 0.2IU 6 0.41U Fic. 3. Influence of estradiol-17 8 and insulin (0.2 IU, 0.4 TU) on the percentage of mitotic uterine luminal epithelial cells (mean + SE of the mean) in control ovariectomized mice and streptozoto- cin-treated ovariectomized mice. The number above each col- umn depicts the number of mice. *, P<0.01 vs. respective vehicle group. a, P<0.01; b, P<0.001 vs. respective oil group. cr oi ,.treptozotocin 400 Hi =o Estradiol 300 Control 200 Blood glucose level (mg/dl) Vehicle Insulin Insulin Vehicle Insulin Insulin 0.2m = 0..4TU 0.2tU) 8=60.41U Fic. 4. Blood glucose levels in control ovariectomized mice and streptozotocin-treated ovariectomized mice receiving estradiol- 17 or insulin. The number above each column depicts the number of mice. *, P<0.05; **, P<0.01 vs. respective vehicle group. a, P<0.01 vs. respective oil group. treated mice (Fig. 3). Insulin administration in association with estrogen did not further increase the mitotic activity of luminal epithelial cells in both control and STZ-treated mice. Blood glucose level Normal blood glucose level was 90.1+8.4 mg/dl (n= 13). Estrogen treatment did not change the blood glucose level, although in insulin-treated control mice estrogen de- creased glucose level (Fig. 4). STZ treatment increased blood glucose levels (409.0+19.6 mg/dl, n=12, at 20 days after STZ injection). Insulin decreased blood glucose levels in a dose-dependent manner in STZ-treated mice. DISCUSSION The present study revealed that insulin stimulated the proliferation of pituitary cells. Growth promoting action of insulin on uterine epithelial cells had been already demon- strated in rats and mice [9, 15]. Rat pituitary tumor cells, GH cells, require insulin for the optimal growth in vitro system [11]. However, as far as we know, stimulatory effect of insulin on the proliferation of normal pituitary cells has never been reported. STZ administration induces diabetes in mice and rats, which is ascertained by serum hyperglycemia and significant loss of body weights. Insulin administration was able to decrease the elevated serum glucose levels in STZ-induced diabetic mice. Our preliminary study showed that STZ (100 mg/kg) injection significantly decreased serum insulin levels (control male mice, 51.1+8.9 ~U/ml; STZ-treated males, 12.0+5.1 #U/ml). In STZ-treated mice, estrogen failed to increase the mitotic activity of pituitary cells. Repeated injection of insulin recovered the reduced responsiveness of pituitary cells to estrogen in STZ-treated diabetic mice to the level observed in normal mice. Insulin (0.4 IU) increased the mitotic activ- ity of luminal uterine epithelial cells in control mice. In- effectiveness of insulin in STZ-treated mice on the prolifera- tion of luminal cells may be accounted for by the reduced responsiveness of uterine epithelial cells to insulin in insulin- deficient mice. Insulin was not able to enhance further the estrogen-induced proliferation of uterine cells. These re- sults indicate that insulin stimulates the cell proliferation in pituitary cells and uterine epithelial cells, but the pathway of signal transduction leading to cell division may be different. Several studies described the proliferation of pituitary cells in rats [4, 22, 23, 28]. Growth hormone-secreting cells (somatotrophs) and PRL cells are the most actively prolife- rating cells in pituitary secretory cells. PRL cells were particularly analyzed in the present study, since PRL secre- tion is known to be regulated by insulin [13, 25]. STZ- treated diabetes decreased PRL secretion in rats [6, 12]. In PRL cells of such diabetic rats the decrease in number of secretory granules and the atrophy of cell organelles were electron microscopically observed [34]. The present study showed that insulin stimulated the proliferation of PRL cells. The enhanced proliferation of PRL cells by insulin may be correlated with the enhanced PRL secretion [29]. Further analysis is needed for the understanding of the secretion and proliferation-coupling of PRL cells. IGF receptors as well as insulin receptors are localized in pituitary glands [10], and in uterine tissues [7]. Our pre- liminary in vitro study indicated that IGF-I was more potent (about 100-fold) in stimulating the proliferation of cultured 448 S. TAKAHASHI, S. OoMIZU AND Y. KOBAYASHI pituitary cells than insulin (Oomizu and Takahashi, unpub- lished observation). Therefore, insulin action may be medi- ated by insulin-like growth factor-I (IGF-I) receptors in pituitary glands. Several studies indicate that estrogen action on the cell proliferation is indirect and mediated by autocrine or para- crine growth factors [24, 27, 32]. Transforming growth factor @ and epidermal growth factor are candidates of growth factors in the pituitary gland and the uterus [3, 16, 19, 20, 31]. Insulin or IGF-I may be another candidate of estrogen-associated growth factors. IGF-I and IGF-I mRNA are detected in pituitary glands [1, 18,21]. Estrogen increased levels of IGF-I mRNA, IGF binding and IGF binding proteins [17]. As mitogenic action of estrogen is well known, the increase in pituitary IGF-I level by estrogen treatment may be closely associated with the estrogen- induced proliferation of pituitary cells. Estrogen adminis- tration may stimulate the secretion of IGF-I from pituitary cells, and in turn IGF-I secreted may stimulate the prolifera- tion of pituitary cells in an autocrine or paracrine fashion. It is highly probable that insulin administered in the present study is able to accelerate the pituitary cell proliferation by the stimulation of intrinsic pituitary IGF-I system. In the uterus IGF-I is also detected [18], and its synthesis is stimu- lated by estrogen [2, 8]. IGF-I receptors are detected in uterine cells [7]. These results strongly suggest the autocrine or paracrine control of IGF-I on the proliferation of uterine cells. Therefore, it is also highly probable that insulin acts on IGF-I receptors, resulting in the stimulation of uterine epithelial cells. Reduced responsiveness to estrogen on prolactin secre- tion had already reported in STZ-treated rats [6, 30, 33], and this reduction is partly due to the alteration in pituitary estrogen receptor system [30, 33]. Most of pituitary secre- tory cells including PRL cells had estrogen receptors [14]. Ineffectiveness of estrogen on the proliferation of pituitary cells in STZ-treated mice may result from altered mechanism of estrogen receptors. In the previous study using STZ-treated or alloxan- treated diabetic rats, the response of uterine epithelial cells to estrogen on the proliferation was significantly reduced, and this was restored by insulin treatment [15]. Their result does not agree with our result. They had used the lower dose of estradiol-17 8 (4 ug/100 g body weight) compared with the dose of estrogen used in the present study. As we clearly found the diminished response in pituitary cells with this estrogen dose, one possible reason for this discrepancy is that the uterine epithelial cells in rats may be more responsive to estrogen than the pituitary cells. Estrogen receptor kinetics and estrogen activity for protein synthesis were altered in the uteri of STZ-induced diabetic rats, and restored by insulin treatment [5]. Thus, STZ treatment in the rat more severely may affect the estrogenic mechanism in the uterus. We preliminarily found that insulin and IGF-I stimulated the proliferation of mouse uterine epithelial cells in vitro (Taka- hashi and Miyake, unpublished observation). Further study on insulin or IGF-I action on the mouse uterine cells is needed. Chronic estrogen administration increases pituitary weights, which mainly results from the hypertrophy and hyperplasia of PRL cells [29]. Gala and Jaques [6] demon- strated that STZ-induced insulin deficiency retarded estrogen-induced pituitary growth in the rat. The retarded growth of pituitary glands in STZ-treated rats is thought to be partly due to the diminished mitotic activity of pituitary cells, since the lower mitotic activity of pituitary cells including PRL cells in STZ-treated mice was shown in the present study. In conclusion, the present in vivo study clearly showed that insulin administration increased the cell proliferation of pituitary cells. In pituitary cells insulin was required for estrogen-induced cell proliferation. Molecular basis of in- sulin-estrogen interaction must be studied using the in vitro system. ACKNOWLEDGMENTS This study was supported in part by the Ryoubiteien Founda- tion. REFERENCES 1 Bach MA, Bondy CA (1992) Anatomy of the pituitary insulin- like growth factor system. Endocrinology 131: 2588-2594 2 Beck CA, Garner CW (1992) Stimulation of DNA synthesis in rat uterine cells by growth factors and uterine extracts. Mol Cell Endocrinol 84: 109-118 3 Borgundvaag B, Kudlow JE, Mueller SG, George SR (1992) Dopamine receptor activation inhibits estrogen-stimulated trans- forming growth factor- a gene expression and growth in anterior pituitary, but not in uterus. Endocrinology 130: 3453-3458 4 Carbajo-Peréz E, Watanabe YG (1990) Cellular proliferation in the anterior pituitary of the rat during the postnatal period. Cell Tissue Res 261: 333-338 5 EkkaE, Vanderheyden I, De Hertogh R (1984) Normalization of estradiol receptor kinetics and hormonal activity in uterus of streptozotocin-induced diabetic rats treated with insulin. En- docrinology 114: 2271-2275 6 Gala RR, Jaques S, Jr (1979) The influence of estrogen on pituitary growth and on prolactin production in vitro in the diabetic rat. Proc Soc Exp Biol Med 161: 583-588 7 Ghahary A, Murphy LJ (1989) Uterine insulin-like growth factor-I receptors: regulation by estrogen and variation through- out the estrous cycle. Endocrinology 125: 597-604 8 Ghahary A, Chakrabarti S$, Murphy LJ (1990) Localization of the sites of synthesis and action of insulin-like growth factor-I in the rat uterus. Mol Endocrinol 4: 191-195 9 Ghosh D, Danielson KG, Alston JT, Heyner S (1991) Func- tional differentiation of mouse uterine epithelial cells grown on collagen gels or reconstituted basement membranes. In Vitro Cell Dev Biol 27A: 713-719 10 Goodyer CG, Lucie de Stéphano, Wei Hsien Lai, Guyda HJ, Posner BI (1984) Characterization of insulin-like growth factor receptors in rat anterior pituitary, hypothalamus, and brain. Endocrinology, 114: 1187-1195 11 Hayashi I (1984) Growth of GHsg, a rat pituitary cell line, in serum-free, hormone-supplemented medium, In “Methods for 12 13 14 15 16 17 18 19 20 21 22 Proliferation of Mouse Pituitary Cells 449 Serum-Free Culture of Cells of the Endocrine System” Ed by DW Barnes, DA Sirbasku, GH Sato, Alan R. Liss, Inc, New York, pp 1-13 Ikawa H, Irahara M, Matsuzaki T, Saito S, Sano T, Aono T (1992) Impaired induction of prolactin secretion from the anterior pituitary by suckling in streptozotocin-induced diabetic rat. Acta Endocrinol 126: 167-172 Keech CA, Gutierrez-Hartmann A (1991) Insulin activation of rat prolactin promoter activity. Mol Cell Endocrinol 78: 55-60 Keefer DA (1980) In vivo estrogen uptake by individual cell types of the rat anterior pituitary after short-term castration- adrenalectomy. Cell Tissue Res 209: 167-175 Kirkland JL, Barrett GN, Stancel GM (1981) Decreased cell division of the uterine luminal epithelium of diabetic rats in response to 17 8 -estradiol. Endocrinology 109: 316-318 Kudlow JE, Kobrin MS (1984) Secretion of epidermal growth factor-like mitogens by cultured cells from bovine anterior pituitary glands. Endocrinology 115: 911-917 Michels KM, Lee W-H, Seltzer A, Saavedra JM, Bondy CA (1993) Up-regulation of pituitary ['7°I] insulin-like growth fac- tor-I (IGF-I) binding and IGF binding protein-2 and IGF-I gene expression by estrogen. Endocrinology 132: 23-29 Murphy LJ, Bell GI, Friesen HG (1987) Tissue distribution of insulin-like growth factor I and II messenger ribonucleic acid in the adult rat. Endocrinology 120: 1279-1282 Nelson KG, Takahashi T, Bossert NL, Walmer DK, McLachlan JA (1991) Epidermal growth factor replaces estrogen in the stimulation of female genital-tract growth and differentiation. Proc Natl Acad Sci USA 88: 21-25 Nelson KG, Takahashi T, Lee DC, Luetteke NC, Bossert NL, Ross K, Eitzman BE, McLachlan JA (1992) Transforming growth factor- @ is a potential mediator of estrogen action in the mouse uterus. Endocrinology 131: 1657-1664 Olchovsky D, Bruno JF, Gelato MC, Song J, Berelowitz M (1991) Pituitary insulin-like growth factor-I content and gene expression in the streptozotocin-diabetic rat: evidence for tissue- specific regulation. Endocrinology 128: 923-928 Sakuma S, Shirasawa N, Yoshimura F (1984) A histometrical study of immunohistochemically identified mitotic adenohy- 23 24 25 26 27 28 29 30 31 32 33 34 pophysial cells in immature and mature castrated rats. docrinol 100: 323-328 Shirasawa N, Yoshimura F (1982) Immunohistochemical and electron microscopical studies of mitotic adenohypophysial cells in different ages of rats. Anat Embryol 165: 51-61 Sirbasku DA (1978) Estrogen induction of growth factors specific for hormone-responsive mammary, pituitary, and kidney tumor cells. Proc Natl Acad Sci USA 75: 3786-3790 Stanley FM (1992) An element in the prolactin promoter mediates the stimulatory effect of insulin on transcription of the prolactin gene. J Biol Chem 267: 16719-16726 Starus DS (1984) Growth-stimulatory actions of insulin in vitro and in vivo. Endocr Rev 5: 356-369 Sutherland RL, Watts CKW, Clarke CL (1988) Oestrogen action. In “Hormones and their actions, Part I” Ed by BA Cooke, RJB P King, HJ van der Molen, Elsevier Science Publishers BV, Amsterdam pp 197-215 Takahashi S, Okazaki K, Kawashima S. (1984) Mitotic activity of prolactin cells in the pituitary glands of male and female rats of different ages. Cell Tissue Res 235: 497-502 Takahashi S, Kawashima S (1987) Proliferation of prolactin cells in the rat: effects of estrogen and bromocriptine. Zool Sci 4: 855-860 Tesone M, Ladenheim RG, Charreau EH (1985) Alterations in the prolactin secretion in streptozotocin-induced diabetic rats. Correlation with pituitary and hypothalamus estradiol receptors. Mol Cell Endocrinol 43: 135-140 Tomooka Y, DiAugustine RP, McLachlan JA (1986) Prolif- eration of mouse uterine epithelial cells in vitro. Endocrinolo- gy 118: 1011-1018 Webster J, Scanlon MF (1991) Growth factors and the anterior pituitary. Bailliere’s Clin Endocrinol Metab 5: 699-726 Weisenberg L, Fridman O, Libertun C, De Nicola AF (1983) Changes in nuclear translocation of estradiol-receptor complex in anterior pituitary and uterus of rats with streptozotocin diabetes. J Steroid Biochem 19: 1737-1741 Yamauchi K, Shiino M (1986) Pituitary prolactin cells in diabetic rats induced by the injection of streptozotocin. Exp Clin Endocrinol 88: 81-88 J En- ZOOLOGICAL SCIENCE 11: 451-454 (1994) © 1994 Zoological Society of Japan Molecular Evolution of Shark C-type Natriuretic Peptides MASAYOSHI TAKANO!, YUICHI SASAYAMA~ and YosHio TAKEr Ocean Research Institute, University of Tokyo, Minamidai, Nakano, Tokyo 164 and *Department of Biology, Faculty of Science, Toyama University, Gofuku, Toyama 930, Japan ABSTRACT— C-type natriuretic peptides (CNP) of varying length were isolated from the atrium or ventricle of a shark, Lamna ditropis and their amino acid sequences were determined. Although the sequence of Lamna CNP was highly homologous to those of other CNPs sequenced to date, the Lamna CNP-41, the longest CNP identified in this study, has one amino acid replacement from those of Triakis scyllia and Scyliorhinus canicula, and three amino acid replacements from that of Squalus acanthias. The degree of similarity of CNP molecules coincides well with their systematic positions in the cladogram of elasmobranchs; Lamna, Triakis and Scyliorhinus belong to the same order, but Lamna and Squalus belong to different orders. The facts that Lamna and Triakis are in different suborders but Triakis and Scyliorhinus are in the same suborder and have identical CNP-41, also support this evolutionary implication. INTRODUCTION C-type natriuretic peptide (CNP) is a member of the natriuretic peptide family first identified in the brain of pig and teleost fishes [2, 4, 9]. In contrast to other members of the peptide family, namely atrial (A-type), B-type and ven- tricular natriuretic peptides (ANP, BNP and VNP) which are cardiac hormones circulating in the blood, CNP has been isolated from the brain in all species from teleost to mam- mals, and its plasma and cardiac concentrations are too low to be detected in mammals [8]. Thus CNP is regarded as a neuropeptide in mammals. However, we have isolated CNP from the heart of two species of dogfish shark, Triakis scyllia and Scyliorhinus canicula [5, 6]. In these fish, plasma and cardiac concentrations of CNP are extremely high, and other cardiac natriuretic peptides, ANP, BNP and VNP, are not identified in their hearts [7]. Furthermore, only CNP cDNA has been cloned from the cDNA library of the heart of spiny dogfish, Squalus acanthias [3]. Therefore it is likely that CNP is the only natriuretic peptide present in elasmobranchs. It is also noted that the amino acid sequence of CNP is more conserved than any other natriuretic peptides, namely ANP, BNP and VNP [8]. Thus, CNP might be an ancestral molecule of the natriuretic peptide family, and other mem- bers might be reproduced by gene duplication. As a prototype of the natriuretic peptide family, it seems of interest to examine chemical evolution of the CNP mole- cule. In previous studies, we have found that amino acid sequences of CNP-22, a mature form stored in the brain, of Triakis and Scyliorhinus are identical, and even proCNP differs in only 3 out of 115 amino acid residues [5,6]. Accepted June 7, 1994 Received April 4, 1994 " Present address: Research Institute, Zenyaku Kogyo Co. Ltd., Ohizumimachi, Nerima, Tokyo 178, Japan > To whom all correspondence should be addressed However, Squalus CNP-22 predicted from the cDNA sequ- ence differs from that of Triakis in 2 amino acid residues, and the difference was much greater at the level of prohormone [3]. Systematically, Triakis and Scyliorhinus belong to the same suborder Scyliorhinoidei, but Squalus is different from the two species at the level of order [1]. We recently have obtained the heart of Lamna ditropis. This fish belongs to the order Lamniformes as do Triakis and Scyliorhinus, but to the suborder different from those sharks. Therefore, we attempted in the present study to isolate CNP from the Lamna heart and to compare its structure with those of other sharks. MATERIALS AND METHODS Isolation of CNP The shark, Lamna ditropis, of approximately 3 m in body length was caught in Toyama Bay and was obtained from fishermen 5 h after capture. The heart was immediately dissected out, the atrium and ventricle separated, and frozen in a deep freezer at —50°C. The atrium (106.4 g) and ventricle (333.2 g) were treated separately. ANP-like peptides in the heart were isolated with protocols described previously [5]. The frozen tissues were crushed in a pulverizer, boiled in 5 volumes (atrium) or 3 volumes (ventricle) of water for 10 min, acidified with AcOH to a concentration of 1 M, and homoge- nized in a Polytron homogenizer (Kinematika, Germany) for 90 sec at maximum speed. The homogenate was centrifuged at 16,000xg for 30 min at 4°C. The supernatant was added to 2 volumes of cold acetone, and centrifuged at 16,000xg for 30min at 4°C. The supernatant was evaporated, reconstituted in 30 ml of 1M AcOH, and added to 2 liters of cold acetone. After centrifugation, the pellet was dissolved in 30 ml of 1M AcOH, and applied onto a column (585 cm) of Sephadex G-25 fine (Pharmacia, Sweden) for desalting. The fractions which contain molecules with Mr>ca. 2,000 were applied onto a column of SP-Sephadex C-25 (1.620 cm), and adsorbed materials were eluted successively with 150 ml each of 1M AcOH, 2M pyridine, and 2 M pyridine-AcOH, pH 5.0. Each fraction was evaporated and assayed for relaxant activity in the 452 M. TAKANO, Y. SASAYAMA AND Y. TAKEI chick rectum as described below. Bioactive fractions were sub- jected to cation-exchange high performance liquid chromatography (HPLC) in an IEC-CM column (7.575 mm, Jasco, Japan). Each bioactive fraction was then subjected to reverse-phase HPLC in an ODS-120T column (4.6X250mm, Tosoh, Japan) with different gradients of CH3CN concentrations. The detailed chromatographic conditions are described in the legend of each figure. The purified material was subjected to amino acid sequencing in a protein sequencer (477A, Applied Biosystems, USA). Validity of the ami- no acid sequence was examined by mass spectrometry (JMS-HX110, JEOL, Japan). ANP-like activity was assayed at each step of purification using a relaxant activity in the chick rectum [10]. New-born male chicks were purchased from Kanagawa Poultry Cooperation (Yokohama) and reared under a infra-red lamp with free access to food and water. The chick was decapitated, rectum immediately isolated, and set up in a trough whose temperature was controlled at 37°C. The rectum — c fe) oO N s oO (3) c ov a ° no He} < O 100 200 Fraction number 1.5 E {= ° oO N ic Oo 4 oO c oO 2 ° 7) Bo} * 05 10 20 30 40 Fraction number was precontracted with 2 x 10° M carbachol (Sigma, USA), and the relaxation was quantified by a displacement transducer connected to a transducer amplifier (1829 and 45347, NEC-Sanei, Japan). ANP- like activity was expressed as equivalents to eel ANP which was used as standard. RESULTS Same molecules were isolated from atrial and ventricular extracts. After Sephadex G-25 chromatography, fractions of 1-70, which contain molecules larger than CNP-22 [5], were pooled and subjected to SP-Sephadex C-25 chroma- tography (Fig. la). Since only the fraction eluted with pyri- dine-AcOH exhibited rectum-relaxant activity, this fraction was subjected to cation-exchange HPLC. The bioactive 0.47 E = ° Ry 760 = - oO _ ® & 3) ec 02+ a Oo = a fetes 5 3) = oD 2 a < oie 2 os = o = fo) = 1 * 0 20 40 Time (min) ob E c s is) & nN x N a ~ is Pe 40 o o 2 20.05 fe) 0.05} a oO § | | SS ) . I 420 g o peal ea] | — 2 th) = < i = = scitian atoll 208 s ae hal V 205 E ae c = Ne <= ” ra a4 440 = ” 3 we 2 aga a s=s om = o O2E = Ca z —lL o < F 0 o 0 0 40 60 Time (min) Fic. 1. Purification of C-type natriuretic peptide (CNP) from Lamna atrium. Solid columns represent relaxant activity in the chick rectum expressed as equivalents to eel atrial natriuretic peptide (eANP). marked by square bracket were subjected to SP-Sephadex C-25 chromatography. A: Sephadex G-25 chromatography of crude atrial extract. Fractions B: cation-exchange high-performance liquid chroma- tography (HPLC) of the fraction eluted with pyridine-AcOH in SP-Sephadex C-25 chromatography. Broken lines show gradient of solvent B (1M NH,OAc : CH;CN=9:1) against solvent A (10 mM NH,OAc : CH3;CN=9: 1). CNP-29, CNP-38 and CNP-41 were recovered, respectively, from fractions marked with bracket 1, 2, and 3. C and D: reverse-phase HPLC of fraction 37 of panel B and a fraction with bioactivity in panel C, respectively. Sequence analysis of bioactive peak in panel D revealed that the peak is that of CNP-29. Broken lines show gradient of CH3CN concentrations. Shark CNP 453 principle was purified only from fractions 31, 32 and 37 of cation-exchange HPLC, although bioactivity was also noted in other fractions (Fig. 1b). A rectum-relaxant principle was isolated from fraction 32 by two steps of reverse-phase HPLC (Figs. 1c, d). Final yield was 2 nmol equivalent to eel ANP as determined by the rectum-relaxant activity and 363 pmol equivalent to eel ANP as determined by absorbance at 220 nm. Sequence analysis of 3/4 of the purified peptide re- vealed that the sequence was H-Phe-Lys-Gly-Arg-Ser-Lys- Lys-Gly-Pro-Ser-Arg-Gly-(Cys)-Phe-Gly-Val-Lys-Leu-Asp- Arg-Ile-Gly-Ala-Met-Ser-Gly-Leu-Gly-(Cys)-OH (Fig. 2). The presence of two cysteine residues was deduced from the similarity to other CNPs thus far sequenced. Thus the peptide was named Lamna CNP-29. The sequence was confirmed by mass spectrometry using the remaining 1/4. A CNP with 12 amino acid residues (Arg-Leu-Leu-Lys-Asp- Leu-Ser-Asn-Asn-Pro-Leu-Arg-) elongated from the N- terminus of CNP-29 was isolated from fraction 37 and thus named Lamna CNP-41. The final yield was 260 pmol as judged by absorbance at 220nm. Sequence analysis of 3/4 of the purified peptide could determine only 28 amino acid residues from the N-terminus. However, it was apparent that the peptide had longer sequence and terminated with the second-half cysteine at the C-terminus, because mass analysis calculated the MH* of 4433 which coincides well with the average mass of predicted sequence of CNP-41 (Mr=4432.3). CNP-38 was also isolated from fraction 31 with the final yield of 312 pmol. Although many other fractions showed bioac- tivity, no bioactive principle could be isolated from those fractions. 200 - ) [o) [s) oa oO Yield of PTH-amino acid ( pmol R @) Wb te i Nn en eee | ee S) 15 25 Cycle number Fic. 2. The yield of phenylthiohydrantoin-derivatized (PTH) amino acid at each cycle of Edman degradation in the sequence analysis of CNP-29. No PTH amino acid was detected at 13th and 29th cycle. The presence of cysteine residues, denoted by (C), was estimated at these cycles from analogy to other CNPs and from the result of mass spectrometry. DISCUSSION We isolated three short forms of CNP from the heart of Lamna ditropis in the present study. In previous attempts to isolate ANP-like peptides from the heart of other sharks, Triakis scyllia and Scyliorhinus canicula, large amounts of proCNP and small amounts of CNP-38 and CNP-39 were isolated [5, 6]. CNP-38 was also isolated in this study, but Class Chondrichthyes Subslass ek eae Superorder ascot Order Lamniformes Squaliformes Suborder Lamnoidei Scyliorhinoidei Squaloidei Family been’ Scyliorhinidae Carcharhinidae sail Subfamily ee ee Scyliorhininae sale cate sr gals Scyliorhinus= Triakis inal yete eocces ON AOQOOCOOQMHOCHHBEMOHHOHOMHS= Fic. 3. Phylogenetic cladogram of 4 species of sharks and their amino acid sequences of CNP-41. The cladogram was depicted after Nelson [1]. Amino acid residues different from Triakis or Scyliorhinus CNP are shaded. 454 M. TAKANO, Y. SASAYAMA AND Y. TAKEI proCNP and CNP-39 were not identified. Instead, CNP-29 and CNP-41 were recovered from the Lamna heart. Several peaks with bioactivity, which may contain other fragments of CNP, were also identified after ion-exchange HPLC. This may indicate that a different processing system is operating in the Lamna heart, or the shorter forms are degradation products of proCNP. The latter is more likely because it took longer to freeze the Lamna heart after its death. In previous studies using Triakis and Scyliorhinus, hearts were frozen on dry ice immediately after isolation from anesthe- tized fish. In addition to Triakis and Scyliorhinus CNP, CNP cDNA has been cloned from the heart of Squalus acanthias [3]. Comparison of the amino acid sequence of CNP-41 between Triakis and other sharks revealed that Triakis CNP-41 is identical to that of Scyliorhinus, is different by one amino acid residue from that of Lamna, and is different by four amino acid residues from that of Squalus (Fig.3). As also shown in Figure 3, Lamna, Triakis and Scyliorhinus belong to the same order (Lamniformes) but Squalus is in a different order [1]. Triakis and Scyliorhinus are the same even at the level of suborder, whereas Triakis and Lamna are in different suborders. It is of interest to note, therefore, that the chemical evolution of CNP molecule is closely related to the cladogram of cartilaginous fishes which is drawn based on the morphological proximity (Fig. 3). During the course of purification, we utilized relaxant activity in the chick rectum as an assay system. We found that the final yield of CNPs quantified by this assay was always much greater than that deduced from absorbance at 220nm. It seems therefore that the shark CNP has much greater relaxant activity than eel ANP which was used as standard for the assay. ACKNOWLEDGMENTS The authors are grateful to Dr. Hideshi Kobayashi of Zenyaku Kogyo Co. Ltd. for critical reading of the manuscript, and to Dr. T. Takao and Professor Y. Shimonishi of the Institute for Protein Research, Osaka University for performing mass spectrometry. We also thank Dr. T. Sato of Misaki Marine Biological Station, Universi- ty of Tokyo for his advice on the classification of elasmobranchs and Mr. H. Kaiya and K. Ukawa of Toyama University for their help in the initial part of purification. This investigation was supported in part by grants from the Ministry of Education, Science and Culture of Japan, from the Fisheries Agency of Japan and from Zenyaku Kogyo Co. Ltd. REFERENCES 1 Nelson JS (1984) Fish of the World. John Willey & Sons, New York, 2nd ed, pp 49-59 2 Price DA, Doble KE, Lee TD, Galli SM, Dunn BM, Parten B, Evans DH (1990) The sequencing, synthesis, and biological actions of an ANP-like peptide isolated from the brain of the killifish, Fundulus heteroclitus. Biol Bull 178: 279-285 3 Schofield JP, Jones DSC, Forrest Jr JN (1991) Identification of C-type natriuretic peptide in heart of spiny dogfish shark (Squalus acanthias). Am J Physiol 261: F734-F739 4 Sudoh T, Minamino N, Kangawa K, Matsuo H (1990) C-type natriuretic peptide (CNP): A new member of natriuretic peptide family identified in porcine brain. Biochem Biophys Res Com- mun 168: 863-870 5 Suzuki R, Takahashi A, Hazon N, Takei Y (1991) Isolation of high-molecular-weight C-type natriuretic peptide from the heart of a cartilaginous fish (European dogfish, Scyliorhinus canicula). FEBS Lett 282: 321-325 6 Suzuki R, Takahashi A, Takei Y (1992) Different molecular forms of C-type natriuretic peptide isolated from the brain and heart of an elasmobranch, Triakis scyllia. J Endocrinol 135: 317-325 7 Suzuki R, Togashi K, Ando K, Takei Y (1994) Distribution and molecular forms of C-type natriuretic peptide in tissues and plasma of an elasmobranch, Triakis scyllia. Gen Comp Endoc- rinol (in press) 8 Takei Y, Balment RJ (1993) Natriuretic factors in non- mammalian vertebrates. In “New Insights in Vertebrate Kid- ney Function” Ed by JA Brown, RJ Balment, JC Rankin, Cambridge University Press, Cambridge, pp 351-385 9 Takei Y, Takahashi A, Watanabe TX, Nakajima K, Sakakibara S, Takao T, Shimonishi Y (1990) Amino acid sequence and relative biological activity of a natriuretic peptide isolated from eel brain. Biochem Biophys Res Commun 170: 883-891 10 Watanabe TX, Noda Y, Chino N, Nishiuchi Y, Kimura T, Sakakibara S, Imai M (1989) Structure-activity relationships of alpha-human atrial natriuretic peptide. Eur J Pharmacol 147: 49-58 ZOOLOGICAL SCIENCE 11: 455-465 (1994) © 1994 Zoological Society of Japan Systematic Study on the Chaenogobius Species (Family Gobiidae) by Analysis of Allozyme Polymorphisms TAKESHI AIZAWA, MAcHIKO HatTsumi* and KEN-IcHI WAKAHAMA Department of Biology, Faculty of Science, Shimane University 1060 Nishikawatsu-cho, Matsue 690, Japan ABSTRACT—Isozyme polymorphisms of eight Chaenogobius species (Family Gobiidae) were studied in order to understand the phylogenetic relationships between them. Fifteen loci from ten enzymes and sarcoplasmic protein were detected by horizontal starch gel electrophoresis. The phylogenetic tree obtained from genetic distances between species essentially agreed with morphological and ecological studies previously reported. In addition, two new features were revealed. First, Nei’s genetic distance between C. Jaevis from Saitama and that from Akita suggested that they are differentiate from each other on the species level. Second, the undescribed taxon, C. sp., from Lake Shinji was closely related to C. laevis from Akita and the genetic distance between them was 0.114. The smallest genetic distance between distinct species of Chaenogobius was 0.103 between C. urotaenia and C. isaza obtained in this study. This shows the possibility that C. sp. is the different species of C. Jaevis, from Akita. INTRODUCTION The family Gobiidae accomplished distinctive adaptive radiation. Each species adapted to various environments and has various life histories. This makes Gobiidae an excellent material for studying the mechanisms of evolution and speciation in Pisces. Chaenogobius, one genus of Gobiidae, consists of several species members that have various life histories, namely, species adapted in brackish water or fresh water, marine species, amphidromous species and land-locked species. The ecological and evolutionary genetic studies of this genus may especially supply us useful information on evolutionary process of adaptation of fishes. The phylogenetic study on the members of this species has been carried out, but not completed, yet [23]. Takagi [31] demonstrated that C.urotaenia (Hilgendorf) and C. castaneus (O’Shaughnessy) are different species. Morpholo- gical and ecological studies showed three types of C. uro- taenia, one lives in freshwater, another lives in brackish water and the other lives in the middlereach type, and now these types are recognized as distinct species and called C.urotaenia (Japanese name, Ukigori), C.sp.2 (Japanese name, Shi- maukigori) and C.sp.1 (Japanese name, Sumiukigori), re- spectively [1, 7, 15, 31, 32]. Takagi [33] discriminated C. laevis (Steindachner) from C. castaneus based on its morphol- ogy. Chaenogobius castaneus has three pairs of pit organs connected by sensory canals and lives in brackish water, whereas C. /aevis has no canal system and lives in freshwater. His study was followed by the description of new morph of Chaenogobius species (Japanese tentative name; Shinjiko- haze by Koshikawa [12]) from Lake Shinji in Japan [13]. This morph lives in brackish water of lower salt condensation Accepted May 10, 1994 Received November 20, 1994 * To whom all correspondence should be addressed. than C. castaneus lives. It is similar to C. laevis for its color pattern and nuptial color of female but different from C. laevis and C.castaneus since shinjiko-haze has two pairs of pit organs connected by sensory canals on its head. This open- ing pattern of sensory canals is quite similar to C. taranetzi Pinchuk distributing Ussuri Bay in Russia and North Korea [9,23]. From these facts there are three possibilities of phylogenetic position of the species from Lake Shinji; name- ly, this species is one geographic variation of C. laevis or C. taranetzi, this is different species of C. laevis or C. taranetzi, and this species and C. taranetzi are the geographic variation of C. laevis. In two decades of population biology, it was turned out that molecular approach is powerful to study phylogenetic relationships of various organisms [6, 18, 22, 27, 35]. Iso- zyme polymorphisms detected by electrophoresis supply use- ful measurement of genetic differentiation between popula- tions or species, in terms of, the genetic distance [16, 17]. Many groups of species were studied on isozyme polymorph- isms [24]. Accumulated data revealed that there were levels of genetic differentiation between local populations, subspe- cies, species or genus [6, 16, 17, 22, 24]. Using these levels, a phylogenetic tree could be constructed by cluster analysis from genetic distances [17, 2,3, 4,5]. For the species under consideration, electrophoretic studies should supply useful information on the phylogenetic relationships of Chaenogo- bius species. In this study we analyzed allozyme polymorphisms of eight members of Chaenogobius species, six taxa mentioned above, C. isaza Tanaka, endemic to Lake Biwa in Japan, and C. heptacanthus (Hilgendorf), marine species, in order to clarify the phylogenetic relationships of this group and to characterize the species from Lake Shinji genetically. 456 T. Aizawa, M. Hatsumi AND K. I. WAKAHAMA MATERIALS AND METHODS Animal sampling Materials used in this experiment were C. castaneus, C. laevis, C. heptacanthus, C. urotaenia, C. sp. 1, C. sp. 2, C. isaza and an undescribed taxon collected from Lake Shinji , Shimane Prefecture in Japan. Since we could not conclude the undescribed taxon from Lake Shinji as C. laevis or C. taranetzi, we called this as C. sp. here. Table 1 shows the collection data of materials and Figure 1 shows the sites of collection. Six populations were kindly supplied by others, or, C. sp. 1 from Daitobetsu river by Dr. A. Goto, Faculty of Fisheries, Hokkaido University, C. laevis from Lake Hachiro-gata by Mr. K. Shibuya, Akita Prefectural Fisheries Consulting Center, C. laevis from Koma river by Mr. A. Iwata, Akasaka Imperial Palace, C. castaneus from Tama river by Mr. I. Kimoto, Tokyo Metropolitan Fisheries Experiment Station, C. isaza from Lake Biwa by Dr. S. Takahashi and Mr. S. Matsuoka, Shiga Prefecture and C. heptacan- thus from Lake Nakaumi, Mawatashi by Mr. T. Kawashima, Mitoya Inland Water Fisheries Branch, Shimane Prefectural Fisheries Ex- periment Station. Sample preparation for electrophoresis Samples of fishes were stored at —25°C before dissection. Liver and lateral muscle were dissected out from each individual melted on ice. Three times or same amount of distilled water was added to liver or muscle, respectively, and homogenized in a microcentrifuge tube by plastic homogenizer on ice. The sample was centrifuged at 10000 g for 15min. at 4°C. The supernatant was absorbed by capilalies and stored at —25°C until electrophoresis. Individuals from which the tissues were removed were fixed in 10% formaldehyde. Identification of three species, C. castaneus, C. laevis and C. sp. was made by their color patterns after fixation and patterns of sensory canals, according to the method by Takagi [34]. Electrophoresis Ten different enzymes and sarcoplasmic protein prepared from the species were analyzed by horizontal starch gel electrophoresis (Table 2). Two buffer systems were used in this experiment. One is citrate-aminopropyl morphorine buffer [8] and the other is citrate- tris buffer [21]. The staining methods of enzymes used were de- scribed by Shaw and Prasad [29] or Selander et al. [28]. Gels were dried between serophan to form films [20] and the isozyme patterns were documented on the films. When one enzyme had two loci, each locus was numbered in order of lower mobility to the anode. TaBLE 1. Collection data of materials Species 1p Locality, prefecture Date N= C. castaneus 1. Lake Nakaumi (Shimo-itou), Shimane Dec. 1989 55 2. linashi River, Shimane Nov. 1990 20 3. Lake Shinji (Matsue-Onsen), Shimane Jan. 1990 20 4. Lake Shinji (Hamasada), Shimane Jan. 1990 14 5. Lake Shinji (Tamayu), Shimane Jan.-Feb. 1990 5 6. Tama River, Tokyo Oct. 1989 C. sp. 7. Lake Shinji (Matsue-Onsen), Shimane Jan.-Feb. 1990 20 8. Lake Shinji (Hamasada), Shimane Jan. 1990 21 9. Lake Shinji (Tamayu), Shimane Jan.-Feb. 1990 34 C. laevis 10. Lake Hachirou-gata, Akita Dec. 1989 75 11. Koma River, Saitama Oct. 1990 20 C. heptacanthus 12. Lake Nakaumi (Mawatashi), Shimane Dec. 1989 34 C. urotaenia 13. Shinshi River, Shimane Feb.-Mar. 1990 5 14. Satoji River, Shimane Apr. Dec. 1990 4 15. Fukaura River, Shimane Dec. 1990 2 16. Motoya River, Shimane Dec. 1990 3 C. sp. 1 17. Daitobetsu River, Hokkaido Nov. 1990 20 C. sp. 2 18. Shinshi River, Shimane Feb.-Mar. 1990 13 19. Satoji River, Shimane Apr. 1990 10 20. Fukaura River, Shimane Dec. 1990 8 21. Ujiki River, Shimane Dec. 1990 10 22. Oku River, Shimane Jan. 1991 10 C. isaza 23. Lake Biwa, Shiga Jan.-Feb. 1991 45 P*; Population number. N**; No. of individuals used for analysis. Biochemical Systematic Study on Chaenogobius 457 Fic. 1. The map of Japan showing sampling localities. Each num- ber is corresponding to the population number in Table 1. RESULTS Fifteen loci were postulated in ten enzymes and sarco- plasmic protein from electrophoretic morph on starch gels among eight taxa of the genus Chaenogobius. Examples of zymograms were shown in Figure 2. Two types of bands were observed in one of the sarcoplasmic proteins. One type showed a single band and the other type showed double bands (Fig 2d, sp-1). Since each species had one of these types alternatively, double bands should not be heterozyotes and may be originated from gene duplication or post- translational modification. We named b to the allele of single band and a to the allele of double bands. One locus, the sarcoplasmic protein 2, was monomorphic and the Ldh locus was also monomorphic with one exception- al individual from the Akita populations of C. laevis (Fig. 2b) but other loci had more than two alleles. Table 3 shows the allelic frequencies at thirteen loci on each population. Each species had fixed allele at five loci, namely, Aat, Ck, Me-1, Me-2 and Sp-1. In these loci Ck, Me-1 Me-2 and Sp-I had only two variants. One type of variants was shared by C. castaneus, C. sp., C. laevis, C. heptacanthus and the other type was shared by C. urotaenia, C. sp. 1, C. sp.2 and C. isaza. Other loci were polymorphic in some species. More than six alleles were observed at the a-Gpd, Gpi-2, Mdh, and Pgm loci. Twenty three populations were polymorphic for any of these four loci. Related species and populations were compared to each other for alleles of these four loci. At the a-Gpd locus, allele f was shared by C. sp. and the Akita population of C./aevis whereas the Saitama population of C. laevis had allele d. In C. castaneus, allele e of this locus was common. Two populations of C. castaneus had allele b at low frequencies while allele b was fixed in C. heptacanthus and C. sp. 1 and was at high frequency in C. urotaenia. C. sp.2 had allele a and C. isaza had allele c at the a -Gpd locus. At the Gpi-2 locus, allele c was observed at high frequency in C. castaneus, C. sp. and C. laevis from Saitama while allele b was observed at high frequency in C. Jaevis from Akita and C. heptacanthus. C. urotaenia, C. sp.1 and C. sp. 2 had allele d at high frequency and allele f at low frequency. C. isaza had two alleles, d and f, at equal frequency at the Gpi-2 locus. At the Mdh locus allele e and allele f were shared by C. castaneus and C. laevis but frequency was different in species. C. sp. had allele e and allele a but not allele f. C. urotaenia, C. sp. 1, C. sp. 2, C. heptacanthus and C. Isaza were monomorphic for the Mdh locus. At the Pgm locus allele e was common in C. castaneus, C. sp. and C. laevis from Akita whereas C. /aevis from Saitama did not have allele e but had allele c and allele b. C. wrotaenia, C. sp. 1, C. sp. 2 and C. isaza shared allele c at the Pgm locus at high frequency with other variants at low frequency. C. heptacanthus had allele d and allele f specifically. The Gpi-/ locus had two alleles, a and b. In almost all species, allele b was fixed. At the Gpi-J locus allele a was TABLE 2. The list of Enzymes and proteins detected Enzyme and protein (Abbreviation) E.C.Number Tissue* Buffer** Asparate aminotransferase (AAT) 2.6.1.1 M CT Creatine Kinase (CK) Del Bo M CT a -glycerophosphate dehydrogenase (a -GPD) 1.1.1.8 M CT Glucosephosphate isomerase (GPI) 5.3.1.9 M CT Isocitrate dehydrogenase (IDH) 1.1.1.42 M, L APM Lactate dehydrogenase (LDH) 1.1.1.27 M APM Malate dehydrogenase (MDH) 1.1.1.37 L APM Malic enzyme (ME) 1.1.1.40 M CT Phosphoglucomutase (PGM) DV SoIl M APM Superoxide dismutase (SOD) 1.15.1.1 L APM Sarcoplasmic protein (SP) — M APM *; M means muscle and L means liver. **; AMP means citrate, aminopropyl mophorine buffer and CT means citrate-tris buffer. 458 T. Aizawa, M. Hatsumi AND K. I. WAKAHAMA a MD H (liver) b LDH Fic. 2. Zymograms of four enzyme systems and a sacroplasmic protein. Thick arrows indicate origin. only found in C. sp. and C. sp.1. The /dh-J locus had four alleles, the Idh-2 had three alleles and the Sod had four alleles. These three loci, however, were not so polymorphic as Other loci. Nei’s genetic distances [16] were calculated from the allele frequencies. Table 4 gives the matrix of average minimal and maximal genetic distances between each pair of species. The genetic distances between populations within species were 0 to 0.018 except 0.194 between the Akita and Saitama populations of C. laevis. This made us to list average genetic distances of each population of C. laevis separately. The genetic distances between species ranged from 0.092 between C. urotaenia and C. isaza to 1.595 between C. heptacanthus and C. sp. 2. Biochemical Systematic Study on Chaenogobius 459 Divergence Time (MYB) 5 (0) . Castaneus sp. laevis, Akita . laevis, Saitama . heptacanthus isaza - urotaenia sp. 2 CEG oO LOMONOnorono - sp. 1 1.5 1.0 0.5 (0) Genetic Distance Fic. 3. Phylogenetic tree showing the relationships among eight species of genus Chaenogobius based on values of genetic distance. A phylogenetic tree was constructed from genetic dis- tances by average distance method (UPGMA) devised by Sneath and Sokal [30] and modified by Nei [17]. Figure 3 shows the phylogenetic tree of nine taxa of Chaenogobius constructed from average genetic distances in this study. The tree was not different topologically from the tree con- structed from genetic distances between populations. Eight species were divided into two groups. One was C. castaneus group including C. castaneus, C. sp. C. laevis and C. heptacanthus and the other was C. urotaenia group including C. urotaenia, C. sp. 1, C. sp. 2 and C. isaza. The genetic distances between two species from different groups were more than 1.2. In C. castaneus group C. heptacanthus, marine species, was separate from other three species, since the genetic distances between C. heptacanthus and other species in C. castaneus group were larger than 0.3. In C. urotaenia group, C. urotaenia and C. isaza, endemic to Lake Biwa, were closely related and the average genetic distance between them was 0.103. Figure 3 shows the curious situation of C. laevis. The Akita population of C. laevis was differentiated from the Saitama population of C. laevis genetically. The Akita population was most closely related to C. sp. from Lake Shinji and closely related to C. castaneus more than the Saitama population of C. laevis. DISCUSSION In the above studies, it is seen that genetic variabilities in populations were low in Chaenogobius species. Expected average heterozygosity ranged from 0.006 in C. sp. 2. to 0.062 in C. isaza and polymorphic loci, from 0% in C. sp. 1 and C. sp. 2 to 26.7% in C. castaneus and C. laevis (Table 5). However these values are comparable to those in populations of Pisces species previously reported [19,24]. This study showed that the application of molecular taxonomy, based on isozyme polymorphisms, is useful to reveal the phylogenetic relationships among morphologically similar Gobiidae species as Masuda et al. [14] showed in their studies of Rhinogobius species. The genus Chaenogobius was divided into two groups in this study (Fig.3). One was C. castaneus group and the other was C. urotaenia group. One group shared the alleles different from the other group’s at four loci, Ck, Me-1, Me-2 and Sp-1 (Table 3). The morphological studies have shown that C. urotaenia, C. sp.1, C. sp. 2 and C. isaza are similar to each other in its lateral line system and its large mouth when they were compared to other Chaenogobius species [1, 23]. Moreover the genetic distance between the two groups was more than 1.2 and this value was large enough to regard that they were genetically differentiated at genus level [18]. It was confirmed that three species of C. castaneus group, C. castaneus, C. sp. and C. laevis, morphologically similar each other, are closely related species. In addition, it was found that two C. laevis populations were highly differenti- ated. The Akita population of C. laevis was closely related to C. sp. and have more similar genetic population structure to C. castaneus than to the Saitama population of C. laevis. We could not find any difference in their sensory canals on their heads between individuals from Saitama and Akita of C. laevis although morphological differences between local populations of C. laevis were observed (Iwata, personal communication). In Lake Shinji, C. castaneus and C. sp. are sympatrically distributed and no hybrid individual was observed in this study. Because hybrid individuals should be easily detected, if they are, since different a-Gpd alleles were fixed in each species. This shows that there is no gene exchange between C.castaneus and C. sp. in Lake Shinji, which confirmed that C. castaneus and C. sp. are different species. This also suggests that C. castaneus and the Akita population of C. laevis are different species because the genetic distance between them was larger than that between C. castaneus and C. sp. And it is possible that the Akita population of C. laevis is different species of the Saitama population of C. laevis since the Saitama population situated on a different cluster from C. castaneus, C. sp. and C. laevis from Akita (Fig. 3). This is supported by the fact that Akita population was different from Saitama population in variants of a-Gpd and Pgm loci. The genetic distance between C. sp. and the Akita population of C. laevis is 0.114. It is difficult to decide from this value whether C. sp. and the Akita population of C. laevis are different species or not. There are species whose genetic identity (I) between closely related species are about 0.9 [36] that corresponds to 0.105 of Nei’s genetic distance. Furthermore C. sp. is different from C. Jaevis from Akita in patterns of sensory canals and pit organs on its head. These facts suggests the possibility that C. sp. and the Akita population of C. laevis are incipient species or subspecies. The genetic distance between populations of C. uro- taenia and C. isaza, sympatric species in Lake Biwa, was from 0.092 to 0.120. This is comparable to the genetic distance between C. sp. and the Akita population of C. laevis from 0.113 to 0.115. The studies on reproductive isolating mechanisms and/or ecological studies on these two species 460 T. Aizawa, M. Hatsumi AND K. I. WAKAHAMA TaBLE3. Allele frequencies at 14 loci in C. castaneus C. sp. C. laevis C. heptacanthus Locus Allele ee 2 3 4 5 6 q 8 9 10 11 12 Aat a 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 — b = == = = = = == os — = = 1.00 Cc at = = = = = — = == = = = d — — = —= = = — eee == Ck a 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 b a — = — = = = — = == — = Gpi-1 a — — — — = — .03 — — — — —_— b 1.00 1.00 1.00 1.00 1.00 1.00 .97 1.00 1.00 1.00 1.00 1.00 Gpi-2 a = = = = — = = = = = = .09 b 01 — — — — — _ — — .93 — .85 c .98 1.00 97 1.00 1.00 1.00 97 1.00 1.00 07. ~—-1.00 .06 d = pe = = ta ae = ats == = = = e .01 — .03 — — — .03 — — — — — f a= a 20 =e a = ee a = = = Es a-Gpd a — — — — — — — — — — — — b 01 03 — — — _ = — — = = 1.00 Cc aS as = a = = uy = ee = = = d — — — — = = = = = — 1.00 — e 99 97 1.00 1.00 1.00 1.00 — — — — — — f — — — — — — 1.00 1.00 1.00 1.00 — — Idh-1 a — = = = — — — — — .03 — — b 16 08 08 — 20 — 1.00 1.00 1.00 88 1.00 1.00 C = —_ = —_ —— ot = = = = —_ a d 84 92 92 + 1.00 80 1.00 — — — 09 _— — Idh-2 a — _ — — — — .03 05 — — — — b 1.00 1.00 1.00 1.00 1.00 1.00 94 95 1.00 1.00 1.00 — c — = = = = = 03 — — — — — Ldh a 1.00 1.00 1.00 1.00 1.00 #£1.00 1.00 1.00 1.00 99 ~=—« 1.00 1.00 b _— _— — — — — — — — 01 — — Mdh a _— — 03 — — —_ 18 .29 22 01 — — b = aes a = = = == sea = — = pales Cc sis = te oa pha a peer Be = = == pa d = = = = == a = == = = = — e 44 45 30 36 .30 50 82 71 78 10 — — f 56 55 67 64 70 .50 — _— — 89 ~—- 1.00 1.00 Me-1 a — = = = = = == = = b 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00° 1.00 1.00 1.00 1.00 Me-2 a 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 ‘1.00 1.00 b ES 2 ee oh = = oe =e = aS ~s 2% Pgm a 01 — — — .10 = = = = = = — b — — — = — _ — — — _— 28 _— c 05 05 _— —_ — — — 02 —_— — 72 — d — — — — — — — _— — — — 03 e 94 95 1.00 1.00 .90 83 1.00 98 1.00 81 — — f — — — — -- -— — — — — 97 g — — — — — 17 — — — 19 -- — Sod a — — — — = = = = = = = — is 3 eiM Rel pid at a =e == ast oek at = Cc a= ae ne pak — = ae uth Oe = eet aa d 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Sp-1 a _— — _— — — = = _- = os — — b 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Biochemical Systematic Study on Chaenogobius population of Caenngobius species C. urotaenia C. sp. 1 C. sp. 2 C. isaza 18 19 20 21 7) 23 100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 100 1.00 1.00 1.00 1.00 1.00 92 89 87 —-:1.00 90 44 08 1 13 10 .56 av 1.00 IO ACD) ACO LOO ao 1.00 MOON ECON SIEO0) 100) 1-00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 LOO 100 MCD 10) soo a2 1.00 1.00 1.00 1.00 1.00 1.00 1.00 L008 1200) ako0) co" “00 1.00 1.00 1.00 1.00 95 1.00 98 =-vboer=shte gat (O05 i = eu jenn te Bete 02 =e endziey 24. 29S 27 = 23 462 T. Aizawa, M. Hatsumi AND K. I. WAKAHAMA Species 1. 1, 3. C. castaneus C. sp. C. laevis, 10* TaBLeE 4. Mean (Min.-Max.) genetic distance 4. 5. 6. C. laevis, 11** CC. heptacanthus C. urotaenia 0.003 0.160 (0.000-0.007) (0.145-0.175) 2. 0.001 (0.000-0.001) 0.204 (0.187-0.221) 0.114 (0.113-0.115) 0.210 0.502 1.565 (0.186-0.226) (0.473-0.523) — (1.541-1.587) 0.198 0.478 1.573 (0.194-0.201) (0.474-0.482) —-(1.546-1.591) 3. = 0.194 0.311 1.568 —_ — (1.558-1.575) 4. — 0.375 1.395 — (1.372-1.425) 5. — 1.315 (1.294-1.335) 6. 0.008 (0.002-0.018) Te *: Lake Hatirou-gata. **; Koma River. TABLE5. Estimates of genetic variability in genus Chaenogobius : Population No. of alleles % Average heterozygosit SPECIES No. per locus polymorpic loci pecan d C. castaneus 1 1.467 26.7 0.061 2 1.267 20.0 0.052 3 1.267 13.3 0.043 4 1.067 6.7 0.031 5 1.200 20.0 0.061 6 1.133 13.3 0.052 Mean 1.234 16.7 0.050 C. sp. 7 1.133 13.3 0.033 8 1.200 6.7 0.036 9 1.067 6.7 0.023 Mean 1.200 8.9 0.031 C. laevis 10 1.467 26.7 0.058 11 1.067 6.7 0.027 C. heptacanthus 12 1.200 6.7 0.022 C. urotaenia 13 1.067 6.7 0.015 14 1.067 6.7 0.015 15 1.067 6.7 0.033 16 1.133 13.3 0.048 Mean 1.084 8.4 0.028 C. sp.1 17 1.200 0.0 0.010 C. sp.2 18 1.067 6.7 0.009 19 1.067 6.7 0.013 20 1.067 6.7 0.015 21 1.067 6.7 0.006 22 1.067 6.7 0.012 Mean 1.067 6.7 0.011 C. isaza 23 1.200 13.3 0.062 Overall mean 1.157 10.7 0.032 Biochemical Systematic Study on Chaenogobius 463 between Chaenogobius species qT. 8. 9. C. sp. 1 C. sp. 2 C. isaza 1.286 1.570 1.529 (1.272-1.301) (1.545-1.589) _—(1.531-1.562) 1.310 1.577 1.554 (1.305-1.319) (1.549-1.593) —-(1.535-1.566) 1.290 1.574 1.549 = (1.566-1.577) a 1.146 1.376 1.352 ae: (1.372-1.386) as 1.567 1.301 1.593 (1.296-1.306) (1.591-1.595) =x 0.237 0.225 0.103 (0.227-0.249) (0.206-0.252) (0.092-0.120) pa 0.414 0.358 (0.401-0.412) ass 8. 0.000 0.235 (0.000-0.001) (0.232-0.242) 9. = should supply significant information to establish the taxono- mic relationship between them. Three species of C. urotaenia, C. sp.1 and C. sp. 2 were confirmed that they were different species each other. Gene- tic distances between them ranged from 0.225 to 0.412 on average. These values are large, indicating that they are different species. C. urotaenia and C. sp. 2, distributing sympatrically at the four rivers in the Shimane Prefecture, had different alleles at the Sod locus and at the Mdh locus from each other and no heterozygote at these loci was observed through this study. Ecological study also showed that they are ethologically isolated [10]. This study revealed phylogenetic position of C. isaza which is specialized to land-locked freshwater and endemic to Lake Biwa. It was closely related to C. urotaenia, amphid- romous species as previously suggested [11]. We used the estimation of divergence time from genetic distance proposed by Nei [18], t=510°D. It suggested that C. isaza was differentiated from C. urotaenia about 0.5 million years ago. This time agrees with the origin of other endemic species in Lake Biwa, which was suggested by fossil records [25, 26] Isozyme polymorphisms supplied the information on population system in some species. The comparison be- tween populations within three amphidromous species, C. castaneus, C. urotaenia and C. sp. 2, showed that the dif- ferentiation of populations did not reflect the geographic distance between populations from different rivers (data was not shown). This suggests that genetic mixture between populations occurs when larvae go down to the sea and they have no behavioral character to return to the river where they were born. This study indicates the effectiveness of applica- tion of isozyme polymorphism to ecological study of Chaeno- gobius species that have various life histories. ACKNOWLEDGMENTS We thank Dr. Akira Goto, Hokkaido University, Mr. Kazuharu Shibuya, Akita Prefectural Fisheries Consulting Center, Mr. Akihisa Iwata, Imperial Household, Mr. Makoto Itoh, Tokyo Metropolitan Bureau of Labor and Economy, Mr. Isao Kimoto, Tokyo Metropoli- tan Fisheries Experimental Station, Dr. Sachiko Takahashi and Mr. Shoichi Matsuoka, Shiga Prefecture, Mr. Takatoshi Kawashima, Shimane Prefectural Fisheries Experimental Station, Mr. Kouichi. Itoh, Mr. Mikio Kadowaki, Mr. Kazuo Fukuda, and Mr. Tomitada Nagashima, Shimane Prefecture for supplying materials. We also thank Dr. Ryozo Kakizawa, Yamashina Institute for Ornithology, and Mr. Ikuo Takabatake, Shimane University, for technical support and Dr. Nobuhiko Mizuno, Ehime University, Dr. Iwao Sakamoto, Shimane Medical Collage, Mr. Toshiki Koshikawa, Akae Junior School, Mr. Akihisa Iwata and Masayoshi Hayashi, Yokosuka City Museum for useful suggestion to complete this study. We thank especially Mr. Takatoshi Kawashima and Mr. Hitoshi Sato, Shimane Prefectural Office for useful discussion and encouragement through this study. We thank Dr. K. M. Yin, Kent State University for his critical reading this manuscript. REFERENCES 1 Akihito, Prince (1984) Genus Chaenogobius. In “The Fishes of the Japanese Archipelago”. Ed by H Masuda, H K Amaoka,K C Araga, T Ueno, T Yoshino, Tokai University Press, Tokyo, pp 264—266 (in Japanese) 2 Avise J C (1974) Systematic value of electrophoretic data. Syst Zool 23: 465-481 3 Avise JC (1976) Genetic differentiation during speciation. In “Molecular Evolution Vol. 7” Ed, by F J Ayala, Sinauer. Sunderland, Mass; pp 106-122 4 Ayala F J (1975) Genetic differentiation during the speciation process. Evol Biol 8: 1-78 5 Ayala F J, Tracy M L, Hedgecock D, Richmond R C (1974) Genetic differentiation during the speciation process in Dro- sophila. Evolution 28: 576-592 6 Buth D G (1984) The application of electrophoretic data in systematic studies. Ann Rev Ecol Syst 15: 501-522 7 Dotu Y (1955) The life history of a Goby, Chaenogobius urotaenia (Hilgendorf).Sci Bull Fac Agr Kyushu Univ 15: 367— 374 8 Clayton J W, Tretiak D N (1972) Amine-citrate buffers for pH control in starch gel electrophoresis. J Fish Res Bd Canada 29: 1169-1172 9 Hayashi M, Sato H (1987) One species of genus Chaenogobius from Shinji Lake, Shimane prefecture. Proc Ann Meet Japan Ichthyol Soc 48 (abstract in Japanese) 10 Ishino K (1987) Freshwater fishes in Japan, their distribution, variation and speciation. Ed by M, Mizuno, A Goto, A Tokai Univ Press, pp 189-197 (in Japanese) 11 Kawanabe H (1978) Some biological problems. Verh Internal Verein Limmol 20: 2674-2677 12 Koshikawa T (1989) Shinjikohaze, Chaenogobius sp. In “Freshwater Fishes of Japan” Ed by H, Kawanabe, N Mizuno, Yama-Kei Publishers Co., Ltd., pp 614-615 (in Japanese) 13 Koshikawa T, Sato H (1986) Synopsis of new recorded goby, Chaenogobius sp. of Lake Shinji. The Freshwater Fishes, 12: 51- 55 (in Japanese) 14 Masuda Y, Ozawa T, Enami S (1989) Genetic differentiation among eight color types of the fresh water goby, Rhinogobius brunneus, from Western Japan. Japan J Ichthyol 36: 30-41 15 Nakanishi T (1978) Comparison of color pattern and meristic 464 16 ily) 18 19 20 21 22 23 24 25 26 T. Aizawa, M. Hatsumi AND K. I. WAKAHAMA characters among the three types of Chaenogobius annularis Gill. Bull Fac Fish Hokkaido Univ 29: 223-232 Nei M (1972) Genetic distance between populations. 106: 283-292 Nei M (1975) Molecular Population Genetics and Evolution. North-Holland Pub.Co., Amsterdam Nei M (1987) Molecular Evolutionary Genetics. University Press, New York Nevo E, Beliles A, Ben-Shlmo R (1984) The evolutionary significance of genetic diversity: Ecological, demographic and life history correlates. In “Evolutionary Dynamics of Genetic Diversity” Ed by G S Mani, Springer-Verlag, pp 13-213 Numachi K (1981) A simple method for preservation and scanning of starch gels. Biochem Genet 19: 233-236 Numachi K, Nagahora S, Iwata M (1979) Genetic demonstra- tion of hybrids between chum and pink salmon in the northwest Pacific. Rep Otsuchi Mar Res Cent Univ Tokyo 5: 87-102 Pasteur N, Pasteur G, Bonhomme F, Catalau J, Britton- Davidian J (1988) Practical Isozyme Genetics. Ellis Horwood limited, England Pinchuk V I (1978) Notes and supplements to the Family Gobiidae in the book by Lindberg and Krasyukova “Fish of the Sea of Japan and the neighboring part of the Sea Ohotsk and the Yellow Sea”, Part 4, 1975 with a description of new species Chaenogobius taranetzi. J Ichtyol 18: 1-14 Powell J R (1975) Protein variation in natural populations of animals. Evol Biol 8: 79-119 Research Group of Natural History of Lake Biwa (1983) Fossil assemblages from the Pleistocene Katata formation of the Kobi- wako group at Ogi-cho, Otsu City, Central Japan. Bull Mizu- nami Fossil Museum 10: 117-142 Research Group for Natural History of Lake Biwa (1986) Freshwater fossil assemblage from the Pleistocene Kobiwako Am Nat Columbia 27 28 29 30 31 32 33 34 35 36 group on the southwest side of Lake Biwa Bull. Mizunami Fossil Museum 13: 57-103 Richardson B J, Baverstock P R, Adams M (1986) Allozyme Electrophoresis. Academic Press, Australia Selander R K, Smith M H, Yang S, Johnson W E, Gentry J B (1971) Biochemical polymorphism and systematics in the genus Peromyscus. 1. Variation in the old-field mouse (Peromyscus polionotus). Stud Genet 6 Univ Texas Publ 7103; 49-90 Shaw C R, Prasad R (1970) Starch gel electrophoresis of enzymes-A compilation of recipes. Biochem Genet 4: 297—320 Sneath P H A, Sokal R R (1973) Numerical Taxonomy. Freeman, San Francisco, pp 188-308 Takagi K (1952) A critical note on classification of Chaenogo- bius urotaenia and its two allies. Japan J Ichthyol 2: 14-22 Takagi K (1966a) Taxonomic and nomenclatural status in chaos of the Gobiid fish, Chaenogobius annularis Gill 1. Re- view of the original description, with special reference to estima- tion of the upper jaw relative length as a taxonomic character. J Tokyo Univ Fish 52: 17-27 Takagi K (1966b) Taxonomic and nomenclatural status in chaos of the Gobiid fish, Chaenogobius annularis Gill Tomiyama with description of the genus Rhodoniichtys, Gen Nov J Tokyo Univ Fish 52: 29-45 Takagi K (1967) Topologie du systeme sensoriel cephalique des Gobioidei. La Mer 5: 131-145 Thorpe J P (1982) The molecular clock hypothesis: Biochemic- al evolution, genetic differentiation and systematics. Ann Rev Ecol Syst 13: 139-168 Thorpe J P (1983) Enzyme variation, genetic distance and evolutionary divergence in relation to levels of taxonomic separation. In “Protein Polymorphism, Adaptive and Taxono- mic Significance” Ed by G S Oxford, D Rollinson, Academic Press, London and New York, pp 131-152 ZOOLOGICAL SCIENCE 11: 465-471 (1994) © 1994 Zoological Society of Japan Speciation of Japanese Pond Frogs Deduced from Lampbrush Chromosomes of their Diploid and Triploid Hybrids HIROMI OHTANI Laboratory for Amphibian Biology, Faculty of Science, Hiroshima University, Higashi-hiroshima 724, Japan ABSTRACT—To examine the hybrid origin hypothesis of Rana porosa porosa cytogenetically, the lampbrush chromo- somes of triparental allotriploids comprising the genomes of R. nigromaculata, R. p. brevipoda, and R. p. porosa were investigated together with those of their intra- and interspecific hybrids. The behavior of their homologous lampbrush chromosomes provided little evidence that chromosomal material from R. nigromaculata is present in the genomic chromosomes of R. p. porosa. On the contrary, it is suggested that R. p. porosa and R. nigromaculata are phylogenetically more distant than are R. p. brevipoda and R. nigromaculata. INTRODUCTION With respect to the differentiation of Rana porosa porosa Cope, Moriya [13] and Kawamura and Nishioka [4-6] prop- osed the hypothesis of hybrid origin between R. nigromacula- ta Hallowell and R. p. brevipoda Ito. This hypothesis was supported by Kuramoto [7], but questioned by Matsui and Hikida [12]. Recently, Nishioka et al. [17] offered support for the hypothesis from electrophoretic analysis, though the support was not without a tinge of interested consideration. Although I was a collaborator of the latter paper, I now question the hybrid origin hypothesis in view of the fact that the lampbrush chromosomes of R. p. porosa closely resemble those of R. p. brevipoda, and yet there are no landmarks derived from R. nigromaculata throughout their axes (unpub- lished). The hybrid origin should be demonstrated by comparing the behavior of lampbrush chromosomes in diploid hybrids between the above-mentioned three taxa, because the num- ber of chiasmata that control the behavior of lampbrush chromosomes changes in accordance with the extent of similarity between the homologues of parental species [9, 14, 20]. Similarly, when the genomic chromosomes of these three taxa are placed together in an oocyte, provided that R. P. porosa receives many dominant and recessive genes from R. nigromaculata as suggested by Kawamura and Nishioka [4], some of the chromosomes of R. p. porosa should join inevitably to those of R. nigromaculata and act as a mediator in formation of trivalents. The lampbrush chromosomes of R. nigromaculata are easily distinguished from those of R. p. brevipoda [16] and R. Pp. porosa (unpublished) by size, type, and position of the landmarks. Thus, the lampbrush chromosomes of tri- parental allotriploid females were examined to cytogenetical- ly determine the relative degree of synaptic affinity among the Accepted May 16, 1994 Received Feburary 7, 1994 chromosomes of these taxa together with those of their intra- and interspecific hybrid females. This paper describes these results and a new hypothesis of the differentiation of R. p. porosa is proposed. MATERIALS AND METHODS The female frogs studied are described in Table 1. The paren- tal frogs for crosses were from the lineages of R. p. brevipoda collected from Konko, Okayama Prefecture, R. p. porosa from Machida, a city of Tokyo, and R. nigromaculata from Hiroshima. The frogs were crossed by artificial fertilization. Autotriploid frogs were produced by cooling the fertilized eggs of R. p. brevipoda to =1°C for 2hrs. Two kinds of allotriploid frogs were produced by inseminating a few diploid ova which brevipoda ? < nigromaculata $ hybrid females spawned with spermatozoa of the two subspecies. Tadpoles were fed on boiled spinach or chard, and frogs were fed on houseflies or tropical crickets. Lampbrush chromosomes were removed from the ovarian eggs of two-year-old females just prior to hibernation (November) accord- ing to Gall’s method with a slight modification [1, 20], and examined under a phase-contrast microscope. The abbreviations B, P, and N refer to brevipoda, porosa, and nigromaculata chromosomal sets, respectively. The letters in parentheses indicate the sources of cytoplasm. Chiasma frequencies per oocyte were compared using Student’s or Aspin-Welch’s t-test. Chi-square test also was used for the comparison of chiasma numbers in two kinds of allotriploids. RESULTS Autotriploid (B)BBB Lampbrush chromosomes from 60 oocytes were analyzed in detail. All the oocytes contained 39 lampbrush chromo- somes consisting of 13 triplets of homologues that belonged to five large chromosomes numbered 1 to 5 and eight small chromosomes numbered 6 to 13. These lampbrush chromo- somes formed eight or more trivalents in all the oocytes; it was notable that those of six oocytes formed exclusively 13 trivalents (Table 2). All the chromosomes other than those 466 H. OHTANI TaBLE 1. Kind, origin, and number of female frogs used in the present study Parental Origin Number of Kind Female Male females Autotriploid R. p. brevipoda R. p. brevipoda 5 (B)BBB (3n=39) Allotriploid brevipoda $- Xnigromaculata $ R. b. brevipoda 5 (B)BBN (3n=39) if Allotriploid brevipoda §. x nigromaculata $ R. b. porosa 4 (B)BPN (3n=39) a Diploid hybrid R. p. porosa R. p. brevipoda 9 (P)PB (2n=26) Z Diploid hybrid R. p. porosa R. nigromaculata 9 (P)PN (2n=26) e Diploid hybrid R. nigromaculata R. p. porosa 8 (N)NP (2n=26) Non-hybrid R. p. porosa R. p. porosa 11 (P)PP (2n=26) ae TaBLE 2. Number of oocytes having tri-, bi- and univalents in homologues in juxtaposition by four to 12 chiasmata and various scombinatlons in the: thr cey iiss batmploids rarely by terminal fusions (Fig. 1A). In 359 other trivalents, No. of No. of No. of : : . Kind two of the three homologues were joined by two to eight trivalents bivalents univalents (B)BBB (B)BBN (B)BPN chiasmata and the rest was joined to one of them by one to 13 0 0 6 three chiasmata, a terminal fusion, or both, in addition (Fig. on ; 4 + 1B). In the remaining 107 trivalents, the three homologues (36) ( 2) (1) were joined in tandem by one or two chiasmata or a terminal BI ) 7 fusion (Fig. 1C). (33) ( 4) ( 2) In chromosome Nos. 1, 4, 6-9 and 12, the chiasma (0) 3) ( 3) 13 frequencies were about 1.5 times higher than those of diploid R. p. brevipoda (Table 4). Those of the remainder were (7) ( 8) ( 4) “I slightly lower than 1.5 times. The number of chiasmata in 8 5 5 4 each oocyte was between 35 and 76 (average, 55.0) in total. (24) (10) ( 5) 7 6 6 1 Allotriploid (B)BBN (21) Ce) (©) Lampbrush chromosomes from 57 oocytes were analy- (as) (14) ( iy ! zed. Eleven of these oocytes, an aneuploid one (3n-2) 5 8 8 1 lacking the nigromaculata chromosomes of Nos. 2 and 8, and (15) (16) ( 8) 10 hexaploid ones, were omitted from the analysis. The 4 9 9 2 remaining 46 oocytes contained 39 lampbrush chromosomes (C2 ae) CY) consisting of one nigromaculata and two brevipoda chromo- ( 3) (20) (0) : 2 somes in each of. the 13 homologue triplets (Fig. 2). These 2 ul a 2 1 chromosomes formed one to five trivalents in 33 oocytes ( 6) (22) (11) (Table 2). The remaining 13 oocytes had no trivalents. All 1 12 12 12 4 the chromosomes other than those of the trivalents formed ) (e) (U2) bivalents and univalents of the same number, or simply : (28) (13) ut © univalents; two oocytes contained only 39 univalents. 0 0 39 2 8 In 64 of the 65 trivalents, a nigromaculata chromosome (39) was always joined to one of the two brevipoda chromosomes Total 60 46 25 by one or two chiasmata, a terminal fusion, or both, in Numbers in parentheses show numbers of lampbrush chromo- addition (Table 3). The remaining one trivalent was seen in somes forming tri-, bi-, or univalents. chromosome No. 6 and arranged three homologues of, in order, brevipoda, nigromaculata and brevipoda, which were each joined by one chiasma in tandem. On the other hand, of the trivalents formed bivalents and univalents of the same in 507 triplets of homologues forming a bivalent and a number. univalent, the bivalent always consisted of two brevipoda The number of trivalents in chromosome Nos. 1 to 13 is chromosomes, and the univalent of a nigromaculata chromo- presented in Table 3. Of the 655 trivalents, 189 joined three some. In the remaining 26 triplets of homologues, they Speciation of Japanese Pond Frogs 467 TABLE 3. Behavior of homologous lampbrush chromosomes in each of the 13 triplets in the three kinds of triploids Chromosome Kinds behavior Chromosome no. Total in a triplet 1 2 3 4 5 6 7 8 Q iQ th WW ig (B)BBB b-b-b 54. 57 58 S55 53 46 52 48 46 47 44 44 51 655 b-b, b Cink Sear ecre tO a ent ys 14a 8 2 4s) 188 lor 6c s--9 125 Total 60 60 60 60 60 60 60 60 60 60 60 60 60 780 (B)BBN b-b-n OS, G2 54 4 3 1 De a3 ve Ohve 25% Wy Ohm aS 64 b-n-b 1 1 b-b, n 38 38 40 40 41 32 42 41 38 41 #38 #37 «41 507 b, b, n ee ee a OM) Selena ees iD | ALND Dt iat yi LD 26 Total 46 46 46 46 46 46 46 46 46 46 46 46 46 598 (B)BPN b-p-n or p-b-n Bre 42557 30-4 Det ee 4, 1 D, 1 Bhat a? 31 b-n-p 1 2 b-p, n LASS: elie 14 ATID on! 1st lis. e13P) 14 183 b-n,p or p-n,b 1 1 iL iL* 5 b, p, n SE Ou OMe Oe MSany FOcOm SSeS sy TSe TO ma nOk a 28 104 Total Tp). DiS) DS SS TS A DS) Oi TiS IBY DY! | PIS. SP) 325 “” indicates a connection between two homologues. Fic. 1. Microphotographs of trivalents of chromosome Nos. 3 (A) and 10 (B and C) in an autotriploid, (B)BBB. Arrows indicate the positions of chiasmata. c and s represent compound- and simple-type giant loops, respectively. Bar=50 4m. remained as univalents. In all the trivalents, nigromaculata and brevipoda chromosomes were joined by 57 (3%) chiasmata in total The abbreviations b, p, and n indicate brevipoda, porosa, and nigromaculata lampbrush chromosomes, respectively. * Joining was effected between brevipoda and nigromaculata chromosomes. except for the terminal fusions. By contrast, joining of two brevipoda chromosomes in the bivalents and trivalents was by 2033 (97%) chiasmata in total except for the terminal fusions. The chiasma frequencies in chromosome Nos. 1 to 13 were about 1.2 times higher than those of diploid R. p. brevipoda except in chromosome Nos. 11 and 13, though the chiasmata for nigromaculata and brevipoda chromosomes accounted for no more than 3% of the total (Table 4). The number of chiasmata in each oocyte was between 0 and 60 (average, 45.4) in total. This average value was different from that of the diploid R. p. brevipoda (t=3.7, P<0.001) or the autotriploid (B)BBB (t=4.4, P<0.0001). Allotriploid (B)BPN Lampbrush chromosomes from 41 oocytes were analy- zed. Sixteen of the 41 oocytes were seven aneuploid oocytes of 3n-1 (three), 3n-2 (two), and 3n-5 (two), and nine hexa- ploid oocytes; all of which were omitted from the analysis. In the remaining 25 normal triploid oocytes in which each triplet of homologues consisted of one brevipoda, one porosa, and one nigromaculata chromosomes, the lampbrush chromosomes were somewhat similar in behavior to those of the other allotriploid (B)BBN (Table 2). In 11 oocytes they formed one to seven trivalents, while all the chromosomes other than those of the trivalents formed bivalents and univalents of the same number (Fig. 3). In six other oocytes they formed 13 bivalents and 13 univalents. The remaining eight oocytes had only 39 univalents. Of the 33 trivalents, 31 joined a nigromaculata chromo- some to either of the brevipoda and porosa chromosomes by One or two chiasmata, a terminal fusion, or both, in addition (Table 3). The remaining two trivalents arranged three 468 H. OHTANI TaBLE 4. Frequency of chiasmata in chromosome Nos. 1 to 13 Chromosome no. Type 1 2 3 4 5 6 7 8 Ge i PP ea iB Total (P)PP ia na nn aLp ORD an DG OD» Dil Be 20-9 Aa 20 39.0 (B)BB* Tr ae eee I eS Sa MOOT BD an 19 20 38.7 (N)NN* HEME a 23 9S 2S TAO ad 2 OB Dm 38.6 (B)BBB TL le tod AG Se SOUR DR Day OR oR 55.0 (B)BBN 60 59 AG AO) 4G 0 se, a2) on 7S EDGED 45.4 (B)BPN ieee ieee OS NN 1S De TE GO 16 ie 35.7 (P)PB SG eo ie Be D5 O2 DD 16 Ai 1 Bil 19 35.8 (P)PN V6” 26 MO Ss oT ge a7 Sion ana notes siecle 25.2 (N)NP MP Oe Al woo ty OO 1S 20- 16 Le 12 26.0 (B)BN* 37, 23 12:57 S310 2146 23) Aes) 210) ions RONMICIEG IEC INNIS 28.9 (N)NB* Wie Deo 8) O27 20 17 19 25 aS 16 16 146 27.6 * Data from [20] Bs gaa 4 pty ee ee eee 3 > oe fh 9 4 s Benge. hae este 7 Gh agty F SDtEC A ‘4 4 “paie es ag Ng si ope iy Bee) My Seng gee ath eg “Aan vo ae ee Re oe. phere Cie vata 3 A a sos * of Wt. ar ha * : . aor: i 1 Breet ahd SE number. B and N represent brevipoda and nigromaculata chromosomes, respectively. Two brevipoda chromosomes form a bivalent and one nigromaculata chromosome forms a univalent in chromosome Nos. 1 to 13 except for5 and 10. The three homologues of chromosome Nos. 5 and 10 form a trivalent which joins a nigromaculata chromosome to one of the two brevipoda chromosomes by a single chiasma. Arrows indicate the positions of chiasmata. Bar=50 um. homologues of, in order, brevipoda, nigromaculata and poro- sa; the three homologues of chromosome No. 11 were each joined by one chiasma in tandem, and those of chromosome No. 12 also were joined by two chiasmata and by a terminal fusion. In 183 of the 188 triplets of homologues forming a bivalent and a univalent, the bivalents consisted of brevipoda and porosa chromosomes, and the univalents of nigromacula- ta chromosomes. In the remaining five triplets of homo- logues, bivalents included a nigromaculata chromosome. In the bivalents and trivalents, joining of nigromaculata and brevipoda or porosa chromosomes was effected by 35 (4%) chiasmata in total except for terminal fusions. In contrast, brevipoda and porosa chromosomes were joined by 857 (96%) chiasmata in total except for the terminal fusions. Speciation of Japanese Pond Frogs aa stn Dee i x BY ns a 4 y Bret: Nea | m ag cart 469 me Y, CBA ty 1 Sot EST he rg : as Eas as PT Yee alt ta 4s, seen cee oar oes Ls topes ul qe Oe Fae Ve ees EQeees ars Fic. 3. Microphotographs of the lampbrush chromosomes in an oocyte of a triparental allotriploid, (B)BPN. The number in each photograph represents the chromosome number. The abbreviation N in photographs 7, 8, 9 and 12 represents a nigromaculata chromosome. Although landmarks did not develop very much in this preparation, nigromaculata chromosomes are infallibly distinguished from those of brevipoda and porosa . Two brevipoda and porosa chromosomes form a bivalent and one nigromaculata chromosome forms a univalent, in chromosome Nos. 1 to 13 except 7, 8, 9, and 12. The homologues of chromosome Nos. 7, 8, 9, and 12 form a trivalent which joins a nigromaculata chromosome by a single chiasma (arrow, in Nos. 7, 8, and 9), and by a terminal fusion (arrow head, in No. 12). These values were not different from those in the other allotriploid (B)BBN (7?=3.0, P=0.08). The chiasma frequencies in chromosome Nos. 1 to 13 were generally much lower than in another allotriploid (B)BBN (Table 4), but the total number of chiasmata in each oocyte (0-67, average=35.7) was not different from that of (B)BBN (t=1.8, P=0.09). Bar=S0 um. Intraspecific hybrid (P)PB In the 30 oocytes examined, all the lampbrush chromo- somes formed 13 bivalents like those of parental subspecies R. p. brevipoda and R. p. porosa (Table 5). The bivalents had one to eight chiasmata. When the chiasma frequency in chromosome Nos. 1 to 13 was compared with those of the two parental subspecies, those of chromosome Nos. 2, 3 and 9 470 H. OnTANI TaBLE5. Number of oocytes having bi- and univalents in various combinations No. of No. of Kind bivalents univalents (P)PP (B)BB* (P)PB (P)PN (N)NP 13 0 50 50 30 24 27 (26) 12, 2 5 3 (24) (2) 11 4 1 (22) (4) Total 50 50 30 30 30 Numbers in parentheses show numbers of lampbrush chromo- somes forming bi- or univalents. * Data from [20] were far lower, but those of the remaining chromosomes were similar (Table 4). The number of chiasmata in each oocyte was between 30 and 44 (average, 35.8) intotal. This average value was different from that of R. p. brevipoda (t=3.3, P= 0.001) or R. p. porosa (t=3.4, P=0.001). Interspecific hybrid (P)PN and (N)NP In 24 and 27 of the respective 30 oocytes examined, all the lampbrush chromosomes formed bivalents of which the homologues were connected by one to six chiasmata, terminal fusions, or both (Table 5). The remaining six and three oocytes had two or four univalents with 12 or 11 bivalents. The chiasma frequency in chromosome Nos. 1 to 13, except for No. 10, was far lower than those of the two parental species, R. p. porosa and R. nigromaculata, and the porosa $- x brevipoda$ intraspecific hybrid (Table 4). The total number of chiasmata in each oocyte was between 15 and 32 (average, 25.2) in the porosa? nigromaculata$ hybrid and between 16 and 33 (average, 26.0) in the reciprocal hybrid. Although the average values did not differ between these reciprocal interspecific hybrids (t=0.8, P=0.45), they were significantly lower than those of the brevipoda? x nigromaculata$ hybrid (t=4.8, P<0.0001) and the nigromaculata $ X brevipoda% hybrid (t=3.5, P<0.001). DISCUSSION Kawamura and Nishioka [4, 5] suggested that the incep- tion of R. p. porosa was natural hybrids between R. nigroma- culata and R. p. brevipoda. They also assumed that the newly divided population received subspecies rank as a result of accumulation of nigromaculata genes; such accumulation made it impossible to distinguish R. p. porosa from inter- specific hybrids between R. p. brevipoda and R. nigromacula- ta by external appearance. However, as White [23] stressed, the presence of three related species, A, B, and C, forming a linear series with B as an intermediate form between the other two, does not necessarily imply that B is a species of hybrid origin. In localized hybrid zones where two populations of closely related species come into contact, there are cases where many hybrid individuals and their descendants are produced [2, 8]. Nevertheless, the fact remains that there is often strong selection against hybrids within the hybrid zone, without the influence of introgression beyond the zone [3, 22]. In R. p. porosa also [17], the electrophoretic patterns caused by introgression of R. nigromaculata genes were found in some of the R. p. porosa collected from two stations where R. p. porosa coexists with R. nigromaculata, whereas they were not found in any of the R. p. porosa collected from eight other stations where R. p. porosa exists alone. Nishioka et al. [17] did not touch upon the latter case in supporting the hybrid origin hypothesis, though the former case was pointed- ly emphasized. Results of the present study offer little support for the hypothesis of hybrid origin of R. p. porosa. While the autotriploid (B)BBB formed trivalents in 655 (84%) of the 780 triplets, both the allotriploids (B)BBN and (B)BPN only formed them in 65 (11%) of the 598 triplets and in 33 (10%) of the 325 triplets, respectively. If the chromosomes of R. p. porosa had received great numbers of dominant and recessive genes from R. nigromaculata, the homologues of the allotri- ploid (B)BPN should have formed far more trivalents through the intermediation of a porosa chromosome. However, those chiasmata for the allospecific chromosomes represented no more than 4% of the total, which was the same as the ratio of the other allotriploid (B)BBN. The chiasma frequencies per oocyte of the reciprocal interspecific hybrids between R. p. porosa and R. nigromacu- lata were 25.2 ((P)PN) and 26.0 ((N)NP), both much lower than those for the two parental species (39.0 and 38.6) and the porosa?xbrevipoda®_ intraspecific hybrid (35.8). Moreover, these values were significantly lower than those of other reciprocal interspecific hybrids between R. p. brevipoda and R. nigromaculata (28.9 and 27.6). Therefore, R. p. porosa and R. nigromaculata must be phylogenetically more divergent than are R. p. brevipoda and R. nigromaculata. I propose that the incipient population of R. p. porosa occurred simply as a result of geographical isolation. Populations of the pond frogs are susceptible to isolation in the area where a range of mountains reaches down to the coastline, since they are typical inhabitants of streams and marshes in the coastal plains. It seems probable that an ancestral distributional range of R. porosa which continuous- ly existed along the Pacific coast of Honshu Island was divided into two geographical isolates, one eventually giving rise to R. p. porosa and the other to R. p. brevipoda, by the disappearance of the habitats in the mountainous Izu-Hakone areas. This isolation might be attributed to the upheaval caused by a collision between the central Honshu region and ancient Izu island (now a peninsula) about 0.5 million years ago [10]. In fact, if Nei’s divergence time (1D=5 million years) [15] is available for a small genetic distance value, the divergence time of the two subspecies will be estimated from their genetic distance data (0.130~0.249, average=0.189) [17] at 0.95+0.2 million years (the average was limited using Chebyshev’s theorem (k=2)). This divergence time seems Speciation of Japanese Pond Frogs 471 not to be so distant from the assumptive time of the collision. In addition, this upheaval has defended the Kanto plain from the invasion of R. nigromaculata [21]. Matsui [11] has reported that Japanese common toad, Bufo japonicus Schlegel, is divided into northeastern and southwestern types on the basis of morphometric variation analyses by the central Honshu region as a dividing line. Such geographic division seems to have affected the populations of Rana japonica Gunther and Rana rugosa Temminck & Schlegel also. Their dendrograms based on genetic distance data [18, 19] suggest it, and the divergence times between the Tokai and Kanto populations are 0.50+0.3 million years in R. japonica and 0.60+0.2 million years in R. rugosa. REFERENCES Gall JG (1966) Techniques for the study of lampbrush chromo- somes. In “Methods in Cell Physiology 2” Ed by DM Prescott, Academic Press, New York, pp 37-60 Gartside DF (1980) Analysis of a hybrid zone between chorus frogs of the Pseudacris nigrita complex in the southern United States. Copeia 1980: 56-66 Gollmann G (1991) Population structure of Australian frogs (Geocrinia laevis complex) ina hybrid zone. Copeia 1991: 593- 602 Kawamura T, Nishioka M (1977) Aspects of the reproductive biology of Japanese anurans. In “The Reproductive Biology of Amphibians” Ed DH Taylor, SI Guttman, Plenum Press, New York and London, pp. 103-139 Kawamura T, Nishioka M (1978) Descendants of reciprocal hybrids between two Japanese pond-frog species, Rana nigroma- culata and Rana brevipoda. Sci Rep Lab Amphibian Biol Hiroshima Univ 3: 399-419 Kawamura T, Nishioka M (1979) Isolating mechanisms among the water frog species distributed in the Palearctic region. Mitt Zool Mus Berlin 55: 171-185 Kuramoto M (1977) Mating call structures of the Japanese pond frogs, Rana nigromaculata and Rana brevipoda (Amphi- bia, Anura, Ranidae). J Herpetol 11: 249-254 Lamb T, Avise JC (1987) Morphological variability in geneti- cally defined categories of anuran hybrids. Evolution 41: 157- 163 Mancino G, Ragghianti M, Bucci-Innocenti S (1979) Ex- perimental hybridization within the genus Triturus (Urodela: Salamandridae). II. The lampbrush chromosomes of F; species hybrids between Triturus cristatus carnifex and T. vulgaris meri- 10 11 12 13 14 15 16 17 18 19 20 21 22 23 dionalis. Caryologia 32: 61-79 Matsuda T (1978) Collision of the Izu-Bonin Arc with central Honshu: Cenozoic Tectonics of the Fossa Magma, Japan. J Phys Earth 26 Suppl: S 409-421 Matsui M (1984) Morphometric variation analyses and revision of Japanese toads (Genus Bufo, Bufonidae). Contrib Biol Lab Kyoto Univ 26: 209-428 Matsui M, Hikida T (1985) Tomopterna porosa Cope, 1868, a senior synonym of Rana brevipoda Ito, 1941 (Ranidae). J Herpetol 19: 423-425 Moriya K (1954) Studies on the five races of the Japanese pond frog, Rana nigromaculata Hallowell. I. Differences in the morphological characters. Jour Sci Hiroshima Univ Ser B Div 1 15: 1-21 Miller WP (1977) Diplotene of chromosomes of Xenopus hybrid oocytes. Chromosoma 59: 273-282 Nei M (1987) Molecular Evolutionary Genetics. Columbia University Press, New York, pp 208-253 Nishioka M, Ohtani H, Sumida M (1980) Detection of chromosomes bearing the loci for seven kinds of proteins in Japanese pond frogs. Sci Rep Lab Amphibian Biol Hiroshima Univ 4: 127-184 Nishioka M, Sumida M, Ohtani H (1992) Differentiation of 70 populations in the Rana nigromaculata group by the method of electrophoretic analyses. Sci Rep Lab Amphibian Biol Hiroshima Univ 11: 1-70 Nishioka M, Sumida M, Borkin LJ, Wu Z (1992) Genetic differentiation of 30 populations of 12 brown frog species distributed in the Palearctic region elucidated by the elec- trophoretic method. Sci Rep Lab Amphibian Biol Hiroshima Univ 11: 109-160 Nishioka M, Kodama Y, Sumida M, Ryuzaki M (1993) Sys- tematic evolution of 40 populations of Rana rugosa distributed in Japan elucidated by electrophoresis. Sci Rep Lab Amphibian Biol Hiroshima Univ 12: 83-131 Ohtani H (1990) Lampbrush chromosomes of Rana nigroma- culata, R. brevipoda, R. plancyi chosenica, R. p. fukienensis and their reciprocal hybrids. Sci Rep Lab Amphibian Biol Hiroshi- ma Univ 10: 165-221 Ohtani H (1994) Polymorphism of lampbrush chromosomes of Japanese pond frog, Rana nigromaculata. Zool Sci 11: 337-342 Szymura JM, Barton NH (1991) The genetic structure of the hybrid zone between the fire-bellied toads Bombina bombina and B. variegata: Comparisons between transects and between loci. Evolution 45: 237-261 White MJD (1978) Modes of Speciation. WH Freeman and Company, San Francisco, pp 323-349 ZOOLOGICAL SCIENCE 11: 473-484 (1994) © 1994 Zoological Society of Japan Karyotypes and Ag-NOR Variations in Japanese Vespertilionid Bats (Mammalia: Chiroptera) Takao Ono! and YosHITAKA OBARA” Department of Biology, Faculty of Science, Hirosaki University, Bunkyo-cho, Hirosaki 036, Japan ABSTRACT— The karyological relationships among 14 species of Japanese vespertilionid bats were investigated paying attention to the chromosomal location and frequency of Ag-NORs. The karyotypes of these bats could be classified into two groups: centromere-cap NOR (cmcNOR) and interstitial NOR (intNOR) lineages. The former contained Myotis, Barbastella, Plecotus and Murina, and the latter Pipistrellus, Nyctalus, Vespertilio and Miniopterus. In the former lineage Ag-NORs were all tiny in size and distributed to many acrocentrics. In contrast, those of the latter were comparatively massive in size and shared by only two pairs at the most. While cmcNORs showed in general intergeneric and inter specific variations as well as intraspecific variations in their distribution pattern, intNORs showed rather stable patterns of distribution even among genera. Detailed G- and NOR-banding analyses suggested that most of the NOR-carrying acrocentrics of the ancestral karyotype might have been involved in Robertsonian rearrangements in the course of karyotype evolution in the Vespertilionidae: cmcNORs may have been transferred to other acrocentrics through cytologic events such as NOR-associations and interchromosomal chromatid exchanges and/or have been deleted through a centric fusion process. The present findings on the type and distribution of Ag-NORs were essentially consistent, except for a certain extent of variations in the location of cmcNORs, with those in the European vespertilionid bats previously reported by Volleth in 1985, and reinforced the karyological dendrogram of Japanese vespertilionid bats proposed by Harada in 1988. Thus, Ag-NORs may also provide useful cytogenetic parameters, at least in the Vespertilionidae, for the determination of phylogenetic relationships. INTRODUCTION The bats of the family Vespertilionidae are among the largest taxa (38 genera, 319 species) in the order Chiroptera [20] and their karyotypes show a marked variation in chromo- some constitution as well as in diploid number [7, 8, 17, 46]. It is well known that Robertsonian rearrangements, i.e., centric fusions/fissions, may have played a major role in chromosomal evolution of the family Vespertilionidae, although other types of rearrangements such as transloca- tions, inversions and additions or deletions of constitutive heterochromatin (C-heterochromatin) are also known in this family [10, 11, 13, 29-31]. Accordingly, most of the chromosome arms of vespertilionid bats have been conserved in the course of diversification of karyotypes [10, 27, 30, 45, 50]. Thus, vespertilionid bats may be especially desirable for elucidation of the relationship between Robertsonian rearrangements and evolutionary processes which pertain to phylogenetic diversification. Recently it has been suggested that variations in the chromosomal location of nucleolus organizer regions (NORs) may provide useful information in reconstructing phylogeny irrespective of the presence or absence of karyotypic altera- tion [1, 21, 35, 54, 55, 57, 58]. The NOR-banding using Accepted May 23, 1994 Received March 2, 1994 " Present address: Chromosome Research Unit, Faculty of Science, Hokkaido University, North 10, West 8, Kita-Ku, Sapporo 060, Japan ? To whom reprint requests should be addressed. silver nitrate visualizes the chromosomal sites of transcrip- tionally active ribosomal RNA gene clusters as Ag-NORs. The Ag-NORs correspond to satellite stalks and secondary constrictions of chromosomes in most animal species [32]. Although a large number of karyotype and G- and C-banding studies have been published on Chiroptera to date, NOR- banding has been done in only a few cases. Volleth [51] examined the chromosomal location of NORs in 21 species of European vespertilionid bats, using silver nitrate staining, and found marked interspecific and/or intergeneric variations in the location of NORs. In some genera, she also found multiple Ag-NORs located on the minute short arms close to the centromere of acrocentric chromosomes. In this paper, such Ag-NORs were tentatively termed centromere-cap NORs (cmcNORs). The relation of cmcNOR variation with Robertsonian rearrangements in this group of bats remains unresolved. In order to shed light on this subject we examined, by means of G-, C- and NOR-banding techniques, the chromo- somes of 14 species of Japanese vespertilionid bats inhabiting northern Honshu and Hokkaido, with special attention to chromosome homoeology and Ag-NORs. MATERIALS AND METHODS Ninety-seven live specimens of 14 species of Japanese vesperti- lionid bats were captured in northern Honshu and Hokkaido. Of these, 73 specimens contributed to the chromosomal findings based on the conventional and/or differential staining (Table 1). Chromosome preparations were made using bone marrow cells 474 T. ONo AND Y. OBARA TABLE 1. Vespertilionid bat specimens examined in this study Number of specimens Species Abbreviation examined Collecting site S + Subfamily Vespertilioninae Tribe Myotini Genus Myotis M. nattereri (Kuhl) Myn 1 Ashiro, Iwate Pref. M. hosonoi Imaizumi Myh 2 — Ichinohe, Iwate Pref. 1 1 Ashiro, Iwate Pref. M. frater kaguyae Imaizumi Myf — 4 Ichinohe, Iwate Pref. 3 — Ashiro, Iwate Pref. M. pruinosus Yoshiyuki Myp — 1 Sawauchi, Iwate Pref. — DZ Ashiro, Iwate Pref. — Takko, Aomori Pref. M. macrodactylus (Temminck) Mym iT 5 Ohwani, Aomori Pref. Tribe Vespertilionini Genus Pipistrellus P. endoi Imaizumi Pie 1 — Ashiro, Iwate Pref. 3 4 San-nohe, Aomori Pref. P. abramus (Temminck) Pia 2 — Hirosaki, Aomori Pref. 5 3 Urawa, Saitama Pref. Genus Nyctalus N. furvus Imaizumi and Yoshiyuki Nf = Iwaizumi, Iwate Pref. N. aviator Thomas Na 2 San-nohe, Aomori Pref. Genus Vespertilio V. superans Thomas Vs 3 5 Hirosaki, Aomori Pref. Tribe Plecotini Genus Barbastella B. leucomelas darjelingensis (Hodgson) Bl 1 — Iwaizumi, Iwate Pref. 1 1 Tamayama, Iwate Pref. Genus Plecotus P. auritus sacrimontis G. Allen Pia — 2 Ohdate, Akita Pref. — 1 Shizunai, Hokkaido 1 — Ohwani, Aomori Pref. Subfamily Miniopterinae Genus Miniopterus M. schreibersi fuliginosus (Hodgson) Mis 1 1 Oga, Akiata Pref. Subfamily Murinae Genus Murina M. silvatica Yoshiyuki Mus 1 1 Fukaura, Aomori pref. 1 — Shizunai, Hokkaido The scientific names of the bats examined were given according to the classification system of Yoshiyuki [56]. of radii and humeri according to Obara and Miyai [39] and Obara [38]. In Myotis pruinosus, chromosome preparations were obtained from the primary culture of tail tissue following the procedure of Obara and Saitoh [41]. G- and C-bandings were carried out according to the ASG method [49] and the BSG method [48], respectively. For NOR- banding the one step silver-staining method [23, 33] was adopted with a slight modification. Identification of the NOR-carrying chromo- somes was done by means of sequential staining of G- and NOR- bandings: after photographing G-banded metaphases, the same metaphases were destained with Carnoy’s fixative and then succes- sively silver-stained and rephotographed. Chromosomes were numbered according to the numbering system of Bickham [11], in which ordinal numbers were given to all of the autosomal arms based on G-banding pattern. Accordingly, a given biarmed chromosome has two arm numbers, e.g., 1/2, 3/4 and 5/6 chromosomes. Since almost half of the chromosomes of the Japanese pipistrelle Pipistrellus abramus have been structurally rear- ranged too much to be given arm numbers in this system, the author’s numbering system [43] was adopted for the Ag-NOR analysis of this Ag-NORs of Japanese Vespertilionid Bats 475 TABLE 2. Cytologic findings of 14 species of bats examined Autosome pair Sex chromosome No. of cells karyotyped Species 2n FN M:SM sSM-ST A x Y Conv. G C N Myn 44 50 3 1 17 M A 8 (21) 5 2 12 (55) Myh 44 52 3 2 16 M A 3 (13) 5 — 7 (33) Myf 44 52 3 2 16 M ST 14 (77) 15 5 21 (125) Mym 44 52 3 2 16 M ST 17 (111) ~=25 — 25 (64) Myp 44 52 3 2 16 SM A 14 (44) 10 4 11 (43) Pie 36 50 7 1 9 A A 7 (45) 14 3 22 (163) Pia 26 44 10 0 2 A A 11 (37) 34 1 35 (165) Nf 44 52 4 2 17 M A 8 (45) 9 — 12 (76) Na 42 50 5 1 15 M A 11 (47) 20 — 26 (160) Vs 38 50 6 1 11 M A 8 (23) 11 — 25 (61) Bl 32 50 9 1 5) M A 6 (26) 10 — 17 (50) Pla 32 50 9 1 5 M A 11 (21) 14 2 19 (57) Mis 46 50 2, 1 19 M A 8 (31) 17 — 15 (91) Mus 44 56 3 4 14 M A 6 (13) 3 — 9 (22) 2n, diploid number of chromosomes; FN, fundamental number; M, metacentric; SM, submetacentric; sSM, small submetacentric; ST, subtelocentric; A, acrocentric; Conv., conventional staining; G, G-band staining; C, C-band staining; N, NOR-band staining. Abbreviations for species name are the same as those in Table 1. Parentheses indicate the number of cells analyzed under microscope. species. : The frequency of Ag-NORs (the mean number of Ag-NORs in each chromosome pair per cell) was calculated for 7-35 metaphases, mainly according to Volleth [51]: in a given chromosome a distinct Ag-NOR was calculated as 1.0 and indistinct one as 0.5, and therefore distinct homozygous Ag-NORs as 2.0, distinct heterozy- gous Ag-NOR as 1.0, indistinct homozygous Ag-NORs as 1.0, indistinct heterozygous Ag-NOR as 0.5 and a pair of distinct and indistinct Ag-NORs as 1.5. XY “ G RESULTS RF G- and C-banding 13 Cytologic findings and analytic data of 14 species of vespertilionid bats examined were summarized in Table 2. The chromosomal findings were essentially consistent with Pd a those of the previous studies [2-4, 25, 27-31, 40-44]. A 22 representative G-banded karyotype of Myotis nattereri is shown in Figure 1 as a standard of the karyotypes of these 14 bat species, since Myotis is regarded as the most “primitive” po & a ee genus among these eight genera examined here [19, 56], and 23 24 25 M. nattereri is considered, from a karyological viewpoint, an original form among the Myotis species of Japan [31]. The Fic. 1. G-banded karyotype of a male Myotis nattereri. Ordinal autosomes consisted of three pairs of large metacentrics [M] numbers indicate autosomal arm numbers. (Nos. 1/2, 3/4 and 5/6), one pair of small submetacentrics 476 T. ONO AND Y. OBARA 19 20 21 22 3h | Cat | Su: ® | i sc Be 2 Hy Se 84 i= Heo / i ae 17 Ghaj GhO@ atin adey 12 14 15 18 Reale ghee 8a Bees oas=| og. 23 24 25 Fic. 2. Comparison of G-banded haploid karyotypes of Myotis nattereri, Nyctalus furvus (Nf), N. aviator (Na), and Vespertilio superans (Vs). Asterisks indicate NOR-bearing chromosomes: No. 8 of Na carries cmcNORs, and No. 15 of Nf, Na and Vs and No. 23 of Vs intNORs, which strictly correspond to the secondary constriction. a | Ke ie w i 1 i ‘ate har hee Sa 21 22 23 . by Baa 8 17 19 a2 é i 8 sae toe Fic. 3. Comparison of G-banded haploid karyotypes of Myotis nattereri (Myn), Barbastella leucomelas darjelingensis (Bl), and Plecotus auritus sacrimontis (Pla). [SM] (No. 16/17) and 17 pairs of medium-to-small acrocentrics [A] (Nos. 7-15 and Nos. 18-25). The sex chromosomes X and Y were a medium-sized M and a small A, respectively. In the standard karyotype all the C-bands were centromeric except for the Y chromosome which was totally heterochromatic. In each species, the chromosome arms were numbered on the basis of G-band homology to the Myotis nattereri chromosomes. Figure 2 shows a composite G-banded karyotype consisting of the haploid sets of four species of the Vespertilioninae (Myotis nattereri, Nyctalus furvus, N. aviator, and Vespertilio superans), in which all chromosome arms could be paired, like a tetraploid cell, in proper quanti- ties among species. The karyotype of Nyctalus furvus (Nf) was almost the same as that of Myotis nattereri (Myn), except for the presence or absence of a distinct secondary constric- tion on the proximal region of the pair No. 15. So, the constitution of uni- and biarmed chromosomes was just the same as Myn and Nf, though some of the A’s of Nf contained comparatively massive C-heterochromatin on their minute Ag-NORs of Japanese Vespertilionid Bats 477 short arms (data not shown). On the contrary, Nyctalus aviator (Na) had fewer A’s than Myn by 2 pairs (Nos. 9 and 10), and the former had more large M’s than the latter by 1 pair (No. 9/10). Similarly, Vespertilio superans (Vs) had fewer A’s than Myn by six pairs (Nos. 7-11 and No. 13), and the former had more large M’s than Myn by three pairs (Nos. 9/10, 11/8 and 13/7). Four biarmed chromosomes, Nos. 1/2, 3/4, 5/6 and 16/17, were common to these four bat species, the chromosomes with the same arm combination necessarily being regarded as homoeologous chromosomes in the same way as all other A’s with the same arm numbers. Thus, the chromosomal relationship of these four bat species could be explained well by Robertsonian rearrangements. Figures 3 and 4 are the composite karyotypes showing the chromosomal relationship of Myn with Barbastella leucomelas darjelingensis (Bl), Plecotus auritus sacrimontis (Pla), and Pipistrellus endoi (Pie). As clearly demonstrated by the pair-matching analysis, Robertsonian rearrangements were the main mechanism of karyotype evolution also in these vespertilionine species. On the other hand, four Myotis species (M. nattereri, M. hosonoi, M. frater kaguyae, and M. macrodactylus) showed quite similar karyotypes with closely resembling chromosome constitution, but differing from each other only in the size of the pair No. 25. The size variation of this pair could be attributed to the duplication of C-heterochromatin as suggested by Harada and Yosida [31]. The other Myotis species, M. pruinosus (Myp) was signifi- cantly different from Myn in the arm ratio of the pair No. 1/2 and the size of the X chromosome. The variation in the arm ratio could be explained by pericentric inversion, and the size variation of the X chromosomes by duplication of C-heterochromatin or translocation of an autosomal element of unknown origin to the X chromosome, as noted by Harada and Uchida [29]. The arm combination of biarmed autosomes of 13 bat species was summarized in Table 3. It is evident from the arm combination analysis that three large M pairs (Nos. 1/2, 3/4 and 5/6) and a small SM pair (No. 16/17) have been well conserved during the course of karyotypic diversification of these bat species except for Miniopterus schreibersi and Pipistrellus abramus. In the former species the pairs Nos. 3/4 and 16/17 were not detected at all. G-banding analysis proved three additional A’s to have been formed from these biarmed pairs through centric fission and pericentric inversion in agreement with Bickham and Hafner [13]. Three large M pairs of Pipistrellus abramus had unusually large C-bands on either side of their centromeres, and the small SM chromo- some corresponding to the pair No. 16/17 was not detected due probably to complicated rearrangements. Other than these three large M pairs (Nos. 1/2, 3/4 and 5/6), Pipistrellus TaBLE 3. Arm combination of the biarmed chromosomes in 13 species of bats examined Sees Combinations of chromosome arms Myotis Myn 1/2 3/4 5/6 16/17 Myh, Myf, Mym 1/2 3/4 5/6 16/17 = 25° Myp Nios 3/4 5/6 Nosy 2 Pipistrellus Pie 1/2 3/4 5/6 16/17' 13/9 11/8 12/10 18/14 Pia 1/2° 3/4° 5/6° 13/9° 14/7* PD Eu) ? Nyctalus Nf 1/2 3/4 5/6 16/17 ~—.25° Na 1/2 3/4 5/6 16/17 ~=9/10 Vespertilio Vs 1/2 3/4 5/6 16/17 9/10 11/8 13/7 Barbastella Bl 1/2 3/4 5/6 16/17 12/9 14/11 15/10 18/13 19/8 20/7 Plecotus Pla 1/2 3/4 5/6 16/17 12/9 14/11 15/10 18/13 19/8 20/7 Miniopterus Mis 1/2 5/6 10 uncertain arm i, pericentric inversion; c, duplicaiton of C-heterochromatin; *, NOR-bearing chromosome. ?, combination. See Table 1 for species abbreviations. 478 Myn Pie Nal /At 3 5 11 J tt a= Die’: \ aw ae et 2 4 6 LAs 8 Hi fm 80 an de 15 19 2073 2 Fic. 4. Comparison of G-banded haploid karyotypes of Myotis nattereri (Myn) and Pipistrellus endoi (Pie). constrictions which locate close to centromere. abramus contained 7 pairs of biarmed chromosomes (Nos. 13/9, 14/7 and 5 pairs of unknown combination) (Table 3). The G-banding pattern of Murina silvatica (Mus) was not obtained. These results are essentially consistent with the earlier study [26] except for discrepancy in the arm combina- tion of Pipistrellus endoi. NOR-banding On the basis of their location on chromosomes, the Ag-NORs detected in this study were classified into two types: cmcNORs and intNORs. The former are located on the minute short arms of A’s, appearing as a centromere cap, and the latter are located within chromosome arms as intersti- b C (A) §-@-8 a ee >? es Fic.5. Typical examples of silver-stained intNORs (A) and cmcNORs (B). The former are from No. 15 chromosomes of Nyctalus furvus (a) and Pipistrellus endoi (b) and No. 23 chromosome of P. endoi (c), respectively, and the latter from Myotis pruinosus, in which each chromosome could not be identified. T. ONO AND Y. OBARA 12 13 18 y—~ ob 78 89 €S; ; oe i act Pi? is pot = ——- SM ST 10 9 a iy Pi. ba br a4. e. ‘9 22) 23) 4424). on, SNA Arrowheads denote secondary p.1., pericentric inversion; SM, submetacentric; ST, subtelocentric; A, acrocentric. tial NORs which correspond to the secondary constrictions, irrespective of their size. Three examples of typical cmcNORs and intNORs are shown in Figure 5. In general, cmcNORs were quite small or fine in size and sometimes of variable occurrence. On the contrary, the one or two pairs of intNORs observed are massive in size and of consistent occurrence. Ag-NORs of the bat species examined here were located only on the A’s irrespective of their types, i.e., “centromeric” or “interstitial”, with the exception of Pipi- strellus abramus. The distribution patterns of cmcNORs and intNORs of 12 bat species analyzed by successive sequential G- and NOR-bandings were summarized in Table 4 for an easy understanding of the interspecific relationship of NOR sites. All five Myotis species examined had multiple cmcNORs (No G-banding analysis was done in Myotis pruinosus). Repre- sentative metaphases of Myotis hosonoi sequentially stained are shown in Figure 6. The NOR patterns and their occur- rence frequency were depicted in Figure 7 with histograms. These Myotis species showed marked variations in NOR distribution, despite their close similarity of karyotypes. The frequency of cmcNORs varied from individual to indi- vidual as well as from species to species, being below 1.0 in many chromosome pairs, while their distribution pattern seemed to be particular, even in consideration of some interindividual variation, to each species, as shown by 1) their steady occurrence in the pair No. 7 of Myotis hosonoi (Myh), 2) their distinct tendency to the small A’s in M. macrodactylus (Mym) and 3) their wide distribution with different combina- tion in M. nattereri (Myn) and M. frater (Myf). Asa whole, Myn, Myh, Myf and Mym carried cmcNORs on 11, 5, 13 and 6 pairs of A’s, respectively. Relatively high frequencies of cmcNORs were found in Nos. 8 and 9 of Myn, Nos. 7 and 23 of Myh, Nos. 8, 18 and 20 of Myf and Nos. 18, 20 and 23 of Mym. Multiple types of cmcNOR were also found in Barbastel- Ag-NORs of Japanese Vespertilionid Bats 479 Fic. 6. Sequentially-stained metaphases of Myotis hosonoi. a, G- banded; b, NOR-banded. Arrowheads denote the NOR sites (a) and cmcNORs (b). la leucomelas (Bl) and Plecotus auritus (Pla): the former had 5 pairs of A’s (Nos. 21-25) in the chromosome complement, all of which carried cmcNORs, and the latter carried cmcNORs on only 4 A’s (Nos. 21-24), in spite of the close similarity of karyotypes (Tables 3 and 4). The frequency of cmcNORs was relatively high in pair No. 24 of B. leucomelas and Nos. 21 and 22 of P. auritus, and the other A’s showed highly variable frequencies of cmcNORs from individual to indi- vidual in both species (Fig. 8). Murina silvatica had multiple cmcNORs similar to those of Myotis species. Unfortunate- ly, no sequential G- and NOR-band stainings were applied to this species. Pipistrellus endoi and Vespertilio superans had intNORs on the pairs Nos. 15 and 23 in common (Table 4 and Figs. 5 and 9). As clearly demonstrated in Figs. 9 and 10, Nyctalus furvus possessed a single pair of intNORs on the pair No. 15, and N. aviator had cmcNORs on the pair No. 8 and intNORs on the pair No. 15. Similarly, Miniopterus schreibersi had cmcNOR on the pair No. 20 and intNOR on the pair No. 23 (Table 4), nevertheless this species carried more A’s than those of Myotis. The intNOR of the pair No. 15 showed a frequency of almost 2.0 in Nyctalus furvus, N. aviator and Vespertilio superans, but under 1.5 in Pipistrellus endoi 4 o 2+ Myn-1 iS) “N i 1 [e4 (e} SEES K nx ay, aes z 123 4 5 6 7 8 91011 12 13 14 15 16 17 18 19 20 21 22 23 24 25 2 Myh-2 | ot 1 (0) 1S) — é 2+ Myh-6 fe) re pe See Se Se Ge 2 xz. 12 3 4 5 6 7 8 91011 12 13 14 15 16 17 18 19 20 21 22 23 24 25 2+ Myf£-6 1 - eres | ro o) 2 (S) ~N a) 1 [e4 9 ib 2 1 Ly os 2h SS Se 123 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 2+ Mym-13 1 a > 2+ Mym-7 12) ~ 9) 1 ie e) Zz 12345 6 7 8 G10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Number of autosome arm Fic. 7. Distribution pattern of cmcNORs in four Myotis species, Myotis nattereri (Myn), M. hosonoi (Myh), M. frater kaguyae (Myf) and M. macrodactylus (Mym). Asterisks indicate the arms of biarmed autosomes, and the numbers following species abbreviation individual numbers. See “materials and methods” for further details. (Table 4), and that of the pair No. 23 of P. endoi showed distinct homozygous Ag-NORs calculated as 2.0. The cmcNOR of the pair No. 8 of N. aviator had a frequency under 1.5. The location of NORs of Pipistrellus abramus was quite exceptional in having intNORs on the proximal region of the long arm of the metacentric pair No. 5 which consists of Nos. 14 and 7 chromosomes of the standard karyotype (Table 4 and Fig. 11). DISCUSSION In general, vespertilionid species show a high degree of the intergeneric variation of karyotypes and a lower degree of 480 T. ONO AND Y. OBARA TABLE 4. Distribution pattern of Ag-NORs in 11 species of bats Number of Number of autosomal arm species specimens analyzed 7 8&8 9 10 IW 12,1385 14> 15> 167 18 19) 20 iy 22 2S ee Myn 1 + + + + + + + + «4+ + + Myh 2 ++ rs + + Myf 4 So ae dE st ES oh ee Mym 5 + + + + + + Pie 6 op Bett Nf 1 A Na 5 + Brot Vs 6 cb A Bl 3 ae SPOS AF Pla 4 ae ak Mis 2 + fb Number of Number of autosome* species specimens analyzed bo (2 Bi oka SS) 2 OES is eS are Oana lel ees) Pia 4 Bits Abbreviations for species name are the same as those in Table 1. + and 2 B1-2 AQ fh fh a) Ag AA Q 8 Siu tere. ae ee al BA~“An 06 AR Ga AA an se aa B r4 2 Bl1-3 15 1 % =i x Serer srs rrr r srry srs s sch “ait 123 4 5 6 7 8 910 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 * ‘. aoe ae it Number of autosome arm io * g8 # & i aa Pe | Fic. 8. Distribution pattern of cmcNORs in three specimens of ————— W#—————_- XX Plecotus auritus sacrimontis (top) and in three specimens of Barbastella leucomelas darjenlingensis (bottom). fy. Dy be 4 4 “a: ono . ) ~< I Fic. 9. NOR-banded karyotypes of three Vespertilionini species, ¢$ Bf a6 C) 94 @e>an es es C Nyctalus furvus (A), N. aviator (B), and Vespertilio superans (C). Arrowheads indicate Ag-NORs. M, large metacentrics. 5 23 Ag-NORs of Japanese Vespertilionid Bats 481 Nf Na Vs 11, Oy" av & N ey 1 af Mes 8 A; > pi if 156,72 Out > hag" 238 ett — i 6 Fic. 10. Comparison of differentially stained NOR-bearing chromosomes of Nyctalus furvus, N. aviator and Vespertilio superans. Asterisks indicate Ag-NORs. Conv., conventional- ly stained; G, G-banded; N, NOR-banded. intrageneric variation [6, 7, 10, 17, 46]. The vespertilionid bats of Japan are no exception to this view: the diploid numbers in eight genera of vespertilionid bats studied here varied widely from 2n=26 to 2n=46. On the other hand, each of five species of Myotis examined had similar karyotypes, showing the same chromosome number and only slightly variable chromosome constitution (Tables 2 and 3). In spite of such remarkable karyotypic uniformity, these Myotis species showed considerable variation in the location of NORs, and in addition their distribution patterns seemed to be species-specific, though a certain degree of intraspecific (inter- and intraindividual) variation in NOR distribution was observed (Table 4 and Figs. 6 and7). A similar relationship has already been found in the European species of Myotis [51]. Taken together, it seems likely that variation in NOR location can be used to trace paths of taxonomic diversifica- Fic. 11. standard karyotype. (arrowheads). tion of the Myotis group. Similar views have been presented for several other taxonomic groups of animals such as Pisces, Amphibia and Rodentia [1, 21, 24, 35, 54, 57, 58]. The distribution pattern of NORs of Myotis nattereri examined by Volleth [51] differs significantly from that of the same species collected from Japan in that the former carry only four pairs of cmcNORs and the latter eleven pairs of cmcNORs. ‘This discordance in NOR pattern may be attri- buted to the subspecies distinctness; the former is the nomi- nate subspecies Myotis n. nattereri, and the latter a Japanese subspecies Myotis n. bombinus {19, 20]. If this is true, the distribution analysis of NORs should be useful in detecting geographic differentiation of the karyotype within a single species or species complex. As summarized in Table 4, intNORs showed a markedly low variation in their distribution, whereas in striking contrast cmcNORs were highly variable in their distribution among species. Further, intNORs were mostly massive in size with steady occurrence, and cmcNORs were distributed to many A’s with various combinations and variable occurrence (Figs. 7 and 9 and Table 4). Since so-called NOR-association makes chromatid exchange physically possible irrespective of homologous or heterologous chromosomes, these findings suggest that cmcNORs of vespertilionid bats may be subject to transfer to another A’s through NOR-association and interchromosomal chromatid exchange, while intNORs may be liable to include cmcNORs into themselves during the course of NOR-association, resulting in growing intNORs and diminishing cmcNORs. The NOR transferred to a given interstitial position of chromosomes may be stabilized in their gene activity as well as in their structural organiza- tion. The steady occurrence of intNORs plainly reflects this point of view. Our view on the marked stability of intNORs is well supported by the observation of Volleth [51] on the NORs of the European Pipistrellus species: the most “primi- tive” pipistrelle P. pipistrellus has no intNORs on any of the chromosomes, but multiple cmcNORs on several A’s, and B A, G-banded metaphase of Pipistrellus abramus. Arrowheads indicate a No. 5 homologue which consists of Nos. 7 and 14 of the B, same metaphase as in (A) after silver-staining, showing intNORs on the proximal site of the long arm 482 T. ONo AND Y. OBARA “advanced” pipistrelles, P. savii, P. kuhli and P. nathusii have a typical secondary constriction on the proximal region of the long arm of the pair No. 15, which is the site of consistently-occurring intNOR. In this context, note that the little brown bats, Myotis, are among the most “primitive” taxa in the family Vespertilionidae [19] and all Myotis species so far examined showed cmcNORs regardless of their dis- tribution in Europe or Japan. So, assuming that the cmcNORs have been conserved intact without altering in the genus Myotis, it follows, in contrast to Volleth’s views, that the ancestral form of NORs of the family Vespertilionidae is not the interstitial type, but a cmc type. The tube-nosed bats Murina, belonging to the subfamily Murinae, are characterized by a projecting tube-like nose. In spite of such a marked phenotypical specialization, Murina aurata closely resembled Myotis nattereri of Vespertilioninae in the NOR type (cmcNORs) as well as in the karyotypic profile. Therefore, Murina is considered to have been de- rived from the ancestral stock of Myotis without alteration in either NOR type nor chromosome morphology, unlike the general tendency of karyotype evolution in vespertilionid bats. Karyotypes of six genera excluding Myotis and Murina have differentiated, more or less,through Robertsonian rear- rangements, as revealed by the pair-matching analysis of G-banded chromosomes. As seen in Tables 3 and 4, these bats could be classified into the following three karyotypic lineages on the basis of the NOR type and arm combination of chromosomes: (1) Pipistrellus-Nyctalus-Vespertilio, (2) Barbastella-Plecotus, and (3) Miniopterus. The first lineage is characterized by the typical intNORs on the homoeologous chromosomes (Table 4, Figs.5, 9 and 10), although their conventional karyotypes differ largely from each other be- cause of chromosome rearrangements by centric fusion. The second lineage, Tribe Plecotini, carries multiple cmcNORs on the smaller A’s (Fig. 8). Their karyotypes could also be regarded as the chromosome change-mediated products resulting from centric fusion as in the first lineage, but here the arm combination of the large M’s was different from that of the first lineage (Table 3), suggesting that the Barbastella-Plecotus lineage has diverged from the ancestral Myotis-like form through its own process different from the first lineage. The third lineage is an offshoot derived from the ancestral stock of Myotis through centric fission, having two pairs of NORs: one pair of intNORs and the other, cmcNORs. The pipistrelle species P. endoi and P. abramus of the first lineage were highly different from each other in their chromosome constitution as well as in their diploid number, although they are congeneric. P. endoi is fairly conservative in its arm constitution, following the general tendency of karyotype evolution in Vespertilionidae, whereas in striking contrast P. abramus attained a higher degree of non- Robertsonian rearrangements [26, 44]. With regard to Ag- NORs, the NOR-bearing chromosomes of P. endoi were identified as Nos. 15 and 23, while those of P. abramus were identified as No.5, formed by the centric fusion between Nos. 14 and 7 chromosomes of the standard karyotype (Fig. 11). However, both species share at least 4 pairs of their large M’s (Nos. 1/2, 3/4, 5/6 and 13/9) in their karyotypes (Table 3). Therefore, P. endoi and P. abramus may have derived from a common ancestor. The vespertilionid bats carry few intNORs on biarmed chromosomes. The thick-thumbed pipistrelle Glischropus tylopus is an exceptional case with an intNOR on the proxi- mal region of the long arm of one M pair [52], and this NOR-bearing chromosome seemed to correspond well with the pair No.5 of P. abramus in G-banding pattern. Such coincidence in the type and location of NORs strongly suggests that G. tylopus is closely related to the lineage of P. abramus. Further, P. abramus (2n=26) remarkably resem- bles the American yellow bat Lasiurus intermedius (2n=26) in the chromomycin Az fluorescence pattern as well as in the chromosome constitution and G-banding pattern ([12], Obara, unpublshed data). This fact may also suggest a close relationship of P. abramus with L. intermedius. The No. 15 chromosomes of Nyctalus furvus, N. aviator, and Vespertilio superans, all of which carry a typical secon- dary constrictions with a high activity of NORs, are probably homoeologous with each other, although the former two exhibit partial loss of the proximal region of the long arm (Figs. 2 and 10), and the pair-matching analysis revealed that the centric fusion rearrangements have occurred at first in the combination of Nos. 9 and 10, and then in the combination of Nos. 11 and 8, and Nos. 13 and 7. Further, Pipistrellus endoi also carries intNORs on the pair No. 15 like those three Vespertilionini species (Fig. 4). Therefore, it is possible that the noctule and particoloured bats Nyctalus and Vesperti- lio are the offshoots from a common ancestor of the “ad- vanced” pipistrelles including P. savii, P. kuhli, P. nathusii and P. endoi: hence, both Nyctalus and Vespertilio belong to the Pipistrellus lineage. This type of secondary constriction and Ag-NORs of the pair No. 15 has been observed also in European species of Nyctalus, Vespertilio and Pipistrellus [15-18, 22, 50, 51]. The relationship between NOR variation and centric fusion in human D- and G-group chromosomes has been investigated by several researchers. They found that in most cases the M’s resulting from centric fusion are lacking in NORs [5, 34, 36, 37, 59]. This pattern of NOR variation through centric fusion resembles that seen in the vesperti- lionid bats examined here. One exceptional case was found in Pipistrellus abramus which carried intNORs on the proxi- mal region of one large M pair formed by centric fusion: the other centric fusion-mediated M’s (11/8 in Vespertilio super- ans and 12/9, 14/11, 15/10, 18/13, 19/8 and 20/7 in Barba- stella leucomelas and Plecotus auritus) were lacking in NORs. Nevertheless, cmcNORs have been present on most of the arms before fusion. According to Stahl ef al. [47] who attempted to explain the relationship between centric fusion and cmcNORs from an electronmicroscopic viewpoint, whether or not cmcNORs remain on the proximal region of Ag-NORs of Japanese Vespertilionid Bats 483 the M’s resulting from centric fusion depends on the point of the breakage/reunion at which centric fusions are induced. Therefore, in the case of vespertilionid bats breakage/reu- nion might have occurred in a point, making exclusion of cemcNORs possible. In the exceptional case of Pipistrellus abramus, breakage/reunion might have occurred in a point which made inclusion of cmcNORs possible. Recently, Harada [26] presented a karyological dendro- gram of Japanese vespertilionid bats based on the G- and C-banding patterns. The present chromosome findings sup- port his classification system, and the findings from the silver staining reinforced this support. The Plecotini Barbastella leucomelas and Plecotus auritus and the Murinae Murina silvatica carried multiple cmcNORs as in Myotis, whereas the Pipistrellus lineage carried intNORs. Thus, the type and location of Ag-NORs are reflected, as the third cytogenetic parameter, by the phylogenetic lineage. So far as cyto- genetic parameters are concerned, the former group has a closer relationship with Myotis than the latter group. The relationship of NOR variation with karyotype evolu- tion in vespertilionid bats could be clarified by the methods of molecular cytogenetic analysis such as in situ hybridization and by further investigation of the present and other bat species. ACKNOWLEDGMENTS The authors are grateful to Mr. Mitsuru Mukohyama of Aomori Prefectural San-nohe High School, for providing and identifying the bat specimens and to Mr. Yoshihiko Misumi of Ichinohe-cho, Iwate Pref., Mr. Azuma Abe of Aomori Prefectural Hirosaki High School and Mrs. Michiko Kiyomiya of Urawa, Saitama Pref., who cooper- ated in collecting the research materials. Our sincere thanks are also due to Emeritus Professor Kazuo Saitoh of Hirosaki University and Dr. Oscar G. Ward of the Department of Ecology and Evolu- tionary Biology, University of Arizona, for reading and refining the manuscript with expert criticism. REFERENCES 1 Amemiya CT, Gold JR (1988) Chromosomal NORs as taxono- mic and systematic characters in North American cyprinid fishes. Genetica 76: 81-90 2 Ando K, Tagawa T, Uchida TA (1977) Considerations of karyotypic evolution within Vespertilionidae. Experientia 33: 877-879 3 Ando K, Tagawa T, Uchida TA (1980) The C-banding pattern of 6 Japanese species of Vespertilioninae bats (Mammalia: Chiroptera). Experientia 36: 653-654 4 Ando K, Harada M, Uchida TA (1987) A karyological study on five species of Myotis and Pipistrellus, with special attention to comparison of their C-band materials. J Mammal Soc Japan 12: 25-29 5 Babu KA, Verma RS (1985) Structural and functional aspects of nucleolar organizer regions (NORs) of human chromosomes. Internat Rev Cytol 94: 151-176 6 Baker RJ (1970) Karyotypic trends in bats. In “Biology of Bats, Vol. 1” Ed by WA Wimsatt, Academic Press, New York and London, pp 65-96 7 Baker RJ, Bickham JW (1980) Karyotypic evolution in bats: 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 extensive and conservative chromosomal evolution in closely related taxa. Syst Zool 29: 239-253 Baker RJ, Patton JL (1967) Karyotypes and karyotypic varia- tion of North American vespertilionid bats. J Mammal 48: 270-286 Bengtsson BO (1980) Rates of karyotype evolution in placental mammals. Hereditas 92: 37-47. Bickham JW (1979) Chromosomal variation and evolutionary relationships of vespertilionid bats. J Mammal 60: 350-363 Bickham JW (1979) Banded karyotypes of 11 species of Amer- ican bats (Genus Myotis). Cytologia 44: 789-797 Bickham JW (1987) Chromosomal variation among seven spe- cies of Lasiurus bats (Chiroptera: Vespertilionidae). J Mam- mal 68: 837-842 Bickham JW, Hafner JC (1978) A chromosomal banding study of three species of vespertilionid bats from Yugoslavia. Genet- ica 48: 1-3 Bush GL, Case SM, Wilson AC, Patton JL (1977) Rapid speciation and chromosomal evolution in mammals. Proc Natl Acad Sci USA 74: 3942-3946 Capanna E (1968) Some considerations on the evolution of the karyotype of Microchiroptera. Experientia 24: 624-626 Capanna E, Civitelli MV (1967) I cromosomi di Pipistrellus savii. Caryologia 20: 265-272 Capanna E, Civitelli MV (1970) Chromosomal mechanisms in the evolution of chiropteran karyotype: Chromosomal tables of Chiroptera. Caryologia 23: 79-111 Capanna E, Romanini MGM (1971) Nuclear DNA content and morphology of the karyotypes in certain Palaearctic Mi- crochiroptera. Caryologia 24: 471-482 Corbet GB (1978) The mammals of the Palaearctic region: A taxonomic review. Cornell Univ. Press, Ithaca, New York, Ist edition Corbet GB, Hill JE (1980) A world list of mammalian species. British Museum (Natural History), London and Cornell Univ. Press, Ithaca, New York, 1st edition Dev VG, Tantravahi R, Miller DA, Miller OJ (1977) Nu- cleolus organizers in Mus musculus subspecies and in the RGA mouse cell line. Genetics 86: 389-398 Dulic B, Soldatovic B, Rimsa D. (1967) La formule chromoso- mique de la noctule, Nyctalus noctula Schreber (Mammalia, Chiroptera). Experientia 23: 945-946 Gold JR, Ellison JR (1983) Silver staining for nucleolar orga- nizing regions of vertebrate chromosomes. Stain Technol 58: 51-55 Gold JR, Amemiya CT (1986) Cytogenetic studies in North American minnows (Cyprinidae). XII. Patterns of chromo- somal nucleolus organizer region variation among 14 species. Can J Zool 64: 1869-1877 Harada M (1973) Chromosomes of nine chiropteran species in Japan (Chiroptera). La Kromosomo 91: 2885-2895 Harada M (1988) Karyotypic evolution in the family Vesperti- lionidae. Honyurui Kagaku 28: 69-83 (in Japanese) Harada M, Ando K, Uchida TA, Takada S (1987) Karyotypic evolution of two Japanese Vespertilio species and its taxonomic implications (Chiroptera: Mammalia). Caryologia 40: 175-184 Harada M, Ando K, Uchida TA, Takada S (1987) A karyo- logical study on two Japanese species of Murina (Mammalia: Chiroptera). J Mammal Soc Japan 12: 15-23 Harada M, Uchida TA (1982) Karyotype of a rare species, Myotis pruinosus, involving pericentric inversion and duplicated translocation. Cytologia 47: 539-543 Harada M, Uchida TA, Yosida TA, Takada S (1982) Karyo- logical studies of two Japanese noctule bats (Chiroptera). Caryologia 35: 1—9 484 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 T. ONO AND Y. OBARA Harada M, Yosida TH (1978) Karyological study of four Japanese Myotis bats (Chiroptera, Mammalia). Chromosoma 65: 283-291 Howell WM (1982) Selective staining of nucleolus organizer regions (NORs). Cell Nucleus 11: 89-142 Howell WM, Black DA (1980) Controlled silver-staining of nucleolus organizer regions with a protective colloidal develop- er: a 1-step method. Experientia 36: 1014-1015 Hurley JE, Pathak S (1977) Elimination of nucleolus organiz- ers in a case of 13/14 Robertsonian translocation. Hum Genet 35: 169-173 Mahony MJ, Robinson ES (1986) Nucleolar organizer region (NOR) location in karyotypes of Australian ground frogs (Fami- ly Myobatrachidae). Genetica 68: 119-127 Mattei MG, Mattei JF, Ayme S, Giraud F (1979) Dicentric Robertsonian translocation in man. Hum Genet 50: 33-38 Mikkelsen M, Basli A, Poulsen H (1980) Nucleolus organizer regions in translocations involving acrocentric chromosomes. Cytogenet Cell Genet 26: 14-21 Obara Y (1982) Comparative analysis of karyotypes in the Japanese mustelids, Mustela nivalis namiyei and M. erminea nippon. J Mammal Soc Japan 9: 59-69 Obara Y, Miyai T (1981) A preliminary study on the sex chromosome variation in the Japanese house shrew, Suncus murinus riukiuanus. Jpn J Genet 56:365-371 Obara Y, Saitoh K (1977) Chromosome studies in the Japanese vespertilionid bats: IV. Karyotypes and C-banding pattern of Vespertilio orientalis. Jpn J Genet 52: 159-161 Obara Y, Saitoh K (1977) Some cytogenetic and cell physio- logic aspects of a chiropteran cell line. Sci Rep Hirosaki Univ 24: 73-80 Obara Y, Tomiyasu T, Saitoh K (1976) Chromosome studies in the Japanese vespertilionid bats. I. Karyotypic variations in Myotis macrodactylus Temminck. Jpn J Genet 51: 201-206 Obara Y, Tomiyasu T, Saitoh K (1976) Chromosome studies in the Japanese vespertilionid bats. II. G-banding pattern of Pipistrellus abramus Temminck. Proc Japan Acad 52(B): 383- 386 Obara Y, Tomiyasu T, Saitoh K (1976) Chromosome studies in the Japanese vespertilionid bats. III. Preliminary observation of C-bands in the chromosomes of Pipistrellus abramus Tem- minck. Sci Rep Hirosaki Univ 23: 39-42 Patton JC, Baker RJ (1978) Chromosomal homology and evolution of phyllostomatoid bats. Syst Zool 27: 449-462 46 47 48 49 50 51 52 53 54 25) 56 57 58 59 Rushton AR (1970) Cytotaxonomy and chromosomal evolu- tion of the bats. Ann Genet Sel Anim 2: 457-479 Stahl A, Luciani M, Hartung M, Devictor M, Berge-Lefrang JL, Guichaous M (1983) Structural basis for Robertsonian trans- locations in man: Association of ribosomal genes in the nucleolar fibrilar center in meiotic spermatocytes and oocytes. Proc Natl Acad Sci USA 80: 5946-5950 Sumner AT (1972) A simple technique for demonstrating centromeric heterochromatin. Exp Cell Res 75: 304-306 Sumner AT, Evans HJ, Buckland RA (1971) New technique for distinguishing between human chromosomes. Nature New Biol 232: 31-32 Volleth M (1985) Chromosomal homologies of the genera Vespertilio, Plecotus and Barbastella (Chiroptera: Vespertilio- nidae). Genetica 66: 231-236 Volleth M (1987) Differences in the location of nucleolus organizer regions in European vespertilionid bats. Cytogenet Cell Genet 44: 186-197 Volleth M, Yong HS (1987) Glischropus tylopus, the first known old-world bat with an X-autosome translocation. Ex- perientia 43: 922-924 White MJD (1978) Modes of speciation. Co, San Francisco, 455pp Yamakage K, Nakayashiki N, Hasegawa J, Obara Y (1985) G-, C- and NOR-banding patterns on the chromosomes of the Japanese grass vole, Microtus montebelli montebelli, with special attention to the karyotypic comparison with the root vole, M. oeconomus. J Mammal Soc Japan 10: 209-220 Yoshida MA, Takagi N, Sasaki M (1983) Karyotypic kinship between the blue fox (Alopex lagopus Linn.) and silver fox (Vulpes vulpes Desm.). Cytogenet Cell Genet 35: 190-194 Yoshiyuki M (1989) A systematic study of the Japanese Chi- roptera. Nat Sci Mus Tokyo, vi+242 pp Yosida TH (1978) A preliminary note on silver-stained nucleo- lar organizer regions in the black and Norway rats. Proc Japan Acad 54(B): 353-358 Yosida TH (1979) A comparative study on nucleolus organizer regions (NORs) in 7 Rattus species with special emphasis on the organizer differentiation and species evolution. Proc Japan Acad 55 (B): 481-486 Zankl H, Hahmann S (1978) Cytogenetic examination of the NOR activity in a proband with 13/13 translocation and in her relatives. Hum Genet 43: 275-279 WH Freeman and ZOOLOGICAL SCIENCE 11: 485-490 (1994) © 1994 Zoological Society of Japan Acoustic Characteristics of Treefrogs from Sichuan, China, with Comments on Systematic Relationship of Polypedates and Rhacophorus (Anura, Rhacophoridae) MasaFumi Matsut! and GAN-FU Wu2 "Graduate School of Human and Environmental Studies, Kyoto University, Sakyo-ku, Kyoto 606, Japan, and *Chengdu Institute of Biology, Academia Sinica, P. O. Box 416, Chengdu, Sichuan, People’s Republic of China ABSTRACT— Advertisement call characteristics of Polypedates chenfui, P. dugritei, and P. omeimontis, all from Sichuan, China, are described. Calls of the three species differ considerably from each other both in temporal and frequency patterns. Acoustically, these three species cannot be differentiated from some Rhacophorus species from Japan and Taiwan, and the systematic relationship of Polypedates and Rhacophorus needs reassessment. INTRODUCTION Rhacophorus Kuhl et van Hasselt, 1822 and Polypedates Tschudi, 1838 represent two major genera among the treefrog subfamily Rhacophorinae [4]. There are, however, con- flicting opinions about the taxonomic relationship of these two genera. Polypedates has long been synonymized with Rhacophorus [13,18], but Liem [12], in revising the family Rhacophoridae, split the two genera on the basis of adult morphology. Later, Dubois [3] placed Polypedates as a synonym of Rhacophorus after reinterpreting Liem’s data on adults [12] and utilizing Inger’s data on tadpoles [6]. On the other hand, Jiang et al. [7] considered that Polypedates and Rhacophorus are separate lineages in China. Similarly, Channing [1] reanalyzed Liem’s data [12] and showed that the two genera were not sister groups. Thus there is at present disagreement about the taxonomic relationship of Polype- dates and Rhacophorus. On the other hand, non-morphological characteristics including acoustic ones have been studied on some members of these two genera, and their bearing on systematics has been discussed [5, 9, 10, 11, 15]. The data hitherto accumu- lated, however, are still insufficient to utilize for outlining phylogenetic relationship of Polypedates and Rhacophorus. Regarding acoustic data of east Asian species, the knowledge of members from Japan and Taiwan is considerable, but nothing is known about Chinese members. Although China is famous for its rich amphibian fauna [21], very few studies have been made on the call characteristics of frog species [16, 17], and members of Polypedates and Rhacophorus are not exceptions. In order to better understand systematic rela- tionship of these two genera, acoustic data on Chinese members are badly needed. In this communication, we will report acoustic characteristics of three rhacophorine species Accepted May 16, 1994 Received Feburary 7, 1994 from China, i. e., Polypedates omeimonitis Stejneger, 1924, P. chenfui (Liu, 1945), and P. dugritei David, 1871, and discuss phylogenetic relationship of the two genera. MATERIALS AND METHODS Recordings of calls were made in the field by the junior author using a cassette tape recorder (Sony WM-D3) with an external microphone (Sony ECM-909A). Ambient temperature was mea- sured at the time of recording. Calls of P. chenfui were recorded on Mt. Emei-shan at the altitude of 1400 m, Sichuan, China, on 13 May 1993. Air tempera- ture at the time of recording was 13.0°C. Males were observed to call simultaneously with P. omeimontis. For the latter species, calls were recorded on Mt. Emei-shan at the altitude of 1400 m, on 13 and 15 May 1993. Air temperatures at the time of recording were 13.0°C and 11.0°C, respectively. Calls of P. dugritei were recorded on Mt. Wa-shan, Sichuan, China, at the altitude of 1600 m on 1 June 1993, and at the altitude of 2600 m on 2 June 1993. Air temperature at the time of each recording was 8.0°C and 4.0°C, respectively. For comparisons, data of the following species of Japanese and Taiwanese Rhacophorus are used: R. arboreus (Honshu, Japan); R. schlegelii (Honshu, Japan); R. v. viridis (Okinawajima and Kume- jima, eastern Ryukyus); R. owstoni (Ishigakijima and Iriomotejima, western Ryukyus); R. prasinatus (northern Taiwan). The recorded calls were analyzed using computer programs, SoundEdit Vers. 2 or SoundEdit Pro (MacroMind-Paracomp, Inc.) by a Macintosh computer. Advertisement calls as defined by Weils [19] were compared among the three species. In the following description, note means a pulse group, note duration the time from the beginning of the first pulse to the end of the last pulse in a note, and pulse repetition rate number of pulses per second. In order to examine relationships among call parameters, analysis-of-covariance (ANCOVA) was performed. Kruskal-Wallis tests with nonpar- ametric multiple comparisons (=Dunn test [20]) or Mann-Whitney U tests were performed to detect the presence or absence of differences in the frequency distributions. The significance level was set at 0.05. 486 M. Matsur AND G.-F. Wu TABLE 1. Call characteristics of three Chinese rhacophorid species (Mean+1SD, followed by sample size) Note Species duration (sec) Polypedates dugritei 4.0°C 10 pulsed note 0.700 1 11 pulsed note 0.895 2 12 pulsed note 0.970 + 0.046 6 13 pulsed note 0.965 +0.025 10 14 pulsed note 1.015 2D Polypedates dugritei 8.0°C 2 pulsed note 0.075 1 3 pulsed note 0.190 1 6 pulsed note 0.310 2 7 pulsed note 0.445 2) 8 pulsed note 0.517 3 9 pulsed note 0.510+0.029 7 10 pulsed note 0.637 + 0.023 3 11 pulsed note 0.700 1 Polypedates omeimontis 11.0°C 3 pulsed note 0.318+0.011 6 4 pulsed note 0.460 + 0.022 42 Polypedates omeimontis 13.0°C 2 pulsed note 0.166 +0.015 4 3 pulsed note 0.325 +0.018 13 4 pulsed note 0.439 +0.007 13 5 pulsed note 0.584 1 Polypedates chenfui 13.0°C 2 pulsed note 0.158 1 4 pulsed note 0.385 +0.027 8 5 pulsed note 0.514+0.018 34 6 pulsed note 0.645 + 0.042 9 Pulse Initial Climax rate frequency frequency (N of pulse/sec) (Hz) (Hz) 14.29 1500 1600 1 1 1 12.30 1275 1525 2 2 2 12.39 +0.60 1391.7+159.4 1675.0+ 41.8 6 6 6 13.48 + 0.34 1445.0+170.3 1657.5+ 39.2 10 10 10 13.80 1350 1625 2D 2 2 26.67 1450 2500 1 1 1 15.79 1150 1450 1 1 1 19.54 1412.5 1875 2 2 2 15.83 1075 1750 2 2, 2 15.50 1191.7 1766.7 3 3 3 17.69+0.97 1192.8+196.7 1764.34 69.0 fh 7 7 15.72 +£0.58 1133.34 57.7 1900.0 + 132.3 3 3 3 15.71 1050 1900 1 1 1 9.45 +0.32 865.8+ 30.1 936.7+ 21.6 6 6 6 8.71+0.41 865.5+ 21.2 9771+ 49.8 42 42 42 12.12+1.16 817.54 53.8 827.5+ 38.6 4 4 4 9.25+0.52 804.64 35.5 884.6+ 68.1 13 13 13 9.11+0.14 845.4+ 32.3 913.14 48.9 13 13 13 8.56 770 850 1 1 1 12.64 2000 2100 1 1 1 10.44+0.70 2120.0+102.1 2348.8+ 53.6 8 8 8 9.74+0.33 2082.1+133.5 2334.4 + 132.7 34 34 34 9.35+0.61 2033.3+111.8 2318.9+ 125.8 9 9 9 Rhacophorid Acoustic Characteristics 487 RESULTS Polypedates chenfui The call recorded at 13.0°C was a well pulsed note (Fig. 1B) and included two to six pulses. The note duration increased with increasing number of pulses, and the mean duration varied from 0.158 sec in the note with two pulses to 0.645 sec in the note with six pulses (Table 1). The duration differed significantly among the notes with different number of pulse (Dunn’s multiple comparison test, P<0.05). The relationship of the number of pulse (X) and the note duration (Y) was expressed as Y =0.126X—0.116 (N=52, r=0.965, P <0.01). The pulse repetition rate tended to decrease with the increment of the pulse number, means varying from 12.64 in the note with two pulses to 9.35 in the note with six pulses. The rate was significantly greater in the note with four pulses than in the notes with five or six pulses (Dunn’s multiple comparison test, P<0.05), but the latter two did not differ from each other. The relationship between the number of pulse (X) and the pulse repetition rate (Y) was Y= —0.682X +13.221 (N=52, r=—0.718, P<0.01). Each pulse had harmonics, and a slight frequency modulation was seen within a note (Fig.1A). The mean dominant frequency in the initial pulse was about 2000-2120 Hz, but it slightly increased to 2100-2349 Hz in the climax pulse. In the climax pulse, the second dominant frequency was about 6000-6500 Hz, and six harmonic bands in total were apparent between 0-—7500 Hz. Average harmonic interval, therefore, was about 1250 Hz, and this value corresponded to the fundamental frequen- cy. The first dominant frequency, therefore, was the second harmonic and the second corresponded to the fifth harmonic of the spectrogram. The dominant frequency of either the initial or the climax pulses did not differ significantly among FREQUENCY (KHZ) : 0.3 0.6 0.9 1.2 TIME IN SEC Fic. 1. Sonagrams (A, C) and sound wave forms (B, D) of adver- tisement calls of Polypedates chenfui (A, B, recorded at 13.0°C) and P. omeimontis (C, D, recorded at 11.0°C). the notes with different number of pulse (Kruskal-Wallis test, P>0.05). Thus, there were insignificant correlations be- tween the number of pulse (X) and dominant frequencies (Y) of either the initial (r= —0.111, P>0.05) or the climax pulse (r=0.097, P>0.05). Polypedates omeimontis Calls of P. omeimontis included several call types, but only the advertisement call [8, 19] is considered here. The advertisement call was a well pulsed note (Fig. 1C, D) and at 11.0°C, it included three or four pulses. The mean note duration of 0.318sec in the note with three pulses was significantly shorter than 0.460 sec in the four pulsed note (Mann-Whitney U test, P<0.05). The mean pulse repeti- tion rate in the three pulsed note (9.45) was significantly larger than that in the four-pulsed note (8.71; Mann-Whitney U test, P<0.05). Each pulse had harmonics, but they are usually not clear in the initial pulse. A weak frequency modulation was seen within a note. Parameters of frequen- cies did not differ between notes with three and four pulses (Mann-Whitney U test, P>0.05). The dominant frequency in the initial pulse was about 866 Hz, but it increased to 936- 977 Hz in the climax pulse. In the climax pulse, the second dominant frequency was about 2300 or 2900 Hz. Seven harmonic bands in total could be traced between 0-7350 Hz. Average harmonic interval was about 1050 Hz and corres- ponded to the fundamental frequency. The first dominant frequency was judged to be the fundamental and the second corresponded to the second or third harmonic of the spectro- gram. The dominant frequency in either the initial or the climax pulse did not differ between the notes with three and four pulses (Mann-Whitney U test, P>0.05). The calls recorded at 13.0°C had basically similar traits, but the number of pulses included in a note tended to be smaller than in 11.0°C (median=three pulses in 13.0°C, compared with four in 11.0°C). The mean note durations varied from 0.166 sec in the note with two pulses to 0.584 sec in the note with five pulses (Table 1), but they were statisti- cally not different (Kruskal-Wallis test, P>0.95). Similarly, the note durations did not differ significantly between the notes with the same number of pulses, and recorded at 13°C and 11.0°C (Mann-Whitney U test, P>0.05). The pulse repetition rates varied from 12.12 in the note with two pulses to 8.56 in the note with five pulses, but their difference was insignificant (Kruskal-Wallis test, P>0.05). In the four- pulsed notes, repetition rate in 13.0°C (9.11) was larger than that in 11.0°C (8.71; Mann-Whitney U test, P<0.05). The dominant frequency slightly increased from 777-845 Hz in the initial pulse to 828-913 Hz in the climax pulse. Domi- nant frequencies in the initial and climax pulses were signi- ficantly lower in some calls recorded at 13.0°C than in 11.0°C (initial pulse in the three-pulsed note, and climax pulse in the four-pulsed note: Mann-Whitney U test, P<0.05). Comparisons with calls of syntopic P. chenfui simul- taneously recorded at 13.0°C, using the four-pulsed note resulted in the followings. The note duration of P. 488 M. Matsur AND G.-F. Wu omeimontis was significantly longer than that of P. chenfui (Mann-Whitney U test, P<0.05), and in the pulse repetition rate P. omeimontis was slightly smaller than in P. chenfui (Mann-Whitney U test, P<0.05). Much greater inters- pecific differences were found in frequency characteristics. The dominant frequency in the climax pulse was significantly lower in calls of P. omeimontis than in P. chenfui (Mann- Whitney U test, P<0.01), so was the dominant frequency in the initial pulse (Mann-Whitney U test, P<0.01). The relationship of the number of pulse (X) and the note duration (Y) was expressed as Y=0.166X—0.201 (N=48, r =0.762, P<0.01) and Y=0.131X—0.079 (N=31, r=0.986, P<0.01) in the calls recorded at 11°C and 13°C, respectively. The slope of the former equation was significantly steeper than the latter (ANCOVA, P<0.05). The latter slope was not significantly different from syntopic R. chenfui. The number of pulse (X) and the pulse repetition rate (Y) had the relationships of Y= —5.951X+ 32.773 (N=48, r= —0.660, P <0.01) and Y= —1.043X+ 13.038 (N=31, r=—0.694, P< 0.01) in the calls recorded at 11°C and 13°C, respectively. The slope of the former regression line was significantly steeper than that of the latter (ANCOVA, P<0.05), which in turn was insignificantly different from the slope in syntopic R. chenfui (ANCOVA, P>0.05). In the calls recorded at 11.0°C, there were significant correlations between the num- ber of pulse (X) and dominant frequencies (Y) of both the initial (Y= —86.167X+ 1226.72, N=48, r=—0.495, P< 0.01) and the climax pulse (Y = —53.048X + 1197.25, N=48, t= —0.353, P<0.01), but correlations were insignificant in the calls recorded at 13.0°C (P>0.05). Polypedates dugritei The call was again a well pulsed note (Fig. 2B, D) and at 4.0°C, it included ten to 14 pulses. The note duration Pr wb eT | 11 a mt || ‘Ie © ; aia Fis ell tal cli FREQUENCY (KHZ) 0 0.3 0.6 0.9 1.2 4 TIME IN SEC Fic. 2. Sonagrams (A, C) and sound wave forms (B, D) of adver- tisement calls of Polypedates dugritei (A, B, recorded at 4.0°C; C, D recorded at 8.0°C). tended to increase with increasing number of pulses, and the mean duration varied from 0.700 sec in the note with ten pulses to 1.015 sec in the note with 14 pulses. However, the duration did not differ significantly (Kruskal-Wallis test, P> 0.05) among the notes with different number of pulses, probably because of small sample size. The mean pulse repetition rate varied from 12.30 to 14.29 but showed no correlation with the number of pulses (r=0.378, P>0.05). Each pulse had clear harmonics, and a frequency modulation was seen within a note. The dominant frequency in the initial pulse was about 1275-1500 Hz, but it increased to 1525-1675 Hz in the climax pulse, and slightly decreased to about 1500 Hz in the final pulse. In the climax pulse, the second dominant frequency was about 4500-4800 Hz. A total of seven harmonic bands could be traced between 0- 11000 Hz, and average harmonic interval was about 1570 Hz. Thus, the first dominant frequency was the fundamental and the second corresponded to the third harmonic. The calls recorded at 8.0°C (Fig. 2C, D) had basically similar traits, but a note tended to have smaller number of pulses (median=nine in 8.0°C and 13 in 4.0°C). Note durations seemed to be shorter than in 4.0°C, but the limited number of corresponding samples prohibited statistical com- parisons. The mean note durations varied from 0.075 sec in the note with two pulses to 0.700 sec in the note with 11 pulses (Table 1), but they did not differ significantly in duration (Dunn’s multiple comparison test, P>0.05), again probably due to small sample size. The pulse repetition rates varied from 15.50 to 26.67, but the mean was insigni- ficantly different from that in 4.0°C (Dunn’s multiple com- parison test, P>0.05). The dominant frequency slightly increased from 1050-1450 Hz in the initial pulse to 1450-2500 Hz in the climax pulse. In the climax pulse, three harmonic bands were apparent at about 1850, 5600, and 9300 Hz. The harmonic interval was thus judged to be about 1850 Hz, and this corresponded to the fundamental and the first dominant frequency. The second and the third dominant frequencies corresponded to the third and fifth harmonic, respectively. The dominant frequency in either the initial or the climax pulse did not differ among the notes with different number of pulses regardless of the temperature difference (Dunn’s mul- tiple comparison test, P>0.05). The relationship of the number of pulse (X) and the note duration (Y) was expressed as Y=0.052X+ 0.303 (N=21, r =0.726, P<0.01) and Y=0.065X—0.042 (N=20, r=0.963, P<0.01) in the calls recorded at 4.0°C and 8.0°C, respective- ly. Similarly, the relationship between the number of pulse (X) and the pulse repetition rate (Y) was expressed as Y= — 0.669X + 22.637 (N=20, r=—0.557, P<0.01) in the calls recorded at 8.0°C, but, as noted above, the relationship was statistically insignificant in the calls recorded at 4.0°C. In the calls recorded at either 4.0°C or 8.0°C, there were insignificant correlations between the number of pulse (X) and dominant frequencies (Y) of either the initial (r=0.076, P>0.05 and r=—0.336, P>0.05) or the climax pulse (r= 0.326, P>0.05 and r= —0.220, P>0.05). Rhacophorid Acoustic Characteristics 489 Interspecific comparisons Although some parameters, such as the dominant fre- quency, did not vary with variant temperatures, others, such as the pulse repetition rate, varied in relation to tempera- tures. Interspecific comparisons of acoustic parameters, therefore, should be made by adjusting parameters at a standard temperature. Because calls of P. chenfui and P. omeimontis were recorded at 13°C, values of parameters at this temperature were calculated from the regression lines for P. dugritei. Figure 3 shows the relationships of the pulse rate and dominant frequency of the climax pulse among the above three Chinese species and some Japanese and Taiwanese species (Table 2). As clearly seen, the three Chinese species differ from each other completely. Polype- dates dugritei had a high repetition rate and moderately high frequency. Polypedates omeimontis, on the other hand, had low repetition rates and low frequencies, which are in con- trast to P. chenfui that was characterized by low repetition rates and high frequencies. Polypedates chenfui was placed near R. schlegelii and R. viridis, while P. omeimontis lay near R. owstoni and R. arboreus. Rhacophorus prasinatus was closest to P. dugritei on this graph. is) oa (=) i=) is) i=) o i=) _ o (=) o DOMINANT FREQUENCY (HZ) a ° °o 5 10 15 20 25 PULSE RATE Fic. 3. Relationships of pulse repetition rate and dominant fre- quency of the climax pulse in the calls of Polypedates chenfui (closed diamond), P. dugritei (closed circle), P. omeimontis (open circle), Rhacophorus schlegelii (open triangle), R. arboreus (closed triangle), R. viridis (closed rectangle), R. owstoni (open diamond), and R. prasinatus (open rectangle). Values are adjusted to the ambient temperature of 13.0°C in all the species, except for P. chenfui and P. omeimontis, both of which represent the raw data. TaBLE2. Parameters for regression lines Y=aX+B, where X =air temperature (in “C) and Y=pulse repetition rate, in Polypedates dugritei and some Japanese and Taiwanese species of Rhacophorus (data from Matsui, unpublished) a B I P. dugritei 1.049 8.930 0.735 R. schlegelii 1.584 —4.554 0.944 R. viridis 1.570 —2.859 0.927 R. arboreus 1.460 —3.302 0.850 R. owstoni 0.452 4.645 0.804 R. prasinatus 1.583 0.156 0.994 DISCUSSION From a cladistic analysis of adult morphology, Jiang et al. [7] classified 14 Chinese rhacophorid species into five genera, which classification is essentially the same as that proposed by Liem [12]. In their classification, all the three species treated in the present paper were classified as Polype- dates, together with P. dennysi, P. hungfuensis, P. nigropunc- tatus, P. leucomystax (type species of the genus), and P. mutus. On the other hand, R. rhodopus and R. reinwardtii (type species of the genus) were placed in Rhacophorus. According to Jiang et al. [7], these two genera commonly possess Y-shaped terminal phalange, unopposable hand, and lack anterior process of hyale, but are distinguished from each other by heart-shaped intercalary cartilage, no skin fold, slightly or half webbed hand, and low vegetation habitat of Polypedates, in contrast to a wing-shaped intercalary cartil- age, a broad skin fold, full hand webbing, and inhabiting on tall trees of Rhacophorus. These differences, however, are not necessarily reliable, because all the Japanese and Taiwanese species currently assigned to Rhacophorus lack broad skin fold and full hand webbing [14]. Also, the habitat of an anuran species is very difficult to specify, and the above Chinese species are not sharply differentiated in this character. As to the shape of the intercalary cartilage, available information is too meager to assess whether or not it could be appropriately used for splitting the two genera. Calls of the three species here reported completely differed from the call of Polypedates leucomystax [15] or from those of several Rhacophorus species from Southeast Asia [2]. Rather, the calls of the three species fairly resembled calls of some Japanese and Taiwanese species of Rhacophor- us (see Fig. 3). Of the three Chinese species treated here, P. omeimontis and P. chenfui are syntopic on Mt. Emei-shan. The two species differ greatly in morphology and ecology; P. omeimonitis is larger and lays an egg nest on the tree or grass [13], while P. chenfui is small and breeds on land. As shown above, these species differ greatly in acoustic characteristics, P. omeimontis with much lower frequency and slightly longer note duration. Interestingly, similar morphological, ecolo- gical, and acoustic relationships are found in the Japanese thacophorids, R. arboreus and R. schlegelii from the Japan mainland [14]. The oviposition site of the larger species, R. arboreus is on the tree or on the grass, whereas the syntopic, smaller R. schlegelii lays an egg mass under the ground and never on the tree [14]. As shown in Fig. 3, calls of R. arboreus and P. omeimontis have similar temporal and fre- quency characteristics, so do calls of R. schlegelii and P. chenfui. Although another species, R. owstoni also lay close to P. omeimontis in Fig.3, their calls are actually quite dissimilar. Rhacophorus owstoni occurs in the western Ryukyus, and has characteristically long calls that are com- posed of slow and fast units [9]. The two unit calls have variant pulse rates, and only the initial, slow unit resembles the call of P. omeimonitis. 490 M. MartsulI AND G.-F. Wu The relationships of the above allopatric species pairs, 1. e., P. chenfui vs. R. schlegelii and P. omeimontis vs. R. arboreus, may be regarded as reflecting results of ecological convergence, but actual close phylogenetic relationships of these pairs are also plausible. This is inferred from the fact that there are some species pairs of Rhacophorus that are allopatric in distribution, similar in call characteristics, and deemed closely related phylogenetically. Rhacophorus mol- trechti from Taiwan and R. owstoni from the western Ryukyus, or R. schlegelii from Japan mainland and R. viridis from the eastern Ryukyus are examples of such pairs [11]. Anyhow, the available evidence indicates that at least three Chinese species of Polypedates reported here cannot be differentiated acoustically from some Rhacophorus species from Japan and Taiwan. Acoustic studies of species from wider regions will surely contribute to better understanding the relationship of Polypedates and Rhacophorus. ACKNOWLEDGMENTS T. Hikida helped laboratory work and critically read the manuscript. This study was partly supported by National Geographic Society (No. 4505-91) to MM. REFERENCES 1 Channing A (1989) A re-evaluation of the phylogeny of Old World treefrogs. S Afr Tydskr Dierk 24: 116-131 2 Dring JCM (1979) Amphibians and reptiles from northern Trengganu, Malaysia, with descriptions of two new geckos: Cnemaspis and Cyrtodactylus. Bull Br Mus Nat Hist (Zool) 34: 181-241 3 Dubois A (1986) Miscellanea taxinomica batrachologica (I). Alytes 5: 7-95 4 Frost DR (1985) Amphibian Species of the World: A Taxono- mic and Geographical Reference. Allen Press, Lawrence 5 Heyer WR (1971) Mating call of some frogs from Thailand. Fieldiana: Zool 58: 61-82 6 Inger RF (1985) Tadpoles of the forested regions of Borneo. Fieldiana: Zool (New Ser) 26: 1-89 7 Jiang S-P, Hu S-Q, Zhao E-M (1987) The approach of the 10 11 12 13 14 15 16 17 18 19 20 21 phylogenetic relationship and the supraspecific classification of 14 Chinese species of treefrogs (Rhacophoridae). Acta Herpe- tol Sinica 6: 27-42 (in Chinese, with English abstract) Kasuya E, Kumaki T, Saito T (1992) Vocal repertoire of the Japanese treefrog, Rhacophorus arboreus (Anura: Rhacophor- idae). Zool Sci 9: 469-473 Kuramoto M (1975) Mating calls of Japanese tree frogs (Rha- cophoridae). Bull Fukuoka Univ Educ III 24: 67-77 Kuramoto M (1986) Call structures of the rhacophorid frogs from Taiwan. Sci Rep Lab Amphibian Biol Hiroshima Univ 8: 45-68 Kuramoto M, Utsunomiya T (1981) Call structures in two frogs of the genus Rhacophorus from Taiwan, with special reference to the relationships of rhacophorids in Taiwan and the Ryukyu Islands. Jpn J Herpetol 9: 1-6 (in Japanese, with English abstract) Liem SS (1970) The morphology, systematics and evolution of the Old World treefrogs (Rhacophoridae and Hyperoliidae). Fieldiana: Zool 57: 1-145 Liu C-C, Hu S-C (1961) Chinese Tailless Batrachians. Kex- ue-chubanshe, Beijing (in Chinese) Maeda N, Matsui M (1989) Frogs and Toads of Japan. Bun- ichi Sogo Shuppan, Tokyo (in Japanese, with English abstract) Matsui M, Seto T, Utsunomiya T (1986) Acoustic and karyoty- pic evidence for specific separation of Polypedates megacephalus Hallowell from P. leucomystax (Boie). J Herpetol 20: 483-489 Matsui M, Wu G-F, Yong H-S (1993) Acoustic characteristics of three species of the genus Amolops (Amphibia, Anura, Ranidae). Zool Sci 10: 691-695 Mou Y, Zhao E-M (1992) A study of vocalization on thirteen species of four genera, Anura. In “Collected Papers on Herpe- tology” Ed by Y-M Jiang, Sichuan Publishing House of Science and Technology, Chengdu, pp 15-26 (in Chinese, with English abstract) Nakamura K, Ueno S-I (1963) Japanese Reptiles and Amphi- bians in Colour. Hoikusha, Osaka (in Japanese) Wells KD (1977) The courtship of frogs. In “The reproduc- tive Biology of Amphibians” Ed by DH Taylor, SI Guttman, Plenum Press, New York, pp 233-262 Zar JH (1984) Biostatistical Analysis, 2nd edition. Hall, New Jersey Zhao, E-M, Adler K (1993) Herpetology of China. Herpetol 10: 1-522 Prentice- Contr ZOOLOGICAL SCIENCE 11: 491-494 (1994) [RAPID COMMUNICATION] © 1994 Zoological Society of Japan Glutamate Substitution for Glutamine at Position 5 or 6 Reduces Somatostatin Action in the Eel Intestine Tosuuro Ursaka!, Keucut YANo*, Motoo YAMASAKI” and Masaaki ANDo!* ‘Laboratory of Physiology, Faculty of Integrated Arts and Sciences, Hiroshima University, Higashi-Hiroshima, Hiroshima 724, and Tokyo Research Laboratories, Kyowa Hakko Kogyo Co. Ltd, 3-6-6 Asahimachi, Machidashi, Tokyo 194, Japan ABSTRACT — Isolating a new somatostatin-like peptide from eel gut, the active residues to potentiate somatostatin action in the eel intestine were determined. In the newly-isolated peptide, the 5th or 6th Gln of the eel somatostatin (eSS-25II) was replaced with Glu. Although this peptide also enhanced the short-circuit current (I,.) across the intestine of the seawater eel, higher concentrations (approximately 30 fold) were required to obtain the same effects as eSS-25II. Since the I, is due to active Cl transport, this indicates that a structure formed by Gln-Gln residues at position 5 and 6 potentiates the somatostatin action to enhance ion and water transport across the intestine of the eel. INTRODUCTION Various somatostatin-related peptides have been iso- lated from diverse vertebrates. They have similar sequences containing 14 amino acids with a disulphide linkage between two Cys residues. In addition, in many vertebrates including eel, N-terminally extended somatostatins were also found [4, 5, 8, 14]. However, the physiological significance of these large somatostatins is controversial. In gastric secretion, the inhibitory effect of somatostatin was greater in short somato- statin (SS-14) than in large form (SS-28) [6, 7, 9]. In contrast, growth hormone secretion from pituitary cells was inhibited more efficiently by SS-28 rather than by SS-14 [11]. Recently, we demonstrated that a large somatostatin (eSS-25I]) isolated from the eel gut enhanced NaCl and water absorption across the intestine at lower concentrations than the short somatostatin (eSS-14II) [13], indicating that the 11 N-terminal amino acid residues potentiate somatostatin action in the eel intestine. However, it is not determined which residue(s) among the 11 amino acids contribute(s) to the potentiation. If a single amino acid substitution alters potency of a polypeptide, that residue must contribute to the activity. Fortunately, such a single-substituted somatostatin Accepted April 28, 1994 Received March 8, 1994 * To whom correspondence should be addressed. was isolated from the eel gut. A primary structure of the newly-isolated peptide was characterized at first, and the potency was compared between this peptide and the eSS- 2511. MATERIALS AND METHODS Purification Bioactive peptides were extracted with 0.1% trifluoroacetic acid (TFA) after boiling the guts (593 g) of 300 eels, Anguilla japonica, as described previously [13]. The extract was evaporated to dryness. The dried material was dissolved in 0.1% TFA (50 ml) and forced through disposable Sep-Pak Cig cartridges (Millipore). The re- tained material was eluted with 50% acetonitrile containing 10% 2-propanol and 0.1% TFA, applied to a column of Toyopearl HW-40F (2.6 cm X<100 cm; Tosoh) and eluted with 1 M acetic acid and 10% 2-propanol at a rate of 1.5ml/min. An aliquot of each fraction (18 ml) was assayed for its ability to enhance the transepithe- lial potential difference (PD) across the eel intestine. Bioactive fractions were pooled and subjected to HPLC separation (LC-6AD system, Shimadzu) with a Cg reverse-phase column (Asahipak C8P-50, Asahi Chemical Industry). The retained material was eluted with a 60-min linear gradient of 0% to 60% acetonitrile containing 10% 2-propanol and 0.1% TFA, and each fraction was bioassayed. Bioactive fractions were further applied to a Cig reverse-phase column (TSKgel ODS-80Ty, Tosoh) and eluted with a 100-min linear gradient of 0% to 20% acetonitrile containing 5% 2-propanol and 0.1% TFA. The active fractions were applied to a cation-exchange column (TSKgel CM-5PW, Tosoh) and eluted with a 35-min linear gradient of 0-0.35 M NaCl in 20 mM phosphate buffer (pH 6.7) containing 10% 2-propanol (see Fig. 1A). Bioactive peak was rechromatographed on a Cjg reverse-phase column (TSKgel ODS-80Ty) with a 50-min linear gradient of 5% to 15% acetonitrile in 5% 2-propanol and 0.1% TFA. Final purification was performed using the same column under isocratic condition (see Fig. 1B). Amino acid compositions were determined with a PICO-TAG amino acid analysis system (Millipore). Amino acid sequence analysis of the peptide was carried out by automated Edman degradation with a gas protein sequencer (PPSQ-10, Shimadzu). To determine molecular mass, fast bombardment mass spectrometry (FAB-MS) was performed with JMS-HX110A (Jeol). 492 T. UESAKA, K. YANO et al. Bioassay An aliquot of each fraction was assayed for its ability to enhance the transepithelial potential difference (PD) across the seawater eel intestine. Details are described elsewhere [2, 13]. Before bioas- say, the cultured Japanese eels, weighing about 220 g, were kept in sea water (20°C) for more than 1 week. After decapitation, the intestine was excised and the outer muscle layers were stripped off according to previous methods (1). The mucosal segment (posterior section) was opened and mounted as a flat sheet between two Ussing-type half-chambers with an exposed area of 0.5cm?. The tissue was bathed in Krebs bicarbonate Ringer’s solution consisting of (mM) 118.5 NaCl, 4.7 KCl, 3.0 CaCh, 1.2 MgSO,, 1.2 KH,PO, and 24.9 NaHCO; and containing 5mM glucose and alanine. The bathing solutions (2.5 ml each) were kept at 20°C and circulated continuously by bubbling with a 95% O, : 5% COp gas mixture (pH 7.4). The PD was recorded through a pair of calomel electrodes with a polyrecorder (EPR-151A, Toa Electronics) as the potential of the serosa with respect to the mucosa. To determine the tissue resistance (R,), rectangular pulses, 30 «A for 500 msec, were applied across the intestinal sheet every 5 min through an isolator (SS-201J, Nihon Kohden) connected to a stimulator (SEN-3301, Nihon Kohden). The I,. was obtained from the ratio of PD to R,. RESULTS AND DISCUSSION Figure 1A shows the elution profile on cation-exchange HPLC for bioactive fractions after partial purification on Cg and Cig reverse-phase HPLC. Two fractions (peaks 1 and 2 eluted at 35 and 26 min, respectively) resulted in enhance- ment of serosa-negative PD. Previous study (13) has already revealed that peak 1 contains 2 kinds of eel somatostatin (eSS-25I and eSS-25II). Since the peak 2 is 2.0 Absorbance at 215 nm Time (min) apparently far from the peak 1 on the chromatogram, a new substance distinct from eSS-25I and eSS-25II must be contained in the peak 2._ Thus we tentatively named it EI-2. The final purification profile of EI-2 is shown in Figure 1B. Amino acid composition of EI-2 is shown in Table 1. Apparently, the composition profile was similar between EI-2 and eSS-25II, indicating that EI-2 also consists of similar amino acid residues as eSS-25II. As shown in Table 2, the amino acid sequence analysis revealed that the amino acid sequence of EJ-2 was almost the same as that of eSS-25II, except for appearance of both Glu and Gln at Sth and 6th cycles. The contents of Glu and Gln were almost identical (1:1) at both Sth and 6th cycle. Although the amino acid residues at 14th and 25th cycle were not determined, these residues may be Cys, since Cys was not detected in this sequence analysis. In FAB-MS, a molecular ion peak was obtained at 2859 m/z (M+H)* in the native EI-2. After reducing the EI-2, the molecular ion peak shifted to 2861 m/z (M+H)*. The two mass increase after reduction seems to indicate an existence of disulphide bond between these two Cys residues in EI-2. Since the molecular ion peak of eSS-25II is 2858 m/z (M+H)* (13), the one mass increase in EI-2 can be explained by a substitution of Glu for Gin at position 5 or 6 in the eSS-25II. In fact, Glu was detected at position 5 and 6 in the sequence analysis (Table 2), and EI-2 was eluted earlier than the eSS-25II in cation-exchange HPLC (Fig. 1A). The latter phenomenon indicates that EI-2 is more negatively charged than eSS-25II at pH 6.7. The reason why both Glu and Gin are detected at 5th and 6th cycle is not clear yet. A most plausible explanation is that B EIl-2 0.2 ‘= _— i 2) a cs 2 0.1 2 o 2 <—c ill [ee ae eee 0 10 20 30 Time (min) Fic. 1. HPLC purification of EI-2 from the eel guts. A. Cation-exchange chromatogram of the bioactive fractions (peaks 1 and 2) after partial purification by reverse-phase HPLC. The respective fractions were eluted with a 35-min linear gradient of 0-0.35M NaCl in 10% 2-propanol and 20 mM phosphate buffer (pH 6.7). Flow rate was 0.5 ml/min. EI-2 was isolated from peak 2. From peak 1, two eel somatostatins (eSS-25I and eSS-25II) had been isolated previously [13]. B. The final purification of EI-2 by the reverse-phase HPLC. EI-2 was eluted isocratically with 12% acetonitrile containing 5% 2-propanol and 0.1% TFA (pH 2.2). Flow rate was 0.5 ml/min. A Novel Somatostatin-like Peptide TaBLeE 1. Amino acid composition of EI-2 Ser Val Asx Glx Arg Lys Ala Gly Cys Phe Tyr Trp Pro’ Thr E]-2 2.1 0.9 DS) 2.4 1.7 pM 1.0 3.1 0.9 0.9 ND 1.0 il Qi) eee Cees) (2) 2) GG) Cy) (i) i) eSS-25II* 2 il 3 3 2 3 1 3 2 1 1 1 1 1 Values indicate the amount of an amino acid relative to Ala, where Ala=1. ND, not determined. Parentheses indicate the nearest integer. Amino acid composition of eSS-25II was obtained from the primary structure characterized previously [5, 13] TABLE 2. Automated Edman degradation of EI-2 Cycle NO. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Residue Ser Val Asp Asn Gln Glu Gly Arg Glu Arg Lys Ala_ Gly Lys Asn Yield (pmol) 30.1 130.4 42.6 846 48.3 41.7 65.7 20.2 51.2 245 55.2 63.7 466 ND 39.8 42.1 Glu Gln 43.3 59.1 Cycle NO. 17 18 19 20 21 22 23 24 25 Residue Phe Tyr Trp Lys Gly Pro Thr _ Ser Yield (pmol) 45 BOs Wil AAO) 783) IL OD AA} INID) ND, not determined. 1 5 10 15 20 25 ( eSS-25I1 eSS-14 11 eSS-251 eSS-141 Fic. 2. Primary structures of somatostatin-like peptides isolated from the eel gut. The intramolecular disulphide linkage between Cys residues is represented by a line. characterized in the previous study [13]. residues that differ from eSS-25II are boxed. the fraction EJ-2 is a mixture of the following structures: Glu°eSS-25II: SVDNE QGRER KAGCK NFYWK GPTSC Glu°eSS-25II: GPTSC At the present time, we have no technique to separate these two peptides further. Even if El-2 is a mixture of Glu°eSS- 2511 and Glu°eSS-25II, the molecular weight should be the same in these peptides. Thus, isolation yield can be esti- mated as 5.6 wg (2 nmol) from 593 g guts. For comparison, SVDNQ EGRER KAGCK NFYWK Ser-Val-Asp-Asn-Gin}Glu|Gly-Arg-Glu-Arg-Lys-Ala-Gly-Cys-Lys-Asn-Phe-Tyr-Trp-Lys Gly-Pro-Thr-Ser-Cys Se eee | Ser-Val-Asp-Asn{Glu}GIn-Gly-Arg-Glu-Arg-Lys-Ala-Gly-Cys-Lys-Asn-Phe-Tyr-Trp-Lys-Gly-Pro-Thr-Ser-Cys 3 —eeeeeeee————EEE———————————— Se) Ser-Val-Asp-Asn-Gin-Gin-Gly-Arg-Glu-Arg-Lys-Ala-Gly-Cys-Lys-Asn-Phe-Tyr-Trp-Lys-Gly-Pro-Thr-Ser-Cys ee Ala-Gly-Cys-Lys-Asn-Phe-Tyr-Trp-Lys-Gly-Pro-Thr-Ser-Cys Cea ees a ee Ser-Val-Asp-Asn-Gin-Gin-Gly-Arg-Glu-Arg-Lys-Ala-Gly-Cys-Lys-Asn-Phe-Tyr-TrpfHyi|Gly-Pro-Thr-Ser-Cys (ee ee ee Ala-Gly-Cys-Lys-Asn-Phe-Tyr-Trp{HyHGly-Pro-Thr-Ser-Cys (ee ee ee eSS-25II, eSS-14II, eSS-25I and eSS-14I] are already Amino acid primary structures of all somatostatin-like peptides isolated from the eel gut are shown in Figure 2. Similarly as eSS-25II, EI-2 also enhanced the serosa- negative PD, I,. and R, (data not shown). Figure 3 shows dose-response curve for the effect of EI-2 on I,., with a threshold concentration of 10-? M and a maximal effect at 3 x10~°M. Since the I,. is due to active Cl” transport (3), the enhancement in I,. means stimulation of active Cl™ transport. Although not measured in the case of EI-2, the previous study (13) has demonstrated that eSS-25II directly 494 T. Ussaka, K. YANo et al. 50 ess-2511 @ EI-2 Oo ess-141 A 40 _ ~ = oO — so <—c< sl — 20 o an i H i Lb liabetads ont 1 é i . ‘ ‘ ‘ Wate lta Net ade hy x 4 rae BARS f F SMITHSONIAN INSTITU’ WON 1 2800 Vane tehtth Shot or epee a i ot Agta FD ’ det et HGS ye ve eels Taha Fer te tery ety Vyig ere wires rf i AR Se Talay Tey sie : uyen Taras i : . Yes seo gatas ay ; ; : rs aes , . u ean Wes : ; : 4 . : yey, ; : , : ‘ das ' H : 4 eT ye Fa : $54 Hye : ! \ ‘ Hee t yee SLA AR G4 a fryer eta ates eds ator ede ge viwret ore ren yreuly Juin git aTcEN rhea Naeele Arts : Vea wees wir f . ‘ Metro fane te , ee sit : : pia Nig ee eat eoeyete ty Ante oe oN (eg * pattie rh tu cllba tas Srihe Sp istnne Vater uy ttre aphedal prmseys. ‘ \ . .