Fiat athe et wt

rorya

¢ faint te

TEAC ia eo eerie

Wind eit

Cesena @h ene

tye? Perrin

MEP seeaty

wide Suva

wet Dre ynigen (ENE

ee

A ell ghakatnd: Siptel ahaa

whale ta i He ali thy! (ona fa He Lares aba Reve Avy it f iy Hay ih dad ani ite ; CHORAL AE? eb ta Nua

nt

Pati a

Wve au

Ay had Acer woe pay

NGA Waynes Wan tp\tice 08 A hohe

segs aMielii’s

Dib ha eat ve

Sneeeett age

Baeeaceel matriers

Diy r pie aasn ute gautenes he aaa

gt EE

ptt

Naw a Cog taste

Le, ty waft " Nghe tsi gce ree eras

ria ie yer gi ne ee

ale ihr i ae

‘ata hy on tye late

Be OC aco

POET a heat A ane Myst upon ine fos

i' eee viet ia att penis

ren)

re PLOirant

seer Auee Pamrei si che Wai tet

Tele wins fy ei sande je and eT ha Bit pein tanta Mae ae ot ane woo tie Vag at et ap Nae mate!

Jetta gt Ne) Be ny! " FRE at " Mg tatueree ey

ce

f

}

ratte a aa

Ae

i 1 Aye ay i é i

ZOOLOGICAL SCIENCE

An International Journal

VOLUME 10 1993

published by

The Zoological Society of Japan

ne

; lnrmol Isnoiiearvinl aA Of SAMUIOV 3 PRR} |

.

vitae

os) te ywieivo? lsoigologs sont

CONTENTS OF VOLUME 10 NUMBER 1, FEBRUARY 1993

REVIEWS

Jenks, B. G., H. J. Leenders, G. J. M. Mar- tens, E. W. Roubos: Adaptation physiolo- gy: The functioning of pituitary melanotrope cells during background adaptation of the amphibian Xenopus laevis

Fingerman, M., R. Nagabhushanam, R. Saro- jini: Vertebrate-type hormones in crusta- ceans: Localization, identification and func- tional significance

Lambert, C. C., D. E. Battaglia: Loss of the paternal mitochondrion during fertilization

ORIGINAL PAPERS Physiology

Yasuyama, K., B. Chen, T. Yamaguchi: Im- munocytochemical evidence for the involve- ment of RFamide-like peptide in the neural control of cricket accessory gland

Kamiya, Y., S. Harada, S. Okoyama, H. Yamamoto, R. Hosono: Developmental and pharmacological studies of acetylcho- linesterase-defective mutants of Caenorhab- AISHCICS ANS Na VOUS ERE RIEN co wissen ein hoc - Cell Biology

Takagi, Y., K. Nimura, Y. Tokusumi, H. Fujisawa, K. Kaji, H. Tanabe: Secretion of mitogenic factor(s) from stocks of Para- mecium tetraurelia, P. caudatum and P. multi- micronucleatum

Kelly, K. L., E. L. Cooper, D. A. Raftos: Cytokine-like activities of a humoral opsonin from the solitary urochordate Styela clava

Biochemistry Sasaki, T., M. Sasaki, J. Enamti: Mouse y- casein cDNA: PCR cloning and sequence analysis

Mita, M., M. Nakamura: Phosphatidylcho- line is an endogenous substrate for energy metabolism in spermatozoa of sea urchins of the order echinoidea

Developmental Biology Itow, T.: Grafting of center cells of horseshoe crab embryos into host embryos at different developmental stages Suzuki, A., S. Nishimatsu, K. Murakami, N. Ueno: Differential expression of Xenopus BMPs in early embryos and tissues (RAPID COMMUNICATION) Koya, Y., K. Takano, H. Takahashi: Ultra- structural observations on sperm penetration in the egg of elkhorn sculpin, Alcichthys alcicornis, showing internal gametic associa- tion Ishijima, S., Y. Hamaguchi, T. Iwamatsu: Sperm behavior in the micropyle of the medaka egg (RAPID COMMUNICATION)

Yoshizato, K., A. Nishikawa, Y. Izutsu, M. Kaiho: Epidermal cells of the tail of an anuran larva are competent to transform into the adult-type cells (RAPID COMMUNI- CATION)

Reproductive Biology Fukumoto, M., T. Numakunai: The acro- some and its differentiation during sper- miogenesis in Halocynthia roretzi (Asci- diacea, Tunicata) Kobayashi, T., Y. Asakawa, H. Iwasawa: Kinetics of first spermatogenesis in young toads of Xenopus laevis (RAPID COM- MUNICATION) Hamlin, G. P., S. D. Kholkute, W. R. Duke- low: Toxicology of maternally ingested car- bon tetrachloride (ccl4) on embryonal and fetal development and in vitro fertilization in mice

i

Nakamura, M., F. Tsuchiya, M. Iwahashi, Y. Nagahama: Reproductive characteristics of precociously mature triploid male masu salm- on, Oncorhynchus masou

Morphology Miyazaki, K., T. Makioka: A case of inter- sexuality in the sea spider, Cilunculus arma- tus (Pycnogonida; ammotheidae) Zizavy, J.,O. Nedved, R. Socha: Metameric color pattern in the bugs (Heteroptera: Lygaeidae, Largidae, Pyrrhocoridae, Rho- palidae): A morphological marker of insect

body compartmentalization .............. 133 Systematics and Taxonomy Suzuki, T., T. Takagi, S. Ohta: N-terminal

amino acid sequences of 440 kDa hemoglo- bins of the deep-sea tube worms, Lamellib-

rachia sp. 1, Lamellibrachia sp. 2 and slender vestimentifera gen. sp. 1. Evolutionary rela- tionship with annelid hemoglobins Sawada, I., K. Koyasu, K. C. Shrestha: Two new species of the genus Staphylocystis (Ces- toda: Hymenolepididae) from the house shrew, Suncus murinus, in Nepal Tsurusaki, N., D. Song: Two new species of Sabacon from Sichuan province, China (Arachnida: Opiliones: Sabaconidae) ....155 Mashiko, K., K. Numachi: Genetic evidence for the presence of distinct fresh-water prawn (Macrobrachium nipponense) popula- tions in a single river system Miura, T., J. Hashimoto: Mytilidiphila, a new genus of nautiliniellid polychaetes living in the mantle cavity of deep-sea mytilid bivalves collected from the Okinawa trough

NUMBER 2, APRIL 1993

REVIEWS Asano, M., K. Shiokawa: Behavior of ex- ogenously-introduced DNAs in_ early embryos of Xenopus laevis Robertson, D. R.: Comparative aspects of intestinal calcium transport in fish and amphibians

ORIGINAL PAPERS Physiology Schlechter, N. L., L. S. Katz, S. M. Russell, C. S. Nicoll: Physiological evaluation of the role of the liver as a mediator of the growth- promoting action of somatotrophin Hara, A., C. V. Sullivan, W. W. Dickhoff: Isolation and some characterization of vitel- logenin and its related egg yolk proteins from Coho Salmon (Oncorhynchus kisutch) ...245 Harada, A., M. Yoshida, H. Minakata, K. Nomoto, Y. Muneoka, M. Kobayashi: Structure and function of the molluscan myoactive tetradecapeptides Noguchi, M., K. Okamoto, S. Nakamura: Effects of proteolytic digestion on the con- trol mechanism of ciliary orientation in cili- ated sheets from Paramecium

Seidou, M., K. Ohtsu, Z. Yamashita, K. Nari- ta, Y. Kito: The nucleotide content of the octopus photoreceptor cells: No changes in the octopus retina immediately following an intense light flash

Kobayashi, W.: Effect of osmolality on the motility of sperm from the lamprey, Lampet- ra japonica

Biochemistry Nakae, H., T. Obinata: Immunocytochemical localization of troponin I and C in the mus- cles of Caenorhabditis elegans (RAPID

COMMUNICATION) _.................-. 375 Shimizu, M., H. Kudo, H. Ueda, A. Hara, K. Shimazaki, K. Yamauchi: Identification

and immunological properties of an olfactory system-specific protein in Kokanee Salmon (Oncorhynchus nerka)

Developmental Biology Komatsu, M., T. Shosaku: Development of the brittle star, Ophioplocus japonicus H. L. Clark. I

Reproductive Biology Amano, T., Y. Okita, T. Yasumoto, M. Hoshi: Maitotoxin induces acrosome reaction and histone degradation of starfish Asterina pecti- niferd SPETM —.. eee reece 307 Harada, T.: Reproduction by overwintering adults of water strider, Aquarius paludum (Fabricius) Asahina, K., K. Aida, T. Higashi: Biosynthe- sis of 17a,20a-dihydroxy-4-pregnen-3-one from 17a-hydroxyprogesterone by goldfish (Carassius auratus) spermatozoa (RAPID COMMUNICATION) .....-----+:+5+5+: 381

Endocrinology Gobbetti, A., M. Zerani, M. M. Di Fiore., V. Botte: Prostaglandins and sex steroids from reptilian (Podarcis sicula sicula) ovarian follicles at different developmental stages

Zairin, M. Jr., K. Asahina, K. Furukawa, K. Aida: Plasma steroid hormone profiles in

ll

HCG-injected male walking catfish Clarias batrachus Kao, Y. H., P. S. Alexander, V. V. C. Yang, J. Y. L. Yu: Annual patterns of testicular development and activity in the chinese bull- frog (Rana rugulosa Wiegmann) Singtripop, T., T. Mori, M. K. Park, K. Shira- ishi, T. Harigaya, S. Kawashima: Prolactin binding sites in normal uterus and the uterus with adenomyosis in mice

Environmental Biology and Ecology Ogino, K., Y. Hirono, T. Matsumoto, H. Ishikawa: Juvenile hormone analogue, S- 31183, causes a high level induction of pre- soldier differentiation in the Japanese damp- wood termite

Systematics and Taxonomy Komai,T., K. Amaoka: A new species of the genus Pasiphaea (Crustacea, Decapoda, Pasiphaeidae) from the North Pacific ....367

NUMBER 3, JUNE 1993

REVIEWS

Ooka, H.: Proliferation of anterior pituitary cells in relation to aging and longevity in the iI. SoscebboboooeudenceborEncnn sotoomnacdse 385

Ricci, N., R. Banchetti: The peculiar case of the giants of Oxytricha bifaria (Ciliata, Hypotrichida): A paradigmatic example of cell differentiation and adaptive strategy

ORIGINAL PAPERS Physiology

Ohnishi, K.: The defective color vision in juvenile goldfish does not depend on used training task as the measure of discrimina- tion : A two- choise response measure ...411

Mikami, K., S. Sato, M. Otokawa, A. Palum- bo, K. Kuwasawa: Cardiotoxic effects of adrenochrome and epinephrine on the mouse cultured myocardium

Suzuki, T., K. Narita, K. Yoshihara, K. Nagai, Y. Kito: Immunochemical detection of GTP-binding protein in cephalopod photore-

ceptors by anti-peptide antibodies

Cell Biology Kurita, T., H. Namiki: Serum-induced cell death

Developmental Biology Matsushita, S.: Effect of the oesophageal mesenchyme on the differentiation of the digestive-tract endoderm of the chick embryo Sivasubramanian, P.: Neural control of mus- cle differentiation in the leg of the fleshfly Sarcophaga bullata (RAPID COMMUNI- CATION)

Reproductive Biology Koike, S., T. Noumura: Immunohisto- chemical expression of inhibin-a subunit in the developing rat gonads Furuya, H. K. Tsuneki, Y. Koshida: development of the hermaphroditic gonad in four species of dicyemid mesozoans_ ...... 455

Endocrinology Tanaka, S.: Hormonal deficiency causing albinism in Locusta migratoria Duan, C., N. Hanzawa, Y. Takeuchi, E. Hamada, S. Miyachi, T. Hirano: Use of primary cultures of salmon hepatocytes for the study of hormonal regulation of insulin- like growth factor I expression in vitro. ...473 Sawada, K., T. Noumura: Metabolism of androgens by the mouse suomandibular gland and effects of their metabolites ....481 Jana, N. R., S. Bhattacharya: Binding of thyroid hormone to freshwater perch leydig cell nuclei-rich preparation and its functional relevance Shiraishi, K., T. Singtripop, M. K. Park, S. Kawashima: Developmental changes of testicular gonadotropin receptors and serum gonadotropin levels in two strains of mice

Bhavior Biology Wassersug, R., A. Izumi-Kurotani: The be- havioral reactions of a snake and a turtle to

abrupt decreases in gravity Ninomiya, K., I. Nohara, T. Toyoda, T. Kimura: The pattern of volatile com- pounds in incubated and fresh preputial fluid of male mice (RAPID COMMUNICA- TION)

Enviromental Biology Ip, Y. K., C. Y. Lee, S. F. Chew, W. P. Low, K. W. Peng: Differences in the responses of two mudskippers to terrestrial exposure

Numata, H., A. H. Saulich, T. A. Volkovich: Photoperiodic responses of the linden bug, Pyrrhocoris apterus, under conditions of con- stant temperature and under thermoperiodic conditions

Systematics and Taxonomy Katayama, T., M. Yamamoto, H. Wada, N. Satoh: Phylogenetic position of acoel tur- bellarians inferred from partial 188 rDNA sequences

NUMBER 4, AUGUST 1993

REVIEWS

Keino, H., S. Kashiwamata: Purkinje- granule-cell interaction in the cerebellar de- velopment of hyperbilirubinemic mutant rats

Inoué, S.: Sleep-promoting substance (SPS) and physiological sleep regulation

ORIGINAL PAPERS Physiology

Burton, D., A. Snow: An in vivo and in vitro comparison of differential melanophore re- sponsiveness in flatfish integumentary color patterns

Yamada, H., R. Horiuchi, K. Gen, K. Yamauchi: Involvement of four hormones in thyroxine deiodination in several tissues of immature yearling masu salmon, Oncorhyn- chus masou

Hirota, K., Y. Sonoda, Y. Baba, T. Yama- guchi: Distinction in morphology and be-

havioral role between dorsal and ventral groups of cricket giant interneurons (RAPID COMMUNICATION) Ikeda, M., K. Tomioka: Temperature de- pendency of the circadian locomotor rhythm in the cricket Gryllus bimaculatus Ghosh, P., S. Bhattacharya, S. Bhattacharya: Acetylcholine stimulates the inhibited acetyl- cholinesterase activity in the brain of a snakehead, Channa punctatus

Cell Biology Shinagawa, Y., H. Abe, K. Saiga, T. Obinata: Increased expression of cofilin in denervated chicken skeletal muscle Yashika, K., K. Matutani, Y. Sekoguti, M. Sakata, P. H. Hashimoto: Changes in ultrastructure of the anal papilla epithelial cell of Aedes togoi Theobald, a sea-water mosquito, in adaptation to different salnities

Developmental Biology Nakamura, M., S. Okinaga, T. Kobayashi, K. Aoki: Sodium fluoride (NaF) releases the two-cell block in mouse embryos Arai, T., M. Wakahara: Hemoglobin transi- tion from larval to adult types in normally metamorphosing, metamorphosed and meta- morphosis-arrested Hynobius retardatus

Nakagawa, S., K. Sumita, T. Hama, T. Mayumi: Production of chimeric mice by exchanging nuclei from blastomeres of the two-cell embryo

Nakagawa, S., R. Adachi, M. Miyake, T. Hama, K. Tanaka, T. Mayumi: Production of normal @~ macular mouse chimeras: The presence of critical tissue in the macular mutant mouse, a model of Menkes’ kinky hair disease

Yokota, Y., K. H. Kato, M. Mita: Morpho- logical and biochemical studies on yolk de- gradation in the sea urchin, Hemicentrotus pulcherrimus

Reproductive Biology Koike, S., T. Noumura: Immunohisto- chemical localizations of TGF-f in the de-

veloping rat gonads Nagasawa, H., Y. Goto: Effects of tumor necrosis factor on pregnancy-dependent mammary tumors in GR/A mice Ueda, H., H. Kudo, M. Shimizu, K. Mochida, S. Adachi, K. Yamauchi: Immunological similarity between an olfactory system- specific protein and a testicular germ cell protein in kokanee salmon (Oncorhynchus nerka)

Endrocrinology Nakajima, K., T. Yanagisawa, S. Tanaka, S. Kikuyama: Absence of methylated thyro- tropin-releasing hormone in the bullfrog (Rana catesbeiana) brain (RAPID COM- MUNICATION)

Systematics and Taxonomy

Matsui, M., G. Wu, H. Yong: Acoustitic characteristics of three species of the genus Amolops (Amphibia, Anura, Ranidae) ..691

Suzuki, H., Y. Sato, S. Fujiyama, N. Ohba: Genetic differentiation between ecological two types of the Japanese firefly, Hotaria parvula: An electrophoretic analysis of allozymes

NUMBER: 5, OCTOBER 1993

REVIEWS Hourdry, J.: Passage to the terrestrial life in amphibians: Events accompanying this eco- logical transition Bagnara, J. T., P. J. Fernandez: Hormonal influences on the development of amphibian pigmentation patterns

ORIGINAL PAPERS Physiology

Ohnishi,K., T. Yamaguchi: Light-, sound-, and ultrasound-induced cercal movements in flying crickets

Oishi, T., K. Ohashi: Effects of wavelengths of light on the photoperiodic gonadal re- sponse of blinded-pinealectomized Japanese quail

Fujii, R., Y. Tanaka, H. Hayashi: Endothe-

lin-1 causes aggregation of pigment in tele- ostean melanophores Fujisawa, Y., M. Shimoda, K. Kiguchi, T. Ichikawa, N. Fujita: The inhibitory effect of a neuropeptide, ManducaFLRFamide, on the midgut activity of the sphingid moth, Agrius convolvuli

Genetics Dinesh, K. R., T. M. Lim, K. L. Chua, W. K. Chan, V. P. E. Phang: RAPD analysis: An efficient method of DNA fingerprinting in

fishes (RAPID COMMUNICATION)... 849 Immunology Tochinai, S.: Strictly thymus-dependent

tolerance induction in immunologically com- petent Xenopus larvae (RAPID COM-

vi MUNICATION) Biochemistry

Sasaki, T., H. Ishikawa: Nitrogen recycling in the endosymbiotic system of the pea

aphid, Acyrthosiphon pisum ............. 779 Sasaki, T., N. Fukuchi, H. Ishikawa: Amino

acid flow through aphid and its symbiont:

Studies with °N-labeled glutamine ....... 787

Developmental Biology Xianyu, Y., N. Haga: Initiation of the ear- liest nuclear event in fertilization of Para- mecium by the microinjection of calcium buffer (RAPID COMMUNICATION)

Fujita, Y., S. Igarashi, H.Fujisawa, K. Yama- su, IT. Suyemitsu, K. Ishihara: Effects of exogastrula-inducing peptides on cell prolif- eration in embryos of the sea urchin Anthoci- daris crassispina

Endocrinology Okimoto, D. K., G. M. Weber, E. G. Grau: The effects of thyroxine and propylthiouracil treatment on changes in body form associ- ated with a possible developmental thyroxine surge during post-hatching development of

the tilapia, Oreochromis mossambicus ...803 Kudo, A., M. K. Park, S. Kawashima: Isola- tion of rat GnRH receptor cDNA having different 5’-noncoding sequence (RAPID COMMUNICATION)

Morphology

Matsuzaki, O., A. Iwama, T. Hatanaka: Fine structure of the vomeronasal organ in the house musk shrew (Suncus murinus)

Systematics and Taxonomy Uchida, K., Y. Ohtani, Y. Sasayama, H. Nam- bu, H. Yoshizawa, S. Akabane, K. Suzuki, N. Suzuki: Levels of calcium in the skin of some amphibians and possible evolutionary implications Kobayashi, M., M. Takahashi, H. Wada, N. Satoh: Molecular phylogeny inferred from sequences of small subunit ribosomal DNA, supports the monophyly of the metazoa

Aoki, J.,S. Hu: Oribatid mites from tropical forests of Yunnan Province in China. II. Families Galumnidae and Galumnellidae

NUMBER 6, DECEMBER 1993

REVIEWS

Kiihn, E. R., K. A. Mol, V. M. Darras: Control strategies of thyroid hormone monodeiodination in vertebrates

Hourdry, J.: Passage to the terrestrial life in amphibians. II. Endocrine determinism

ORIGINAL PAPERS Physiology Iga,T., TIT. Mio: Leucophores of the dark- banded rockfish Sebastes inermis. I. Adrenergic mechanisms that control the

movements of pigment ................... 903 Muramoto, A.: Axotomy-induced long-

lasting firing in an identified crayfish

MOLONEULON Mhtoth ii. seroma Roath. 913

Okada, J., K. Kuwasawa: Valve dilator neurons of the lateral arteries of the isopod crustacean Bathynomus doederleini (RAPID COMMUNICATION) ................. 1057

Cell and Molecular Biology Yonezawa, Y., R. Hirai, T. Noumura, H. Kondo: A novel growth factor highly mitogenic on immortalized cells from rat serum: Purification and its properties ....925

Genetics Shimizu, T.,S. Masaki: Genetic variability of the wing-form response to photoperiod in a subtropical population of the ground cricket, Dianemobius fascipes

Developmental Biology Iwamatsu, T., R. A. Fluck, T. Mori: Mecha- nical dechorionation of fertilized eggs for experimental embryology in the medaka

Reproductive Biology Hihara, F., Y.Hihara: Electron dense parti- cles appeared in the microvilli zone of the cuboidal cells of the ventral receptacle in Drosophila melanogaster mated female

Nagasawa, H., M. Hasegawa, K. Yamamoto, S. Sakamoto, T. Mori, A. Nagumo, H. Tojo: Normal and neoplastic growth of mammary glands and circulating levels of prolactin and growth hormone in mouse whey acidic pro- tein promoter/human growth hormone (mWAP/hGH) transgenic mice .......... 963

Endocrinology Watanabe, Y. G., M. Shirai: In vitro evi- dence for a neural factor(s) involved in the proliferation of adenohypophysial primor- dial cells in fetal rats ..................... 971

Morphology Ishii, T., Y. Saito, Y. Taneda: Histological differences between the color patterns of two strains of the compound ascidian Polyandro- carpa misakiensis — ............e.eeeeeeeees 977

Environmental Biology and Ecology Kobayashi, S., H. Numata: Photoperiodic re- sponses controlling the induction of adult diapause and the determination of seasonal form in the bean bug, Riptortus clavatus ...................5. 983

Vil

Hashimoto, J., T. Miura, K. Fujikura, J. Ossa- ka: Discovery of vestimentiferan tube- worms in the euphotic zone (RAPID COM- MUNICATION) ieee eee aan SA 1063

Systematics and Taxonomy Kim, B. K., T. Aotsuka, O. Kitagawa: Evo- lutionary genetics of the Drosophila mon- tium subgroup. II. Mitochondrial DNA VanlatiOny reine yet 5 esi phe eee 991 Tsurusaki, N., S. Nakano, H. Katakura: Karyotypic differentiation in the phytopha- gous ladybird beetles Epilachna vigintiocto- maculata complex and its possible relevance to the reproductive isolation, with a note on supernumerary Y chromosomes found in E. PUSCULOSA © Seasons cmisemcial sa cue) sigiess eeu 997 Hasumi, M., H. Iwasawa: Geographic varia- tion in the pes of the salamander Hynobius lichenatus: A comparison with tetradactyl Hynobius hidamontanus and _pentadactyl Hynobius nigrescens ............-++++-- 1017 Ikezawa, H., S. F. Mawatari: A systematic study of three species of Celleporina (Bryozoa, Cheilostomata) from Hokkaido, Japan with special reference to their early ASLO PCMH Aetna! Mic Se hes ieee et 1029 Nishikawa, T., X. Turon: Taxonomy and geographic variation of Pyura microcosmus (Savigny) from Lusitanian waters and the

Red Sea (Urochordata, Ascidiacea) .... 1045 Authormindex: mcinereeessceee cones. ccs 1069 Acknowledgment .................-2-000- 1074 [BN PICHIA) bt el ee Reno bccn too cat oqo 1076 Contents of ZOOLOGICAL SCIENCE, Vol.

DINOS NEG 25'S Re Oe aceon ne 1

an

Development Published Bimonthly by the Japanese Society of

Developmental Biologists

i é 5 Distributed by Business Center for Academic Growth & Differentiation Societies Japan, Academic Press, Inc.

Papers in Vol. 35, No. 6. (December 1993)

64. REVIEW: Y. Maeda: Pattern Formation in a Cell-Cycle Dependent Manner during the Development of Dictyostelium discoideum

65. R. Amikura, S. Kobayashi, K. Endo and M. Okada: Nonradioactive In Situ Hybridization Methods for Drosophila Embryos Detecting Signals by Immunogold-silver or Immunoperox- idase Method for Electron Microscopy

66. T. Iwamatsu, Y. Toya, N. Sakai, Y. Terada, R. Nagata and Y. Nagahama: Effect of 5-Hydroxytryptamine on Steroidgenesis and Oocyte Maturation in Pre-ovulatory Follicles of the Medaka Oryzias latipes

67. M. Yamashita, J. Jiang, H. Onozato, T. Nakanishi and Y. Nagahama: A Tripolar Spindle Formed at Meiosis I Assures the Retention of the Original Ploidy in the Gynogenetic Triploid Crucian Carp, Ginbuna Carassius auratus langsdorfii

68. T. Akiyama and M. Okada: A Mutation in the Drosophila Profilin Homolog Gene Affects Gametogenesis and Bristle Development in Both Sexes

69. H. Kajiura, M. Yamashita, Y. Katsu and Y. Nagahama: Isolation and Characterization of Goldfish cdc2, a Catalytic Component of Maturation-Promoting Factor

70. S. H. Cheng and T. W. Mak: Molecular Characterization of Three Murine Hox11-Related Homeobox Genes, T/x-1, -2, and -3, and Restricted Expression of T/x-1 during Embryogenesis

71. <A. Jayadeep, S. Sailesh, P. Reddanna and V. P. Menon: Prostaglandin Metabolism during Cell Aggregation in the Regenerating Vertebrate Appendage

72. I. Yazaki: A Novel Substance Localizing on the Apical Surface of the Ectodermal and the Esophageal Epithelia of Sea Urchin Embryos

73. P. Hardman, B. J. Klement and B. S. Spooner: Growth and Morphogenesis of Embryonic Mouse Organs on Non-Coated and Extracellular Matrix-Coated Biopore Membrane

74. T.-C. J. Wu, L. Wang, M. H. Jih and Y.-J. Y. Wan: Activin Receptor Type II Gene Expression is Induced During Embryonic Development and Differentiation of Murine F9 Embryonal Carcinoma Cells

75. K. Yamada, T. Yamamoto, K. Akasaka and H. Shimada: cis-Elements and Protein Factors Related to Regulation of Transcription of Arylsulfatase Gene during Sea Urchin Development

76. Y. Yasuda, M. Nagao, M. Okano, S. Masuda, R. Sasaki, H. Konishi and T. Tanimura: Localization of Erythropoietin and Erythropoietin-Receptor in Postimplantation Mouse

Embryos

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.

BRANCHES NOW OPEN IN USA AND EUROPE

Narishige. The complete range for micromanipulation

This large range

includes micro- manipulators, microelectrode pullers and micro- forges, micro- injectors, microgrinders Shown here is a small and stereotaxic selection of instruments from instruments. the extensive Narishige range.

or over 30

years Narishige

have been developing their

MM-31 Miniature Micromanipulator

SN-3 Stereotaxic Instrument

extensive range of precision ; ; instruments for Physiology, US-I 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.

MP-I Micromanipulator

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 ewritccw micromanipulators. more information.

REPAIRS, AFTER SALES SERVICE AND TECHNICAL SUPPORT Narishige Scientific Instrument Laboratory | NARISHIGE | 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 404 Glen Cove Avenue, Sea Cliff, New York 11579 U.S.A. Tel: +1 (516) 759 6167 Fax: +1 (516) 759 6138

SH PE-2 Microelectrode Puller

ergonomic design.

YES

von OLYMPUS

is the answer.

xcellent optics.

Y-shaped ergonomic design ensures maximum comfort. Easy operation. Easy recognition. And fatigue-free use. To the question of what constitutes the ideal microscope from the user’s point of view, our answer is

a unique Y-shaped design. We invite you to discover unprecedented operational ease and comfort.

Excellent optics provide superior image quality.

Developed through the application of more than 70 years of experience and what the latest technology has to offer, Olympus’ own UIS (Universal Infinity System) infinity-corrected optical system assures exceptional image quality. You benefit from both higher contrast and outstanding flatness throughout the field of view.

System versatility answers diverse research requirements. Frame rigidity and stability have been increased dramatically through use of computer simulation (FEM analysis). You'll find that reliability has been enhanced, too, when the time comes for system expansion, via devices for image processing, microscopic light metering, photography and so on.

OLYMPUS OPTICAL CO.,LTD. OLYMPUS (JAPAN) CO.,LTD.

The BX Series was developed with your needs and our ideals in mind.

For free catalog, please contact: OLYMPUS (JAPAN) CO.,LTD. Ryumeikan Bldg., 3-4, Kanda-surugadai, Chiyoda-ku, Tokyo Tel. (03) 3251-8971

SYSTEM MICROSCOPE

BX40/BX50 |

NEW

“ZOOLOGICAL. SC] ENCE

An International Jou

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,

Eplideshi Kobayashi (Tokyo) Tokyo 113, Japan. Phone (03) 3814-5675 Managing Editor: Officers:

Tsuneo Yamaguchi (Okayama) President: Hideo Mohri (Chiba)

Assistant Editors: Yoshihisa Kamishima (Okayama) Masaki Sakai (Okayama) Akiyoshi Niida (Okayama)

Secretary: Hideo Namiki (Tokyo) Treasurer: Makoto Okuno (Tokyo) Librarian: Masatsune Takeda (Tokyo)

Editorial Board:

Howard A. Bern (Berkeley) Walter Bock (New York) Aubrey Gorbman (Seattle) Horst Grunz (Essen) Robert B. Hill (Kingston) Yukio Hiramoto (Chiba) Susumu Ishii (Tokyo) Yukiaki Kuroda (Tokyo) John M. Lawrence (Tampa) Koscak Maruyama (Chiba) Roger Milkman (Iowa) Kazuo Moriwaki (Mishima) Richard S. Nishioka (Berkeley) Chitaru Oguro (Toyama) Tokindo S. Okada (Okazaki) Andreas Oksche (Giessen) Hidemi Sato (Nagano) Mayumi Yamada (Sapporo)

Ryuzo Yanagimachi (Honolulu) Hiroshi Watanabe (Tokyo)

ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and communications in the broad field of Zoology. The journal appears bimonthly. An annual volume consists of six numbers of more than 1200 pages including an issue containing abstracts of papers presented at the annual meeting of the Zoological Society of Japan.

MANUSCRIPTS OFFERED FOR CONSIDERATION AND CORRESPONDENCE CONCERN- ING EDITORIAL MATTERS should be sent to: Dr. 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 Honkogagome, 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 1992 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 10: 1-11 (1993)

@,1993 Zoological Society of Japan

Adaptation Physiology: the Functioning of Pituitary Melanotrope Cells during Background Adaptation of the Amphibian Xenopus laevis

Bruce G. JENKS, HANS J. LEENDERS, GERARD J. M. MARTENS

and Eric W. Rousos

Department of Animal Physiology, University of Nijmegen, Toernooiveld, 6525 ED Nijmegen, The Netherlands

INTRODUCTION

The vertebrate hypothalamo-hypophyseal sys- tem is responsible for converting neuronal in- formation into an endocrine output, the release of hormones into the blood. This process is known as neuroendocrine integration. A major source of neural input to the neuroendocrine system comes from sensory information concerning conditions in the environment (e.g. light, temperature, presence of stressors). Neuroendocrine transducer cells of the hypothalamo-hypophyseal system are res- ponsible for integrating the diverse neuronal in- formation to produce and release endocrine sig- nals. These signals in turn coorrdinate the func- tioning of peripheral organs (e.g. reproductive organs, adrenal glands, thyroid gland) thus allow- ing the animal to adapt to changes in the environ- ment.

An example of such a neuroendocrine transduc- er cell is the melanotrope cell of the pituitary pars intermedia. In amphibians this cell is responsible for converting neuronal information concerning color of background into an endocrine output. The cell produces the multifunctional precursor protein proopiomelanocortin (POMC), which is processed to give a number of peptides including melanophore-stimulating hormone (a-MSH). This peptide hormone stimulates the dispersion of the

Received August 24, 1992

black pigment melanin in dermal melanophores thus causing skin darkening. The hormone is released in animals on a black background. Among amphibians, the South Afircan clawed toad Xenopus laevis shows a particularly strong adaptive response to background color. The skin pigment goes very rapidly from the fully aggre- gated to the fully dispersed state in animals trans- ferred from a white to a black background, and rapidly aggregates upon transfer from black to white background (Fig. 1). The adaptation of Xenopus to a black background involves activation of biosynthetic and secretory processes in the melanotrope cells, whereas these processes inacti- vate during white-background adaptation [17, 22, 32, 65]. Therefore, by simply changing back- ground color, it is possible to manipulate the activity of Xenopus melanotrope cells. This ease of manipulation is an attractive feature of these cells for studies aimed at elucidating cellular and molecular mechanisms of cell activation and in- activation. Indeed, Xenopus melanotrope cells have been the object of study in a number of laboratories over many years, starting with classic studies of Hogben and Slome (reviewed in [64]). This model system and related topics were exten- sively reviewed several years ago [23, 31, 39, 60]. The intention of this survey is to examine recent developments in the analysis of Xenopus in- termediate lobe melanotrope cells, particularly in relation to the physiological functioning of these

2 B. G. Jenxs, H. J. LEENDERS et al.

cells.

Dynamics of Background Adaptation: the Demon- stration of Short-term and Long-term Adaptive Mechanisms

It has generally been thought that a-MSH is the only factor involved in stimulating pigment disper- sion in dermal melanophores of Xenopus during black background adaptation. An analysis of events during white to black background transfer, however, has shown that there is a major discre- pancy between the degree of pigment dispersion in dermal melanophores and plasma a-MSH levels [67]. The pigment becomes fully dispersed within a few hours of transfer to black background and yet it takes several days for plasma a-MSH to reach its maximum level (see Fig.1). These results suggested that a factor other than a-MSH must be

pigment dispersion (m.i.)

plasma MSH

POMC biosynthesis

black

Fic. 1.

responsible for the rapid pigment dispersion seen during the early stage of black background adapta- tion. The discovery that the rapid pigment disper- sion can be blocked by the f-adrenergic receptor antagonist propranolol indicates that a /- adrenergic receptor is involved [67]. Jn vitro experiments show that this receptor is working at the level of the dermal melanophore cells in caus- ing pigment dispersion. Denervated skin explants lack the rapid short-term response, which indicates that the short-term mechanism depends on in- nervation of the skin. Probably, this innervation is a component of the autonomic nervous system releasing (nor)adrenalin in the vicinity of the mela- nophores [24]. In long-term-adapted animals (several days or longer on black background) the release of a-MSH from the pars intermedia is critical for maintaining pigment dispersion. This is shown by the fact that neurotransmitters such as

aan K@sss 24 hours

|

A summary of the dynamics of various aspects of background adaptation by the amphibian Xenopus laevis.

Changes in the degree of pigment dispersion in dermal melanophores, in plasma MSH levels and in the level of POMC biosynthesis in pars intermedia melanotrope cells are shown. The discrepancy between the rapid pigment dispersion and slow increase in plasma a-MSH (indicated by white arrows) led to studies that established the involvement of a #-adrenergic mechanism causing pigment dispersion during the early stages of black background adaptation [67]. The relatively slow decrease in POMC biosynthesis (black arrow) in black to white background adaptations is, in part, probably due to the fact that an enzymatic system for the degradation of RNA must first be

synthesized [4]. See text for further discussion.

Functioning of Pituitary Melanotrope Cells 3

dopamine and GABA, which are known to inhibit in vitro release of a-MSH, cause skin whitening when injected into fully black-background- adapted animals [61, 63]. Treatment of long-term black-adapted animals with propranolol has no effect on pigment dispersion, which is in sharp contrast to the effect of this #-blocker in short- term black-adapted animals [67]. The importance of a-MSH in long-term regulation of the mela- nophores is further illustrated by an analysis of plasma a-MSH levels in animals adapted for sever- al weeks to white, light grey, dark grey and black backgrounds [52, 66]. Grey animals, which have only partially dispersed pigment in dermal mela- nophores, were found to have plasma e-MSH levels that are intermediate between the low levels found in white-adapted animals and the high levels of black-adapted animals.

Physiological Significance of Short-term and Long- term Mechanisms

It is generally accepted that in most amphibians pigment cells are regulated primarily by the endo- crine system (a-MSH) while in fishes the nervous system is a more important regulatory system. This seems to fit the locomotory behavior of these animals as amphibians tend to be more stationary than fish and thus may not need rapid color adaptation. The presence of §-adrenergic recep- tors on amphibian melanophores, and thus the potential for nervous control of these cells, was established in early studies [14, 15, 30] but the physiological significance of these receptors re- mained obscure. The idea has been forwarded that the f-adrenergic receptor might be activated as a result of acute handling stress, thus causing transi- tory darkening, termed “excitement darkening” [6]. It is apparent that, in the species X. laevis at least, this receptor is also involved in the regula- tion of pigment migration during adaptation to a dark background.

In considering the significance of the /- adrenergic regulatory mechanism in Xenopus it is important to realize that the capacity of melano- trope cells to synthesize a-MSH is acquired only very slowly, taking a number of days to reach maximum level during adaptation to black back- ground [22, 32, 34, 44]. Hormone stores possessed

by animals on white background are, however, depleted within 24 hr during adaptation to black background [22, 62]. For Xenopus, the regulation of dermal melanophores must be viewed as a cooperative effort; the #-adrenergic mechanism plays an important role in the short-term, during which time the melanotrope cells of the neuroen- docrine system acquire the capacity to produce enough a-MSH to maintain pigment dispersion in the long-term. This cooperation ensures that Xenopus is capable of both rapid and sustained pigment dispersion.

Melanotrope Cell Morphology: the Demonstration of Cell Recruitment during Background Adapta- tions

There have been several studies on the mor- phology and ultrastructure of melanotrope cells of white- and black-background-adapted Xenopus [17, 21, 65]. These studies have shown that cells of white animals are small and inactive when com- pared to those of black animals. A question that remained was how melanotrope cells meet an intermediate (submaximal) demand for a-MSH, such as in animals on grey background. Two possible physiological mechanisms are: (1) all melanotropes are activated to an intermediate level of secretion to meet the submaximal demand for hormone, or (2) a subpopulation of melano- trope cells becomes fully acitvated to meet this demand, whereas the other cells remain relatively inactive. In the first case, the melanotrope cells act as a homogeneous population whereas in the latter situation they are heterogeneous. The results of an analysis at the light and electron microscopic level showed that the pars intermedia of grey-adapted animals is composed of a mixture of inactive and fully active melanotrope cells [52]. It was there- fore concluded that melanotrope cells respond as a heterogeneous cell population to an increased de- mand for e-MSH, with progressively more cells being recruited to the active state as the physiolo- gical demand for a-MSH increases.

Physiological Significance of Recruitment

Two general strategies may be used by endo- crine cells to increase the rate of hormone release.

4 B. G. Jenks, H. J. LEENDERS et al.

One is a simultaneous activation of the entire population of endocrine cells, the activation of oxytocin-producing neuroendocrine cells being a good example [49]. Here, the strength of stimula- tion (suckling) determines the degree of activation achieved in the homogeneous cell population. It appears that the heterogeneous response, exem- plified by the intermediate lobe melanotrope cells of Xenopus, is the more general phenomenon among endocrine cells. Morphometric and bio- chemical evidence has been given that this mechanism holds for hormone secretion by gona- dotropes [47, 48], lactotropes [35], pancreatic /- cells [54] and follicular thyroid cells [11]. There may be an energetic advantage of this mechanism because an increased demand for hormone is met through activation of relatively few endocrine cells.

It appears from studies on mammalian lacto- trope and somatotrope cells that the mechanism regulating cell recruitment involves differences in the sensitivity for regulatory factors among the

individual endocrine cells. This has been discoy- ered through an analysis of the effects of secreta- gogues on hormone release from individual cells, measured using the reverse-plaque assay [35, 47, 48] and, more recently, the sequential cell im- munoblot assay [1, 26]. Using a modification of this latter assay it has been shown that Xenopus melanotropes display different sensitivities to dopamine, one of the a-MSH release-inhibiting factors [50].

Ultrastructure of Intermediate Lobe Nerve Termi- nals: Coexistence of Dopamine, GABA and Neuro- peptide Y

The intermediate lobe melanotrope cells of X. laevis are regulated by multiple factors, both stimulatory and inhibitory (see Fig. 2). It had been assumed that each regulatory factor would be present in separate neuronal networks [23]. Analysis at the ultrastructural level, however, has revealed that GABA and NPY are present within

pars nervosa

pars intermedia

folliculo- stellate cell

TRH

melanotrope cell

Fic. 2.

A summary of the classical neurotransmitters and neuropeptides involved in the regulation of e-MSH

secretion from the melanotrope cell of Xenopus laevis. Abbreviations: TRH, thyrotropin releasing hormone; CRH, corticotropin releasing hormone; NPY, neuropeptide Y; DA, dopamine; GABA, y-aminobutyric acid; D5, dopamine D, receptor; G,, GABA, receptor; G,, GABA, receptor. Stimulatory (+) or inhibitory (—) action on the secretory process is indicated. NPY, DA and GABA coexist in the same nerve terminals [51, 59] with NPY and DA being present within the same secretory vesicle [59].

Functioning of Pituitary Melanotrope Cells 5

the same nerve terminals within the intermediate lobe [51]; a recent study showed that these same varicosities also contain dopamine [59]. GABA is confined to small electron-lucent vesicles whereas dopamine and NPY are found in electron-dense vesicles. The varicosities have been shown to make close contact with both the melanotrope cells and the other major cell type of the pars interme- dia, the so-called folliculo-stellate cells.

Physiological Significance of Coexistence

At present it is difficult to attribute a functional significance to the fact that the three e-MSH release inhibitory factors occur in the same nerve terminals. In another case of coexistence, that of acetylcholine and vasoactive intestinal peptide (VIP) in autonomic nerve terminals of the cat salivary gland, it has been shown that the transmit- ter substances involved can be differentially re- leased [36]. Low frequency stimulation leads to release of acetylcholine alone while with high frequency stimulation both the neurotransmitter and the neuropeptide are released. Moreover, it has been shown that VIP potentiates the stimula- tory action of acetylcholine on the salivary gland [36]. Concerning the Xenopus pars intermedia, the possibility of differential release of the coexist- ing inhibitory factors has yet to be examined. An analysis of the effect of combinations of these factors on secretion of a-MSH showed only addi- tive effects [27], thus ruling out interactions lead- ing to potentiation. Nonetheless, there is some evidence for differential action of these regulatory factors; inhibitions induced by GABA and dopa- mine are rapidly reversed when treatment with the factor is terminated whereas recovery from NPY- induced inhibition is achieved only very slowly [28]. Altogether these results suggest that dopa- mine and GABA are appropriate short-term in- hibitors while NPY might be used for long-term inhibition of the melanotrope cell.

The Folliculo-Stellate Cell has an Inhibitory Action on the Release of ac-MSH from the Melanotrope Cell

Folliculo-stellate cells are glial-like cells struc- turally related to the astrocytes of the central nervous system. They are evenly distributed

throughout the pars intermedia and constitute about 5% of the cell content of the lobe. These cells possess long extensions that make intimate contacts with virtually every melanotrope cell [53, 56]. The inhibitory action of folliculo-stellate cells on the secretory process of melanotropes was demonstrated in studies with 7-day cultured neurointermediate lobes of Xenopus [28]. The cultured tissue is devoid of neuronal influences due to the degradation of nerve terminals, as demon- strated ultrastructurally and confirmed by im- munocytochemical analysis showing a complete lack of dopamine, GABA and NPY. Giving such tissue a depolarizing pulse of K* in superfusion gives rise to a biphasic response with respect to the secretion of a-MSH. The short duration stimula- tory phase could be attributed to a direct K*- evoked depolarization of the melanotrope cell. The subsequent inhibitory phase, however, can only be attributed to K*-induced activation of an inhibitory mechanism emanating from another cell type within the cultured lobes. Ultrastructural analyses show that in such tissue the folliculo- stellate cell remains viable and it was therefore concluded that this cell type must be the source of the K*-induced inhibition. The mechanism of action of the folliculo-stellate cell on the melano- trope cells has still to be established. The stellate cell could release an a-MSH release inhibiting factor. Alternatively, it could inhibit the secretory activity of the melanotrope cells by inducing changes in the extracellular ionic environment of the melanotropes. Precedent for the first option can be found in the literature. Folliculo-stellate cells of the mammalian pars distalis have been reported to produce and secrete paracrine factors that regulate secretory activity of endocrine cells in this organ [5, 12, 13]. In support of the second option is the observation that bovine intermediate lobe folliculo-stellate cells are involved in ion transport [10]. Further, a high level of staining indicative of Na*/K*-ATPase activity has been observed between melanotropes and _ folliculo- stellate cells of the frog pars intermedia [56]. It was suggested that this activity might reflect stel- late cell regulation of the extracellular ionic en- vironment of the melanotropes.

6 B. G. Jenks, H. J. LEENDERs et al.

Physiological Significance of the Inhibition by Fol- liculo-Stellate Cells

In view of the structural relationship between folliculo-stellate cells and the melanotropes, activation of an inhibitory mechanism via the folliculo-stellate cell would be expected to affect a number of melanotrope cells simultaneously. Possibly, this indirect mechanism of inhibition is used when a general inhibition of the pars interme- dia is required, as opposed to the recruitment mode of regulation where cells are individually activated or inactivated. The question arises as to how the inhibitory mechanism emanating from the folliculo-stellate cell is itself activated and inacti- vated. While we initially thought that NPY might be of importance in this respect [28], subsequent analyses have shown that NPY can act directly on the melanotrope cell to inhibit a-MSH release. Further research will be required to fully under- stand the role of the folliculo-stellate cell in the regulation of the secretory process.

Analysis of Secretion: the Demonstration of Fast and Slow Secretory Pathways

The biosynthetic and secretory activity of mela- notrope cells of Xenopus can be set to a very high level simply by placing the animal for several days on black background. This fact allows the use of in vitro pulse-chase labelling methods to follow the biosynthetic and secretory process of Xenopus melanotropes. A number of studies have been devoted to an analysis of the release of radiolabelled peptides [27, 33, 43, 57]. pulse-chase labelling experiment with Xenopus neurointermediate lobes, conducted in combina- tion with tissue superfusion methods to analyze the dynamics of the secretory process, it was shown that there are two phases in the release of radiolabelled products [43]. The first phase com- prises a rapid increase (reaching a maximum 3 hr after the pulse) followed by a rapid decrease; the second phase is a low basal secretion that persists for many hours of superfusion. In a subsequent analysis, using dual-labelling protocols to follow simultaneously the release associated with both phases, it was concluded that each phase reflects

In a

the functioning of a distinct pathway of secretion [68]. The first, termed the fast pathway, concerns peptides that are released within 6hr of their biosynthesis. The second pathway, designated the slow pathway, pertains to peptides remaining in the melanotrope cells for up to 2 days before being released. It was shown that in the presence of dopamine the secretory peptides of the fast path- way are shunted to the slow pathway and are subsequently released from this latter pathway when the dopamine treatment is terminated.

At present, it is not possible to say whether the two secretory pathways of Xenopus melanotrope cells occur in one cell type or, alternatively, reflect the functioning of two different cell types within the tissue. There is morphological evidence for both possibilities. In support of the one cell-type explanation is the observation that there are two types of secretory granules in Xenopus melano- tropes, electron-dense granules and somewhat larger, fibrous, electron-lucent granules [17]. In the biosynthetically active melanotrope cells of black- adapted animals the dense granules dominate but there also are, within these cells, lucent granules; in the inactive cells of white-adapted animals the granules are almost exclusively of the lucent type. Morphological indication that the pathways might reflect the activity of two cell types is provided by the observation that the melanotrope cell popula- tion of the pars intermedia of black-adapted ani- mals is heterogeneous; the tissue possesses both active and inactive malonatrope cells [52].

Physiological Significance of Two Secretory Path- ways

The existence of multiple secretory pathways in endocrine cells seems to be a general phe- nomenon, although the physiological significance of these pathways is unknown. Multiple pathways have been reported for cells producing thyrotropin [25], prolactin [7, 45], somatotropin [7, 8, 58] and parathyroid hormone [16]. An important criterion for attributing physiological significance to such pathways is to establish that they are independent- ly regulated. This would appear to be the case for the pathways in prolactin cells [9] and parathyroid hormone producing cells [16, 46]. In both cases the pathways concern newly synthesized hormone

Functioning of Pituitary Melanotrope Cells GT

versus hormone sequestered in mature secretory compartments. For the prolactin cell the mature pathway is cyclic- AMP-dependent while release of newly synthesized hormone is independent of the cyclic nucleotide. A similar situation might exist for Xenopus melanotrope cells; the slow pathway appears to be more sensitive to stimulation by 8-bromo-cyclic-AMP than the fast pathway [68]. Interestingly, in mouse melanotrope cells 8- bromo-cyclic-AMP gives clear differential effects, the cyclic nucleotide analogue being more effective in stimulating mature peptides than newly synthe- sized peptides [29].

In considering the physiological significance of independently regulated secretory pathways it is worthwhile to focus on cells that produce multi- functional precurgor proteins such as POMC. The existence of independently regulated pathways in POMC-producing cells might endow the cells with the ability to manipulate the peptide composition, and thus the physiological effect(s) of their secre- tory signal. In this way the cell itself could be multifunctional, responding with different sets of peptides to different physiological demands. Stud- ies with the melanotrope cell of X. /aevis may help to establish the full potential of POMC-producing cells to participate in multiple regulatory pro- cesses.

Analysis of POMC Gene Expression: Regulation is at the Level of both Gene Transcription and mRNA Stability

POMC gene expression in the pars intermedia is 20 to 30 times higher in animals adapted to a black background than in animals adapted to a white background. Using a steady-state kinetic model for mRNA degradation, the half-life of POMC mRNA in Xenopus melanotrope cells was deter- mined during the process of background adapta- tion [4]. During induction of the POMC gene (i.e. following white to black background transfer) the half-life of POMC mRNA was found to be 3- to 4-fold longer than during de-induction of the gene (transfer from a black to a white background). This difference in the stability of POMC mRNA is not sufficient to account for the 20- to 30-fold differences in the steady-state levels of POMC

mRNA in fully black- versus fully white-adapated animals. Therefore, it seems that background adaptation involves not only changes in the stabil- ity of POMC mRNA but also changes in the transcriptional activity of the POMC gene.

Physiological Significance of Dual Control Sites in the Regulation of POMC Gene Expression

The stabilization of POMC mRNA in animals on a black background would seem physiologically relevant. This mechanism will help maintain a high level of POMC mRNA under this environ- mental condition, and thus would add to the efficiency of the POMC expression system. Like- wise, the decreased stability of POMC mRNA in animals transferred from a black to a white back- ground seems a physiologically appropriate res- ponse to the new environmental situation. The shutting down of the POMC expression system, viewed at the level of POMC biosynthesis, is a slow process (see Fig. 1). Part of the reason for this may be that in the animals adapting to a white background RNA degrading enzymes must first be synthesized. This is indicated by the fact that treatment of animals with mRNA synthesis inhibi- tors block the degradation of POMC mRNA observed after black to white background transfers

[4].

Analysis using Differential Hybridization: a Battery of Genes is Coexpressed with the POMC Gene

Proper functioning of a peptide-secreting cell requires not only production of precursor proteins for secretory peptides but also expression of genes associated with the secretory function. Such genes include those encoding for precursor-processing enzymes and for structural and regulatory proteins involved in the translocation, sorting, packaging and release of secretory products. To identify genes coexpressed with the POMC gene in Xeno- pus melanotrope cells advantage has been taken of the fact that the transcriptional process is very active in cells of black-adapted animals but inac- tive in white-adapted animals. The method used, termed differential hybridization, involved screen- ing of a Xenopus pituitary cDNA library using single-stranded cDNA probes synthesized from

8 B. G. Jenks, H. J.

LEENDERS et al.

remove neurolntermedilate lobe

Isolate Isolate mRNA mRNA

triplicate filters of neurointermediate lobe cDNA flbrary from black animais

Fic. 3. An overview of the differential screening approach being used to find genes important to the transducer

function of the melanotrope cell of Xenopus laevis. The

approach is based on the premiss that non-POMC genes

that are more highly expressed in black- than in white-adapted animals would be important in supporting the

regulated secretory pathway leading to the release of P

RNA of biosynthetically active and inactive mela- notropes, i.e. from black and white animals, res- pectively (Fig. 3). Screening approximately 20,000 colonies of the library resulted in the identification of 130 differentially hybridized clones [41]. Of these, 92% were found to be related to POMC mRNA and discarded from further analysis. The 10 remaining clones presumably code for proteins essential for the secretory activity of the Xenopus melanotrope cell. Thus far only one of these clones has been fully characterized. It has proved to be highly similar (97% identical) to a partially sequenced porcine and human pituitary protein named 7B2 [18, 55]. The Xenopus 7B2 cDNA clone was used to isolated a human pituitary cDNA clone encoding the entire structure of hu- man 7B2 [40]. A comparison of the amino acid sequences of Xenopus and human 7B2 shows that the 7B2 protein, and in particular the N-terminal region, is remarkably well conserved during 350 million years of vertebrate evolution. The overall

OMC-derived peptides.

degree of amino acid sequence identity between Xenopus and human 7B2 is 83%, with most differ- ences being conservative amino acid substitutions. This degree of conservation is much higher than that between the POMC proteins of these two species (55%) [42].

While POMC and 7B2 are clearly coexpressed, there are some differences in the nature of the expression of these proteins in Xenopus melano- trope cells. First, there are large differences in the steady-state levels of mRNA for the two proteins; in black-adapted animals the amount of mRNA coding for POMC in the pars intermedia is about 25-fold higher than that coding for 7B2 [41]. Moreover, the differences in steady-state levels of 7B2 and POMC mRNA between black- and white- adapted animals are about 8-fold and 25-fold, respectively [2]. From an analysis of the time course in the expression of the genes in animals transferred from a white to a black background it is evident that there are also important differences in

Functioning of Pituitary Melanotrope Cells 9

the dynamics of the expression of the two genes. The level of POMC mRNA was significantly in- creased within 16 hours of transfer whereas it took over 40 hours for significant increases in the level of 7B2 mRNA to be detected [2]. It will be interesting to investigate the molecular mecha- nism(s) underlying the differences in activity be- tween the 7B2 and POMC genes. It could be that the POMC promotor contains elements that are responsible for the very high level of POMC gene expression in the intermediate lobe. Alternatively, it is possible that POMC mRNA contains se- quences responsible for a long half-life in this tissue while 7B2 mRNA is less stable because it lacks such sequences.

Biosynthetic studies have revealed that 7B2 is a precursor protein in Xenopus [3]. During pulse incubations there was synthesis of a 25 kDa form of 7B2 which was converted, during subsequent chase incubations, to an 18 kDa protein. Chemical cleavage and peptide mapping have shown that processing takes place in the C-terminal region [3], a region that possesses three pairs of basic amino acid residues [40]. Such pairs often function as cleavage sites in the enzymatic processing of pre- cursor proteins. Xenopus 7B2 is apparently se- questered in the regulated secretory pathway of the melanotrope cell and is possibly stored within the same granules as the POMC-derived peptides. This is shown by the observation that the 18 kDa form of the protein is released [3] and that this release can be inhibited or stimulated by factors established to inhibit or stimulate, respectively, the secretion of POMC-derived peptides from Xenopus melanotrope cells [2]. Therefore, 7B2 is not only coexpressed with POMC but it is also, at the level of secretion, coordinately regulated with POMC-derived peptides. The concept of coordi- nate regulation can be extended to the level of gene expression. Long-term treatment (3 days) of neurointermediate lobes with apomorphine (D2 dopamine receptor agonist), GABA or NPY leads to decreases not only in the level of POMC mRNA but also to concomitant decreases in the level of mRNA coding for the 7B2 protein [2].

Physiological Significance of the Coexpressed 7B2 Protein

The function of the 7B2 protein is unknown. In mammalian tissues it has been shown to be present in cells containing secretory granules, such as neurons and endocrine cells [20, 38]. Ultrastructu- ral studies show this protein to be stored within secretory granules [37, 38]. Therefore, it is con- ceivable that 7B2 has an important function in the secretory pathway of these cells, such as being a component of the exocytotic machinery or a sor- tase directing secretory proteins to the regulated pathway of secretion. The suggestion has been made that the 7B2 protein is a member of the granin (chromogranin-secretogranin) family [19]. On the other hand, the fact that 7B2-derived products are released makes an extracellular func- tion, i.e. as an endocrine or paracrine factor, also a distinct possibility.

Concluding Remarks

The analysis of the pituitary melanotrope of Xenopus shows that this cell, in fulfilling its neuroendocrine integrative function, is extremely complex. The cell acts as a neuroendocrine inter- face under the control of multiple factors, some of which coexist within the same nerve terminals. The regulatory factors clearly have differential effects on the melanotrope cells, some of these factors stimulating the melanotrope cell secretory process (viz. the neuropeptides CRH and TRH) while others (dopamine, GABA, NPY) inhibit secretion (see Fig. 2). Even among the inhibitory factors there appears to be differential action. Dopamine and GABA give rise to rapid and reversible inhibition of secretion, and might thus be appropriate inhibitors for short-term adapta- tions. Physiological background adaptations have been shown to involve the activation or inactiva- tion of individual melanotrope cells, depending on the demand for e-MSH. Because both dopamine and GABA are delivered to the melanotrope cells via synaptic communication, they could very well be involved in regulating the (de)recruitment of individual cells. NPY, in contrast, may have a more general effect on the pars intermedia. This

10 B. G. Jenxs, H. J. LEENDERS et al.

neuropeptide acts more slowly than dopamine and GABA in inducing inhibition and, once estab- lished, this inhibition is more sustained than that induced by the two classical neurotransmitters.

There are several challenging areas for future research. One area of interest concerns the phy- siological significance of the multiple secretory pathways of the amphibian melanotrope cell. A first step in examining this will be to determine the peptide content of the pathways. This will give insight into the potential biological activities associated with these secretory signals. Also of immediate interest will be studies on the effect of the various regulatory factors on the pathways, to determine if they are independently regulated. On the molecular front, further characterization of genes co-expressed with the POMC gene in the Xenopus melanotrope should ultimately reveal a host of gene products that are essential to the proper physiological functioning of this neuro- endocrine transducer cell.

REFERENCES

1 Arita J, Kojima Y, Kimura F (1991) Endocrinology 128: 1887-1894

2 Ayoubi TAY, Jenks BG, Roubos EW, Martens GJM (1992) Endocrinology 130: 3560-3566

3 Ayoubi, TAY (1991) Regulation of proopiomelano- cortin and 7B2 gene expression in the pituitary of Xenopus laevis. Thesis, University of Nijmegen, Nijmegen, The Netherlands, 89 pp

4 Ayoubi TAY, van Duijnhoven HLP, van de Ven WJM, Jenks BG, Roubos EW, Martens GJM (1990) J Biol Chem 265: 15644-15647

5 Baes M, Allaerts W, Denef C (1987) Endocrinology 120: 685-691

6 Burgers ACJ, Boschman TAC, van de Kamer JC (1953) Acta Endocrinol 14: 72-82

7 ChenTT, Kineman RD, Betts JG, Hill JB, Frawley LS (1989) Endocrinology 125: 1904-1909

8 Chun CC, Hoeffler JP, Frawley LS (1988) Life Sciences 42: 701-706

9 Dannies PS (1982) Biochem Pharmacol 31: 2845— 2849

10 Ferrara N, Gospodarowicz D (1988) Biochem Bio- phys Res Commun 157: 1376-1382

11 Gerber H, Peter HJ, Bachmeier C, Kaempf J, Studer H (1987) Endocrinology 120: 91-96

12 Gospodarowicz D, Abraham JA, Schilling J (1989) Proc Nat Acad Sci (USA) 86: 7311-7315

13

14 15

16

17 18

19

20

21

22

23

24

Gospodarowicz D, Lau K (1989) Biochem Biophys Res Commun 165: 292-298

Graham JDP (1961) J Physiol (London) 158: 5P—6P Hadley ME, Goldman JM (1970) Am J Physiol 219: 72-77

Hanley DA, Wellings PG (1985) Can J Physiol Pharamacol 63: 1139-1141

Hopkins CR (1970) Tissue Cell 2: 59-70

Hsi KL, Seidah NG, De Serres G, Chretién M (1982) FEBS Lett 147: 261-266

Huttner WB, Gerdes HH, Rosa P (1991) Trends Biochem Sci 16: 27-30

Iguchi H, Natori S, Nawata H, Kato K, Ibayashi H, Chan JSD Seidah ND, Chretién M (1987) J Neurochem 49: 1810-1814

Jenks BG, Martens GJM, van Helden HPM, van Overbeeke AP (1985) Biosynthesis and release of melanotropins and related peptides by the pars intermedia in Xenopus laevis. In “Current Trends in Comparative Endocrinology”. Ed by B Lofts, WN Holmes, Hong Kong University Press, Hong Kong, pp 149-152

Jenks BG, van Overbeeke AP, McStay BF (1977) Can J Zool 55: 922-927

Jenks BG, Verburg-van Kemenade BML, Martens GJM (1988) Proopiomelanocortin in the amphibian pars intermedia: a neuroendocrine model system. In “The Melanotropic Peptides: Volume 1” Ed by ME Hadley, CRC Press, Boca Raton, Florida, pp 103- 126

Jenks BG, van Zoest ID (1990) Melanotrope cell function in the neuroendocrine regulation of dermal melanophores. In “Biology and Physiology of Amphibians” Ed by W Hanke, Gustav Fischer Ver- lag, Stuttgart, pp 219-228

Keith LD, Tam B, Ikeba H, Opsahl Z, Greer MA (1986) Neuroendocrinology 43: 445-452

Kendall ME, Hymer WC (1987) Endocrinology 121: 2260-2262

Kongsamut S, Shibuya I, Douglas WW (1991) Neuroendocrinology 54: 599-606

Koning de HP, Jenks BG, Scheenen WJJM, Rijk de EPCT, Caris RTJM, Roubos EW (1991) Neuroen- docrinology 54: 68-76

Leenders HJ, Jenks BG, Rélo AL, Roubos EW (1990) J Neuroendocrinol 2: 563-566

Lerner AB, Shizume K, Buding I (1954) J Clin Endocrinol Metab 14: 1463-1490

Loh YP, Myers B, Wong B, Parish DC, Lang M, Goldman ME (1985) J Biol Chem 260: 8956-8963

Loh YP, Gainer H (1977) J Gen Physiol 70: 37-58 Loh YP, Elkabes S, Myers B (1988) Regulation of proopiomelanocortin biosynthesis in the amphibian and mouse pituitary intermediate lobe. In “The Melanotropic Peptides: Volume 1” Ed by ME Had-

34

35

36

37

38

39

40

41

42

43

44 45

46

47

48

49

50

51

52

Functioning of Pituitary Melanotrope Cells 11

ley, CRC Press, Boca Raton, Florida, pp 85-102 Loh YP, Jenks BG (1981) Endocrinology 109; 54— 61

Lucque EH, Munoz de Toro M, Smith PF, Neill JD (1986) Endocrinology 118: 2120-2124

Lundberg JM, Hokfelt T (1983) Trends Neurosci 6: 325-333

Marcinkiewicz M, Benjannet S, Seidah NG, Cantin M, Chretién M (1987) Cell Tiss Res 250: 205-214 Marcinkiewicz M, Benjannet S, Cantin M, Seidah NG, Chretién M (1986) Brain Res 380: 349-356 Martens GJM, Jenks BG, van Overbeeke AP (1981) Comp Biochem Physiol 69C: 75-82

Martens GJM, Weterings KAP, van Zoest ID, Jenks BG (1987) Biochem Biophys Res Comm 143: 678-684

Martens GJM, Bussemakers MJG, Ayoubi TAY, Jenks BG (1989) Eur J Biochem 181: 75-79 Martens GJM, Civelli O, Herbert E (1985) J Biol Chem 260: 13685-13689

Martens GJM (1988) The pro-opiomelanocortin gene in Xenopus laevis: structure, expression and evolutionary aspects. In “The Melanotropic Pep- tides: Volume 1” Ed by ME Hadley, CRC Press, Boca Raton, Florida, pp 67-84

Martens GJM (1988) FEBS Lett 234: 160-164 Mena F, Clapp C, Aguayo D, Martinez-Escalera G (1989) Neuroendocrinology 49: 207-214

Morrissey TJ, Cohen DV (1979) J Cell Biol 83: 521- 528

Neill JD, Smith PF, Lucque EH, Munoz de Toro M, Nagy G, Mulchahey JJ (1986) In “Neuroendocrine Molecular Biology” Ed by G Fink, AJ Harmar, KW Kerns, Plenum Press, New York, pp 325-340

Neill JD, Frawley LS (1983) Endocrinology 112: 1135-1137

Poulain DA, Wakerley B (1982) Neuroscience 7: 773-808

Rik EPC T de, Terlou M, Cruijsen PMJM, Jenks BG, Roubos EW (1992) Cytometry 13: 863-871 Rijk EPCT de, Jenks BG, Vaudry H, Roubos EW (1991) Neuroscience 38: 495-502

Rijk EPCT de, Jenks BG, Wendelaar Bonga SE

53

54

55

56 Si 58

59

60

61

62

63

64

65

66

67

68

(1990) Gen Comp Endocrinol 79: 74—82

Rijk EPCT de, Cruijsen PMJM, Jenks BG, Roubos EW (1991) Endocrinology 128: 735-740

Schuit FC, In ’t Veld PA, Pipeleers DG (1988) Proc Nat Acad Sci (USA) 85: 3865-3869

Seidah NG, Hsi KL, De Serres G, Rochemont J, Hamelin J, Antakly T, Cantin M, Chretién M (1983) Arch Biochem Biophys 255: 525-534

Semoff S, Hadley ME (1978) Gen Comp Endocri- nol 35: 329-341

Shibuya I, Kongsamut S, Douglas WW (1991) Proc R Soc Lond 243: 129-137

Stachura ME (1986) Mol Cell Endocrinol 44: 37-45 Strien FJC van, Rijk EPCT de, Heymen PSH, Hafmans TGM, Roubos EW (1991) Histochemistry 96: 505-510

Tonon MC, Danger JM, Lamacz M, Leroux P, Adjeroud S, Anderson A, Verburg-van Kemenade BML, Jenks BG, Pelletier G, Stoeckel L, Burlet A, Kupryszewski G, Vaudry H (1988) Multihormonal control of melanotropin secretion in cold-blooded vertebrates. In “The Melanotropic Peptides: Volume 1” Ed by ME Hadley, CRC Press, Boca Raton, Florida, pp 127-170

Verburg-van Kemenade BML, Jenks BG, Smits RJM (1987) Neuroendocrinology 46: 289-296 Verburg-van Kemenade BML, Jenks BG, Lenssen FJA, Vaudry H (1987) Endocrinology 120: 62-68 Verburg-van Kemenade BML, Tonon MC, Jenks BG, Vaudry H (1986) Neuroendocrinology 44: 446- 456

Waring H (1963) Color Change Mechanisms of Cold-Blooded Vertebrates. Academic Press, New York

Weatherhead B, Whur P (1972) J Endocrinol 53: 303-310

Wilson JF, Morgan MA (1979) Gen Comp Endocri- nol 38: 172-182

Zoest ID van, Heijman PS, Cruijsen PMJM, Jenks BG (1989) Gen Comp Endocrinol 76: 19-28 Zoest ID van, Leenders HJ, Jenks BG, Roubos EW (1990) Comp Biochem Physiol 96C: 199-203

x

te GS tee enka cpm ahd MAES 120 es

Nie PRAY F # bile id Tied Tat a

(PEO) oth

A LL eed ar ote TORR aaa He » oe v5 al Bik RPM

me

et of. ney j “ih

A tit TOMES ES, Myc hn lee 2 whee: Ce ii A ear ate oped mann oe . th ; oF Ras ti

EF Jorf ipa I avieeet ce: TWEE OO PREY Greet iP Thiet) See y- yortesa is ‘ee

RO ee ane oe AB D>etg ab Worl eke ieee iv 1S 74h

fi mabe i rae f , A dey utes, Mi esest , Je

i Yi fo ptehunie

f i Sethe 7 . eel x ue it 7 Kee ; “ ’ +4 i ? - va = u J 4 ath A re = is + cy rt zy grail “ } Ate , - MEM etre

oa

avian) oye: een

he Aertel HP LGOR),

Fi Sabana

Neath: tid t ARIE -OR IES

ie arent UAE. a Sil dekh A TREO Sine ob Lptgimy pare patient

(chia! 4 TA cbs

Hoa be ORG ct nba aiatdilacsenr tins: HT ae nf ABH Dedel CUE pei Nak SRA OD Rae | Ayelet } aN Ot Watts Pte ‘ ee VAT verde AVL! ie it te tb ng

Fou (R01 all

wed oT wa “ a sito Saghlsey heer pte) ' a sae ean ke, , Sauer Ss ats iri Gea ear : Het > bite iothates EAS gt: ti [

f aa * Brains: Cpa bad wensrel Aa | a ceases Sarthe. = Ma SUE ta. bi : teh ES teers SPR mys

bre a iy aff int? Fit ee a alae even Py A a) ste q f SAA Y. pe a Tet i abn t t207) ore (od

Ng "4 Rut Heat +> »

te has 3 Joe tay ‘ np: feat #10 4

Lai as Ny ety ev a ae Then ur bs Reims

J r ue? ve i:

i? nat AF i

el Ti oe ip ‘ahi “Aine ¥ ; Kile Soka - hg Rod iB es

t nite 4 “dang q : Fe |

“B. Wore Ba i S vil ert ieee J et [Or alae trad neg Ta it .

ZOOLOGICAL SCIENCE 10: 13-29 (1993)

© 1993 Zoological Society of Japan

REVIEW

Vertebrate-type Hormones in Crustaceans: Localization, Identification and Functional Significance

MILTON FINGERMAN, RACHAKONDA NAGABHUSHANAM

and RACHAKONDA SAROJINI

Department of Ecology, Evolution, and Organismal Biology, Tulane University, New Orleans, Louisiana 70118, U.S.A.

INTRODUCTION

In a 1983 review of the literature relating to invertebrate neuropeptides Greenberg and Price [49] referred to invertebrates as having two sets of neuropeptides. One set they called “native neuropeptides.” These are the ones that were originally isolated from invertebrates and have been shown to have specific physiological or biochemical roles in these animals. In contrast, the compounds that Greenberg and Price called “naturalized neuropeptides” are those originally discovered in vertebrates and later identified in invertebrates by various techniques, including chromatography, radioimmunoassay, and immuno- cytochemistry. At the time Greenberg and Price wrote their review very little information had been obtained about what the physiological and biochemical roles of these “naturalized” com- pounds might be, and for many of them their roles remain obscure. Peptides are now known to be the largest class of neuroregulatory compounds in both invertebrates and vertebrates [92]. The object of this reivew is to present the more recent evidence for the presence in crustaceans of not only verte- brate-type neuropeptides but of all vertebrate-type peptidergic, proteinaceous, steroid, and lipid- derived hormones. However, the classical neuro- transmitters, such as norepinephrine and dopa-

Received September 14, 1992

mine, are not treated herein because (a) they are so ubiquitious in the animal kingdom that conclud- ing whether one or more of them is vertebrate-type or invertebrate-type does not seem justified and (b) their presence and roles in crustaceans have been extensively reviewed in a recent publication from this laboratory [40]. Where information is available concerning possible physiological roles of these vertebrate-type compounds in crustaceans, that information will be presented also.

In this review we shall use “hormone” as it has been defined by Norman and Litwack [88]. They stated that “any substance that operates at the cellular level, generated either externally or inter- nally, which conveys to that cell a message to stop, start, or modulate a cellular process will come under the purivew of modern endocrinology.” Also, Norman and Litwack divided hormones into three classes, based on the distance of action: (a) endocrine hormones, which are the classical hor- mones such as lutenizing hormone that travel to distant target cells; (b) paracrine hormones, such as the classical aminergic neurotransmitters (e.g. acetylcholine), which travel only a short distance to neighboring cellular targets; and (c) autocrine hormones, such as prostaglandins, that are synthe- sized and released by the same cell on which they act.

Invertebrate nervous systems, including those of crustaceans, are useful models for analyses of

neuronal functioning, including peptidergic

14 M. FINGERMAN, R. NAGABHUSHANAM AND R. SAROJINI

neurotransmission. Technical improvements in such techniques as high performance liquid chro- matography, peptide sequencing and radioimmuno- assay have greatly enhanced our ability to identify the vertebrate-type hormones in invertebrate tis- sues, even when these compounds are present in extremely low concentrations.

ENDORPHINS

Localization and identification

The presence of substances in the mammalian brain that have opiate-like activity was clearly established in 1975 when Hughes et al. [57] pub- lished the structures of two such endogenous opioids, the pentapeptides methionine enkephalin and leucine enkephalin extracted from the porcine brain. The name “endorphin” is often used, as we have done here, for the entire class of compounds endogenous in mammals that have morphine-like activity [43]. Since 1975, endorphins have been found in several non-nervous organs of vetebrates such as the gut, adrenal medulla, and pancreas [43]. The first demonstration of the presence of an opioid in an invertebrate came from a study with a locust, Locusta migratoria. Radioimmunoassays (RIA) showed its nervous system contains both methionine enkephalin-like and leucine enkepha- lin-like substances [50]. With respect to crusta- ceans, the first opioid study was done in 1981 by Mancillas et al. [75]. Through immunocytochemis- try they found leucine-enkephalin-like immuno- reactivity in all the retinular cells of the spiny In addition, such immunoreactivity was also apparent in nerve fibers in chiasm 3 that pass from the medulla terminalis to terminate in the medulla interna. No immuno- reactivity against antibodies of another opioid, beta-endorphin, was found. Later, other crustaceans were similarly studied. Jaros et al. [58], using the shore crab, Carcinus maenas, obtained immunocytochemical evidence for a leucine enkephalin-like substance in the eyestalk, i.e. in the sinus gland, lamina ganglionaris, medulla externa, medulla interna, and medulla terminalis. In addition, high performance liquid chromatogra- phy (HPLC) of sinus gland extracts of Carcinus

lobster, Panulirus interruptus.

maenas revealed the presence of substances that appeared to be methionine enkephalin, leucine enkephalin, and methionine enkephalin-Arg°-Phe’ [58]. Likewise, Fingerman ef al. [39] showed by immunocytochemistry the presence of leucine en- kephalin-like and methionine enkephalin-like material in the eyestalk of the fiddler crab, Uca pugilator. Immunoreactivity to both substances was apparent in the retinular cells, lamina ganglio- naris, sinus gland, optic peduncle, the three chiasms and medulla terminalis. However, in the X-organ of the medulla terminalis, immunoreac- tivity against only the methionine enkephalin anti- bodies, but not leucine enkephalin, was seen. Using the mantid shrimp, Squilla mantis, Marino et al. [77] showed that the subesophageal ganglia contain a peptide that is opioid-like because it is able to displace D-Ala’-D-Leu°-enkephalin (DA- DLE) from rat brain membranes in a receptor binding assay, and gives positive results in the standard guinea pig ileum assay for opioids. In this test, opioids inhibit the twitch response of the electrically stimulated ileum. Furthermore, the action of this opioid-like substance is inhibited by naloxone, an opioid receptor blocker. This pep- tide, through use of proteolytic enzymes, gave evidence that it is synthesized as part of a larger inactive peptide and subsequently released in the active form. Because this opioid does not coelute with methionine enkephalin or leucine enkephalin, nor does it cross-react with antibodies to methionine enkephalin or give positive results in RIA for methionine enkephalin, leucine enkepha- lin, or beta-endorphin, Marino et al. suggested this peptide from Squilla mantis is a “native” neuro- peptide. Piccoli et al. [91], also using the DADLE receptor binding assay, found opioid-like DADLE displacement with extracts of brain from Carcinus maenas and brains.plus dorsal organ from Squilla mantis, suggestive of the presence of opioid pep- tides. In an abstract, Dircksen et al. [27] reported that leucine enkephalin-like immunoreactivity, as seen by light and electron microscopic immuno- cytochemistry, was apparent in the pericardial organs and their segmental nerve connections to the thoracic ganglia of Carcinus maenas and in the pericardial organs of the crabs Cancer pagurus, Portunus puber, and Maja squinado.

Vertebrate-type Hormones in Crustaceans 15

Carcinus maenas has also been used in other studies involving enkephalins. Amiard-Triquet er al. [1] by immunofluorescence showed the pre- sence of methionine enkephalin-like material in the hepatopancreas (midgut gland) of this crab. The immunofluorescent material was located mainly basally in hepatopancreas tubule cells of crabs taken from nonpolluted water. However, in crabs from a polluted site or those exposed to Cd, Pb, Cu or Zn for 1-3 weeks the immuno- fluorescence was located mainly apically in these tubule cells. No immunoreactivity was seen in the lumina of these tubules, which suggests that this immunoreactive material is either not secreted into the lumen or loses its immunoreactivity upon being secreted. Also with Carcinus maenas, Dircksen [26] found by immunocytochemistry that leucine enkephalin-like immunoreactive axons are present in the pericardial organs, neurohemal organs that are cardiostimulatory. These immunoreactive axons enter the pericardial organs via segmental nerves 1, 3 and 7. No colocalization of this material with proctolin, FMRFamide or crusta- cean cardioactive peptide was seen. More recent- ly, both methionine enkephalin and leucine en- kephalin were found in extracts of the thoracic ganglion of Carcinus maenas [73]. Their presence was established by RIA, HPLC and sequence analysis of the opioids in thoracic ganglion ex- tracts. An immunocytochemical study by Rothe et al. [94] of the eyestalks of Carcinus maenas re- vealed immunoreactivity to both methionine en- kepahlin and leucine enkephalin in the sinus gland (more leucine enkephalin than methionine en- kephalin being present in this neurohemal organ); only leucine enkephalin in the medulla terminalis, lamina ganglionaris, medulla externa and medulla interna; and only methionine enkephalin in the chiasm between the medulla interna and medulla terminalis. In addition, by HPLC analysis a sub- stance that coelutes with leucine enkephalin was identified in sinus gland extracts.

With another crab, the land crab, Gecarcinus lateralis, Leung et al. [70] by HPLC and RIA analyses found a methionine enkephalin-like sub- stance in the eyestalk, and methionine enkephalin- like and leucine enkephalin-like substances in the brain. Colleti-Previero et al. [22] found three

peptidases in the hemolymph of the crayfish, Asta- cus fluviatilis, that rapidly degrade leucine en- kephalin.

Another opiate first found in mammals is called f-endorphin. It is a 31-amino acid polypeptide. Sephadex G-50 column chromatography of ex- tracts of the hepatopancreas of the red swamp crayfish, Procambarus clarkii, revealed three peaks of immunoreactive f-endorphin-like mate- rial. These immunoreactive peaks were identified by immunocytochemistry and RIA [52]. Cells of the hepatopancreas of this crayfish contain en- zymes capable of degrading $-endorphin [123].

Functions

A Behavior

The first study of a possible role of an opioid in any crustaceans was that of Maldonado and Miralto [74] with Squilla mantis. They found that the defensive response, rapid flexure of the abdo- men, which can be experimentally induced by an electrical shock, can be modified by morphine. This drug increases the threshold current required to elicit the response. The loss of sensitivity is dose-related. Naloxone blocks the action of this opioid; coinjection of morphine and naloxone does not result in a loss of sensitivity.

In a more recent study of locomotor activity of Gecarcinus lateralis, Martinez et al. [79] found that when the crabs were first placed into the activity monitoring chambers locomotor activity increased. FK 33 824, a stable methionine enkephalin analog, markedly enhances this initial activity; but nalox- one blocks this excitatory action of the opioid.

In studies with another crab, Chasmagnathus granulatus, Lozada et al. [72] found that this crab assumes a defensive posture, with both chelae extended and the body elevated, when an electric shock, 50 Hz, one second duration, and of at least 8V, is given. Morphine produces a dose- dependent reduction in the sensitivty of this crab to the electrical shock. When naloxone was co- injected with morphine no reduction in sensitivity occured, which suggests that opioid receptors are indeed involved in this reduction of sensitivity to the electric shock. In other behavioral studies with this crab [10, 117], it was found that a danger stimulus in the form of a passing shadow elicits an

16 M. FINGERMAN, R. NAGABHUSHANAM AND R. SAROJINI

escape response that habituates after repeated stimulation, and that this habituation appears to be mediated by endorphins. Support for this sugges- tion that habituation is the result of endorphin release is the observation that after an habituation session there is an analgesic effect on the defensive response of this crab to an electrical shock. B_ Pigmentary effectors

Crustaceans have two types of pigmentary effec- tors: (a) chromatophores which are responsible for color changes and (b) the retinal pigments, which control the amount of light impinging on the rhabdom. Translocation of the pigments in the chromatophores and at least in the distal and reflecting pigments is clearly regulated by neuro- hormones. (a) Chromatophores

The physiology of crustacean chromatophores has been reviewed recently [38]. In several species dual control of the chromatophores by pigment- dispersing and pigment-concentrating neurohor- mones has been demonstrated. In this laboratory we have been particulary intersted in identifying the neuroregulators that control the release of crustacean neurohormones. In the fiddler crab, Uca pugilator, 5-hydroxytryptamine stimulates re- lease of red pigment-dispersing hormone while norepinephrine stimulates release of black pig- ment-dispersing hormone, and dopamine stimu- lates release of red pigment-concentrating hor- mone and black pigment-concentrating hormone. Methionine-enkephalin, but not leucine- enkephalin, was found to stimulate black and red pigment concentration in intact Uca pugilator, but not in isolated legs [93]. Methionine-enkephalin also stimulated the release of black and red pig- ment-concentrating hormones from isolated eye- stalks [13, 93]. The opioid antagonist, naloxone, blocked the effects of methionine-enkephalin in intact crabs and on isolated eyestalks. Methionine enkephalin, like the classical aminergic neurotrans- mitters, has no direct effect on these chromato- phores, as evidenced by the lack of effect on the chromatophores in isolated legs. Presumably, therefore, these compounds exert their effect only indirectly, by stimulating release of the appropri- ate specific chromatophorotropic neurohormone. However, surprisingly, 8-endorphin was found to

produce black pigment dispersion in both intact Uca pugilator and in isolated legs [93]. Additional evidence for the involvement of opioid-like pep- tidergic substances in crustacean color changes was provided by Martinez et al. [80], who found that although injection of the stable methionine- enkephalin analog FK 33 824 alone into Gecarcinus lateralis has no effect on the chroma- tophores, whereas coinjection of this analog with eyestalk extract strongly potentiates the black pig- ment-dispersing and red pigment-concentrating ac- tions of the eyestalk extract. This potentiation is blocked by naloxone. (b) Distal retinal pigment

Norepinephrine produces light adaptation of the distal retinal pigment of Uca pugilator whereas dopamine induces dark adaptation [62, 64], pre- sumably by stimulating release respectively of the light-adapting and dark-adapting neurohormones. Methionine enkephalin, like dopamine, produces dark adaptation of this pigment, but leucine en- kephalin has no effect on this pigment [63]. C_ Blood glucose

The crustacean hyperglycemic hormone (CHH) is found in the sinus gland. In several crustaceans CHH release has been found to be triggered by 5-hydroxytryptamine [40]. Recently Liischen et al. [73] and Rothe et al. [94] provided evidence that synthetic leucine-enkephalin, endogenous se- quenced leucine-enkephalin isolated from thoracic ganglia of Carcinus maenas, and purified leucine- enkephalin-like material (not yet sequenced) from sinus glands of Carcinus maenas inhibit CHH release. Injection into intact Uca puglilator of synthetic leucine-enkephalin or either the sinus gland or thoracic gland material from Carcinus maenas results in a decrease in the hemolymph glucose concentration. Also, eyestalks of Carcinus incubated in the presence of synthetic leucine enkephalin release less than the basal level of CHH as compared with eyestalks incubated in saline alone. Injection into intact fiddler crabs of naloxone prior to injection of leucine-enkephalin blocks the hypoglycemic action of this opioid. In contrast, leucine-enkephalin does not affect the hemolymph glucose concentration in eyestalkless fiddler crabs, which is consistent with the hypoth- esis that the enkephalin is acting by reducing CHH

Vertebrate-type Hormones in Crustaceans 17

output from the sinus glands in the eyestalks.

HYPOTHALAMIC AND HYPOPHYSIAL HORMONES AND NEUROPHYSIN

Localization and identification

In what appears to have been the first effort to identify a “naturalized” peptide in a crustacean, Martin and Dubois [78] by use of an immuno- fluorescence technique demonstrated the pres- ence of a somatostatin-like substance in the brain, subesophageal ganglion, and ventral nerve cord of the isopod, Porcellio dilatatus. However, no localization of this somatotropin release- inhibiting hormone-like material was apparent in the sinus glands. Van Herp and Bellon-Humbert [120], using the prawn, Palaemon serratus, pro- vided immunocytochemical evidence for neuro- physin-like and arginine vasopressin-like sub- stances in the organ of Bellonci, a component of the eyestalk. However, whereas the neurophysin-like material was present at all times, the vasopressin- like substance was present only at ecdysis. Later, Van Deijnen et al. [119] did an immunocytochemi- cal study of the eyestalk of the crayfish, Astacus leptodactylus, with antibodies against melanocyte- stimulating hormone, vasotocin, and oxytocin. The immunoreactivity they observed to antivaso- tocin was the most widespread while that to anti- oxytocin was the least widespread. The sinus gland reacts positively to antimelanocyte-stimulating hormone and antivasotocin. In addition, cells or fibers reactive: (a) to antivasotocin are in chiasms 1 and 2, the medulla externa, medulla interna, the medulla terminalis neuropil and X-organ of the medulla terminalis, (b) to antialpha-melanocyte- stimulating hormone are in the lamina ganglio- naris, chiasm 1, medulla externa, medulla interna and X-organ of the medulla terminalis; and (c) antioxytocin in the neuropils of the medulla exter- na and medulla terminalis. In a related study, Mattson and Spaziani [82], using arginine vaso- pressin antiserum in an RIA with eyestalk extract of the crab, Cancer antennarius, found vasopressin- like peptides are indeed present.

Mizuno and Takeda [84] used immunocytochemi- stry to identify in several crustaceans neurons

immunoreactive to arginine vasotocin/arginine vasopressin antibodies. Such neurons were seen in brains of the sea louse, Gnorimosphaeroma rayi, and the crab, Hemigrapsus sanguineus, but not in the brain or subesophageal ganglion of Procam- barus clarikii or the crab, Helice tridens. Mizuno and Takeda [83] also looked for oxytocin-like immunoreactivity in the nervous systems of these four crustaceans, but none was seen. In a pre- liminary study done earlier Takeda ef al. [108] examined only isopod central nervous systems by immunocytochemistry with these neurohypophy- sial hormone antibodies. Arginine vasotocin/argi- nine vasopressin immunoreactivity was apparent with Ligia exotica, Porcellio scaber, and Armadilli- dium vulgare, but no oxytocin immunoreactivity was apparent.

A human somatotropin-like substance was iden- tified by RIA in the hemolymph and extracts of the stomach and hepatopancreas of Palaemon serratus [113]. The amounts in the three tissues varied with the molting cycle, being highest in all three tissues during proecdysis.

Functions

Few attempts have been made to identify possi- ble roles in crustaceans of hypophysial com- pounds. Sarojini et al. [99] found that oxytocin elvates the hemolymph glucose concentration in the freshwater crab, Barytelphusa cunicularis and the prawn, Macrobrachium kistnensis. ‘he gly- cogen levels in the hepatopancreas, muscles and integument of both species showed concomitant decreases, which suggests that these glycogen stores were hydrolyzed to provide the glucose that appeared in the hemolymph. Later, Mattson and Spaziani [82] reported that lysine vasopressin, argi- nine vasopressin, vasotocin, and oxytocin mimic the action of molt-inhibiting hormone, inhibiting ecdysteroid production by Y-organs of the crab, Cancer antennarius. Charmantier et al. [17] found that larval and postlarval lobsters, Homarus amer- icanus, grow more rapidly when injected with human somatotropin. This hormone did not bring on precocious molting activity, but the specimens that received the hormone became longer and weighed more than the controls. The hormone appeared to stimulate tissue growth without in-

18 M. FINGERMAN, R. NAGABHUSHANAM AND R. SAROJINI

fluencing the water content of these test indi- viduals.

Zukowska-Arendarczyk [127] tested ovine folli- cle-stimulating hormone and luteinizing hormone on the sand shrimp, Crangon crangon. Both intact and eyestalk ligated females were used. Eyestalk ligation was used to eliminate any influence of the gonad-inhibiting hormone of the sinus gland in the eyestalk. However, the responses when these gonadotropins were injected were qualitatively the same among the ligated and non-ligated indi- viduals, although the ligated individuals did show accelerated vitellogenesis and oocyte growth as compared with the non-ligated shrimp. Follicle- stimulating hormone causes the number of somatic cells in the ovary to increase but does not affect the oogonia or oocytes. In contrast, luteinizing hor- mone affects only the germinal epithelium; the oogonia increasing in number and size and the oocytes undergoing accelerated vitellogenesis, but the somatic cells are not affected. So, both hor- mones stimulate this ovary, but in different ways.

Although chorionic gonadotropin is not an hypophysial hormone, in mammals it is functional- ly similar to luteinizing hormone, and this for this reason is treated here. Bomirski and Klek- Kawinska [9] found that human chorionic gonado- tropin stimulates oogenesis in Crangon crangon. Similarly, Souty and Picaud [106] found that hu- man chorionic gonadotropin stimulates vitel- logenesis by the fat body in the isopod, Jdotea balthica. Maruo et al. [81], using RIA and a radioreceptor assay, found a human chorionic gonadotropin-like substance in the stomach and hepatopancreas of the lady crab, Ovalipes ocella- tus. However, when these investigators tested this substance from the crab stomach in a mouse uterus bioassay, this substance exhibited no biological activity.

GASTRIN/CHOLECYSTOKININ

Localization and Identification

Gastin and cholecystokinin are peptides that have the same five amino acids at their carboxyl end [68]. Consequently, antisera raised against either of these peptides generally are immunoreac-

tive to both. It is now recognized that both gastrin and cholecystokinin belong to a single peptide family called the gastrin/cholecystokinin (G/ CCK) family. Not only do gastrin and cholectysto- kinin function as vertebrate gastrointestinal hor- mones, but both are also present in vertebrate nervous systems, probably functioning there as neurotransmitters. By RIA Larson and Vigna [67] showed the presence of G/CCK-like material in the digestive tract of Cancer magister (the highest concentration occurring in the stomach), and also in the digestive tract of Upogebia pugettensis. This material was not detected in the hepatopancreas of Cancer nor in the digestive tracts of two barnacles, Balanus sp. and Pollicipes polymerus. G/CCK- like material of Cancer magister was analyzed by Sepahdex G-50 chromatography [66] from the stomach, hemolymph and carapace. Three immunoreactive molecular forms of G/CCK-like peptides were found.

Scalise et al. [102], using immunocytochemistry, provided supporting evidence for the presence of G/CCK-like material in the stomach of Cancer magister. This material appears to be synthesized in the gastric epithelial cells and then secreted into the procuticle which lines the stomach lumen. Bellon-Humbert et al. [6] reported that by immunocytochemistry and with an antibody to cholecystokinin 8 they observed that neurosecre- tory cells in the medulla externa and medulla terminalis X-organ of Palaemon serratus are immunoreactive to this antibody.

Van Deijnen er al. [119] who, as mentioned above, used a wide variety of mammalian antisera to test for “naturalized’ peptides in eyestalks of Astacus leptodactylus found immunoreactivity against a gastrin (C-terminal) antibody in the medulla externa nd medulla terminalis X-organ. They [119] also used an antibody to cholecystoki- nin that was raised against the middle portion of the molecule which does not crossreact with gas- This anticholecystokinin revealed the pre- sence of reactive cells in the lamina ganglionaris, medulla externa, medulla interna, medulla termi- nalis (both in the X-organ and the rest of the ganglion), and also in the sinus gland. By im- munofluorescence and RIA Turrigiano and Selver- ston [114] showed G/CCK-like material is in the

trin.

Vertebrate-type Hormones in Crustaceans 19

input nerve and neuropil of the stomatogastric ganglion, commissural ganglia, brain and eyestalks Panulirus interruptus.

Favrel et al. [37], by immunocytochemistry, showed the presence of G/CCK-like peptides in the eyestalk and stomach of Palaemon serratus. In the eyestalk, cells in the medulla externa and medulla terminalis X-organ and the sinus gland were labelled. In the stomach, epithelial cells and the cuticle were labelled also. Like with Cancer magister, Sephadex G-S0 fractionation of eyes- talks, stomach and hemolymph of Palaemon serra- tus revealed multiple molecular forms of G/CCK- like material. The quantity of G/CCK-like pep- tides, as measured by RIA, in the eyestalks showed no significant variation with the molting cycle, but in the hemolymph the concentration was highest, during late proecdysis, just before ecdysis; and for the stomach concentration peaks were apparent during postecdysis and during proecdy- sis. Van Wormhoudt ef al. [121], using shrimps, Penaeus japonicus and Penaeus stylirostris, and an RIA for G/CCK-like peptides found, after frac- tionation on a Sephadex R G-50 SF column, four molecular forms of these peptides in the hemolymph of each shrimp. The amount of each form varies with the species. Four G/CCK-like peptides that cross-reacted with a specific C- terminal G/CCK antiserum were also isolated from extracts of the stomach of the lobster, Nephrops norvegicus, by Favrel et al. [35]. Their molecular weights are in the range 1000-2000 daltons. These four peptides show very weak crossreactivity with human gastrin, about 0.03%. Amino acid analysis showed none of these pep- tides has the same five terminal amino acids at the carboxyl end as human gastrin and cholecystoki- nin, these five being necessary for full im- munoreactivity. Instead they had only three or four of the requisite amino acids.

Functions

Because of the established roles of gastrin and cholecystokinin as regulators of digestion among vertebrates, studies of the possible roles of G/ CCK-like peptides in crustaceans have centered on similar functions. Larson and Vigna [66] reported that G/CCK-like material from Cancer magister

does not stimulate secretion of either gastric acid or pancreatic amylase in the rat. This suggests that the crab material while sufficiently like vertebrate G/CCK to react with the antisera is nevertheless sufficiently different so as to be incapable of elicit- ing a response in the rat. Working with Palaemon serratus, Favrel and Van Wormhoudt [36] found that injection of gastrin 17 results in increased incorporation of leucine into proteins of the hepa- topancreas.

Turrigiano and Selverston [114] found that elec- trical stimulation of the stomatogastric input nerve, which contains G/CCK immunoreactive fibers, results in the release of G/CCK-like pep- tide into the stomatogastric ganglion of Panulirus interruptus. Also, application of cholecystokinin to this ganglion results in increases in the spike frequency and number of spikes per burst of the pyloric rhythm and also activates the gastric mill. In view of these results, these investigators sug- gested that in this lobster G/CCK-like peptide functions as a neuromodulator in the stomatogas- tric ganglion. When Penaeus japonicus and Penaeus stylirostris are fed, the amount of G/ CCK-like material in their hemolymph increases [121]. This increase indicates that these peptides may help regulate the digestive process. Also, in Panulirus interruptus the hemolymph concentra- tion of G/CCK-like peptide increases after feeding (up to fourfold), and the effect lasts about four hours [115]. Concomitant with this increase is an increase in gastric mill activity: Also, the cholecys- tokinin antagonist, proglumide, when injected within the first two hours after feeding, causes a prolonged decrease in gastric mill activity.

Sedlmeier and Resch [103] tested the efficacy of gastrin and cholecystokinin in vitro, using the hepatopancreas of the crayfish, Orconectes limo- sus. Both gastrin and cholecystokinin were found to stimulate release of amylase and proteases into the incubation medium. Favrel et al. [35], using three of the four G/CCK-like peptides they iso- lated from extracts of the stomach of Nephrops norvegicus, found that two of the three stimulate protease secretion in vitro from the hepatopan- creas of Orconectes limosus.

20 M. FINGERMAN, R. NAGABHUSHANAM AND R. SAROJINI

INSULIN

Localization and identification

Although early studies [41, 53, 69] showed that injected mammalian insulin does not affect the hemolymph glucose concentration in crustaceans, interest in the possible presence and roles of insulin-like peptides in crustaceans continues. Davidson et al. [25] reported that the hepatopan- crease of Carcinus maenas shows insulin-like activ- ity in a mouse diaphragm glycogen synthesis test, and that some cells in the hepatopancreas, but not in the gut, of Carcinus maenas and Homarus gammarus have staining characteristics of mamma- lian insulin-secretory pancreatic islet beta cells. But Davidson et al. [25] cautioned about hastily drawing firm conclusions when using light micro- scopic staining features and histochemical staining reactions devised for mammalian cells on tissues of invertebrates. Likewise, Falkmer [34] had earlier found in a study with the use of light microscopy that such insulin-producing cells are absent from the gut of the blue crab, Callinectes sapidus, also.

Sanders [96], using an RIA and guinea pig anti-insulin serum showed that the hepatopan- creas, gut and hemolymph of Homarus americanus contain insulin-like peptides. The hepatopancreas contains the highest concentration of immunoreac- tive material while the eyestalk has none. Frac- tionation of hepatopancreas, hemolymph and gut on a Sephadex G-50 column revealed molecular heterogeneity of the insulin-like material. The hepatopancreas contains the highest proportion of high molecular weight insulin-like material, the gut and hemolymph having relatively more of the lower molecular weight insulin-like material. In view of these results, Sanders [96] hypothesized that a large insulin-like peptide is synthesized in the hepatopancreas, secreted into the gut where it is broken down to smaller units, reabsorbed into the hepatopancreas and then secreted into the hemolymph.

Functions

Sanders [97], using Homarus americanus, found that bovine insulin, hepatopancreas extract, and hemolymph induce glycogenesis in abdominal

muscle of this lobster in vitro, but gut extracts have no such effect. However, injection of these in- sulin-like peptides from the hepatopancreas does not affect the glucose concentration in the hemolymph of Homarus americanus [98]. High glucose concentrations in the hemolymph did not stimulate release of insulin-like peptides into the hemolymph, nor did hepatopancreas extracts speed the rate of glucose clearance from the hemolymph. These findings are consistent with the earlier results from several laboratories mentioned above [41, 53, 69] that mammalian insulin has no effect on the hemolymph glucose concentration in This crustacean insulin-like material presumably functions to regulate glycogenesis in muscle cells, but has no role in hemolymph gluco- stasis. Insulin does not stimulate release of amy- lase or proteinase from the isolated hepatopan- creas of Orconectes limosus [103]. Consistent with the results of Sanders [96, 97] are the observations of Baker and Carruthers [4, 5] that porcine insulin stimulates glucose transport into muscle cells of the barnacle, Balanus nubilis. Baker and Car- ruthers also found with these muscle cells that insulin lowers the cytosolic ionized calcium level, reduces the cyclic AMP content and increases the cyclic GMP content. These observations are con- sistent with the earlier observations of Bittar et al. [7] that insulin stops sequestration of soldium ions in muscle cells of the barnacles, Balanus nubilus and Balanus aquila, and also increases the activity of the guanylate cyclase system. Insulin injected into these cells produces an increase in the rate of sodium efflux and injection of guanosine triphos- phate into muscle cells pre-exposed to insulin also produces an increase in sodium efflux.

crustaceans.

SUBSTANCE P

Localization and identification

In the vertebrate central nervous system sub- stance P apparently serves as a neurotransmitter that mediates sensory input, e.g. transmission of pain impulses to the brain. By immunocytoche- mistry substance P-like material was found in the eyestalk of Panulirus interruptus [75]. This im- munoreactivity material is in the lamina ganglio-

Vertebrate-type Hormones in Crustaceans 21

naris, medulla externa, medulla interna, medulla terminalis (in the X-organ as well as the rest of this ganglion) and sinus gland. In eyestalks of Uca pugilator, substance P-like immunoreactivity was seen in the retinular cells, lamina ganglionaris, medulla externa, medulla interna, sinus gland and optic peduncle [39].

Goldberg et al. [46, 47] by immunocytochemis- try identified cells immunoreactive to substance P antibody in the stomatogastric nervous systems of Panulirus interruputs, Homarus americanus and Cancer borealis. In these three decapods the neuropil of the stomatogastric ganglion is strongly immunoreactive, but none of the stomatogastric ganglion somata showed any immunostaining. Also, in the connective ganglia some somata and the neuropil were positively immunoreactive. Marder [76] noted that the stomatogastric ganglion of Cancer irroratus shows the same pattern of substance P-like immunoreactivity as do the stomatogastric ganglia in these other three de- capods studied by Goldberg et al. [46, 47].

Sanderman et al. [95] by immunocytochemistry demonstrated substance P-like immunoreactivity in cells of the brains of the crab, Leptograpsus variegatus, and the crayfish, Cherax destructor. Four large neurons in the brain of each of these crustaceans are immunoreactive, two in the proto- cerebrum and two in the deutocerebrum. These paired immunoreactive protocerebral cells are, more specifically, in the dorsal cell clusters that lie on each side of the median protocerebrum. Each of these cells extends through the ipsilateral and contralateral olfactory lobes to terminate among the lateral somata of the olfactory lobe, not in the neuropil. The cells from each side of the brain are mirror images of each other. Each deutocerebral cell runs only ipsilaterally, terminating within the neuropil of the olfactory lobes.

Functions

Substance P was found to increase the amounts of amylase and proteases released in vitro from the hepatopancreas of Orconectes limosus [103].

CALCITONIN and CALCITONIN GENE-RELATED PEPTIDE

Localization and identification

Calcitonin is a small polypeptide secreted by the ultimobrachial glands in jawed fishes, amphibians, reptiles and birds; and in mammals by the so-called C cells of the thyroid gland. As a hormone, calcitonin produces hypocalcemia in vertebrates. Calcitonin is also present in the vertebrate central nervous system, presumably functioning there as a neuromodulator. The gene that codes for calcito- nin also produces another polypeptide, called cal- citonin gene-related peptide (CGRP). This sub- stance is present in central and peripheral nervous systems of vertebrates, where it appears to func- tion as a neuromodulator, and in several other regions of the vertebrate body, including the heart, spleen and urogential tract as well [89].

Bellon-Humbert er al. [6] by use of immuno- cytochemistry and antibodies against human, pig and salmon calcitonins demonstrated the presence of a calcitonin-like substance in the medulla termi- nalis X-organ of Palaemon serratus. Arlot- Bonnemains et al. [3] by RIA with anticalcitonin antibodies from salmon showed the presence of a calcitonin-like polypeptide in the hemolymph of Palaemon serratus. The amount of calcitonin-like material in the hemolymph varies with the molting cycle, being highest during postecdysis, and lowest during intermolt The hemolymph calcium level, however, was maximal during proecydsis and lowest during postecdysis.

Using RIA, a radioreceptor assay and anti- salmon and antihuman calcitonin antibodies, Fun- chereau-Peron et al. [42] found a calcitonin-like peptide in the eyestalks of Nephrops norvegicus, Homarus americanus, Palaemon serratus, and Penaeus vanamei. In addition, other organs of Nephrops were examined. The hepatopancreas of this lobster contains a higher concentration of calcitonin-like peptide than any of the eyestalks examined from these four species. The foregut of Nephrops also contains a high concentration, like- wise higher than in these eyestalks, but less than in the hepatopancreas. This calcitonin-like material of Nephrops that was detected by the antisalmon

22, M. FINGERMAN, R. NAGABHUSHANAM AND R. SAROJINI

calcitonin antibodies exists in two forms. One has a molecular weight of 4500 daltons, somewhat higher than vertebrate calcitonins. Salmon calcito- nin is 3432 daltons, and human calcitonin is 3418 daltons. However, in addition to this 4500 dalton molecule, a higher molecular weight calcitonin- like substance was also found. The latter was found by gel filtration to have a molecular weight greater than 5,000 daltons.

Graf et al. [48], using extracts of an entire amphipod, Orchestia cavimana, in an RIA found a calcitonin-like substance, the concentration of which changes with the molting cycle. The maxi- mum amount is present at ecdysis, while the least is found early in intermolt. The concentration increases markedly late in proecdysis. The least is seen early in intermolt.

Arlot-Bonnemains et al. [2] with an RIA and human anti-CGRP antibody found immunoreac- tive material in several tissues of Nephrops norve- gicus, including the hepatopancreas, foregut, eyes- talk, brain and heart. Of the tissues examined, the hepatopancreas has the highest concentration of this CGRP-like material, followed by the foregut, which led these investigators to suggest the CGRP- like substance may function in crustaceans as a regulator of gastrointestinal function.

Sasayama et al. [100] demonstrated by im- munocytochemistry with antibodies to salmon cal- citonin and rat CGRP the presence of cells im- munoreactive to both antibodies in the central nervous system of Armadillidium vulgare. CGRP immunoreactive cells are present in the brain and circumesophageal connectives while calcitonin immunoreactive cells are in the brain only. More recently, Cameron and Thomas [14] by use of RIA showed the presence of salmon calcitonin-like material in several tissues, including the hepato- pancreas, brain, heart and hemolymph of the blue crab, Callinectes sapidus. The hepatopancreas contains the highest concentration of immunoreac- tive material. The concentration of immunoreac- tive material in the hemolymph varies with the molting cycle, being highest during proecdysis, but the calcitonin-like material in the other tissues does not vary with the molting cycle. The im- munoreactive material in the hemolymph and hepatopancreas consists to a great extent of a 27.2

kilodalton protein with an amino acid composition very similar to those of the 17.5 kilodalton human calcitonin precursor and the 22 kilodalton calcito- nin-like molecule isolated earlier from the hemolymph and hepatopancreas of Nephrops nor- vegicus by Van Wormhoudt and Fouchereau- Peron [122]. However, in Callinectes there is also some immunoreactive material of relatively low molecular weight. These high molecular weight substances of 22 and 27.2 kilodaltons likely corres- pond to the one found by Funchereau-Peron et al. [42] in Nephrops that was larger than 5,000 dal- tons, whereas the low molecular weight material from Callinectes likely corresponds to the 4,500 dalton calcitonin-like compound found by Fouchereau-Peron et al. in Nephrops [42].

Functions

The roles of calcitonin and CGRP in crustaceans have not been examined extensively. In a study of Palaemon serratus by Arlot-Bonnemains et al. [3] they found that injection of salmon calcitonin during preecdysis does not alter the hemolymph calcium concentration. Sellem et al. [104] investi- gated the effect of salmon calcitonin on the con- centration of calcium in the hemolymph of Orches- tia cavimana. They found that the calcium level decreased when the calcitonin was injected during intermolt and early proecdysis but had no effect during late proecdysis or during postecdysis. Nor- mally, the hemolymph calcium level is high during proecdysis because of calcium resorption from the old cuticle, and is low during intermolt. As mentioned above, Graf er al. [48] found that calcitonin-like material is at its lowest concentra- tion in Orchestia cavimana at the start of intermolt and increases rapidly during proecdysis to peak at ecdysis. So, the exogenous salmon calcitonin was effective when the endogenous calcitonin-like activity was low, but ineffective when the endog- enous activity was high.

MISCELLANEOUS PEPTIDES

Localization and identification

A few “naturalized” peptides have been the subject of only one or two investigations. These

Vertebrate-type Hormones in Crustaceans 23

substances will be treated together in this section. Van Deijnen ef al. [119], in the immunocyto- chemical study of the eyestalks of Astacus lepto- dactylus already referred to above, found cells immunoreactive to three antisera in addition to those mentioned earlier. Cells or fibers showing immunoreaction (a) to antisecretin are in the lami- na ganglionaris, medulla externa, medulla interna, medulla terminalis and sinus gland; (b) to antiglu- cagon are in the lamina ganglionaris, medulla externa, medulla interna, medulla terminalis and chiasm 1 and (c) to antiglucose-dependent insuli- notropic peptide (antigastric inhibitory peptide) are in the medulla externa, medulla interna, medulla terminalis and chiasm 1. Charmantier-Daures et al. [18] did an immunocytochemical study of the eyestalks of Homarus gammarus that involved antibodies against neuropeptide Y and the atrial natriuretic factor. A few cells that are immunoreactive against the atrial natriuretic factor antibody are present in the medulla externa and medulla termi- nalis whereas the sinus gland and a few cells and numerous fibers in the medulla interna and medul- la terminalis react to neuropeptide Y antiserum.

Functions

Turrin et al. [116] using an RIA for atrial natriuretic peptide found that the amount of this substance in the hemolymph of the mangrove crab, Ucides cordatus, is related to the environmental salinity. When crabs are transferred from 26 parts per thousand sea water, which is isosmotic with the hemolymph of this crab, to 34 parts per thousand sea water the concentration of this peptide in the hemolymph icreases significantly. These investiga- tors suggested that this peptide has a role in the osmoregulation of this crab, perhaps to accelerate sodium excretion under when the crab is in a hyperosmotic environment. Sedlmeier and Resch [103] in a study already referred to above found that glucagon, like insulin, does not produce in vitro release amylase or proteinase from the hepa- topancreas of Orconectes limosus whereas bombe- sin, like gastrin, cholecystokinin, and substance P does stimulate release of these enzymes in vitro.

EICOSANOIDS

Localization and identification

This section will be devoted to those derivatives of fatty acids collectively known as eicosanoids, the best known of which are the prostaglandins. Prostaglandins were first identified in mammalian semen and thought secreted by the prostate gland (hence their name), although it is now clear they are secreted into the semen by the seminal vesi- cles. However, prostaglandins are by no means limited to the reproductive system. Prostaglandins are produced in virtually all mammalian tissues, and function as chemical messengers. Other eico- sanids include the leukotrienes, thromboxanes and such hydroxyunsaturated fatty acids as the barna- cle hatching substance.

The barnacle hatching substance is released by adult barnacles into their mantle cavity where it stimulates hatching of nauplii that underwent their early development in the mantle cavity. While several barnacle tissues can synthesize the hatching substance, the epidermis appears to be the major site of synthesis. Preliminary evidence led Clare et al. [19-21] to suggest that the hatching substance of Semibalanus balanoides is a prostaglandin-like compound. In 1985, the same year as the third publication on the subject by Clare et al. [21], Holland et al. [56] showed that this hatching sub- stance of Semibalanus balanoides is an eicosanoid, but not a prostaglandin. Prostaglandins are cyclic compounds whereas this hatching substance is from the linear, acyclic, pathway of unsaturated fatty metabolism for eicosanoid synthesis, and is therefore more closely related to the leukotrienes. Nevertheless, a histochemical staining technique for prostaglandin synthetase gives positive results with ovaries and developing egg masses of Semi- balanus balanoides, and extracts of these egg mas- ses have thin layer chromatography characteristics similar to those of the primary prostaglandins [55]. Holland et al. [56] proved by thin layer chroma- tography, gas chromatography, and mass spec- trometry that the hatching substance of Semibala- nus balanoides is 10, 11, 12-trihydroxy-5, 8, 14, 17-eicosatetraenoic acid. Later, Hill et al. [54] showed that the hatching substance of another

24 M. FINGERMAN, R. NAGABHUSHANAM AND R. SAROJINI

barnacle, Elminius modestus, has a different struc- ture. It appears to be an eicosanoid also, apparently a monohydroxyeicosapenataenoic acid isomer, probably 8-hydroxyeicosapentaenoic acid. Although this substance is different from the hatching substance identified by Holland et al. [56] for Semibalanus balanoides, the hatching sub- stance of Elminius modestus is effective not only with Elminius but also with Semibalanus bala- noldes.

Hampson et al. [51] found that hemolymph cells of Carcinus maenas are capable of synthesizing eicosanoids such as 5-hydroxyeicosatetraenoic acid, prostaglandin E, and thromboxane B>. The calcium ionophore A23187 was used to stimulate this eicosanoid production

Functions

Casterlin and Reynolds [15] used the crayfish, Cambarus bartonii, in one study and in another they [16] used Homarus americanus and Penaeus duorarum. The aim of both studies was to deter- mine whether injection of prostaglandin E; would alter the preferred temperatures of these crusta- ceans. This prostaglandin does indeed evoke an increase in the preferred water temperatures. The test chambers were designed so that each specimen could control the water temperature, thereby re- vealing its temperature preference. The crayfish showed a 3.4°C increase in temperature preference with 0.5 mg of prostaglandin, the lobster 4.7°C, and the shrimp 4.5°C, the lobster and shrimp receiving 0.1 mg. Prostaglandins may have a role in thermoregulation in these crustaceans, even though it would be minor because they are basical- ly ectotherms, perhaps by stimulating a thermo- regulatory control center.

The crab, Rhithropanopeus harrisii, produces a rhythmic abdominal flexion during and after copulation. This behavior can be induced by exposing males or females to the crab’s semen or homogenates or extracts of the crab’s seminal vesicles [44]. This abdominal flexion can also be induced by prostaglandin F, and prostaglandin E>. However, prostaglandins have not yet been iden- tified in crab seminal fluid.

Prostaglandins may have a role in regulating ion

transport across crustacean gills. Siebers ef al.

[105] found that indomethacin, an inhibitor of prostaglandin synthesis, produces a decrease of the transbranchial potential across isolated gills of Carcinus maenas when applied to the basolateral side of the gills. Concomitantly, the rate of chloride ion influx decreased significantly in the presence of indomethacin.

Spaziani et al. [107] incubated ovaries of Pro- cambarus paeninsulanis, with the prostaglandin precursor arachadonic acid, and then isolated from these ovaries a newly synthesized compound that comigrated on thin layer chromatography plates with prostaglandin F,,. Synthesis of this prosta- glandin was inhibited by indomethacin. The rate of synthesis was higher when ovaries in the early stages of vitellogenesis were used rather than later stages. As suggested by these investigators, prostaglandins may have a normal role in regulating vitellogenesis in this crayfish.

Koskela et al. [61], working with the tiger prawn, Penaeus esculentus, found that prostaglan- din E, produces a decrease in the length of the intermolt cycle; but does not stimulate ovarian development. Furthermore, neither 17a- hydroxyprogesterone nor 17/-estradiol affects the duration of the molting cycle or ovarian develop- ment of Penaeus esculentus.

STEROIDS

Localization and identification

Several steroids that are ordinarliy identified with mammals have been found in crustaceans. Donahue [28], in what appears to have been the first report that a “naturalized” steroid is present in crustaceans, found that the ovaries of Panulirus argus contain an estrogenic substance. Ovarian extracts elicited growth of the vaginal epithelium in mature, ovariectomized rats. Later, Donahue [29] by fluorimetric analysis and the rat vaginal assay he used previously [28] showed that eggs of Homarus americanus, at the time the eggs attach to the pleopods, contain an estrogenic substance. Donahue [31] identified this estrogenic substance as alpha-estradiol. Lisk [71] later reported that eggs of Homarus americanus contain estradiol- 178. Estrone was not found. Whole body extracts

Vertebrate-type Hormones in Crustaceans 25

of the euphausiacean, Euphausia superba, were found to contain the steroids progesterone, testos- terone and esterone [87]. Jeng et al. [59] by use of gas chromatography, RIA, and a bioassay based on weight increases of the mouse uterus found that the ovaries of Parapenaeus fissurus contain both estrone and estradiol-17f, with estrone being the major estrogen present. In a recent study of the blue crab, Callinectes sapidus [101], neither 17/- estradiol nor estrone was detected in males or females. However, a compound similar to estriol is present in the hemolymph, hepatopancreas, Ovaries, testes and eyestalks.

Gilgan and Idler [45] found that the testes and androgenic glands of Homarus americanus are each capable of converting androstenedione to testosterone. This conversion requires the pre- sence of the enzyme 17f$-hydroxysteroid dehyd- rogenase (oxidoreductase). Blanchet et al. [8] showed this enzyme must also be present in the testes and vasa deferentia of male Carcinus maenas because extracts of these organs converted andros- tenedione, dehydroepiandrosterone and estrone respectively to testosterone, androst-5-ene-3f, 17f-diol and estradiol-178. Burns et al. [11] showed that testes of Homarus americanus can convert progesterone to 20a-dihydroprogesterone. This conversion requires the enzyme steroid 20- ketone reductase. The same investigators [12] also showed that testosterone is present in the hemolymph and testes of Homarus americanus.

Tcholakian and Eik-Nes [109, 110] found that the androgenic gland of Callinectes sapidus has the enzymatic capability to convert progesterone to testosterone, 11-deoxycorticosterone, 20a-hydro- xyprogesterone, and 4‘-androstenedione. Also, intact male crabs converted A°-pregnenolone to progesterone, A*-androstenedione, testosterone and 11-deoxycorticosterone.

Teshima and Kanazawa [111, 112] found that the ovaries of the crab, Portunus trituberculatus, can convert progesterone to 11-ketotestosterone, 116-hydroxyandrost-4-ene-3, 17-dione, testoster- one, 17a-hydroxyprogesterone and deoxycorticos- terone. Kanazawa and Teshima [60] in another study of steroids used Panulirus japonica. In this study they injected radioactive cholesterol and then found that the hepatopancreas, ovaries and

hemolymph contained newly synthesized prog- esterone, 17a-hydroxyprogesterone, andros- tenedione and testosterone, and in addition the hepatopancreas contained deoxycorticosterone and corticosterone. Vitellogenic ovaries of Penaeus monodon metabolize progesterone main- ly to four Sa-pregnane derivatives (Sa-pregnan- 3,20-dione, 20a-hydroxy-Sa-pregnan-3-one, 3/- hydroxy-Sa-pregnan-20-one and 5a-pregnan-3/, 20a-diol) plus two minor metabolites, 20a- hydroxypregn-4-en-3-one and 1,4-pregnadiene- 3,20-dione; whereas vitellogenic ovaries of Nephrops norvegicus metabolize progesterone to only one metabolite, 20a-hydroxypregn-4-en-3- one [126].

Ollevier et al. [90] analyzed the hemolymph of Astacus leptodactylus for nonecdysteroid steroids. In both sexes pregnenalone, 17a-hydroxypregne- nalone, testosterone and cholesterol were clearly identified, and there was indication that andros- tenedione, S5a-dihydroxytestosterone, 11-ketotes- tosterone and 11$-hydroxytestosterone were also present. However, with the techniques they used, progesterone, 17a- hydroxyprogesterone and estrogens could not be detected in either sex, but females alone have 68-hydroxyprogesterone.

Couch et al. [23], using RIA, found progester- one but not testosterone, in mandibular organs of Homarus americanus. Later Couch et al. [24], again with RIA, and using female specimens re- ported the presence of estradiol-178 and pro- gesterone in the mandibular organ, green gland, hepatopancreas, ovary and hemolymph of Homar- us americanus. However, the estradiol was not detectable in any of these tissues if the female had immature ovaries. Progesterone was always pre- sent in the mandibular organs; its concentration is unrelated to the state of the ovaries. But in all other tissues progesterone was not detectable or

dehydroepiandrosterone,

present only in low concentrations when the ovar- ies were immature.

Van Beek and De Loof [118] analyzed by RIA total body extracts of adult female brine shrimp, Artemia sp. Progesterone and pregnenolone were low during vitellogenesis, but high at the beginning and end of each vitellogenic cycle. The amount of 5a-dihydroxytestosterone was very low at the start

26 M. FINGERMAN, R. NAGABHUSHANAM AND R. SAROJINI

of vitellogenesis, peaking at the end of vitellogene- sis. The testosterone level was low during the vitellogenic cycle, and highest in nonvitellogenic ovaries. Estrone was higher during vitellogenesis than at other times, but the fluctuation was not large. Estradiol increased during vitellogenesis, peaking at the end of vitellogenesis. Pregnenalone was present in higher concenterations than any of the other steroids.

In a study designed to detect both conjugated and unconjugated steriods Fairs ef al. [32], using gas chromatography/mass spectrometry, ex- amined tissues of Nephrops norvegicus. Found in the ovary in unconjugated form were S5a- dihydroxytestosterone, testosterone, pregneno- lone and 20a-hydroxypreg-4-en-3-one. Eggs and hemolymph contained unconjugated 17/-estradiol while the ovary and hemolymph contained conju- gated 5a-dihydrotestosterone. Fair ef al. [33] then analyzed the steroids in the ovary of Penaeus monodon by the same techniques. Conjugated pregnenolone, conjugated and unconjugated de- hydroepiandrosterone, unconjugated progester- one, conjugated 17-estradiol and conjugated and unconjugated estrone were present in these ova- ries, peaking during mid to late vitellogenesis, but conjugated testosterone was present only in ma- ture ovaries.

Functions

Donahue [30], who showed that eggs of Homar- us americanus contain an estrogen, looked for a possible role for this substance. Berried females undergo ecdysis only after the larvae have hatch- ed. When he stripped newly laid eggs from the swimmerets, the majority of the lobsters under- went a premature ecdysis. He suggested the eggs contain a moltinhibiting factor. He then pro- ceeded to inject estradiol into male and female lobsters, and found that molting activity was inhi- bited. Danahue concluded by suggesting that ecdysis of berried females is normally inhibited by egg estrogen and perhaps also ovarian estrogen.

Kulkarni er al. [65] found that progesterone stimulates ovarian maturation in the prawn, Para- penaeopsis hardwickii. The ovary size, oocyte size, and ovarian fat content all increased. With another species of this prawn, Parapenaeopsis sty-

lifera, Nagabhushanam et al. [85] found that 17- hydroxyprogesterone induces spawning at 20°C whereas normally spawning does not occur until the water temperature reaches 30°C. Also with male Parapenaeopsis hardwickii Nagabhushanam and Kulkarni [86] observed that testosterone injec- tions produce hypertrophy and hyperplasia of the androgenic glands which develop to a functional state only in male crustaceans. There was a concomitant stimulation of testicular develop- ment, presumably under the influence of androgenic gland hormone released as a conse- quence of the effect of the testosterone on the andorgenic gland.

Yano [124] observed, using the greasyback shrimp, Metapenaeus ensis, that progesterone in- jections induce ovarian maturation and spawning of the eggs from these ovaries. Later, Yano [125] using the kuruma prawn, Penaeus japonicus, con- cluded that injection of 17a-hydroxyprogesterone induces vitellogenesis, as evidenced by an increase in the amount of vitellogenin in the hemolymph of the specimens treated with this hormone.

Koskela et al. [61], as noted above, reported that injection of prostaglandin E> esculentus results in a significant decrease in the length of the molting cycle but does not affect ovarian development. However, neither 17a- hydroxyprogesterone nor 17/-estradiol affected the molting cycle or the ovaries of this shrimp.

into Penaeus

CONCLUDING REMARKS

Despite the large body of knowledge that has accumulated in recent years concerning the iden- tity of many vertebrate-type hormones in crusta- ceans, largely as a result of improved immunologic and chromatographic procedures as well as the possibilities offered by recombinant DNA technology/molecular biology, the precise func- tions and mechanisms of action of these com- pounds remain to be established. In regard to the vertebrate-type peptides in crustaceans, the biochemical and physiological roles of these pep- tides (including their receptor systems) in crusta- ceans need to be firmly established in order to eliminate any doubt as to which ones are normally involved in the vital functions of this large group of

Vertebrate-type Hormones in Crustaceans 27

invertebrates.

Similarly, with reference to the vertebrate-type steroids in crustaceans we need to firmly establish their roles, and to determine their relative impor- tance vis-a-vis the “native” steroids, the ecdyster- oids, in the regulation of molting and reproduc-

tion.

Not only are ecdysteroids the crustacean

molt-promoting hormones, but there is also evi- dence that at least in some crustaceans, vitel- logenesis 1s ecdysteroid-dependent.

12

REFERENCES

Amiard-Triquet C, Amiard J-C, Ferrand R, Andersen AC, Dubois M P (1986) Ecotoxicol Environ Safety 11: 198-209

Arlot-Bonnemains Y, Fouchereau-Peron M, Ches- nais J, Taboulet J, Milhaud G, Moukhtar MS (1991) Gen Comp Endocrinol 83: 1-6 Arlot-Bonnemains Y, Van Wormhoudt A, Favrel P, Fouchereau-Peron M, Milhaud G, Moukhtar MS (1986) Experientia 42: 419-420

Baker PF, Carruthers A (1980) Nature 286: 276- 279

Baker PF, Carruthers A (1983) J Physiol 336: 397- 431

Bellon-Humbert C, Van Herp F, Van Wormhoudt A (1984) Ann Soc Roy Zool Belg 114 (Suppl 1): 164

Bittar EE, Schultz R, Harkness C (1977) J Mem- brane Biol 34: 203-222 : Blanchet M-F, Ozon R, Meusy J-J (1972) Comp Biochem Physiol 41B: 251-261

Bomuirski A, Klek-Kawinska, E (1976) Gen Comp Endocrinol 30: 239-242

Brunner D, Maldonado H (1988) J Comp Physiol 162A: 687-694

Burns BG, Sangalang GB, Freeman HC, McMenemy M (1984) Gen Comp Endocrinol 54: 422-428

Burns BG, Sangalang GB, Freeman HC, McMenemy M (1984) Gen Comp Endocrinol 54: 429-432

Butler TA, Fingerman M (1985) Amer Zool 25: 102A

Cameron JN, Thomas P (1992) J Exp Zool 262: 279-286

Casterlin ME, Reynolds WW (1978) Pharmacol Biolchem Behav 9: 593-595

Casterlin ME, Reynolds WW (1979) Life Sci 25: 1601-1604

Charmantier G, Charmantier-Daures M, Aiken DE (1989) CR Acad Sci Paris 308 Ser III: 21--26

18

20

21

39

40

Charmantier-Daures M, Danger J-M, Netchitailo P, Pelletier G, Vaudry H (1987) CR Acad Sci Paris 305 Ser III: 479-483

Clare AS, Walker G, Holland DL, Crisp DJ (1982) Mar Biol Letters 3: 113-120

Clare AS, Walker G, Holland DL, Crisp DJ (1984) In “Advances in Invertebrate Reproduction, Vol III” Ed by W Engels, Elsevier, Amsterdam, p 569 Clare AS, Walker G, Holland DH, Crisp DJ (1985) Proc Roy Soc Lond 224B: 131-147 Coletti-Previero M-A, Mattras H, Zwilling R, Pre- viero A (1985) Neuropeptides 6: 405-415

Couch EF, Adejuwon CA, Koide SS (1978) Amer Zool 18: 610

Couch EF, Hagino N, Lee JW (1987) Comp Biochem Physiol 87A: 765-770

Davidson JK, Falkmer S, Mehrotra BK, Wilson S (1971) Gen Comp Endocrinol 17: 388-401 Dircksen H (1990) In “Frontiers in Crustacean Neurobiology Advances in Life Sciences” Ed by K Wiese, W-D Krenz, J Tautz, H Reichert, B Mul- loney, Birkhauser Verlag, Basel, pp 485-491 Dircksen H, Stangier J, Keller R (1987) Gen Comp Endocrinol 66: 42

Donahue JK (1940) Endocrinology 27: 149-152 Donahue JK (1948) Proc Soc Exp Biol Med 69: 179-181

Donahue JK (1955) Research Bulletin No. 24 State of Maine Department of Sea and Shore Fisheries Donahue JK (1957) Research Bulletin No. 28, State of Maine Department of Sea and Shore Fisheries

Fairs NJ, Evershed RP, Quinlan PT, Goad LJ (1989) Gen Comp Endocrinol 74: 199-208

Fairs NJ, Zuinlan PT, Goad LJ (1990) Aquacul- ture 89: 83-99

Falkmer S (1969) In “Diabetes Proc Int Diab Fed 6th Congr Stockholm 1967” Ed by J Ostman and RDG Milner, Excerpta Med Found Int Congr Ser 172: 55-66

Favrel P, Kegl G, Sedlmeier, D Keller R, Van Wormhoudt A (1991) Biochimie 73: 1233-1239 Favrel P, Van Wormhoudt A (1986) Bull Soc Zool France 111: 21

Favrel P, Van Wourmhoudt A, Studler JM, Bellon C (1987) Gen Comp Endocrinol 65: 363-372 Fingerman M (1988) In “Endocrinology of Selected Invertebrate Types” Ed by H Laufer, RGH Downer, ARLiss Inc, New York, pp 357- 374

Fingerman M, Hanumante MM, Kulkarni GK, Ikeda R, Vacca LL (1985) Cell Tissue Res 241: 473-477

Fingerman M, Nagabhushanam R (1992) Comp Biochem Physiol 102C: 343-352

28

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

M. FINGERMAN, R. NAGABHUSHANAM AND R. SAROJINI

Florkin M, Duchateau G (1939) CR Soc Biol 132: 484-486

Fouchereau-Peron M, Arlot-Bonnemains Y, Milhaud G, Moukhtar MS (1987) Gen Comp Endocrinol 65: 179-183

Frederickson RCA, Geary LE (1982) Prog Neuro- biol 19: 19-69

Gerhart DJ, Clare AS, Eisenman K, Rittschof D, Forward RB, Jr (1990) In “Progress in Compara- tive Endocrinology” Ed by A Epple CG Scanes MH Stetson Wiley-Liss, New York, pp 598-602 Gilgan MW, Idler DR (1967) Gen Comp Endocri- nol 9: 319-324

Goldberg D, Nusbaum MP, Marder E (1986) Soc Neurosci Abstr 12: 242

Goldberg D, Nusbaum MP, Marder E (1988) Cell Tissue Res 252: 515-522

Graf F, Fouchereau-Peron M, Van Wormhoudt A, Meyran JC (1989) Gen Comp Endocrinol 73: 80- 84

Greenberg, MJ, Price DA (1983) Ann Rev Physiol 45: 271-288

Gros C, Lafon-Cazal M, Dray F (1978) C R Acad Sci Paris 287D: 647-650

Hampson AJ, Rowley AF, Barrow SE, Steadman R (1992) Biochim Biophys Acta 1124: 143-150 Hara M, Kouyama H, Watabe S, Yago N (1985) Zool Sci (Tokyo) 2: 971

Hemmingsen AM (1924) Skand Arch Physiol 46: 56-63

Hill EM, Holland DL, Gibson KH, Clayton E, Oldfield A (1988) Proc Roy Soc Lond 234B: 455- 461

Holland DL (1987) In “Barnacle Biology” Ed by AJ Southward Balkema, Rotterdam, pp 227—248 Holland DL, East J, Gibson KH, Clayton E, Oldfield A (1985) Prostaglandins 29: 1021-1029 Hughes J, Smith TW, Kosterlitz HW, Fothergill LA, Morgan BA, Morris HR (1975) Nature 258: 577-579

Jaros PP, Dircksen H, Keller R (1985) Cell Tissue Res 241: 111-117

Jeng SS, Wan WC-M, Chang CF (1978) Comp Endocrinol 36: 211-214

Kanazawa A, Teshima S (1971) Bull Jap Soc Sci Fish 37: 891-898

Koskela RW, Greenwood JG, Rothlisberg PC (1992) Comp Biochem Physiol 101A; 295-299 Kulkarni GK, Fingerman M (1986) Comp Biochem Physiol 84C: 219-224

Kulkarni GK, Fingerman M (1987) Pigment Cell Res 1: 51-56

Kulkarni GK, Fingerman M (1991) Biochem Physiol 99C: 47-52

Kulkarni GK, Nagabhushanam R, Joshi PK (1979)

Gen

Comp

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

Indian J Exp Biol 17: 986-987

Larson BA, Vigna SR (1983) Regul Pept 7: 155- 170

Larson BA, Vigna SR (1983) Gen Comp Endocri- nol 50: 469-475

Larsson L-I, Rehfeld JF (1977) Cytochem 25: 1317-1321

Leinen R, McWhinnie MA (1971) Gen Comp Endocrinol 16: 607-611

Leung MK, Kessler H, Whitefield K, Murray M, Martinez EA, Stefano GB (1987) Cell Molec Neurobiol 7: 91-96

Lisk RD (1961) Can J Biochem Physiol 39: 659- 662

Lozada M, Romano A, Maldonado H (1988) Phar- macol Biochem Behav 30: 635-640

Liischen W, Buck F, Willig A, Jaros PP (1991) Proc Natl Acad Sci USA 88: 8671-8675 Maldonado H, Miralto A (1982) J Comp Physiol 147: 455-459

Mancillas JR, McGinty JF, Selverston AI, Karten H, Bloom FE (1981) Nature 293: 576-578 Marder E (1986) In “The Crustacean Stomatogas- tric System” Ed by AI Selverston, M Moulins Springer-Verlag, New York, pp 263-300

Marino G, Palmisano A, Di Marzo V, Melck D (1985) Peptides 6 (Suppl 3): 403-406

Martin G, Dubois MP (1981) Gen Comp Endocri- nol 45: 125-130

Martinez EA, Murray M, Leung MK, Stefano GB (1988) Comp Biochem Physiol 90C: 89-93 Martinez EA, Vassell D, Stefano GB (1986) Comp Biochem Phsyiol 83C: 77-82

Maruo T, Segal SJ, Koide SS (1979) Endocrinolo- gy 104: 932-939

Mattason MP, Spaziani E (1985) Peptides 6: 635- 640

Mizuno J, Takeda N (1988) Comp Biochem Phy- siol 91A: 733-738

Mizuno J, Takeda N (1988) Comp Biochem Phy- siol 91A: 739-747

Nagabhushanam R, Joshi PK, Kulkarni GK (1980) Indian J Mar Sci 9: 227

Nagabhushanam R, Kulkarni GK (1981) Aquacul- ture 23: 19-27

Nikitina SM, Savchenko ON, Kogan ME, Ezhkova NS (1977) Zh Evol Biokhim Fiziol 13: 443-447 Norman AW, Litwack G (1987) Hormones, Academic Press, New York, pp 2-7

Okimura Y, Chihara K, Abe H, Kita T, Kashio Y, Sato M, Fujita T (1987) Regul Pept 17: 327-337 Ollevier F, De Clerck D, Diederik H, De Loof A (1986) Gen Comp Endocrinol 61: 214-228 Piccoli R, Melck D, Spagnuolo A, Vescia S, Zanet- ti L (1985) Comp Biochem Physiol 80C: 237-240

J Histochem

92

93

94

95

96

oy

98

99

100

101

102

103

104

105

106

107

108

109

Vertebrate-type Hormones in Crustaceans 29

Platt N, Reynolds SE (1988) In “Comparative Invertebrate Neurochemistry” Ed by GG Lunt, RW Olsen, Cornell University Press, Ithaca New York, pp 175-226

Quackenbush LS, Fingerman M (1984) Comp Biochem Physiol 79C: 77—84

Rothe H, LuSchen W, Asken A, Willing A, Jaros PP (1991) Cmp Biochem Physiol 99C: 57-62 Sandeman RE, Sandeman DC, Watson AHD (1990) J Comp Neurol 294: 569-582

Sanders B (1983) Gen Comp Endocrinol 50: 366- 373

Sanders B (1983) Gen Comp Endocrinol 50: 374- 377

Sanders B (1983) Gen Comp Endocrinol 50: 378- 382

Sarojini R, Jahagirdar VG, Nagabhushanam R (1980) J Adv Zool 1: 54-61

Sasayama Y, Katoh A, Oguro C, Kambegawa A, Yoshizawa H (1991) Gen Comp Endocrinol 83: 406-414

Sasser ER, Singhas CA (1992) Aquaculture 104: 367-373

Scalise FW, Larson BA, Vigna SR (1984) Cell Tissue Res 228: 113-119

Sedimeier D, Resch G (1989) XIth Inter Symp Comp Endocrinol Malaga Spain Abst P-315 Sellem E, Graf F, Meyran J-C (1989) J Exp Zool 249: 177-181

Siebers D, Mustafa T, Bottcher K (1991) Biol Bull 181: 312-314

Souty C, Picaud J-L (1984) Gen Comp Endocrinol 54: 418-421

Spaziani EP, Hinsch GW, Edwards SC (1991) Amer Zool 31: 23A

Takeda N, Mizuno J, Fujii K (1986) Abstract Second Symposium on Biology of Terrestrial Iso- pods Urbino Italy

Tcholakian RK, Eik-Nes KB (1969) Gen Comp

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

Endocrinol 12: 171-173

Tcholakian RK, Eik-Nes KB (1971) Gen Comp Endocrinol 17: 115-124

Teshima S, Kanazawa A (1970) Bull Jap Soc Sci Fish 36: 246-249

Teshima S, Kanazawa A (1971) Gen Comp Endoc- rinol 17: 152-157

Toullec J-Y, Van Wormhoudt A (1987) CR Acard Sci Paris 305 Ser III: 265-269

Turrigiano GG, Selverston AI (1989) J Neurosci 9: 2486-2501

Turrigiano G, Selverston A (1990) In “Frontiers in Crustacean Neurobiology: Advances in Life Scien- ces” Ed by K Wiese, W-D Krenz, J Tautz, H Reichert, B Mulloney, Birkhatiser Verlag, Basel, pp 407-415

Turrin MOA, Sawaya MI, Santos MCF, Veiga LV, Mantero F, Opocher G (1992) Comp Biochem Physiol 101A: 803-806

Valeggia C, Fernanadez-Duque E, Maldonado H (1989) Brain Res 481: 304-308

Van Beek E, De Loof A (1988) Comp Biochem Physiol 89A: 595-599

Van Deijnen JE, Vek F, Van Herp F (1985) Cell Tissue Res 240: 175-183

Van Herp F and Bellon-Humbert C (1982) CR Acad Sci Paris 295 Ser III: 97-102

Van Wormhoudt A, Favrel P, Guillaume J (1989) J Comp Physiol 159B: 269-273

Van Wormhoudt A, Fouchereau-Peron M (1987) Biochem Biophys Res Comm 148: 463-470 Watabe S, Hara M, Yago N (1985) Zool Sci (Tokyo) 2: 971

Yano I (1985) Aquaculture 47: 223-229

Yano I (1987) Aquaculture 61: 49 57

Young NH, Quinlan PT, Goad LJ (1992) Gen Comp Endocrinol 87: 300-311 Zukowska-Arendarczyk M (1981) Marine Biol 63: 241-247

4

Ha ol LER PATRAS A at

eS Atte

aan Fb oS, Renta, | et ie he tuted t

Bein Ne yiores aN

pe Met ay: Oh tid

Pelee! -“Oltaah

Anse tt a ee Vat, BT cs ivorzi al Ke ae banat £6.

GMA corte tie A Caieiahat

is UN vo tH

i i LjcAt vt tracey yar pe eee. grebiol th 9} A oR Vianney! soon het tie ais basne RUA CL ORY PE th. Apcbe Coe ne { poieddesuyousluht th atone oe SiR, cs ¥ oes clue AOtk viet} AP! CO eile), aie. a EARS ry BEF "Supe ‘ ; pit sei nes } Siwacis “4 ii Ne iF Rh pe Sh sire COR : be (BRP A deeedt ot i 4598 nay At! , oe =a aay! ioctl a stab nave, dale HS nepist hme. Ol ‘ a if Y * 7 it “yee We Vie ey enpa igh. foe a Fh: Geely a yal OHS, ne pigs wth : : ; : yt ; : 4 iQ Alaa oxnor ae EL agS e bse | oye, Fae ae et 7 ‘ctouateut towne SE Ml RO a . al sii 4 ore 1 é‘ Se Cs Tea AE es ‘ rgn() Kicshony P i th aighl cto hel te count qmnay nom ae +4052 cay eck ! eho a baer Lb onl i fis e aes J ¥ ¢ henna 4 as

nmgasg ts hiv Wes

at sent oerreya]

Tidy ue SH OT han wahht -

sae

sae ol ae WY, " evn ant ie

ee 2 Vsheas,

“at

Fes

hth a

AG | patel | i) bea 2 vin re

as UST) Es i ‘SH rh ys.

‘4 ie oe ap a“ be rrereey.¥ fiat. Lei RS J ye Ph ci

‘Lan te Tekjea ETS

; wenariy FD + im : a tet tue Fle 4 = -t . Ls, Liveatk 40s leer | Yartice, Pee'D York .dy Ve “hare & je tly * Unt (eet, Regal 4 ‘ ey era u at was (hfel il od Ge ei ah. magitag eA cieny Pil sey Cleves

ZOOLOGICAL SCIENCE 10: 31-37 (1993)

© 1993 Zoological Society of Japan

REVIEW

Loss of the Paternal Mitochondrion during Fertilization

CuHarLes C. Lampert! and Davip E. BatTraGLia2

'Department of Biology California State University Fullerton, CA 92634, and *Department of Obstetrics and Gynecology SJ-70 University of Washington Medical School Seattle Washington 98195, U.S.A.

INTRODUCTION

“If any apology is needed for presenting so strictly a morphological study of such an apparent- ly threadbare subject as the fertilization of the ovum, I might say that the impulse to make it came from an experimental study ....” [22].

Spermatozoa are small ceils specialized for the single function of delivering the haploid sperm nucleus to the egg and aiding in inserting this genome into the egg. As a specialized delivery vehicle, the sperm cell loses most of its original cytoplasm during spermatogenesis, leaving us with a basic means of locomotion (generally a tail with a functional axoneme), the nucleus and a means of generating ATP, the mitochondrion or several small mitochondria. Most sperm are divided into three functional regions: head with nucleus and acrosome, midpiece with mitochondria and tail with central axoneme [13, 15, for reviews].

Although the paternal mitochondrion generally makes no genetic or functional contribution to the zygote [5, 8], in most animals the sperm mitochon- drion or multiple mitochondria enter the egg along with the nucleus, centriole and tail during fertiliza- tion [23 for review]. This occurs even though most of the spermatid mitochondria are shed during spermiogenesis. Here we review the apparently rare cases where the sperm mitochondrion is eli- minated during fertilization. Hopefully this will foster additional interest in this area and additional cases will be uncovered. This process which often

Received September 16, 1992

involves mitochondrial sliding along the sperm tail, occurs sporadically through-out the animal king- dom in such evolutionarily distant organisms as polychaetes and ascidians.

Annelida

Lillie [22] appears to have been the first to call attention to loss of the paternal mitochondrion during the fertilization of animal eggs. He studied fertilization in the polychaete worm Nereis limbata and demonstrated that the sperm mitochondrion remains on the egg surface after sperm incorpora- tion into the egg. The sequence of fertilization in this species begins with sperm binding to the egg vitelline coat (VC) via attachment of the tip of the acrosome. The egg reacts to sperm contact within 2-3 min by releasing egg jelly from its cortical granules and forcing supernumerary sperm from the VC surface. Subsequently a fertilization cone forms at the site of contact with the fertilizing sperm. The fertilization cone disappears leaving the sperm on the the egg surface until 40-50 min after insemination when it is rapidly internalized by the egg. The midpiece with its mitochondrion remains on the egg surface (Fig. 1)

A similar process occurs in another polychaete, Platynereis megalops [14]. Here fertilization is internal following copulation and the eggs are extruded with sperm attached and the jelly form- ing. Penetration, as in Nereis requires over 30 minutes and results in the midpiece being left on the egg surface after the completion of sperm penetration. However, the large fertilization cone seen during Nereis fertilization is not formed in this

32 C. C. LAMBERT

AND D. E. BATTIAGLIA

Fic. 1.

Fertilization in the polychaete Nereis limbata (22) showing the sperm head and tail entering the egg and

leaving the mitochondrion on the surface of the vitelline coat.

species.

In all other known cases of fertilization the sperm mitochondrion enters the fertilization cone and is taken into the zygote cytoplasm [23].

Mollusca

During the fertilization of most molluscs the sperm mitochondrion enters the egg and may even persist in the egg cytoplasm [24]. The sperm mitochondrial genome seems to be inherited to a high degree in the bivalve Mytilus edulis [10, 35] but this is clearly not the general case in bivalves.

Chitons with their heavily sculptured egg hulls have evolved a unique way to fertilize the egg without forcing the wide midpiece through the VC [1]. Chiton sperm are typical primitive sperm except for the locations of the mitochondria where one lies lateral to the nucleus as in the ascidian sperm and the other encircles the base of the nucleus [1]. In Tonicella lineata the sperm passes into cupules in the hull which give the sperm access

to the underlying VC. Sperm-egg fusion occurs between the acrosomal membrane and an egg microvillus which penetrates into a pore in the VC. This composite tube formed from the fusion of the two membrane bound cylinders then serves as the connection through which the sperm chromatin mass is injected into the egg without even a nuclear envelope. This leaves the mitochondria and tail behind on the VC surface (Fig. 2).

Laternula limicola is a brackish water bivalve in which a golgi derived “temporary acrosome” moves from the sperm tip to the midpiece during spermatogenesis, finally taking up a position post- erior to the mitochondrion in the midpiece [17]. During fertilization, the sperm first encounters a 24 wm thick jelly coat, then penetrates a 7 um thick metachromatic layer before reaching the VC. The sperm then passes through the VC to the egg interior leaving the midpiece with its mitochon- drion and temporary acrosome on the outside (Fig. 3). The temporary acrosome remains unchanged

Male Mitochondrial Loss 33

througout the entire process [17].

acrosome similar to that of Laternula [16] but we know of no studies of fertilization in this species.

Fic. 2. Fertilization in the chiton Tonicella lineata (24). In this mollusc the sperm acrosome fuses with an egg surface microvillus and the nucleus is injected through the composite tube leaving the tail and 3 mitochondria on the egg surface.

Another bivalve, Lyonsia ventricosa also has a temporary

sa ned Se ca ape se

Fic. 3. Fertilization in the bivalve mollusc Laternula limicola (10). Here the mitochondrion (M) remains embedded in the vitelline coat with the “temporary acrosome” (TA) as the remainder of the sperm enters the egg cytoplasm.

Bryozoa

Both external and internal fertilization can be found in the bryozoans. In several brooded species fertilization occurs within the ovary during oogenesis. Markus [25] reported that only the nucleus enters the egg in several species, a finding confirmed by Matawari [26]. In a more detailed study, Temkin [31] found that sperm of De- ndrobeania lichenoides, Hippodiplosia insculpta and Tricellaria gracilis leave both the midpiece and tail on the VC surface at fertilization; only the sperm head actually enters the oocyte. In another study [6], both the sperm nucleus and midpiece can be seen within the oocyte of Chartella papyracea, a species with precocious intra-ovarian fertilization. Thus there may be interspecific variation of

Fic. 4. Sperm penetration in the asteroid Pisaster ochraceus showing the usual pattern with the mitochondrion

(arrow) entering the fertilization cone and egg.

34 C. C. LAMBERT AND D. E. BATTAGLIA

whether the paternal mitochondrion is discarded or incorporated during the fertilization of bryo- zoans [31].

Echinodermata

In most echinoderms the sperm mitochondrion experiences a change in orientation at the time of the acrosome reaction [2, 29] but is generally incorporated into the egg during fertilization. Ex- ceptions to this have been seen in a holothurian and an asteroid which seem to demonstrate facultative rather than obligatory sperm mito- chondrial loss during fertilization. Colwin and Colwin [4] report that while the sperm mitochon- drion usually enters the egg of the holothurian Thyone briareus, occasionally the midpiece re- mains outside as the sperm is incorporated.

In the asteroid Pisaster ochraceus careful observation reveals that there are three patterns of

Fic. 5.

mitochondrial behavior during fertilization to be found among different males in the populations around San Juan Island, Washington [Battaglia, unpublished observations). In the most common pattern the head and midpiece remain intact and pass through the fertilization cone together (Fig. 4). Normally 9-12 min elapse from the time of initial sperm-egg fusion until the entire flagellum enters. The second pattern is seen with the sperm from about 30% of the males where most of their sperm enter the egg jelly, undergo the acrosome reaction, and enter the fertilization cone, but the midpiece becomes separated from the head and remains stationary in the jelly as the sperm moves past this fixed point (Fig. 5). An intermediate pattern is seen in a small number of cases where the midpiece becomes separated from the head and lags behind but is eventually captured by the fertilization cone and incorporated with the rest of

Sperm penetration in the asteroid Pisaster ochraceus showing the case where the mitochondrion (arrow)

remains on the surface of the vitelline coat. This pattern was seen in 30% of the males.

Fic. 6. Sperm penetration in the asteroid Pisaster ochraceus showing the case where the mitochondrion (arrow) separates from the nucleus but is eventually incorporated in the fertilization cone and taken into the egg

cytoplasm.

Male Mitochondrial Loss 35

the sperm (Fig. 6). These patterns were found in the sperm from different males and independent of the source of oocytes. Since these observations were made at the peak of the breeding season and led to perfectly normal development we conclude that they reflect normal modes of fertilization rather than testicular degeneration or some other pathological condition.

Urochordata

The fate of the sperm mitochondrion is now known for several members of the invertebrate complex commonly called the lower chordates. In the cephalochordate Branchiostoma the sperm has a well developed acrosome and a single mitochon- drion in the midpiece. This mitochondrion clearly enters the egg at fertilization [11]. Although the sperm of the appendicularian tunicates is quite small it also has a well developed acrosome and as in cephalochordates the sperm mitochondrion en- ters the egg at fertilization (Linda Holland, 1992, personal communication).

Ascidian sperm lack a conspicuous acrosome and midpiece. Instead of a midpiece these sperm have the mitochondrion next to the nucleus in the head [20, for review of sperm structure]. The first detailed account of fertilization in ascidians re- ported that the mitochondrion entered the egg along with the nucleus and tail [27] but subsequent studies [7, 32, 33] report that sperm between the VC surface and egg surface lack the mitochon- drion. Richard Cloney (personal communication, 1975 and 3) first observed the mitochondrion of Molgula occidentalis to remain on the VC surface as the sperm moved past this fixed point. This pattern of fertilization was subsequently seen in Ascidia ceratodes and other ascidians [9, 18, 19]. Figure 7 shows an Ascidia ceratodes sperm head within the inner layer of the VC surface while the mitochondrion remains on the outer surface of this layer between the follicle cells. Mitochondrial sliding can be induced in vitro by a number of treatments including alkaline SW and low Na* SW [18, 19]. Mitochondrial migration down the tail results in the sperm moving past this point when the sperm mitochondrion sticks to the cover slip in vitro. This sliding occurs at the rate of 6.7 u~m/sec in vitro but 3.1 ~m/sec in vivo into the egg and

Fic. 7. Sperm penetration in the ascidian Ascidia cera- todes. Here the mitochondrion (arrow) remains on the vitelline coat surface between the follicle cells as the remainder of the sperm penetrates through the egg envelopes to reach the egg surface.

appears to be a rate limiting step in fertilization (Yim, Abadi and Lambert, unpublished data). Mitochondrial sliding is thought to drive the sperm through the VC and perivitelline space and into the egg [15, 20 for reviews]. In addition to the mechanical forces generated by mitochondrial slid- ing, sperm surface proteases are also involved [28 and Koch, Norton and Lambert, unpublished data].

Conclusions

The interplay between the physical and biochemi- cal capabilities of the sperm and egg is instructive as to how the remarkable process of fertilization is accomplished. The management of sperm mitochondria during fertilization is rarely consi- dered in studies of sperm-egg interaction. Gener- ally it may be assumed that all of the mitochondria that survive in the zygote are of maternal origin. Thus, the prevention of paternal mitochondrial replication after fertilization may be a universal feature of all species. The importance of disposing of paternal mitochondria during fertilization is speculative; however, the mechanisms underlying the success of this phenomenon are interesting. Recent evidence suggests that mitochondrial DNA has an extremely high rate of mutation which may be exacerbated by free radical formation [34]. It is

36 C. C. LAMBERT AND D. E. BATTAGLIA

clear that the sperm mitochondrion is a potent source of free radicals due to its extremely high metabolic activity. Thus, oxidative damage to the unprotected mitochondrial genome may be exten- sive in mature, active sperm. It is conceivable that this could result in inefficient/abnormal zygote mitochondria if the replication of these damaged organelles were allowed to occur after fertilization [12]. Most animal eggs appear to deal with this scenario by inactivating and destroying the sperm mitochondria after they have been incorporated into the egg cytosol. Fertilization in the mammal is a good example of a class of animals where active destruction of sperm mitochondria is mediated by the egg [30]. The complexity of the mammalian midpiece precludes its expulsion during sperm incorporation. It contains many mitochondria which rapidly swell and are destroyed upon entry into the egg cytosol. Thus, all mitochondria in the mammalian embryo are considered to be of mater- nal origin (although recent evidence suggests that 0.1% of human sperm mitochondria are replicated in the fertilized egg) [34]. However, the mechanisms attendant to the in- activation or destruction of mitochondria within the egg may not exist in all species, particularly in animals where mitochondrial expulsion is observed as described in this review. In these species it is the physical elimination of this organelle that prevents its incorporation into the egg, thereby avoiding any replication of the paternal mitochondrial genome. The results are the same, but the mechanisms that accomplish these phenomena may be different. For example, the ascidian sperm appear to have a complex mechanism that utilizes cytoskeletal elements in the sperm to actively expel the mitochondrial compartment [21]. In contrast, the expulsion of sperm mitochondria in some starfish and polychaetes utilizes a more pas- sive approach; the physical constraints imposed by the egg extracellular matrix aid in holding the mitochondrial compartment in place while the rest of the sperm slides past. One unifying principle to all of these species is that so-called “primitive” sperm are involved, thereby suggesting that mitochondrial expulsion can only be accomplished in species possessing a simple sperm architecture.

27

REFERENCES

Buckland-Nicks J, Ross K, Chia FS (1988) Gamete Research 21: 199-212

Christen R, Schackmann RW, Shapiro BM, (1982) J Biol Chem 257: 14881-14890

Cloney RR (1989) Urochordata-Ascidiacea. In “Reproductive Biology of Invertebrates Vol IV Part B” Ed by Adiyodi and Adiyodi, Oxford and IBH, New Delhi, pp 391-451

Colwin LH, Colwin AL (1956) Biol Bull 110: 243- 257

Dawid IB, Blackler AW (1972) Dev Biol 29: 152- 161

Dryrynda PEJ, King PE (1983) J Zool Lond 200: 471-492

Ezell SD (1963) Exp Cell Res 30: 615-617

Giles RE, Blanc H, Cann HM, Wallace DC (1980) Proc Nat Acad Sci USA 77: 6715-6719 Godknecht, A, Honegger TG (1991) Dev Biol 143: 398-407

Hoeh WR, Blakley KH, Brown WM (1991) Science 251: 1488-1490

Holland LZ, Holland ND (1992) Biol Bull 182: 77- 97

Hurst LD, Hamilton WD (1992) Proc Roy Soc Lond B 247: 189-194

Jamieson BG (1991) Fish evolution and systematics: evidence from spermatozoa Cambidge University Press, Cambridge

Just EE (1915) J Morph 26: 217-233

Koch RA, Lambert CC (1990) J EM Techniques 16: 115-154

Kubo M, Ishikawa M (1978) J Submicr Cytol 10: 411-421

Kubo M, Ishikawa M, Numakunai T (1979) Proto- plasma 100: 73-83

Lambert CC (1989) J Exp Zool 249: 308-315 Lambert CC, Epel D (1979) Dev Biol 69: 296-304 Lambert CC, Koch RA (1988) Develop Growth Diff 30: 325-336

Lambert CC, Lambert G (1984) Dev Biol 106: 307-— 314

Lillie FR (1912) J Exp Zool 12: 413-459

Longo FJ (1987) .“Fertilization” Chapman and Hall, New York

Longo FJ, Anderson E (1969) J Exp Zool 172: 97- 119

Markus E (1938) Bryozarios marinhos brasieleros II. Univ Sao Paulo Fac Filos Cien Let, Bol Zool 2: 1-196

Matawari S (1952) Misc Rep Res Inst Nat Resour (Tokyo) 28: 17-27

Meves F (1913) Archiv Mikroschkop Anat 82: 215- 260

28

29

30 31

Male Mitochondrial Loss 37

Pinto MR, Hoshi M, Marino R, Amoroso A, De Santis R (1990) Molec Repro Dev 26: 319-323 Schroeder TE, Christen R (1982) Exp Cell Res 140: 363-371

Szollosi DG (1965) J Exp Zool 159: 367-378 Temkin M (1991) Fertilization and early develop- ment of gymnolaemate bryozoans Ph D Dissertation

32

University of Southern California

Ursprung H, Schabtach E (1965) J Exp Zool 159: 379-384

Villa L (1977) Acta Embryol Exp 2: 179-193 Wallace DC (1992) Science 256: 628-632

Zouros E, Freeman, KR, Ball AO, Pogson, GH (1992) Nature 359: 412-414

rm ‘

mai ST nana sea castag one a gl

rt

i aei st TAL! ws yt

ess _ i‘. ; Led age 9 uy ay na otal f ey von

ye a * tes al Bf. 1 i: need ie ;

, * a CLA Lett wn. i

cen oe meer § y “ahr eee ah ie Ln ie ess a al ee fi COUN eS Nephi atieets re Soria wh sate 2 4 YSN Pe eee ry ricpinnuENg, Wats " : rans 4 7 SOSH, Sey eas ~ ed sd rat! : eectned ; mty (ati i ' { ;

ae eK hs

Hires prank bo ytcrawighstinni . ') p RSEMGIE Vas i, ‘

or i bp Aaipodt e ~ Reeth. pn What" a win 1A Cotitin, ALES

6

mn YR! +

« Lee eial pH. Hi eiser Aww Lg

«At

& Devriede Ph Sey Sate ibe na Due Si FLD) tzp al

\.. Fe a Ho Het ey Be = Pd Wi had USA ee "Glee h onan ait

wie, Whitey KIM.

i4 43 » +h ie Heed Lo. boolean ™% tracey 2B Hanmi wa Lom ty 247 ee > i bad tine é

s, he Hie sa if

herr Wei ee ri Y G26 peat >i ‘The Hitt a a ie a) K i a Lconhoee a 1 > ; Niiewa hee i” a i - i rabce Pe ap re . * , ' ss ra! (ity ieee y ' -tmtilaert if ua i ore i? ike ae ‘ ge i] Last Nee f aT (ayes 2 NAD yt Sh ( iy (ine (RT ¢ t Min Wie tet re on! : ol & t rene

Mien “Wy +

ZOOLOGICAL SCIENCE 10: 39-42 (1993)

© 1993 Zoological Society of Japan

Immunocytochemical Evidence for the Involvement of RFamide-like Peptide in the Neural Control of Cricket Accessory Gland

Koust YAsuyAMA!, BIN CHEN’ and TsuUNEO YAMAGUCHI2*

‘Department of Biology, Kawasaki Medical School, Kurashiki 701-01, and Department of Biology, Faculty of Science, Okayama University, Okayama 700, Japan

ABSTRACT— Using anti-RFamide antiserum, innervation of the accessory gland of the male cricket was investigated. RFamide-like immunoreactive fibers were found over the musculature of this organ, i.e., in the thin muscle winding around each glandular tubule as well as in the thick muscle surrounding the opening of each tubule to the ejaculatory duct. This result suggests that in addition to the previously reported proctolinergic system, another peptidergic system mediated by an RFamide-like substance is involved in regulation or modulation of contractile activity of the accessory gland.

INTRODUCTION

Spermatophore formation in the male cricket requires the rapid release of ingredients produced by the accessory gland at a certain period in the mating cycle [7]. This implies that contraction of the muscles of the accessory gland to extrude the secretions may be directly triggered by neural signals. In our previous morphological and elec- trophysiological investigations it was shown that the accessory gland is innervated by two types of neurons originating in the terminal abdominal ganglion [5, 11, 12]. The first type is the dorsal unpaired median neuron (DUMR7 neuron), which induces contraction of the accessory gland [5] and has a proctolin-like immunoreactivity [12]. The second type is the ordinary paired neuron (LC neuron), which has an inhibitory effect on the contraction of the accessory gland, and is likely to be serotonergic [5].

Recent investigations indicate FMRFamide im- munoreactive fibers innervating the midgut of Drosophila \arvae [8] and the dual peptidergic innervation of the blowfly hindgut by proctolin, FMRFamide and related peptides [2]. In the locust, FMRFamide and related peptides were

Accepted September 22, 1992 Received September 2, 1992 ** To whom reprint requests should be sent:

shown to have modulatory effects on the oviduct [6], on the heart [3], and on the foregut [1]. In the isolated locust foregut, FMRFamide was reported to potentiate relaxation induced by serotonin and to inhibit the contraction caused by proctolin [1,9].

The present report describes the presence of FMRFamide immunoreactive fibers innervating the glandular tubules of the cricket accessory gland, suggesting that an RFamide-like peptide has an important role in regulation or modulation of contractile activity of the accessory gland.

MATERIALS AND METHODS

Adult (Gryllus bimaculatus) reared in the laboratory, were used for the experi- ments. The accessory glands were fixed for 6-12 hr in 4% paraformaldehyde in 0.1 M phosphate buffer at 4°C. Rabbit anti-FMRFamide antiserum (Cambridge Research Biochemical) used for this experiment is specific to the C-terminal (RFa- mide), and positive reaction to this antiserum demonstrates the existence of RFamide-like pep- tides. Antiserum was diluted 1 : 2000 in phosphate buffered saline [5] (PBS, pH 7.4) with 0.5% Triton X-100 and 0.5% bovine serum albumin (BSA).

Light microscopic immunocytochemistry was employed using the avidin-biotin-peroxidase com- plex (ABC, Vector Lab.) method as described by Yasuyama et al [12]. Paraffin sections were incu-

male crickets

40 K. YASUYAMA, B. CHEN AND T. YAMAGUCHI

bated in primary antiserum for 48 hr at 4°C. The secondary antibody, a biotinylated goat anti-rabbit IgG, was applied to the sections at a 1 :200 dilution in PBS for 1 hr at room temperature. Incubation in the PBS-diluted ABC reagent (1:100) was for 1 hr at room temperature. Each incubation was followed by three 10min washes in PBS. The peroxidase activity was visualized by treatment with 0.05M Tris buffer (pH=7.2) containing 0.025% 3,3’-diaminobenzidine tetrahydrochloride (Sigma) and 0.005% HO. The sections were dehydrated and mounted in Canada Balsam. For whole mounts, the accessory glands were incu- bated in the primary antiserum for 72 hr at 4°C. Each of the solutions of the secondary antibody (1 :200) and the ABC reagent (1 : 100) was applied for 8-10 hr at 4°C. Each incubation was followed by 12-24 hr wash in PBS. After visualization of the peroxidase activity, the preparations were de- hydrated and mounted.

For electron microscopic immunocytochemistry, the post-embedding protein A-gold method was employed as described by Yasuyama et al [12]. Ultrathin sections were incubated in primary anti- serum for 48 hr at 4°C. The protein A-gold com- plexes (Janssen Pharmaceutical; particle size 15 nm) were applied at a concentration of 1:40 in PBS with 1% BSA. The sections were stained with uranyl acetate and lead citrate, and were examined with a Hitachi H 500 electron microscope.

The specificity of the immunoreactivity was tested by preincubating the anti-RFamide anti- serum with synthetic FMRFamide (Funakoshi). The test peptide was added at 10-*M concentra- tion to the working dilution (1 : 2000) of antiserum, and the solution was incubated for 24 hr at 4°C. Immunoreactivity was eliminated by such treat- ment.

RESULTS AND DISCUSSION

The accessory gland of the male cricket is com- prised of more than 600 slender tubules arising from the lobed anterior end of the ejaculatory duct. Single-layered thin muscles wind around most of each glandular tubule’s length, and multi- layered thick muscles surround the opening of each tubule to the ejaculatory duct [11]. On the

basis of the arrangement of these two types of muscles, it is likely that the antagonistic action of muscles, i.e. the contraction of single-layered mus- cles and relaxation of multi-layered muscles, causes the release of the secretions into the ejaculatory duct. RFamide-like immunoreactive fibers were found in both single- and multi-layered muscles (Figs.1 and 2). In the single-layered muscles associated with a glandular tubule, the immunoreactive fibers had many varicosities and ran rather straightly along the tubule’s long axis to the distal end. Projection of short thin branches from the thick fibers were observed at rather regular intervals (Fig. 1). In the multi-layered muscles surrounding the opening of each tubule, the RFa-like immunoreactive fibers ran tortuously among the muscle fibers. Figure 2 shows the immunoreactive fibers circularly arranged around the glandular tubule.

Electron microscopic immunocytochemistry rev- ealed the widely distributed RFa-like immunoreac- tive axons in the accessory gland muscles. RFa- mide-like immunoreactivity was restricted to large electron-dense granular vesicles with a mean dia- meter of 141.9 nm (n=56) contained in the axons (Fig. 3). This immunoreactivity was also found in granular vesicles (141.5 nm in mean diameter, n= 86) accumulated in the nerve branch (Br3) after ligation ({11]; data not shown). Our earlier inves- tigations have revealed that Br3 carries innerva- tion from the terminal abdominal ganglion to the accessory gland, and ligation of Br3 causes a heavy accumulation of large electron-dense granular vesicles [11, 12]. It is, therefore, suggested that RFa-like immunoreactive granular vesicles found in the accessory gland musculature are transported through the Br3.

Our recent investigation using the protein A- gold method combined with retrograde HRP label- ing showed that proctolin-like immunoreactivity is present in the DUMR7 neurons innervating the accessory gland musculature [12]. The RFa-like immunoreactive fibers demonstrated in this experi- ment are probably derived from neurons which are distinct from the DUMR7 neurons. This conclu- sion is based on the following facts. First, the size of the vesicles with proctolin-like immunoreacti- vity found in the axons associated with the acces-

RFamide-like Peptide in Accessory Gland 41

Fic. 1. RFamide-like immunoreactive fibers in a wholemount preparation of a glandular tubule of the accessory gland. Note projection of short thin branches (arrows) from the thick immunoreactive fibers. Scale bar, 20 um. Fic. 2. RFamide-like immunoreactive fibers (arrows) within the proximal part of the accessory gland tubule, where multi-layered muscle fibers are located. Arrowheads indicate the lumen of the glandular tubule. Scale bar, 20

pm.

Fic. 3. An axon (ax) containing RFamide-like immunoreactive vesicles and running within the single-layered muscle (mf) which envelops the accessory gland tubule. Immunogold labeling was restricted to the vesicles. bm, basement membrane; gc, glandular cell. Scale bar, 0.5 ~m.

sory gland musculature was 100-115 nm in dia- meter, while the RFamide-like immunoreactivity was restricted to the vesicles with a diameter of about 140nm. Second, the anti-RFamide anti- serum did not stain the DUMR7 neurons, but did stain the bilaterally symmetrical neurons in the terminal abdominal ganglion [10]. These suggest that a separate peptidergic innervation is supplying the accessory gland, and not only proctolin but also FMRFamide-like peptide have important roles in regulation or modulation of neuro- or myogenic activities.

In the locust, Evans and Cournil have described a pair of FLRF-like immunoreactive neurons, which are located in the suboesophageal ganglion and send their axons through the ventral nerve cord to the terminal abdominal ganglion and fur- ther, in the male locust, to the genital nerve arising

from this ganglion [4]. In the locust there is, also, a possibility that the RFa-like immunoreactivity fibers found in the accessory gland musculature derive from the neurons located in the higher ganglion. Furthermore, a possibility that the RFa- mide-like peptide is contained in the LC neurons cannot be eliminated, since anti-RFamide anti- serum revealed the lateral cluster of neurons around the site where the LC neurons were located in a cluster (Yasuyama, unpublished observation). Further experiments are being carried out both to identify the RFa-like immunoreactive neurons in- nervating the accessory gland and to elucidate their functional roles in regulation of the mechan- ical activity of the accessory gland.

42 K. YASUYAMA, B. CHEN AND T. YAMAGUCHI

cally and neurogenically evoked contractions. J In- sect Physiol 35: 251-264

6 Lange AB, Orchard I, Te Brugge VA (1991) Evi- dence for the involvement of a SchistoFLRF-amide- like peptide in the neural control of locust oviduct. J

ACKNOWLEDGMENTS

This work was supported in part by a Grant-in-Aid from the Ministry of Education, Science and Culture of

Kimura T, Yasuyama K, Yamaguchi T (1089) Proc- tolinergic innervation of the accessory gland in male cricket (Gryllus bimaculatus): detection of proctolin and some pharmacological properties of myogeni-

Japan. Comp Physiol A 168: 383-391 7 Loher, W. (1974) Circadian control of sperma- tophore formation in the cricket Teleogryllus com- REFERENCES modus Walker. J. Insect Physiol 20: 1155-1172.

Banner SE, Osborne RH (1989) Modulation of 8 White K, Hurteau T, Punsal P (1986) Neuropep- 5-HT and proctolin receptors by FMRFamide in the tide-FMRFamide-immunoreactivity in Drosophila: foregut of the locust Schistocerca gregaria. J Insect development and distribution. J Comp Neurol 247: Physiol 11: 887-892 430-438 Cantera R, Nassel DR (1991) Dual peptidergic 9 Wood SJ, Osborne RH, Casttell KJ, Banner ST innervation of the blowfly hindgut: a light- electron (1990) FMRFamide and related peptides modulate microscopic study of FMRFamide and proctolin the actions of 5-hydroxytryptamine and proctolin on immunoreactive fibers. Comp Biochem Physiol 99C: the foregut of the locust, Schistocerca gregaria. 517-5251 Biochem Soc Trans 18: 384-385 Cuthbert BA, Evans PD (1989) A comparison of | 10 Yasuyama K, Chen B, Yamaguchi T (1992) Comp the effects of FMRFamide-like peptides on locust Physiol Biochem 9: 224 heart and skeletal muscle. J exp Biol 144: 395-415 11 Yasuyama K, Kimura T, Yamaguchi T (1988) Mus- Evans PD, Cournil I (1990) Co-localization of culature and innervation of the internal reproductive FLRF- and vasopressin-like immunoreactivity in a organs in the male cricket, with special reference to single pair of sexually dimorphic neurones in the the projection of unpaired median neurons of the nervous system of the locust. J Comp Neurol 292: terminal abdominal ganglion. Zool Sci 5: 767-780 331-348 12 Yasuyama K, Kimura T, Yamaguchi T (1992) Proc-

tolin-like immunoreactivity in the dorsal unpaired median neurons innervating the accessory gland of the male cricket, Gryllus bimaculatus. Zool Sci 9: 53-64

ZOOLOGICAL SCIENCE 10: 43-51 (1993)

Developmental and Pharmacological Studies of Acetylcholinesterase- defective Mutants of Caenorhabditis elegans

Y. Kamiya, S. Harapa, S. Oxoyama!, H. YAMAMOTO

and R. Hosono?

Department of Biochemistry, and ‘Department of Anatomy, School of Medicine, Kanazawa University, Kanazawa, Ishikawa 920, Japan

ABSTRACT—Three genes (ace-1, ace-2, and ace-3), which code for acetylcholinesterase (class A, B, and C) have been identified in the nematode Caenorhabditis elegans. Here we investigate the developmental and pharmacological properties of mutants retaining only one of distinct classes of acetylcholinesterase. Both class A and B acetylcholinesterases share about one-half of the total enzyme activity throughout development and maximum specific activity at the first larval stage, and are sufficient for maintaining normal acetylcholine levels. Although class B acetylcholinesterase is distributed at the head and the body region as seen in the wild type, the class A enzyme is biased towards the head region. Class C acetylcholinesterase occupies only a few percent of the total activity, is mainly distributed in the body, shows maximum specific activity at the late larval stage and is clearly different in pharmacological response from class A and B acetylcholinesterases. We propose that class

© 1993 Zoological Society of Japan

B is the major acetylcholinesterase, and that classes A and C are supplementary.

INTRODUCTION

Acetylcholinesterases (AChE) have been exten- sively studied in various organisms (for reviews, see [5, 13]). It has been proposed that the enzyme has roles additional to the hydrolysis of acetylcho- line (ACh), such as the regulation of membrane excitability, permeability, general metabolism or the inactivation of neuropeptides. However, many problems remain to be solved, for example, why the enzyme is present in neurons which has neither choline acetyltransferase, the synthetic enzyme for ACh, nor any identified cholinergic input, and in neurons, which use transmitters other than ACh. These findings raise the question of what role AChE would be playing in noncholinergic cells.

Caenorhabditis elegans is a desirable organism for such studies [15]. Because genetic and bio- chemical properties of AChE are extensively stud- ied [3, 8, 10]. In C. elegans, three genes relating to AChE activity have been identified: ace-1, ace-2

Accepted September 24, 1992 Received August 24, 1992 2 To whom correspondence should be addressed.

and ace-3. These three classes of AChE (A, B, and C) corresponding to the three ace-1, ace-2, and ace-3 genes have been characterized biochemical- ly, molecularly and genetically [9, 12]. Mutations of any of the three ace genes do not result in any visible defects. In the ace-1 ace-2 double mutant, movement is abnormal and the ace-I ace-2 ace-3 triple mutant is lethal at the post-embryonic stage [3, 8, 10]. Therefore, any one of the three ace genes is sufficient for survival, and ace-1 or ace-2 are necessary for normal movement.

Although much is known about the kinetic prop- erties of the three classes of AChE, the develop- mental or pharmacological differences among them are not understood. To elucidate the func- tional role of each class of AChE, we studied the developmental changes in AChE activity in rela- tion to enzyme localization and ACh levels. Sever- al organophosphates inhibit C. elegans AChE activity but the correspondence between the extent of the enzyme inhibition and the fatal effect is not always coincident [9]. The relationship between enzyme inhibition and pharmacological action has not yet been systematically studied. An orga- nophosphate, trichlorfon {(2,2,2-trichloro-1-hy-

ai

44 Y. Kamiya, S. Harapa et al.

droxyethyl)-phosphonic acid dimethyl ester}, is a potent nematocide, which probably functions by inhibiting AChE. We studied the response to trichlorfon of mutants defective in two of three ace genes. Our experiments indicate that class A and class B AChE have similar properties but class C AChE is greatly different from the two classes AChE in expression, distribution and response to organophosphates. In addition, we describe that class B AChE is indistinguishable from the wild type in distribution and the staining intensity, suggesting that class B is the major AChE of the three.

MATERIALS AND METHODS

Nematode strains. The following genes and mutations were provided from J.Rand (Oklahoma Medical Research Foundation, USA): ace-I (p1000) X, ace-2 (g72) I, ace-3 (dc2) Il, ace-1 (p1000) ace-2 (g72), ace-2 (g72) ace-3 (dc2), ace-3 (dc2) ace-I (p1000). C. elegans was cultured on NGM agar as described by Brenner [2], except that in some cases, a higher concentration of bacto- peptone (20 mg/ml) was added to the medium.

Histochemical staining of AChE. AChE in whole mounts of nematodes was stained by the modified method developed by Karnovsky and Roots [11] and by Culotti et al. [3]. Nematodes, permeabilized with 95 % acetone and extensively washed, were suspended in a mixture of 10 mg acetylthiocholine iodide, 65 ml 100mM malate buffer(pH 6.0), 0.5 ml 100 mM sodium citrate, 1 ml 30mM CuSO,, 1 ml H,O and 1ml 5mM potassium ferrocyanide, and were shaken at 37°C for 30 min. The reaction was stopped by diluting and washing. A DAB-nickel solution(3,5- diaminobenzoic acid 4 mg, ammonium nickel 60 mg, and 10ml 10mM Tris(pH7.6)) containing 0.003% HzO, were added and shaken for 3 min at room temperature to intensify the staining. The DAB-nickel was removed and the slides were washed several times with 10 mM Tris (pH 7.6).

Acetylcholinesterase assays. Organisms were suspended in two volumes of 50 mM Tris pH 7.6 and 0.03% Triton X-100 and homogenized in

liquid nitrogen. The homogenates were mixed with class specific assay mixtures [10] as follows. Wild-type AChE was assayed in 95 mM sodium phosphate buffer (pH 8.0), 0.32 mM 5-dithiobis-2- nitrobenzoic acid and 0.47 mM acetylthiocholine iodide. In addition to this assay mixture, 0.13% deoxycholate for assay of class A AChE and 0.03% Triton X-100 for class B AChE, and 0.03% Triton X-100, 0.013% deoxycholate and 2.5 uM neostigmine bromide for class C AChE were added. The final reaction volume was 1585 yl. The reaction was started by adding the homogen- ate at 25°C and the absorbance at 412 nm was read [4]. Homogenates from the ace-J ace-2 mutants were added to a final concentration of 0.3 mg/ml and from the remaining strains, of 0.03 mg/ml.

Acetylcholine assay. Acetylcholine was assa- yed by the enzymatic conversion of ACh to [**P] choline in the presence of [y--*P] ATP, acetylcho- linesterase and choline kinase [14].

Assay of paralysis and its recovery. Worms were suspended in M9 buffer containing trichlor- fon. Paralysis was followed by measuring the frequency of propagating waves along their bodies as described elsewhere [6]. For recovery, com- pletely paralyzed worms were washed with M9 buffer and resuspended in the buffer.

RESULTS

Developmental changes in AChE activity

Changes in the activity of each class AChE with development were compared (Fig. 1). Class A and B AChE activities were maximum at the first larval (L1) stage and the activity decreased gradually with the progress of development. Class C AChE occupied less than 2% of the total AChE through- out development and reached the maximum activ- ity at the later larval stage. Thus, the synthesis of the class C enzyme might be differently regulated from that of the other two.

Developmental changes in AChE distribution

To determine a more detailed developmental pattern of AChE expression, organisms were

Acetylcholinesterase in C. elegans 45

AChE activity (nmotes/min/mg protein)

0 10 20m 0 40 50 60 70 hatching Hours Embryo | L1} jose Th Meesy tee [ener Adult

Fic. 1. Changes in AChE activities throughout development. Organisms were homogenized and suspended in the class-specific AChE mixtures [8]. AChE activity was measured at 25°C by the thiocholine method [11]. The embryonic stage is arbitrarily plotted. The mean values of three assays are presented with bar for standard deviation. ace-2 (g72) ace3 (dc2) (&), ace-3 (dc2) ace-I (p1000) (A), ace-1 (p1000) ace-2 (g72) (@).

Fic. 2. AChE staining of the second stage larvae. Wild type (A), ace-2 (g72) ace-3 (dc2) (B), ace-3 (dc2) ace-1 (p1000) (C), ace-1 (p1000) ace-2 (g72) (D). Animals are oriented anterior side to the left and the ventral side on

top. Bar=100 pm.

46 Y. Kamiya, S. Haraba et al.

stained with acetylthiocholine by the method of Karnovsky and Roots [11]. Culloti et al. [3], who pioneered AChE staining with C. elegans, found that AChE was abundant in nearly all cells at the early larval stage, then was reduced gradually, but remained mainly in the area of the nerve ring, ventral ganglion, the pharyngeo-intestinal valve and the tail neural region as development progres- sed. We obtained the same results with the wild-type animal and compared the changes in the staining pattern of the three distinct classes of AChE with the ace double mutants from embryo to adult. Representative staining of the second stage larva is shown (Fig. 2). The AChE staining was first detectable on one of two-blastomeres and increased to stable, high levels in all blastomeres at around the 24-cell stage (data not shown). AChE expression in the three double mutants approx- imately coincided with that of wild type, though the extent of the staining differed. In the newly hatched wild-type larva, nearly all somatic cells and the nerve ring were intensely stained (Fig. 6 A-D). In the ace-3 ace-I1 double mutant, the staining pattern was similar to that of wild type. In the ace-2 ace-3 double mutant, the head region was stained as in wild type but the body region was less

intensely stained. In the ace-J ace-2 mutant, the non-neural cells were scarcely stained but the nerve ring and the ventral nerve cord were detect- ably stained. At the second larval stage in all strains, nonneuronal cells were weakly stained but that of neuronal cells remained intense (Fig. 2). The staining intensity of the positive regions be- tween ace mutants was similar to that of the first larval stage.

Developmental changes in ACh levels

At three developmental stages, the ACh levels of three mutants were compared with those of wild-type animals (Fig. 3). The ACh levels in the wild type were high at the larval stage, but de- creased to about 25% at the adult stage. ACh levels of the ace-2 ace-3 and the ace-3 ace-l mutants were similar to those of wild-type animals. In the ace-J ace-2 mutant, abnormal elevation of ACh was observed at the adult stage, though it was normal at the early larval stage. Therefore, class A and B AChE are sufficient but class C AChE is insufficient for maintaining normal ACh levels.

Inhibition of AChE by trichlorfon

Of the organophosphates tested, trichlorfon was

Acetylcholine (nmoles/mg protein )

0.2 0.4

L] wild type ace-2 ace-3 ace-3 ace-/ ace-! ace-2

LZ wild type ace-2 ace-3 ace-3 ace-/ ace-/ ace-2

wild type ace-/ ace-2

wild type ace-2 ace-3 ace-3 ace-/ ace-/ ace-2

M wild type ace-2 ace-3 ace-3 ace-/ ace-/ ace-2

Fic. 3.

0.6 0.8 1.0 1.2 1.4 1.6

Changes in ACh levels of wild type and ace double mutants. The indicated time at the adult stage means

hours after hatching. L1, first stage larva; L2, second stage larva; A, adult; M, asynchronous population. Standard deviations of mean values of three assays are indicated by vertical bars.

Acetylcholinesterase in C. elegans 47

ACHE activity ( %)

0) 107 10°

10°

Trichlorfon(M)

Fic. 4. (p1000) (4), and ace-1 (p1000) ace-2 (g72) (@).

the most prominent nematocide [7]. Although several organophosphates have been identified as AChE inhibitors, trichlorfon has not yet been tested [9, 12]. As shown in Figure 4, crude extracts of class A and B AChE were similarly inhibited but class C AChE was not, at the concentrations of trichlorfon tested.

Survival sensitivity of ace mutants to trichlorfon

Survival of the ace mutants was tested in the presence of trichlorfon (Table 1). Wild-type anim- als failed to propagate their progeny above 20 uM trichlorfon. Mutation of the ace-1 or ace-2 genes did not influence survival. However, mutation of the ace-3 gene brought about much higher sensitiv- ity to the drug. These results suggest that the trichlorfon sensitivity of wild-type animals is deter- mined by the class C AChE that is encoded by the ace-3 gene. This hypothesis was tested with mutants preserving one of the three ace genes. The sensitivity of the ace-1 ace-2 double mutant did not differ from that of the wild-type. However, both the ace-3 ace-1 and the ace-2 ace-3 mutants were hypersensitive to trichlorfon, supporting the above hypothesis. The ace-2 ace-3 mutants were slightly but reproducibly more sensitive to trichlor-

Inhibition of three classes AChE by trichlorfon. Wild type (©), ace-2 ( g72) ace-3 (dc2) (A), ace-3 (dc2) ace-1

TABLE 1. Sensitivity of ace mutants to trichlorfon Mutations pa Nie wild type 20

ace-1 (p1000) 20 ace-2 (g72) 20 ace-3 (dc2) 2 ace-2 (g72) ace-3 (dc2)

ace-3 (dc2) ace-1 (p1000) 2 ace-1 (p1000) ace-2 (g72) 40

To determine sensitivity, three fourth stage larvae were put onto NGM agar plates containing various concentrations of trichlorfon (0, 0.4, 1.0, 2.0, 4.0, 10, 20, 40 and 100 uM) and grown at 20°C. The highest concentration of trichlorfon is indicated in which worms, in triple experiments, produced F progeny within ten days.

fon than the ace-3 ace-1 mutants.

Paralytic sensitivity of ace mutants to trichlorfon

When C. elegans was exposed to trichlorfon, the animals gradually became immovable. The pat- tern of trichlorfon-induced paralysis differed in each ace double mutant (Fig. 5). Animals retain- ing only the ace-3 genes were slowly paralyzed as

48 Y. Kamiya, S. HARADA et al.

B

Frequency of sinusoidal waves ( % )

0 20 40 60 Time (min)

Fic. 5.

changes in the sinusoidal frequency were followed.

Time (min)

Paralysis by trichlorfon and recovery. (A) Newly hatched larvae were exposed to 1 mM trichlorfon and (B) Recovery from the paralysis was followed after an

extensive wash of the completely paralyzed worms. The sinusoidal activity of each strain means a relative value to

that prior to trichlorfon exposure.

The mean values of seven worms are presented with bar for standard

deviation. Wild type (©), ace-2 (g72) ace-3 (dc2) (A), ace-3 (dc2) ace-1 (p1000) (4), and ace-I (p1000) ace-2

(g72) (@).

were wild-type animals, but the ace-3 defective mutants were instantaneously paralyzed (Fig. 5A). Animals preserving either the ace-/ or ace-2 gene alone recovered from the trichlorfon-induced pa- ralysis in the same manner as wild-type animals, though a portion of the population did not. On the other hand, animals retaining only the ace-3 gene had difficulty recovering from paralysis (Fig. 5B). It is likely that the higher concentration of trichlor- fon induced a slow but irreversible inactivation of class C AChE, whereas the inactivation of class A and B AChE is completely reversible.

Effects of trichlorfon on the localization of AChE staining

Since trichlorfon principally inhibits AChE activity, the pattern of the AChE staining will diversely change during the paralytic and recovery states after trichlorfon exposure. As described above, newly hatched larvae of wild-type and double mutants of the ace genes showed character- istic AChE staining (Fig. 6 A-D). However, the staining was completely lost in any strain during paralysis as shown in the wild-type animal (Fig. 6E). Staining along the ventral nerve cord recov-

ered partially accompanying recovery from the paralysis, but staining at the head region did not recover (Fig. 6F-I). Thus movement recovery does not need complete restoration of the AChE staining.

Effects of trichlorfon treatment on ACh levels

Changes in ACh levels after trichlorfon expo- sure were also studied (Fig. 7). To prepare ani- mals in the same state of paralysis, each strain was incubated at different concentrations of trichlor- fon. At the paralyzed state, the ACh levels were slightly elevated in the ace-1 ace-3 and ace-3 ace-2 mutants as seen in the wild type, but greatly elevated in the ace-1 ace-2 mutant. Therefore, the extent of the paralysis did not apparently correlate with the accumulation of ACh. However, since the ACh level was assayed after washing the animals to remove trichlorfon, it is conceivable that the difference in ACh levels among respective mutants reflects the difference in reversibility of the inactivated enzyme.

Acetylcholinesterase in C. elegans 49

DISCUSSION

Yo understand the roles that the three classes of AChE may play, we studied the enzyme activity, histochemistry and ACh levels in mutant and wild-type C. elegans. Classes A and B AChE have several similar properties. That is, the develop- mental pattern of the expression of the two en- zymes is comparable (Fig. 1) and the inhibitory pattern by trichlorfon is similar (Fig. 4), although both enzymes seem to functionally overlap but are

Fic. 6. Changes in AChE staining caused by trichlor- fon. Newly hatched larvae became paralyzed, then recovered as in Figure 5. Animals before trichlorfon treatment (A-D). A completely paralyzed wild type (E). Animals recovered from the paralysis (F-I). Wild type (A, E, F), ace-2 (g72) ace-3 (dc2) (B, G), ace-3 (dc2) ace-I (p1000) (C, H), and ace-1 (p1000) ace-2 (g72) (D, 1). Animals are oriented anterior side to the left and the ventral side on bottom. Bar =200 pm.

not coincident. For example, mutants retaining class A AChE are more sensitive to trichlorfon than those containing only class B AChE. The staining pattern of AChE of whole mutant animals containing only class B was similar as that of the wild type. However, the staining of animals con- taining only class A AChE is faint in the body region. Classes A and B AChE differ in Km values [9, 12] and the extent of membrane association [9]. Solubilized class A AChE is as inhibited by tri- chlorfon as the crude enzyme but solubilized class

50 Y. Kamiya, S. HarapDa et al.

Time

(min)

1.0

: 0 wild type 60

= Ss 0 ace-2 ace-3 60

-3 = 0 ace-3 ace-/ 60

ace-/ ace-2 _° ce-/ ace-2 60

Acetylcholine (nmoles/mg protein )

2.0 3.0

Fic. 7. Changes in ACh levels by trichlorfon exposure. The newly hatched larvae of the wild type and the the ace-/ ace-2 mutant were exposed to 1 mM, and the ace-2 ace-3 and the ace-3 ace-J mutants to 0.1 mM trichlorfon, respectively. Under these conditions, four strains reached in approximately the same paralyzed state. At 60 min after exposure, animals were collected for the assay of ACh levels. Standard deviations of the mean values of

three assays are indicated by vertical bars.

AChE is less inhibited (unpublished results).

Class C AChE was found to be greatly different from classes A and B AChE in expression, dis- tribution and response to trichlorfon. The class C enzyme is resistant to trichlorfon but the strain retaining only the enzyme is hardly recovered from the trichlorfon-induced paralysis. The activity of class C AChE is one-fiftieth lower than the activity of A and B AChE, peaks at the early L4 stage, and distributed in the body rather than in the head region.

For normal locomotion of C. elegans, it is essen- tial that either the ace-/ or the ace-2 gene 1s intact, suggesting that a decrease in the total AChE activity rather than in the specific AChE, leads to locomotive abnormalities. In other words, the reduction of ACh hydrolysis and therefore, the accumulation of ACh, results in the movement abnormality. However, the extent of the elevation of the total ACh levels and the movement abnor- mality is not always correlated because, in the ace-1 ace-2 mutant, the movement is abnormal throughout development, but the ACh levels are normal at the early larval stage. Also, under the trichlorfon-induced paralyzed states with wild-type animals and ace double mutants, the extent of the ACh elevation of the ace-] ace-2 mutant greatly differs from those remaining. However, the rule conforms at the adult stage but not at the early

larval stage, because, in the ace-J ace-2 mutant, ACh levels are normal at the L1 and L2 stage, but are elevated at the adult stage.

The present findings can not fully explain why three types of AChE are present in C. elegans. We showed that slight elevation of ACh results in a movement abnormality, indicating that the conse- quent accumulation of ACh levels leads to lethal over-stimulation of some control and peripheral neurons. The presence of heterogeneous types of AChE, whose localization, expression and abund- ance differ, is useful for the prompt removal of stagnated ACh. Animals containing only class B AChE are indistinguishable from the wild type in the distribution and the staining intensity. Class A AChE normally presents at the anterior region of the body but is less abundant at the posterior region. Although class C AChE is much less abundant, it is detectable at the posterior region of the body. Therefore, class B AChE may play a major role and class A and C AChE a regionally supplementary role in the hydrolysis of ACh. Thus, the presence of different types of AChE may be useful for regulating the local concentration of ACh in the synapse. The electrical organ of Torpedo californica has also structurally distinct molecular forms of AChE [1]. They are differently distributed in the electric organ and suggested that they may play different roles on the hydrolysis of

Acetylcholinesterase in C. elegans 51

ACh.

ACKNOWLEDGMENTS

We wish to thank Dr. S.Kuno for suggestions and critical reading of the manuscript. We are grateful to Mrs. R. Kitamura and Mr. S. Matsudaira for technical support. Supported in part by funds for Medical Treat- ment of Elderly, School of Medicine Kanazawa Universi- ty, 1990 and Grant in Aid for Special Project Research from the Ministry of Education, Science and Culture of Japan.

REFERENCES

1 Abramson SN, Ellisman MH, Deerinck TJ, Maulet Y, Gentry MK, Doctor BP, TaylorP (1989) Differ- ences in structure of the molecular forms of acetyl- cholinesterase. J Cell Biol 108: 2301=2311

2 Brenner S (1974) The genetics of Caenorhabditis elegans. Genetics 77: 71-94

3 Culotti JG, von Ehrenstein G, Culotti MR, Russell RL (1981) A second class of acetylcholinesterase- deficient mutants of the nematode Caenorhabditis elegans. Genetics 97: 281-305

4 Ellman GL, Courtney KD, Andres JV, Feather- stone RM (1961) A new and rapid colorimetric determination of acetylcholinesterase activity. Biochem Pharmacol 7: 88—95

5 Greenfield S (1984) Acetylcholinesterase may have novel functions in the brain. Trends Neurosci. 7: 364-368

6

11

12

13

Hosono R (1978) Sinusoidal movement of Caenorhabditis elegans in liquid phase. Zool Mag 87: 191-195

Hosono R, Sassa T, Kuno S (1989) Spontaneous mutations of trichlorfon resistance in the nematode Caenorhabditis elegans. Zool Sci 6: 697-708 Johnson CD, Duckett JG, Culotti JG, Herman RK, Meneely PM, Russell RL (1981) An acetylcho- linesterase-deficient mutant of the nematode Caenorhabditis elegans. Genetics 97: 261-279 Johnson CD, Russell RL (1983) Multiple molecular forms of acetylcholinesterase in the nematode Caenorhabditis elegans. J Neurochem 41: 30-46 Johnson CD, Rand JB, Herman RK, Stern BD, Russell RL (1988) The acetylcholinesterase genes of C. elegans: Identification of a third gene (ace-3) and mosaic mapping of a synthetic lethal phenotype. Neuron 1: 165-173

Karnovsky MJ, Roots L (1964) A “direct-coloring” thiocholine method for cholinesterase. J Histochem Cytochem 12: 219-221

Kolson DL, Russell RL (1985) New acetylcho- linesterase-deficient mutants of the nematode Caenorhabditis elegans. J Neurogenet 2: 69-91 Lehmann J, Fibiger HC (1979) Acetylcholinesterase and the cholinergic neuron. Life Sci. 25: 1939-1947 McCamann RE, Stetzler J (1977) Radiochemical assay for ACh: modifications for sub-picomole measurements. J Neurochem 28: 669-671

Wood WB ed (1988) The nematode Caenorhabditis eleganspp. 1-667 Cold Spring Harbor Laboratory, New York

Bia (¥ r ths ies : . c , : fr Ye Quetta. detedaeanged ae 4 err CaP an E yatt too oade baal at Haale reittelbweowag n ia abies eo

; oie (Reel) wl inst: ii PIO ee Dp pr ; Fi vores ll At Sod ipaes wat (ner io

Cn iy a al ete

OF?~T RA 26 fae lao Eau nibAndwrsad at tabsrn ny ark i Fs tanta Ne AA nmenslt, havea’ ee Aaed CET ope RE) y edebtienata 2° ctivewe eh (1800) 2a idiagll-s de aepangios = teeta a ‘a i “eadt hw a ce i oie: wamiatacertl, ; sevovint suresemre A ort. 1s ‘SPbiterye? fey abry detndennetes a eaves ay int 5x. ‘ “i elon Vwherrelorn mepliiota (OM! Fon edt Tiiandat” aN nis sonking a phot Oy oftheir tia tea? Apel atysayer “er mi — inthe (owas. Earn, S} : be wri ORS. aes tt pal ibnad 2D ti adie i j ostetytooe afl (ane fiat seam Ro) sod LF har AP Feat She: My: ts * niche | wih ie on * yin Srl Sh See ithe Wa lent” ‘ Ps hag Le iwi saath pies he PA MES aby <i A RN ‘iY ne TERRY) a tee eee er" 0mm tourna er ; ; CLES

1 raheem ah %y ért a Nive? S84 Meesitit L Di omeadsane sca iysigts x! Jneitiets tigi, wet) (5804) CE ile if _ AE gorlewt «51 bint certo Si TB Pea Rtas Bena irri cara

[Rte Ox Dedede pecans Woh eal aomkides 9 A senqerplod ye Ot athy

: ; md relied : aa hee ert athe tM Saeki toothy baveniiats (et a na Ay ee | si oe Ch wT alk Aetn. (PECL A es dehineia guls hese ;

br 08 Epes sy ak: seiteatrk sisi

soooba yh | CTE) ) E gaeaih Pe Se eer ean ad

i <iiiniialtin Bight: Binet

joel yolk, apmapisomibom A Sh -7ot vane Shed SE kw ae | ere viekeor “ (raat ; ( Desh Ait VAROY) Ley tea OB F Sw ‘eth i } Wwe oP Bie ae to } ce AE ree . “3 » | oy “AB f

AL NES) aren of vou

(A Vieeovis my : a ii pay H “* pei) onal ii: Mi fi “itap ial f

z ! Wvetorituit nt she re ni thet papeneslia ke D warn ‘ hugh ist : t Mic «4 hand. ie Ch ae ood Cr 7 | an i ; ‘(loro eee ee a yeivehig Ls feces

eiiapye pee]

fidence Ti she ae

il rool ACHE pte FF ’ xl in tee digit Appan ae

ZOOLOGICAL SCIENCE 10: 53-56 (1993)

© 1993 Zoological Society of Japan

Secretion of Mitogenic Factor(s) from Stocks of Paramecium tetraurelia, P. caudatum and P. multimicronucleatum

YOSHIOMI TAKAGI, KUmMIKO NIMURA, YUMIKO TOKUSUMI,

Hiromi FusisawA, Kazuniko Kagt!

and Hiroyuki TANABE-

Department of Biology, Nara Women’s University, Nara 630, 'Tokyo Metropolitan Institute of Gerontology, Sakaecho, Tokyo 173, *Government Industrial Research Institute, Takamatsu 761, Japan

ABSTRACT—We recently isolated and purified the Paramecium growth factor (ParGF). The starting material (crude sample) was a ca.100-fold concentrate obtained by ultrafiltration of a cell-free medium

of an early stationary mass culture of the jumyo mutant of Paramecium tetraurelia.

Its mitogenic

function has been assessed from the restoration of the reduced fission rate of the jumyo mutant in daily reisolation cultures. To determine whether stocks of P. tetraurelia other than the jumyo mutant and those of other Paramecium species secrete similar substance(s), we assayed crude samples of those stocks for mitogenic function using the jumyo mutant. The results obtained were positive for almost all the stocks of P. tetraurelia, P. caudatum and P. multimicronucleatum studied, suggesting that ParGF, or a functional homologue, is common to Paramecium and act as a mitogen on “non-self” cells as well as

on “self” cells (i.e. those producing the factor).

INTRODUCTION

Paramecium cells divide 3—5 fissions a day at 25°C in bacterized Wheat-Grass-Powder medium. The jumyo mutant [3], which was first isolated as a mutant having a short clonal lifespan [4], divides 0.5-1.5 fissions a day in daily reisolation cultures but 3-5 fissions a day in mass culture. This characteristic of the mutant provided the clue that led to the finding of Paramecium growth factor (ParGF) and to a method for its biological assay [5]. ParGF has been collected and purified from a cell-free medium of an early stationary-phase culture of jumyo cells and has been assayed in daily reisolation cultures of these cells. We investigated whether the cell-free medium derived from stocks of P. tetraurelia other than the jumyo mutant and those of other Paramecium species can restore the reduced fission rate of the jwmyo mutant in daily reisolation culture.

Accepted August 21, 1992 Received July 17, 1992

MATERIALS AND METHODS

Stocks

All the stocks except for G3 were from the collections of Y.T. at Nara Women’s University. Figure 1 shows the lineages of 12 stocks of P. tetraurelia as well as the mating types for some stocks, the genotypes of the jumyo locus [3] and the nd169 locus [1] for all the stocks. Stocks G3 (mating type V) and NK15 (VI) of P. caudatum syngen 3 and stocks CH313 (III) and CH312 (IV) of P. multimicronucleatum syngen 2 were used. G3 was provided by Dr. M. Fujishima of Yama- guchi University.

Culture

Cells from a given stock were washed and cloned according to the method of Sonneborn [2]. The culture medium was a 1% phosphate-buffered medium of 5 g/] Wheat-Grass-Powder (Pines In- ternational, Inc., USA), inoculated with Klebsiella pneumoniae 1-2 days before use.

For the mass culture, 1,000 ml flasks, each

54 Y. TAKAGI AND K. Nimura et al.

St51 (VII) KL-DD6 (VIII) 4-110 (VIII) PATIENCE EY ieb pe og” ndind Mutagenesis d4-SL4 (VII) d4N-527 (VIII) jumjum, cee fe” ndind Conjugation F

"hum, + nd

Autogamy cS St ole at al | 17a 12b 16c 15d 19c 18c 20b

jum/jum, yee jum/jum, nd/ind Ee HN ee pe” ndind Fic. 1. Lineages of P. tetraurelia stocks with their mat- ing types and genotypes of the jumyo and nd169 loci. The two genes are abbreviated jum and nd. d4-SL4 is the jumyo mutant.

containing 400ml culture medium, were used. After introduction of 1-10 cells/ml, the individual flasks were incubated at 26°C as still cultures. When the culture first became clear, indicating exhaustion of bacteria, it was considered to be at the early stationary-phase of growth.

For the daily reisolation culture, a single cell was placed in 400 wl of culture medium in a well of a depression slide housed in a moist chamber.

Preparation of the crude sample

Cells in the early stationary-phase culture were removed by filtering the medium through two folds of two sheets of filter paper (TOYO). The resulting cell-free fluid was concentrated about 100-fold by ultrafiltration in a Diaflow cell (Ami- con) with a Diaflow ultrafiltration membrane YM10 (Amicon) that had a nominal cut-off molecular weight of 10,000. The crude sample obtained was stored in a freezer (—25°C) and thawed in a refrigerator before use.

Assay of the crude sample for mitogenic function

Cells were taken from a clonal culture of the jJumyo mutant and placed in 12 (or 18) wells of depression slides, each containing 400 yl of culture medium, then cultured as daily reisolation lines for

5 days, half the lines being designated ex- perimental and half control. For each ex- perimental culture, a crude sample was added daily so as to raise concentration 1.5-fold of that in the original cell-free culture medium (when the sample was a 100-fold concentrate, 6 l was added; an 80-fold concentrate, 7.5 ul was added). When a line died out, it was replaced by one of the other living lines.

Fission rate was compared between two groups of lines for both the daily averages and the total averages on the 5th day. The crude sample was considered effective as a mitogen when the total average of the daily fission rates in the ex- perimental group was significantly higher (t-test).

RESULTS

Secretion of mitogenic factor from stocks of P. tetraurelia other than the jumyo mutant

Of the 12 stocks of P. tetraurelia with various combinations of mating types and the genotypes of the jumyo and nd169 loci (see Fig. 1), 11 stocks (except 17a) were examined for the secretion of mitogenic factor, or a functional homologue of ParGF. The crude sample collected from the early stationary-phase culture of each stock effectively restored the fission rate of jumyo cells in daily reisolation culture. The results for 6 representa- tive stocks with various combinations of mating types and genotypes are shown in Figure 2. The lines of jumyo cells supplemented with the crude sample from the indicated stock showed signi- ficantly higher fission rates than unsupplemented control lines, although the restored level of the fission rate was not so high as the wild-type level. The crude sample prepared from _bacterized medium that had not been inoculated with Para- mecium cells showed no such effect (data not shown).

These results indicate that the ability to secrete mitogenic factor, probably ParGF or a homologue, has nothing to do with differences in mating types and genotypes of the jumyo and nd169 loci.

Secretion of Mitogenic Factor in Paramecium 55

i St51 (VII) ,

3

0

Fission rate

3

2

(0)

0 1 2 3 4 5 6 , 12b bit 0 1 2 3 4 5 6

d4-SL4 (VII)

’ KL-DD6 (VIII)

3

2

1

0 T 1 3 4 3) 6 0 1 2 3 4 5) 6

15d h 16c

3

2

1

0 3 4 5 6 io} 1 2 3 4 5 6

Days of reisolation culture

Fic. 2. Mitogenic activity of crude samples prepared from P. tetraurelia stocks with various genetic backgrounds (see Fig. 1). Controls (open circles) are for 6 or 9 daily reisolation lines of jumyo cells cultured in depression slide wells that contained 400 pl of culture medium. Experiments (closed circles) are for 6 or 9 daily reisolation lines supplemented daily with the crude sample prepared from the stock indicated. Daily reisolation cultures were begun on day 0 and maintained for 5 days. The fission rate is given in log,N where N is the number of cells on the day following reisolation. Bars indicate standard deviations.

G3 (V) NK15 (VI) 4 4 o 3 3 - Ss So S 2 2 D Pe ay 1 1 (0) T T T T T 1 0 0 1 2 3 4 5 6 10} 1 2 3 4 S) 6

Days of reisolation culture Fic. 3. Mitogenic activity of crude samples prepared from P. caudatum stocks of different mating types. Same as in Figure 2.

restored the fission rate of jumyo cells in daily reisolation culture. These results strongly suggest that the mitogenic factor, probably ParGF or a

CH313 (III) CH312 (IV)

4 4 @) 3 3 ~ Ss Some 5 2 2 a7) eZ ey

1 1

(0) (0)

(0) 1 2 3 4 5 6 (0) 1 2 3 4 5 6

Days of reisolation culture Fic. 4. Mitogenic activity of crude samples prepared from P. multimicronucleatum stocks of different mating types. Same as in Figure 2.

homologue, is secreted into the surrounding medium from every stock of Paramecium species.

56 Y. TAKAGI AND K. Nimura et al.

Secretion of mitogenic factor from stocks of P. caudatum and P. multimicronucleatum

The crude sample collected from the early stationary-phase culture of two stocks of P. cauda- tum (Fig. 3) and two of P. multimicronucleatum (Fig. 4), all of different mating types, effectively

DISCUSSION

ParGF first was identified by the use of the Jumyo mutant both as a donor in mass culture and as a recipient in daily reisolation cultures for the assay [5]. ParGF therefore has been regarded as a substance recognized by “self” cells. This study shows that mitogenic factor(s) is secreted not only by a variety of P. tetraurelia cells which are of different mating types and are homozygous domi- nant and recessive for the jumyo locus (Figs. 1, 2), but also by cells with different mating types of P. caudatum (Fig.3) and P. multimicronucleatum (Fig. 4). Since what we have tested for the mitogenic activity in this study is not the purified ParGF but crude samples, the restoration of the reduced fission rate of jumyo cells is not due to ParGF itself but to interaction of molecules pre- sent in the crude samples. However, it is highly probable that a functional homologue of ParGF is secreted by a number of Paramecium species and acts on jumyo cells; ParGF may be a substance that is recognized by “non-self” cells as well as by “self” cells and all Paramecium cells may produce both ParGF and its receptor.

The fact that the jumyo mutant divides slowly in daily reisolation culture, where the substances secreted are abandoned daily, but divides rapidly in mass culture, where the substances secreted are accumulated, suggests that the mutant may be slow either in the rate of production of ParGF or in the effectiveness to respond to ParGF, or both. Although we repeatedly prepared crude samples from the same stock and tested their mitogenic activity with the use of this assay method, the levels of the fission rate restored were not consis- tent and thus comparison of the kinetics among different stocks was impossible. Therefore, the above-mentioned possibilities both remain open.

For that, we need the improvement of the assay method that should be accurate enough to com- pare the quantity and quality of ParGF as well as the use of purified ParGF.

We are so far unsuccessful in developing a new assay method in which wild type cells, not jumyo mutant cells, are used. When wild type cells are cultured in a bacterized medium in which Wheat- Grass-Powder is reduced to 1/16 that of the standard, they divide as slowly as jumyo mutant cells in daily reisolation cultures. But, this low fission rate can be restored not only by crude samples but also by concentrates of bacterized medium in which Paramecium cells had not been inoculated (data not shown). Therefore, the low fission rate induced by nutritional restriction can- not represent the low fission rate of the jumyo mutant in daily reisolation culture.

ACKNOWLEDGMENTS

We are grateful to Dr. Masahiro Fujishima of Yama- guchi University for his providing the stock G3 and to Drs. Geoffrey Beale of Edinburgh University, UK, and Donald Cronkite of Hope College, USA, for improving the manuscript. This work was supported by a Grant-in- Aid for Science Research (01540596) and one for Cooperative Research (A) (03304001) to Y.T. from the Ministry of Education, Science and Culture of Japan.

REFERENCES

1 Nyberg D (1978) Genetic analysis of trichocyst discharge of the wild stocks of Paramecium tetraure- lia. J Protozool 25: 107-112

2 Sonneborn TM (1970) Methods in Paramecium research. In “Methods in Cell Physiology Vol 4” Ed by DM Prescott, Academic Press, New York, pp 241-339

3 Takagi Y, Izumi K, Kinoshita H, Yamada T, Kaji K, Tanabe H (1989) Identification of a gene that shortens clonal life span of Paramecium tetraurelia. Genetics 123: 749-754

4 Takagi Y, Suzuki T, Shimada C (1987) Isolation of a Paramecium tetraurelia mutant with short clonal life-span and with novel life-cycle features. Zool Sci 4: 73-80

5 Tanabe H, Nishi N, Takagi Y, Wada F, Akamatsu I, Kaji K (1990) Purification and identification of a growth factor produced by Paramecium tetraurelia. Biochem Biophys Res Commun 170: 786-792

ZOOLOGICAL SCIENCE 10: 57-64 (1993)

© 1993 Zoological Society of Japan

Cytokine-Like Activities of a Humoral Opsonin from the Solitary Urochordate Styela Clava

KAREN L. Kexry!, Epwin L. Cooper!

and Davip A. RAFTos~

"Department of Anatomy and Cell Biology, School of Medicine, University of California, Los Angeles Los Angeles, California 90024, U.S.A and *Immunobiology Unit, School of Biological and Biomedical Sciences, University of Technology, Sydney N.S.W. 2007, Australia

ABSTRACT— Previously, we have identified and purified an opsonin from the plasma of the solitary urochordate, Styela clava. Here we show that this opsonin exhibits cytokine-like activies in tunicates. The opsonin enhances incorporation of *H-thymidine into cultures of pharyngeal tissues from S. clava and acts as a powerful chemoattractant. Cytokine-like activity is also indicated by the ability of the S. clava molecule to activate tunicate phagocytes. In addition, we show that the opsonin is produced by S. clava hemocytes and that its production is stimulated by zymosan. These data suggest that the opsonin characterized here is identical to a tunicate, cytokine-like molecule (tunIL-1/) previously identified by its ability to stimulate mouse thymocyte proliferation.

INTRODUCTION

Opsonins play a key role in invertebrate humor- al defense systems. We have previously identified a humoral opsonin in the tunicate, Styela clava [8]. Our studies indicate that preincubation of target cells wih Styela clava plasma increases the capacity of tunicate hemocytes for phagocytosis. Opsoniza- tion is inhibited by the carbohydrates mannan, N-acetyl-D-galactosamine, and galactose, and by the divalent cation chelator, EDTA. Such data suggest that the S. clava hemolymph may contain a C-type lectin [6]. This lectin-like activity has recently been isolated from S. clava hemolymph [9]. Gel filtration yielded a fraction with strong opsonic activity that was associated with a single, electrophoretically-resolved protein. The fraction- ated opsonin is sensitive to tryptic digestion and heat denaturation, and its biological activity is dose-dependent. Molecular characterization indi- cated that the opsonic protein is a 17.5kDa monomer with a pI of 7.0.

Accepted October 1, 1992 Received June 4, 1992

We postulate that the humoral opsonin from S. clava may subserve multiple inducible biological activities during inflammatory reactions. Such a diversity of function is typical of inflammatory proteins in mammals. For instance, the mamma- lian inflammatory cytokine, IL-1, has multiple activities including the enhancement of IL-2 and IL-2 receptor expression by T-cells, the induction of slow wave sleep, increased thymocyte and fibroblast proliferation and the regulation of chemotaxis and hematopoiesis [5]. IL-1 also appears to be involved with communication be- tween the immune and nervous systems [3, 4].

Similarly, complement components in mammals subserve diverse effector functions. C3 has the capacity to opsonize foreign targets, initiate the subsequent lytic cascade of complement compo- nents, release an anaphylatoxin that is both chemo- attractive and stimulatory and is involved in the activation of phagocytes and lymphocytes [7].

The propensity of inflammatory protein in mam- mals to have multiple biological activities may also be reflected in tunicates. Here we test a purified opsonin from S. clava for its ability to stimulate cell proliferation, chemotaxis, and cell activation.

58 K. L. Ketty, E. L. Cooper anp D. A. Rarros

MATERIALS AND METHODS

Tunicates

Styela clava was purchased from Marinus Inc., Long Beach, CA. Tunicates were maintained at 15°C in an aerated, sandbed-filtered, 180-liter aquarium filled with artificial sea water (Instant Ocean, Aquarium Systems, Mentor, OH; 3.4% w/v). Approximately 2 ml of Marine Invertebrate Diet (Carolina Biological, Gladstone, OR) were added to the aquarium every two days. Tunicates were allowed to adjust to conditions in the aquarium for one week prior to experimentation.

Hemocyte Suspensions and Plasma

Tunicates were bled by severing the stolon and collecting 1.5 ml of the exuding hemolymph in chilled, polystyrene, centrifuge tubes containing 1.5 ml of sterile-filtered, artificial sea water (ASW) (Instant Ocean; 3.4% w/v). The resulting hemo- cyte suspensions were allowed to stand for five minutes so that large debris and cell aggregates sank. The upper 2 ml of each suspension were then transferred to fresh tubes. The number of hemocytes in suspension was determined and the volume adjusted with ASW. Hemocytes from individual tunicates were not pooled for any assay.

Plasma was obtained by harvesting hemolymph in the absence of ASW. Cells and debris were removed by centrifugation (800g, 15 minutes, twice) and the resulting plasma stored frozen (— 20°C).

Proliferation of Pharyngeal Explants

Tunicate Tissue Culture Medium: Tunicate tis- sue culture medium (T-RPMI) was prepared in sterile-filtered ASW and contained, per liter, 454 mg RPMI 1650 powder (with L-glutamine, without sodium bicarbonate; Sigma Chemicals, St. Lous, MO), 10° units penicillin sulfate (Sigma Chemi- cals), and 100mg streptomycin sulfate (Sigma Chemicals) [10]. This medium was then sterilized by filtration and stored at 4°C.

Tunicate Tissue Culture Protocol: The stiulatory effects of purified opsonin were tested in cultures of pharyngeal tissue from S. clava. Pharyngeal tissue is considered to be a hematopoietic field in

tunicates thereby providing an excellent source to study cell proliferation. The procedure used to culture pharyngeal explants was similar to that of Raftos et al. [10]. Pharyngeal tissue was diced into 3x3X1mm? explants and rinsed three times in T-RPMI. Rinsed explants were transferred to 96-well-flat-bottomed culture plates (1 explant/ well; Costar, Pleasanton, CA) contianing T-RPMI (200 yl/well). Explants were cultured at 15°C in normal atmosphere.

Incubation of Tunicate Explants with Purified Opsonin: Explants were equilibrated to culture conditions for 3 days prior to adding purified opsonin. After this period, explants were transfer- red to either fresh T-RPMI or T-RPMI containing 1 «g/ml of purified opsonin (200 wl/well) [9]. Ex- plants were cultured at 15°C in normal atmosphere for a further 3 days.

Quantification of *H-thymidine Uptake: Ex- plants were removed from tissue culture plates and pooled in 2 ml of the same medium in which they had been cultured. Tissues were then incubated at 15°C with 5 ~Ci/ml methyl *H-thymidine (H- TdR, 20 Ci/mmol; ICN Radiochemicals, Costa Mesa, CA). After 18hr, excess *H-TdR was removed by washing the explants three times (1 hr per wash) in 4ml ASW on a hematology mixer (Fisher Scientific, Pittsburgh, PA). Wahsed ex- plants were transferred to scintillation vials (1 explant/vial) containing 200 yl trypsin (2% w/v in ASW) and digested overnight at 37°C. Two millili- ters of scintillation cocktail (Ecolite; ICN Radiochemicals) was added to each vial so that the incorporated radioactivity could be quantified with a Beckman LS 3150P Scintillation Counter. Data are presented as proliferation indexes (P.I.) such that:

P.J.=(counts per minute in experimental trials) / (counts per minute in controls).

Mean values for each trial were calculated from P.I.s for individual trials.

Hemocyte Migration Assay

The migration of tunicate hemocytes was quan- tified using chemotaxis chambers in which Trans- well inserts (6.5mm diameter, 5.0 ~m pore size polycarbonate filter, tissue culture treated; Milli-

S. clava Cytokine 59

pore, Bedford, MA) and 24-well tissue culture plates (Costar) formed the upper and lower wells respectively. One hundred microliter aliquots of cell suspensions (2 10° cells/ml) were added to the upper wells of the chemotaxis chambers. When required, purified opsonin was added to a final concentration of 1.0 “g/ml to the upper well immediately after the addition of cells. The lower well was filled with 600 «1 ASW with or without purified opsonin (1 g/ml). Cells were allowed to migrate through Transwell membranes for 2 hr (15°C). After incubation, the chemotaxis cham- bers were agitated vigorously, the Transwell in- serts removed, and the plates centrifuged (200g, 5 min. 4°C). Cells that had migrated to the lower wells were counted in 16 randomly selected 400 x fields of view using an inverted (tissue culture) microscope. Data are presented as migration indexes (M.I.) such that:

M.I.=(cells per field of view in experimental trials)/(cells per field of view in controls).

Mean values for each trial were calculated from M.I.s for individual trials.

Zymosan Stimulation of Opsonin Production

Zymosan-treated Hemocyte Cultures: Two hun- dred microliters of S. clava hemocyte suspensions (210° hemocytes/ml in ASW) were added to 48-well tissue culture plates. The plates were centrifuged (200g, 5 min.) and the supernatant removed and replaced with 200 yl either T-RPMI or zymosan (50 “zg/mlin T-RPMI). The cells were incubated 48 hr at 15°C. Following incubation, the plates were centrifuged (300g, 5 min.) and the supernatants removed and pooled. Pooled super- natants were centrifuged (S00 xg, 10 min.) to re- move cells and particulate zymosan and then filter- sterilized (0.22 ~m; Costar).

Assays of Supernatants: Culture supernatants were analyzed in two ways. First, they were tested for opsonic acitivity as described previously by Kelly, et al. [9]. Second, the supernatants were concentrated thirty-fold using Centriprep 10 cen- trifugal concentration units (Amicon, Danvers, MA) and then assayed for protein content by the Bradford method (Protein Determination Kit, Biorad, Richmond, CA) and subjected to SDS-

PAGE (see below).

Activation of Phagocytes by Purified Opsonin

Pre-incubation of Latex Beads: 800 pl aliquots of latex bead (Sigma LB 30, Sigma Chemicals) suspensions were washed (16,000 g, 6 min) and resuspended in 200 ul of ASW or purified opsonin. The beads were incubated at room temperature with constant agitation for 1.5 hr. Suspensions were then washed thrice (16,000g, 6 min) and resuspended in ASW to a final concentration of 5 x 10° beads/ml.

Phagocytosis of Latex Beads: Cover glasses (22 mm X22 mm, Gold Seal, Fisher Scientific) were suspended between moistened strips of filter paper in glass petri dishes. Each cover slip was overlaid with 100 1 of hemocyte suspension (3 x 10° cells/ ml). Hemocytes were allowed to adhere to cover slips for 2 hr (15°C). Cover slips were then washed thrice (800 ul ASW), and 50 ul of purified opsonin (S00 ng/ml) or ASW were added. Cultures were allowed to incubate for 30 min (15°C) before being overlaid with 50 pl of pre-incubated latex beads (5 10° beads/ml). These cultures were incubated for a further 30 min at 15°C before excess latex beads were removed by dipping each cover slip in ASW ten times. The cells were then fixed in 200 pl of absolute methanol (15 min, 4°C) and washed in distilled water. Cover slips were inverted onto microsope slides and sealed.

Quantification of Phagocytosis: Phagocytosis was assessed by phase contrast microscopy (1,250 magnification, oil immersion). A minimum of 200 hemocytes from at least four fields of view per cover slip were inspected. Data were recorded as the percentage of hemocytes that had phagocy- tosed any number of latex beads (% phagocytic cells) relative to the total cell population. Data are presented as phagocytic stimulation indexes (PSI) where:

PSI=(% phagocytic cells in experimental trials) / (phagocytic cells in controls).

Mean values for each trial were calculated from PSIs for individual trials.

SDS-PAGE SDS-PAGE was performed according to the

60 K. L. Ketty, E. L. Cooper anp D. A. Rarros

method of Laemmli [1] using 12% separating gels. Dalton VII-L markers (Sigma Chemicals) were used for calibration. Styela clava plasma (total protein concentration 70 ul/ml) and G-S50 gel filtration-purifed opsonin (49 wg protein/ml) [9] were diluted 1:1 in 2xSDS/sample buffer with 2% v/v 2-mercapoethanol while tunIL-1£ (200 ug protein/ml) [2] and zymosan-stimulated super- natant (260 ug protein/ml) were diluted 1:4 in sample buffer. Gels were silver-stained using an AG-5 kit according to the manufacture’s instruc- tions (Sigma Chemicals).

Purification of Opsonin and TunIL-18

The procedures used to purify opsonin have been described in detail elsewhere [9]. Thirty milliliters of concentrated tunicate plasma was subjected to gel filtration. Plasma was sterilized by filtration (0.22 wm) and concentrated thirty-fold using Centriprep 10 centrifugal concentration units (Amicon). Concentrated plasma was then applied to a gel filtration chromatography column (bed volume 150 ml) packed with Sephadex G-50 (Sig- ma Chemicals) in phosphate-buffered saline. Hemolymph fractions were eluted with PBS, and fractions of 2.5 ml were collected and immediately sterilized by filtration. The hemolymph fractions were assessed for opsonic activity, and the purity of the opsonic fraction was confirmed by a single band on silver-stained 12% SDS-PAGE gel.

The procedures used to purify tunIL-18 have also been described in detail elsewhere [2]. Tuni- cate IL-1-like proteins were isolated from 100 ml of tunicate hemolymph. After concentration by ultrafiltration (PM 10 membrane, Amicon, Dan- vers, MA), hemolymph was applied to a Biogel AcA 54 gel filtration column (Pharmacia, Piscat- away, NJ) which was eluted with phosphate buf- fered saline. Fractions were assayed for IL-1 activity in a mouse thymocyte proliferation assay. Peaks of IL-1 activity were pooled, concentrated and applied to a PBE 84 chromatofocusing column (Pharmacia). The chromatofocusing column was eluted with a linear pH gradient (pH 8.4—pH 4.0) and fractions were assayed for IL-1-like activity as above. Two predominant species, one being tunIL-1f, were isolated. The purity of both spe- cies was confirmed on silver-stained 12% SDS-

PAGE gels.

Statistical Analysis

Statistical analyses were performed with the Mystat software package (Systat, Inc., Evanston, IL). The statistical significance of differences between mean values was determined by the Stu- dent’s t-test [14]. Differences were considered to be significant for probabilities of less than 5.0%.

RESULTS

Effect of Purified Opsonin on?H-thymidine Uptake

Incubation of pharyngeal explants from S. clava with purified opsonin resulted in a significant (P< 0.05) increase in the level of *H-thymidine uptake when compared to controls incubated with T- RPMI (Fig. 1). Pharyngeal explants which had been cultured for three days with purified opsonin (1 «g/ml in T-RPMI) followed by 18 hr with 7H- thymidine (5 ~Ci/ml) exhibited a proliferation in- dex of 39.8+4.8.

40

30

P.I. ag

T-RPMI + opsonin

T-RPMI

Fic. 1. Proliferation indexes (+SEM, n=20 explants) measured by *H-thymidine uptake for S. clava pharyngeal explants incubated in the presence or absence of purified S. clava opsonin.

Purified Opsonin Enhances Hemocyte Migration

When added to the lower wells of chemotaxis chambers, purified S. clava opsonin significantly (P <0.05) increased the number of tunicate hemo- cytes migrating through Transwell membranes

S. clava Cytokine 61

ug/ml of protein in top:bottom well

Fic. 2. Migration indexes (+{SEM, n=5) for hemo- cytes incubated in Transwell chambers containing either ASW in both the upper and lower wells, ASW in the upper well and 1 ug/ml S. clava opsonin (in ASW) in the lower well, 1 “g/ml S. clava opsonin (in ASW) in the upper well and ASW in the lower well, and 1 g/ml S. clava opsonin (in ASW) in both the upper and lower wells.

(Fig. 2). Approximately twice as many cells (migration index 1.8+0.05) entered lower wells containing 1 ng/ml of opsonin relative to controls containing ASW alone (migration index 1.0+ 0.05). A reciprocal pattern was evident when purified opsonin was added to the upper wells of chemotaxis chambers. The addition of 1 ng/ml of opsonin to the upper wells reduced migration by 50% (migration index 0.5+0.05) relative to ASW- controls (P<0.05). The addition of purified opso- nin to both the upper and lower wellls (1 x g/ml) yielded no significant change (P>0.05) in the number of tunicate hemocytes migrating through the Transwell membranes (migration index 1.0+ 0.04) compared to ASW-controls.

Effect of Zymosan-stimulated Hemocyte Super- natants on Phagocytosis

Incubation of yeast in supernatants from zymo- san-stimulated hemocyte cultures significantly (P < 0.05) increased their phagocytosis by tunicate hemocytes when compared to yeast incubated in T-RPMI alone (Fig. 3). The phagocytic stimula- tion index for zymosan-stimulated-hemocyte su- pernatants was twice (PSI=2.1+0.05) that of T-

PSI

control supernatant supernatant

zymosan

Yeast Incubation Media

Fic. 3. Phagocytic stimulation indexes (+SEM, n=6) for hemocytes exposed to yeast that had been pre- incubated with either T-RPMI, zymosan-stimulated hemocyte supernatant, or supernatant from cells cultured in the absence of zymosan (control super- natant).

RPMI controls (PSI=1.0+0.09). Incubation of yeast in supernatants of hemocyte cultures that had not been exposed to zymosan also increased phagocytosis (PSI=1.4+0.09) relative to T-RPMI controls (P<0.05). However, this level was signi-

kDa 45.

36.

24. .

14. 1 2 3 4 lane Fic. 4. Silver-stained, reducing SDS-PAGE of Styela clava plasma (lane 1), zymosan-stimulated hemocyte supernatant (lane 2), tun-IL 1 (lane 3), and G-50 column purified opsonin from S. clava plasma (lane

4). Position of molecular weight markers are shown at the left (kDa).

62 K. L. Ketiy, E. L. Cooper AND D. A. RaFros

ficantly lower (P<0.05) than that for the phagocy- tosis of yeast which had been opsonized by zymo- san-stimulated hemocyte supernatants.

Characterization of Zymosan-stimulated Hemocyte Supernatant

SDS-PAGE of zymosan-stimulated hemocyte supernatant resolved a number of proteins with a prominent band at 17.5 kDa under reducing condi- tions (Fig. 4, lane 2). A similar 17.5 kDa protein was evident in whole S. clava plasma (Fig. 4, lane 1). The major protein evident in zymosan- stimulated hemocyte supernatant migrated to the same position as both purified S. clava opsonin and tunIL-1 (Fig. 4, lanes 4 and 3 respectively).

Activation of Phagocytes by Purified S. clava Opsonin

Pre-incubation of hemocytes with purified S. clava opsonin significantly (P<0.05) increased their capacity to phagocytose latex beads when compared to hemocytes pre-incubated in ASW (Fig. 5). Hemocytes which had been incubated with purified opsonin had a PSI of 3.0+0.22 in the

3.5

2.0 PSI 1.5 1.0 0.5 has ASW: opsonin: opsonin ASW Incubation media of hemocytes and latex beads (hemocytes:beads) Fic. 5. Phagocytic stimulation indexes (+SEM, n=5)

for tunicate hemocytes exposed to latex beads under the following conditions: ASW-incubated cells with ASW-incubated beads (ASW : ASW), opsonin-incu- bated cells with ASW-incubated beads (opsonin: ASW), ASW-incubated cells with opsonin- incubated beads (ASW : opsonin). Hemocytes were cultured 30 minutes at 15°C with either ASW or opsonin before the addition of latex beads.

presence of ASW-incubated latex beads while con- trol (ASW-incubated) hemocytes exhibited a PSI of 1.0+0.15. The addition of latex beads pre- incubated in purified opsonin to ASW-incubated hemocytes did not significantly (P>0.05) alter the phagocytic stimulation indexes (1.2+0.14) when compared to controls comprised of ASW- incubated hemocytes exposed to ASW-incubated latex beads.

DISCUSSION

We have shown that an opsonin from Styela clava exhibits several functional characteristics which are typical of cytokine-like molecules. First, the S. clava opsonin stimulates the proliferation of tunicate cells. Differences in cell proliferation between control and experimental cultures of S. clava pharyngeal tissue are readily quantified by uptake of *H-thymidine [11, 12]. Raftos et al. [13] have shown that *H-thymidine incorporation by cultured explants in inhibited by irradiation and competition with non-isotopic thymidine. This suggests that there is a direct relationship between cell proliferation and *H-thymidine uptake [12]. Hence, the enhanced incorporation of °*H- thymidine in pharyngeal cultures containing S. clava opsonin indicates that this molecule can regulate the proliferation of S. clava cells in vitro.

Purified opsonin also acts as a chemoattractant for S. clava hemocytes (Fig. 2). The addition of purified opsonin to the lower well of chemotaxis chambers significantly increased the level of migra- tion of hemocytes to the lower well. Conversely, when opsonin was added to the upper well of the chamber, migration to the lower well was inhibited while no net movement was evident when equal concentrations of opsonin were present in both the upper and lower wells of the chemotaxis chambers. This type of analysis indicates that the S. clava opsonin acts as a powerful chemoattractant for tunicate hemocytes.

In addition to its proliferative and chemotactic activities, the S$. clava molecule activates tunicate phagocytes. Pre-incubation of S. clava hemocytes with purified opsonin enchanced the level of pha- gocytosis of latex beads. This enhanced phagocytic activity cannot be explained by the opsonization

S. clava Cytokine 63

of latex beads by the S. clava protein. The S. clava protein acts as a carbohydrate specific lectin [8] which, as shown here, is not opsonic for abiotic targets such as latex. Latex beads incubated with Opsonin are not ingested at a greater rate than that of untreated controls. Hence, the increased pha- gocytosis of unopsonized latex beads by hemocytes in the presence of S. clava opsonin must reflect the ability of this protein to metabolically activate tunicate hemocytes.

The capacity of S. clava opsonin to regulate cellular processes such as proliferation, chemota- xis, and hemocyte activation indicates that it may play a central role in adaptive inflammatory reac- tions. However, adaptive reactions such as chemo- taxis would be dependent upon the differential production of opsonin. To fulfill that prerequisite, we have shown that the secretion of S. clava opsonin is induced by antigenic challenge. Assays of super- natants from hemocytes cultured with zymosan revealed an enhanced secretion of opsonic mole- cules relative to controls cultured in the absence of zymosan. SDS-PAGE indicated that this induced Opsonic activity was associated with a 17.5kDa protein that is of identical electrophoretic mobility to purified S. clava opsonin. It is reasonable to conclude that the S. clava opsonin is produced by hemocytes in repsonse to stimulation by the target, zymosan.

The functional and physiochemical properties identified here suggest S. clava opsonin is identical to a cytokine-like protein, designated tunIL-12, that has been isolated from the same species [2, 11]. Both the opsonin studied here and tunIL-1£ are 17.5 kDa proteins with pI’s of 7.0-7.4 [9, 11]. Both molecules are opsonic [13], and exhibit simi- lar sensitivity to inhibition by N-acetyl-D- galactosamine, galactose, and mannose [8]. They enhanced the proliferation of pharyngeal explants in vitro above the normal levels [11], act as che- moattractants (D.A. Raftos, personal communica- tion), and activate phagocytic cells [13]. Moreov- er, production of both S. clava opsonin and tunIL- 18 by S. clava hemocytes is stimulated by zymosan [13]. These data suggest that tunicates express an opsonic protein with cytokine-like activities which can mediate adaptive inflammatory responses.

ACKNOWLEDGEMENTS

We thank Rick Fairhurst and Steve Sarafian for their contributions. This study was supported in part by the National Science Foundation (Grant #DCB 90 05061) and by BRSG funds from the UCLA Schools of Medi- cine and Dentistry. David A. Raftos was a Fulbright Postdoctoral Fellow and a recipient of a Frederik B. Band Scholarship in Marine Invertebrate Immunology administered by The American Association of Immuno- logists.

REFERENCES

1 Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K (1989) Current Protocols in Molecular Biology section 102 John Wiley and Sons New York

2 Beck G, Vasta GR, Marchalonis JJ, Habicht GS (1989) Characterization of interleukin-1 activity in tunicates. Comp Bioch Physiol 92B: 93-98

3 Berkenbosch J, van Oers J, del Rey A, Tilders F, Besedovsky HO (1987) Corticotoropin-releasing factor-producing neurons in the rat activated by interleukin-1. Science 238: 524-526

4 Besedovsky H, del Rey AE, Sorkin E, Dinarello CA (1986) Immunoregulatory feedback between interleukin-1 and glucocorticoid hormones. Science 233: 652-659

5 Dinarello CA, Cannon JG, Mier JW, Bernheim HA, LoPreste G, Lynn DL, Love RN, Webb AC, Auron PE, Reuben RC, Rich A Wolff SM, Putney SD (1986) Multiple biological activities of human recombinant interleukin-1. J Clin Invest 77: 1734- 1739

6 Drickamer K (1988) Two distinct classes of carbohy- drate-recognition domains in animal lectins J Biol Chem 263: 9557-9560

7 Fearon DT, Wong WW (1983) Complement ligand- receptor interactions that mediate biological re- sponses. Ann Rev Immunol 1: 243-271

8 Kelly KL, Cooper EL, Raftos DA (1993) A humor- al opsonin from the solitary urochordate Styela clava. Dev Comp Immunol, 17: 29-39

9 Kelly KL, Cooper EL, Raftos DA (1992) Purifica- tion and molecular characterization of a humoral opsonin from the solitary urochordate Styela clava. Comp Bioch Physiol, 103B: 749-753

10 Raftos DA, Stillman DL, Cooper EL (1990) In vitro culture of tissue from the tunicate Styela clava. In Vitro 26: 962-970

11 Raftos DA, Cooper EL, Habicht G, Beck G (1991) Invertebrate cytokines: Tunicate cell proliferation stimulated by an endogenous hemolymph factor. Proc Natl Acad Sci USA 88: 9518-9522

12

13

64 K. L. Ketty, E. L. Cooper anp D. A. RAFros

raftos, DA, Stillman DL, Cooper EL (1991) Inter- leukin-2 and phytohaemagglutinin stimulate the pro- liferation of tunicate cells. Immunol and Cell Biol 69: 225-234

Raftos DA, Stillman DL, Cooper EL, Habicht G, Beck G (1992) Invertebrate cytokines II: Release of

Interleukin-1-like molecules from tunicate hemo- cytes stimulated with zymosan. Lymphokine and Cytokine Research, 11: 235-240

Sokal RR, Rohlf FJ (1981) Biometry 2nd Ed, WH Freeman and Co, New York, 2nd edition

ZOOLOGICAL SCIENCE 10: 65-72 (1993)

Mouse y-Casein cDNA: PCR Cloning and Sequence Analysis!

TETSUO SASAKI, MASATO SASAKI and JUMPEI ENAMI

Research Laboratory, Zenyaku Kogyo Co., Ltd., 2-33-7, Ohizumi, Nerima, Tokyo 178, Japan

ABSTRACT—A partial amino acid sequence of mouse y-casein was determined on a gas-phase amino acid sequencer. The N-terminal sequence (23 residues) was obtained using native y-casein, and the C-terminal sequence (13 residues) was determined with a corresponding peptide. The C-terminal peptide was isolated by affinity chromatography on an anhydrotrypsin agarose column after digestion of y-casein with Achromobacter lysyl-endopeptidase. A cDNA (507 base pairs) encoding the mature protein of y-casein was produced by polymerase chain reaction (PCR) amplification of lactating mammary gland first strand cDNA with oligonucleotide primers designed from the N-terminal (KHEIKDK-) and the C-terminal (-AHYTRFY) amino acid sequences. The full-length cDNA including 5’- and 3 -noncoding regions (920 base pairs) was cloned by the rapid amplification of cDNA ends (RACE) method of Frohman et al. [7]. Comparison of full-length cDNAs of mouse y-casein and rat y-casein showed 80% identity at the nucleotide level. The amino acids in the coding region of both

© 1993 Zoological Society of Japan

y-caseins were 75% identical.

INTRODUCTION

Caseins are the major milk proteins produced by lactating mammary glands, and their synthesis and secretion are controlled by the coordinated actions of multiple hormones [20]. Tissue- and stage- specific expression of casein genes is regulated by hormones, adipocyte-mammary epithelial cellinter- actions, and the extracellular matrix [2, 17, 21]. In order to study the specific contributions of elements which play major roles in the control of casein production, casein genes have been cloned and characterized from several species. These include bovine, murine, and rat f-casein as well as bovine alpha-S1 casein [9, 14, 15, 22, 23]. As a further step toward the goal of understanding the molecular mechanisms regulating milk protein gene expression, mouse y-casein cDNA _ was cloned and sequenced. Here we present the results including the deduced amino acid sequence.

Accepted July 30, 1992 Received Jyly 8, 1992

" The sequence reported in this paper has been deposited in the DDBJ, EMBL and GenBank data- bases (accession no. D10215).

MATERIALS AND METHODS

Purification of mouse y-casein

Mouse jy-casein (formerly known as pp22, a phosphoprotein of 22,000 daltons) was purified from SHN mouse skim milk as described previous- ly [5]. Fractions obtained through successive chro- matography on Sephadex G-100 and DEAE- cellulose columns were dialyzed against distilled water and freeze-dried. Purified y-casein was stored at —20°C until characterized.

Amino acid sequence analysis

The N-terminal amino acid sequence was deter- mined using the intact protein. Mouse y-casein was dissolved in 50% acetonitrile in water to a final concentration of 1mg/ml. About 150 pmol of y-casein was applied on a 4-mm square membrane of polyvinylidene difluoride (PVDF; Immobilon- P, Millipore Corporation), and it was oven-dried at 55°C. The sample was placed in the reaction chamber of a Shimazu PSQ-1 automatic gas-phase amino acid sequencer [6], and amino acid sequen- cing was done by Edman degradation in an automat- ic mode as recommended by the manufacturer. PTH-amino acids were analyzed with an on-line

66 T. Sasaki, M. SASAKI AND J. ENAMI

HPLC separation system. The C-terminal amino acid sequence was determined using a purified C-terminal peptide fragment [13, 16]. This frag- ment was purified from lysyl endopeptidase- digested y-casein by affinity chromatography on a column of anhydrotrypsin agarose (Takara Shuzo, Kyoto), which specifically binds arginyl or lysyl C-terminated peptides under acidic conditions.

Eighteen nmol (400 4g) of mouse y-casein was incubated at 37°C with 8 xg of lysyl endopeptidase (EC3.4.21.50) from Achromobacter lyticus (Wako, Osaka; sample to enzyme weight ratio=50:1) in 25mM_ Tris-HCl (pH8.9) containing 1mM EDTA. At 0, 18, and 22hr after starting the incubation, an aliquot of the reaction mixture was applied to a reverse-phase HPLC column (# ODS-2, GL Sciences, Tokyo) to monitor the ex- tent of digestion. Subsequent procedures were done at 4°C. The reaction was stopped by addition of diisopropyl fluorophosphate (DFP) to a final concentration of 0.1mM. After the pH of the enzyme-digested product was adjusted to 5.0 with acetic acid, 100 »g of the product was applied to a column (1013 mm) of anhydrotrypsin agarose equilibrated with 50 mM ammonium acetate (pH 5.0) and 20mM CaCl. The column was eluted with the same buffer at a flow rate of 0.2 ml/min, and the eluate was collected as the “pass-through” fraction. After the base line stabilized, the elution buffer was changed to 5mM HCl (pH 2.5). Lysyl C-terminated peptides eluted under these condi- tions were collected as “bound” fractions.

Since the C-terminal peptide fragment was pre- sent in the “pass-through” fraction, this fraction was further purified by reverse-phase HPLC. The “pass-through” fraction was mixed with an equal volume of 0.1% (v/v) trifluoroacetic acid (TFA), and it was applied to a Cjg-silica column (# ODS-2, GL Sciences, Tokyo) equilibrated with 0.1% TFA. Peptides were eluted with a linear gradient of acetonitrile from 0% in 0.1% TFA at 0 min to 100% in 0.1% TFA at 100 min at a flow rate of 0.5 ml/min. Absorbance was monitored at 215 nm. The fraction corresponding to the major peak was freeze-dried, and the amino acid sequence was determined as described above except that an arylamine-P VDF membrane (Sequelon-AA, Milli- pore Corporation) was used as a support. The

freeze-dried sample was dissolved in 20 ul of 30% aqueous acetonitrile solution, and it was applied to the membrane and heated at 55°C until completely dry. After the membrane cooled to room tempera- ture, 5 wl of 1% (w/v) 1-ethyl-3-(3-dimethylami- nopropyl) carbodiimide (EDC) in 15% acetonitrile and 0.1M 2-(N-morpholino)ethanesulfonic acid (MES), pH 5.0 (supplied in Covalent Attachment Kit from Millipore Corporation), were applied to the membrane, and the reaction was allowed to proceed for 20 min at room temperature. Then the membrane with the peptide coupled at its C- terminal and side chain carboxy! groups was placed in the reaction chamber of the sequencer, and peptide sequence analysis was performed as above.

Alkaline phosphatase treatment of mouse y-casein

In order to dephosphorylate amino acid re- sidues, mouse y-casein was treated with alkaline phosphatase as described by Draetta and Beach [4]. Mouse y-casein (1 nmol; 22 ug) was dissolved in 11 wl of buffer consisting of 100 mM Tris-HCl (pH 8.0), 5mM MgCl, and 100 mM NaCl. Then the mixture was incubated with 0.4 unit of calf intestinal phosphatase (Sigma) at 37°C for 2 hr. The amino acid sequence was determined using 400 pmol of the alkaline phosphatase-treated mouse y-casein as described above.

Reverse transcription

Total RNA was extracted from 10-day lactating BALB/c mouse mammary glands by the acid guanidinium thiocyanate-phenol-chloroform meth- od of Chomezynski and Sacchi [3]. First strand cDNA was synthesized in a 20 wl reaction mixture containing 4.9 4g of total RNA, 800ng of oligo(dT);213 primer (Pharmacia) and 200 units of SuperScript™ reverse transcriptase (GIBCO BRL) with a buffer system supplied by GIBCO BRL. After incubation at 45°C for one hr, the mixture was heated at 95°C for 5 min to inactivate the enzyme, and it was rapidly cooled on ice.

Synthesis of oligonucleotides

Oligonucleotide primers for polymerase chain reaction (PCR) and DNA sequencing were synthe- sized using a Cyclone Plus DNA Synthesizer

Mouse y-casein cDNA Sequence Gil

(MilliGen/Biosearch).

PCR

The reaction mixture (50 jl) consisted of 0.5 wl of the first strand cDNA and 49.5 yl of PCR cocktail (GeneAmp’™ DNA Amplification Kit, Perkin Elmer Cetus) containing 20 pmol each of sense and antisense primers, and 0.2 mM each of dNTPs. Both primers were attached with restric- tion sites at their 5’ ends for the purpose of cloning into plasmid vectors. The mixture was heated at 95°C for 5 min and rapidly cooled on ice. Then 1.5 units of Thermus aquaticus (Taq) DNA poly- merase (Perkin Elmer Cetus) was added. PCR amplification was carried out for 30 cycles in a DNA Thermal Cycler (Perkin Elmer Cetus) using a step program (50°C, 3 min; 72°C, 2 min; 94°C, 1 min) followed by a 10-min final extension at 72°C. A second amplification reaction was performed in 50 yl of PCR cocktail containing 0.5 yl or 1 pl of the first amplification product and 20 pmol each of the same or internal primers under the same conditions as the first reaction.

Rapid amplification of cDNA ends (RACE) method

The 5’-noncoding region of cDNA was amplified by the RACE method of Frohman et al. [7]. First strand cDNA was synthesized using the antisense primer C1 (for sequence, see Fig. 2). Then deoxy- adenosine was polymerized to the 3’-end of the first strand cDNA using dATP and terminal deoxy- nucleotidyl transferase as described below. PCR was performed with an oligo(dT),7 primer (U1 as shown in Fig. 2), and a gene-specific antisense primer (N2) located upstream to that used in reverse transcription.

Terminal deoxynucleotidyl transferase treatment

The reaction mixture (20 yl) contained 10 # 1 of the reverse transcription product, 100 mM sodium cacodylate buffer (pH 7.2), 8mM MgCl, 0.2 mM dithiothreitol (DTT), 0.25 mM dATP, and 10 units of terminal deoxynucleotidyl transferase (Takara Shuzo, Kyoto). After incubation at 37°C for 5 min, the reaction mixture was heated at 65°C for 5 min to denature the enzyme.

DNA sequence analysis

PCR amplification products were cloned into a restriction enzyme- and bacterial alkaline phos- phatase-treated pBluescript II vector (Stratagene) using a DNA ligation kit (Takara Shuzo, Kyoto). The plasmid was amplified in XL1-Blue as a host. Plasmid DNA was purified by the alkaline SDS method [1], and plasmid sequencing was per- formed in both directions by the dideoxy chain termination method using Sequenase Version 2.0 kit (U.S. Biochemical) and alpha-[**P]dCTP (Amersham; 3,000 Ci/mmole) [19].

Data analysis

Protein and DNA sequences were analyzed with PC/GENE software (Release 6.01, IntelliGenetics Inc., California) in an NEC-9801 personal com- puter.

RESULTS AND DISCUSSION

Analysis of native mouse y-casein on a gas- phase sequencer determined twenty-three N- terminal amino acid residues. The sequence was Lys-His-Glu-Ile-Lys- Asp-Lys- X-X-X-Glu-Glu- Ser-Ser-Ala-Ser-Ile-Tyr-Pro-Gly-Lys-X-Lys. The letter “X” indicates that the amino acid was not identified during this cycle of sequencing. Since an unidentified “X” amino acid was present as a single peak at the same position in each of these cycles, we hypothesized that the same amino acid was present in modified form in these cycles. In addi- tion, measurement at 322 nm instead of the stand- ard 269 nm revealed the presence of a dehydro- serine peak in each cycle where an unidentified peak was present. A dehydroserine peak detected at 322 nm is usually associated with a serine peak at 269nm [8]. Casein is a well-known phos- phoprotein with a high content of phosphoserine. Therefore, the possibility that this amino acid residue was phosphorylated serine was examined. When native y-casein was treated with alkaline phosphatase [4] and analyzed on a PSQ-1 gas- phase sequencer, serine peaks appeared at the positions where “X” amino acids were found. The height of the unidentified peak diminished corres- pondingly. Thus, it was concluded that the “X”

68 T. Sasaki, M. SASAKI AND J. ENAMI

amino acid residue was either pure phosphoserine Accordingly the C-terminal peptide was purified or a mixture of serine and phosphoserine (data not _ from the lysyl endopeptidase-digested product us-

shown). ing anhydrotrypsin-agarose affinity chromatogra- Determination of the C-terminal amino acid phy. Figure 1A shows the elution profile after sequence is an important step for obtaining the affinity chromatography. When the “pass-

y-casein CDNA sequence by the PCR method. through” fraction of the affinity chromatography

A

: fini

Q 10 20 30 40

minute

B

0 20 : 60 80 100 minute

Fic. 1. Isolation of the C-terminal peptide of mouse y-casein. Lysyl endopeptidase-digested mouse y-casein (100 g) was applied to a column of anhydrotrypsin agarose. The C-terminal peptide in the “pass-through” fraction was further purified by reverse-phase HPLC as described in “Materials and Methods.” A, Elution profile of lysyl endopeptidase-treated y-casein from an anhydrotrypsin agarose column. Large arrow points the position of the “pass-through” fraction. Elution buffer was changed to 5mM HCl at the point indicated by the small arrow. Vertical scale bar represents 1 x 10 * absorbance unit at 278 nm. B, Separation of the C-terminal peptide. Total “pass-through” fractions were combined and applied to a reverse-phase HPLC column. Elution was performed with a gradient of 0% to 100% acetonitrile in 0.1% TFA. Arrow indicates the position of the major peak. Vertical scale bar represents 0.1 absorbance unit at 215 nm.

Mouse y-casein cDNA Sequence 69

was applied to a reverse-phase HPLC column, one major peak was observed (Fig. 1B). To determine the sequence of this peptide, an arylamine-PVDF membrane was selected instead of a PVDF mem- brane for sample support. An arylamine group can potentially react with both C-terminal and side chain carboxyl groups (aspartic acid and glutamic acid) of peptides via carbodiimide activation. Be- cause the peptide is covalently bound to the mem- brane, the chance that the progressively shortened peptide will detach is greatly reduced.

Twelve residues distal to the first residue were determined using the arylamine-PVDF mem- brane. The sequence was ?-Tyr-Thr-Phe-Pro-Asn- Ala-His-Tyr-Thr-Arg-Phe-Tyr. To obtain the in- ternal amino acid sequences, each peptide frag- ment was purified from the lysyl endopeptidase- digested product, and they were applied to a reverse-phase HPLC column (chromatograms of the HPLC are not shown). Four major peaks were

detected, and their amino acid sequences were analyzed. The sequences were Ile-Ile-Asp, Asp-Tyr-Thr-Phe-Pro-Asn-Ala, Asp-Tyr-Thr-Gln, and Tyr-Ile-Gln-Tyr-Gln-Gln-Val-Thr-Ile-Pro- Gln-Leu-Pro-GlIn-Ala-Leu-His-Pro-Gln-Ile. Ba- sed upon these results, oligonucleotide primers were synthesized for PCR cloning (Fig. 2). The sense primer N1 was an oligonucleotide mixture corresponding to the N-terminal amino acid sequ- ence, and a restriction enzyme site was attached to N1 for cloning. The antisense primer Cl was an oligonucleotide mixture corresponding to the C- terminal amino acid sequence. A restriction en- zyme site was similarly attached to Cl for cloning (Fig. 2). After PCR with 20 pmol each of these oligonucleotides primers and first strand cDNA as a template, a DNA fragment of about 500 base pairs (bp) was obtained. The product was then cloned into plasmid vectors, and a partial nu- cleotide sequence was determined. The sequence

Mouse y-Casein (coding region)

<— N2

U3 A —_> Ul NO N1 ee —_ B Sense Primer (LysH N1 5’ -TT-GGATCC-AAAC Baail G NO

la Gil - 8 . si =6:G

IN eine aia te3)

Universal Primer

Ul U2

Us 8 BB”

EcoRI Baol1

<= ———— Ci U1 <— U2 GlulleLysAspLys-) GAAATAAAAGACAA- 3’ G c G T

-GATTCGACCTGTTACCG- 5’

5” -GA-GAATTC-GGATCC-TTT-CTCGAG-TCTAGA-Tj7- 3’ EcoR1 Bani Xhol Xbal 5” -CC-TTT-CTCGAG-TCTAGA- 3’ Xhol Xbal

-GA-GAATTC-GGATCC-TTT- 3’

Fic. 2. Primers used for PCR amplification. A, Maps of the PCR primers. Sense and antisense primers are denoted by (—) and (<), respectively. The universal primer U1 was used for both the sense and antisense primer. B, Sequences of primers depicted in Figure 2A. Primers N1 and Cl were mixtures of oligonucleotides corresponding to the N- and C-terminal peptide sequences, respectively. “N” in the sequence of the primer C1 means a mixture of A, C, G and T. Restriction sites attached to each primer also are shown.

70 T. Sasaki, M. SASAKI AND J. ENAMI

closely resembled that of rat y-casein cDNA, and it was, therefore, concluded that this clone repre- sented part of mouse ;-casein.

The 5’-noncoding region was amplified by the rapid amplification of cDNA ends (RACE) method [7], and poly(dA) was tailed to the first strand cDNA using terminal deoxynucleotidyl transferase. PCR was performed using 1 yl of this poly(dA)-tailed first strand cDNA as a template, 23 pmol of the antisense primer N2, a mixture of 2.2 pmol of the universal primer U1, and 19 pmol of the universal primer U3 (Fig. 2). Subsequently, the product was cloned into plasmid vectors. Se- quencing of a clone revealed that the insert con- sisted of artificial (dA)so-tail, 5’-noncoding (57 bp) and coding (54 bp) regions of mouse y-casein.

For amplification of the full length y-casein cDNA, the sense primer NO corresponding to the 5’-noncoding region was synthesized based on the above results (Fig. 2). PCR was done using 20 pmol of the sense primer NO, 20 pmol of the universal primer U2, and 0.5 yl of the first strand cDNA, which was reverse-transcribed using the universal primer Ul. A DNA fragment of about 900 bp was obtained, and its size was close to that expected for rat y-casein CDNA. A second PCR was done with 15 pmol of the sense primer NO, 15 pmol of the universal primer U2, and 1 wl of ten-fold diluted primary PCR product as a tem- plate. A DNA fragment of identical size was obtained.

After cloning this fragment into a Psfl- and Xhol-treated plasmid vector, its nucleic acid se- quence was determined in both directions. Analysis of cDNAs obtained from eight independent plas- mids revealed that four had an identical sequence. The other four clones differed at several positions. The latter were probably caused by Taq poly- merase-induced misincorporations of nucleo- tides during PCR amplification. Figure 3 shows the nucleotide sequence of mouse y-casein cDNA from the four identical clones. The deduced amino acid sequence is also shown. The size of mouse y-casein cDNA was 920bp including the 5’- noncoding region (56 bp), the coding region (552 bp), and the 3’-noncoding region (312 bp). Mouse y-casein cDNA and rat y-casein cDNA [12] were 80% identical at the nucleotide level and 75% at

the encoded amino acid level. A polyadenylation signal was present near the 3’-end. The amino acid sequences of four internal peptides as well as a C-terminal peptide purified from lysyl endopepti- dase-digested product on a gas-phase peptide se- quencer were also determined. These were in complete agreement with the amino acid sequence deduced from the mouse y-casein cDNA. Although the first amino acid residue of the C- terminal peptide was not identified on the gas- phase peptide sequencer, the cDNA sequence showed that it was asparatic acid. This conclusion is reasonable, because asparatic acid cannot be identified on a gas-phase sequencer when peptides are treated with arylamine to fix to the membrane. The total number of deduced amino acid residues is 184 including a signal peptide of 15 residues, and the calculated molecular weight of mature mouse y-casein is 19,446 daltons.

Although Hennighausen and Sippel [10] isolated a cDNA for mouse y-casein, the sequence was not reported. The restriction map of their y-casein cDNA differed from the present y-casein cDNA. Consequently, their y-casein cDNA is probably different from ours. They also reported the isola- tion of a cDNA for mouse 6-casein [10], and a partial amino acid sequence of the 0-casein, Ser- Ser-Ser-Ser-Ser-Glu-Glu-Ser-Ser-Glu-Glu-Val, is identical to the 53rd-64th residues of the present y-casein [11]. In addition, the restriction map of the 6-casein resembles that of the present y-casein. Further studies including sequence analysis of 0- casein cDNA are needed to determine whether or not these cDNAs are identical.

In casein phosphoprotein, the common se- quence for serine phosphorylation sites (-Ser-Ser- Glu-Glu-) [14], where both serine residues are phosphorylated, is found in three regions in mouse y-casein (Fig. 3). .In addition, there are three consecutive phosphorylated serine residues near the N-terminal sequence of mouse y-casein. According to Mercier [18], casein kinase phos- phorylates the serine residue in the sequence, Ser-X-Y, where Y is either a phosphorylated serine or an acidic residue, usually glutamic acid. It is likely, therefore, that 16 of 29 serine residues (55%), including the above mentioned three con- secutive serine residues, are phosphorylated in

MOUSE RAT

MOUSE RAT

MOUSE RAT

MOUSE RAT

Mouse y-casein cDNA Sequence

AT CAT CAT CYA CCT ATT CCT CTC GTC TTC CAT TTG GGG AGC AAG GAA CAA GTA ACC

a aaa waf\ cod soo oul 1 } 20 Met Lys Phe Phe Ile Phe Ala Cys Leu Val Val Val Ala Leu Ala Lys llis Glu Ile Lys ATG AAG TTC TTC ATC TTT GCC TGC CTG GTG GTT GTT GCT CTG GCA AAG CAC GAA ATA AAG CA ce g a(ticer OG 3 T Wo Coo e e e e e e Thr e e e Ala Ala e e e e e Ala Va e

AC Tt === oT) Gee TTG aie “A> 3G: e e Pro e e e e e — e e Val e Leu e e Tyr e Gln Gly 50 60

G > AG oo[6 o09 Cae 5 {Me TA e C) C) Ser . C) C) e e Asp ° s e Pro e Ile ° se e Tyr 110 120

a(e G ot o50 aofl o C Cc . Ser e C) e Thr ° e e e e e e e e e ° e © Ile 150 160

Gln Ile Pro Val Ser Tyr Trp Tyr Pro Ser Lys Asp Tyr Thr Phe Pro Asn Ala His Tyr CAG ATA CCT GTG AGC TAC TGG TAC CCC AGT AAG GAT TAC ACC TIC CCA AAT GCT CAC TAT

Sie Goo cbs oP oon (Co eho ao Oban oon ODE apo Noo (ooo ocr a tomornlt e e Arg e e Ser e Ala e e e e e e e e Thr e Arg e

Thr Arg Phe Tyr *### ACG AGA TTC TAC TAA GATTCCTAAAATAACTGCTGCTACCTAGCTATAG GCT TGA GTAGAAAATTGGTTTC

Se at ees SS ee SS eee Se eee SiS se eS Hod D6 Sooo oo oD HEE oO Ob A Oe eo _ la BOR MME GMGG WING CC====<=s=sssssssee5s5=5= CATCACCTCCTACTATTAACATTCCCTAAGATGCTITCTTGAAA

=cSoo0nc055 CoccoscvnsasaPoacoflccc sans o@foconn cna cso oneflasscananoo sooo cbfooa of\s

GTTCCAACTTTGGACTGTGGAGCCGTATTAACTGCATTGTTTAATTGAAGTCTTTTCAGTACTACTTCTGAGATAAGAA 2o(@@sseencnnoPcaoocs fo oftconacoocs [Mooooecgsases BoocnscconsoscuceuoDs f\ccboonsone

CCAAAATATCTCTCGCCAAGACATTAAAGTTTTAAACGC— Poly(A) Fococeeadodon Poacoonsvegpe7eND040009000 — Poly(A)

71

120 117

180 177

240 237

300 297

360 351

420 411

480 471

540 531

611 556

667 620

746 695

825 174

864 813

Fic. 3. Alignment of mouse y-casein amino acid and nucleotide sequences with those of rat y-casein. Mouse and rat y-caseins were aligned using the computer software, PC/Gene. Positions of identical residues in both y-caseins are labeled as ( - ) for a nucleotide and (@) for an amino acid residue. Amino acids are shown in three-letter code. A sequence gap is marked by (—). Amino acid residues corresponding to the termination codons are

indicated by (***).

residues are written in bold letters.

A polyadenylation signal (ATTAAA) is underlined. Potential phosphorylated serine

72 T. SASAKI, M. SASAKI AND J. ENAMI

mouse y-casein.

In conclusion, the nucleotide sequence of cDNA encoding mouse y-casein and the deduced amino acid sequence were presented. Based on this information, cloning of the 5’-flanking region of mouse y-casein gene is in progress. Identification of regulatory sequences may be useful for under- standing transcriptional control of milk protein genes.

ACKNOWLEDGMENTS

We are grateful to Dr. Hideshi Kobayashi, director of this research laboratory, for continuous encouragement. We thank Dr. A. Iwamatsu (KIRIN Brewery Co., Ltd.) for helpful advice on peptide sequencing, Drs. K. Kohmoto and F. Aoki (Department of Agriculture, Tokyo University) for providing information on dephos- phorylation, and Dr. R. Shiurba for critically reviewing the manuscript.

REFERENCES

1 Birnboim HC, Doly J (1979) A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res 7: 1513-1523

2 Blum JL, Zeigler ME, Wicha MS (1987) Regulation of rat mammary gene expression by extracellular matrix components. Exp Cell Res 173: 322-340

3. Chomczynski P, Sacchi N (1987) Single-step method of RNA isolation by guanidinium thiocyanate- phenol-chloroform extraction. Analyt Biochem 162: 156-159

4 Draetta G, Beach D (1988) Activation of cdc2 protein kinase during mitosis in human cells: Cell cycle-dependent phosphorylation and subunit rear- rangement. Cell 54: 17-26

5 Enami J, Nandi S (1977) A sensitive radioimmu- noassay for a component of mouse casein. J Im- munol Methods 18: 235-244

6 Findlay JBC, Pappin DJC, Keen JN (1989) Auto- mated solid-phase microsequencing. In “Protein se- quencing” Ed by JBC Findlay, MJ Geisow, IRL Press, Oxford, pp 69-84

7 Frohman MA, Dush MK, Martin GR (1988) Rapid production of full-length cDNAs from rare trans- cripts: Amplification using a single gene-specific oligonucleotide primer. Proc Natl Acad Sci USA 85: 8998-9002

8 Geisow MJ, Aitken A (1989) Gas- or pulsed liquid- phase sequence analysis. In “Protein sequencing” Ed by JBC Findlay, MJ Geisow, IRL Press, Oxford, pp 85-98

9 Gorodetsky SI, Tkach TM, Kapelinskaya TV (1988)

12

13

14

16

17

20

21

22

23

Isolation and characterization of the Bos taurus B-casein gene. Gene 66: 87-96

Hennighausen LG, Sippel AE (1982) Characteriza- tion and cloning of the mRNAs specific for the lactating mouse mammary gland. Eur J Biochem 125: 131-141

Hennighausen LG, Steudle A, Sippel AE (1982) Nucleotide sequence of cloned cDNA coding for mouse ¢ casein. Eur J Biochem 126: 569-572 Hobbs AA, Rosen JM (1982) Primary structure of the casein messenger RNA species. Nucleic Acids Res 10: 8079-8098

Ishii S, Yokosawa H, Kumazaki T, Nakamura I (1983) Immobilized anhydrotrypsin as a specific affinity adsorbent for tryptic peptides. Methods En- zymol 91: 378-383

Jones WK, Yu-Lee L-Y, Clift SM, Brown TL, Rosen JM (1985) The rat casein multigene family: Fine structure and evolution of the -casein gene. J Biol Chem 260: 7042-7050

Koczan D, Hobom G, Seyfert M-H (1991) Genomic organization of the bovine alpha-S1 casein gene. Nucleic Acids Res 19: 5591-5596

Kumazaki T, Terasawa K, Ishii S (1987) Affinity chromatography on immobilized anhydrotrypsin: General utility for selective isolation of C-terminal peptides from protease digests of proteins. J Biochem 102: 1539-1546

Li ML, Aggeler J, Farson DA, Hatier C, Hassell J, Bissell MJ (1987) Influence of a reconstituted base- ment membrane and its components on casein gene expression and secretion in mouse mammary epithe- lial cells. Proc Natl Acad Sci USA 84: 136-140 Mercier J-C (1981) Phosphorylation of caseins, present evidence for an amino acid triplet code posttranslationally recognized by specific kinases. Biochimie 63: 1-17

Sanger F, Nicklen S, Coulson AR (1977) DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74: 5463-5467

Topper YJ, Freeman CS (1980) Multiple hormone interactions in the developmental biology of the mammary gland. Physiol Rev 60: 1049-1106

Wiens D, Park CS, Stockdale FE (1987) Milk protein expression and ductal morphogenesis in the mammary gland in vitro: Hormone-dependent and -independent phases of adipocyte-mammary epithe- lial cell interaction. Dev Biol 120: 245-258 Yoshimura M, Oka T (1989) Isolation and struc- tural analysis of the mouse #-casein gene. Gene 78: 267-275

Yu-Lee L-Y, Richter-Mann L, Couch CH, Stewart AF, Mackinlay AG, Rosen JM (1986) Evolution of the casein multigene family: Conserved sequences in the 5’ flanking and exon regions. Nucleic Acids Res 14: 1883-1902

ZOOLOGICAL SCIENCE 10: 73-83 (1993)

Phosphatidylcholine Is an Endogenous Substrate for Energy Metabolism in Spermatozoa of Sea Urchins of the Order Echinoidea

MasatTosuHi Mira! and Masaru NAKAMURA~

"Department of Biochemistry, Teikyo University School of Medicine, Itabashi-ku, Tokyo 173, Japan and *Department of Biology, Faculty of Medicine, Teikyo University, Hachioji, Tokyo 192-03, Japan

ABSTRACT—A study was conducted to examine an endogenous substrate for energy metabolism in spermatozoa of six species of sea urchin, Anthocidaris crassispina, Echinometra mathaei, Pseudocentro- tus depressus, Strongylocentrotus intermedius, Strongylocentrotus nudus and Temnopleurus hardwickii, which belong to the order Echinoidea. These spermatozoa contained cholesterol and several kinds of phospholipids, including phosphatidylcholine (PC), phosphatidylserine, phosphatidylethanolamine and cardiolipin. After dilution of dry sperm from these species in seawater, a rapid decrease in the level of PC was observed, but other phospholipids remained at constant levels. The preferential hydrolysis of PC was related to the properties of phospholipase A>. Ultrastructural study showed that lipid bodies were present within mitochondria of the sperm midpieces of A. crassispina, E. mathaei and P. depressus. After incubation in seawater, the lipid bodies became small or disappeared. Thus it is concluded that spermatozoa of sea urchins of the order Echinoidea commonly use PC located in the

© 1993 Zoological Society of Japan

lipid bodies within mitochondria as a substrate for energy metabolism.

INTRODUCTION

It is known that spermatozoa are stored for months as immotile cells in male sea urchins [7, 24]. After being spawned in seawater, spermato- zoa begin movement, and respiration is activated. The energy for swimming is produced by mitochondrial respiration, and ATP is utilized almost exclusively by the dynein ATPase of the flagellar axoneme [4-6]. The transportation of high-energy phosphate from the midpiece to the tail is associated with a creatine-phosphate shuttle [27, 28]. However, it seems unlikely that sea urchin spermatozoa are capable of using an exog- enous substrate for energy metabolism, since hardly any nutrients are present in seawater. Previous studies have shown that sea urchin spermatozoa obtain energy for swimming through oxidation of endogenous phospholipids [20-22, 25]. Sperm respiration also supports the phospholipid meta-

Accepted November 4, 1992 Received September 21, 1992

bolism [20-22, 25]. Furthermore, it has been shown that following incubation in seawater of spermatozoa from Hemicentrotus pulcherrimus, a sea urchin of the order Echinoidea, the content of endogenous phosphatidylcholine (PC) decreases significantly, with no change in the levels of other phospholipids [16, 18, 19]. Thus, H. pulcherrimus spermatozoa possibly use PC as a substrate for energy metabolism. It is important to determine whether PC is a common preferred substrate for energy metabolism in sea urchin spermatozoa. For further clarification of energy metabolism using PC, the present study examined the substrate for energy metabolism in spermatozoa of six species of sea urchin of the order Echinoidea; Anthocidaris crassispina, Echinometra mathaei, Pseudocentrotus depressus, Strongylocentrotus intermedius, Stron- gylocentrotus nudus and Temnopleurus hardwickii.

Recently, PC has been shown to be abundant in H. pulcherrimus sperm midpieces [13]. Following the initiation of motility, the PC content of sperm midpieces decreases significantly, while that in sperm heads and tails does not change [13]. Fur-

74 M. Mira AND M. NAKAMURA

thermore, the sperm midpiece of H. pulcherrimus has been shown to contain lipid bodies in the intramembrane space of the mitochondrion [14]. Interestingly, the lipid bodies become small gradu- ally, coincident with a decrease in the level of PC [14]. Therefore, an ultrastructural examination of the sperm midpieces of A. crassispina, E. mathaei and P. depressus was also carried out.

MATERIALS AND METHODS

Sea urchins

Six species of sea urchin belonging to the order Echinoidea: A. crassispina, E. mathaei, P. depres- sus, S. intermedius, S. nudus and T. hardwickii, were collected at Asamushi (Aomori, Japan), Mis- aki (Kanagawa, Japan), Tateyama (Chiba, Japan) and Kagoshima (Kagoshima, Japan). E. mathaei consists of four different types (A, B, C, and D types) which are distinguishable by color pattern of spines, shape of spicules and so on [29, 30]. A type of E. mathaei was used in this study. Spermatozoa were obtained by forced spawning induced by injection of either 0.5 M KCl or 0.1 M acetylcho- line into the coelomic cavity. Semen was always freshly collected as “dry sperm” and kept undi- luted on ice. The number of spermatozoa was counted with a hemocytometer.

Incubation of spermatozoa

Dry sperm were diluted 100-fold in artificial seawater (ASW) consisting of 458 mM NaCl, 9.6 mM KCl, 10mM CaCl, 49mM MgSO,, 10 mM Tris-HCl at pH 8.2. Following dilution and in- cubation at 20°C, each sample was centrifuged at 3,000 x g for 5 min at 0°C.

Analysis of lipids

Total lipids were extracted from spermatozoa by the method of Bligh and Dyer [13] and analyzed by high-performance thin-layer chromatography (HPTLC), according to the method of Macala et al [12] with some modifications as described in pre- vious papers [18, 22]. The amounts of PC, phos- phatidylserine (PS), phosphatidylethanolamine (PE), cardiolipin (DPG), cholesterol (Ch), and free fatty acid (FFA) in sea urchin spermatozoa

were determined from the standard curves of the respective authentic lipids. The mass of PC in ug was converted to nmol using the relation, 1 “g= 1.27 nmol, based on_ 1-palmitoyl-2-arachidonyl- PC. Also, 1 ug of FFA was calculated to be 3.27 nmol as arachidonic acid.

Analysis of fatty acid composition of PC

Isolated PC on a thin-layer chromatography (TLC) plate was subjected to methanolysis by heating with 5% HCl-methanol at 85°C for 2 hr, as described previously [16, 17]. The fatty acid methylesters were extracted with n-hexane, fol- lowed by evaporation under a stream of No. The residues were dissolved in a small amount of n-hexane and analyzed using a GC-R1A gas-liquid chromatograph (GLC) (Shimadzu Instruments, Kyoto, Japan).

Assays of glycogen and glucose

Before and after incubation of dry sperm in ASW, spermatozoa were homogenized with 0.6 M perchloric acid. The homogenate was used for determination of glycogen by the enzymatic method [8]. The acidified homogenate was centri- fuged at 10,000xg for 10min at 4°C, and the supernatant was used for estimation of glucose after neutralization to pH6.5-7.0 with KOH. Glucose was measured enzymatically according to Kunst et al [9]. Readings were made at 340 nm at 20°C using a UVIDEC 430B spectrophotometer (Japan Spectroscopic Co., Tokyo, Japan).

Oxygen consumption

Oxygen consumption in a sperm suspension was measured polarographically with an oxygen con- sumption recorder (MD-1000, Iijima Electronics MFG Co., Aichi, Japan). Twenty-five microliters of dry sperm was incubated in 2.5 ml ASW in a closed vessel of the oximeter at 20°C. The diluted spermatozoa were left exposed to air until deter- mination of oxygen consumption at the desired time. Total oxygen consumption was calculated from the respiratory rate and incubation period, as described previously [20].

Estimation of phospholipase and lipase activity

Dry sperm were homogenized with 10mM

Energy Metabolism in Sea Urchin Sperm 75

CaCl, 10 mM MgCl, 1 mM dithiothreitol and 50 mM Tris-HCl at pH7.5. The homogenate was incubated with 4.75 kBq 1-palmitoyl-2-[1-'*C]ara- chidonyl-PC (1.9 GBq/mmol), 4.75 kBq 1-palmi- toyl-2-[1-'*C]arachidonyl-PE (1.9 GBq/mmol) or 9.25 kBq [carboxyl-'4C]-triolein (4.1 GBq/mmol) for 1 hr at 20°C in a total volume of 0.4 ml. The lipids were extracted according to Bligh and Dyer [3]. The radioactivity in the FFA fraction sepa- rated by TLC was measured by liquid scintillation spectrometry. The protein content of the homogenate was measured by the method of Low- ry et al [11], using bovine serum albumin as the standard.

Electron microscopic observation

Before and after dilution of dry sperm in ASW and incubation for 30 min at 20°C, the spermato- zoa were prefixed in 2.5% glutaraldehyde-ASW solution for 40-60 min at 4°C; a volume of sperm suspension was mixed with the same volume of cold 5% glutaraldehyde in 80% ASW. The fixed spermatozoa were rinsed with cold ASW and post-fixed with 1% OsO, in ASW for 2 hr at 4°C. Samples were washed in distilled water, and then immersed in saturated aqueous uranyl acetate for 4 hr for block staining. After dehydration in a graded series of ethanol solutions, the specimens were embedded in epoxy resin and ultrathin- sections were cut on a Reichert Ultracut ultrami- crotome. After staining with lead citrate, they were observed using a Hitachi 7000 electron microscope.

Reagents

The phospholipid, Ch and FFA standards were purchased from Sigma Chemical. 1-Palmitoyl-2- [1-"*C]arachidonyl-PC, 1-palmitoyl-[1-'4C]-arachi- donyl-PE and [carboxyl-C]triolein were obtained from Du Pont-New England Nuclear. All reagents and solvents were of analytical grade. HPTLC and TLC plates (silica gel 60) were obtained from E. Merck (Darmstadt, Germany).

RESULTS

Lipid content

Previous studies have shown that the lipids in H. pulcherrimus spermatozoa are composed of sever- al kinds of phospholipid and Ch [16, 17]. Similar phospholipids and Ch were also detected in A. crassispina, E. mathaei, P. depressus, S. intermed- ius, S. nudus and T. hardwickii (Fig. 1). Among the phospholipids, PC, PS, PE and DPG were identified in these spermatozoa. PC was present at high concentrations. Triglyceride and cholesterol ester were present at extremely low levels (less than 1 »g/10’ sperm).

When dry sperm of A. crassispina, E. mathaei, P. depressus, S. intermedius, S. nudus and T. hardwickii were diluted 100-fold in ASW and incubated for 1 hr at 20°C, the PC content de- creased significantly following the initiation of flagellar movement (Figs. 1 and 2). Although a slight increase in FFA was also observed during incubation, the levels of other phospholipids and Ch remained almost constant. During incubation in ASW for 1 hr, about 11, 9 and 5 nmol PC were consumed by 10° spermatozoa of A. crassispina, E. mathaei and P. depressus, respectively (Fig. 2a). Amounts of FFA accumulated were about 5, 5 and 2 nmol in 10’ spermatozoa of these species (Fig. 2b).

Fatty acid composition of PC

Fatty acid components of PC in dry sperm of A. crassispina, E. mathaei, P. depressus, S. inter- medius, S. nudus and T. hardwickti were of the unsaturated type for the most part, such as oleic (18:1), eicosamonoenoic (20:1), arachidonic (20: 4) and eicosapentaenoic acid (20:5) (Table 1). Polyenoic fatty acids constituted more than 50% of the total fatty acid moiety of PC in these spermato- zoa, except that the PC in T. hardwickii spermato- zoa contained about 35% polyenoic fatty acids. In contrast, the saturated fatty acids in PC were present at only 20-30%, palmitic (16:0) and stearic (18:0) acids being predominant.

Glycogen and glucose content

Glycogen was present in spermatozoa of the six

76 M. Mira AND M. NAKAMURA

Content (ug/10°% sperm)

Fic. 1.

Changes in lipid levels after incubation of spermatozoa of A. crassispina (a), E. mathaei (b), P. depressus (c),

S. intermedius (d), S. nudus (e) and T. hardwickii (f). Before (clear) and after (dotted) 100-fold dilution and incubation of dry sperm in ASW for 1 hr at 20°C, lipids were extracted and analyzed by HPTLC. Each value is the mean of four separate experiments. Vertical bars show S.E.M. P values are compared with those prior to incubation by means of Student’s ¢ test. *P<0.1; **P<0.05.

species, but at extremely low levels (Table 2). Glucose was present in a trace amount (Table 2). After incubation in ASW for 1hr, the glycogen content of 7. hardwickii spermatozoa decreased significantly, but there was little change in sperma- tozoa of the other species.

Oxygen consumption

Since oxygen is required for oxidation of the lipid, the amount of O, consumed by the sperma- tozoa was measured at various intervals after dilu- tion in ASW (Fig. 3). The rate of O2 consumption decreased gradually during long-term incubation. About 0.55, 0.40 and 0.20 umol O, was consumed

during incubation of 10’ spermatozoa of A. crassi- spina, E. mathaei and P. depressus for 1 hr, respec- tively.

Phospholipase activity

Since hydrolysis of PC in H. pulcherrimus sper- matozoa occurs via the action of phospholipase A> [13, 16, 18], an experiment was conducted to examine the properties of the phospholipase A> in spermatozoa of the other species of sea urchin. The homogenates of dry sperm from A. crassi- spina, E. mathaei, P. depressus, S. intermedius, S. nudus and T. hardwickii were incubated with 1-palmitoyl-2-[1-'"C]arachidonyl-PC, 1-palmitoyl-

Relative lipid content (nmol/10° sperm)

Energy Metabolism in Sea Urchin Sperm Wi

2-[1-'*C]-arachidonyl-PE or [carboxyl-'4C]-triolein for 1 hr at 20°C, followed by extraction and separa- tion of FFA by TLC. The radioactivity of FFA from PC released by phospholipase A, in these species of spermatozoa was 2-3 times higher than that from PE (Table 3), suggesting that the phos- pholipase A» in sea urchin spermatozoa of the order Echinoidea has high substrate specificity for PC. In contrast to phospholipase, there was only a trace amount of radioactivity of FFA from TG released by lipase (Table 3).

Observation of sperm midpieces

The sea urchin spermatozoon consists of a head, a midpiece and a tail. Previous studies have shown

Fic. 2. Changes in levels of phosphatidylcholine (a) and free fatty acid (b) following incubation of sea urchin spermatozoa. Dry sperm of A. crassispina (™), E. mathaei (A) and P. depressus (@) were diluted 100-fold and incubated in ASW at 20°C.

0 15 30 45 60 Each value is the mean of four separate experi- Incubation time (min) ments. Vertical bars show S.E.M. TaBLeE 1. Fatty acid composition of phosphatidylcholine in sea urchin spermatozoa Percentage Fatty acid A. crassispina EE. mathaei PP. depressus S. intermedius S. nudus T. hardwickii

14:0 2.3+0.9 1.1+0.1 1.3+0.2 0.3+0.1 2.1+0.2 2.1+0.4 15:0 n.d. 0.3+0.1 0.1 tr. 1.9+0.2 2.7+0.4 16:0 12.6+0.7 19.6+0.1 12.1+0.5 12.7+0.6 22.8+0.3 11%3},5) 3E Il.) 16:1 4.2+1.0 n.d. 3.4+0.3 2.4+0.3 1.5+0.1 4.1+0.5 18:0 7.9+0.8 7a Ili 3.1+0.1 1.1+0.2 6.6+0.7 13}.5)35 2.5) 18:1 9.8+0.4 5.1+0.3 8.3+0.9 7.8+0.6 6.7+0.4 NS){3ae25) 18:2 1.5+0.2 1.5+0.1 1.1+0.3 0.8+0.1 1.3+0.2 Zsa) 5H) 18:3 0.3+0.1 lAseO.3 0.5+0.1 n.d. 0.6+0.1 1.0+0.3 18:4 8.9+0.3 4.5+0.7 14.3+0.5 7.4+0.5 n.d. n.d. 20:1 4.6+0.2 8.0+0.4 13.4+0.5 8.3+0.3 7.4+0.2 8.4+1.2 20:2 n.d. 2.4+0.7 n.d. 4.9+0.5 2.9+0.1 DD jpaes 20:3 7.8+0.6 3.2+0.5 6.0+0.6 tr. 2.1+0.6 1.0+0.5 20:4 (n—6) 27.1+0.9 Wd sya= Mel 23.4+1.5 17.2+0.4 20.0+0.6 Yaz Well 20:5 (n—3) 12.3+0.2 16.8+1.5 11.2+0.5 32.8+1.3 18.4+0.2 10.6+2.7 22:4 0.5+0.1 0.9+0.1 0.4+0.1 0.9+0.1 0.7+0.2 0.9+0.3 22:5 n.d. 0.7+0.1 0.1 0.2+0.1 1.3+0.2 DaV 7 22:6 tr. 2.3+0.6 1.4+0.8 3.2+0.7 3), 7/ae0)3) 7.0+0.5 Saturated 22.8+0.8 30.6+0.9 16.6+0.6 14.2+0.7 33.5+0.7 36.6+2.8 Monoenoic 18.6+1.0 13.1+0.2 25.1+1.6 18.4+0.9 15.5+0.4 DSO Eien al Polyenoic 58.5+1.1 56.2+0.9 58.3+2.2 67.4+1.5 50.8+0.4 SS ol se 3),2

Each value is the mean+S.E.M. obtained in three separate experiments.

n.d., not detectable.

tr., trace amount (less than 0.1%);

78 M. Mira AND M. NAKAMURA

TABLE 2. incubation with seawater

Glycogen (ug/10° sperm)

Changes in the levels of glycogen and glucose in sea urchin spermatozoa following

Glucose (nmol/10’ sperm)

Species dry sperm incubation for 1 hr dry sperm incubation for 1 hr

A. crassispina 0.5+0.1 0.5+0.1 tr. tr. E. mathaei 0.6+0.1 0.7+0.1 tr. tr. P. depressus 0.3+0.1 0.2+0.1 tr. tr. S. intermedius 0.5+0.2 0.4+0.1 ii, tte S. nudus 0.4+0.1 0.3+0.1 tr. tr. T. hardwickii 3.7+0.4 fig | ae(0).5)* tr. (ir.

Dry sperm were diluted 100-fold in ASW and incubated for 1 hr at 20°C.

Each value is the mean+

S.E.M. obtained from three separate experiments. tr., trace amount (less than 0.1 nmol/10° sperm) P

values are compared with those prior to incubation by means of Student’s ¢ test.

Oxygen consumption (jmol 03/109 sperm)

O 15 30 45 60 Incubation time (min)

Fic. 3. Oxygen consumption in sea urchin spermato- zoa. Dry sperm of A. crassispina (@), E. mathaei (A) and P. depressus (@) were diluted 100-fold and incubated in ASW at 20°C. Each value is the mean of three separate experiments. Vertical bars show S.E.M.

that the sperm midpiece of H. pulcherrimus con- tains lipid bodies in the intramembrane space of the mitochondrion [14]. In longitudinal sections through the spermatozoa of A. crassispina (Fig. 4a), E. mathaei (Fig. 4b) and P. depressus (Fig.

FPR<Ome

4c), the lipid bodies were observed within a single toroidal mitochondrion in the midpiece. Although a region between the mitochondrial outer and inner membranes was occasionally empty, the lipid bodies were irregular in profile and low-electron density. After 30 min of incubation in ASW, the lipid bodies in sperm midpieces of A. crassispina (Fig. 5), E. mathaei (Fig. 6) and P. depressus (Fig. 7) had shrunk or disappeared. Although these spermatozoa did not possess the lipid globules which have observed in Brissopsis lyrifera [1], Echinarachinus parma [26] and Glyptocidaris cre- nularis [15], the midpiece of A. crassispina occa- sionally contained spherical inclusion bodies with- in mitochondrial matrix (data not shown). How- ever, these inclusion bodies did not change during incubation.

DISCUSSION

This study showed that spermatozoa of A. cras- sispina, E. mathaei, P. depressus, S. intermedius, S. nudus and T. hardwickii, sea urchins of the order Echinoidea, contained PC at high concentra- tions (Table 1). The PC content decreased without any change in the content of other phospholipids following initiation of flagellar movement (Figs. 1 and 2). These spermatozoa contained triglyceride (Table 1) [16, 17], glycogen (Table 2) [20] and glucose (Table 2) [20] at very low levels. Thus it appears that PC is a common endogenous sub- strate for energy metabolism in spermatozoa of

Energy Metabolism in Sea Urchin Sperm 79

TaBLe3. Activities of phospholipase A> and lipase in sea urchin spermatozoa

Activity (nmol hydrolyzed/hr/mg protein)

Substrate

A. crassispina EE. mathaei PP. depressus S. intermedius S. nudus T. hardwickii PC 24+3 25-3 21+3 2342 PN ae 72 29 +2 PE 12+1 11+1 6+1 8+1 8+1 15+2 TG tr. tr. tr. tr. tr. tr.

Dry sperm were homogenized with 10 mM CaCl, 10 mM MgCl, 1 mM dithiothreitol and 50 mM Tris-HCl at pH7.5. The homogenate was incubated with 1-palmitoyl-2-[1-'*C]arachidonyl-PC, 1-palmitoyl-2-[1- 4C]arachidonyl-PE or [carboxyl-'*C]triolein for 1 hr at 20°C. Total lipids were extracted following by measurement of the radioactivity in the fatty acid separated by TLC. Each value is the mean+S.E.M.

obtained from three separate experiments. tr., trace amount (less than 0.1 nmol hydrolyzed/hr/mg protein).

"FE 4a oe

Fic. 4. Longitudinal section through a spermatozoon of A. crassispina (a), E. mathaei (b) and P. depressus (c). Arrow heads show lipid bodies. F: flagullum, G: acrosomal granule, M: mitochondrion, N: nucleus, SF:

submitochondrial fossa. 24,000.

urchins belonging to the order Echinoidea.

The data also showed the presence of phospholi- pase A>» activity in spermatozoa of these sea urchin species, but lipase activity was very low (Table 3).

The phospholipase A, was found to have strict substrate specificity for PC. These results suggest that the preferential hydrolysis of PC among phos- pholipids is due to the properties of phospholipase

80 M. Mita AND M. NAKAMURA

Fic. 5.

Longitudinal (a, b) and transverse (c, d) sections through the mitochondrial region of A. crassispina

spermatozoa before (a, c) and after incubation in ASW for 30 min (b, d). Arrow heads show lipid bodies. F:

flagellum, M: mitochondrion, N: nucleus.

A>. Previous studies have also indicated that H. pulcherrimus spermatozoa possess the f-oxidation and tricarboxylic acid cycle system [16, 18, 20], and thus it follows that the fatty acid obtained by hydrolysis of PC is metabolized to produce ATP for flagellar movement in sea urchin spermatozoa.

In this study, about 11 nmol PC was consumed in 10’ spermatozoa of A. crassispina after incuba- tion for 1 hr (Fig. 2a). In E. mathaei and P. depressus, PC consumption was about 9 and 5 nmol/10’ spermatozoa, respectively. Since PC is composed of two fatty acid moieties, about 22, 18 and 10 nmol fatty acids/10° spermatozoa are pro- duced respectively from PC in A. crassispina, E.

x 42,600.

mathaei and P. depressus during incubation for 1 hr. However, about 5, 5 and 2 nmol fatty acid remained as FFA following 1 hr of incubation of A. crassispina, E. mathaei and P. depressus sper- matozoa (Fig. 2b), Thus it is possible that about 17, 13 and 8 nmol fatty acid/10’ spermatozoa is utilized, respectively, to produce energy in these species. The chain length of fatty acid in PC of these spermatozoa was generally 20 carbons, as in the case of arachidonic and eicosapentaenoic acids (Table 1). From this, the amount of O5 required for PC metabolism during incubation for 1 hr was determined to be 0.50 “mol in 10° spermatozoa for A. crassispina, 0.38 «mol for E. mathaei and 0.24

Energy Metabolism in Sea Urchin Sperm 81

Fic. 6. Longitudinal (a, b) and transverse (c, d) sections through the mitochondrial region of E. mathaei spermatozoa before (a, c) and after incubation in ASW for 30 min (b, d). Arrow heads show lipid bodies. F:

flagellum, M: mitochondrion, N: nucleus. X 42,600.

pmol for P. depressus. These values are quite consistent with those actually determined for O, consumption; 0.55, 0.40 and 0.20 “mol O, per 1 hr in 10’ spermatozoa of A. crassispina, E. mathaei and P. depressus, respectively (Fig. 3).

Upon being spawned in seawater, sea urchin spermatozoa begin flagellar movement and re- spiration is activated. The initiation of sea urchin sperm motility requires external Na‘ and is associ- ated with Na‘-dependent acid extraction [23]. Following dilution in seawater, the intracellular pH (pHi) of sea urchin spermatozoa rises from 6.8 to 7.4 [2, 5, 10]. Internal alkalization leads to activation of dynein ATPase, resulting in initiation

of motility [5]. Also, the activation of phospholi- pase A, and fatty acid oxidation have been shown to increase with a rise in pH from 6.5 to 7.5 [19]. Thus, PC metabolism is activated following an increase in pHi of spermatozoa, coincident with initiation of motility and activation of respiration.

Our data also showed that several lipid bodies were present within the mitochondria of A. crassi- spina, E. mathaei and P. depressus spermatozoa (Fig. 4). Similar lipid bodies located between the mitochondrial outer and inner membrane have been observed in H. pulcherrimus spermatozoa [14]. These lipid bodies shrink and disappear after incubation [14], and the present data for spermato-

82 M. Mita AND M. NAKAMURA

Fic. 7.

Longitudinal (a, b) and transverse (c, d) sections through the mitochondrial region of P. depressus

spermatozoa before (a, c) and after incubation in ASW for 30 min (b, d). Arrow heads show lipid bodies. F:

flagullum, M: mitochondrion, N: nucleus. 42,600.

zoa of A. crassispina (Fig. 5), E. mathaei (Fig. 6) and P. depressus (Fig. 7) confirm this. These findings suggest that the disappearance of the lipid bodies is correlated with the decrease in the level of PC. Presumably, the lipid bodies within mitochondria of echinoid spermatozoa are reser- voirs of PC used as an endogenous substrate.

ACKNOWLEDGMENTS

The authors are grateful to Dr. I. Yasumasu, Waseda University, and to Dr. Y. Nagahama, National Institute for Basic Biology, for their encouragement and valuable advice. Thanks are also extended to Dr. K. Osanai and the staff of Asamushi Marine Biological Station, Tohoku

University, Dr. K. Inaba and the staff of Misaki Marine Biological Station, University of Tokyo, Dr. S. Nemoto and the staff of Tateyama Marine Laboratory, Ochano- mizu University, and Dr. H. Tousuji, Kagoshima Uni- versity, for affording us the opportunity to utilize their facilities and for their kind assistance in collecting the sea urchins. This study was supported in part by a Grant-in Aid (No. 03740396 to M.M.) from the Ministry of Educa- tion, Science and Culture of Japan.

REFERENCES

1 Afzelius BA, Mohri H (1966) Mitochondria respir- ing without exogenous substrate: A study of aged sea urchin spermatozoa. Exp Cell Res 42: 10-17

Bibring TJ, Baxandall J, Harter CC (1984) Sodium-

Nw

10

11

12

13

14

15

16

Energy Metabolism in Sea Urchin Sperm 83

dependent pH regulation in active sea urchin sperm. Dev Biol 101: 425-435

Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and publication. Can J Biochem Physiol 37: 911-917

Christen R, Schackmann RW, Shapiro BM (1982) Elevation of intracellular pH activates sperm re- spiration and motility of sperm Strongylocentrotus purpuratus. J Biol Chem 257: 14881-14890 Christen R, Schackmann RW, Shapiro BM (1983) Metabolism of sea urchin sperm. Interrelationships between intracellular pH, ATPase activity, and mitochondrial respiration. J Biol Chem 258: 5392— 5399

Gibbons BH, Bibbons IR (1972) Flagellar move- ment and adenosine triphosphate activity in sea urchin sperm extracted with triton X-100. J Biol Biol 54: 75-97

Gray J (1928) The effect of dilution on the activity of spermatozoa. J Exp Biol 5: 337-344

Keppler D, Decker K (1984) Glycogen. In “Methods of Enzymatic Analysis Vol 6” Ed by HU Bergmeyer, VCH Publishers, Weinheim, pp 11-18 Kunst A, Draeger B, Ziegenhorn J (1984) D- Glucose. In “Methods of Enzymatic Analysis Vol 6” Ed by HU Bergmeyer, VCH Publishers, Weinheim, pp 163-172

Lee HC, Johnson C, Epel D (1983) Changes in internal pH associated with initiation of motility and acrosome reaction of sea urchin sperm. Dev Biol 95: 31-45

Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the Folin phenol reagent. J Biol Chem 193: 265-275

Macala LJ, Yu RK, Ando S (1983) Analysis of brain lipids by high performance thin-layer chromatogra- phy and densitometry. J Lipid Res 24: 1243-1250 Mita M, Harumi T, Suzuki N, Ueta N (1991) Localization and characterization of phosphatidy- Icholine in sea urchin spermatozoa. J Biochem 109: 238-242

Mita M, Nakamura M (1992) Ultrastructural study of an endogenous energy substrate in spermatozoa of the sea urchin Hemicentrotus pulcherrimus. Biol Bull 182: 298-304

Mita M, Nakamura M (1992) Lipid globules at the midpieces of Glyptocidaris crenularis spermatozoa and their relation with energy metabolism. Mol Reprod Develop (in press)

Mita M, Ueta N (1988) Energy metabolism of sea urchin spermatozoa, with phosphatidylcholine as the preferred substrate. Biochim Biophys Acta 959: 361-369

17

20

21

22

23

24

25

26

27

28

29

30

Mita M, Ueta N (1989) Fatty chain composition of phospholipids in sea urchin spermatozoa. Comp Biochem Physiol 92B: 319-322

Mita M, Ueta N (1990) Phosphatidylcholine meta- bolism for energy production in sea urchin sperma- tozoa. Biochem Biophys Acta 1047: 175-179

Mita M, Ueta N, Harumi T, Suzuki N (1990) The influence of an egg-associated peptide on energy metabolism in sea-urchin spermatozoa: the peptide stimulates preferential hydrolysis of phosphatidyl- choline and oxidation of fatty acid. Biochem Bio- phys Acta 1035: 175-181

Mita M, Yasumasu I (1983) Metabolism of lipid and carbohydrate in sea urchin spermatozoa. Gamete Res 7: 133-144

Mohri H (1957) Endogenous substrates of respira- tion in sea-urchin spermatozoa. J Fac Sci Univ Tokyo, Sec IV, 8: 51-63

Mohri H (1964) Phospholipid utilization in sea- urchin spermatozoa. Publ Staz Zool Napoli 34: 53- 58

Nishioka D, Cross N (1978) The role of external sodium in sea urchin fertilization. In “Cell Repro- duction” Ed by ER Dirksen, DM Prescott, CF Fox, Academic Press, New York, pp 403-413 Rothschild Lord (1956) The physiology of sea urchin spermatozoa. Action of pH, dinitrophenol, dinitrophenol+versene, and usnic acid or O> up- take. J Exp Biol 33: 155-173

Rothschild Lord, Cleland KW (1952) The physio- logy of sea-urchin spematozoa. The nature and location of the endogenous substrate. J Exp Biol 41: 66-71

Summers DG, Hylander BL (1974) An ultrastructur- al analysis of early fertilization in the sand dollar, Echinarachnius parma. Cell Tiss Res 150: 343-368 Tombes RM, Brokaw CJ, Shapiro BM (1987) Creatine kinase dependent energy transport in sea urchin spermatozoa: Flagellar wave attention and theoritical analysis of ~P diffusion. Biophys J 52: 75-86

Tombes RM, Shapiro BM (1985) Metabolite chan- neling: A phosphocreatine shuttle to mediate high energy phosphate transport between sperm mitochondrion and tail. Cell 41: 325-334

Tsuchiya M, Nishihara M (1984) Ecological dis- tribution of two types of the sea urchin, Echinometra mathaei (Blainville), on Okinawan reef fiat. Galaxea, 3: 131-143

Uehara T, Shingaki M (1985) Taxonomic studies in the four types of the sea urchin, Echinometra mathaei, from Okinawa Japan. Zool Sci 2: 1009 (Abstract)

aa

Pie Rika)

es

it i tit

ae ir)

iy

ee Per Soest vets i U

va SY J H a 3) . We *- ae eh: f ~ ‘ a yey

aa

mivaoqe fry? oe ae an =

ie

Pr ae ul Sy Yu ijt wale?

(lay iy)

ci ie re

<3T fig ? eta apie ied

Seog tsi +) = ie oie ‘

i Bae” soe tori »

mate aelyicty Te oi WHrise auschigeen dy is

a aA br ae pani Seah S- aTy

a

eon C itt , toi | + c= > rit Pee . a * sh f - Fie F ‘ sear J Ty eg ~ at > io cated

sgul-pail) sonra

Feo) AC 25% bight Lene A iF A Wie J if

) ifeendn ‘1 ey {701 Ni eie

iti { tee a nN fy ake

ZOOLOGICAL SCIENCE 10: 85-92 (1993)

© 1993 Zoological Society of Japan

Grafting of Center Cells of Horseshoe Crab Embryos into Host Embryos at Different Developmental Stages

Tomio Itow

Department of Biology, Faculty of Education, Shizuoka University, Ohya 836, Shizuoka 422, Japan

ABSTRACT—The center cells of horseshoe crab embryos at the early gastrula stage induced secondary embryos after grafting into embryos at same stage and into late blastula stages. However, when the center cells of embryos at early gastrula stage were grafted into embryos at the stage of early cleavage, secondary embryos were not formed. Secondary embryos were not induced after grafting center cells into embryos at the late gastrula stage or after. These results indicate that the center cells of the early gastrula cannot induce secondary embryos at stages other than the early gastrula and late blastula.

However, center cells of embryos at stages later than the early gastrula did induce secondary embryos. These center cells of later embryos were grafted into embryos at the early gastrula and late blastula stage, and secondary embryos were induced. This indicates that center cells of embryos after the early gastrula stage retain the ability of embryonic induction.

INTRODUCTION

The mechanism of induction by the primary organizer is now one of the most exciting problems to be analysed in developmental biology [1, 9, 10, 13, 14]. However, reports of the primary organizer in animals except Chordata are sparse [2, 6, 7].

Surface cleavages occur in horseshoe crab embryos as in other Arthropoda, and the blasto-

Fic. 1.

pore appears at the early gastrula stage. Surface cells migrate toward the blastopore and enter there. A cell mass is formed beneath the blasto- pore at the early gastrula stage (Fig. 1). The cell mass at the gastrula stage is called the primary cumulus in spider embryos. Morphological obser- vations and previous experimental studies have shown that the germ disc is formed around the cell mass, and that cell mass is later situated at the

A: Anembryo of horseshoe crab at the early gastrula stage (Stage 7). B: An embryo of horseshoe crab at the

late gastrula stage (Stage 10). C: An embryo of horseshoe crab at the early neurula stage (Stage 12). CC; center

cells. PC; posterior cumulus.

Accepted March 26, 1991 Received August 29, 1992

86 T. Itow

posterior end of the ventral plate (the embryonic area) and successively forms segment primordia [3, 4,5, 12]. The cell mass at the posterior end of the embryonic area is sometimes called the growth zone.

If the cell mass of the embryo at the early gastrula stage is electrically cauterized, the treated embryo cannot develop further. If the cell mass is divided into two or three pieces by electrocauter- ization or cell dissociation, the treated embryo develops into double or triple embryos [8, 11]. These experiments suggest that the cell mass plays a central role in embryogenesis, and that it is similar to the mesodermal teloblast of Annelida and Crustacea. Therefore the mass is referred to as the center cells. To find out more about the role of the center cells, we attempted grafting experi- ments [6, 7]. We found that secondary embryos could be induced by interspecific grafts of center cells, indicating that center cells have the capaci- ties of primary induction.

It would be very interesting to know whether the center cells (the primitive cumulus) of the early gastrula can induce secondary embryos in embryos at different stages. Can the center cells (the growth zone) of stages later than the early gastrula induce secondary embryos in embryos at the early gastula stage? This paper deals with this question.

MATERIALS AND METHODS

Adult horseshoe crabs, Tachypleus (Carcinos- porpius) rotundicauda, were collected in the Gulf of Siam in Thailand and sent to Japan. Limulus polyphemus were obtained from Woods Hole Marine Laboratory in Massachusetts, the Duke University Marine Laboratory in North Carolina, and the Gulf Specimen Company in Florida. They were transferred to Shizuoka University where the present study was conducted. Eggs were fertilized by artificial insemination. The developmental stages were identified according to the normal criteria [12].

For the grafts, a small hole (diameter about 0.05—0.075 mm) was opened in the chorion of host horseshoe crab embryos at the late blastula (Stage 6) and early gastrula stage (Stage 7). The hole was positioned on the side opposite to the center cells.

The center cells and other tissues were cut off the donor embryos, and inserted through the hole. In the case of embryos at stages earlier than the gastrula, the surface layer and surface cells of the same volume as the grafted center cells were removed at the early gastrula stage. For the injection of homogenized center cells, center cells of 25-50 eggs at different stages were removed and homogenized in 0.5 ml distilled water or sea water. The concentration of homogenates of center cells from embryos at Stage 7 was about 0.5-5.0 mg (dry weight)/ml. Absorption granules (Collagen, Spongel [Yamanouchi Co., Tokyo]) were added to the resulting homogenates, and the solutions were again homogenized. More than 10 ° ml of solu- tion including the granules was then injected into host embryos at different stages, on the side oppo- site to the center cells. The treated embryos were cultured in a small laboratory dish in about 10 ml sea water containing antibiotics such as 0.5 «g/ml streptomycin and 0.5 unit/ml penicilline.

Normal and treated embryos were vitally stained with 1/20,000-1/400,000 neutral red and observed using a streomicroscope. Secondary embryos were judged to be at stage 20, Stage 21 (the stage of hatching), and the stage after hatching. When the treated embryo had an extra ventral plate (embryonic area), where there were often extra appendages, it was assumed a secondary embryo had been formed.

Some normal and treated embryos were fixed in Bouin’s or Carnoy’s solution, embedded in cel- loidin and paraffin and sectioned at 5-10 um. The sections were stained with Mayer’s haematoxylin and eosin.

RESULTS

Grafts of center célls at the early gastrula into embryos at same stage

The grafts of center cells were made in the American horseshoe crab, Limulus polyphemus and in the Asian species, Tachypleus rotundi- cauda. The results were similar in both species (Table 1). formed using mainly Limulus polyphemus, the results of experiments using both species are de-

Although the experiments were per-

Center Cell Grafts into Different Stages 87

TABLE 1. the same stage

Results of transplantation of crude center cells from early gastrulae (Stage 7) into embryos at

No of embryos No of embryos

Total number developed with secondary DONOR—HOST of operated upto St. 20 embryos

on embryos (percent of (percent of

the total) the developed) Limulus — Limulus 162 83 (51.2) 18 (21.7) Tachypleus — Tachypleus 191 20 (10.5) 4 (20.0) Limulus — Tachypleus 166 22 (13.3) 3 (13.6) Tachypleus — Limulus 211 38 (18.0) 6 (15.8) Non-center cells — Limulus 107 47 (43.9) 0 ( 0.0) [Control] Tas_e2. Results of transplantation of crude center cells and injection of homogenized center cells.

The center cells at Stage 7 were grafted into embryos at the same stage

No of embryos No of embryos

METHODS Total number developed with secondary OF of operated upto St. 20 embryos GRAFTINGS on embryos (percent of (percent of the total) the developed) TRANSPLANTATION 1536 188 (12.2) 37 (19.7) INJECTION 569 243 (42.7) 23 ( 9.5) Injection of homogenized 121 72 (59.5) 1 ( 1.4)

non-center cells [Control]

The control for transplantation experiments is shown in Table 1.

TABLE 3. (Stage 6) and Stage 7

Total number

Grafts of center cells from early gastrulae (Stage 7) into embryos at the late blastula stage

No of embryos developed

No of embryos with secondary

See RVs of operated upto St. 20 embryos on embryos (percent of (percent of the total) the developed) Late blastula (Stage 6) 1878 504 (26.8) 71 (14.1) Early gastrula (Stage 7) 1503 504 (33.5) 60 (11.9) imjection Of soa wavs (Cartel 0) a mal ne queers He pha hehe eee ntti Stage 6 362 137 (37.8) 0 ( 0.0) Stage 7 906 373 (41.2) 1 ( 0.3) scribed together. nique.

The rate of formation of secondary embryos following the injection of homogenized center cells was lower than that resulting from the transplanta- tion of intact center cells (Table 2). However, as the injection method is simpler, the following section describes the results obtained by this tech-

When the center cells were grafted into embryos at the late blastula stage (Stage 6), secondary embryos were induced at a rate similar to that in embryos at the early gastrula stage (Stage 7) (Table 3 and Fig. 2).

The secondary embryos induced after grafting

88 T. Itow

A _— _ __

Fic. 2. A: Secondary embryo after gafting of the cortical cytoplasm (surface layer) of an embryo at Stage 3 into an

embryo at Stage 7. B: Histological features of the secondary embryo. C: Secondary embryo after grafting of the midgut gland (liver) of an adult into an embryo at Stage 7. SE; secondary embryo. ns; nervous system.

center cells from the early gastrula stage (Stage 7) into embryos at the same stage had the following characteristics. The size and structure of secon- dary embryos differed. The smaller ones had only a small ventral plate without appendages. The larger ones had the same size and form as the host embryos. The alimentary canals and hearts of secondary embryos often fused with those of the

Cc

host embryos. The medial body axis of the secon- dary embryo was sometimes oriented in the oppo- site direction to that of the host embryo. These characteristics differed from those of double and triple embryos induced by electrocauterization and cell dissociation [11]. In the latter case, the posterior ends of induced double and triple embryos faced towards the point where the center

TasLe4. Grafts of center cells from early gastrulae (Stage 7) into embryos at different stages

No of embryos No of embryos

STAGE OF Total number developed with secondary HOST EMBRYOS of operated upto St. 20 embryos on embryos (percent of (percent of the total) the developed) Immediately after fertilization 40 21 (52.5) 2 ( 9.5) Early cleavage (Stages 1 & 2) 230 62 (27.0) 0 ( 0.0) Late cleavage (Stages 3 & 4) 118 30 (25.4) 1 ( 3.3) Early blastula (Stage 5) 149 66 (44.3) 2 ( 3.0) Late blastula and early gastrula 1005 194 (19.3) 26 (13.4) (Stages 6 & 7) Late gastrula and early neurula 57 29 (50.9) 0 ( 0.0) (Stages 10 to 12) Later stages (Stages 14 to 17) 26 18 (69.2) 0 ( 0.0) Injection of sea water [Control] Immediately after fertilization 59 24 (40.7) 2 ( 8.3) Stages 1 & 2 113 13 (11.5) 0 ( 0.0) Stages 3 & 4 111 41 (36.9) 0 ( 0.0) Stage 5 55 26 (47.3) 0 ( 0.0) Stages 6 & 7 1268 510 (40.2) 1 ( 0.2) Stages 10 to 12 349 231 (66.2) 0 ( 0.0) Stages 14 to 17 307 274 (89.3) 0 ( 0.0)

Center Cell Grafts into Different Stages

cells had been. In the secondary embryos resulting from grafting, the clear junction part was often observed in the region between the host embryo and the secondary one. These characteristics were not observed in the double and triple embryos induced by electrocauterization and cell dissocia- tion. More detailed characteristics of the secon- dary embryos have been described in the previous

paper [6].

Grafts of center cells from the early gastrula into embryos at different stages

When the center cells (primitive cumulus) of embryos at the early gastrula stage (Stage 7) were homogenized and injected into embryos at early cleavage stages (Stage 1 and Stage 2), secondary embryos were rarely induced, except in embryos injected immediately after fertilization (Table 4).

For the control experiments, we opened small holes in the chorions of embryos at different developmental stages, without grafting. In these cases, secondary embryos were not induced, ex- cept in embryos immediately after fertilization. When the host embryos were operated on im- mediately after fertilization, 21 of 40 treated embryos (52.5%) developed, and 2 of them

TABLE 5.

89

(9.5%) had secondary embryos. The proportion of treated embryos which developed secondary embryos did not differ significantly from that of the control embryos with small holes but no grafts.

When the host embryos were in the late cleavage stages (Stage 3 and Stage 4) and the early blastula stage (Stage 5), secondary embryos were induced at a very low rate (Fig. 3). The characteristics of the secondary embryos did not differ from those of secondary embryos induced after grafting center cells of early gastrulae into embryos at the early gastrula stage (Stage 7).

The center cells of early gastrulae (Stage 7) were grafted into embryos at stages later than Stage 7 (Table 4). Secondary embryos were never formed.

Grafts of surface cells at stages earlier than gastrulae into eggs at early gastrula stage

The grafting of cortical cytoplasm (surface layer) and surface cells (blastoderm cells) from unfertil- ized eggs and embryos at the early cleavage stages (Stage 1 and Stage 2) into the embryos at late blastula and early gastrula stages (Stages 6 and 7) never resulted in the formation of secondary embryos (Table 5). The grafting of center cells of embryos at late cleavage stages (Stage 3 and Stage

Grafts of the cortical cytoplasm (surface layer), surface cells and center cells at different stages into late blastulae (Stage 6) and early gastrulae (Stage 7).

The whole tissue of larvae and

different tissues from adults were also grafted into embryos at the same stages

Total number

No of embryos

No of embryos developed

with secondary

STABE OF DONOR EMBRYO of operated upto St. 20 embryos on embryos (percent of (percent of the total) the developed) Unfertilized eggs 76 25 (32.9) 0 ( 0.0) Early cleavage (Stages 1 & 2) 120 113 (94.2) 0 ( 0.0) Late cleavage (Stages 3 & 4) 116 66 (56.9) 5 ( 7.6) Blastula (Stages 5 & 6) 159 84 (52.8) 3 ( 3.6) Early gastrula (Stages 7) 1153 238 (20.6) 28 (11.8) Late gastrula (Stages 9 to 11) 360 95 (26.4) 9 (9.5) PPA Later stages (Stages 13 to 20) 629 294 (46.7) BD) (99) Ist & 2nd larvae 179 61 (34.1) 9 (14.8) Tissues of adults cartilage 238 76 (31.9) 4 ( 5.3) midgut gland 55 14 (25.5) 4 (28.6) ovary 110 42 (38.2) 5 (11.9) blood 80 39 (48.8) 0 ( 0.0)

Stage 2

So,

Fic. 3.

4) and blastula stages (Stage 5 and Stage 6) into embryos at Stage 6 and Stage 7 resulted in the formation of secondary embryos, but the rate of formation was very low (Table 5). The character- istics of the secondary embryos did not differ from those of secondary embryos induced after grafting center cells of early gastrulae into embryos at the early gastrula stage (Stage 7).

Grafts of center cells of stages later than early gastrulae into embryos at the early gastrula stage

When the center cells (growth zones) of embryos at stages later than early gastrulae were grafted into embryos at the late blastula stage (Stage 6) and early gastrula stage (Stage 7), secon- dary embryos were induced at a high rate (Table 5). Cells other than center cells did not induce secondary embryos after similar grafting. For example, when the posterior cumulus was grafted into embryos at Stage 6 or Stage 7, 18 of 92 treated embryos developed and none of them had secon- dary embryos.

The homogenates of whole tissues of larvae induced secondary embryos at a high rate after grafting into embryos at Stage 6 or Stage 7. Several tissues of adult horseshoe crabs were also able to induce secondary embryos, including cartil- age (endoskeleton), midgut gland (liver) and ov-

Stage 13

ot

Summary of results

ary. Injection of blood did not induce secondary embryos.

The characteristics of the secondary embryos induced after grafts of tissues of adults, larvae and embryos at stages later than Stage 7 did not differ from those of secondary embryos induced after grafting center cells of early gastrulae into embryos at the early gastrula stage (Stage 7).

The above-mentioned results are summarized in Fig. 3.

DISCUSSION

Secondary embryos were formed at a high rate after grafts of center cells of embryos at the early gastrula stage (Stage 7) into embryos at the late blastula stage (Stage 6) and Stage 7 (Table 1) [6, 7]. The rate of induction of secondary embryos in embryos receiving grafts at Stage 6 was as high as that in grafted Stage 7 embryos. This must mean that the cells of the embryos at the late blastula stage (Stage 6) have the capacity to respond to the signal for embryonic induction.

Secondary embryos rarely developed after graft- ing center cells of Stage 7 embryos into the embryos at other stages, except Stage 6. This indicates that only embryos at Stages 6 and 7 have the capacity to respond to the signal for embryonic

Center Cell Grafts into Different Stages 91

induction.

When small holes were opened, without graft- ing, on the chorions of embryos immediately after fertilization, secondary embryos sometimes de- veloped in the treated embryos. I interpret this phenomenon as follows. The chorions of embryos immediately after fertilization are still very soft. The distribution of substances in these eggs is thought not to be stable, and the distribution of cells is determined later, according to the distribu- tion of substances. The shock of opening a hole seems to deragne the normal distribution of subst- ances and cells, and treated embryos may some- times develop secondary embryos.

When the center cells were grafted into embryos at the stages of late cleavage and early blastula (Stages 3 to 5), secondary embryos were some- times induced, although at a low rate. I interpret these results as follows. One possibility is that embryos at Stages 3 to 5 have a weak capacity to respond to the signal for embryonic induction. Another posibility is that the grafted center cells or the substance for induction are retained in treated embryos until the latter develop to Stages 6 and 7.

The cortical cytoplasms (surface layers) of un- fertilized eggs and embryos at early cleavage stages (Stages 1 and 2) did not induce secondary embryons in embryos at Stage 6 or Stage 7. The cortical cytoplasms of such eggs may not yet pos- sess the induction substance for embryogenesis.

On the other hand, the surface cells of embryos at late cleavage and blastula stages (Stages 3 to 6) induced secondary embryos in embryos at Stage 6 and Stage 7, although the rate of formation of secondary embryos was low. The finding may mean that the surface cells of embryos at Stages 3 to 6 contain the substance for embryonic induc- tion.

When center cells of embryos at the early gastru- la stage (Stage 7) were grafted into embryos at the late gastrula stage and after, secondary embryos were never induced. But the center cells (growth zone) of embryos at the late gastrula, or later, stages did induce secondary embryos after grafting into embryos at Stage 6 or Stage 7. This must indicate that the center cells of embryos at stages later than the early gastrula stage (Stage 7) still contain the substance for embryonic induction.

The homogenates of whole tissues of larvae and some tissues of adults induced secondary embryos after injections into embryos at early gastrula. This indicate that the substance for embryonic induction is contained in tissues of larvae and adults.

REFERENCES

1 Cooke J (1989) Mesoderm-inducing factors and Spemann’s organizer phenomenon in amphibian de- velopment. Development 107: 229-241

2 Holm A (1952) Experimentelle Untersuchungen uber die Entwicklung und Entwicklungsphysiologie des Spinnen Embryos. Zool Bidrag Uppsala 29: 293-424

3 Itow T (1984) The induction of malformed embryos by inhibition of cell proliferation in the horseshoe crab, Tachypleus tridentatus. Acta Embryol Morpho exp NS 5: 177-193

4 Itow T (1985) The effect of Ca**-free sea water on the body segmentation in the horseshoe crab (Che- licerata, Arthropoda). Acta Embryol Morph exp NS 6: 15-29

5 Itow T (1986) Inhibitors of DNA synthesis change differentiation of segments and increase segment number in horseshoe crab embryos. Roux’s Arch Devl Biol 195: 323-333

6 Itow T, Kenmochi S, Mochizuki T (1991) Induction of secondary embryos by intra- and interspecific grafts of center cells under the blastopore in horseshoe crabs. Develop Growth & Differ 33: 251- 258

7 tow T, Mochizuki T (1988) Interspecific trans- plantation of germ disc cells from the American to the Asian horseshoe crab eggs. Proc Arthropod Embryol Soc Jpn 24: 15-16

8 Itow T, Sekiguchi K (1979) Induction of multiple embryos with NaHCO; or calcium free sea water in the horseshoe crab. Roux’s Arch Devi Biol 187: 245-254

9 Kimelman D, Kirschner M (1987) Synergistic induc- tion of mesoderm by FGF and TGF f and the identification of an mRNA coding for FGF in the early Xenopus embryo. Cell 51: 869-877

10 Kintner CR, Dodd J (1991) Hensen’s node induced neural tissue in Xenopus ectoderm. Implications for the action of the organizer in neural induction. Development 113: 1495-1505

11 Sekiguchi K (1966) Determination in the develop- ment of the horse-shoe crab. Jap J Exp Morphol 20: 84-88 (In Japanese)

12 Sekiguchi K (1973) A normal plate of the develop- ment of the Japanese horse-shoe crab, Tachypleus tridentatus. Sci Rep Tokyo Kyoiku Daigaku, Sec B,

92 T. Itrow

15: 152-162 326: 197-200 13 Slack JMW, Darlington BG, Heath JK, Godsave SF 14 Smith JC (1987) A mesoderm-inducing factor is (1987) Mesoderm induction in early Xenopus produced by a Xenopus cell line. Development 99:

embryos by heparin binding growth factors. Nature J il4t

ZOOLOGICAL SCIENCE 10: 93-101 (1993)

Ultrastructural Observations on Sperm Penetration in the Egg of Elkhorn Sculpin, Alcichthys alcicornis, Showing Internal Gametic Association

Yasunori Koya!, KAzuNoRI TAKANO? and Hiroya TAKAHASHI

‘Department of Biology, Faculty of Fisheries, Hokkaido University, Minato-cho 3-1-1, Hakodate 041, Japan, and *Sesoko Marine Science Center, University of the Ryukyus, Sesoko 3422, Motobu-cho, Okinawa 095-02, Japan

ABSTRACT—Early fertilization process has been investigated ultrastructurally in elkhorn sculpin, Alcichthys alcicornis. In the eggs before immersion in sea water, spermatozoon was seen to be attached to the ooplasmic membrane at the site of an inner opening of the micropylar canal. By 10sec after immersion in sea water, a fertilizing spermatozoon penetrated the ooplasm for about half the head length, and vesiculation occurred on the fused plasma membrane at the apical region of sperm head. By 30 sec after the immersion, the fertilizing spermatozoon showed a change in the tail axoneme, indicating a loss of its motility. By 1-2 min after the immersion, spermatozoon further penetrated the ooplasm up to its middle piece. It is worthy to note that a portion of ooplasm containing no significant cytoplasmic organelles other than ribosomes appeared around the penetrated sperm head and came to expand during the sperm penetration to form a triangular region surrounding the sperm head. This ooplasmic region may have a role in an active engulfment of immotile spermatozoon into the ooplasm. By 3 min after the immersion, sperm nuclear membrane was separated from chromatin and became vesiculated starting from the anterior region of sperm head caudalwards. Details of the events in early fertilization

© 1993 Zoological Society of Japan

process were presented.

INTRODUCTION

An extensive work on the ultrastructural level has been accumulated dealing with early events occurring in the course of fertilization in teleosts. Initial stages of sperm-egg fusion and the forma- tion of fertilization cone have been described in mummichog, Fundulus heteroclitus [2], and com- mon carp, Cyprinus carpio [8]; the process of sperm penetration has been traced, though partial- ly, in several species of teleosts [3, 7, 10, 13, 17, 18, 22, 24]; and changes of sperm nuclear membrane and the formation of the second polar body and blastodisc have been observed in rose bitterling, Rhodeus ocellatus ocellatus [18] and the mum- michog [3], respectively. However, few studies have so far examined a complete spectrum of these continuous events of early fertilization in any

Accepted October 15, 1992 Received August 5, 1992

single species of fishes. This seems to be due mainly to a difficulty of determining an actual time of occurrence of these successive events, since these events may proceed rather asynchronously in a mass of eggs after insemination. Only the work of Iwamatsu and Ohta [6] has revealed, using denuded and polyspermic eggs of medaka, Oryzias latipes, a series of ultrastructural changes from an initial stages of sperm-egg fusion to the formation of male pronucleus.

Elkhorn sculpin, Alcichthys alcicornis, has a unique mode of reproduction called “internal gametic association” [14]. In this teleost species, sperm is introduced into ovarian cavity by copula- tion and enters the micropylar canal of eggs ovu- lated into the ovarian cavity, but actual sperm-egg fusion does not occur within the ovary until the eggs have been released into sea water. In eggs freshly stripped from copulated females, sperm remains to be attached to the ooplasmic surface, and an actual fertilization reaction is initiated quite

94 Y. Koya, K. TAKANO AND H. TAKAHASHI

synchronously when the eggs are immersed in sea water. Availability of a lot of sperm-associated eggs from a single, copulated female allows sam- plings of these eggs in accurate time sequences after immersion of the eggs in sea water. Thus the eggs of the elkhorn sculpin appear to provide a useful model system for the time-course studies of early events occurring during fertilization reaction in teleost fishes. In the present study, ultrastruc- tural features were described on gametes in the course of serial changes from “attachment” through “binding” to the early step of male pro- nucleus formation, with special reference to a characteristic ooplasmic movement having a pos- sible relation with sperm penetration into the ooplasm.

MATERIALS AND METHODS

Adult males and females of the elkhorn sculpin used in the present study were captured by a trammel set along the shore of Usujiri, southern Hokkaido, during spawning season (March-April) in 1989-1991. Mature eggs and sperm were obtained by manually pressing the abdomen of mature fish. Spermatozoa, unfertilized eggs and fertilized eggs sampled at various intervals of time after immersion in sea water at 10°C were fixed with Karnovsky’s fixative for 5 hr at room tempera- ture. For transmission electron microscopy, after being washed with cold 0.1M cacodylate buffer (pH 7.4), the specimens were post-fixed with 1% OsO, in 0.1M cacodylate buffer for 2 hr at 4°C. After dehydration through a graded ethanol series, the specimens were embedded in Epon. Ultrathin sections were prepared with a Reichert-Jung Ultracut E ultramicrotome, stained with uranyl acetate and lead citrate or with lead citrate alone, and examined with a Hitachi H-7000 electron microscope. Epon-embedded eggs were sectioned at about 1m in thickness, and stained with toluidine blue for light microscopy. For scanning electron microscopy, the fixed specimens were washed with 0.1 M cacodylate buffer and dehyd- rated in a graded ethanol series followed by isoamyl acetate. Then the specimens were dried with liquid CO, in a critical point drier. The dried specimens were coated with gold and examined

with a Hitachi S-2300 electron microscope.

RESULTS

Sperm morphology

The mature spermatozoon of the elkhorn scul- pin measures about 37 zm in total length. Its head is of thick spatula-shape, being about 2.9 um long, 1.4 wm wide and 0.8 um thick, and is devoid of acrosomal structure (Fig. 1). The sperm nucleus is composed of homogeneously condensed chroma- tin. On one side of the flattened nucleus, two centrioles are present slightly posterior to the middle of its length. From the distal centriole a groove with an axonema runs posteriorly along the

Fic. 1. sperm head and middle piece. Numerous mitochon-

TEM micrograph of a sagittal section of the

dria (m) are seen in the middle piece. Arrow indicats a cytoplasmic sleeve surrounding the pro- ximal portion of the sperm tail. 15,000.

Sperm Penetration in the Egg of Sculpin 95

long axis of the head. A middle piece conjoins the posterior end of the head, with many small-sized mitochondria aggregating compactly. It is noted that, on the centriolar side of the nucleus, several mitochondria extend their distribution anteriorly along the axonema in the head portion. A short, thin cytoplasmic sleeve surrounds the proximal portion of the sperm tail (Fig. 1 arrows). The flagellum has a typical 9+ 2 arrangement of micro- tubules.

Micropylar apparatus

The mature egg of elkhorn sculpin is globular in form measuring about 1.2 mm in diameter, and is covered with a vitelline envelope of about 40 um thick. A micropylar apparatus is located at the

Fic. 2. (a), SEM micrograph of the animal pole of an egg, Showing a micropylar apparatus consisting of a vestibule and a canal (arrow). 285. (b), TEM micrograph of an inner opening of the micropylar canal. Arrow indicates an ooplasmic projection beneath the micropylar canal. 30,000.

Pie!

animal pole of the egg. The micropyle is composed of a micropylar vestibule forming a shallow de- pression of about 70 ~m in diameter on the cho- rion, and a micropylar canal which opens in the center of the micropylar vestibule (Fig. 2a). The micropylar canal is about 5.5 ~m in diameter at its outer opening, and tapers gradually toward the inner opening of about 1.4 ~m in diameter which allows only one spermatozoon to pass through (Fig. 2b). An ooplasmic projection containing a few ribosomes exists just beneath the micropylar canal, and protrudes into the micropylar canal for about 0.6 um (Fig. 2b).

Morphological changes of the egg after immersion in sea water

In the elkhorn sculpin, sperm is introduced by

Fic. 3. TEM micrograph of the micropylar canal of an egg before immersion in sea water. A spermatozoon is attached to the ooplasmic projection, but a fusion of its plasma membrane with ooplasmic membrane has not yet occurred. 20,000.

96 Y. Koya, K. TAKANO AND H. TAKAHASHI

copulation into the ovarian cavity. In the eggs stripped from copulated females, many spermato- zoa had entered the micropylar canal, and a single spermatozoon was seen to be attached to the plasma membrane of ooplasmic projection in the inner opening of the micropylar canal (Fig. 3). However, fusion between sperm and ooplasmic membrane had not yet occurred at this stage, since no spermatozoon remained on the ooplasmic sur- face around the inner opening of the micropylar canal after the removal of the vitelline envelope from freshly stripped eggs.

Eggs from copulated females were immersed into sea water immediately after stripping. Five sec after immersion, the spermatozoon attached to ooplasmic membrane did not show any change. It was remarked that the ooplasmic projection had disappeared around the attached spermatozoon which came to lie on the flat surface of ooplasm. Membrane fusion between spermatozoon and egg could not be observed yet at that time. Cortical alveoli were distributed closely abutting against the ooplasmic membrane, without showing an actual membrane fusion.

Membrane fusion between spermatozoon and egg had already completed 10-15 sec after immer- sion. Fertilizing spermatozoon penetrated the ooplasm for about half the head length (Fig. 4), and a vesiculation occurred on the plasma mem- brane at the apical region of sperm head fused to ooplasmic membrane. As the vesiculation ad- vanced, the sperm plasma membrane disappeared completely to expose the sperm nuclear membrane to ooplasm. In the ooplasm around the penetrated sperm head, there were vesicular membranous structures which appeared to originate in the vesiculated plasma membrane of spermatozoon and egg (Fig. 4). Part of ooplasm came to pro- trude along the unpenetrated part of sperm head to form the so-called fertilization cone. Furthermore, an ooplasmic region containing no particular cyto- plasmic organelles appeared surrounding the penetrating sperm head (Fig. 4). Membranes of cortical alveoli near the micropylar region were fused to the ooplasmic membrane, and the cortical alveoli were broken down to release their contents into the space between the vitelline envelope and ooplasmic membrane.

Si <3

¢ Seen) OS (ae ton >»: ~ 2 yee

WS } oud KAY a

Fic. 4. Electron micrograph of the egg by 10sec after immersion in sea water. An fertilizing spermato- zoon penetrates the ooplasm and membrane fusion between spermatozoon and egg has begun. Several vesiclar structures are seen near the penetrating sperm head (arrows). 20,000.

Twenty to thirty sec after immersion, the sper- matozoon penetrated the ooplasm for more than half the head length. The membrane of sperm head had completely disappeared (Fig. 5). The ooplasm surrounding the penetrating spermato- zoon contained few cytoplasmic organelles. At this stage, the sperm tail which remained outside of ooplasm was shown to be completely lacking in the typical 9+2 structure of its flagellar microtu- bles, and consisted of numerous fibrillar structures running through its length (Fig. 5). Such a struc- tural change of flagellar microtubles seemed to accompany the progress of fertilization reaction in the penetrating spermatozoon. The breakdown of cortical alveoli was advancing further near the micropylar region, but not in other regions of the cortical cytoplasm of the egg.

Sperm Penetration in the Egg of Sculpin 97

Fic. 5. Electron micrograph of the fertilizing spermatozoon by 30 sec after immersion of the egg in sea water. The flagellar microtubles are degenerated. Sperm mitochondria (m) remain outside the egg. 15,000.

5 mS Loa

Fic. 6. TEM micrographs of eggs by 2 min after immersion in sea water. (a) Sperm head and mitochondria (m) have penetrated the ooplasm. The ooplasm surrounding the penetrating spermatozoon forms an electron dense, triangular zone (ed!) containing lybosomes only. 12,000. (b) The vesicular structures (arrows) are arranged in

a row near the sperm nucleus (n). 45,000.

One to two min after immersion, the spermato- zoon penetrated the ooplasm further up to its middle piece. Sperm mitochondria, the basal part of axial filaments, and sperm nucleus covered with the nuclear membrane were all exposed to oo-

plasm (Fig. 6a). Several vesicles were seen to be arranged in a row in the ooplasm near the sperm head (Fig. 6b). A portion of the ooplasm contain- ing only ribosomes as a significant cytoplasmic organelles formed a rather electron-dense, trian-

98 Y. Koya, K. TAKANO AND H. TAKAHASHI

Fic. 7. Electron micrographs of eggs by 3 min after immersion in sea water. (a) The penetrated sperm nucleus (n) in the ooplasm. The nuclear membrane of spermatozoon becomes vesiculated begining from the anterior region of sperm head (arrows). 60,000. (b) The chromatin of the penetrated sperm nucleus is dispersed. A fragment of nuclear membrane is visible near the centriol (arrow). 60,000.

gular zone expanding from the apex of sperm head to the associated surface ooplasm (Fig. 6a). Three min after immersion, the spermatozoon penetrated the ooplasm much deeper, and an electron dense layer was seen over the sperm head. The nuclear membrane of fertilizing spermatozoon was separated from the chromatin aggregation, and became vesiculated beginning from the ante- rior region of sperm head (Fig. 7a). In that region the chromatin began to disperse into ooplasm, forming a granular structure. In the eggs at this stage of fertilization, granular chromatin, cent- riole, and the basal part of axial filaments of spermatozoon were seen in the ooplasm. A single

Fic. 8. TEM micrograph of an egg by Smin after immersion in sea water. The second polar body undergoing second meiotic division is seen in the perivitellin space. Numerous microtubles are pre- sent in the cytoplasmic bridge. 4,000.

Sperm Penetration in the Egg of Sculpin 99

vesicle which seemed to be a part of vesiculated nuclear membrane was seen in ooplasm near the spermatozoan centriole (Fig. 7b).

Five min after immersion, fertilizing spermato- zoon could not be discerned in ooplasm even by electron microscopy. A second polar body under- going second meiotic division was seen in the perivitelline space near the micropylar region (Fig. 8). The second polar body was connected with ooplasm by a cytoplasmic bridge in which a num- ber of microtubles ran vertically. The surface of the second polar body was complicatedly crum- pled.

DISCUSSION

Fertilization is defined as the fusion of a male and a female gamete followed by nuclear fusion between the two gametes [1]. In the present study on the elkhorn sculpin, it was confirmed that spermatozoa entered the micropylar canal of ovu- lated eggs and were attached to the ooplasmic membrane without showing any membrane fusion between spermatozoon and egg until after the egg had been released into sea water. It is of signifi- cance to refer to such a peculiar reproductive mode as “internal gametic association” in order to distinguish it from the actual oviparity [14].

Spermatozoa of teleostean fishes are different in shape in different species, though they lack an acrosomal structure without exception. Spermato- zoon of the elkhorn sculpin has a relatively long middle piece and a long, flattened head. Sperma- tozoa of a similar shape were reported to occur also in Oligocottus maculosus [20], Cottus han- giongensis and C. nozawae [19], suggesting that such a shape of spermatozoa may be common to fishes belonging to the family Cottidae. However, the spermatozoa of the elkhorn sculpin have a larger number of mitochondria as compared with those of other cottid species. Such a feature may reflect the fact that the spermatozoa of the sculpin can be motile for a very long time in ovarian fluid of the same species (data not shown).

It has been shown in many teleosts that there is a specialized conformation to attach the spermato- zoon to the surface of ooplasm beneath the micropylar canal. Kudo [9] showed that the con-

formation in common carp, Japanese dace, Tribo- lodon hakonensis, and ayu, Plecoglossus altivelis was an ooplasmic projection extending up into the micropylar canal and had a function as a receptor for fertilizing spermatozoon. The existence of such a conformation has been demonstrated also in mummichog [2], chum salmon, Oncorhynchus keta [12], rose bitterling [18], and zebrafish, Brachy- danio rerio [5]. The egg of the elkhorn sculpin also had a cytoplasmic protrusion on the egg surface beneath the micropylar canal, and a spermatozoon came to attach to this protrusion after passing through the canal. In the elkhorn sculpin, how- ever, such cytoplasmic protrusion was observed to exist not only just beneath the micropylar canal but in other regions of ooplasmic surface near the micropyle. The present study could not confirm whether the cytoplasmic protrusion has a function of sperm receptor or not. It has been reported that polyslermy occurs in the medaka [6], and the rose bittering [15] when the eggs have been denuded mechanically. If the eggs of the elkhorn sculpin have a similar nature to those of the medaka and rose bitterling, it seems likely that the cytoplasmic protrusion beneath the micropylar canal does not have any special function for fertilization.

In the eggs stripped freshly from copulated female elkhorn sculpin, many spermatozoa have entered the micropylar canal and a spermatozoon have been attached to the egg surface, but the membrane fusion between spermatozoon and egg does not occur yet. The spermatozoon did not remain to be attached to the egg surface after removal of the vitelline envelope when observed by a scanning electron microscope. This indicates that the attachment between egg and spermato- zoon in the ovarian lumen is relatively loose. In mammals the egg-sperm association includes two successive steps which are referred to as attach- ment and binding [21]. The attachment is a relatively loose, nonspecific association, while the binding is a tenacious, species-specific association [21]. Therefore the attachment of spermatozoon and egg in the ovary of the elkhorn sculpin may correspond to the nonspecific attachment of the gametes in mammals.

Membrane fusion between spermatozoon and egg has been studied in various species of animals.

100 Y. Koya, K. TAKANO AND H. TAKAHASHI

It is known that the inner acrosomal membrane fuses with ooplasmic membrane in several marine invertebrates [4]. On the other hand, in mammals, a fusion between the plasma membranes of sperm head and egg microvilli occurs at the post-nuclear cap region of spermatozoon after the acrosomal reaction has completed [25]. Ultrastructural observations on membrane fusion between sper- matozoon and egg in teleosts which is lacking in an acrosome on sperm head have been carried out in the medaka [6], rose bitterling [15, 17, 18], mum- michog [3], common carp [11], and zebrafish [22]. However, information about gamete membrane fusion in teleosts has been relatively fragmentary as compared with that in other animal groups, possibly because of the fact that the membrane fusion between spermatozoon and egg in teleosts occurs in a short time after gamete association. As mentioned before, in the case of denuded eggs of the medaka [6], the ooplasmic membrane comes to fuse with the sperm plasma membrane in various regions one min after insemination, giving rise to polyspermy. Ohta [17] demonstrated that, in the rose bitterling, the fusion occurred between the microvillus membrane of a sperm entry site on the egg surface and the sperm head membrane at the portion in front of the centrioles. In the elkhorn sculpin, membrane fusion between spermatozoon and egg began by 10sec after immersion in sea water. The membrane fusion occurred with a vesiculation of apical membrane of the sperm head and the ooplasmic membrane, and the vesiculation extended successively to the whole surface of the penetrated sperm head, regardless of the location of centrioles existing in the middle region of sperm head. Therefore, the membrane fusion in the elkhorn sculpin seems to occur in a similar manner to that in the medaka, and begins at the portion having no direct relation to the position of sperm centrioles.

By 30 sec after immersion in sea water, a typical 9+2 axonemal structure of the tail of fertilizing spermatozoon had disappeared. Since the micro- tubles of sperm tail are indispensable for sperm motility, such a change may indicate that the spermatozoon has lost it motility and the sperm tail is undergoing degeneration by this time. The ooplasm surrounding the penetrating spermato-

zoon was initially lacking in significant cytoplasmic organelles exhibited a relatively uniform structure. As the spermatozoon penetration proceeded, this part of ooplasm continued to expand inside to form a triangular region with accumulating ribo- somes. Moreover, several vesicles, which were formed possibly by vesiculations of spermatozoon and ooplasmic membrane, were seen to be arranged in a row near the deeply penetrated sperm head, suggesting that the ooplasm was acting to draw the fertilizing spermatozoon inside into the egg. Such a change of ooplasm was never observed in unfertilized eggs even when they were immersed in sea water. These findings indicate that the ooplasm surrounding a fertilizing sper- matozoon may have a role in active engulfment of the spermatozoon which has lost its motility by degeneration of tail axoneme. Wolenski and Hart [23] showed that the treatment of zebrafish eggs with cytochalasin B or D prevented the incorpora- tion of fertilizing spermatozoon into eggs, and suggested the presence of actin-containing fila- ments in ooplasm which might play a role in the sperm incorporation. In the present study, although the precise mechanism of ooplasmic movement during penetration was not clarified, it is quite possible that ooplasmic microfilaments may be involved in the observed ooplasmic move- ment.

In medaka egg, sperm nuclear membrane is broken down by vesiculation along the lateral aspect of sperm nucleus by 3 min after insemina- tion [6], while in rose bitterling egg, sperm nuclear membrane disappears first at the apical region of the head by 5 min after insemination [18]. In elkhorn sculpin egg, vesiculation of the outer and inner nuclear membrane occurs by 3 min after immersion of the egg in sea water. The vesicula- tion of sperm nuclear membrane started at the anterior region of sperm head and finished at the region near the centriole, like in the case of the medaka and the rose bitterling.

The process of formation of the second polar body in teleostean egg has been described in artificially activated eggs of the rose bitterling [16]. In this fish, the second polar body begins to rise by 15 min after activation, and completes its separa- tion from the egg proper by 30 min later. In the

Sperm Penetration in the

present study, a separating polar body was notice- able ultrastructurally 5 min after immersion of the eggs in sea water. The process of formation of the second polar body was confirmed to be generally similar to that found in the rose bitterling [16].

In present study, we clarified in detail the ultra- structural change of the elkhorn sculpin gametes in early fertilization processes. Moreover, we could suggest an ooplasmic movement for sperm penetration and the destiny of sperm tail during sperm penetration. Information obtained by the present study may be important to solve problems concerning the whole process of fertilization in teleosts.

ACKNOWLEDGMENTS

We wish to thank Dr. H. Munehara, Hokkaido Uni- versity, for his useful suggestions. For sampling of the specimens, we are greatly indebted to the staff of Usujiri Fisheries Laboratory, Hokkaido University. This work

was supported in part by a grant of the Akiyama Founda- tion.

REFERENCES

1 Balinsky BI (1970) An introduction to embryology. W B Saunders, Co, Philadelphia, Pennsylvania, pp 107-137

2 Brummett AR, Dumont JN (1979) Initial stages of sperm penetration into the egg of Fundulus hetero- clitus. J Exp Zool 210: 417-434

3 Brummett AR, Dumont JN, Richter CS (1985) Later stages of sperm penetration and second polar body and blastodisc formation in the egg of Fundu- lus heteroclitus. J Exp Zool 234: 423-439

4 Colwin AL, Colwin LH (1961) Changes in the spermatozoon during fertilization in Hydroides hexa- gonus (Annelida). Il. Incorporation with the egg. J Biophys Biochem Cytol 10: 255—274

5 Hart NH, Donovan M (1983) Fine structure of the chorion and site of sperm entry in the egg of Brachydanio. J Exp Zoo 227: 277-296

6 Iwamatsu T, Ohta T (1978) Electron microscopic observation on sperm penetration and pronuclear formation in fish egg. J Exp Zool 205: 157-180

7 Iwamatsu T, Ohta T (1981) Scanning electron microscopic observation on sperm penetration in teleostean fish. J Exp Zool 218: 261-277

8 Kudo S (1980) Sperm penetration and the formation of a fertilization cone in the common carp egg. Develop Growth & Differ 22: 403-414

9 Kudo S (1982) Ultrastructure of a sperm entry site beneath the micropylar canal in fish eggs. Zool Mag

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Egg of Sculpin 101

91: 213-220

Kudo S (1983) Response to sperm penetration of the cortex of eggs of the fish, Plecoglossus altivelis. Develop Growth & Differ 25: 163-170

Kudo S and Sato A (1985) Fertilization cone of carp eggs as revealed by scanning electron microscopy. Develop Growth & Differ 27: 121-128

Kobayashi W, Yamamoto TS (1981) Fine structure of the micropylar apparatus of the chum salmon egg, with a discussion of the mechanism for blocking polyspermy. J Exp Zool 217: 265-275

Kobayashi W, Yamamoto TS (1987) Light and electron microscopic observation of sperm entry in the chum salmon egg. J Exp Zool 243: 311-322 Munehara H, Takano K, Koya Y (1989) Internal gametic association and external fertilization in the elkhorn sculpin, Alcichthys alcicornis. Copeia 1989: 673-678

Ohta T (1985) Electron microscopic observations on sperm entry and pronuclear formation in naked eggs of the rose bitterling in polyspermic fertilization. J Exp Zool 234: 273-281

Ohta T (1986) Electron microscopic observations on the process of the polar body formation in artificially activated eggs of the rose bitterling. J Exp Zool 237: 263-270

Ohta T (1991) Initial stages of sperm-egg fusion in the freshwater teleost, Rhodeus ocellatus ocellatus. Anat Rec 229: 195-202

Ohta T, Iwamatsu T (1983) Electron microscopic observations on sperm entry into eggs of the rose bitterling, Rhodeus ocellatus. J Exp Zool 227: 109- 119

Quinitio GF (1989) Studies on the functional mor- phology of the testis in two species of freshwater sculpin. Doctoral Dissertation, Hokkaido Univ Stanley HP (1969) An electron microscope study of spermiogenesis in the teleost fish Oligocottus macu- losus. J Ultrastruct Res 27: 230-243

Wassarman PM (1987) The biology and chemistry of fertilization. Science, 235: 553-560

Wolenski JS, Hart NH (1987) Scanning electron microscope studies of sperm incorporation into the zebrafish (Brachydanio) egg. J Exp Zool 243: 259- 273

Wolenski JS, Hart NH (1988) Effects of cytochala- sins B and D on the fertilization of zebrafish (Brachydanio) eggs. J Exp Zool 21: 473-515 Wolenski J S, Hart NH (1988) Sperm incorporation independent of fertilization cone formation in the danio egg. Develop Growth & Differ 30: 619-628 Yanagimachi R, Noda YD (1970) Physiological changes in the postnuclear cap region of mammalian spermatozoa: a necessary preliminary to the mem- brane fusion between sperm and egg cells. J Ultra- struct. Res 31: 486-493

Mi

é =" Siw Ae AAG pe

ti mol inh inet =, ia, seenealh:; Net od %, Pavan - ay a) Nes ried 4 eg mit to 2 17tk

: a TAY at ariel wd Mi yetati a oie 5

1)

USS Ta DON PSR CUT IO (TY ACME BRN ‘ie jt

+ br de TE TU te as bey eal Hh Fe hh gga

‘Sled oAR ah 2 Mineo qpateeeban!

rtiitiil sak LR) Sl area t if! lee vein oe |

am Ais Sd heath JOGA Tebregy ay Woe

seo wit teva orn. sleds Hostels 2; ‘b, i AK TESPAT RTT ros UT s 5 ere eo

Tye ATBeT ip Upto Belin

Ifa a ely aan

arrest “ht arperened obt- fi. Y pple 'as “ie My) ater oe tSeky iu uy it cmd tr» ' ; nieste a nin in FL Rr yiery qa") ot aie) “et /

Oy areas eehey Uden ST pel nap late! ina pereerey leon garelbes he epee emir TAS kD 3 2e Do ten Qeee say

+ a rn | Tl @s0al § 4 ‘ G batt Ee: ae ee eee tee } “ } i re : ' ‘ & ' it i te i J a ae i + oe: FES ; io “st

au} p weeds Pooase CAE Py 1 Pheri (ary “08 pet $ feat Seeeeldkin ered yachts ot * «'%e oe OO Se al é ) ( 4 MM TG Vie fF }. Boies aN > tt T3teu £ nine re pl year hale td . b A ‘ in’ ‘ ‘ie a wide Wy i i Ths I shed F ; j ia ‘ A ‘ . & rk eT ‘ - > ry ' bie Ww r ‘ ' i} h : vi ] 7 t nv : ’ ‘ ‘ ' ? Photo eS’

Angin ie

te) fray: nt

attinile’ evita gutter ty apneic

‘tere ican caries re eon

Tb wor ol de eee Hoot ie, Soe elie vans tix

Ja VLR PY ‘hy fa

is scons sane ote to eredhpees fo ree Haiakbhod ite aaa fad Ais ra wes, SST: Ate

bits Hamewse ant to im

Ay (thts spc " a) Ae uy “Orv mat Pes bi gry et werd

“rs wont, #aoks Reith 7 Guusebonpts +

OF 5 ai ie tyeastt ae aC. Oe

fe, Wis, det! kb Bab ney abe Leaeiati Me ine

(2PRI beak pee : eee

eins) DAT i de sole ogee

Titel Ty gee wii’ rena oON

cht ey! tag fla cat whe pared » tr : 1 LARA OPS CORE Pumocats. ane betmogeionl |

ly tes Shy ANON

nis WGA 328 enact ER o 4) : Hifast a2 ; se Yioterati he! oi Ake ie ils

ih ot rh’ wy DOS thesia F fee tM ei OT Oy eee Sh] ho sstthanies ining

Tdeties bast OL Gat aah

eas a

t rl thers P 192 : PS) Se a i Viren rt eT ve Lovin Urey W Ste trade

ve hemes, Hee aa) onipnery

ZOOLOGICAL SCIENCE 10: 103-109 (1993)

© 1993 Zoological Society of Japan

The Acrosome and Its Differentiation during Spermiogenesis in Halocynthia roretzi (Ascidiacea, Tunicata)

Makoto Fukumoto! and TAKAHARU NUMAKUNAL-

‘Biological Institute, College of General Education, Nagoya City University, Mizuho-ku, Nagoya 467, Japan and *Marine Biological Station, Tohoku University, Asamushi, Aomori 039-34, Japan

ABSTRACT—The spermatozoa of Halocynthia roretzi have architectural features characteristic of ascidian spermatozoa that have been previously described. They have an elongated head (approximate- ly 5.5 4m in length) and a single mitochondrion which is closely applied laterally to the nucleus. They lack a midpiece. An acrosome is present at the apex of the sperm head. The acrosome of H. roretzi spermatozoa is a depressed sphere, approximately 125 x 12540 nm.

At an early stage of spermiogenesis a vesicle (about 50 nm in diameter) appears in a blister at the apex of each spermatid. During spermiogenesis this vesicle becomes larger (approximately 100 nm in diameter). The vesicle in the blister contains moderately electron-dense material located along the vesicle membrane. At the completion of spermiogenesis the vesicle is most probably flattened by the

elongating nucleus and transforms into an acrosome.

INTRODUCTION

Recent studies have shown that the spermatozoa of ascidians have an acrosome(s), albeit a small one [1, 5, 9]. In spite of various studies on fertilization in H. roretzi, there is still a debate over whether or not an acrosome is present in this species.

In Ciona intestinalis, the morphological changes in sperm associated with fertilization are becoming clear. An acrosome reaction occurs via vesicula- tion on the surface of the chorion or after penetra- tion through the chorion. In the perivitelline space, apical processes protrude from the apex of the sperm head. Gamete fusion occurs between these apical processes and egg plasmalemma, re- sulting in the incorporation of the sperm head into the egg via its anterior region [6-8].

In order to develop a satisfactory understanding of the general mechanism of ascidian fertilization, further studies on morphological aspects of ferti- lization in other ascidian species are indispensable.

Accepted July 27, 1992 Received June 12, 1992 " To whom reprint request should be addressed.

In this paper, the authors describe the acrosome and its differentiation during spermiogenesis as a preliminary step in the analysis of morphological aspects of fertilization in H. roretzi.

MATERIALS AND METHODS

The solitary ascidian, Halocynthia roretzi (Type C)[13, 14] was collected near Asamushi Marine Biological Station, Aomori, Japan.

For TEM observations on spermiogenesis and on the differentiated spermatozoa, the testes and the sperm duct were prepared for electrom micros- copy according to a method described by Fukumo- to [2]. Electron micrographs were taken on a Hitachi H-300 electron microscope operated at 75 kV.

RESULTS

The acrosome in Halocynthia roretzi sperm

A fully differentiated spermatozoon of H. roret- zi is approximately 40 4m length. It consists of a head and a tail. It lacks a midpiece. The head (approximately 5.5 um long) contains an elon-

104

gated nucleus and a single mitochondrion which flanks the nucleus (Fig. 1). An acrosome is present at the apex of the head (Fig. 1, arrow).

The fully differentiated acrosome is approx- imately 12512540 nm (Figs. 2, 3 and 4). At the region where the acrosome is located, the inner and outer nuclear membranes are in extremely close contact with each other and form a “pedes- tal” for the acrosome (Figs. 2, arrow and 3).

Acrosome Differentiation during Spermiogenesis

The early spermatid is spherical and contains a round nulcues and a single mitochondrion. The chromatin exists as thin strands which are present throughout the nucleus. A fairly well developed Golgi apparatus with its associated vesicles is pre- sent in the cytoplasm (Fig. 5, arrow). The vesicle contains moderately electron-dense material (Fig. 6).

In spermatids mid-way through spermiogenesis, the nucleus has undergone some antero-posterior elongation. The future apex of the sperm corres- ponds to the blister (Fig. 7, arrow) and the future proximal end of the sperm head corresponds to the region where the flagellum extends out. During this stage the chromatin strands begin to become

M. Fukumoto AND T. NUMAKUNAI

oriented parallel to the antero-posterior axis of the spermatid (Fig.7). At this stage of sper- miogenesis, the blister protrudes out from the apex of the sperm head. Moderately electron-dense material in the vesicle is located along the vesicle membrane (Fig. 8).

In spermatids at later stages of spermiogenesis, the nucleus further elongates antero-posteriorly in a helical configuration. This morphological change of the nucleus is accompanied by a change in the arrangement of the chromatin strands in the nuc- leus; chromatin strands which were arranged para- llel to one another along the longitudinal axis of the nucleus coil up in a definite pattern (Fig. 9). A blister continues to be present in the anterior region (Fig. 9, arrow). A vesicle is present in a blister (Fig. 10). The nuclear membranes at the blister form an indentation, in which both the inner and the outer nuclear membranes are in close contact with each other (Figs. 8 and 10). The plasmalemma enclosing the blister is decorated by fuzzy material on its external surface (Figs. 8 and 10).

At the completion of spermiogenesis, the vesicle in the blister becomes a depressed sphere (Figs. 11 and 12) and transforms into an acrosome. This

Fic. 1.

Sagittal section through the head of a differentiated sperm. An acrosome (arrow) is present at the apex

Fics. 2 and 3.

Fic.

Fic.

Fic.

Fic.

Fic.

Fic.

Fic.

between the plasmalemma and the nuclear membranes. Fuzzy material (F) which decorates the outside of the plasmalemma enclosing the apex is thicker than it is on other parts of the head. M, mitochondrion; N, nucleus. The bar indicates 1 um.

Sagittal and cross sections through the apex of a differentiated sperm head, respectively. An acrosome is present between the plasmalemma and nuclear membrane, where the inner and the outer nuclear membranes make close contact with each other to form a “pedestal” (arrow) for the acrosome. F, fuzzy material. The bar indicates 200 nm, which is also applicable to the Figs. 3 and 4.

4. Frontal slightly oblique section through the apex of the head. An acrosome (A) is recognized as a sphere in the central region of the pear-shaped pedestal. F, fuzzy material.

5. An early spermatid in which a blister is present (arrow). G, Golgi complex; M, mitochondrion; N, nucleus. The bar indicates 2.5 “m (This is also the magnification for Figs. 7 and 9).

6. Enlargement of the blister region. A veiscle which contains electron-dense material is present in the blister. Fuzzy material covers the outside of the plasmalemma enclosing the blister. F, fuzzy material; N, nucleus. The bar indicates 200 nm (This is also the magnification for Figs. 8 and 10).

7. Longitudinal section through a spermatid at a middle stage of spermiogenesis. The nucleus has begun to elongate antero-posteriorly. A blister (arrow) is present at the apex. M. mitochondrion; N, nucleus.

8. Enlargement of a blister region in a spermatid midway through spermiogenesis. A vesicle with a central electron lucent region is present in the blister. The inner and the outer nuclear membranes are in close contact with each other (arrow). Fuzzy material (F) decorates outside of the plasmalemma enclosing the blister.

9. Longitudinal section through a spermatid at a late stage of spermiogenesis where the elongated nucleus has a helical configuration. A blister (arrow) is present at the apex. M, mitochondrion.

10. Enlargement of a blister region in a late stage spermatid. The blister projects more than that at early stages of spermiogenesis. A vesicle is present in the blister. Fuzzy material (F) decorates outside of the plasmalemma enclosing the blister.

105

Acrosome in Halocynthia roretzi

106 M. Fukumoto AND T. NUMAKUNAI

Acrosome in Halocynthia roretzi 107

ae)

and transverse sections through the apex of the head of a sperm that has

Fics. 11 and 12. Sagittal

almost completed maturation. The vesicle in the blister becomes an ellipsoid. Where the vesicle is

morphological change of the vesicle most likely occurs as a consequence of the elongation of the nucleus. Moderately electron-dense material (MEDM) is present in the space between the plasmalemma and the nuclear membranes (Fig. 11). A cross section at the level of the MEDM reveals that this material is located at one side of the sperm head (Fig. 13) below the acrosome (Fig. 0).

DISCUSSION

During acrosome differentiation in other animal species, proacrosomal vesicles derived from Golgi complex(es) coalesce to form one acrosomal vesi- cle. In ascidians, a similar process has been reported. In Pyura haustor and Styela plicata, two vesicles usually appear in a blister at the apex of these spermatids in an early stage of sper- miogenesis. They fuse to form a single acrosome- like structure in the apex of differentiated sperma- tozoa [3]. In Molgula manhattensis, at least three or four vesicles which are moderately electron- dense appear in the blister at the apex of spermatid in early stages of spermiogenesis. They fuse with each other to form an acrosomal vesicle [4]. In H. roretzi, 1n most cases, one vesicle appears in the blister of young spermatids. The vesicle becomes larger during the course of spermiogenesis (com- pare Figs. 6 and 8). This probably occurs via the coalescence of smaller vesicles. These findings are consistent with reports on acrosome differentiation in other animal species. With respect to the origin of the vesicle in the blister, no direct evidence has been found to show that it is derived from the Golgi apparatus. In view of the fact, however, that early spermatids have a fairly well-developed Gol-

located, the inner and the outer nuclear membranes come into close contact with each other to form a “pedestal” (arrow) for the vesicle. Moderately electron-dense material (MEDM) is present in the space between the plasmalemma and the uclear membranes near the apex. Fuzzy material (F) covers the outisde of the plasmalemma enclosing the apex. N, nucleus. The bar indicates 200 nm, which is also applicable to Figs. 12 and 13.

Fic. 13. Transverse section through the head at the level of the MEDM. The MEDM occupies almost half of the space around the nucleus.

108 M. FUKUMOTO AND T. NUMAKUNAI

gi apparatus and Golgi-derived vesicles, it seems safe to assume that they are derived from Golgi vesicles.

Recently, Lambert and Koch [12] have argued that the vesicles in ascidian spermatozoa should be called ‘apical vesicles’ until criteria for their posi- tive identification as acrosomes can be more close- ly met. However, if a vesicle(s) is present in the proper location for an acrosome, this would be one kind of evidence for the existence of an acrosome, because acrosomes are anteriorly located vesicles in the sperm of other animals. Furthermore the vesicle in ascidian spermatozoa is not an oddity which is only found in one or two species, but a feature of all species examined in the Orders of Pleurogonan and Enterogonan ascidians [5, 9]. These are two good reasons for applying the term “acrosome” to these vesicles. We can conclude that spermatozoa of H. roretzi have an acrosome as has been confirmed in other ascidian species [9].

Recently it has been found that caffeine or theophylline induces morphological changes in the acrosome of Ciona intestinalis spermatozoa. These morphological changes have been referred to as an acrosome reaction. The acrosome reaction occurs by fusion between the acrosomal outer-membrane and the plasmalemma enclosing the acrosome. The fusion seems to proceed along the peripheral margin of the acrosome, resulting in vesiclulation [8]. Morphological studies on the acrosome reac- tion in H. roretzi are now in progress.

In H. roretzi, the egg is enclosed by a relatively thick acellular chorion (vitelline coat). For suc- cessful gamete fusion, spermatozoa must pass through the chorion. Sperm might utilize chorion lysin(s) in order to pass through the chorion. Indeed, there is evidence that spermatozoa of H. roretzi contain proteases that act as lysins [10, 15]. However, it is not clear where the lysins are located in spermatozoa. Recent studies on ferti- lization in H. roretzi reveal that spermatozoa can pass through the chorion (vitelline coat) with an intact acrosome and have loose their acrosome in the perivitelline space (Fukumoto, in prepara- tion). This suggests that chorion lysin(s) are associated with the surface of the plasma mem- brane of sperm head. The chemical nature and the precise role of the acrosomal substance remain

obscure.

In H. roretzi, Kubo et al. [11] reported that an acrosome was formed from Golgi-derived vesicles midway through spermiogenesis. They speculated that during further differentiation this acrosome was translocated to the space between the plas- malemma and the nuclear membranes near the apex of the sperm head, where it was recognized as an accumulation of moderately electron-dense material (MEDM). Although the MEDM is not a vesicular structure, they considered it homologous to an acrosome. However, they concluded that the acrosome had dissappeared before the completion of spermiogenesis, because they could not find an acrosome in mature spermatozoa. In this study an MEDM was seen in the same region of spermatids as has been described by Kubo et al. [11] during spermiogenesis (Figs. 11 and 13). In contrast to their descriptions, the MEDM is not homologous to an acrosome (Fig. 11). The precise role and fate of the MEDM remains unknown at present. Although Kubo et al. [11] found the blister at the anterior tip of spermatid, they did not find the vesicular “precusor of the acrosome” in the blister and the acrosome at the apex of the differentiated sperm head, which are described here.

ACKNOWLEDGMENTS

The authors are grateful to Professor Gary Freeman, of the University of Texas at Austin, for his valuable suggestions and for reading the manuscript. The author (M.F.) wishes to thank Professor Kenji Osanai, the director of the Asamushi Marine Biological Station, for facilities. We also wish to thank Mr. S. Tamura and Mr. M. Yashio, of Asamushi Marine Biological Station, for their help in collecting materials.

This work was supported in part by a Grant-in-Aid from the Ministry of Education, Science and Culture, Japan.

REFERENCES

1 Cloney RA, Abbott LC (1980) The spermatozoa of ascidians: acrosome and nuclear envelope. Cell Tiss Res 206: 261-270

2 Fukumoto M (1981) The Spermatozoa and Sper- miogenesis of Perophora formosana (Ascidia) with Special Reference to the Striated Apical Structure and the Filamentous Structures in the Mitochon- drion. J Ultrastruct Res 77: 37-53

10

Acrosome in Halocynthia roretzi

Fukumoto M (1983) Fine Structure and Differentia- tion of the Acrosome-like Structure in the solitary Ascidian, Pyura haustor and Styela plicata. Develop Growth Differ 25: 503-515

Fukumoto M (1985) Acrosome Differentiation in Molgula manhattensis (Ascidiacea, Tunicata). J Ultrastruct Res 92: 158-166

Fukumoto M (1986) The acrosome in ascidians. I. Pleurogona. Int J Invert Reprod Dev 10:335-346 Fukumoto M (1988) Fertilization in ascidians: apical processes and gamete fusion in Ciona intestinalis spermatozoa. J Cell Sci 89: 189-196

Fukumoto M (1990) Morphological aspects of asci- dian fertilization: Acrosome reaction, apical proces- ses and gamete fusion in Ciona intestinalis. Invert Reprod Develop 17: 147-154

Fukumoto M (1990) The Acrosome Reaction in Ciona intestinalis (Ascidia, Tunicata). Develop Growth Differ 32: 51-55

Fukumoto M (1990) Review: Morphological Aspects of Ascidian Fertilization. Zool Sci 7: 989— 998

Hoshi M, Numakunai T, Sawada H (1981) Evidence for participation of sperm proteases in fertilization of the solitary ascidian, Halocynthia roretzi: effects

11

13

14

15

109

of protease inhibitors. Dev Biol 86: 117-121

Kubo M, Ishikawa M, Numakunai T (1978) Dif- ferentiation of apical structures during sper- miogenesis and fine structure of the spermatozoon in the Ascidian Halocynthia roretzi. Acta Embryol Exp 3; 289-295

Lambert CC, Koch RA (1988) Review: Sperm Binding and Penetration during Ascidian Fertiliza- tion. Develop Growth Differ 30: 325-336 Numakunai T, Hoshino Z (1973) Biology of the ascidian, Harocynthia roretzi (Drasche), in Mutsu Bay. I. Differences of spawning time and external features. Bull Mar Biol Stat Asamushi, Tohoku Univ 14: 191-196

Numakunai T, Hoshino Z (1974) Biology of the ascidian, Halocynthia roretzi (Drasche), in Mutsu Bay. II. One of the three types which has the spawning season and the time different from two others. Bull Mar Biol Stat Asamushi, Tohoku Univ 15: 23-27

Sawada H (1990) Sperm penetration through the vitelline coat in ascidian. In “Advances in Inverte- brate Reproduction 5” Ed by M Hoshi, O Yamashi- ta, Elsevier Science Publishers BV (Biomedical Di- vision), pp 111-116

aad DOW itee, aiekiow ala ob areca ead

ee Leta eo By ing, Zeta

(HIOR Tad ae he Dt. cepclidat let Hate. i boty gnitEar icone learn io noes rad ad iy ODay? ary Th sori Ort Pate Amar ayer: EF Aaa SE Ritorrbcmacnnstrinsne inn 8 ern te ohiahia ke ow! offered’: ee a oer reed «hl

Le Aw ra eT Tit Sear Ag gaibinagh

‘Oe tke ee ret wy Bia eed: (ln yantormt Jttel) Spenser Pe ; P ‘a a Oy ier iil

Teer ee ee ioe tat ct ou: fee eek} ics foupet adrenal. f- nos hard tetas Sem wtetaed i HROURA TUE) cel phe, \ roo nine aata rl nein

ihe per we

; ge). Goth). , he

ty tht GA EMG ‘ot ate jaulrbns

bee map sefaing Adare etd! 1eht aides vt nr P 74- oe 4 rad Ol FG ienoG wee (UR ab eee ont grvich Cutter 24 eos slaty y va a fal! nfs APA gt often i a . i 3 *

a: eae

tl bates oy pai eon tedreeialiateate td

Same sMES tassomstey tenet fara) it Mast

¥? ttt e aS

so era e st at ia YY eH * er Le F Nera ital doe ( bites ba tits de Aw prtohe the MED Dati eh Oe te am

tn brn a Dake a

eri eo Pes wes Thy Wy Jitly OF A

72 Ft ly oe a reallly § a6 rr ie pore “4 aN “1

+) (oC aCe sav rm hr) Oy are

‘ti

fu

rt : Lone Oe = = po ‘

Sari gt hal ate fon 7 Vexube a2 y A~ijete “st =) ) iS18a= Pie ri bel ore Big

ay alpey) Tarra: ge

- : | stare rnth] Othe a

oe ee py eins a Py, 4 ; Te - . i

PKR NCR . an i? j rey fi

2) lege pest Pog egy va.

ye Wy her Oe penis 3 (erie Bie

wot 0 7-5

ZOOLOGICAL SCIENCE 10: 111-116 (1993)

Toxicology of Maternally Ingested Carbon Tetrachloride (CCl) on Embryonal and Fetal Development and in vitro Fertilization in Mice

G. P. HAMLIN, S. D. KHOLKUTE and W. R. DUKELOW

Endocrine Research Center, Michigan State University EAST Lansing, MI 48824 U.S.A.

ABSTRACT— Oral administration of carbon tetrachloride (CCl,) for five consecutive days beginning on either day 1, 6 or 11 of pregnancy in mice had no effect on maternal body weight, liver and kidney weight or pregnancy. The various neonatal parameters e.g. pup weight and crown rump length were also not affected. No malformations were detected in any pup on day 1 post-partum. The development of pups (incisor eruption on day 11 and eye opening on day 14) was normal. CCl, at concentrations of 0.05 and 0.5 mM had no significant effect on in vitro fertilization while higher doses caused a significant decrease. A significant linear relationship between 1, 2, 5 and 10mM CCl, and a decrease in the fertilization rate were found. Similarly all the doses of CCl, (1 to 10 mM) resulted in a significant increase in abnormal ovum forms. A significant linear relationship was found between the dose and percent of abnormal forms. Thus CCl, had no in vivo reproductive toxic effects following administra-

© 1993 Zoological Society of Japan

tion of 1/10 and 1/100 LDs 9 dose.

However, addition of CCl, in culture medium in concentrations

found in blood of lethally intoxicated rats, adversely affected the in vitro fertilization rate and caused an

increased incidence in abnormal ova.

INTRODUCTION

Carbon tetrachloride (CCl4) is a colorless vola- tile organic compound commonly used as an in- dustrial solvent, a fumigant and in the manufacture of fluorocarbons. CCl, has been detected in ground and municipal water, air, aquatic organ- isms and food. Thus, it becomes a potential health hazard [7, 17, 21]. CCl, is readily absorbed and distributed to all organs following ingestion or inhalation. The liver is the main organ responsible for biotransformation and detoxification [13, 14]. Evidence for a genotoxic or mutagenic action of CCl, has largely been negative [9, 10].

Reproduction may be sensitive to a number of environmental factors, physical, chemical or emo- tional. Studies of reproductive effects of CCl, in the past have tended to use varying exposure routes and very high doses resulting in maternal toxicity and conflicting results [3, 4, 11, 15, 16].

The present study reports the effects of oral

Accepted October 17, 1992 Received July 27, 1992

ingestion of CCl, on mice for 5 consecutive days beginning on either Day 1, 6 or 11 of pregnancy in B6D2F1 mice. In this strain the first 5 days of gestation are characterized by sequential cleavage of the fertilized oocytes that generates a hatched blastocyst. Implantation is completed on Day 5. Days 6 to 10 are characterized by organogenesis. Finally, Days 11 to 15 are characterized by growth of the fetus. These time periods were chosen in an attempt to isolate detrimental effects on various stages of development and to determine the most susceptible stage for CCl, insult. The effects of CCl, on in vitro fertilization was also evaluated.

MATERIALS AND METHODS

Animals

Female B6D2F1 mice were obtained from one of two sources. Six week old females used for in vivo experiments were obtained from the Jackson Laboratories (Bar Harbor, ME). In some experi- ments, mice were bred on site from parental strains. All animals were housed in a 12 hr light:

112 G. P. HAMuin, S. D. KHOLKUTE AND W. R. DUKELOW

12 hr dark photoperiod at 27°C. Food and water was available ad libitum except where mentioned.

In vivo experiments

Mating procedure: On the evening of their arrival, females were placed with fertile males in a ratio of 1:1 and checked on three subsequent mornings for the presence of vaginal plugs. Most females with plugs were obtained on the third morning as expected from the Whitten effect of induction of estrus [22]. Females were randomly assigned to one of three groups (control, 1/100 LDs0, 1/10 LDs9) on day 1 of pregnancy (=day of plug) and housed in groups of 5 until day 15 at which time they were separated. A minimum of four animals per group per replicate were housed for the two replicates.

CCl, solutions: At room temperature, CCl4 (catalog number CAS 56-23-5, spectropho- tometric grade, Mallinckrodt Int., Paris, KY) was dissolved in corn oil on the first day of treatment and stored at room temperature in light protected, sealed bottles. All solutions were stirred using a teflon coated magnetic bar on each day immediate- ly before use. Animals received either 0.2 ml corn oil day J (control), 1/100 LDso (82.63 mg/kg) or 1/10 LDso (826.3 mg/kg) of CCl, in 0.2 ml corn oil. The LDs9 of 8263 mg/kg was obtained from the Materials Safety Data Sheet OH504310 (Occu- pational Health Servies, Inc., New York, New York). All procedures using CCl, were performed in a certified fume hood.

Experimental protocol: Gavage_ protocol: Females were gavaged by placing a curved gavage needle (size 18, Perfektum, New Hyde Park, NY) attached to a 1 ml glass syringe into the stomach through the esophagus. The volume of 0.2 ml was given per animal.

Maternal parameters: Maternal weight (g) was determined on days 1, 8 and 15 of pregnancy and the 22nd day postpartum. On that day the females were sacrificed by cervical dislocation, the kind- neys and liver removed, blotted dry, and weighed. A portion of the liver and the kindneys (one sliced longitudinally, the other horizontally) were stored in formalin for histological processing.

Neonatal parameters: On the day of birth, litter size was determined and each neonate was

weighed. Crown-rump (C-R) length was measured and individuals were sexed and checked for ob- vious birth abnormalities e.g. extra/missing digits, cleft palate, exencephaly and spina bifida. After recording these parameters, litter size was reduced to 6, where possible leaving equal numbers of males and females.

Neonatal weight and C-R length were measured again on day 8, 15 and 22. In addition, lower incisor eruption and eye opening was assessed in all pups on day 11 and 15 respectively. On day 22, neonate were weaned and two of each sex allowed to grow to maturity (6 wk), the others were sac- rificed by CO, asphyxiation. At 6 wk the four remaining youngs were sacrificed by cervical dis- location and the ovaries or testes removed, blotted dry and stored in Bouin’s fixative.

Statistical analysis: All in vivo parameters were analyzed by one way ANOVA. When statistical significance between control and treatments were found pair-wise analysis of treatment and control was performed using Dunnett’s ¢ test [6].

In vitro experiments

Culture Medium: Brinster’s medium [2] for oocyte culture (BMOC-3, Gibco, Grand Island, New York) was used for incubation of oocytes and sperm in the center well of Falcon organ tissue culture dishes (Becton-Dickinson and Co., #3037, Cockeysville, MD). Alternatively, BMOC medium was prepared in the laboratory and used in the outer well of the tissue culture dishes.

Superovulation: Six to eight-week-old B6D2F1 females were injected s.c. with 8 I.U. pregnant mare serum gonadotropin (PMSG) followed 45-49 hr later with 8 I.U. human chorionic gonadotropin (hCG) s.c. The PMSG and hCG were obtained from Sigma Chemical Co. (St. Louis, MO).

CClz: The BMOC and BMOC-3 media were added to the culture dishes 16 hr before IVF. CCl, was dissolved in t-butanol (Catalog number B 2138, Sigma Chemical Co.) to a concentration 100 times the final concentration in a total volume of 10 ul to be added directly to the 1 ml BMOC-3 medium at the time stated. Final concentrations of CCl, in the center well of organ tissue culture dishes were 0.05, 0.5, 1, 2, 5 or 10 mM.

Gamete recovery and IVF: Twelve hours after

Reproductive Toxicology of CCl, in Mice 113

females received hCG, 2 or 3 males of 6-10 months of age were sacrificed by certical disloca- tion for each IVF experiment and a single epididy- mis was placed directly into the center well of a organ dish containing 0.5 ml BMOC-3. The cauda epididymis was then repeatedly punctured with a 25 G needle before the dish was returned to the incubator for 1-1.5 hr.

Approximately 45 min after sperm collection the superovulated females were sacrificed by cervical dislocation for collection of the cumulus mass containing the oocytes. One side of each repro- ductive tract (ovary, oviduct and uterine horn) was placed in the outer well of the organ dishes con- taining 3ml of BMOC. The cumulus mass was removed from the oviduct under a dissecting microscope, quickly washed and placed in the center well of a treatment or control organ culture dish. The cumulus masses were randomly placed in either treatment or control across replicates to avoid the same female donating to the same treat- ment group.

The organ dishes were placed in the incubator after cumulus mass addition while the CCl, solu- tions were prepared. CCl, solutions were added to the organ culture dish containing the cumulus mass, with care to avoid directly adding the 10 yl CCl, or 10 wl t-butanol (control) on top of the cumulus mass.

Immediately after the addition of the cumulus mass, 50 sl of sperm suspension was added directly to the center well of the dishes taking care that each replicate was inseminated from a different, randomly selected male. All treatment and control dishes were replaced in the incubator for 26 hr before assessment of 2 cell oocytes (which repre- sent fertilized eggs) was conducted.

Statistical analysis: All statistical analysis of IVF experiments used Chi square and Bonferronni Chi square contingency tables [6].

RESULTS

In vivo experiments

Maternal Parameters: Replicate treatement data sets were pooled for tests of statistical signi- ficance. A total of 60 female B6D2F1 mice were

mated with a pool of fertile male B6D2F1 mice. Thirty-five females were designated as pregnant due to the presence of a plug. Of these 31 were assigned to experimental groups. All 31 treatment females maintained pregnancy to term and a total of 294 pups were delivered. Results are expressed as the mean+SEM.

No significant difference (P>0.1) in maternal body weight across treatments on the days meas- ured were seen. Dams were sacrificed on day 22 postpartum (day 1=day of parturition) and liver and kidney weights were recorded, body weight was also recorded to enable calculation of relative organ weights. Relative and absolute liver and kidney weights and maternal weights were not significantly different (P>0.1) across treatment groups (data not shown). Tissue samples of liver and kidney were prepared and preserved in forma- lin for possible histopathological examination at a later date.

Neonatal parameters: Pup weight and C-R length were measured on days 1, 8, 15 and 22 postpartum. A significant difference between con- trol and treated groups was seen on day 15 post- partum (data not shown). No other significant differences were seen across treatment groups in pup weight or C-R length.

No malformations were detected in any pup on day 1 postpartum (observations not presented). The normal development of pups was assessed from observation of incisor eruption on day 11 and eye opening on day 14.

In vitro experiments

CCl, was added directly to the center well of organ culture dishes, containing the cumulus be- fore addition of sperm, at concentrations of 0.05, 0.5, 1,2, 5 and 10mM. The total number of two cell embryos, total number of ova and percent fertilization for all CCl, trials are shown in Table 1.

The vehicle for delivery of CCl, was t-butanol and the percent fertilization in 1% t-butanol was not significantly different from the control. CCl, at concentrations of 0.05 and 0.5 mM had no signi- ficant effect on fertilization while 1, 2, 5 and 10 mM CCl, caused a significant decrease in fertiliza- tion. A significant linear relationship (P<0.001) between 1, 2, 5 and 10 mM CCl, and a decrease in

114 G. P. HaAmuin, S. D. KHOLKUTE AND W. R. DuKELOwW

TABLE 1. Effects of CCl, (mM) and t-butanol (v/v) On mouse in vitro fertilization

Dose (mM) 2 cell total ova % fertilization Control 298 321 92.8 1% t-butanol 215 240 89.6 0.05 179 197 90.9 0.5 169 186 90.9 1.0 49 109 45.0* 2.0 29 94 30.9% 5.0 23 75 MLS 10.0 9 75 12.0*

* significantly different from t-butanol, P<0.001

TaBLeE 2. Effect of CCl, (mM) on ovum morphol- ogy in mouse in vitro fertilization

Dose (mM) abnormal total ova % abnormal 1% t-butanol 0 175 0

1.0 9 109 8.0° 2.0 15 94 16.0* 5.0 69 105 65.7* 10.0 53 100 53.0°

* significantly different to t-butanol, P<0.01 > significantly different to t-butanol, P<0.001

fertilization was found.

In the presence of 1, 2, 5 and 10mM CCl, abnormal ovum forms were observed. The num- ber of abnormal forms, total number of ovum and percent abnormal forms are listed in Table 2. All doses of CCl, resulted in a significant increase in abnormal ova forms and a significant linear rela- tionship (P< 0.001) was found to exist between the dose of CCl, and percent of abnormal ova forms.

DISCUSSION

The results of the present study demonstrates that oral administration of CCl, on days 1 to 5, 6 to 10 and 11 to 15 of gestation had no adverse effects on reproduction of mice. Furthermore, there were no overt signs of maternal toxicity.

Exposure of pregnant mice to 1/10 LDso and 1/ 100 LDs9 (which is equivalent to 826.3 mg/kg and 83.63 mg/kg or 0.52 ml/kg and 0.05 ml/kg, re- spectively) over the first five days of gestation had no effect on fetal mortality and development or

postnatal normality and development. The aver- age litter size of all groups was not statistically significant and no stillbirths were observed indicat- ing that CCl, had no effect on implantation, re- sorption or induction of any embryo lethal effects. Average C-R length of pups measured on days 1, 8, 15 and 22 postpartum was not significantly different across treatments. No significant differ- ence in average pup weight was seen across treat- ments except on day 15 when only the high dose was significantly different from the control. This difference is not interpreted as biologically signi- ficant as no other difference in pup weight was seen and the day 15 pup weights within litters were more dispersed than at other times. Also, C-R length is a more sensitive measure of neonatal growth and no differences were seen in this para- meter.

Recent reviews [8, 18] have reported that expo- sure to chemicals during the preimplantation period can lead to teratogenesis. Methyl nitro- sourea has been shown to act on preimplantation embryos resulting in teratogenesis visualized at birth [3]. This study provides evidence that mater- nal exposure during the preimplantation period to CCl, at 1/10 LDso or 1/100 LD59 does not result in teratogenesis in mice.

In conclusion, this study clearly demonstrates that exposure of maternal mice by oral gavage to 1/10 LDso and the 1/100 LDso of CCly once a day from day 1 through 6 of gestation has no embryotoxic, embryolethal or teratogenic effects. These results agree quite well with the literature in that CCl, induction of embryotoxic and or embryolethal effects occur at concentrations where significant overt maternal toxicity is seen [4, 16]. However, this report is significant in that no other studies were found where the effects of CCly were determined during the preimplantation period and most previous studies demonstrated high levels of maternal toxicity.

An additional set of experiments examined the effect of CCl, on fertilization utilizing the mouse in vitro fertilization system. In these experiments CCl, at 0.05 mM and 0.5 mM was found to have no statistically significant effect on the percent of fertilization. While the concentrations of 1mM through 10 mM had a markedly significant effect

Reproductive Toxicology of CCl, in Mice 115

on fertilization. A linear relationship between dose and inhibition of fertilization was established with the lowest effective dose of 1 mM inhibiting fertilization by approximately 50%. Interestingly, the inhibition of fertilization was not just marked by a reduction in 2 cell embryos and an increase in one cell ova. An ovum exhibiting a very dense constricted cellular/nuclear region and a dust like circle of degraded cell constituents surrounding the dense body, was observed. The appearance of these forms was statistically significant at all doses of CCl, from 1 mM to 10 mM and a dose-related increase in these forms was established.

The results of this IVF study celarly demonstrate a dosedependent effect of CCl, indicating that the effect seen is a direct solvent effect and not a specific one. However, Dahlstrom-King et al. [5] studied eight chlorinated hydrocarbons in the iso- lated rat hepatocyte system and found that toxicity did not relate to physiochemical properties (no- tably the oil/water partition coefficient) of the chemicals. In fact, the authors found that the in vitro toxicity did not correlate well with the in vivo toxicity of the eight chemicals. The appearance of the abnormal ovum in this study corresponds well with morphological changes in hepatocytes re- ported to be due to solvent action [1]. However, it is interesting to note that at 1mM CCl, a 50% reduction in fertilization occurs with only a small highly variable percent of abnormal forms seen, while at 10 mM CCl, there is an approximate 80% reduction in fertilization with a more consistent 58% appearance of abnormal forms. This could indicate that at the lower effective levels of 1 mM and 2 mM CCl,, there are other mechanisms apart from the solvent effect in operation resulting in the inhibition of fertilization.

One of the mechanisms could be a disruption of Ca** homeostasis which is critical to fertilization and proposed as a mechanism of CCl, induced hepatotoxicity [1, 14, 19]. Obviously more re- search is required to determine the effect of CCl, on the IVF system and perhaps some insight into the general mechanism of CCl, cytotoxicity may be gained.

Although, our in vivo studies failed to reveal any delerious effects of CCl, on pregnancy, the dos- ages employed were low. Nonetheless Kim et al.

[12] reported that a single oral dose of 25 mg/kg CCl, in the rat can result in a plasma concentration of 0.3mM. Furthermore, Srivastava et al. [19] have reported concentrations of 0.5-2.5 mM CCl, in the blood of intoxicated animals. Thus a severe decrease in fertilization rate may occur if the ovum were exposed to levels of CCl, above 1 mM and this in turn may adversely affect pregnancy follow- ing high dose chronic exposure.

ACKNOWLEDGMENTS

This work was supported in part by NIH grants HD07534 and ES04911. Appreciation is expressed to Mrs L. M. Cleeves for preparation of this manuscript.

REFERENCES

1 Berger ML, Sozeri T (1987) Rapid halogenated hydrocarbon toxicity in isolated hepatocytes is medi- ated by direct solvent effects. Toxicology 45: 319- 330

2 Brinster RL (1971) In vitro culture of the embryo. In “Pathways to contraception” Ed. by AI Sherman, CT Thomas, Springfield, IL, pp 245-277

3 Cilman MR (1971) A preliminary study of the teratogenic effects of inhaled carbon tetrachloride and ethyl alcohol consumption in the rat. Diss. Abstr. Int. B32: 2021-2042

4 Clemedson C, Schmid B, Walum E (1989) Effects of carbon tetrachloride on embryonic development studied in the postimplantaion rat embryo culture system and in chick embryos. Toxic. in Vitro 3: 271- 275

5 Dahlstrom-King L, Couture J, Lamoureaux C, Vail- lancourt T, Plaa A (1990) Dose-dependent cytotox- icity of chlorinated hydrocarbons in isolated rat hepatocytes. Fund Appl Toxicol 14: 833-841

6 Gill JL (1987) Design and analysis of experiments in the animal and medical sciences Vol 1. The Iowa State University Press, Ames, Iowa, pp 117-132

7 Gilli GR, Bono R, Scursatone E (1990) Volatile halogenated hydrocarbons in urban atmosphere and in human blood. Arch Env Health 45: 101-106

8 Jannaccone PM, Bossert N, Connelly C (1987) Dis- ruption of embryonic and fetal development due to preimplantation chemical insults: a critical review. Am J Obstet Gynecol 157: 476-484

9 International Agency for Research on Cancer (IARC) (1987a) Genetic and related effects: an updating of selected IARC Monographs from Volumes 1 to 42. IARC Monogr Eval Carcinog Risk Chem Hum Suppl 6: 136-137

10

11

12

13

14

15

16

116

International Agency for Research on Cancer (IARC) (1987b) Overall evaluations of carci- nogenicity: an updating of IARC monographs Volumes 1 to 42. [ARC Monogr Eval Carcinog Risk Hum Suppl 7: 143-144

Khominska ZB (1974) Change in function of ovaries and adrenal glands in rats under experimental toxic hepatitis Fiziologicheskii Zhurnal 20: 747-751

Kim HJ, Odend’hal S, Bruckner J (1990) Effect of oral dosing vehicles on the acute hepatotoxicity of carbon tetrachloride in rats. Toxicol App! Pharma- col 102: 34—49

Neal RA (1980) Metabolism of toxic substances In “Toxicology: the basic science of poisons” Ed by J Doull, CD Klassen, MO Amdur, 2nd ed, MacMillan Publishing Co Inc, New York, pp 56-70

Recknagel RO, Glende E, Dolak J, Waller R (1989) Mechanisms of carbon tetrachloride toxicity. Phar- mac Ther 43: 139-154

Roschlau VG, Rodenkirchen H (1969) Histologiche Untersuchungen uber die Diaplazentare Wirkung von Tetrachlorkohlenstoff und Allylalkohol auf Mausefeten. Exp Path 3: 255-263

Schwetz BA, Leong B, Gehring P (1974) Embryo and fetotoxicity of inhaled carbon tetrachloride, 1, 1-dichloroethane and methyl ethyl ketone in rats.

G. P. Hamuin, S. D. KHoLKuTE

17

18

19

20

21

22

AND W. R. DUKELOW

Toxicol Appl Pharmacol 28: 452-464

Silkworth JB, McMartin D, Rej R, Narang R, Stein V, Briggs R, Kaminsky L (1984) Subchronic expo- sure of mice to Love Canal soil contaminants. Fund Appl Toxicol 4: 231-239

Spielmann H, Vogel R (1989) Unique role of studies on preimplantation embryos to understand mechanisms of embryotoxicity in early pregnancy. Crit Rev Toxicol 20: 51-64

Srivastava SP, Chen N, Holtzman J (1990) The in vitro NADPH-dependent inhibition of CCl, of the ATP-dependent calcium uptake of hepatic micro- somes from male rats. J Biol Chem 265: 8392-8399 Takeuchi IK (1984) Teratogenic effects of methyl nitrosourea on pregnant mice before implantation. Experientia 40: 879-881

Wallace LA, Pellizzani E, Hartwell T, Sparacino C, Whitmore R, Sheldon L, Zelon H, Perritt (1987) The TEAM Study: personal exposures to toxic substances in air, drinking water and breath of 400 residents of New Jersey, North Carolina and North Dakota. Env Res 43: 290-307

Whitten WK (1958) Modification of the oestrous cycle of the mouse by external stimuli associated with the male. Changes in the oestrous cycle deter- mined by vaginal smears. J Endocrinol 17: 307-313

ZOOLOGICAL SCIENCE 10: 117-125 (1993)

Reproductive Characteristics of Precociously Mature Triploid Male Masu Salmon, Oncorhynchus masou

Masaru NaKaAmural/, Fumio Tsucuiya?, MAsAo IwAHASHI? and YosHITAKA NAGAHAMA®

‘Department of Biology, Faculty of Medicine, Teikyo University, Hachioji, Tokyo 192-03, *Niigata Prefecture Inland Water Fisheries Experimental Station, Koide-cho, Kitauonumagun, 946, and *Laboratory of Reproductive Biology,

National Institute for Basic Biology, Okazaki, 444 Japan

ABSTRACT —In order to clarify the reproductive characteristics of triploid males, changes in serum steroid hormone levels and testicular development in precociously mature triploid males of masu salmon, Oncorhynchus masou, were examined during sexual maturation. Testicular development in triploids differs from that in diploids. The gonadosomatic index (GSI) in triploids in August was high (3.9%), similar to that in diploids (3.0%). Although germ cells at all stages of spermatogenesis were observed in the testes of diploids, the testes of triploids consisted of many cysts of primary spermatocytes and some degenerating germ cells. In the spawning season in October, GSI in triploids dropped suddenly (1.0%), though GSI in diploids increased (4.8%). At that time, a large number of degenerating figures of spermatogenic germ cells were found in the testes of triploids. Very few sperm were seen in the sperm fluid. However, most of the sperm cells revealed abnormalities in morphology, such as two tails, two heads, and heads of various sizes. Changes in serum testosterone (T), 11-ketotestosterone (11-KT) and 17a,206-dihydroxy-4-pregnen-3-one (17a,20-diOH) in triploid males were similar to those in mature diploid males. T and 11-KT levels were high in August and October, then they dropped in November after the spawning season. 17a,20-DiOH levels in both triploid and diploid males were only high in October during the spawning season. From these results, it seems likely that the cause of abnormality of spermatogenesis in triploid males is mechanical difficulty involved in chromosome separation at meiosis I due to triploidy per se, rather than the reduced levels of testicular steroid hormones. ;

© 1993 Zoological Society of Japan

INTRODUCTION

Recently it has become possible to manipulate chromosome sets by means of physical or chemical treatment of eggs at an early developmental stage [7, 12]. This technique of chromosome set man- ipulation is expected to be applicable to practical fish culture in a variety of ways, such as sex determination or improvement of breed [7]. Tri- ploid fish can be induced by retaining the second polar body of eggs just after fertilization. Triploid fish are thought to become sterile as a consequence of having an odd number of chromosome sets. This sterile triploidy may bring about benefits to

Accepted November 4, 1992 Received September 18, 1992

practical fish culture for the reason that sterile fish avoid the muscle deterioration and death which accompany sexual maturation [10].

Induction of triploidy has been attempted in several fishes so far. However, characteristics of sexually mature triploids depend on sex and spe- cies. In a previous paper, we reported the gonadal development and changes of steroid hormones in triploid female rainbow trout, Oncorhynchus mykiss [6]. In the present study, we deal with testi- cular development and changes in steroid hor- mones in the triploid male masu salmon, Onco- rhynchus masou.

MATERIALS AND METHODS

Land-locked masu salmon (Yamame) used in

118

the present study were obtained from the Niigata Prefecture Inland Water Fisheries Experimental Station. They have been cultivated systematically in ponds. Mature eggs and sperm were obtained from randomely selected 2-year-old females and males. To induce triploidy, just after artificial insemination, eggs were immersed in warm water at 27-31°C for 10-15 min. After this temperature shock, eggs were removed and put into water at about 12°C. Fry were fed on the regular diet for salmon culture from the first feeding.

Precociously mature triploid males (10 to 13 month-old) were autopsied 3 times, August 30, October 22, and November 25, 1986. For each sample, body weight, body size, and testicular weight were measured. To confirm triploidy, a blood smear from each individual was prepared. After staining with 1% Giemsa solution, long diameters of erythrocytes were measured. Testes of triploids were fixed in Karnovsky’s solution for 2 hr. Next, they were postfixed in 1% OsO, buf- fered to pH 7.4 with 0.1 M cacodylate buffer for 2 hr. After dehydration, they were embedded in epoxy resin. For histological observations, one micrometer sections were stained with 1% toluidine blue in 0.1 M phosphate buffer.

For the investigation of sperm, sperm fluid was collected in the October spawning season. Smears of sperm fluids were prepared and stained with Delafield’s hematoxilin.

OS os

Fic. 1.

© O08

ee of diploid (A) and spied (B) masu eae

M. NAKAMURA, F. TsucuiyA et al.

For the measurement of steroid hormone levels in serum, blood from each fish was collected by cutting the caudal region. The serum was sepa- rated by centrifugation, frozen at —20°C, and stored until use. Serum levels of testosterone (T), 11-ketotestosterone (11-KT) and 17a,20¢- dihydroxy-4-pregnen-3-one (17a,208-diOH) were measured by radioimmunoassay [3, 8, 13, 14].

Precociously mature diploid males were sac- rificed in the same manner and at the same time, and they served as the control.

Data were analyzed using Student’s +-test (erythrocyte diameter) and two-way analysis of variance (ANOVA, GSI and body weight), accepted at P<0.05.

RESULTS

Diameter of erythrocytes

Erythrocytes of diploid and triploid males are pictured in Figs. 1A and 1B, respectively. Dia- meter of erythrocytes of triploids is 19.22+0.81 ym. This value is significantly larger (P<0.01) than that of diploids (14.62+0.95 um). Triploids were distinguished from diploids by the size differ- ence in erythrocytes.

Growth and external view of fish

Body weight and total length of triploid and diploid males is shown in Table 1. Diploid and

ON

ao ©

as Ss

x 610.

Triploid Male Masu Salmon 119

TaBLe 1. Comparison of body weight (B.W.) and total length (T.L.) of triploid and diploid males in each sampling time

Aug. 30 Oct. 22 Nov. 25

Triploid B.W. (g) T.L. (cm) Diproid B.W. (g) T.L. (cm)

36.743.1 48.1+2.3 51.3+3.0 14.5+0.2 16.7+0.4 17.6+0.3 32.7+2.5 51.3+3.7 55.4+5.0 14.7+0.3 17.4+0.3 18.0+0.5

Each value is the mean+SE.

triploid males grew similarly and showed no differ- ences in body weight and total length at any sampling time.

Triploid individuals were not different from diploids in external view (Fig. 2). Male secondary sex characters such as dark color on the skin of body and each fin appeared normally in triploids in the October spawning season, similar to diploid males. About 70% of triploid males matured precociously one year after fertilization. This value was almost the same as that of diploid males.

Changes in gonadosomatic indices (GSI=(gonad weight/ body weight) x 100)

Changes in GSI in diploid and triploid males are shown in Figure 3. Changes in GSI in triploids

differ from those in diploids. GSI in triploids was high, 3.9+0.6%, in August, as was GSI in diploids (3.0+0.9%). In October, GSI in diploids in- creased further to 4.8+0.8%. In contrast, GSI in triploids dropped markedly to 1.0+0.7%, even in the spawning season (P<0.01). GSI in diploids dropped rapidly (1.5+0.2%) in November after the spawning season. GSI in triploids dropped even more (0.2+0.01%) (P<0.01).

Gonadosomatic index (%)

Aug. 30 Oct.22 Nov.25

Fic. 3. Changes in the gonadosomatic index (mean+ SE) of diploid (---@---) and triploid (—™—) masu salmon. *Significantly different from diploid (P< 0.01).

Fic. 2. The secondary sex character of dark skin appears in both diploid (D) and triploid (T) male masu salmon.

M. NAKAmuRA, F. TsucuiyA et al.

=

250 AL

y

ie

. Y 7, Ys

EX

2 a

5

ee ie = res Utne Mamaareeencgt oe NS

’

Se

Diploid testes in August (A),

October (D and E) and November (G). 590.

>

Fic. 4. Photomicrographs of testes from diploid and triploid male masu salmon. October (C) and November (F). Triploid testes in August (B)

Triploid Male Masu Salmon 121

Testes

Cysts of spermatogenic germ cells at various stages were contained in the testes of diploids in August (Fig. 4A). A small amount of matured sperm had already been seen in the central region of the lumen. Active spermatogenesis was also seen in the testes of triploids (Fig. 4B). However, spermatogenic germ cells consisted mainly of pri- mary spermatocytes (Fig. 4B). Secondary sper- matocytes, spermatids, and sperm were not observed in triploids, although some clusters of degenerating spermatocytes were observed in the lumen (Fig. 4B). The testes of triploids in the spawning season in October were noticeably dif- ferent in size and color from those of diploids (Fig. 5). Testes of diploids were white in color and were packed with matured sperm (Fig. 5). In contrast, those of triploids were semitranslucent. Sperm fluid filled the efferent ducts. The testes of triploids (Figs. 4D and 4E) were also extremely different in histology from those of diploids (Fig. 4C). The amount of spermatogenic germ cells was markedly less in triploids (Figs. 4D and 4E) than in diploids (Fig. 4C). Moreover, these germ cells revealed necrosis. Some necrotic cells were ingested by hypertrophied Sertoli cells lining the

inner wall of the lumen (Fig. 4D). Although some sperm were recognized in sperm fluid in triploids (Fig. 6B), they were fewer in number and notice- ably different in morphology from those of dip- loids (Fig. 6A). Sperm of triploids were morpho- logically abnormal, such as having two heads, two tails, or various sizes of heads. Large amounts of sperm still remained in the testes of diploids in November (Fig. 4F). In the testes of triploids, the inner wall of the lobules contained only some spermatogonia (Fig. 4G).

Changes in serum steroid hormone levels

Changes in T, 11-KT, and 17a,208-diOH levels in triploids and diploids are shown in Figure 7A- 7C. High levels of serum T (diploid, 8.0 ng/ml; triploid, 7.4ng/ml) and 11-KT (diploid, 5.7 ng/ ml; triploid 5.0 ng/ml) were observed when the two groups of males were compared in August. Levels of these steroids rose further in October (diploid, T 37.0 ng/ml and 11-KT 48.1 ng/ml; tri- ploid, T 30.8 ng/ml and 11-KT 33.3 ng/ml). T and 11-KT levels rapidly decreased in November after the breeding season in both diploid (T 0.4 ng/ml and 11-KT, not detectable) and triploid (T 0.5 ng/ ml and 11-KT 0.4ng/ml). Serum 17a@,206-diOH was not detectable in either diploids or triploids in

Fic. 5.

Testes from diploid (D) and triploid (T) male masu salmon in October.

122 M. NAKAmuRA, F. Tsucuiya et al.

B a <5 ae

Fic. 6. Photomicrograph of sperm from diploid (A) and triploid (B) masu salmon. Small head (small arrow) and, two heads and two tails sperm (large arrow) are seen. 1040.

August. A sharp increase of this steroid occurred in October in diploids (8.7 ng/ml) and triploids (15.9 ng/ml). 17a,208-DiOH dropped in Novem- ber (diploid 0.7 ng/ml and triploid 0.2 ng/ml).

DISCUSSION

Morphological characteristics of testicular de- velopment in triploid males have been reported in several fishes so far. In most cases, abnormalities of spermatogenesis were observed. Despite well-

Triploid Male Masu Salmon 123

{ { A

Testosterone ( ng/ml )

Aug. 30 Oct. 22 Nov.25

11-Ketotestosterone (ng/ml)

Aug.30 Oct.22 Nov.25

(ng/ml )

17a,20B -Dihydroxy—4—pregnen-3-one

Oct.22

Aug. 30 Nov.25

Fic. 7. Changes in serum testosterone (A), 11- ketotestosterone (B) and 17a,20@-dihydroxy-4- pregnen-3-one (C) of diploid (---@---) and triploid (—m_) male masu salmon.

developed testes, matured spermatozoa were not produced in channel catfish (Ictalurus punctatus) [17], nor in Atlantic salmon (Salmo salar) [2] or in

loach (Misgurnus anguillicaudatus) [11]. A few sperm were formed in the testes of some species such as Biwa gudgeon (Gnathopogon caerulescens) [16], rainbow trout (Oncorhynchus mykiss) [4], and Pacific salmonids [1]. The cause of the inabil- ity to produce sperm normally in triploids is thought to be the disruption of meiosis due to the odd number of chromosome sets [12].

In the present study, it is clearly shown that the process of testicular development in triploid males is remarkably different from that of diploids. De- spite the spawning season, GSI in triploids was markedly low. Sperm in triploids was less in amount and contained cells with morphological abnormalities, such as two heads, two tails, or different sized heads. In addition, secondary sper- matocytes and spermatids were not found, though a large number of primary spermatocytes were formed. These facts indicate that the transforma- tion from primary spermatocytes to secondary spermatocytes was suspended. Thus, it seems likely that the large amount of degenerating germ cells that appeared in the testes of triploids in October originated from the primary spermato- cytes which failed to be transformed into second- ary spermatocytes.

A few papers have dealt with the relationship between testicular development and steroid hor- mones in triploid fishes [1, 4, 5]. In triploid males, blood steroid hormone levels have been found to be high, just as is the case for mature diploid males. In the triploid males of masu salmon, changes of serum T, 11-KT and 17a,206-diOH showed no significant differences in comparison with those of mature diploid males. From these results, it is concluded that the production of steroid in triploid males occurs similarly to those in diploid males, even though the development of germ cells is hindered during the process of sper- matogenesis. Thus, it is presumed that abnormali- ties of testicular development in triploids are re- lated to mechanical difficulties involved in chromo- some separation at meiosis I due to triploidy per se, rather than the reduced level of steroid hor- mones.

On the other hand, production of steroid hor- mones and ovarian development in_ triploid females differs among species. Blood steroid hor-

124

mone levels in female rainbow trout [4, 6] and female masu salmon (Nakamura ef al., unpub- lished data) were extremely low, even in the spawning season. The ovaries of these female triploids were composed of numerous cysts con- taining many small oocytes and degenerating oocytes. In contrast, triploid female of tilapia [9] revealed high levels of steroid hormones and had well-developed yolky oocytes in their ovaries. Thus, it is still not known why the reproductive characteristics in triploid fish are different between males and females and among species.

Serum 17a,206-diOH levels in triploid males increased only in the spawning season. Changes in this steroid are similar to those of diploid males in salmonids [9, 18], but there are no reports on the production of 17a,208-diOH in triploid males of other species. This steroid is thought to be in- volved in spermiation [13]. 206-Hydroxysteroid dehydrogenase (206-HSD), which is the enzyme essential for the conversion from 17a-hydroxy progesterone to 17a,206-diOH, was found to be localized on sperm [14]. Thus, it is interesting to note that 17a,208-diOH production in triploids was as high as it was in diploids, though the amount of sperm in the testes of triploids was reduced. The testes of triploid males may provide a suitable material for the study of 17a,20-diOH production.

ACKNOWLEDGMENTS

We would like to thank Dr. J. Specker, Department of Zoology, The University of Rhode Island, for reading the manuscript.

REFERENCES

1 Benfy TJ, Dye HM, Solar II, Donaldson EM (1988) The reproductive endocrinology of triploid Pacific salmonids. Western Regional Conference on Com- parative Endocrinology p 30

2 Benfy TJ, Shutterlin AM (1984) Growth and gonad- al development in triploid landlocked Atlantic sal- mon (Salmo salar). Can J Fish Aquat Sci 41:1387- 1392

3 Kagawa H, Young G, Adachi S, Nagahama Y (1982) Estradiol-178 production in amago salmon (Oncorhynchus rhodurus) ovarian follicle: Role of the theca and granulosa cells. Gen Comp Endocri-

10

11

12

13

14

15

16

M. Nakamura, F. Tsucuiya et al.

nol 47: 440-448

Lincoln RF, Scott AP (1984) Sexual maturation in triploid rainbow trout, Salmo gairdneri Richardson. J Fish Biol 25: 385-392

Nakamura M, Nagahama Y, Iwahashi M, Kojima M (1985) Reproduction of triploid rainbow trout, Salmo gairdneri. Proceedings of the 10th Annual Meeting of Japan Society of Comparative Endocri- nology p 18

Nakamura M, Nagahama Y, Iwahashi M, Kojima M (1987) Ovarian structure and plasma steroid hor- mones of triploid female rainbow trout. Nippon Suisan Gakkaishi 53: 1105

Onozato H (1987) Senshokutai Kougakuteki Shu- houniyoru Ikushu. In: “Kaiyugyo no Seibutugaku” Ed by K Morisawa, K Wada, T Hirano, Japan Science Societies Press, Tokyo, pp 235-251 (in Japanese)

Sakai N, Ueda H, Suzuki N, Nagahama Y (1989) Steroid production by amago salmon (Oncorhyn- chus rhodurus) testes at different developmental stages. Gen Comp Endocrinol 75: 231-240

Satoh S, Tomita M,Iwahashi M, Nakamura M (1991) Reproduction of triploid tilapia, Oreochro- mis niloticus. Scientific Reports of Niigata Prefec- ture Inland Water Fisheries Experimental Station. 16: 73-81 (in Japanese)

Solar II, Donaldson EM, Hunter G (1984) Induc- tion of triploidy in rainbow trout (Salmo gairdneri Richardson) by heat shock, and investigation of early growth. Aquaculture 42: 57-67

Suzuki R, Nakanishi T, Oshiro T (1985) Survival, growth and sterility of induced triploids in the cyprinid loach Misgurnus anguillicaudatus. Bull Jap Soc Sci Fish 51: 889-894

Thorgaard GH (1983) Chromosome set manipula- tion and sex control in fish. In “Fish Physiology Vol IXB” Ed by WS Hoar, EM Donaldson, Academic Press, New York, pp 405-434

Ueda H, Kagawa H, Nagahama Y (1985) Involve- ment of gonadotropin and steroid hormones in spermiation in the amago salmon, Oncorhynchus rhodurus, and goldfish, Carassius auratus. Gen Comp Endocrinol 59: 24-30

Ueda H, Kambegawa A, Nagahama Y (1984) In vitro 11-ketotestosterone and 17a,208-dihydroxy-4- pregnen-3-one production by testicular fragments and isolated sperm of rainbow trout, Salmo gaird- neri. J Exp Zool 231: 435-439

Ueda H, Young G, Crim LW, Kambegawa A, Nagahama Y (1983) 17a,20-Dihydroxy-4-pregnen- 3-one: Plasma levels during sexual maturation and in vitro production by the testes of amago salmon (Oncorhynchus rhodurus) and rainbow trout (Salmo gairdneri). Gen Comp Endocrinol 51: 106-112 Ueno K (1985) Sterility and secondary sexual char-

Triploid Male Masu Salmon 125

acter of triploid Gnathopogon elongatus caeru- lescens. Fish Genetics and Breeding Science 10: 37- 41 17 Wolters WR, Libey GS, Chrisman GL (1982) Effect of triploidy on growth and gonadal development of channel catfish. Trans Am Fish Soc 111: 102-105 18 Young G, Crim LW, Kagawa H, Kambegawa A,

Nagahama Y (1983) Plasma 17a,20£-dihydroxy-4- pregnen-3-one levels during sexual maturation of amago salmon (Oncorhynchus rhodurus): Correla- tion with plasma gonadotropin and in vitro produc- tion by ovarian follicles. Gen Comp Endocrinol 51: 96-105

soci atBae hy — i nani abet aren

o. naeoquleny inten anh ete. eres bps: al whe CIP Mabe taj PLAN ww) OES _ Thee ugh trait ; tr fy ‘intl sewriy fied fe al

satitor rin ha vt itbeeer WhArcae o- S32 Sao.

: ae ‘ 1 dipeneearag : i hi Wile +e fap pat retag a, jompign Scie 5 ta oe ts Wt heortrng oh eed Peal” -- wit = BG if A a ae vies in be -aniaegs ieee id, Agee cs ‘ aay ; 1907) those bal beware ; : Ae MOAN TE wrod 1 Segui, has a aie Cablaliig Slo Pip

ii 2 sreiieg ecerace FT ISET | itl Kk

© ti 7 dated E ) saree “as - af ee th pe . ; ‘* 4 4 “ty rly ri ; ' ’ ba Tice a 7 vie A Pape + eg) +. 1 ISA aa ct ic Peas ornate} a a a a mo, tae ai tien = an ee ; = we te tal tind by i U Nid). aires) > Metin ~ Shes, ed tape = ; hia anes Se eee : . ; } ee 3 i efx Sere : Be Wes i ; x . ak 4 ARF 'p J z at) “yee tT reer ae) PME ” ih uM me , A a-tripicaly by Anal = J elaniowt: 4 rig Aue : tony 2rea Aare a - = wt’, 4 sty . : a beh ating | ee Pay 8a i ie Lite ' ey 4 ' i ee a! = . + ou b = =i : y i ay Le | © ' is ; - ‘ Go aa ite (oe it the 4 ine ig Ah <4) 5 See : y O° wy i Sah D Ai , dicks nrg oS) Se The Tees “t) ; bar “Sedurne) nee Tai wits Creleceirenlhll . , Poy MATT So

ZOOLOGICAL SCIENCE 10: 127-132 (1993)

© 1993 Zoological Society of Japan

A case of Intersexuality in the Sea Spider, Cilunculus armatus (Pycnogonida; Ammotheidae)

Katsumi MryazaAkr and TosHiki MAKIOKA

Institute of Biological Sciences, University of Tsukuba, Tsukuba, Ibaraki 305, Japan

ABSTRACT— An intersexual specimen of the sea spider, Cilunculus armatus, having female and male characteristics together with some intermediate ones was collected from Sagami Bay. Eggs are seen through the cuticle of each femoral segment of walking legs. At the same time, cement gland, peculiar to male, is also found on the same segment of each walking leg, but somewhat atrophied and distally shifted. Genital pore opening on the second coxal segment of each walking leg as seen in female is of intermediate size between the sizes of both sexes. Oviger is intermediate in size and shape, and the number and distributional pattern of ovigerous compound spines are also intermediate.

INTRODUCTION

A number of cases of the sexual abnormality, such as the gynandromorphism and the intersex- uality, have been reported in arthropods, especial- ly in insects [7], malacostracan crustaceans [3, 4] and spiders [12].

Pycnogonids are dioecious, except for a her- maphrodite species, Ascorhynchus corderoi (1). No intersexuality has been described and gynan- dromorphs have been known only in the following four cases. Losina-Losinski [9] reported a bilateral gynandromorphic specimen of Asc. abyssi from the Greenland Sea. Child [5] examined various grades of gynandromorphism in Anoplodactylus portus from the Pacific entrance of the Panama Canal. Child and Nakamura [6] found a gynandro- morph of Ano. gestiens from Sagami Bay, and described it with a few possible gynandromorphic specimens of Ano. jonesi (loc. undescribed).

In the present study, we describe for the first time an intersexual specimen of Cilunculus arma- tus, found among over 1500 specimens newly col- lected from Sagami Bay.

Accepted July 29, 1992 Received June 22, 1992

' Present address: Department of Biology, Keio Uni- versity, Hiyoshi 4-1-1, Kohoku-ku, Yokohama 211, Japan.

MATERIALS AND METHODS

A specimen of Cilunculus armatus (Bohm, 1879) with characters of both sexes was collected by dredge from the sandy bottom at about 40m in depth off the coast of Shimoda, near the top of Izu Peninsula, Central Japan, 24 March 1990. After fixation with 70% ethanol, the specimen was ex- amined under a stereomicroscope, compared with eight normal male and eight normal female speci- mens collected by the same dredging.

RESULTS

We newly collected 1535 specimens of Ciluncu- lus armatus, and found only one specimen with characters of both sexes. We also reexamined other 1009 preserved specimens which had been collected from the same locality, but could not find any more.

Sexual characters

The present specimen has both ovarian eggs as a female character and cement gland as a male character in the femoral segment of each walking leg, other than the regenerating first right (1R) leg (Fig. 1). There are also some curiosities in several sexual characters of the present specimen especial- ly on morphological features of cement gland, genital pore, oviger, and ovigerous compound

128 K. MryAZAKI AND T. MAKIOKA

Fic. 1. Walking legs. Setae and spines are omitted. A, Normal male; B, Normal female; C, Intersexual specimen; D, 1R leg of intersexual specimen. ac, accessory claw; c, claw; ct, efferent tube of cement gland; f, femur; Ic, first coxa; It, first tibia; ov, ovary; pr, propodus; 2c, second coxa; 2t, second tibia; t, tarsus; 3c, third coxa. Bar=1.0

mm. TaBLE 1. Comparison of some sexual characteristics in an intersexual specimen with those in normal males and females + —=present. — =absent Normal males wee Normal females Oviger 6-8th long intermediate short segments hairy a little hairy not hairy 6th seg. swollen intermediate not swollen Distribution of Ws2eils7 S878 ie 7) SSS sD, compound spines Cement gland + at =: (somewhat reduced) Genital pore 3rd & 4th legs all legs all legs (exc. regenerating 1R) small intermediate large (50x 40 ~m) (75 x50 4m) (120 x 100 «m) Gonad Ovary = ap = Testis ar not visible —

spine.

Cement gland was found in all the normal males examined and in the present specimen. In normal male, cement gland was well-developed in the

femoral segment of each walking leg, and its long external tube with an opening was protruded dor- sally at distal one-third of this pedal segment (Fig. 2A). The present specimen had somewhat

Pycnogonid Intersexuality 129

Fic. 2. Comparison of cement glands. A, Normal male; B, Intersexual specimen. cg, cement gland; ct, efferent tube of cement gland; f, femur; It, first tibia; 3c, third coxa. Bar=0.25 mm.

atrophied cement gland with shorter tube bearing an opening in the femoral segment of each walking leg other than the 1R leg (Fig. 2B). The tube was shifted more distally than normal one (Fig. 2B). Genital pore with single cuticular lid was situ- ated on the ventral surface of the second coxae of

all the walking legs in female, and on the same position of the third and fourth walking legs in male. The female genital pore was larger than male one (Table 1, Fig.3A, C). The present specimen had genital pore in all the walking legs other than the 1R leg, whose size was intermediate between that in male and female (Table 1, Fig. 3B).

Male oviger was larger than female one (Table 2, Fig.4A, C), and its sixth segment was prominen- tly stout and hairy (Fig. 5A). Oviger in the present specimen was mostly similar in shape and in size to the female one (Fig.4B, C), but its sixth segment was more hairy than that in female (Fig. 5B). Several apical segments of the oviger had some compound spines (Fig. 5), arranged in different patterns between sexes: In each of normal males, the seventh ovigerous segment had no spines, the eighth two, the ninth one, and the tenth two (0:2:1:2) (Fig. 5A), while in each of normal females, these spines were arranged in 3:3:1:2 (Fig. 5C). In the present specimen, however, they were arranged in 3:2:1:2 (Fig. 5B).

TABLE 2. Comparison of lengths of external body parts in intersexual specimen with those in normal 8 males and 8 females. Mean+SE (Range). measured in mm Normal males ear Normal females Proboscis 2.10+0.071 (1.89—2.42) PAPA 2.31+0.060 (2.11—2.63) Trunk 2.67+0.112 (2.05—3.11) 2.71 2.80+0.092 (2.26—3.05) Abdomen 1.13+0.061 (0.92—1.39) 1.18 1.15+0.041 (0.95—1.32) Chelifore 0.44+0.011 (0.39—0.50) 0.47 0.43+0.016 (0.37—0.50) Palp 2.35+0.075 (1.97—2.68) 2.82 2.59+0.093 (2.21—3.08) Oviger 3.83+0.085 (3.45—4.13) 2.84 2.40+0.076 (2.13—2.82) Lateral process 0.39+0.019 (0.32—0.47) 0.37 0.35+0.020 (0.26—0.45) Leg Ist coxa 0.46+0.023 (0.37—0.58) 0.45 0.38+0.019 (0.32—0.47) 2nd coxa 0.74+0.038 (0.58—0.92) 0.76 0.69+0.026 (0.58—0.79) 3rd coxa 0.53+0.025 (0.37—0.63) 0.68 0.64+0.023 (0.58—0.76) femur 1.39+0.039 (1.27—1.53) LSS) 1.53+£0.036 (1.37—1.68) Ist tibia 1.36+0.043 (1.18—1.50) 1.47 1.44+0.025 (1.34—1.53) 2nd tibia 1.29+0.054 (1.00—1.50) 1.45 1.47+0.024 (1.42—1.58) tarsus 0.16+0.010 (0.13—0.21) 0.16 0.17+0.005 (0.16—0.18) propodus 0.93+0.030 (0.79—1.08) 0.97 0.98+0.015 (0.95—1.05) claw 0.60+0.019 (0.53—0.68) 0.58 0.62+0.021 (0.55—0.71) accessory claw 0.34+0.013 (0.29—0.37) 0.26 0.33+0.015 (0.29—0.39) Cement gland tube 0.35+0.015 (0.26—0.42) 0.18 —

130 K. MIyAZAKI AND T. MAKIOKA

A B Cc 3c - V4 we of cy Q ale as

Fic. 3.

Comparison of genital pores (arrows). A, Normal male; B, Intersexual specimen; C, Normal female. 1c,

first coxa; 2c, second coxa; 3c, third coxa. Bar=0.1 mm.

B cs

Leen

Fic. 4. Comparison of ovigers. Setae and compound spines are omitted. A, Normal male; B, Intersexual specimen; C, Normal female. Bar=0.5 mm.

Any male gonadal element could not be found by the present external observations.

Several sexual characters of the present speci- men and normal males and females are summa- rized in Table 1.

Measurements of external body parts

In Cilunculus armatus, females were mostly lar- ger than males (Table 2). Measurements of va- rious parts of the present specimen lay mostly within the female range (Table 2). Normal female leg was stouter than normal male one (Fig. 1A, B). The leg of the present specimen was approximately similar in shape to the female leg (Fig. 1C).

Measurements of some external body parts of the present specimen and normal males and females are summarized in Table 2.

DISCUSSION

In the present specimen, we detected some male, female, and also intermediate characters. Existence of cement glands is a typical male char- acter in Cilunculus armatus as in other pycnogo- nids, but in the present specimen, these glands are somewhat atrophied. The ovarian eggs and the genital pores situated on all the walking legs are female characters. Intermediate sexual characters are seen in size and shape of ovigers and genital pores, and, at the same time, in distributional patterns of ovigerous compound spines (see Table 1).

There are some descriptions incompatible with each other on the distributional pattern of these spines in Cilunculus armatus. Ortman [11] noticed compound spines on a probably female oviger; seventh and eighth segments with no spines, ninth with one, tenth with two (0:0:1:2). Loman [8] described the distributional pattern of the spines in both male and female ovigers as 1:2:1:2. Schim- kewitsch [13] correctly illustrated female pattern as 3:3:1:2. Utinomi [14] described a female speci- men with the 0:1:1:2 spines. Nakamura [10] recognized the difference in male and female pat- terns as 0:1:1:2.and 3:1:1:2, respectively.

In the present study, we determined correct numbers and distributional patterns of male and female ovigerous compound spines in Cilunculus armatus, and found an intermediate pattern in the present specimen.

Bacci [2] defined two main sexual abnormalities; the intersex as “an individual of unisexual species whose reproductive organs and (or) secondary sex characters are partly of one sex and partly of the

Pycnogonid Intersexuality 131

Fic. 5. 0.1 mm.

other although it does not show genetically diffe- rent parts”, and the gynandromorph as “an indi- vidual of a unisexual species containing a mosaic of genetically male and genetically female cells”. According to Bacci’s [2] definition, we diagnose the present specimen as an intersexual one.

Child and Nakamura [6] reported a gynandro- morphic specimen in other pycnogonid, Anoplo- dactylus gestiens. The specimen had many female characteristics and a few male ones. Existence of the cement gland and the oviger is peculiar to males in Anoplodactylus species, but the gynan- dromorphic specimen has an atrophied cement gland only in the femoral segment of the fourth left leg and the ovigers, the left one being slightly deformed. The gynandromorphic specimen of Ano. gestiens has some close resemblances with the present intersexual specimen of Cilunculus armatus.

As the intact preservation of the present interse- xual specimen is needed, histological, karyological and biochemical details were left untouched.

ACKNOWLEDGMENTS

We gratefully acknowledge Professor Emeritus Jan H. Stock, University of Amsterdam, for his helpful advice and kind encouragement during the present study. Thanks are also due to Mr. Hajime Ueda, Mr. Yasutaka Tuchiya, and Mr. Toshihiko Sato, crew of the collecting

Comparison of apical parts of ovigers. A, Normal male; B, Intersexual specimen; C, Normal female. Bar=

ship “Tsukuba”, for their constant help in collecting sea spiders, and other staffs of the Shimoda Marine Research Center, University of Tsukuba, for their hospitality. This paper is the Contribution from the Shimoda Marine Research Center, University of Tsukuba, No. 543.

REFERENCES

1 Arnaud F, Bamber RN (1987) The biology of Pycnogonida. Adv Mar Biol 24: 1-96

2 Bacci G (1965) Sex Determination. Pergamon, New York

3 Charniaux-Cotton H (1960) Sex determination. In “The Physiology of Crustacea, Vol. 1”. Ed by TH Waterman, Academic Press, New Yolk, pp 411-447

4 Charniaux-Cotton H (1975) Hermaphroditism and gynandromorphism in malacostracan crustacea. In “Intersexuality in the Animal Kingdom”. Ed by R Reinboth, Springer-Verlag, Berlin, Heidelberg, New York, pp 91-105

5 Child CA (1978) Gynandromorphs of the pycnogo- nid Anoplodactylus portus. Zool J Linn Soc Lond 63: 133-144

6 Child CA, Nakamura K (1982) A gynandromorph of the Japanese pycnogonid Anoplodactylus gestiens (Ortmann). Proc Biol Soc Wash 95: 292-296

7 Lauge G (1985) Sex determination: Genetic and epigenetic factors. In “Comprehensive Insect Phy- siology, Biochemistry and Pharmacology, Vol. 1, Embryogenesis and Reproduction”. Ed by GA Ker- kut, LI Gilbert, Pergamon Press, Oxford, New York, Toronto, Sydney, Paris, Frankfurt, pp 295- 318

8

10

11

132

Loman JCC (1911) Japanische Podosomata; Beit- rage zur Naturgeschichte Ostasiens, herausgeben von F. Doflein. Abh Kaiser Bayersch Akad Wiss Suppl 2: 1-18

Losina-Losinsky LK (1964) Pantopoda from the collections of the expeditions of the F Litka in 1955 and the Ob in 1956 Trudy Ark Antarktic Nauchno- Issledov Inst Leningrad 259: 330-339 (in Russian) Nakamura K (1987) The Sea Spiders of Sagami Bay. Ed by Biological Laboratory Imperial Household, Maruzen, Tokyo

Ortmann AE (1890) Bericht uber die von Herrn Dr.

12

13

14

K. MIYAZAKI AND T. MAKIOKA

Doderlein in Japan gesammelten Pycnogoniden. Zool Jahrb (Syst) 5: 157-168

Roberts MJ, Parker JR (1973) Gynandry and in- tersexuality in spiders. Bull Brit Arach Soc 2: 177- 183

Schimkewitsch W (1929) Pycnogonida (Pantopoda). Fauna SSSR Izdateljstvo Academii Nauk SSSR, Part 1: I-CXV+1-224 (in Russian and Latin) Utinomi H (1955) Report on the Pycnogonida collected by the Soyo-Maru Expedition made on the continental shelf bordering Japan during the years 1926-30. Pub Seto Mar Biol Lab 5: 1-42

ZOOLOGICAL SCIENCE 10: 133-140 (1993)

Metameric Color Pattern in the Bugs (Heteroptera: Lygaeidae, Largidae, Pyrrhocoridae, Rhopalidae): A Morphological Marker of Insect Body Compartmentalization

JAN Zrzavy!”, OLpkIcH NEDvED!” and RApomir Socua!

Institute of Entomology, ‘Academy of Science of the Czech Republic and, *Faculty of Biological Science, University of South Bohemia, Branisovska 31, 370 05 Ceské Budéjovice, Czech Republic

ABSTRACT— Diversity of abdominal color patterns in some coreoid bugs (Heteroptera: Pentatomo- morpha: Lygaeidae, Largidae, Pyrrhocoridae, Rhopalidae) is used as a morphological marker to reveal the developmental organization of the abdominal segments. Dominant metameric arrangement of these color patterns is documented by comparison of 160 species, and those results are supplemented by detailed analysis of two model species (Pyrrhocoris apterus, Dysdercus cingulatus), including a preliminary clonal analysis of mosaic epidermis in the P. apterus mutant strain. This allows the conclusion that (a) the intrasegemental transversal line, marked by contrasting epidermal coloration in the model species, has some properties of the intrasegmental compartmental boundary known in the Drosophila segments; (b) this boundary has a considerable role in cuticular color-pattern formation, being hardly exceedable for the cuticular spots established in one half of the segment; and (c) a great majority of the cuticular color patterns, although polyphyletically evolved, follows the same organiza- tional scheme, and their diversity can thus be regarded as constrained by the compartmental

© 1993 Zoological Society of Japan

boundaries.

INTRODUCTION

The segmentation of the insect body during development have been studied predominantly in Drosophila, a species possessing an extremely ad- vanced body plan (for reviews see [1, 6, 8, 14, 18, 25]). The use of Drosophila as a model allows us to employ the latest methodologies of molecular genetics and developmental biology. Yet many findings derived from Drosophila may pertain only to cyclorrhaphous flies, and may not be sufficient for a general scheme of the insect (or even arthro- pod and articulate) body plan. A broader compar- ative survey of the segmentation processes in other insects or arthropods is therefore required. Sander [23] considers spontaneous or experimentally- induced teratological “monsters” a basis for ontogenetic analyses of arthropods which are still unapproachable by genetic methods (i.e., a vast majority of cases). We believe, however, that

Accepted November 4, 1992 Received July 1, 1992

phylogenetic studies of diversity patterns can also provide a similarly valuable source of evidence.

The studies of Minelli and Bortoletto [19], Emerson and Schram [3], and Nijhout [21] are of particular interest in this respect. These studies deal primarily with morphology rather than gene- tics or developmental biology, proceeding by three principal steps: (1) description of the diversity pattern, (2) determination of the regularities in- volved in this diversity, and (3) confrontation of these pattern regularities with presumably relevant developmental processes. The discovery of a con- venient morphological trait that would serve as a marker of the sought developmental rules is the critical point in this approach.

Cuticular coloration of some bugs belonging to the superfamily Coreoidea s. lat. (Pentatomonor- pha; [24, 29] and references therein) displays con- spicuous red-, ochre- or orange-and-black wasp- like patterns on the abdominal segments. The light (non-melanized) and dark (melanized) halves of the individual segments usually alternate regularly. This “half-segment” arrangement of the color pat-

134 J. ZRZAvY, O. NEDVED AND R. SocHa

tern is interesting in terms of the anterior-posterior pattern-formation of the insect body. The coreoid bugs should therefore provide a good model group for developmental studies.

In the present paper, we attempt to examine the color patterns in a variety of species (160 spp.). The results are supplemented by more detailed analysis of the model species [Pyrrhocoris apterus (Linnaeus, 1758), Dysdercus cingulatus (Fabricius, 1775); both Pyrrhocoridae], including a prelimin- ary clonal analysis of the epidermis in the P. apterus mutant strain mosaic (mo) [20, 26-27].

MATERIALS AND METHODS

Material

The insects used for comparative studies were obtained from the American Museum of Natural History (New York), National Museum (Praha), Moravian Land Museum (Brno), University of Connecticut. (Storrs), and Charles University (Praha).

160 species of five higher taxa of the Coreoidea s. lat., viz. Lygaeinae (Lygaeidae: 34 species), Rhyparochrominae (Lygaeidae: 43 spp.), Largidae (9 spp.), Pyrrhocoridae (57 spp.) and Rhopalidae (17 spp.), were examined. Unless otherwise stated, only the adults were studied. Individual variability and sexual dichroism were tentatively checked in approximately 35% of the examined species and found negligible; therefore they were omitted from further analysis.

Two model species, D. cingulatus and P. apter- us, were used for more detailed morphological and ontogenetic analysis. Specimens of both species were obtained from laboratory stock cultures. Be- sides cuticular color patterns, the epidermal were also examined in all stages of postembryonic de- velopment. The coloration of epidermis was either observed through a transparent non-melanized cutile (larvae of both species, adults of D. cingula- tus), or the animals were beforehand dissected (adults of P. apterus).

Morphological analysis and statistical evaluation

The topological distribution of color spots over the abdominal segments was studied. Abdominal

segments II to VII were chosen for study, since they are clearly visible and mainly uniform in shape. The abdominal surface was divided into a grid of 120 approximately square-like fields (Fig. 1A), in which the presence vs. absence of mela- nized cuticle (1 vs. 0) was recorded. An individual field was considered to be melanized if (a) it was completely melanized, or (b) it contained a small melanized spot (a “melanization centre”) not ex- ceeding the field borders, or (c) a large melanized spot originating outside the field covered more than 50% of its area. Data obtained by trans- formation of existing color patterns to the square grid (Fig. 1A) could be compared due to the similar body shapes of the examined species.

The transformed color patterns were tested sta- tistically to reveal regularities in color pattern diversity within the examined taxa. Lygaeinae, Rhyparochrominae, Largidae, Pyrrhocoridae, and Phopalidae were analysed separately. To identify pairs of fields with parallel melanization, the con- cept of “conjunctions” was introduced. Fields #1 and #2 were in conjunction if #1 was melanized Just when +2 was melanized. In other words, a conjunction displays a completely concerted mela- nization of two fields. The fields melanized in less than 10 percent of the examined species within each taxon were excluded from analysis.

The conjunctions fell into four categories: (A) metameric conjunctions between fields on the same (anterior or posterior) halves of different abdominal segments, (B) conjunctions between two fields on the same half of a single segment, (C) conjunctions between two fields on different halves of a single segment, and (D) conjunctions between fields on different halves of different segments (Fig. 1A). Since the number of all possi- ble conjunctions differs for each category (3,000 for A and D, ‘540 for B, 600 for C), direct comparison of the frequency of the individual categories of conjunctions was not possible. The observed numbers of the conjunctions (N.ps) were tested against the expected numbers of the con- junctions (Nexp, reflecting 3,000 :540:600 :3,000 ratio if the probability of all categories of conjunc- tions was equal) by Y? test. The null hypothesis that categories A+B (i.e., those revealing the “half-segment” arrangement) are equal to C+D

Color Pattern Metamerism in the Bugs 135

was also tested by the y° test (against 3,540 : 3,600 ratio).

Clonal analysis of mosaic epidermis

A laboratory stock of mosaic (mo) mutant of P. apterus, possibly transposon-mediated, was analy- sed [20, 26-27]. The epidermis of mo bugs dis- plays an irregular mosaic of pigmented cells (red anterior, yellow posterior) with transparent (whit- ish) cell clones lacking pteridine pigments. The distribution of transparent cell clones was consi- dered an apparent marker of normal behaviour of epidermal cell lineages, as the individuals of the mo Strain do not differ from those of the wild-type in any fundamental morphological, developmental and ecophysiological features. Moreover, when mo insects are compared to X-ray-treated animals (for example, [11]), little or no incidence of ex- traordinary cell deaths should be expected.

Last-instar larvae, whose epidermis is well visi- ble through a transparent cuticle and whose epidermal clones are the best-developed, were studied. Only abdominal segments IV-VI were analysed (both dorsal and ventral surfaces). Since the extent of transparent clones varies consider- ably in the mo strain, only specimens with distinct,

sharply outlined patches were anlaysed. Each abdominal segment consists of two distinctly col- ored areas, red (approximately on the anterior two thirds of the segment) and yellow (the posterior third). A grid of six parallel transversal lines (A, R;, Ro, RY, Y, P) was superimposed over the segment so that two lines (A, P) fitted the interseg- mental boundaries, one (RY) fitted the boundary between the red and yellow parts of the segment, two (R;, R2) were laid on the red area, and one (Y) on the yellow area (see Fig. 1B). The distribu- tion pattern of the cell clones was quantified as the number of the patches crossed by the grid lines (see camera lucida drawing in Fig. 1B).

Several hundreds of bugs of both sexes were used for basic screening, and those with suitably distinct clones (10 males, 10 females) were selected for a detailed morphological analysis, which was made on more than 400 uncolored patches.

RESULTS

Development of color patterns in the model species

Postembryonic development of abdominal col- oration in D. cingulatus and P. apterus is summa-

eee Fal TF PP Hn i HNO ATTA

Lone CAC

Fic. 1A. A method for transformating the color pattern into grid form (exemplified by Spilostethus saxatilis [Lygaeidae]). Do: dorsal, Ve: ventral surface; A-D: categories of conjunctions (A: conjunctions between fields on the same halves of different abdominal segments, B: conjunctions between two fields on the same half of a single one segment, C: conjunctions between two fields on different halves of a single segment, D: conjunctions between fields on different halves of different segments).

Fic. 1B. A method of evaluation of the transparent clones (black) in mo mutant of P. apterus. IV-VI: abdominal segments; dashed areas: red epidermis; open areas: yellow epidermis; A, R1, Ro, RY, Y, P: arbitrary grid lines (A, P: intersegmental boundaries; RY: boundary between the red and yellow parts of the segment; R,, R»: red area; Y: yellow area); 0, 7, 6, 2, 6, 0: numbers of the clones crossed by the grid lines (in camera lucida drawing).

136 J. ZrzavyY, O. NEDVED AND R. SocHA

Fic. 2.

Postembryonic development of abdominal color patterns in D. cingulatus (left) and P. apterus (right). L:

larvae; A: adults; a: anterior part of a segment; p: posterior part of a segment; c: cuticle; e: epidermis; black areas: melanized cuticle; open areas: transparent cuticle; dashed areas: red epidermis; dotted areas: white or

yellow epidermis.

rized in Figure 2. Both species show epidermal color patterns in addition to the cuticular ones. Whereas the epidermis is red or orange-red in the anterior parts of segments, it is either white (D. cingulatus, ventral side) or yellow (P. apterus, around the entire body) in the posterior parts of segments in both larvae and adults. Posterior boundaries of cuticular melanization in adult D. cingulatus completely coincide with the boundaries between red and white epidermal cells (Fig. 2). Similarly, the cuticular melanized spots, which are still distinct in the larvae of P. apterus, never reach over the epidermal boundaries from the red to the yellow regions (Fig. 2).

A similar epidermal coloration was previously described by Lawrence [11-12] and Lawrence and Green [15] in Oncopeltus fasciatus (Lygaeidae: Lygaeinae). Undoubtedly this subdivision of seg- ment epidermis is widespread among coreoid bugs, especially among the Pyrrhocoridae (Zrzavy, in prep.). In conclusion, cuticular and epidermal color patterns are highly congruent in the model species. A cuticular spot established in one half of a segment usually does not exceed its intersegmen- tal as well as intrasegmental boundaries. The line dividing the segment into anterior and posterior parts (marked by different coloration of the epidermis of the model species) plays an important role in the anterior-posterior arrangement of cu-

ticular coloration.

Clonal restriction in epidermal color pattern forma- tion

As shown in Table 1, no clones seem to lie on the intersegmental boundaries in P. apterus; but any intrasegmental line is not an insurmountable barrier.

The red-yellow boundary appears to be the least exceedable, compared to the other lines. We are aware that our simple morphological method is not fully objective since not only the clones but also

TaBLeE1. Distribution of colorless epidermal clones on the abdomen of mo mutant of P. apterus (10 males, 10 females; c. 7 patches per a segment in average). IV, V, VI: abdominal segments; A, R,, Ro, RY, Y, P: arbitrary grid lines (A: anterior boundary of a segment; P: posterior boundary of it; R;, Ro: two lines within the red region of a segment; RY: red/yellow intrasegmental bound- ary; Y: a line within the yellow region of a segment; see Fig. 1B); +: some number of clones might be overlooked

Ao oRy oo Reo (RY oid Y oy ee

IV 0 61 54° 96 342 aioe Vv 1 37 44. 23. 284° 0 wlssee VI O 37 31, . 21 > 1742 Omeeel06e Da 1 i 1A 70) Ts 1 415+

Color Pattern Metamerism in the Bugs 137

TABLE 2.

General pattern of color conjunctions in Lygaeinae, Pyrrhocoridae and Rhopalidae Ny:

total observed number of conjunctions within a taxon; N.,,:expected number of conjunctions if the probability of all categories is equal (3,000:540:600:3,000 ratio—see Material and Methods); A, B, C, D: categories of conjunctions (Fig. 1A); x7 test—**, ***: significant on the levels of probability P=0.010, and 0.001, respectively

Lygaeinae Pyrrhocoridae Rhopalidae (34 spp.) (57 spp.) (17 spp.) Nobs Nexp Nobs Nexp Nops Nexp A 41 33.2 79 68.9 46 37.0 B 14 6.0 25 12.4 11 6.7 C 6 6.6 20 13.8 15 7.4 D 18 33.2 40 68.9 16 37.0 A+B 55 39.2 104 81.3 57 43.7 C+D 24 39.8 60 82.7 31 44.4 A:B:C:D FORK Rx Per * KK * KK **K

(A+B) :(C+D)

their fragments could be enumerated, while very small transparent (i.e., whitish) clones on the yellow background may be overlooked. Hence, the number of clones within the yellow area (as well as the depth of “RY depression”) may have been slightly underestimated.

General organization of color patterns in the coreoid bugs

The color arrangement in the coreoid families is obviously non-random, but organized according to the anterior-posterior polarity of each segment in Lygaeinae (Lygaeidae), Pyrrhocoridae, and Rho- palidae. A significant majority of the color-pattern conjunctions connects either fields lying on the same (anterior or posterior) halves of different segments, or fields lying on the same half of a single segment. Thus, from the four possible categories of conjunctions (see Material and Methods), those marked A and B clearly pre- dominate. On the contrary, the C-category con- junctions connecting fields of the opposite halves of the same segment are less frequent. This data indicates an obviously metameric “half-segment” organization of abdominal color patterns in the taxa studied here (Table 2).

However, the dorsal parts of the abdomen (i.e., those covered by wings in resting position) reveal this metameric color pattern to a much lesser

extent. This is because there is, particularly in Rhopalidae, a strong tendency of the abdominal dorsum to display another type of coloration, associated with the aposematic function during flight.

No general metameric arrangement is developed in the Rhyparochrominae (Lygaeidae) and Lar- gidae.

DISCUSSION

Phylogenetic status of the metameric color patterns

Three coreoid taxa examined (Lygaeinae, Pyr- thocoridae, Rhopalidae) apparently tend to gener- ate very similar metameric arrangements of the color pattern. However, each differs from others in some details. For example, the lateral sectors (“connexiva”) of both Lygaeinae and Rhopalidae display a clear metamerism, but while the mela- nized spots usually appear on the anterior halves of the segments in Lygaeinae, they are found on the posterior in Rhopalidae.

It can be assumed that the lygaeines, pyrrhocor- ids, and rhopalids do not constitute a monophyle- tic taxon, rather that they represent independent phylogenetic lines of the coreoid bugs. Thus, although our knowledge of phylogenetic rela- tionships among the coreoid bugs is rather scant

138 J. Zrzavy, O. NEDVED AND R. SOCHA

[24, 29], and, further, the evolution of their color patterns has never been rigorously analysed, the hypothesis of the homology of their color patterns should be rejected [32]. The majority of the coreoid bugs has cryptic coloration without any conspicuous patterns on the abdominal segments. It appears that the wasp-like abdominal color pattern results from aposematic function of the coloration, which certainly does not belong to the coreoid groundplan. For example, the Rhypar- ochrominae (a group cladistically related to Pyr- rhocoridae, Largidae, Rhopalidae and some other families) and the Largidae (a pyrrhocorid sister group) show complex (but generally cryptic) pyr- rhocorid-like color patterns on the pronotum and forewings and generally uniform coloration of abdomen [32].

In conclusion, the wasp-like patterns un- doubtedly are not homologous. Instead, the metameric color patterns seem to have arisen independently at least three times during the evolution of the Coreoidea s. lat.—(a) in Lygaeinae, (b) in Pyrrhocoridae, and (c) in Rho- palidae [32]. Yet in spite of the undoubtedly polyphyletic origin of the “wasp-like” color pat- terns in the Coreoidea s. lat., their diversity is strongly constrained to the formation of a single

general organization of color fields, with iteration of the “half-segment” units.

Morphogenetic nature of the color patterns metamerism

The results acquired by clonal analysis of the abdominal epidermis in P. apterus are quite com- patible with those of Lawrence [11-13] and Lawr- ence and Green [15]. They showed by clonal analysis of the abdominal epidermis of O. fasciatus that (a) segment epidermis is of a polyclonal nature and epidermal cell lineages do not cross intersegmental boundaries; (b) although many clones are sufficiently large to span the entire segment length, no clone comprises both the ante- rior and posterior margins of the same segment; (c) clones are usually laterally elongated; and (d) clones usually do not cross the intrasegmental boundary between bands of red and white epidermis. Concluding, regions of red and white (or yellow) epidermis in the bugs seem to repre- sent further morphogenetic units within the abdo- minal segments, and they may represent relatively independent cell lineages (Fig. 3; see Table 1).

These conclusions can, moreover, be applied as well to cuticular color spots of coreoid bugs, which are laterally arranged and do not span the segment

Fic. 3. Comparison of body compartmentalization of the Drosophila and coreoid bugs. A: Drosophila; B: O. fasciatus; C: P. apterus;; D: D. cingulatus ; E. coreoid bugs in general; 1: cuticular coloration (orig.); 2: epidermal coloration ([{2, 11-13, 15], orig.); 3: position of a spiracle ((7], orig.); 4: clonal restriction [6, 11-16, orig.]; 5: engrailed expression pattern [2, 7, 14], a: anterior compartment; p: posterior compartment. Markers available for

comparative morphological analysis are boxed.

Color Pattern Metamerism in the Bugs 139

length. Both color patterns, epidermal and cuticu- lar, seem to be organized with respect to the same line, the “red-white boundary”, and the evolution of the aposematic wasp-like coloration only further reveals this basic two-fold organization of the abdominal segment.

This intrasegmental boundary coincides with the line least exceedable by cell lineages, that is, with the possible intrasegmental compartmental bound- ary, analogous to that of Drosophila (Fig. 3) [1, 6- 10, 14, 16-18, 25, 30]. The integument of each body segment of Drosophila is derived from two cell polyclones, the anterior and posterior com- partments, each of which is characterized by a unique combination of active selector genes, and therefore forms a distinct component of the pat- tern. The compartment seems to be a functional unit for subsequent patterning (including also col- or-pattern formation: for example, the anterior compartments are melanized, the posterior ones non-melanized).

Using an antibody binding to the product of the segmentation gene engrailed (en), Campbell and Caveney [2] have shown that en-positive cells in O. fasciatus form a continuous band localized im- mediately anterior to the segment border. This pattern of en expression is identical to that of posterior compartments in Drosophila [7, 9-10, 14], as well as other flies [28], bees [4-5] and locusts [22], and even the crustaceans [22] and centipedes [31]. The size and location of the en-positive cell band of the O. fasciatus corres- ponds precisely to a distinct band of white- pigmented epidermal cells [2]. In other words, white pigmentation of the epidermis is a conspic- uous morphological marker of the posterior com- partment (Fig. 3).

Conclusions

(A) The red-yellow intrasegmental boundary in P. apterus seems to display clonal properties similar to the red-white boundary in O. fasciatus, and is probably homologous to it (Figs. 3B-C).

(B) The intrasegmental transversal line, marked by contrasting epidermal coloration in the model species, shares the basic properties of the intrasegmental compartment boundary of Dro-

sophila segments, and is probably homologous to it (Figs. 3A-C).

(C) The red-yellow boundary takes a consider- able role in cuticular color-pattern formation, being hardly exceedable by the cuticular spots established in one half of the segment (Fig. 3).

(D) The great majority of the cuticular color patterns, although independently evolved, re- spects the same organization scheme. This diversi- ty can be thus regarded as constrained by the compartment boundaries, both inter- and in- trasegmental. The uniform organizational scheme of abdominal color patterns in the coreoid bugs seems to serve as a good morphological marker of body compartmentalization.

Consequently, the coreoid bugs can provide a good model for comparative studies of body metamerism, since their compartmentalization seems to be distinctly marked by wild-type body coloration in a large number of species. There is surely some factual bridge between Drosophila developmental biology and comparative insect morphology, however careful one must be in mak- ing such generalizations.

ACKNOWLEDGMENTS

We are indebted to Prof. C. W. Schaefer and Prof. J. A. Slater (University of Connecticut, Storrs), Dr. R. T. Schuh (American Museum of Natural History, New York), Prof. P. Stys (Charles University, Praha), Dr. L. Hoberlandt (National Museum, Praha), Dr. J. Stehlik (Moravian Land Museum, Brno), Prof. F. Sehnal, Dr. M. Jindra, Dr. V. Novotny, and Dr. P. Svacha (Institute of Entomology, Ceské Budéjovice), and Mr. T. Steyskal (University of South Bohemia, Ceské Budéjovice) for their kind assistance during preparation of this paper. This study was supported by grants from the Czechoslo- vak Academy of Sciences and Hasselblad Foundation (Sweden).

REFERENCES

1 Akam M (1987) The molecular basis for metameric pattern in the Drosophila embryo. Development 101: 1-22

2 Campbell GL, Caveney S (1989) engrailed gene expression in the abdominal segments of Oncopeltus: gradients and cell states in the insect segment. Development 106: 727-737

3 Emerson MJ, Schram FR (1990) The origin of crustacean biramous appendages and the evolution

10

11

12

18

140

of Arthropoda. Science 250: 667-669

Fleig R (1990) engrailed expression and body seg- mentation in the honeybee Apis mellifera. Roux Arch Dev Biol 198: 467-473

Fleig R, Walldorf U, Gehring WJ, Sander K (1988) In-situ localization of the transcripts of a homeobox gene in the honey bee Apis mellifera L. (Hymenop- tera). Roux Arch Dev Biol 197: 269-274 Garcia-Bellido A, Lawrence PA, Morata G (1979) Compartments in animal development. Sci Am 241: 102-110

Hama C, Ali Z, Kornberg TB (1990) Region- specific recombination and expression are directed by portions of the Drosophila engrailed promoter. Gene Dev 4: 1079-1093

Kauffman SA, Shymko RM, Trabert K (1978) Con- trol of sequential compartment formation in Dros- phila. Science 199: 259-270

Kornberg T (1981) Compartments in the abdomen of Drosophila and the role of the engrailed locus. Dev Biol 86: 363-372

Kornberg TB, Siden I, O'Farrell PH, Simon M (1985) The engrailed locus of Drosophila: in situ localization of transcripts reveals compartment spe- cific expression. Cell 40: 45-63

Lawrence PA (1973) A clonal analysis of segment development in Oncopeltus (Hemiptera). J Embryol Exp Morphol 30: 681-699

Larence PA (1975) The structure and properties of a compartment border: the intersegmental boundary in Oncopeltus. In “Cell Patterning” Ed by S Bren- ner, Ciba Foundation Symposium 29, Elsevier, Am- sterdam, pp 3-23

Lawrence PA (1981) The cellular basis of segmenta- tion in insects. Cell 26: 3-10

Lawrence PA (1991) The Making of a Fly. Black- well Scientific Publications, Oxford, pp 228 Lawrence PA, Green SM (1975) The intersegmen- tal boundary in Oncopeltus. J Cell Biol 65: 373-382 Lawrence PA, Green SM, Johnston P (1978) Com- partmentalization and growth of the Drosophila abdomen. J Embryol Exp Moxphol 43: 233-245 Lawrence PA, Martinez-Arias A (1985) The cell lineage of segments and parasegments in Drosophi- la. Phil Trans R Soc Lond, B, 312: 83-90 Martinez-Arias A, Lawrence PA (1985) Paraseg- ments and compartments in the Drosophila embryo. Nature 313: 639-642

19

20

21

22

23

24

25

26

27

28

29

30

31

32

J. ZRZAVY, O. NEDVED AND R. SOCHA

Minelli A., Bortoletto S. (1988) | Myriapod metamerism and arthropod segmentation. Biol J Linn Soc 33: 323-343

Némec V, Socha R (1988) Phosphatases and pter- idines in Malpighian tubules: a possible marker of the mosaic mutant in Pyrrhocoris apterus (Heterop- tera, Pyrrhocoridae). Acta Entomol Bohemoslov 85: 321-326

Nijhout HF (1991) Development and Evolution of Butterfly Wing Patterns. Smithsonian Institution Press, Washington-London, pp 297

Patel NH, Kornberg TB, Goodman CS (1989) Ex- pression of engrailed during segmentation in grass- hopper and crayfish. Development 107: 201-212 Sander K (1988) Studies in insect segmentation: from teratology to phenogenetics. Development 104 (Suppl): 112-121

Schuh RT (1986) The influence of cladistics on heteropteran classification. Annu Rev Entomol 31: 67-93

Scott MP, O’Farrell PH (1986) Spatial program- ming of gene expression in early Drosophila embryogenesis. Annu Rev Cell Biol 2: 49-80 Socha R (1987) Unusual presence of pigmented granules in the Malpighian tubules in a mosaic mutant of Pyrrhocoris apterus (Heteroptera, Pyr- thocoridae). Acta Entomol Bohemosloy 84: 241- 245

Socha R, Némec V (1992) Pteridine analysis in five colour-body mutations of Pyrrhocoris apterus (Heteroptera: Pyrrhocoridae). Acta Entomol Bohe- moslov 89: 195-203

Sommer R, Tautz D (1991) Segmentation gene expression in the housefly Musca domestica. De- velopment 113:419-430

Stys P, Kerzhner I (1975) The rank and nomencla- ture of higher taxa in recent Heteroptera. Acta Entomol Bohemoslovy 72: 65-79

Szabad J, Schiipbach T, Wieschaus E (1979) Cell lineage and development in the larval epidermis of Drosophila melanogaster. Dev Biol 73: 256-271 Whitington PM, Meier T, King P (1991) Segmenta- tion, neurogenesis and formation of early axonal pathways in the centipede, Ethmostigmus rubripes (Brandt). Roux Arch Dev Biol 199: 349-363 Zrzavy J (1990) Evolution of the aposematic color pattern in some Coreoidea s. lat. (Heteroptera). Acta Entomol Bohemoslov 87: 470-474

ZOOLOGICAL SCIENCE 10: 141-146 (1993)

N-Terminal Amino Acid Sequences of 440 kDa Hemoglobins of the Deep-sea Tube Worms, Lamellibrachia sp.1, Lamellibrachia sp.2 and Slender vestimentifera gen. sp.1 Evolutionary Relationship with Annelid Hemoglobins

Tomouiko Suzuxkt’, TAKASHI TAKAGL and SuGuRU OHTA?

‘Department of Biology, Faculty of Science, Kochi University, Kochi 780, *Biological Institute, Faculty of Science, Tohoku University, Sendai 980, and *Ocean Research Institute,

University of Tokyo, Tokyo 164, Japan

ABSTRACT—The deep-sea tube worm Lamellibrachia, belonging to the phylum Vestimentifera, contains two types of extracellular hemoglobins, a 3,000 kDa hemoglobin and a 440 kDa hemoglobin. The latter hemoglobin is composed of four heme-containing chains with molecular masses of 16-18 kDa. We have collected Lamellibrachia sp.1, Lamellibrachia sp.2 and Slender vestimentifera gen. sp.1 from the deep-sea cold-seep or hydrothermal areas at a depth of 1100-1400 m. The four constituent chains of the 440 kDa hemoglobin were isolated from each of the three tube worms by reverse-phase chromatography, and the N-terminal amino acid sequences of 16-44 residues were determined by automated protein sequencer. The amino acid sequences of the homologous chains showed high homology (76-85%), suggesting that they are closely related. The sequences also showed 45-49% homology with annelid hemoglobins. A phylogenetic tree constructed from hemoglobin sequences showed that the tube worm Lamellibrachia, the polychaete Tylorrhynchus and the oligochaete Lumbricus diverged from a common ancestor at almost the same time, about 450 million years before

© 1993 Zoological Society of Japan

present.

INTRODUCTION

The deep-sea tube worms Riftia and Lamellib- rachia, belonging to the phylum Vestimentifera, are found in hydrothermal vent or cold seeps at a depth of 600—2,500m. These animals are sus- tained by mutual symbiosis with sulfide-oxidizing bacteria [3], and their blood, containing abundant extracellular hemoglobin, has a function to trans- port sulfide (H2S) to internal bacterial symbionts, as well as to facilitate oxygen transport compatible with high oxygen demand [1].

The tube worm Lamellibrachia sp.1, referred as Lamellibrachia sp. in previous papers [17-20], has two types of giant extracellular hemoglobins, a 3,000 kDa hemoglobin and a 440 kDa hemoglobin.

Accepted August 18, 1992 Received July 9, 1992

The former hemoglobin is composed of four 16-18 kDa heme-containing chains and two 24-26 kDa linker chains, while the latter consists of only four heme-containing chains. Two of the four heme- containing chains are common to both hemoglo- bins. So far, we isolated most of the constituent chains of the two hemoglobins by a reverse-phase chromatography, and determined the complete amino acid sequences of a 16kDa heme- containing chain [18] and a 24kDa linker chain [19]. The sequence results suggested that the hemoglobin of Lamellibrachia sp.1 is closely re- lated to those of annelids.

In this report, we isolated four heme-containing chains of 440 kDa hemoglobin newly from the two tube worms, Lamellibrachia sp.2 and Slender ves- timentifera gen. sp.1, and determined the N- terminal sequences of all the chains. Phylogenetic relationship among the tube worms and annelids is

142 T. Suzuki, T. TAKAGI AND S. OHTA

discussed.

MATERIALS AND METHODS

The tube worm Lamellibrachia sp.1 and Slender vestimentifera gen. sp.1 (both undescribed) were collected from the cold-seep area located off Saga- mi Bay at a depth of 1110-1170 m, Japan, by a Japanese submersible SHINKAI 2000. Lamellib- rachia sp.2 (undescribed) and Slender vestimenti- fera gen. sp.1 were collected from the hydrother- mal area of the Okinawa Trough, eastern part of the Iheya Ridge at a depth of 1405 m, Japan.

Hemoglobin was purified from frozen animals as described previously [17]. Two hemoglobin com- ponents, a 3,000 kDa hemoglobin and a 440 kDa hemoglobin, were isolated on a gel filtration col- umn of Superose 12 (Pharmacia, 1 x30 cm), in 50 mM phosphate buffer (pH 7.2) containing 150 mM NaCl at a flow rate of 0.5 ml/min. The constituent polypeptide chains of the reduced 440 kDa hemog- lobin were separated by a Cosmosil 5C;jg-300 col- umn (4.6150 mm, Nacalai Tesque) with a linear gradient of acetonitrile in 0.1% trifluoroacetic acid [17].

The amino acid sequence of carboxymethylated protein was determined directly by use of an automated sequencer (Applied BioSystems 477A Protein Sequencer).

RESULTS AND DISCUSSION

We separated 3,000 kDa and 440 kDa hemoglo- bins from the three tube worms, Lamellibrachia sp.1, Lamellibrachia sp.2 and Slender ves- timentifera gen. sp.1, by a high-performance gel filtration column of Superose 12. A typical elution profile for the latter species is shown in Fig. 1. The 3,000 kDa hemoglobin is susceptible to dissocia- tion, and its several dissociation products emerge after the 440 kDa hemoglobin as small peaks. The hemoglobins of Lamellibrachia sp.1 and Slender vestimentifera gen. sp.l1 were isolated mainly in the oxygenated form, but that of Lamellibrachia sp.2 was almost in the met (oxidized) form.

The extracellular giant hemoglobins of annelids and tube worms comprise four heme-containing chains, which can be separated into two distinct

440K

SS EEE eee

0 20. 40 min.

Fic. 1. Separation of the 3,000kDa and 440kDa hemoglobins of Slender vestimentifera gen. sp.1 on a gel filtration column of Superose 12. The column was equilibrated and eluted with 50 mM phosphate buffer containing 150 mM NaCl at a flow rate of 0.5 ml/min. Absorbance was monitored at 540 nm.

strains, A and B as suggested by Gotoh er al. [7]. However, the nomenclature of the chains is not unified among the researchers, so it is strongly recommended to propose a common name, like alpha and beta chains of vertebrate hemoglobins, for the corresponding chains to make clear their evolutionary relationship. Gotoh et al. [8, 9] were the first to propose a common name, a, A, b and B for each chain (see Table 1). But not all of the researchers agreed with the nomenclature, mainly because the difference between capital and small characters is difficult to appreciate in the spoken language [9]. Therefore we modified the nomencl- ature proposed by Gotoh et al. [8, 9], and intro- duced a new name (A1, A2, B1 and B2) for each of the heme-containing chains, based on the amino acid sequence homology between the chains, in this paper (Table 1).

The four heme-containing chains of the tube worm 440 kDa hemoglobin were separated by a reverse-phase chromatography. A typical elution profile for Slender vestimentifera sp.1 is shown in

Evolution of Tube Worm Hemoglobins 143

TABLE 1. hemoglobins

Nomenclature for four heme-containing chains of annelid and tube worm

Chain Name (Earlier Nomenclature)

Source Reference Strain A Strain B Lumbricus I II Il] IV Vinogradov et al. [23] Lumbricus d b c a Fushitani et al. [5] Tylorrhynchus I IIA IIC IIB Suzuki et al. [13] Lamellibrachia I Ill IV II Suzuki ef al. [17] Proposed name a A b B Gotoh et al. [8, 9] Al A2 Bl B2 This paper

A(220nm)

0 20

40

Time (min)

Fic. 2.

Separation of the constituent chains of the 440 kDa hemoglobin from Slender vestimentifera gen. sp.1 by

reverse-phase chromatography. The column (Cosmosil 5C;s-300) was eluted with a linear gradient of acetonitrile in 0.1% TFA at a flow rate of 1 ml/min. (a), reduced 440 kDa hemoglobin; (b), rechromatography of chains Al

and B2 after carboxymethylation.

Fig. 2. In this case, chains Al and B2 were not separated (Fig. 2a), but they were separated by rechromatography after carboxymethylation of cysteine residues (Fig. 2b).

The N-terminal amino acid sequences of 16-44 residues of the isolated chains of tube worm 440 kDa hemoglobin were determined by an auto- mated protein sequencer and aligned in Fig. 3 with those of the polychaete Tylorrhynchus [13, 14, 16,

21] and the oligochaete Lumbricus [5, 12]. The amino acid sequences (total 134 residues com- pared) of the homologous chains from the three tube worms showed high homology (76-85%) with each other, suggesting that they are closely re- lated. In addition, the sequence of tube worm hemoglobin has significant homology (45-49%) with those of annelid hemoglobins, which is com- parable to the homology (43%) between the

144 T. Suzuki, T. TAKAGI AND S. OnTA 10 20 30 40

Al chain Lam.sp.1 DeNILQRLKVKMQWAKAYG FGAERAKFGNSLWTSIFNYAP Lam.sp.2 EKCNDLERIKVKMQWAKAYS FSANRAKFGDALWANVFNYAP Ves.sp.1 DNcNILQRIKMKMQWGKAYG TGAKRAEFGDALWANVFNYAP Tyl. TDcGILQRIKVKQQWAQVYS VGESRTDFAIDVFNNFFRTNP Lum. EcLVTEGLKVKLQWASAFG HAHQRVAFGLELWKGILREHP

A2 chain Lam.sp.1 YECGPLQRLKVKRQWAEAYG SGNDREEFGHFIWTHVF Lam.sp.2 DHVcCGPLQRLKVKRQWAEAYG SGNRREDFGHY IWAHVF Ves.sp.1 DTHVcGPLQRLKVKRQWAEAYG SGGRREDFGHYIWAHVF Tyl. SSDHCGPLQRLKVKQQWAKAYG VGHERVELGIALWKSMF Lum. KKQcGVLEGLKVKSEWGRAYG SGHDREAFSQAIWRATF

Bl chain lLam.sp.1 SKFCSEGDATIVIKQW Lam.sp.2 EASDHCHYEDAE I VMKEW Ves.sp.1 TVVSDDcCSYEDAD I VMKEW Tyl. DTCcSIEDRREVQALW Lum. DEHEHCCSEEDHRIVQKQW

B2 chain Lam.sp.1 SSNScTTEDRREMQLMWANVWSAQFTGRRLA I AQAVFKDLFA Lam.sp.2 SNHCTTEDRREMQLMWGNVWSAQFTGRRLA I AQAVFKDLFD Ves.sp.1 SNHCTTEDRKEMQIMWSNVWHAQFTGRRLA I AQAVFNDLFA Tyl. DDCcSAADRHEVLDNWKGIWSAEFTGRRVAI GQAI FQELFA Lum. ADDEDCcSYEDRREIRHIWDDVWSSSFTDRRVAI VRAVFDDLFK

* * Fic. 3. Alignment of N-terminal amino acid sequences of four heme-containing chains of three tube worms with

those of Tylorrhynchus and Lumbricus. Asterisks indicate the invariable residues in all chains. Lam. sp.1, Lamellibrachia sp.1; Lam. sp.2, Lamellibrachia sp.2; Ves. sp.1, Slender vestimentifera gen. sp.1; Tyl.,

Tylorrhynchus ; Lum., Lumbricus.

polychaete and oligochaete hemoglobins.

In annelid-like giant hemoglobins, Cys residues play a particularly important role in the subunit assembly of the giant molecule. They are all participating in either intra- or interchain disulfide bridges [5, 15]. All of the chains of the tube worm hemoglobins conserves Cys-7 that would be used for the formation of intrachain disulfide bridge.

The deep-sea tube worms were placed in a new phylum, Vestimentifera, on the basis of their uni- que outward appearance, such as the very long trunk region and absence of a mouth, gut and anus [10]. Since there is no fossil record on tube worms, it is very hard to get an evolutionary relationship among the tube worms and other invertebrate animals directly. So far both of the 18S ribosomal RNA sequence [4] and hemoglobin sequence [17, 18] suggest that the tube worm and annelid are closely related. In order to get more information on the evolution of the tube worms, we con- structed a phylogenetic tree for the hemoglobin sequences by an unweighted pair group clustering method (Fig. 4). Standard errors are given at each branching point, to help evaluation of the tree [11].

Fig. 4a shows a phylogenetic tree constructed

from the partial hemoglobin sequences (total 134 residues) of the four chains of the three tube worms and two annelids (see Fig. 3). The tree apparently indicates that the tube worms, polychaete Tylorrhynchus and oligochaete Lum- bricus diverged from a common ancestor at almost the same time, and the radiation of tube worms now examined is a relatively recent event. Judging from the cluster of the tube worms, the Slender vestimentifera gen. sp.1 may be included in the genus Lamellibrachia.

We have determined the complete amino acid sequences of the four heme-containing chains of Lamellibrachia sp.1 440kDa hemoglobin, very recently [22]. Fig. 4b shows a phylogenetic tree constructed from the four complete sequences (total 576 residues) of Lamellibrachia sp.1, Tylor- rhynchus and Lumbricus hemoglobins. This tree must be more reliable than that of Fig. 4a, but the branching pattern for the two trees was quite similar, supporting the accuracy of the tree in Fig. 4a.

Of great interest is the divergence time of tube worms and annelids. The evolutionary rate of hemoglobins is roughly estimated to be constant in vertebrates, and Goodman et al. [6] calculated the

Evolution of Tube Worm Hemoglobins 145

Myr BP

Q@ 134 redidues

b

576 redidues

40 50 G0 . FO

% Sequence Identity

Fic. 4.

Lumbricus

Tylorrhynchus

Lamellibrachia sp. 1

Lamellibrachia sp. 2

Slender vest. gen. sp. 1

Lumbricus

Tylorrhynchus

Lamellibrachia sp. 1

Phylogenetic trees for the four heme-containing chains of annelid and tube worm hemoglobins. The tree was

constructed by an unweighted pair-group clustering method using the Poisson corrected sequence difference matrix. Standard errors at the branching points are indicated as boxes. (a), a tree constructed from the partial amino acid sequences of the four chains compared in Fig. 3 (total 134 residues); (b), a tree from the complete sequences of the four chains consisting of total 576 amino acid residues.

divergence time of human alpha and beta globins with 44% sequence identity to be about 450 mil- lion years before present (Myr BP). In addition, the phylogenetic tree constructed from hemoglo- bin sequences shows a good correlation with that from classical taxonomy [6].

Just as in vertebrates, all of the members be- longing to the phylum Annelida expresses abun- dant hemoglobin, indicating that hemoglobin is a physiologically important molecule. Annelida consists of two major classes, polychaete and oli- gochaete, but at present it is very difficult to

estimate the exact divergence time of the two classes from the poor or incomplete fossil records of annelids. Therefore, to introduce a tentative time scale for the evolution of annelid and tube worm hemoglobins, we used the same evolution- ary rate as in vertebrate hemoglobins. Finally, it was estimated that the tube worms, polychaete and oligochaete diverged at almost the same time, about 450 Myr BP, and that the radiation of three tube worms occurred around 100 Myr BP (see Fig. 4). These divergence dates are not unreasonable values, because the fossil records indicate that

146

most of representatives of the living phyla and classes of invertebrates appeared about 450-500

Myr BP [2].

In addition, Fushitani et al. [5]

roughly estimated the divergence time of Lumbri- cus and Tylorrhynchus to be about 450 Myr BP using hemoglobin sequences.

11

REFERENCES

Arp AJ, Childress JJ (1981) Blood function in the hydrothermal vent vestimentiferan tube worm. Sci- ence 213: 342-344

Brasier MD (1979) In “The Origin of Major In- vertebrate Groups” Ed by MR House MR, Academic Press, New York, pp 103-159 (cited in Ref. 4)

Childress JJ, Felbeck H, Somero GN (1987) Sym- biosis in the deep sea. Scientific Amer 256: 106-112 Field KG, Olsen GJ, Lane DJ, Giovannoni SJ, Ghistelin MT, Raff EC, Pace NR, Raff RA (1988) Molecular phylogeny of the animal kingdom. Scien- ce 239: 748-753

Fushitani K, Matsuura MSA, Riggs AF (1988) The amino acid sequences of chains a, b, and c that form trimer subunit of the extracellular hemoglobin from Lumbricus terrestris. J Biol Chem 263: 6502-6517 Goodman M, Moore GW, Matsuda G (1975) Darwinian evolution in the genealogy of hemoglo- bin. Nature 253: 603-608

Gotoh T, Shishikura F, Snow JW, Ereifej K, Vinog- radov SN (1987) Two globin strains in the giant annelid extracellular haemoglobins. Biochem J 241: 441-445

Gotoh T, Suzuki T (1990) Molecular assembly and evolution of multi-subunit extracellular annelid hemoglobins. Zool Sci 7: 1-16

Gotoh T, Suzuki T, Takagi T (1991) Nomenclature of the major constituent chains common to annelid extracellular hemoglobins. In “Structure and Func- tion of Invertebrate Dioxygen Carriers” Ed by SN Vinogradov, OH Kapp, Springer-Verlag, Berlin, pp 279-283

Jones ML (1985) On the Vestimentifera, new phy- lum: six new species, and other taxa, from hyd- rothermal vents and elsewhere. Biol Soc Wash Bull 6: 117-158

Nei M, StephensJC, Saitou N (1985) Methods for computing the standard errors of branching points in an evolutionary tree and their application to molecu- lar data from humans and apes. Mol Biol Evol 2: 66-85

12

13

14

15

16

17

18

19

20

21

22

23

T. Suzuki, T. TAKAGI AND S. OHTA

Shishikura F, Snow JW, Gotoh T, Vinogradov SN, Walz DA (1987) Amino acid sequence of the monomer subunit of the extracellular hemoglobin of Lumbricus terrestris. J Biol Chem 262: 3123-3131 Suzuki T, Furukohri T, Gotoh T (1985) Subunit structure of extracellular hemoglobin from the polychaete Tylorrhynchus heterochaetus and amino acid sequence of the constituent polypeptide chain (IIC). J Biol Chem 260: 3145-3154

Suzuki T, Gotoh T (1986) The complete amino acid sequence of giant multisubunit hemoglobin from the polychaete Tylorrhynchus heterochaetus. J Biol Chem 261: 9257-9267

Suzuki T, Kapp OH, Gotoh T (1988) Novel S-S loops in the giant hemoglobin of Tylorrhynchus heterochaetus. J Biol Chem 263: 18524-18529 Suzuki T, Takagi T, Gotoh T (1982) Amino acid sequence of the smallest polypeptide chain contain- ing heme of extracellular hemoglobin from the polychaete Tylorrhynchus heterochaetus. Biochim Biophys Acta 708: 253-258

Suzuki T, Takagi T, Ohta S (1988) N-Terminal amino acid sequence of the deep-sea tube worm hemoglobin remarkably resembles that of annelid haemoglobin. Biochem J 255: 541-545

Suzuki T, Takagi T, Ohta S (1990) Primary struc- ture of a constituent polypeptide chain (AIII) of the giant haemoglobin from the deep-sea tube worm Lamellibrachia. A _ possible H,S-binding site. Biochem J 266: 221-225

Suzuki T, Takagi T, Ohta S (1990) Primary struc- ture of a linker subunit of the tube worm 3000-kDa hemoglobin. J Biol Chem 265: 1551-1555

Suzuki T, Takagi T, Okuda K, Furukohri T, Ohta S (1989) The deep-sea tube worm hemoglobin: Sub- unit structure and phylogenetic relationship with annelid hemoglobin. Zool Sci 6: 915-926

Suzuki T, Yasunaga H, Furukohri T, Nakamura K, Gotoh T (1985) Amino acid sequence of polypeptide chain IIB of extracellular hemoglobin from the polychaete Tylorrhynchus heterochaetus. J Biol Chem 260: 11481-11487

Takagi T, Iwaasa H, Ohta S, Suzuki T (1991) Primary structure of 440 kDa hemoglobin from the deep-sea tube worm Lamellibrachia. In “Structure and Function of Invertebrate Dioxygen Carriers” Ed by SN Vinogradov, OH Kapp, Springer-Verlag, Berlin, pp 247-252

Vinogradov SN, Shlom JM, Hall BC, Kapp OH, Mizukami H (1977) The dissociation of Lumbricus terrestris hemoglobin: a model of its subunit struc- ture. Biochim. Biophys. Acta 492: 136-155

ZOOLOGICAL SCIENCE 10: 147-154 (1993)

© 1993 Zoological Society of Japan

Two New Species of the Genus Staphylocystis (Cestoda: Hymenolepididae) from the House Shrew, Suncus murinus, in Nepal

Isamu Sawapa!, Kazuniro Koyasu~ and

KRISHNA CHANDRA SHRESTHA®

‘Biological Laboratory, Nara Sangyo University, Sango, Nara 636, *the Second Department of Anatomy, School of Dentistry, Aichi-Gakuin University, 1-100 Kusumoto-cho, Chikusa-ku, Nagoya 464, Japan and *Department of Zoology, Pri-Chandra Campus, Kathmandu, Nepal

ABSTRACT—Two new species of the cestode parasite, Staphylocystis (Staphylocystis) kathmanduensis sp. nov. and S. (S.) trisuliensis sp. nov. are described from the house shrews, Suncus murinus of Kathmandu and Trisuli, respectively. The former is related to, but different from S. (S.) delicata Sawada et Koyasu, 1991 in the length and number of the rostellar hooks, and the size of the rostellum. The latter is related to, but different from S. (S.) dsinezumi Sawada et Koyasu, 1990 in the rostellar hooks. The house shrew, Suncus murinus, one kind of the commensal mammals, are widly distributed in Asia and are found infected with a great number of different cestodes. The difference between the two cestode species infecting Suncus murinus collected respectively at Kathamandu and Trisuli is

discussed according to the hosts’s behavior patterns.

INTRODUCTION

The cestode parasites of the house shrew, Sun- cus murinus, in Nepal are little known except the one reported by Sawada and Koyasu [11], who described a new species, Pseudhymenolepis nepalensis from Suncus murinus collected at Kath- mandu. Since then, no attempts have been made to study the cestode parasites of Suncus murinus, although it is quite commonly found in Nepal. This paper reports anothor two new hymenolepi- did cestodes obtained from Suncus murinus col- lected at Kathmandu and Trisuli, and discusses the difference between the two new species from the point of view of the host’s behavior patterns [1, 3].

MATERIALS AND METHODS

Seven specimens of Suncus murinus were col- lected with traps at Kathmandu and Trisuli in

Accepted September 5, 1992 Received March 30, 1992

March, 1991, and were examined for cestodes in connection with the previous investigation (Fig. 1). The shrews were autopsied immediately after capture, and their guts were removed and fixed in Carnoy’s fluid, and maintained until the investiga- tion in Japan. The methods used have been described in the previous paper [10]. All measure- ments are given in millimeters unless otherwise stated.

Staphylocystis (Staphylocystis) kathmanduensis sp. nov. (Fig. 2-6)

From March 17 to 31, 1991, four house shrews, Suncun murinus, were captured at Kathmandu. One of them harbored five mature specimens of this cestode.

Description (based on five specimens): Small- sized hymenolepidid; mature worm length 7.1—8.3 and maximum width 0.8-0.9. Metamerism dis- tinct; margin slightly serrate. Scolex round, 0.221-

148

I

INDIA

. SAWADA, K. KoyaAsu AND K. C. SHRESTHA

CHINA

SAGARMATHA (Mt.Everest)

Trisulie

BHUTAN

* Kathmandu

200Km BANGLADESH

Fic. 1. Map of Nepal showing the localities of the house shrews collected.

SEO ey

Cy,

Fics. 2-6. Staphylocystis (Staphylocystis) kathmanduensis sp. nov. 6: Egg.

2: Scolex.

3: Rostellar hooks.

4: Rostellar hooks magnified. 5: Mature segment drawn from a projected

microphotographic negative, dorsal view. 6: Egg.

Two New Species of the Genus Staphylocystis 149

0.235 in length by 0.290-0.456 in width. Rostel- lum pyriform, 0.056 long by 0.070 wide, armed with a single row of 13 thorn-shaped hooks 0.018 long. Hook handle short; guard bluntly round at its end, shorter than blade; blade slender, sharp at its end, curved toward guard. Rostellar sac slightly elongated, 0.161-0.189 long by 0.119—0.126 wide, extending past posterior margin of suckers. Suck- ers discoid, 0.111 in diameter.

Genital pores unilateral, situated a little anterior to middle of segment margin. Testes three in number, oval, 0.070-0.084 long by 0.028-0.035 wide, arranged in a transverse row, one poral and two aporal. Cirrus sac pryiform, 0.105—0.126 long by 0.035-0.042 wide, extending beyond longitu- dinal excretory canals. Internal seminal vesicle 0.049-0.056 long by 0.028-0.035 wide, occupying almost whole of cirrus sac. External seminal vesicle 0.070 long by 0.021—0.028 wide. Ovary transversely elongate, bilobate, 0.105—0.140 wide. Seminal receptacle large, dorsal to ovary, 0.112- 0.140 long by 0.035—0.042 wide. Vitelline gland bilobate, 0.049-0.070 long by 0.028-0.035 wide. Eggs elliptical, 0.049-0.053 in major axis and 0.032 in minor axis. Embryophore 0.032 by 0.028. Onchospheres_ spherical, 0.028 in diameter; embryonic hooks 0.014 long.

Host: Suncus murinus (Insectivora: Soricidae).

Habitat: Small intestine.

Locality and date: Kathmandu, Nepal; March SISO:

Type specimens: Holotype, Nara Sangyo Univ. Lab. Coll. No. 9300; paratypes, 9301-9302.

Remarks: About 22 species of Staphylocystis (Staphylocystis) have been recorded from the Sori- cidae [9, 10, 12, 13]. Of these, the species armed with 10-15 rostellar hooks ranging in length from 0.015 to 0.021 are: S. (S.) minutissima (Meggitt, 1927) Yamaguti, 1959 [4]; S. (S.) pauciproglottis (Neiland, 1953) Yamaguti, 1959 [6]; S. (S.) sun- cusensis Olsen et Kuntz, 1978 [8]; S. (S.) curiosiha- mata Sawada et Koyasu, 1990 [10]; S. (S.) naga- noensis Sawada et Koyasu, 1990 [10]; and S. (S.) delicata Sawada et Koyasu, 1991 [12]. The present new species most closely resembles S. (S.) delicata in the shape of the rostellar hooks. However, the species is distinguished from S. (S.) delicata by the larger number (13 against 10) and longer size

(0.018 against 0.014) of the rostellar hooks, and the larger rostellum (0.056 by 0.070 against 0.028 by 0.035).

Staphylocystis (Staphylocystis) trisuliensis sp. nov. (Fig. 7-12)

On March 20 and 21, 1991, three house shrews, Suncus murinus, were captured at Trisuli. All of them were found infected with one or two mature cestodes.

Description (based on four specimens): Small- sized hymenolepidid; mature worm 9.2—10.3 long by 0.8-0.9 wide. Mature segment serrate and wider than long. Scolex 0.140-0.175 long by 0.266-0.280 wide, sharply demarcated from neck. Rostellum oval, 0.056-0.070 long by 0.070-0.091 wide, armed with a single row of 21-22 chelate- shaped hooks 0.018 long. Hook handle compara- tively long; blade long, slender and pointed; guard shorter than blade and thick. Rostellar sac oval, 0.126-0.161 long by 0.070-0.140 wide. Suckers discoid, 0.119—0.026 in diameter.

Genital pores unilateral, located a little anterior to middle of segment margin. Testes three in number, oval, 0.098-0.105 long by 0.035-0.049 wide, arranged in a transverse row, one poral and two aporal. Cirrus sac pyriform, 0.126—0.140 long by 0.042 wide, extending beyond longitudinal ex- cretory canals. Internal seminal vesicle 0.091- 0.105 long by 0.035—0.042 wide, occupying almost whole of cirrus sac. External seminal vesicle 0.105-0.126 long by 0.042-0.049 wide. Ovary transversely elongated, bilobate, 0.154—0.175 wide. Voluminous seminal receptacle measuring 0.119-0.140 long by 0.070-0.098 wide. Vitelline gland irregularly lobate, situated in posterior field of segment, 0.070-0.091 long by 0.035-0.042 wide. Eggs elliptical, 0.039-0.042 in major axis and 0.028-0.032 in minor axis, with at each pole a round projection provided with polar filaments. Onchospheres spherical, 0.025 in diameter; em- bryonic hooks 0.011-0.014 long.

Host: Suncus murinus (Insectivora; Soricidae).

Habitat: Small intestine.

Locality and date: Trisuli, Nepal; March 20 and Al, MEM.

Type specimens: Holotype, Nara Sangyo Univ.

150 I. SawaDA, K. Koyasu AND K. C. SHRESTHA

0.2 mm

Fics. 7-12. Staphylocystis (Staphylocystis) trisuliensis sp. nov. 7: Scolex 8: Scolex magnified. 9: Rostellar hooks. 10: Rostellar hooks magnified. 11: Mature segment drawn from a projective microphotographic negative, dorsal view. 12: Egg.

TABLE 1. A comparison of related species of Staphylocystis (Staphylocystis) armed with 18-24 rostellar hooks ranging in length from 0.020 to 0.029 mm from the Insectivora

Rostellar hooks

Species ; Host number length (mm) 1. S. (S.) chrysochloridis [2] 16-18 0.029 Chrysochloria capensis Ch. aurea 2. S. (S.) furcata [14] 22-28 0.026-0.028 Sorex araneus

Suncus murinus Neomys fodiens 3. S. (S.) dsinezumi [10] 23 0.020 Crocidura dsinezumi . (S.) sindensis [5] 20 0.022-0.023 Suncus murinus sindensis

i)

Lab. Coll. No. 9303; paratypes, 9304-9309. Remarks: Out of the 22 known species of

Staphylocystis (Staphylocystis) from the Soricidae

[9, 10, 12, 13], four; S. (S.) chrysocholoridis

Two New Species of the Genus Staphylocystis 151

(Janicki, 1904) Spassky, 1950 [2], S. (S.) furcata (Stieda, 1862) Spassky, 1950 [14], S. (S.) sindensis Nama, 1976 [5] and S. (S.) dsinezumi Sawada et Koyasu, 1990 [10] are armed with 18-24 rostellar

TABLE 2. Suncus spp. and their cestode parasites in Asia ([1, 7, 13], the present study)

Locality Suncus spp.

Japan Kyushu Suncus murinus temmincki Okinawa 4

Taiwan

Taoyuan Hsien

Nantou Hsien

Ping Toung County

China Southern China Hainan Dao

Vietnam Saigon-Cholon Nha Trang Con Son Island

Bangladesh Mymensingh

Thailand Chanthaburi

Pakistan

Karachi

Myanmer Rangoon

S. murinus swinhoei

4

4

. murinus

4

. murinus

Za

4

. murinus

. murinus

. murinus tytleri . murinus sindensis

a

. etruscus . stoliczkanus

. murinus

Cestode parasites

*

Vampirolepis jakounezumi Sawada et Hasegawa, 1991

V. okinawaensis Sawada et Hasegawa, 1991

V. grascilistrobila Sawada et Harada, 1989

Stapiyloaystts (Staphylocystis) suncusensis Olsen et Kuntz,

Rodentolepis sp. Uchikawa, Sakumoto et Kinjo, 1981

V. sunci Sawada et Harada, 1989

V. gracilistrobila Sawada et Harada, 1989 V. sessilihamata Sawada et Harada, 1989 S. (S.) suncusensis Olsen et Kuntz, 1978 S. (S.) delicata Sawada et Koyasu, 1991

S. (S.) furcata (Stieda, 1862) Spassky, 1950 V. microscolex Sawada et Koyasu, 1991

V. nana (Siebold, 1852) Spassky, 1954

Raillietina (Raillietina) madagascariensis (Davaine, 1869) Fuhrmann, 1920

Syn. R. (R.) siriraji Chandler et Pradatsundarasar, 1957 V. jacobsoni (Linstow, 1907) Schmidt, 1986 Hymenolepis mujibi Bilgees et Malik, 1974

*

*

S. (S.) minutissima (Meggitt, 1927) Yamaguti, 1959 S. (S.) furcata (Stieda, 1862) Spassky, 1954 S. (S.) solitaria (Meggitt, 1927) Yamaguti, 1959

152

Afghanistan

Jalalabad, Laghman S.

Singapore

Malaysia Sabah Province Sarawak

Indonesia Kalimantan

Java Island Flores Island

India Sanchore Jadhpur

Allahabad

Bombay Khrhja Lucknow South India

Sri Lanka Horton Plains

Nepal Kathmandu

Adhabar Trisuli

Philippines Palawan Island

Luzon Island

Northern Marianas Guam Island

* Unknown

I. SAwaDA, K. KoyAsu AND K. C. SHRESTHA

murinus

. murinus . etruscus malayanus

murinus

S. hosei

S.

AnNHHALD

. etruscus

ater

. murinus . murinus

. mertensi

. murinus sindensis . murinus sindensis

. murinus

. murinus . murinus

4

. striatus

dayi

. stoliczkanus

. murinus montanus

. etruscus

. murinus

. etruscus . stoliczkanus

. murinus

murinus

. occultidenus . palawanensis . luzoniensis

murinus

V. jacobsoni (Linstow, 1907) Schmidt, 1986 Hymenolepis sunci Vaucher et Tenora, 1971

* * *

V. jacobsoni (Linstow, 1907) Schmidt, 1986

*

(S.) sanchorensis Nama et Kichi, 1975 (S.) sindensis Nama, 1979 . bhali (Singh, 1958) Schmidt, 1986 . molus Srivastava et Capoor, 1979 . allahabadensis Srivastava et Pandey, 1982 (S.) indicus Nanda et Malhotra, 1990 V. jacobsoni (Linstow, 1907) Schmidt, 1986 Pseudhymenolepis guptai Gupta et Singh, 1987

A Ss Ss) SS} KA)

Pseudhymenolepis suncusi Gupta et Sinha, 1984

*

*

V. montana Crusz et Sanmugasunderam, 1971 Pseudhymenolepis eisenbergi Crusz et Sanmugasunderam, 1971

*

Pseudhymenolepis nepalensis Sawada et Koyasu, 1991 S. (S.) kathmanduensis sp. nov.

*

*

S. (S.) trisuliensis sp. nov.

Two New Species of the Genus Staphylocystis

AQ KS 5h NS

0.01mm

Fic. 13. Comparison of rostellar hook in shape among four related species. 1: S. (S.) chrysochloridis [2] 2: S. (S.) furcata [14] 3: S. (S.) dsinezumi [10] 4: S. (S.) sindensis [5]

hooks ranging in length from 0.015 to 0.025 (Table 1). The present new species most closely resem- bles S. (S.) dsinezumi in the number and length of the rostellar hooks. However, the shape of the rostellar hooks separates this new species from S. (S.) dsinezumi (Fig. 13).

DISCUSSION

There are quite a number of different cestodes infecting Suncus murinus in Asia (Table 2) ((1, 7, 13], the present study). The following is thought to be one of the reasons. Because the behavior patterns of predation displayed by Suncus murinus are similar to those of commensal mammals, Rat- tus norvegicus and Mus musculus [1, 3], their eating habits are thought to overlap with each other resulting probably in a diversity in the in-

2000m

Indian plains

Kathmandu

153

termediate hosts of the cestodes infecting it. So, varying with the area where Suncus murinus lives, the species of cestodes infecting it differ as much.

Even though the areas of Kathmaudu and Trisu- li are separated by less than 30 km, the species of tapeworms differ. Suncus murinus cannot take low temperature (below 0°C) and the winter in Kath- mandu (1350 m) is extremely harsh. The harsh coldness of winter causes their population crash, thus greatly decreasing the number of individuals which can, after surviving the cold season, bear offspring. Nonetheless, when May comes round, they come to appear in various places. This fact suggests, in order to recover the population crash, they represent annually a presence of dramatic fluctuation among the individuals which have sur- vived the winter season. Between the Indian Plains and Kathmandu, and Kathumandu and Tri- suli there are ranges of 2000 m plus mountains, so it cannot be presumed that Suncus murinus mi- grates between the three areas (Fig. 14), but in the past there have been numerous cases in which Suncus murinus was introduced to each of the areas, so many hereditary changes can be recog- nized. Evidence for this can be seen in the fact that Pseudhymenolepis nepalensis Sawada et Koyasu, 1990 were found infecting Suncus murinus in Kath- mandu but not those in Trisuli.

Suncus murinus can be found in Kathmandu and Trisuli, wherever there are human dwelling and people often feed them. Looking at this type of

SAGARMATHA (Mt.Everest) 8848m

NEPAL

2000m

A

1350m Trisuli

300m

Fic. 14. Topographical map showing the heights above sea level of Kathmandu and Trisuli.

154 I. SawabDA, K. KoyaAsu AND K. C. SHRESTHA

environment, over many years Suncus murinus of both areas have formed characteristic population, and since the type of intermediate hosts for ces- todes in both areas is fixed, it can be seen that species of cestodes infecting Suncus murinus in the two areas differ from each other.

ACKNOWLEDGMENTS

We hereby wish to acknowledge our indebtedness to Mr. A. M. Vaidya, Pension SAKURA, Kathmandu, Nepal, for kind help in collecting shrews.

REFERENCES

1 Corbet GB, Hill JE (1991) A world list of mamma- lian species. 3rd Nat Hist Mus Publ and Oxford Univ Pres, London and Oxford, pp 243

2 Janicki C (1904) Zur Kenntnis einiger Saugetierces- toden. Zool Anz 27: 770-782

3 Marshall JD, Quy DV, Gibson FL, Dung TC, Cavanaugh DC (1967) Ecology of plague in Viet- nam. 1. Role of Suncus murinus. Proc Soc Exp Biol 124: 1083-1086

4 Meggitt FJ (1927) On cestodes collected in Burma. Parasitology 19: 141-153

5 Nama HS (1976) On a new species of Staphylocystis Villot, 1877 (Cestoda: Hymenolepididae) from Sunucus murinus sindensis. Acta Prasitol Pol 24: 19- DD

6

10

11

12

13

14

Neiland KA (1953) Helminths of Northwestern mammals. Part V Observations on cestodes of shrews with the descriptions of new species of Liga Weinland, 1857, and Hymenolepis Weinland, 1858. J Parasitol 39: 487-494

Ohbayashi M (1985) Helminth parasites of the Soricidae. In “SUNCUS MURINUS”. Ed by S Oda J Kitoh, K Ohta, G. Isomura Jpn Sci Soc Press, Tokyo, pp 88-93

Olsen OW, Kuntz RE (1978) Staphylocystis (Staphylocystis) suncusensis sp. n. (Cestoda: Hyme- nolepididae) from the musk shrew, Suncus murinus (Soricidae) from Taiwan, with a key to the known species of Staphylocystis Villot, 1877. Proc Helmin- thol Soc Wash 45: 182-189

Sawada I, Harada M (1990) Cestodes of field micromammalians (Insectivora) from Central Hon- shu, Japan. Zool Sci 7: 467-475

Sawada I, Koyasu K (1990) Further studies on cestodes from Japanese shrews. Bull Nara Sangyo Univ 6: 187-202

Sawada I, Koyasu K (1991) Pseudhymenolepis nepalensis sp. nov. (Cestoda: Hymenolepididae) pa- rasitic on the house shrew, Suncus murinus (Sorici- dae), from Nepal. Zool Sci 8: 575-578

Sawada I, Koyasu K (1991) Further studies on cestode parasites of Taiwanese shrews. Bull Nara Sangyo Univ 7: 131-142

Schmidt GD (1986) Handbook of tapeworm iden- tification. CRC Press, Inc., Florida, pp 675

Stieda L (1862) Ein Beitrag zur Kenntnis der Taenien. Arch Naturg 28: 208-209

ZOOLOGICAL SCIENCE 10: 155-159 (1993)

© 1993 Zoological Society of Japan

Two New Species of Sabacon from Sichuan Province, China (Arachnida: Opiliones: Sabaconidae)

Nosuo Tsurusakt! and DAxIANG SONG?

"Department of Biology, Faculty of Education, Tottori University, Tottori, 680 Japan, and *Institute of Zoology, Academia Sinica, 19 Zhongguancun Lu, Haitien, Beijing, China

ABSTRACT— Two new species of the genus Sabacon, S. martensi n. sp. and S. gonggashan n. sp. are described based on the specimens collected from Mt. Gong-ga-shan of the Da-xue-shan Mountains, Sichuan Province, southern China. The two species are similar to each other and to some species of the genus from Nepal-Himalayas (S. chomolongmae, S. dhaulagiri, and S. unicornis) and Japan (S. dentipalpe and S. imamurai) in having dorsolateral spurs on the fixed fingers of male chelicerae.

INTRODUCTION

The genus Sabacon (Sabaconidae, Ischyropsali- doidea, Palpatores) is a group of soil harvestmen, with about 40 species, and is disjunctively found in the Holarctic regions. The distribution covers the eastern (2 spp.) and western areas (5 plus some spp.) of North America (7 described spp. in total [5, 1], Cokendolpher pers. comm., 1992), southern part of Europe (6 spp. [3]), the Nepal-Himalayas (7 spp. [2]), Siberia (2 spp. [4]), and East Asia including Japan (10 spp. [6, 7]). From China, however, only one species, Sabacon okadai Suzu- ki, has been recorded [6].

During an examination of the opilionids col- lected from the Da-xue-shan Mountains, Sichuan Province, southern China, we found two unde- scribed species of the genus. Descriptions of the species will be presented here.

All the specimens examined are deposited in the collection of the Institute of Zoology, Academia Sinica, Beijing.

Genus Sabacon Simon, 1879 Sabacon martensi n. sp. (Figs. 1-2)

Material. Male holotype, West slope of Mt.

Accepted October 15, 1992 Received September 17, 1992

Gong-ga-shan, Kangding, Sichuan Province, Chi- na, 2 September 1982, Zhang Xue-zhong leg.

Measurements (in mm). Male holotype: cepha- lothorax, 0.96 long, total body length 2.3.

Length of palp and legs: Palp (femur/patella/ tibia/tarsus; total): 0.88/0.98/1.10/0.50; 3.46. Legs (femur/patella/tibia/metatarsus/tarsus; to- tal): Leg I: 1.75/0.79/1.78/2.40/2.40; 9.12. Leg II: 2.57/0.99/2.80/3.76/4.10; 14.22. Leg III: 1.68/0.78/1.60/2.73/2.80; 9.59. Leg IV: 2.32/ 0.94/2.24/3.60/3.72; 12.82.

Male. Body (Fig. 1A-B) relatively small, poorly sclerotized. Eye tubercle low, slightly canalicu- late, unarmed. Second thoracic tergite with pair of postocular spines. Abdominal tergites represented by small, weakly sclerotized plates (scutum laminatum/dissectum) with few scattered setae. Abdominal sternites poorly sclerotized, with mi- nute black setae. Chelicera (Fig. 1C-E) with basal article flat, dorsally not elevated; fixed finger dor- solateraly with a conspicuous black spur, ventrally with a small knob. Palp (Fig. 1F) relatively slen- der, patella distally with row of six ventromesal teeth.

Penis (Fig. 2) 2.13 mm long (including glans), 0.16 mm wide at base; pigmented, laterally with a row of several denticles on each side.

Coloration: Venter brownish white, with coxae, abdominal sclerites pale brown; dorsum brownish white with tergites and cephalothorax brown. Eye tubercle dark brown. Chelicerae, palps and legs

156 N. TsuRUSAKI AND D. SONG

Fic. 1. Sabacon martensi n. sp., holotype male. A-B: Lateral (A) and dorsal (B) views of body. C-E: Dorsal (C), ectal (D), and mesal (E) views of left chelicera. F: Mesal view of left palp. Scales=1 mm.

light brown.

Female. Unknown.

Distribution: Known only from the type locality.

Etymology. The specific epithet is given in honor of Prof. Jochen Martens, Mainz, Germany, for his eminent contributions to the systematics of Saba- con species.

Remarks. This species has affinities with some species of the genus from the Nepal-Himalayas and Japan (S. chomolongmae, S. dhaulagiri, and S. dentipalpe, etc.). These species share a dorso- lateral spur on both chelicerae and a similar mor- phology in the penes. The new species, however, can be easily distinguished from any other de- scribed species of the genus, by its penis having a row of denticles on both lateral margins of the shaft.

Fic. 2. Sabacon martensi n. sp., holotype male. A-B: Ventral and lateral views of penis. C-E: Ventral (C), lateral (D), and dorsal (E) views of distal part of penis. Scales=1 mm for A-B, 0.5 mm for C-E.

New Harvestmen from China 157

Fic. 3. Sabacon gonggashan n. sp. Dorsal view of body. A, holotype male; B, paratype female. Scale=1 mm.

Fic.4. Sabacon gonggashan n. sp. A, Anterior ventral surface of male (holotype) body. B-C: Mesal (B) and ectal (C) views of male left chelicera. D, Mesal view of male left palp. E-F: Mesal (E) and ectal (F) views of female (paratype) chelicera. G, Mesal view of female left palp. Scale=1 mm.

158 N. TsuRUSAKI AND D. SONG

Sabacon gonggashan n. sp. (Figs. 3-5)

Material. Male holotype, male paratype and two female paratypes, West slope of Mt. Gong-ga- shan, Kangding, Sichuan Province, China, 2 September 1982, Zhang Xue-zhong leg.

Measurements (in mm). Male holotype (one of the female paratypes in parentheses): cephalotho- rax 1.50 (1.60) long, 2.1 (2.67) wide; abdomen 2.18 (2.82) wide; total body length 4.32 (4.56).

Length of palp and legs: Palp (Fe/Pa/Ti/Ta; total): 1.05/1.06/1.12/0.56; 3.79 (1.52/1.45/1.74/ 0.75; 5.46). Legs (Fe/Pa/Ti/Mt/Ta; total): Leg I: 1.83/0.89/1.66/2.46/2.41; 9.25 (legs I to III of the paratype female measured are absent). Leg II: 2.16/1.00/2.20/3.08/3.88; 12.32. Leg III: 1.80/ 0.80/1.56/ 2.72/2.72; 9.6. Leg IV: 2.40/1.01/ 1E95//32547/3290 2285 (2252/0295) 2elG/ S57 0S 22: 12.55).

Male. Body (Fig. 3A) poorly sclerotized. Eye tubercle low, slightly canaliculate above, unarmed. Second thoracic tergite with pair of postocular spines. Abdominal tergites slightly sclerotized, with few scattered setae, disposed in a manner “scutum laminatum”. Abdominal sternites weakly sclerotized, with few scattered setae. Genital operculum pointed anteriorly, set with numerous black setae (Fig. 4A). Chelicera (Figs. 4B-C) with three slit sensilla on ectal surface of basal segment; with a black cone-shaped spur on apico-ectal sur- face of fixed finger. Palp (Fig. 4D) relatively slender; patella with five large ventromesal teeth followed by a few small denticles. Legs short, with fine setae.

Penis (Fig. 5A-D) 2.87 mm long, simple and slender, laterally with a pair of setae near the apical end of the shaft.

Coloration: Cephalothorax yellowish brown, marked blackish brown near margins, eye tubercle brown, abdominal tergites light brown. Venter yellowish white, lightly mottled brown.

Female. Similar to male but with tergites repre- sented by small median sclerites (Fig. 3B). Che- licera without an apico-ectal spur on fixed finger (Fig. 4E-F). Palp (Fig. 4G) slender, without ven- tral denticles on patella. Ovipositor (Fig. 5E), 2.09 mm long, elongate, scattered with numerous

Fic. 5. lateral (B) views of penis (holotype male). C-D: Ventral and lateral views of distal part of the penis. E, Ventral view of ovipositor (paratype female). Scale=1 mm for A-B, E, and 0.5 mm for C-D.

Sabacon gonggashan n. sp. Ventral (A) and

short setae. Coloration as in male, but with tergites brown.

Distribution. Known only from the type locality.

Etymology. The specific epithet is a noun in apposition.

Remarks. This species is characterized by its simple and elongate penis. The presence of apico- ectal spurs on the fixed fingers of male chelicerae suggests its close affinity with S. martensi from the same mountain; some Nepal-Himalayan congen- ers, such as Sabacon chomolongmae, S. dhaulagiri, S. unicornis [2]; and some Japanese counterparts belonging to the dentipalpe-group [7] like S. denti- palpe and S. imamurai. Phylogenetic proximity to the dentipalpe-group, a group which was recog- nized by Suzuki [7] within Japanese species of Sabacon, is also inferred from its elongate oviposi-

New Harvestmen from China 159

tor with numerous setae.Although the present species, as well as S. martensi, were found on the west slope of Mt. Gong-ga-shan, it is uncertain if the two species are sympatric.

ACKNOWLEDGMENT

We are grateful to Mr. James C. Cokendolpher, Lubbock, Texas, for his review of the draft.

REFERENCES

1 Cokendolpher JC (1984) A new species of Sabacon Simon from Oregon (Arachnida: Opiliones: Sabaco- nidae). Can J Zool 62: 989-991.

2 Martens J (1972) Opiliones aus dem Nepal- Himalaya, I. Das Genus Sabacon Simon (Arachnida:

Ischyropsalididae). Senckenbergiana biol 53: 307- 323.

Martens J (1983) Europdische Arten der Gattung Sabacon Simon 1879 (Arachnida: Opiliones: Sabaco- nidae). Senckenbergiana biol 63: 265-296.

Martens J (1989) Sibirische Arten der Gattung Saba- con Simon 1879 (Arachnida: Opiliones: Sabaconi- dae). Senckenbergiana biol 69: 369-377.

Shear WA (1975) The opilionid genera Sabacon and Tomicomerus in America (Opiliones, Troguloidea, Ischyropsalidae). J Arachnol 3: 5-29.

Suzuki S (1941) Opiliones from Manchoukuo and North China, with a description of the interesting genus Sabacon (Ischyropsalidae). (In Japanese with English description and summary). Bull Biogeogr Soc Jpn 11: 15-22.

Suzuki S (1974) The Japanese species of the genus Sabacon (Arachnida, Opiliones, Ischyropsalididae). J Sci Hiroshima Univ (B-1) 25: 83-108.

spn

Fe eit a a agen’. hegoibite es

100 sb wont ei (oe bane

np? -ahinton AY PUNT nemic are

SAL ds | "Sa toel venti iat f ‘fee moa ab as A cudinnie (edt. fra $e esegemgleg Ct ‘yimbewe VOMTy pola 3

TRUM) edoil meaebeyerin gonad - (ant

Wuiled, ayes Lin Hin aT (ML Ae eee

pet) anole): Orie fC UPI tay OF fiulseies | (oa inegerry iad

asafvil aks ceil Jae cel) eT A ine I

i yo eoahhirest 2 whe “eee ree jel. (Giles aay Houle ainsy tts Hae J ve hotter ho een eel dikes ee ree GE in|

, , r Ci SPSL, at Py ess ray PS, P bitiars reel -ceaistea biel F bin goels ‘ aie . t 17 } = 1) yo) leristrl eft ns? = is? oF

_stneh gab bile. ubienisatA ) pret

Fre (pa 1) { = . ~ n ee H & Aus aa bd

pt

waa

hans: eT)

Theg i og

Dut} OES Aga de

eon oF

c eon way u J

ZOOLOGICAL SCIENCE 10: 161-167 (1993)

© 1993 Zoological Society of Japan

Genetic Evidence for the Presence of Distinct Fresh-water Prawn (Macrobrachium nipponense) Populations in a Single River System

Kazuo Masuiko! and KEn-1cHt NUMACHI”

‘Biological Laboratory, Teikyo University, Hachioji, Tokyo 192-03, and ?Ocean Research Institute, University of Tokyo, Minamidai, Nakano-ku, Tokyo 164, Japan

ABSTRACT— Genetic difference between two groups of individuals in the freshwater prawn Macro- brachium nipponense, which exhibit characteristically different egg sizes within a single river system, was examined electrophoretically and morphologically. Out of 16 prospective loci, PGM and MPI-1 were polymorphic and allelic frequencies were significantly different between the two groups in both loci. In addition, rostral tooth count (number of spines on the dorsal margin of rostrum), which is regarded to be genetically controlled on the basis of crossing experiments, varied significantly between the two groups. These facts indicate that the two groups of individuals have genetically differentiated.

INTRODUCTION

A previous study on the freshwater prawn Mac- robrachium nipponense (de Haan) [14] revealed that egg size remarkably varied from one local population to another, with an apparent rela- tionship between egg size and the hydrogeographic feature of the habitats. Relatively large eggs (approximately 0.10 mm? per egg) were associated with freshwater lakes and rivers, small eggs (0.05 mm7*) with estuaries, and intermediate-sized eggs with brackish-water lagoons. Such populations with different-sized eggs were occasionally found even within a single water system, for instance, the Sagami River in central Japan where individuals with large and small eggs were found in the upper basin of the river and in the estuary, respectively [12]. Crossing experiments between these two groups of individuals, where the size of eggs pro- duced by F, hybrids was intermediate between those of the parents, suggested that the different egg sizes among local populations of this species are controlled as a quantitative genetic trait [15]. The two groups were also different in larval phy- siological characteristics [13]. These facts all imply

Accepted October 15, 1992 Received August 24, 1992

that M. nipponense is now differentiating into distinct local populations through modification of some life-history traits.

In the present study, the genetic difference between the two groups of M. nipponense in the Sagami River was further confirmed by means of both biochemical and morphological analyses. Electrophoretically different forms of enzyme with the same metabolic function (isozyme), especially at the same gene loci (allozyme), offer very useful information about population structures and sys- tematics [e.g., 19, 23]. Chow and Fujio [4] have attempted this approach to M. nipponense, inde- pendently of our study without discriminating egg size, among a few local populations. Similarly, several authors [2, 6, 9, 10, 26] have made electro- phoretic analyses among conspecific populations or different species in some taxonomic groups of freshwater prawns and shrimps. However, multiple conspecific populations coexisting in the same water system with different life-history traits was beyond the scope of those foregoing investigations, except the study by Chow et al. [5] on an inland-water prawn Palaemon paucidens. In the present study, in addition to the electropho- retic analysis, a morphological trait (rostral tooth count) was compared between the two groups of M. nipponense in the Sagami River with respect to

162 K. MASHIKO AND K. NuMACHI

its genetic bases.

MATERIALS AND METHODS

Adults were collected from the estuary (station Banyu) and the upper basin (stations Ogura and Lake Tsukui, approximately 35 and 40 km up- stream, respectively, from the river mouth) of the Sagami River during the period 1985-1989. Topography of these collecting sites has been described previously in detail with a map [12].

Samples were kept frozen at below —50°C or kept alive in aquaria until electrophoresis. Ap- proximately 0.4 g of abdominal muscle was minced and homogenized in 0.5 ml distilled water contain- ing 25 wg NADP and centrifuged. The super- natant was absorbed into a filter paper wick and inserted into gel blocks. Horizontal starch-gel electrophoresis was run at the constant current of 4 to 5mA/cm? (approximately 30 V/cm) in a cold chamber. During electrophoresis the gel was cooled with a glass-plate pan containing ice water, which was set on the gel. Usually electrophoreses were stopped when the marker of amid black 10B migrated 5cm from the origin. The enzymes surveyed, along with their buffer systems, are summarized in Table 1. The MES/TEA buffer system (pH 7.0) was newly developed for separa- tion of MPI by one of us (K.N.), where the gel contained 0.6 mg ATP/ml. After electrophoresis, each gel was sliced horizontally into 1 mm-thick sections and stained. The enzyme was detected according to Numachi [21], excepting ACP, HK, AO, MPI and GAPDH which were detected according to Redfield and Salini [22]. In the staining for PGM, MPI, HK, GPI and PEP, agar- ose solution (final concentration of 0.9%) contain- ing reactants was overlaid on the gel. The stained starch gels were dried and preserved according to Numachi [20] for later analysis. The agarose gel was dried on a gel bond film.

The number of teeth on the dorsal margin of rostrum (excluding the top spine of the rostrum itself) was counted for the two groups of indi- viduals collected from the estuary (Banyu) and the upper basin (Ogura) of the Sagami River. Besides the field-collected individuals, this trait was ex- amined for F, individuals by reciprocal crossings

between and within the two groups (hybrid and control, respectively). Those F, individuals were raised for about 6 months after hatching under nearly identical laboratory conditions. The proce- dures to obtain the hybrids and rear them have been described previously in detail [15].

RESULTS

In electrophoresis, a total of 17 genic loci were routinely assayed (Table 1). Among them, PGM and MPI-I showed polymorphic banding patterns according to the definition of polymorphism that frequency of the most common allele was no greater than 0.95 in at least one locality. HK-/ also appeared to be polymorphic, but its banding pat- tern was not always definitely scorable. All other loci were judged to be monomorphic, and the proportion of polymorphic loci was 0.13 (excluding HK-1).

Phenotypes and interpreted genotypes in PGM and MPI-I are shown in Figure 1, where one or two bands appeared in each individual, showing monomeric protein structure in both loci. The allelic variants were alphabetically designated in the order of faster anodal migration. Relative mobility of each allele in comparison to the most common one (1.00) was: a, 1.84; b, 1.47; c, 1.00 in PGM, and a, 1.09; b, 1.04; c, 1.00; d, 0.95; e, 0.91 in MPI-I. Allelic frequency was not affected by the sex in either locus.

Table 2 shows the observed and expected genotypic frequencies and allelic frequencies of PGM and MPI-1 in individuals collected from the three stations of the Sagami River. All the geno- types of PGM and MPI-/ were at Hardy-Weinberg equilibrium in each station (P>0.05 in all, chi- square test). However, allelic frequencies was significantly different between the small-egg (sta- tion Banyu) and the large-egg (stations Ogura and Tsukui) groups of individuals for the alleles b and c in PGM and the allele ain MPI-I (P<0.05 in all of them, chi-square test). This means that the two groups of individuals, one with large-sized and the other with small-sized eggs, consist of distinct gene pools. Between the two groups of individuals laying large eggs at Ogura ad Tsukui, on the other hand, no difference in allelic frequency was

Splitting Macrobrachium Populations 163

TABLE 1. The enzymes and loci surveyed, and corresponding buffer systems. Numerals suffixed to abbreviated enzyme name represent one of multiple banding zones (loci) assigned in the order of faster anodal migration. The abbreviation of enzyme name is based on Shaklee et al. [25]

Enzyme (E.C.Number) Locus Buffer Phosphogluconate dehydrogenase (1.1.1.44) PGDH TBE* Glycerol-3-phosphate dehydrogenase (1.1.1.8) G3PDH ibid Acid phosphatase (3.1.3.2) ACP ibid Valyl-leucine peptidase (3.4.—.—.) PEPA ibid Esterase (3.1.1.—) EST-3 ibid Hexokinase (2.7.1.1) HK-1 TME** Lactate dehydrogenase (1.1.2.7) LDH ibid Isocitrate dehydrogenase (1.1.1.42) IDHP-1 ibid Aldehyde oxydase (1.2.3.1) AO ibid Phosphoglucomutase (5.4.2.2) PGM CAPM*** Asparate aminotransferase (2.6.1.1) AAT-1 ibid Malate dehydrogenase (1.1.1.37) MDH ibid Malic enzyme (1.1.1.40) MEP ibid Superoxide dismutase (1.15.1.1) SOD-2 ibid Mannose-6-phosphate isomerase (5.3.1.8) MPI-1 MES/TEA**** Glucose-6-phosphate isomerase (5.3.1.9) GPI-1 ibid Glyceraldehyde-3-phosphate dehydrogenase (1.2.1.12) GAPDH ibid

* Tris-borate-EDTA, pH 8.7 [11] ** Tris-maleate-EDTA, pH 7.4 [8] modified by Numachi [21] *** Citrate-N-(3-Aminopropyl) morpholine, pH 6.0 [7] modified by Numachi [21] **** 2-(N-Morpholino)ethanesulfonic acid (MES) 21.32g, Triethanolamine 16.96 g/1000ml D.W. for electrode, and its 1/10 dilution for gel, pH 7.0

PGM

n : 7. | ee ©

O

b/e a/c b/b c/e a/c C/e a/a C/ce bfe a/c c/d

Fic. 1. Banding patterns of Phosphoglucomutase (PGM) and Mannose-6-phosphate isomerase (MPI), with inter- preted genotypes for the loci of PGM and MPI-1. Gels for the reaction of MPI were usually cut off 10 mm toward the anode from the origin.

164

K. MASHIKO AND K. NUMACHI

TABLE 2. Observed vs. expected genotypic frequencies and allelic frequencies at the loci of PGM and MPI-I for small-egg (station Banyu) and the large-egg (stations Ogura and Tsukui) populations of M.

nipponense in the Sagami River.

The expected genotypic frequency is from the Hardy-Weinberg

equilibrium. The total number of specimens in each locality is indicated by the symbol n.

Locus Station Genotype Allelic frequency alc b/b ble cle a b c Banyu Obs. 3 19 22 13 0.03 0.53 0.45 n=57 Exp. 1.3 15.8 26.8 11.4 PGM Ogura Obs. 0 54 16 0 0.0 0.89 0.11 n=70 Exp. 0.0 54.9 14.1 0.9 Tsukui Obs. 0 32 6 1 0.0 0.90 0.10 n=39 Exp. 0.0 31.3 pal 0.4 alc bib ble cle c/ld_ cle a b c d e Banyu Obs. 14 1 8 33 1 1 0.12 0.09 0.78 0.01 0.01 n=58 Exp 10.9 0.5 Si 35% OWL 0.9 MPI-1 Ogura Obs. Z 0 13 55 0 5 0.01 0.09 0.87 0.0 0.03 n=75 Exp 13 06 11.6 56.0 0.0 4.0 Tsukui Obs. 0 6 32 0 1 0.01 0.08 0.90 0.0 0.01 n=40 Exp. 0.7 0.5 5.8 32.4 0.0 0.0 ~ ‘15 ° 5 } 14 e e ° (2) A 13 e ° eee oo Oo @0 oo (oe) fe} ae ee ° = rs) 12 ° AA © © © CO CORR0Oe @ OO ”gRpEse CgoeegCeo 0 ee ACSA CC® «200 ems © 000 0 co) co) ° ie) - II a se © 0c © 0.0080 e”ERHe © eanmangpCe weaned cHO0eCe® 88 000mD0m 0 OO (On ° @ 10 ° e e e ° ° e i) 20 30 40 50 60 70 Body length (mm)

Fic. 2. Rostral tooth counts plotted against body length in individuals collected from station Ogura in the Sagami River. Solid circles, open circles and open triangles indicate males, females and sexually unidentified juveniles,

respectively.

observed in both loci. Genetic identity and dis- tance [17] between the large and the small-egg populations of M. nipponense in the Sagami River were 0.989 and 0.011, respectively, which were calculated from the results on 16 loci (HK-J ex- cluded). The values of average heterozygosity, calculated from actual allelic frequencies, in the large-egg and the small-egg populations were 0.055 and 0.025, respectively.

Prior to analyzing the rostral tooth count, ontogenetic change of this trait was examined using specimens collected from the station of Ogura (Fig. 2). This count fluctuated from 10 to 15 among individuals larger than 20mm in body length, but was almost independent of the body size of individuals. In the regression equation for this relationship, Y=12.1-0.0063X, r=—0.08, where X and Y mean the body length (mm) and

Splitting Macrobrachium Populations 165

60

Frequency (%)

40

60 cs

Fic. 3.

HL NL

HS NS

Frequency distributions of the rostral tooth count in six groups of individuals, NL: natural large-egg

population at station Ogura (mean 11.8+SD 0.8, n=93), NS: natural small-egg population at station Banyu (12.6 +0.9, n=87), HL: F, hybrids between females of the large-egg population and males of the small-egg population (12.4+0.8, n=66, HS: F, hybrids between females of the small-egg population and males of the large-egg population (12.4+0.7, n=105), CL: control of F, individuals by crosses within the large-egg population (11.7+ 0.9, n=45), and CS: control of F, individuals by crosses within the small-egg population (12.7+1.0, n=46). Numbers on the horizontal axes of histogram represent the ranks of the rostral tooth count.

rostral tooth count, respectively, the slope of —0.0063 did not significantly differ from zero (P> 0.3, ANCOVA). Moreover, the tooth count did not significantly differ between the two sexes (P> 0.5, student’s t-test). Thus, comparison of the mean value of this morphological trait among populations is justified even when the composition of individual body size and sex differ among them.

Figure 3 shows the rostral tooth count in the six groups of individuals, NL: natural large-egg population at Ogura; NS, natural small-egg population at Banyu, HL: F, hybrids between females of the large-egg population and males of the small-egg population, HS: F; hybrids between females of the small-egg population and males of the large-egg population, and CL and CS: controls of F, individuals obtained by crosses within the large-egg and the small-egg populations, respec- tively. The histogram for F; individuals in each type of cross was drawn from the results accumu- lated for more than three parental females. All individuals examined here were larger than 20 mm in body length. The mean rostral tooth counts for NL and NS, 11.8 and 12.6 (12 and 13 in mode),

respectively, were significantly different from each other (P< 0.001, student’s t-test). This difference persisted between CL (11.7 in mean and 12 in mode) and CS (12.7 and 13; significant difference in the mean value, P<0.001). On the other hand, the mean tooth count was 12.4 (12 in mode) in both hybrids (HL and HS), which was nearly, though not precisely, intermediate between the values of CL and CS, as expected from quantita- tive inheritance. The differences between HL and CL and between HS and CS were significant (P< 0.001 and P<0.05, respectively). Thus the differ- ent rostral tooth counts in the two groups of field-collected individuals (NL and NS) are consi- dered to have a genetic basis as a quantitative trait, like different egg sizes as mentioned at the begin- ning.

DISCUSSION

Both allozymic and morphological data indicate that the two groups of M. nipponense occurring in the Sagami River are genetically different, and these two groups of individuals appear to be more

166 K. MASHIKO AND K. NUMACHI

or less reproductively isolated within the single river system. The reproductive isolation between them has been also suggested from another fact that, while experimentally crossed F; hybrids lay intermediate-sized eggs between parental popula- tions, no female carrying such intermediate-sized eggs has been found in the field [15]. Since the two groups of individuals easily intercrossed in the laboratory, and since viability and fertility do not decline significantly in F; hybrids [15], any post- mating isolation factor does not seem to work effectively between them. It is likely that, while individuals of this species are apt to move down- stream especially during planktonic larval stage, spatial separation of habitats at the estuary and at the upper basin of river plays an important role of the reproductive barrier between the two groups of M. nipponense in the Sagami River. However, the detailed mechanism of this reproductive isolation remains to be solved in the future.

It is generally recognized that in natural heter- ogenous environments a single species consists of genetically diversified local populations or races, fundamentally due to the depression of free gene flow among them [e.g., 1, 24]. For such local populations, Nei’s genetic distance is calculated to be, very roughly speaking, no greater than 0.05 in many organisms [18]. This holds also in inland- water prawns and shrimps [e.g., 3, 6, 26]. Accord- ing to this criterion, the degree of differentiation between the two groups of M. nipponense in the Sagami River (genetic distance 0.011) is consi- dered to be within the level of local populations. This suggests that, though the two groups of M. nipponense remarkably differ in some life- historical and morphological traits, they have differentiated rather recently. It was presumed previously that the divergence of this species from the presumptive original population with small eggs to those with large and intermediate-sized eggs took place within recent, at most, several thousand years [16]. This consideration comes from the established geological recognition that most of the coastal lagoons in Japan (sea-relict lakes), where the populations with large and in- termediate-sized eggs were specifically found [14], were originally formed by alluvial actions after the most recent (Jomon-period) marine transgression

of the Holocene. The electrophoretic data in the present study well coincide with the previous view.

ACKNOWLEDGMENTS

We are grateful to Drs. K. Tatsukawa and T. Kobayashi and other members of the Ocean Research Institute, University of Tokyo for their kind assistance, and also to Prof. M. Peters of Teikyo University for reading the manuscript and improving the English. This work was supported in part by Grants-in-Aid for Scien- tific Research (Nos. 59740328 and 6074036) to K. M. from the Ministry of Education, Sciences and Culture of Japan, and in part by funds (Cooperative Programs Nos. 89008, 90112 and 9111) provided by Ocean Research Institute, University of Tokyo.

REFERENCES

1 Ayala FJ (1975) Genetic differentiation during the speciation process. Evol Biol 8: 1-78

2 Benzie JHH, Silva PK (1984) The taxonomic rela- tions of the Atyidae (Decapoda, Caridea) of Sri Lanka determined by electrophoretically detectable protein variation. J Crust Biol 4: 632-644

3 Boulton AJ, Knott B (1984) Morphological and electrophoretic studies of the Palaemonidae (Crus- tacea) of the Perth region, West Australia. Aust J Mar Freshw Res 35: 769-783

4 Chow S, Fujio Y (1985) Population genetics of the palaemonid shrimps (Decapoda: Crustacea) I. Genetic variability and differentiation of local populations. Tohoku J Agri Res 36: 93-108

5 Chow S, Fujio Y (1985) Biochemical evidence of two types in the fresh water shrimp Palaemon paucidens inhabiting the same water system. Nippon Suisan Gakkaishi 51: 1451-1460

6 Chow S, Fujio Y, Nomura T (1988) Reproductive isolation and distinct population structures in two types of the freshwater shrimp Palaemon paucidens. Evolution 42: 804-813

7 Clayton JW, Tretiak DN (1972) Amine-citrate buffers for pH control on starch electrophoresis. J Fish Res Board Can 29: 1169-1172

8 Gomori G (1955) Preparation of buffers for use in enzyme studies. In “Methods of Enzymology” Ed by SP Colowick and NO Kaplan, Academic Press, New York, pp 138-146

9 Hedgecock D, Stelmach DJ, Nelson K, Lindenfelser ME, Malecha SR (1979) Genetic divergence and biogeography of natural populations of Macro- brachium rosenbergii. Proc World Maricul Soc 10: 873-879

10 Ikeda M, Kijima A, Fujio Y (1992) Divergence between two species in Paratya compressa (Decapo-

11

12

13

14

15

16

17

18

19

Splitting Macrobrachium Populations

da: Atyidae). Nippon Suisan Gakkaishi 58:819-824 Kraus AP, Neeley CL (1964) Human erythrocyte lactate dehydrogenase: Four genetically determined variants. Science 145: 595-597

Mashiko K (1983) Differences in the egg and clutch sizes of the prawn Macrobrachium nipponense (de Haan) between brackish and fresh waters of a river. Zool Mag 92: 1-9

Mashiko K (1983) Evidence of differentiation be- tween the estuarine and upper freshwater popula- tions inhabiting the same water system in the long- armed prawn Macrobrachium nipponense (de Haan). Zool Mag 92: 180-185

Mashiko K (1990) Diversified egg and clutch sizes among local populations of the freshwater prawn Macrobrachium nipponense (de Haan). J Crust Biol 10: 306-314

Mashiko K (1992) Genetic egg and clutch size variations in freshwater prawn populations. Oikos 63: 454—458

Mashiko K (1992) Laying few large eggs or many small eggs in freshwater prawns of the genus Mac- robrachium. The Heredity (The Iden), Special issue 4: 7-16 (in Japanese)

Nei M (1972) Genetic distance between popula- tions. Am Nat 106: 283-292

Nei M (1990) Molecular Evolutionary Genetics, translated into Japanese by T Gojyobori and T Narita. Baifukan Press, Tokyo, pp 180-217 Numachi K (1974) Genetic characters of popula- tions. In “Biological Study on Marine Resources”

20

21

22

23

24

25

167

Ed by M Nishiwaki, Tokyo University Press, Tokyo, pp 5-36 (in Japanese)

Numachi K (1981) A simple method for preserva- tion and scanning of starch gels. Biochem Genet 19: 233-236

Numachi K (1983) Genetic Analysis on the Growth and Survival of Pacific Abalone. Report of Resear- ches (Subject No. 00548050) by the Grant from the Ministry of Education, Science and Culture of Japan. Ocean Research Institute Tokyo University, pp 1-47 (in Japanese)

Redfield JA, Salini JP (1980) Techniques of starch- gel electrophoresis of penaeid prawn enzymes (Penaeus spp. and Metapenaeus spp.). CSIRO Aust Div Fish Oceanogr Rep 116: 1-20

Richardson BJ, Baverstock PR, Adams M (1986) Allozyme Electrophoresis: A Handbook for Animal Systematics and Population Studies. Academic Press, New York, pp 271-346

Selander R, Smith MH, Yang SY, Johnson WE, Gentry JB (1971) Biochemical polymorphism and systematics in the genus Peromyscus. I. Variation in the old-field mouse (Peromyscus polionotus). Univ Texas Publ 7103: 49-90

Shaklee JB, Allendorf FW, Morizot DC, Whitt GS (1990) Gene nomenclature for protein-coding loci in fish. Trans Am Fish Soc 119: 2-15

Trudeau TN (1978) Electrophoretic protein varia- tion in Macrobrachium ohione and its implications. Proc World Maricul Soc 9: 139-145

a

ravuins oye

parla, 2 4, Phat ene ¥ wah 6 vo) r ' ; i A =ifit We uy os F wie Arey) ie (te ~ f ay ey i he ba) £ bake mh. aia mt Lass fy eile a ahh? mi - ' : ” { F, : \

Morweree? bee ‘hon oy fete tb ilak ona ae >) jovani 1O0. 3 dpe ah ANss lose Jans Nee ty

¢

+ teas Pelee ment Nose :

ipa Gale! A 30 ROlRy Aboa bea Gems A yeti Mads bg me

‘ i Vite Ni vetoed , at ie Ppa “hy ThA. yege

petted CPT afey

wt 7 4p Aa : at Are rule yw) fs i cy mn

bi rey the pore dak

‘ ws é > Las geo SEPy é pte g ‘ls saosin, mink paren y i

1

O77 sb Dok uae Be

went] aya geile ray)

i “ UK 2 eit PP a a | apd ‘ Khe 2h pe, Lea Nak PRR

Alnnolve ne dotuly bee gas eorses ins oid eevhit) — ‘ag

ah winds (Ey Meath an pi Oy hibesunie ee

aWer ; Sele K oF tld Dl apeaee TLR ony Sree ¥ ’ BlOQRe 16 i StyAas ps

its A diel aki Vbulse a Ty. (Dea

4 ah Lies 1 af? he r we Lo - # ish i ¢ 7 _-_ + ‘s ; J : 4 1 ' ye? pam

ZOOLOGICAL SCIENCE 10: 169-174 (1993)

© 1993 Zoological Society of Japan

Mytilidiphila, a New Genus of Nautiliniellid Polychaetes Living in the Mantle Cavity of Deep-sea Mytilid Bivalves Collected from the Okinawa Trough

Tomoyuki Miura! and JuN HasHIMoTo~

‘Faculty of Fisheries, Kagoshima University, 4-50-20, Shimoarata, Kagoshima 890, and "Japan Marine Science and Technology Center, 2-15 Natsushima-cho, Yokosuka 237, Japan

ABSTRACT—Nautiliniellid polychaetes living in the mantle cavity of deep-sea mytilid bivalves collected from the Minami-Ensei Knoll and the Iheya Ridge of the Okinawa Trough, were studied taxonomically. These polychaetes including two different species were unique in the family in having more than 20 simple hooks per parapodium. They were considered to belong to a single new genus, Mytilidiphila. Mytilidiphila enseiensis n. sp., found living in the mantle cavity of undescribed mytilid bivalves close to the genus Adula, has more than 35 simple hooks per parapodium on middle setigers. Myztilidiphila okinawaensis n. sp., found living in the mantle cavity of undescribed deep-sea mussels close to the genus Bathymodiolus, has also numerous simple hooks on each parapodium. However, the distal end of simple hook of M. okinawaensis is rounded and distinguishable from the pointed one of M.

enseiensis.

INTRODUCTION

The deep-sea biological community at the Mina- mi-Ensei Knoll of the Okinawa Trough was firstly studied by Japan Marine Science and Technology Center in 1988 and 1989 in a series of surveys using the deep towed camera system and the deep submergence research vehicle (DSRV) Shinkai 2000 [2]. The community found at the Minami- Ensei Knoll were composed of typical hydrother- mal vent organisms including vestimentiferans, mytilid bivalves, vesicomyid clams and lithodid crabs, some of which have been described as new to science [7, 8]. Many specimens of deep-sea mytilid bivalves also were sampled in our recent dives, and during measurement and dissection of these mussels on the research vessel Natsushima, we collected nautiliniellid worms living in their mantle cavity. Most nautiliniellid worms attached so stiffly to or penetrated so deeply into the gill filaments of the host mussels that we had to dissect the host tissue for collection.

Accepted October 16, 1992 Received August 6, 1992 " To whom all correspondences should be addressed.

The polychaete family Nautiliniellidae has in- cluded four genera and five species from both Pacific [4-6] and Atlantic Oceans [1]. These nautiliniellid polychaetes were found living in the mantle cavity of the deep-sea bivalves of hydro- thermal vents and cold seeps. Concerned with their commensal or parasitic life, the nautiliniellid worms have simple-structured body with less de- veloped prostomium, numerous similar segments, a few kinds of setae and pygidium lacking anal cirri. However the number of prostomial anten- nae, the presence of dorsal cirri and/or projected setae on the second segment, the parapodial struc- ture and the setal composition show remarkable interspecific variability. These morphological characters are thus important for their classifica- tion at both generic and specific levels. In the previously known species, the Atlantic species, Petrecca thyasira Blake is unique in having modi- fied parapodia with elongated notopodia and reduced tentacular segment without dorsal cirri nor setae [1]. The nautiliniellid polychaetes we collected from the Okinawa Trough also have reduced tentacular segment. However they differ from the Atlantic species in the parapodial struc- ture. They also are distinct from the other

170 T. Miura AND J. HASHIMOTO

nautiliniellid species in the setal composition. Consequently, a new genus and two new species are described in this study. The types are depo- sited in the National Science Museum, Tokyo (NSMT) and Japan Marine Science and Technol- ogy Center (JAMSTEC).

DESCRIPTION

Family Nautiliniellidae Miura & Laubier Mytilidiphila, new genus

(Japanese name: Igaiyadorigokai-zoku, new) Type species.—Mytilidiphila enseiensis, new spe- cies. Gender.—Feminine. Diagnosis. — Body long, vermiform, tapering post- eriorly with numerous setigers, flattened ventrally and arched dorsally in cross-section. Prostomium short with a pair of antennae, without eyes. Second or tentacular segment well fused with prostomium, having ventral cirri and embedded neuroacicula, with or without setae. Parapodia subbiramous with dorsal and ventral cirri; notopo- dium supported by a single stout notoaciculum, without setae; neuropodia supported by a single stout neuroaciculum, with numerous simple, hooked setae. Pygidium cylindrical without appendages. Remarks.—The family Nautiliniellidae is unique in the class Polychaeta, in the combination of one or tow pairs of prostomial antennae, reduced tentacular segment, lack of palps, subbiramous parapodia and simple neuropodial hooks as well as the association with deep-sea bivalves [1, 4-6]. The genus Mytilidiphila also is referred to the

family, in having these morphological characters and ecological feature.

In the family, the genus Natsushima Miura &

Laubier has additional simple bifurcate setae and differs from four other genera having only simple neuropodial hooks [5]. The four genera divided into two groups by the thickness of the notoacicu- la. The genera Nautiliniella Miura & Laubier and Mytilidiphila have the notoacicula as thick as the neuroacicula [4], while Shinkai Miura & Labuier and Petrecca Blake have slender notoacicula [1, 5]. The genus Mytilidiphila has only a pair of prosto- mial antennae and differs from Nautiliniella with two pairs. The genus Mytilidiphila also differs from all other genera of the family in having more than 20 simple hooks on a single parapodium at least in some setigers, instead of up to 10. In other point of view, each nautiliniellid genus is tend to be associated with a certain group of bivalves: Shinkai and Nautiliniella with Calyptogena spp., Petrecca with Thyasira sp., Natsushima with a solemyid bivalve, Mytilidiphila with mytilid bivalves. Etymology.—The genus name is made by the combination of the family name of the host bivalves, Mytilidae, and a Latinized Greek suffix, -phila.

Mytilidiphila enseiensis, new species (Japanese name: Ensei-igaiyadori, new) (Fig. 1a-g)

Material.— Minami-Ensei Knoll, Okinawa Trough, DSRV Shinkai 2000 Dive 616, 1 June 1992, 28° 23.5°N, 127° 38.5'E, 625 m, associated with unde- scribed deep-sea mytilid bivalves close to the genus

TABLE 1. Measurements of complete specimens of Mytilidiphila enseiensis n. sp.

Body Length Body Width

Number of

No. (mm) mm) Segments Notes

2-1 20.8 1.3 121 Holotype

2-2 11.4 1.2 80 caudal end regenerated

2-3 14.1 1.2 81

2-4 13.5 0.6 79

2-5 27.8 ils) 164 Largest paratype (ovigerous) 5-1 8.5 1.2 54 caudal end regenerated

5-2 12.8 12 77

Nautiliniellids from the Okinawa Trough 171

Fic. 1.

Mytilidiphila enseiensis n. sp. (Holotype): a, Anterior end, dorsal view; b, Posterior end, dorsal view; c,

Anterior end enlarged, dorsal view; d, Same, ventral view; e, Parapodium of the 11th segment (or parapodium 10), anterior view; f, Parapodium of the 27th segment (or parapodium 26), anterior view; g, Hooks.

Adula, holotype (NSMT Pol. H-351) and 36 para- types (JAMSTEC).

Description.—Holotype 20.8mm long, 1.3mm wide including parapodia, with 121 setigers. Largest paratype ovigerous, 27.8 mm long, 1.5 mm wide, with 164 setigers (measurement of all com- plete specimens shown in Table 1). Body flattened ventrally and arched dorsally. Integument

smooth. Preserved specimen pale.

Prostomium short, with a pair of short cirriform antennae, without eyes (Fig. la, c, d). Second segment well fused with prostomium, defined by embedded acicula and ventral cirri as well as projected simple hooks (Fig. 1d). Mouth opening situated between prostomium and first fully de- veloped setigerous segment. Foregut with well-

172 T. Miura AND J. HASHIMOTO

developed muscular part (Fig. 1a).

Pygidium simple, without anal cirri (Fig. 1b).

Parapodium subbiramous throughout body; notopodium supported by a single stout notoacicu- lum; neuropodium well-developed, supported by a single stout neuroaciculum (Fig. le, f); dorsal cir- rus short, conical, situated distally on notopodium, sometimes distal tip of notoacicula penetrating dorsal cirrus; ventral cirrus very short, conical, situated ventro-posteriorly on neuropodium.

Setae consisting of simple hooks only; number of hooks on each parapodium varying 6-10 on first 3 setigers, reaching more than 35 on parapodium 26. Hooks simple, slender, with inflated subdistal stems and slightly curved, pointed distal ends (Fig.

1g).

Fic. 2.

Remarks.—Mytilidiphila enseiensis differs from most of nautiliniellid species in the absence of dorsal cirri on the second segment and the pre- sence of equally stout noto- and neuroacicula. Petrecca thyasira Blake also has less developed parapodia without dorsal cirri on the second seg- ment [1]. However this Atlantic species differs from M. enseiensis in having branchia-like long notopodia and occurring of only one or two simple hooks on each parapodium, instead of short conic- al notopodia and more than 30 simple hooks on a parapodium respectively. Shinkai longipedata

Miura & Ohta has up to 10 hooks on some anterior parapodia [6] and it has been the maximal number of hooks occurring on a parapodium in the pre- viously known nautiliniellid species.

a,d: 1.0mm

| nEnEEEEEEEEEEeemeeenemeneeseeeeee

b,c: 0.3mm

e,f: 0.3mm

SeeeEnOEEEEEEEeeemeeeeeemanemmemmnt

g g: 0.02mm

Mytilidiphila okinawaensis n. sp. (Holotype): a, Anterior end, dorsal view; b, Same, enlarged, dorsal view; c,

Same, ventral view; d, Posterior end, dorsal view; e, Parapodium of the 11th segment (or parapodium 10), anterior view; f, Parapodium of the 36th segment (or parapodium 35), anterior view; g, Hooks.

Nautiliniellids from the Okinawa Trough 173

Most examined specimens had one or two sim- ple hooks on a single parapodium of the second segment (average number: 0.96 for 37 specimens), though six of them did not have any.

The specimens of Mytilidiphila enseiensis were found to harbor all four dissected bivalves from the Minami-Ensei Knoll. The bivalves are ranging from 163 to 175mm in length. Two of them contained only one nautiliniellid polychaete, while the other two had 16 and 21 individuals. Other than these nautiliniellid polychaetes, two speci- mens of polynoid polychaetes tentatively identified as Branchipolynoe pettiboneae Miura and Hashi- moto [3] also were collected from two specimens of the same mytilid bivalve.

Etymology.—The species name is derived from the Minami-Ensei Knoll, the type-locality.

Mytilidiphila okinawaensis, new species (Japanese name: Okinawa-igaiyadori, new) (Fig. 2a-g) Material — Minami-Ensei Knoll, Okinawa Trough, DSRV Shinkai 2000 Dive 549, 5 June 1991, 28° 23.5'N, 127° 38.5’E, 701 m, associated with unde- scribed deep-sea mussel close to the genus Bath- ymodiolus, holotype (NSMT Pol. H-352) and 1 paratype (JAMSTEC); Iheya Ridge, Okinawa Trough, Dive 614, 30 May, 1992, 27° 33.0’N, 126° 58.0°E, 1395 m, associated with undescribed deep-

sea mussel, 1 paratype (JAMSTEC). Description.—Holotype 15.2mm long, 1.4mm wide including parapodia, with 95 setigers. Para- type of Dive 549, 9.4 mm long, 1.2 mm wide, with 84 setigers. Body flattened ventrally and strongly arched dorsally. Integument smooth. Preserved specimen pale.

Prostomium short, ventrally with a pair of very short cirriform antennae, without eyes (Fig. 2a-c). Second or tentacular segment well fused with prostomium, achaetous, defined by embedded aci- cula and ventral cirri (Fig. 2c). Ventral cirri of tentacular segment inserted in front of neuro- podium. Mouth opening situated between prosto- mium and first fully developed setigerous segment. Foregut with well-developed muscular part (Fig. 2a).

Pygidium rounded, without anal cirri (Fig. 2d).

Parapodium subbiramous throughout body, with inflated notopodium supported by a single stout notoaciculum, and smaller neuropodium sup- ported by a single stout neuroaciculum (Fig. 2e, f); dorsal cirrus short, conical, situated distally on notopodium; ventral cirrus short conical, situated ventro-posteriorly on neuropodium.

Setae consisting of simple hooks only; number of hooks on each parapodium varying 4-7 on first 3 setigers, thereafter immediately increasing and reaching to 20 on parapodium 10. Hooks simple, slender; distal end slightly curved, with rounded tips (Fig. 2g).

Remarks.— Mytilidiphila okinawaensis differs from M. enseiensis in the shape of simple hooks and in the presence of setae on the second seg- ment. The simple hooks of M. okinawaensis are gradually tapered with rounded distal ends, while those of M. enseiensis have inflated subdistal stems and pointed distal ends. The simple hooks are present on the second segment in most specimens of M. enseiensis, but absent in M. okinawaensis. Etymology.—The species name is derived from the Okinawa Trough, the type-locality.

ACKNOWLEDGMENTS

The authors wish to thank Dr. Yoshiyuki Nozaki of Ocean Research Institute, University of Tokyo for his kind offer of a part of the material examined in this study. Thanks are also due to the staff of JAMSTEC for their assistance at the sampling by means of the deep submergence research vehicle Shinkai 2000. This work was supported in part by grants-in-aid for the first author from the Ministry of Education, Culture and Science, Japan (No. 03640630) and Kato Foundation for the Promotion of the Bioscience Researches.

REFERENCES

1 Blake JC (1990) A new genus and species of Polychaeta commensal with a deep-sea thyasirid clam. Proc Biol Soc Washington 103: 681-686

2 Hashimoto J, Fujikura K, Hotta H (1990) Observa- tions of deep sea biological communities at Minami- Ensei Knoll. JAMSTECTR Deepsea Res 167-179 (In Japanese with English abstract)

Miura T, Hashimoto J (1991) Two new branchiate scale-worms (Polynoidae: Polychaeta) from the hydro- thermal vent of the Okinawa Trough and the volcanic seamount off Chichijima Island. Proc Biol Soc

174 T. Mrura AND J. HAasHImoTo

Washington 104: 166-174

Miura T, Laubier L (1989) Nautilina calyptogenicola, a new genus and species of parasitic polychaete on a vesicomyid bivalve from the Japan Trench, repre- sentative of a new family Nautilinidae. Zool Sci 6: 387-390

Miura T, Laubier L (1990) Nautiliniellid polychaetes collected from the Hatsushima cold-seep site in Saga- mi Bay, with descriptions of new genera and species. Zool Sci 7: 319-325

6 Miura T, Ohta S (1991) Two polychaete species from

the deep-sea hydrothermal vents in the middle Oki- nawa Trough. Zool Sci 8: 383-387

Okutani T, Fujikura K (1990) A new turbinid gastro- pod collected from the warm seep site in Minami- Ensei Knoll, West of the Amami-Oshima Island, Japan. Venus (Jap J Malacology) 49: 89-90

Takeda M, Hashimoto J (1990) A new species of the genus Paralomis (Crustacea, Decapoda, Lithodidae) from Minami-Ensei Knoll in the mid-Okinawa Trough. Bull Natn Sci Mus Tokyo, ser A, 12: 79-88

ZOOLOGICAL SCIENCE 10: 175-178 (1993)

[RAPID COMMUNICATION]

© 1993 Zoological Society of Japan

Differential Expression of Xenopus BMPs in Early Embryos and Tissues

ATSUSHI SUZUKI, SHIN-ICHIRO NISHIMATSU, KAZUO MURAKAMI

and Naoto UENO

Institute of Applied Biochemistry, University of Tsukuba, Ibaraki 305, Japan

ABSTRACT— Expression levels of mRNA for Xenopus bone morphogenetic proteins (BMPs) in early embryos and adult organs were examined. Reverse transcription- polymerase chain reaction (RT-PCR) that used specific primers for each BMP subtype revealed that mRNAs for BMP-2 and BMP-4 are distributed in a wide variety of adult tissues including lung, heart, and kideney. Unex- pectedly, distribution of mRNA for BMP-7 was found to be limited to ovary and early embryos. The result suggests that Xenopus BMP-7 may have a specific role in ovary and early embryos.

INTRODUCTION

A growing body of evidence suggest that peptide growth factors play essential roles in animal development [4]. We have previously isolated genes encoding proteins related to activin, a member of transforming growth factor-@ (TGF-/) family proteins from Xenopus laevis [12]. Among several activin related genes, one gene was found to code an extremely similar protein to mammalian bone morphogenetic protein-2 (BMP-2) [14], which was originally discovered for its ability to induce bone and cartilage formation in vivo. Subsequently, we have cloned cDNAs for BMP-2, -4 and -7 from Xenopus oocyte cDNA library using BMP-2 gene as a probe [8]. DNA sequence analysis has shown that their primary structures are highly conserved between amphibian and mammalian.

Accepted August 31, 1992 Received July 17, 1992

In the present study, ontogeny and tissue dis- tribution of these Xenopus laevis BMPs were examined by reverse transcription PCR [6] that is a rather sensitive method to detect mRNAs from minute embryonic tissues.

MATERIALS AND METHODS

Xenopus embryos and tissues

Xenopus laevis embryos were obtained by arti- ficial fertilization. The embryos were staged according to Nieuwkoop and Faber [7]. Adult organs are surgically removed from anesthetized male and female Xenopus laevis. All embryos and tissues were immediately frozen in liquid nitrogen and kept at —80°C until use.

Reverse transcription PCR

Total RNA was purified from embryos and adult tissues by acid guanidinium thiocyanate-phenol- chloroform extraction (AGPC) method [2] and guanidinium-CsCl method [1], respectively. RNAs from adult tissues were from mixture of several individuals. cDNA was synthesized from total RNA as follows. A 20 sl of reverse transcrip- tion mixture containing 500 ng of total RNA, 1x RT buffer (S50 mM Tris-HCl, pH 8.3/75 mM KCl/ 3 mM MgCl), 100 pmol of random hexamer prim- ers, 0.5 mM of each dNTP, 10 mM DTT, 20 U of RNasin (Promega), and 100 U of MMLV reverse transcriptase (BRL) was incubated at 37°C for 60

176 A. SUZUKI AND S. NIsHMaTsuU ef al.

min, and heated to 95°C for 10min. PCR was performed at a final concentration of 1xPCR buffer (10 mM Tris-HCl, pH 8.8/50 mM KCI/1.5 mM MgCl,/0.1% Triton X-100)/0.2 mM each of dNTP/5 wCi of [a-**P]dCTP (3000 Ci/mmol)/5 pmol of each primer/0.225U of Taq DNA polymerase (Wako Chemical, Japan) in a total volume of 10 ul. The mixture was overlaid with mineral oil and then amplified with a thermal cycler (Perkin-Elemer/Cetus). The amplification profile involved denaturation at 95°C for 30sec, primer annealing at 58°C for 30 sec, and extension at 72°C for 1 min. A 4 yl of the PCR sample was electrophoresed on 5% polyacrylamide gel and the PCR fragment was visualized by autoradiography. Appropriate bands were cut out from the dried gel and radioactivity was determined by Cerenkov counting. For quantitative analysis, optimal cycle number for each BMP was determined. Cycle of 22 was chosen for all BMPs because it is in the range of exponential increase of PCR products depending on cycle numbers and resulting radioac- tivities of PCR products are relative to RNA amounts under the amplification condition. Prim- ers used in PCR were as follows: BMP-2 (5- CAATGGTCGCTGGGATCCAC-3’ and 5’-TC- ACAACATTTTTGCCAGGCGT-3’), BMP-4 (5’-CATCATGATTCCTGGTAACCGA-3’ and 5’-CTCCATGCTGATATCGTGCAG-3’), BMP- 7 (5'-AAGGCCACGAACGCAGA-3' and 5- TGTCAGTTCATCTCCAAGT-3’), EF-la (5’- CCTGAATCACCCAGGCCAGATTIGGTG - 3’ and 5'--GAGGGTAGTCTGAGAAGCTCTCCA- CG-3’).

RESULTS AND DISCUSSION

Newly established RT-PCR method enabled us to detect mRNAs in small preparations of embryos and adult tissues compared to conventional methods such as Northern blot analysis and RNase protection assay. In addition, the method has minimized the time required for the assay. Pre- sence of BMP mRNAs in developing embryos were demonstrated by this method using respec- tive specific primers. Figure l(a) schematically represents the precursor structures of BMPs and the regions chosen for designing oligonucleotide

o.% ROA DD OQOVON O

BMP-2 wememOwewew ~~ ws BMP-4 CY BMP-7 ~~ ~eoe

0 0 2 4 6 8 10 12 14 16 18 20 (Hour)

1 5 8 9 10 12) 4519 (Stage)

Fic. 1. Levels of BMP mRNAs during early Xenopus embryogenesis. (a) Primer positions used for RT- PCR. Each primer set was specifically designed to detect each BMP mRNAs and denoted by arrows. The propeptide structures that include mature re- gion (shaded bars) derived from the full-length cDNAs are represented by bars. (b) Total RNA from staged embryos was subjected to RT-PCR using specific primers for each BMP mRNA. Xeno- pus embryos‘ were cultured in 0.1X Steinberg’s solution at 23°C. Developmental stages are shown to the top of the figure. (c) Schematic representa- tion of (b). Radioactivity of appropriate band was measured by Cerenkov counting and plotted against hours after fertilization (BMP-2, squares; BMP-4, circles, BMP-7, triangles).

primers. Specific nucleotide sequences were chosen for each BMP to prevent the primers from cross-hybridization. Figure 1(b) shows the accu-

Xenopus BMPs in Embryos and Tissues 177

mulation of mRNA for Xenopus BMP-2, -4 and -7 during early embryogenesis. The result shows that all BMP mRNAs are maternally present and that the levels of the BMP mRNAs change drastically during development (BMP-4 transcript in early embryos before stage 9 becomes detectable by longer exposure). It should be noted that genes for BMPs are differentially regulated in early develop- ment although their primary structures are closely related to each other. Especially, homology of amino acid sequence of Xenopus BMP-2 and BMP-4 are 82% in their predicted mature pro- teins. Independent gene regulation of BMP genes in early embryos implies that they have distinct functions in embryogenesis. The result is very similar to that obtained by Northern blot analysis of poly(A)+ RNA purified from staged early embryos. This method, however, provided a sharper image of accumulation and decline of the mRNAs by plotting radioactivities of the PCR products (Figure 1(c)). Existence of BMP-2 [13] and BMP-4 [9] in embryo was also confirmed by Western blot analysis that used specific antibodies against their respective sequences.

Tissue distribution of BMP mRNAs was also examined. RNA was purified from adult tissues

BMP-2

BMP-4_ ..

BMP-7

EF-1la

and subjected to RT-PCR assay for BMP mRNAs. As shown in Figure 2, distributions of mRNAs for BMP-2 and -4 are very similar to each other and the transcripts were detected in a wide variety of tissues of not only mesoderm-origin but also ectoderm-orgin. The result is in a good agreement with a previous observation on mammalian BMP-2 and -4 [15]. Interestingly, however, mRNA for Xenopus BMP-7 was detected only in ovary. In addition to the presence of MBP-7 mRNA in early embryo before neurula, this observation strongly suggests that BMP-7 has embryo-specific role(s) in Xenopus which is as yet unknown. It remains to be examined whether localization of mammalian counter part of Xenopus BMP-7 transcript is also limited to ovary and early embryos. It is known that a heterodimer of BMP-2 and -7 (OP-1) subunits is present in mammalian bone extracts [11]. Therefore, BMP-7 may modulate specificity and potency of BMP activities in ovary and early embryos by forming a heterodimer with another BMP peptide.

BMPs are structurally related to gene product of Drosophila dppc [10] which is a locus responsible for dorso-ventral specification and formation of imaginal disks in Drosophila embryo. Moreover,

Seu wHeouvuns &

Fic. 2. Distribution of BMP mRNAs in adult Xenopus tissues. Total RNA (500 ng) from adult Xenopus tissues was subjected to RT-PCR using specific primers for each BMP mRNA and EF-lamRNA. EF-la was used to confirm

both RNA quality and quantity.

RT-PCR for EF-1a mRNA was performed at 19 cycles.

178 A. SUZUKI AND S. NISHMATSU et al.

it has been proposed that Xenopus BMP-4 may function as a posterior-ventralizing factor in Xeno- pus [3, 5]. The presence of BMP-7 mRNA in “organizer field”, dorsal lip of early gastrula (Dr. K. Cho, personal communication) suggests that Xenopus BMP-7 whose structure is similar to BMP-4 may also be an essential factor in pattern formation of amphibian development.

ACKNOWLEDGEMENTS

We wesh to thank Dr. R. Makino for technical advice on RT-RCR. This study was supported in part by Grant-in-Aid for Scientific Research (No. 02558017) from Ministry of Education, Science and Culture, Japan and grants from Chichibu Cement Co. and Kowa Foundation for Promotion of Life Sciences.

REFERENCES

1 Chirgwin JM, Przybyla AE, MacDonald RJ, Rutter WJ (1979) Biochemistry 18: 5294-5299

2 Chomcezynski P, Sacchi N (1987) Anal Biochem 162: 156-159

3 Dale L, Howes G, Price BMJ, Smith JC (1992) Development 115: 573-585

4 Jessell TM, Melton DA (1992) Cell 68: 257-270

5 Jones CM, Lyons KM, LaPan PM, Wright CVE,

10

12

13

14

15

Hogan BLM (1992) Development 115: 639-647 Makino R Sekiya T, Hayashi K(1990) Technique 2: 295-301

Nieuwkoop PD, Faber J (1967) In “Normal Table of Xenopus laevis (Daudin) 2nd”. North-Holland, Am- sterdam

Nishimatsu S, Suzuki A, Shoda A, Murakami, Ueno N (1992) Biochem Biophys Res Commun 186: 1487-1495

Nishimatsu S, Takebayashi K, Suzuki A, Murakami K, Ueno N (1992) Growth Factors, in press Padgett RW, St Johnston RD, Gelbart WM (1987) Nature 325: 81-84

Sampath TK, Coughlin JE, Whetstone RM, Banach D, Corbett C, Ridge RJ, Ozkaynak E, Oppermann H, Rueger DC (1990) J Biol Chem 265: 13198- 13205

Ueno N, Asashima M, Nishimatsu S, Suzuki A, Murakami K (1991) In “Frontiers in muscle re- search”. Ed. by E Ozawa, T Masaki, Y Nabeshima Elsevier, Amsterdam, pp 17-28

Ueno N, Shoda A, Takebayashi K, Suzuki A, Nishimatsu S, Kikuchi T, Wakimasu M, Fujino M, Murakami K (1992) Growth Factors 7: 233-240 Wozney JM, Rosen V, Celeste AJ, Mitsock LM, Whitters MJ, Kriz RW, Hewick RM, Wang EA (1988) Science 242: 1528-1534

Wozney JM, Rosen V, Byrne M, Celeste AJ, Moutsatsos I, Wang EA (1990) J Cell Sci Suppl 13: 149-156

ZOOLOGICAL SCIENCE 10: 179-182 (1993)

[RAPID COMMUNICATION]

© 1993 Zoological Society of Japan

Sperm Behavior in the Micropyle of the Medaka Egg

Sumio IsHwIMA, YUKIHISA HAMAGuUCHI and TaKAsHi Iwamatsu!

Biological Laboratory, Faculty of Science, Tokyo Institute of Technology, O-okayama, Meguro-ku, Tokyo 152 and ‘Department of Biology, Aichi University of Education, Kariya 448, Japan

ABSTRACT— Sperm behavior in the micropyle of the medaka egg was examined using videomicroscopy. Sper- matozoa swimming near the egg, whose progressive speed was similar to that of spermatozoa freely swim- ming in a medium, entered the micropylar canal after being trapped in the funnel-shaped micropylar vestibule or entered directly. The spermatozoa stalled in the micropylar vestibule entered the micropylar canal with a time lag of less than 0.5 sec. Proceeding at a fairly high speed for a short time, the spermatozoa gradually slowed down when the flagellar movement of the spermatozoa became obstructed by the surface of the micropylar canal, finally reaching the plasma membrane. The spermatozoa which directly entered the micropylar canal maintained progressive speed until their flagellar move- ment became obstructed by the surface of the micropylar canal. They reached the plasma membrane through the micropylar canal in the same manner as the spermatozoa experiencing collision with the surface of the micropylar vestibule.

INTRODUCTION

In teleost fish, spermatozoa can reach the egg’s plasma membrane only by means of a narrow channel, the micropyle of the vitelline envelope. Since the base of the micropyle is as wide as the diameter of the sperm heads, spermatozoa must pass in single file through the micropyle. This has been thought to be the mechanism for blocking polyspermy [2] because medaka eggs do not show a rapid electrical block to polyspermy [6].

Since observation of sperm entry into the micropyle is difficult, little is known about the

Accepted November 20, 1992 Received October 19, 1992

exact behavior of spermatozoa in the micropyle. Recent videomicroscopy and the transparency of the medaka egg enabled us to record on videotape the movement of the sperm head in the micropyle of the medaka eggs.

The purpose of the present study is to record and analyze the behavior of spermatozoa in the micropyle of the medaka egg. On the basis of this data, the mechanism for blocking polyspermy in the medaka egg is discussed.

MATERIALS AND METHODS

Medaka oocytes and spermatozoa were pre- pared by the method described earlier [4] and were kept in Ringer’s solution (6.5 g NaCl, 0.4g KCl, 0.113 g CaCl, and 0.15 g MgSO,-7H2O per liter of deionized water, pH=7.3 adjusted with NaHCOs;) before use. In each experiment, the gametes were freshly prepared and used within 30min. The observation chamber was constructed by gluing two strips of silicon rubber (1 mm thick) onto a glass slide parallel to each other. An egg was put in the chamber with a small amount of Ringer’s solution, and the chamber was covered with a coverslip after removal of the Ringer’s solution around the egg. To observe an oblique view of the egg micropyle, the egg was rotated by sliding the coverslip. Sperm movement was observed using a Nikon Optiphot microscope equipped with a plan 40x objective, 10 eyepieces and a National video camera (WV-1300A, Matsushita Com- munication Industrial Co., Ltd, Yokohama, Japan). While observing the egg displayed on a

180 S. IsHuima, Y. HAMAGUCHI AND T. IwAMATSU

National TN-96 monitor (Matsushita Communica- tion Industrial Co., Ltd), approximately 0.1 ml of sperm suspension was introduced into the observa- tion chamber from an open end. The image was recorded on video tape with a Panasonic video tape recorder (NV-FS900, Matsushita Com- munication Industrial Co., Ltd).

For detailed field-by-field analysis, images of the head were traced from the video monitor onto transparent plastic sheets using a fine-point mar- ker. The progressive speed of free-swimming spermatozoa was determined by measuring the distance between the positions of the sperm head at the beginning and the end of a time interval.

RESULTS AND DISCUSSION

Movement characteristics of medaka spermatozoa

Medaka spermatozoa swam at a maximum trans- verse displacement of the sperm flagella of 5.14+ 1.73 um (mean+S.D., n=16) while rotating ab- out their longitudinal axis at a rate of 3.77 per second as previously reported by Ishijima et al. [3]. The progressive speed of medaka spermatozoa swimming near the micropyle was 74.0+14.6 um/ sec (mean+S.D., n=19), whereas that of sperma- tozoa freely swimming in the medium was 74.9+ 31.3 wm/sec (mean+S.D., n=25). No difference was found between them, suggesting that sperm movement, especially progressive speed, is not stimulated by the presence of the micropyle.

The progressive speed of medaka spermatozoa was low as compared to that (191.4 ~m/sec) of sea urchin spermatozoa [1], whose flagellum is similar in size to that of medaka spermatozoa. The progressive speed is roughly proportional to the square of the maximum transverse displacement of the flagellum. The ratio of the progressive speed to the square of the maximum transverse displace- ment of the flagellum does not differ between medaka (2.8; maximum transverse displacement of 5.14 um) and sea urchin spermatozoa (3.0; max- imum transverse displacement of 8.0 ~m) [1]. This suggests that the low progressive speed of medaka spermatozoa is mainly due to the narrow ampli- tude of their flagellar waves. The small transverse displacement of the beating flagella of medaka

spermatozoa may be critical for sperm progress in the small diameter of the micropyle (see below).

Progressive speed of medaka spermatozoa in the micropyle

Most spermatozoa attached to the surface of the micropylar vestibule, stayed there for approx- imately 424 msec (S.D.=243 msec, n=20) and then entered the micropylar canal, although some spermatozoa entered directly. Both the progres- sive speed of the medaka spermatozoa and the size of the micropyle of the medaka egg varied; the micropyle had a mean diameter (d,) of 5.30 wm (S.D.=1.25 wm, n=21) at the end of its vestibule and a mean diameter (d,) of 3.19 um (S.D.=0.84 ym, n=21) at the base of the micropylar canal (Fig. 1). Thus, the profile of the progressive speed of medaka spermatozoa in the micropyle varied. In general, three phases in the time course of the progressive speed were recognized (Fig. 2). The first phase was a decrease in the progressive speed of the spermatozoa from the free-swimming speed

Fic. 1.

A phase-contrast videomicrograph of the mic- ropyle of the medaka egg. d,, diameter of the micropylar vestibule at its base; d., diameter of the micropylar canal at its end. Two spermatozoa line up along the micropylar canal (arrowheads). Bar, 20 ~m.

Spermatozoa in Micropyle of Medaka Egg 181

799

wa

Time (sec) Fic. 2. A typical profile of the time course of the progressive speed of medaka spermatozoa in the micropyle.

Progressive speed (m/sec )

100 3) oa 2 80 a rer \ e) E @ 60 mo} @ @ a ° o 40 2 u ° Ooo, © hoo} oS 20 Qa

to) 3 4 5 6 7 8 9 10

Inner diameter of the micropyle ( um )

Fic.3. The relationship between the progressive speed and the diameter of the micropyle. Two examples are shown.

182 S. IsHuima, Y. HAMAGUCHI AND T. IWAMATSU

to nearly zero; the spermatozoa which swam freely were trapped in the micropylar vestibule and wandered through it or ceased any progress for less than 0.5sec as mentioned above. The second phase was the recovery of the progressive speed; the spermatozoa entered the micropylar canal and proceeded without obstruction from the micropy- lar canal. The final phase was the decrease of the progressive speed until it was reduced to zero; the spermatozoa made their way through the micropy- le because the flagellar movement of the spermato- zoa was mechanically inhibited by the surface of the micropylar canal, and finally reached the egg’s plasma membrane. The inhibitory effect of the micropyle on the progressive speed is shown in Fig. 3. Progressive speed rapidly decreased with a decrease in the diameter of the micropyle. It took 689 +428 msec (n=16) for the spermatozoa to pass through the micropylar canal (the second and third phases).

No difference was found in the time course of the spermatozoa secondarily entering the micropy- le compared to that of the first (data not shown).

Most spermatozoa were trapped in the micropy- lar vestibule before entering the micropylar canal, suggesting that the micropylar vestibule is very useful in capturing the spermatozoa.

In sperm-egg interaction, there are various mechanisms blocking polyspermy [2]. In medaka, the micropyle plays an important role in blocking polyspermy because only one spermatozoon reached the plasma membrane in our study. The blocking of the polyspermy with the micropyle is probably due to the mechanical inhibition of sperm progression of the surface of the micropyle. The

mean diameter of the micropyle at its end (3.19+ 0.84 um) is almost the same as that of the head of spermatozoa (3.71+0.60 ~m, n=17), so that sperm must pass in single file through the micropy- le. The decrease in progressive speed (Figs. 2 and 3), especially slow progression of further sperm, may contribute to the effective polyspermy block. There was no difference in sperm behavior in the micropyle between the spermatozoa entering im- mediately or later. Many spermatozoa could pass through the micropyle which had been isolated from the medaka egg (data not shown). These observations suggest that substances inhibit sperm behavior are not secreted by the plasma membrane and the micropyle.

Most medaka spermatozoa swimming freely in the medium rolled in a clockwise direction on their longitudinal axis when an observer viewed the cell from the anterior ends although some spermatozoa rolled in the other direction [3]. In contrast, the sense of the micropyle is counterclockwise [5]. Therefore, it is unlikely that the chirality of the micropyle is involved in sperm movement when passing through the micropyle.

REFERENCES

1 Gray J, Hancock GJ (1955) J Exp Biol 32: 802-814

2 Hart NH (1990) Int Rev Cytol 121: 1-66

3 Ishijima S, Hamaguchi MS, Naruse M, Ishijima SA, Hamaguchi Y (1992) J Exp Biol 163: 15-31

4 Iwamatsu T (1975) Bull Aichi Univ Educ (Nat Sci) 24: 113-144 (in Japanese)

5 Iwamatsu T, Onitake K, Yoshimoto Y, Hiramoto Y (1991) Develop Growth Differ 33: 479-490

6 Nuccitelli R (1980) Develop Biol 76: 499-504

ZOOLOGICAL SCIENCE 10: 183-187 (1993)

[RAPID COMMUNICATION]

© 1993 Zoological Society of Japan

Epidermal Cells of the Tail of an Anuran Larva Are Competent to Transform into the Adult-Type Cells

KATSUTOSHIYOSHIZATO!, Akio NISHIKAWA’, Yumi Izutsu!

and Masayosu! KAIHo?

"Molecular Cell Science Laboratory, Zoological Institute, Faculty of Science, Hiroshima University, Kagamiyama 1-3-1, Higashihiroshima-shi, Hiroshima 724, "Department of Physiology, Saitama Medical School, Morohongo 38, Moroyama, Iruma-gun, Saitama 350-04, and *Safety Research Laboratories, Yamanouchi Pharmaceutical Co., Ltd.,

Azusawa 1-1-8, Itabashi-ku, Tokyo 174, Japan

ABSTRACT—The epidermis of an anuran tadpole shows the region-specific metamorphic fate: the body epidermis transforms into the adult epidermis, while the tail epidermis commits apoptosis. The only explanation for this difference has been that the tail lacks basal cells which are competent to transform into germinative cells of the adult epidermis. Human blood group antigen A was found to be a specific molecular marker for adult-type epidermal cells. A sharp threshold was apparent at the body-tail junction in expression of the antigen. Utilizing this marker, we succeeded in demon- strating the presence of a specific population of epidermal cells in the tail which can differentiate into the adult-type germinative cells under the direct action of thyroid hormone. This was an unexpected result in that the tail makes provision for the adult life. We propose a hypothesis on the regulation of region-specific metamor- phic changes of the anuran larval epidermis.

INTRODUCTION

The metamorphic fate of the epidermis of an anuran tadpole depends on its location in the head-tail axis. Changes in the epidermis of the tail contrast with those in the epidermis of the body (head and body trunk region): the former is to be lost during metamorphosis, while the latter is to survive and transform into the adult type [10]. The

Accepted December 16, 1992 Received November 27, 1992 " To whom correspondence should be addressed.

present study was carried out to understand the reason for this different metamorphic fate of the two tissues at the cellular level.

The epidermis of a tadpole of bullfrog, Rana catesbeiana, is composed of three types of cells: apical, skein and basal cells [8, 10]. Apical and skein cells are larva-specific in that they terminate their life at the metamorphosis. Basal cells are the larva to adult cell in that they transform their characters from the larval type into the adult one. They are most likely the progenitor of germinative cells of the adult skin [8].

The only explanation given hitherto for the region-dependent metamorphic change in the epidermis described above has been the lack of basal cells in the tail and their presence in the body [8]. To verify if this explanation is correct, we looked for a specific probe that detects the adult-type epidermal cells, but not larval cells. We found that the human group A antigen is expressed specifically in the suprabasal cells of postmetamor- phic skin and is qualified for the probe. Utilizing this probe, the present study could show the presence of a specific population of epidermal cells also in the tail that can differentiate into the adult-type germinative cells under the direct action of thyroid hormone. These cells are most likely basal cells. We suggest that the tail epidermis is competent to transform into the adult type.

184 K. Yosuizato, A. NISHIKAWA et al.

(A)

MATERIALS AND METHODS

The bullfrog, Rana catesbeiana, was used in the present study. Tadpoles were obtained from a local supplier and staged according to Taylor and Kollros [9].

Epidermal cells were prepared from the tail and the body of tadpole as described previously [6]. Cells were cultured in medium of RPMI-1640 (GIBCO) containing 7% fetal calf serum (Com- monwealth Serum Laboratory) which had been treated with charcoal (Wako Pure Chemical Indus- tries LTD.) to eliminate thyroid hormone as described previously [11]. RPMI medium was diluted with distilled water to 70% of its original strength. Cells were maintained at 24°C in a humidified incubator gassed with 5% CO, and 95% air. They were precultured for 2 days in the absence of 3,3’,5-triiodo-L-thyronine (T3, Sigma Chemical Co.) and media were changed (day 0) with fresh medium with (experimentals) or without (controls) T3. When the culture extended more than 4 days, media were changed at day 4.

Histological studies were performed as follows. The skin was removed from tadpoles, fixed with 10% neutralized formalin, embedded in paraffin and sectioned 4 um thick. Staining with toluidine blue was performed according to Robinson and Heintzelman [8]. Sections for electron microscopy were prepared according to Kaiho et al. [2] and were viewed through a JE-100SX electron micro- scope. Sections for immunohistochemistry were also processed as described above. Cells for this aim were fixed with 10% neutralized formalin or methanol. The antigen A was detected by the method of Kaiho and Ishiyama [1] using anti-A antiserum (rabbit, Takeda) and Vecstastain ABC kits (Vector Lab.) as an indicator system.

Fic. 1. Immunostaining of the epidermis by antisera against human blood group antigen A. (A) Cross section of the adult dorsal body skin. a, Section stained with toluidine blue, showing a typical five- or six- layered epidermal cells on the dermis. b, Immunostained section, showing that granular cells are positive but outermost keratinized and inner- most germinative cells are negative. c, Negative control section stained with the anti-A antiserum which had been immunoabsorbed by human group A blood cells. Black granules in the epidermis were

: ae © a melanins of pigment cells. Magnification: 200. (B) Immunostaining of the dorsal body skin of tadpoles during spontaneous metamorphosis. The number indicated in each figure shows the metamor- phic stage defined by Taylor and Kollros [9]. A few positive cells appeared first at stage XX. The black-colored cells in the dermis are melanocytes. Magnification: 200. (C) An electron micrograph of a cell which reacts with anti-A antiserum. Mag- nification: 2000. (D) A section of the body-tail junction of the skin at stage XXI. The section was immunostained by anti-A antiserum. Black bodies under the epidermis are melanocytes. Note the clear immunological difference of the tail and the body epidermis. Magnification: 150.

Metamorphic Fate of Tadpole Skin 185

RESULTS AND DISCUSSION

It was found that the human blood group antigen A (A) is a useful probe for the analysis of transition of the epidermis during metamorphosis. Keratinizing cells in the stratum spinosum of the adult body skin of bullfrog were immunostained by the anti-A antiserum (Fig. 1A). This reaction was specifically immunoabsorbed by human A group blood cells. The keratinized outermost layer of the stratum corneum and the innermostlayer of the stratum basale were negative to the antiserum. These indicate that keratinizing daughter cells of basal cells in the germinative layer express human blood group antigen A.

Contrary to the adult skin, we found that there exist no epidermal cells positive to the antiserum in the entire skin of tadpoles at pre- and prometa- morphic stages (Fig. 1B). A few positive cells appeared in the epidermis of the body at stage XX, which were flat in shape (Fig. 1C). The presence of A-positive cells at stage XX indicates the appearance of adult-type germinative cells in the body skin at this critical stage of metamorphosis, because they are differentiated daughter cells of germinative cells as shown in FigurelA. The A-antigen expressing cells increased dramatically in their numbers at stage XXI through stage XXV, when the tail is lost completely, suggesting that basal cells are actively transforming into the germinative cells during climax stages of meta- morphosis. It, therefore, can be concluded that the anti-A antiserum is a useful probe for the analysis of the conversion of basal cells from the larval type to the adult one during metamorphosis.

As expected, the tail epidermis did not contain the cells reactive to anti-A antibodies at stage XXI (the climax metamorphic stage) (Fig. 1D), indicat- ing the absence of cells transforming into the adult type. As shown in the figure, the body and the tail epidermis was clearly separated at stage XXI for their reactivity to the antibody.

In vitro studies on epidermal cells clearly de- monstrated that thyroid hormone (TH) is directly involved in the conversion of the larva-type epidermis to the adult type. The back skin was treated first with ethylenediaminetetraacetate (EDTA) to remove apical cells. Skein and basal

cells were isolated from the EDTA-treated skin by digesting it further with trypsin and EDTA [6]. Cells thus obtained contained skein cells and basal cells with ratios of about 70% and 30%, respec- tively. They were cultured for 4 days in the presence of T3 at the concentration of 10° M. The body cells were induced to express the antigens reactive to anti-A antiserum (Fig. 2, b and d). The identical in vitro study was performed for primary tail epidermal cells which were almost a homogeneous population of skein cells. Unex- pectedly, tail epidermal cells were also induced to express the adult-type specific antigen as body cells (Fig. 2, a and c). Nishikawa ef al. [5] recently obtained the similar result utilizing other adult- specific antigen as a prove. Epidermal keratin of Xenopus laevis with a molecular weight of 63 K is expressed in the adult but not in the larva. Tail epidermal cells of X. /aevis were cultured in the presence of TH and shown to express 63 KDa

BS Vas

EX x a ; ————

Fic. 2. Induction of anti-A positive cells in cultured epidermal cells. Eight hundred thousand epidermal cells obtained from tadpoles at stage X were cul- tured in 22-mm dishes at 24°C for 2 days. Media were changed and the culture was continued for additional 4 days in the presence or absence of 10 * M T;. Cells were immunostained with anti-A anti- serum. a and c¢, tail cells; b and d, body cells; a and b, in the absence of T;; c and d, in the presence of T;. Anti-A positive cells are colored brown. Com- pletely keratinized cells are not stained (arrow- heads). Bars represent 100 ~m.

186 K. YosHizaTo, A. NISHIKAWA et al.

keratin.

It is reasonable and expectable that the epidermis of the body contains the cells which express adult type characters under the influence of TH. Basal and skein cells in the primary culture seem to respond differently to TH. The former is stimulated to proliferate and then transform into the antigen-expressing adult-type cell, while the latter is fallen into the cell death by the direct action of TH [4]. The result of in vitro experiment for body epidermal cells coincides with that of in vivo experiment shown in Figure 1.

On the other hand, the in vitro result for the tail cells is against expectation in the following four reasons: the tail epidermis apparently need not provide for the adult-type cell, because its life ends up at metamorphosis [10]; the tail epidermis does not express the adult-type character in vivo at the climax stage of metamorphosis as shown above in Figure 1D; the tail epidermal primary culture does not contain basal cells as mentioned above; it was shown that the tail epidermis lacks basal cells [8]. However, the result shown in Figure 2 presents clear evidence that the tail contains a population of cells which is competent to convert to the adult type in response to TH.

The cells in the tail that expressed A-antigen in response to TH should be basal cells, because skein cells are to be subjected to the programmed cell death by the direct action of TH [4]. Two possibilities arise to explain the result of in vitro experiment of tail cells. One is that basal cells might be present in the tail cell population with so low frequencies as not to be detected micoscopi- cally. This minor population of basal cells is stimulated by TH to proliferate and differentiate into the adult-type cells. The time course experi- ment on the A-antigen expression was performed. Few cells in the body but not in the tail expressed the antigen at day 0 (Fig. 3). These cells might be contaminated epidermal glandular cells which are known to exist in the body but not in the tail and produce A-antigens (Izutsu and Yoshizato, manu- script in submission). During 4 to 6 days in culture T3 markedly enhanced the appearance of positive cells in both the tail and the body as in the case shown in Figure 2. It was noteworthy that a few positive cells appeared at day 4 to day 6 depending

6

SP ane

SLGANO’

and body epidermal cells in the absence of TH. Hundred thousand epidermal cells from tadpoles at stage X were cultured in 11-mm dishes for 2 days at 24°C. Media were changed (day 0) and the culture was continued for additional 6 days with a medium change at day 4. Cells were immunostained with anti-A antiserum. a and b, body cells; c and d, tail cells; a and c, day 0; b and d, day 6 without T;. The bar equals 100 «2m.

on experiments in a T3-free culture of tail cells which contained about 10° cells (Fig. 3d). This supports the possibility described above that very few basal cells are included in the tail epidermis. Also a few additional positive cells appeared in T3-free cultures of body cells, suggesting that the “differentiation” of the larval basal cell into the adult germinative cell could occur spontaneously at a very low level.

The second possibility is described below and is related to the difference between in vivo and in vitro status of epidermal cells. The reactivity of tail epidermal cells to anti-A antiserum was quite different between in vivo and in vitro experiments. As shown in Figure 1D, the tail epidermis in vivo does not show the positive reactivity even at the

Metamorphic Fate of Tadpole Skin 187

climax stages of metamorphosis in which the plasma level of TH is high enough [7]. There might be some unknown factor(s) in vivo which suppresses the expression of A-antigen in the tail epidermis even in the presence of TH.

The character of the epidermis as tail or body seems to be determined by some mesenchymal factor. Utilizing transplantation techniques, Kinoshita et al. [3] revealed that the epidermis combined with the tail dermis undergoes degen- erative changes during metamorphosis, irrespec- tive of the origin of the epidermis (tail or body), while the tail epidermis combined with the body dermis transforms into the body-type and diffe- rentiates into the adult type. Based upon these we propose the second possibility that skein cells are not a homogeneous population: some skein cells could transform into basal cells. The epithe- liomesenchymal interaction is postulated for this transformation. The body mesenchyme is required for the epidermal skein cell to “differentiate” into the basal cell. In contrast, the tail mesenchyme inhibits this transformation. When tail epidermal cells are removed from its mesenchyme and are cultured, some of them transform into basal cells

because they are free from the influence of the mesenchyme. Thus, the tail epidermal cell in culture is competent to express A-antigens when exposed to TH.

REFERENCES

1 Kaiho M, Ishiyama I (1987) Zool Sci 4: 627-634

2 Kaiho M, Nakamura T, Kumegawa M (1975) Anat Rec 183: 405-420 3 Kinoshita T, Sasaki F, Watanabe K (1986) Cell Tissue Res 245: 297-304 4 Nishikawa A, Kaiho M, Yoshizato K (1989) Dev Biol 131: 337-344 5 Nishikawa A, Shimizu-Nishikawa K, Miller L (1992) Dev Biol 151: 145-153 6 Nishikawa A, Yoshizato K (1985) Zool Sci 2: 201- 211 7 Regard E, Taurog A, Nakashima T (1978) Endocri- nology 102: 674-684 8 Robinson DH, Heintzelman MB (1987) Anat Rec 217: 305-317 9 Taylor AC, Kollros JJ (1946) Anat Rec 94: 7-23 10 Yoshizato K (1989) Int Rev Cytol 119: 97-149 11 Yoshizato K, Kikuyama S, Shioya N (1980) Biochem Biophys Acta 627: 23-29

ts | ‘ednine re

Rin sureninl sl) ane son we vod sevecrad i): Hos! CeMreeaits biel aie’? aed Se salie pera rmit@ aq igiten-F eee’ aT’ rey ane eee felis

i >. sh %s te i ial berinnas isa ) Pi c o m ,: A i ve ee et dus ae ;

ro Thad Ware CMBNG-anaeglabal At re “4 > | gis * ti ithe ridin hex

2 : MeN, aia Loud na sa ame

= . : bi ¢ ty A : ¢ ig Niece oe A { ° - ca ’ t f i> vend 2 t »y i =. , y ie refit 4 . vat pet { : } Whliva nord ~ an i Ht i - . ‘ i? - 4 : i t tT , : f i f b) M ‘ ‘ fF ’ ’ ‘ } tae i

ZOOLOGICAL SCIENCE 10: 189-192 (1993)

[RAPID COMMUNICATION]

© 1993 Zoological Society of Japan

Kinetics of First Spermatogenesis in Young Toads of Xenopus laevis

Touru Kosayasut’*, YoHKo Asakawa! and Hisaakt IwASAWwA

1

'Biological Institute, Faculty of Science, and Graduate School of Science and Technology, Niigata University, Niigata 950-21, Japan

ABSTRACT—The kinetic profile of first spermatogene- sis was examined at 22+1°C in young toads of Xenopus laevis. The most progressed spermatogenic stage was primary spermatogonia in newly metamorphosed toads. Secondary spermatogonia were observed 3 days after metamorphosis. First leptotene and pachytene sper- matocytes appeared 17 and 20 days after metamorphosis, respectively. Then first diplotene spermatocytes, meiotic division and round spermatids were observed 42 days, and first mature spermatozoa 70 days, after metamorpho- sis. On the basis of these results, the difference between the time required for first spermatogene- sis in young toads and cycling spermatogenesis in adult toads is discussed.

INTRODUCTION

The kinetics of spermatogenesis in amphibians has been studied by several investigators, but the materials used in their studies were mostly sub- adult and adult animals [2, 4, 9, 17]. The formation of spermatozoa in juvenile frogs has been reported as precocious spermatogenesis in several species [5, 7, 22, 23], but these papers simply describe the phenomenon. Recently, we defined the kinetics of cycling spermatogenesis at 22°C in adult toads of Xenopus laevis [2]. In the present study, the time after metamorphosis at which the most rapidly progressed spermatogenic cells appeared at 22°C is examined in young toads of this species, and difference between these

Accepted August 18, 1992 Received June 8, 1992

results and the kinetics of cycling spermatogenesis in adult toads was discussed.

MATERIALS AND METHODS

Numerous fertilized eggs were obtained from a few pairs of male and female toads Xenopus laevis injected with a human chorionic gonadotropin (Gonatropin, Teikoku Zoki Co., Tokyo). The eggs, embryos, and larvae were kept in dechlori- nated tap water at 22+1°C. The larvae and young toads were fed on a commercial diet for carp. These details were previously described [3, 8].

Until the time of the first appearance of sperma- tozoa after metamorphosis, three to four young toads were decapitated every day, and the testes were fixed in Bouin’s solution. The testes were then embedded in Paraplast (Sherwood Medical, St. Louis, U.S.A.), and sectioned serially at 4 ~m. The sections were stained with Carazzi’s hema- toxylin and eosin.

RESULTS AND DISCUSSION

In this study, the average body length just after metamorphosis was 15 mm, and was 20 mm at the end of experiment (82 days after metamorphosis). As previously reported [8], seminiferous tubules were clearly observed in young toads just after metamorphosis. At this time (0 day after meta- morphosis), the seminiferous tubules were filled with primary spermatogonia. The days on which

190 T. Kosayasul, Y. ASAKAWA AND H. Iwasawa

TaBL—E1. The most rapidly progressed sperma- togenic stage of first spermatogenesis in Xenopus laevis

FeGnee DEES

0 primary spermatogonium

3-16 secondary spermatogonium

17-19 primary spermatocyte (leptotene)

20-29 zygotene

30-41 pachytene

42-64 diplotene-round spermatid

65-69 elongated spermatid

70-82 spermatozoon

the most rapidly progressed spermatogenic stages were observed are shown in Table 1. The first appearance of spermatocytes and spermatids was observed on 17 days and 42 days after metamorpho- sis, respectively. Elongated spermatids were seen 65 days after metamorphosis. Seventy days after metamorphosis, the first spermatozoa were observed, though the number is a few (Fig. 1a). When the round spermatids differentiated into the elongated spermatids (from 45 to 70 days after metamorphosis), pycnotic germ cells were fre- quently observed (Fig. 1b).

From the results shown in Table 1, the duration of each spermatogenic process was calculated. The duration of leptotene, zygotene, and pachytene spermatocytes was 3, 10, and 12 days. The progression from the diplotene spermatocytes to the round spermatids was accomplished within one day. The duration of spermiogenesis was 28 days, whereas the progression of elongated spermatids to spermatozoa was 5 days.

In a previous study, we determined the time required for the progression of cycling sperma- togenesis at 22°C in adult X. laevis ({2], Koba- yashi, unpublished). Although the time required for the progression from elongated spermatids to spermatozoa was similar between young and adult toads (young toads: 5 days, adult toads: 4 days), the present study (Table 1) indicates that the progression from round spermatids to elongated spermatids in young toads was greatly delayed compared with that in adult toads [2] (young toads:

ere sb? @, & 2 rl 0 @ 4°:

ean . is a Te ae Fic. 1. First spermatogenesis in X. Jaevis. a: First spermatozoa (arrow) are seen 70 days after meta- morphosis. b: Pycnotic substances (arrow head) are

frequently seen from 45 to 70 days after metamorpho- sis. Scale bar: 30 ~m.

23 days, adult toads: 4 days). In this study, furthermore, during spermiogenesis the degenera- tion of germ cells, probably round to elongated spermatids, was frequently observed. From these facts, it is suggested that the delay of spermiogenic progression in first spermatogenesis is caused by the failure of nuclear elongation of spermatids. It seems, therefore, that the time required for the progression from round to elongated spermatids in first spermatogenesis is the real time required for the progression from round to elongated sperma- tids plus the time required for the appearance of round spermatids which could be differentiated into elongated spermatids. Thus, it is calculated that first appearance of round spermatids which

Kinetics of First Spermatogenesis 191

could be differentiated into elongated spermatids is 19 days after the first appearance of round spermatids, if the duration of progression from round to elongated spermatids is similar to that in adult toads as in mice [21]. To define the dynamics in first spermatogenesis, further studies will be necessary.

In X. laevis and Cynops pyrrhogaster, it is known that in an in vitro cell culture system progression from primary spermatocytes to elon- gated spermatids was accomplished but sper- miogenesis failed to be completed [1]. When nuclear elongation proceeds in vivo, in X. laevis, a few of microtubules contained in spermatids were cast off along with cytoplasm and numerous microtubules in Sertoli cells were situated approxi- mately parallel to the axis of spermatids elongation in the area which immediately surround the spermatids [18]. In X. Jaevis [19], furthermore, progression from premeiotic S spermatogenic cells to mature spermatozoa was accomplished in an in vitro organ culture system. Thus, in X. laevis, it seems that Sertoli cells play the primary role for the nuclear elongation of spermatids. Therefore, the failure of the progression from round to elongated spermatids may be caused by the lack of functional Sertoli cells and/or abnormality of spermatids, though the functional characterization of Sertoli cells and spermatogenic cells was not clarified in details.

In Rana esculenta it was suggested that the completion of spermiogenesis required androgen [15, 16]. Concerning the experiment which dealt with precocious spermatogenesis in Rana nigroma- culata, we pointed out that androgen was needed for the completion of spermiogenesis [10, 11, 24, 25]. In immature rats, on the other hand, it was reported that an increase of luteinizing hormone binding sites in Leydig cells was induced by prolactin (PRL), though this phenomenon was accompanied by the decrease of PRL-binding sites [12]. On this point, in anurans, it is known that serum PRL levels are high at the climax of metamorphosis [23]. In X. laevis, it was reported that gonadotropin- and PRL-producing cells were observed at premetamorphic stages [13, 14]. Furthermore, it was reported that PRL binding sites was present in testes [6]. Although, at

present, the information on these pituitary hor- monal levels and gonadal steroid levels in young toads of X. laevis is few, defined steroid producing cells in the testes were not observed electron microscopically 70 days after metamorphosis (Kobayashi, unpublished). Therefore, it is possi- ble that androgen is involved in spermiogenic process. In this study, however, whether or not the relationship between PRL and androgen pro- duction during first spermatogenesis is not known.

In conclusion, the present study indicates in X. laevis that the duration of spermatogenesis in first spermatogenesis in young toads is much longer than that of cycling spermatogenesis in adult toads. This delay may be caused by the failure of nuclear elongation of spermatids.

REFERENCES

1 Abe S-I (1987) Int Rev Cytol 109: 159-209

2 Asakawa Y, Kobayashi T, Iwasawa H (1990) Zool Sci 7: 1088 (Abstract)

3 Asakawa Y, Kobayashi T, Iwasawa H (1991) Biomed Res 12: 269-272

4 Callan HG, Taylor JH (1967) J Cell Sci 3: 615-626

5 Goto G (1940) Acta Anat Nippon 22: 30-43 (Abstract in English)

6 Guardabassi A, Muccioli G, Pattono P, Bellussi G (1987) Gen Comp Endocrinol 65: 40-47

7 Iwasawa H, Kobayashi M. (1976) Copeia 1976: 461- 467

8 Iwasawa H, Yamaguchi K (1984) Zool Sci 1: 591- 600

9 Kalt MR (1976) J Exp Zool 195: 393-408

10 Kobayashi T, Iwasawa H (1988) Copeia 1988: 1067- 1071

11 Kobayashi T, Iwasawa H (1986) Zool Sci 3: 387-390

12 Morris PL, Saxena BB (1980) Endocrinology 107: 1639-1645

13. Moriceau-Hay D, Doerr-Schott J, Dubois MP (1979) Gen Comp Endocrinol 39: 322-326

14 Moriceau-Hay D, Doerr-Schott J, Dubois MP (1982) Cell Tissue Res 225: 57-64

15 Rastogi RK, Iela L, Saxena PK, Chieffi G (1976) J Exp Zool 196: 151-166

16 Rastogi RK, Iela L, Delrio G, diMeglio M, Russo A, Chieffi G (1978) J. Exp Zool 206: 49-64

17 Rastogi RK, Iela L, diMeglio M, Russo A, Chieffi G (1983) J Zool Lond 201: 515-525

18 Reed SC, Stanley HP (1972) J Ultrastruct Res 41: 277-295

19 Risley MS, Miller A, Bumcrot RA (1987) Biol

20

21

22

192

Reprod 36: 985-997

Rosenkilde P (1985) In “Metamorphosis”, Ed by M Balls and M Bownes, Clarendon Press, Oxford, pp 221-259

Sung WK, Komatsu M, Jagiello GM (1986) Gamete Res 14: 245-254 Swingle WW (1921) J Exp Zool 32: 235-331

23

24

25

T. KopayasHi, Y. ASAKAWA AND H. Iwasawa

Takahashi H (1971) Zool Mag (Tokyo), 80: 15-21 (Abstract in English)

Tanaka S, Iwasawa H, Wakabayashi K (1988) Zool Sci 5: 1007-1012

Tanaka S, Iwasawa H (1989) Sci Rep Nugata Univ Ser D 26: 1-11

193

INSTRUCTIONS TO AUTHORS

ZOOLOCIGAL SCIENCE publishes contri- butions, written in English, in the form of (1) Reviews, (2) Articles, and (3) Rapid Communica- tions for short reports of timely and unusual interest. A Review is usually invited by the Editors. Those who submit reviews should consult with the Editor-in-Chief or the Managing Editor in advance. Articles of less than 6 printed pages and Rapid Communications less than 3 printed pages will be published free of charge. Charges will be made for extra pages (10,000 yen/page). A Rapid Communication cannot exceed 4 printed pages. No charge will be imposed for invited reviews up to 15 printed pages. No free reprints of Articles and Rapid Communications are available. To the author(s) of an invited review 50 reprints are provided gratis. Submission of papers from nonmembers of the Society is welcome. However, page charges (10,000 yen/page) will be made to nonmembers.

A. SUBMISSION OF MANUSCRIPT

The manuscript should be submitted in tripli- cate, one original and two copies, each including all illustrations. Rough copies of line drawings and graphs may accompany the manuscript copies, but the two copies of continuous-tone prints (photo- micrographs, etc.) should be as informative as the original. The manuscript should be sent to:

Dr. Tsuneo Yamaguchi, Managing Editor, Zoological Science, Department of Biology, Faculty of Science, Okayama University, Okayama 700, Japan.

B. CONDITIONS

All manuscripts are subjected to editorial re- view. A manuscript which has been published or of which a substantial portion has been published elsewhere will not be accepted. It is the author’s responsibility to obtain permission to reproduce illustrations, tables, etc. from other publications. Accepted papers become the permanent property of ZOOLOGICAL SCIENCE and may not be reproduced by any means, in whole or in part,

without the written consent of both the Zoological Society of Japan and the author(s) of the article in question.

C. ORGANIZATION OF MANUSCRIPT

The desirable style of the organization of an original paper is as follows: (1) Title, Author(s) and Affiliation (2) Abstract (3) Introduction (4) Materials and Methods (5) Results (6) Discussion (7) Acknowledgments (8) References (9) Tables (10) Illustrations and Legends. The author is not obliged to adhere rigidly to this organization. He or she may modify the style when such modifica- tion makes the presentation clearer and more effective. In a Rapid Communication, combina- tion of some of these sections is recommended. There is no restriction on the style of review articles.

D. FORM OF MANUSCRIPT

Manuscripts should be typewritten and double spaced throughout on one side of white typewrit- ing paper with 2.5cm margins on all sides. Abstract not exceeding 250 words, tables, figure legends and footnotes should be typed on separate sheets. All manuscript sheets must be numbered successively. The use of footnotes to the text is not recommended.

1. Title page

The first page of manuscript should contain title, authors’ names and addresses of university or institution, abbreviated form of title (40 characters or less, including spaces), and name and address for correspondence, and any footnotes. Authors with different affiliations should be identified by the use of the same superscript on name and affiliation. If one or more of the authors has changed his or her address since the work was carried out, the present address(es) to be pub- lished should be indicated in a footnote. In addition, a sub-field of submitted papers to be used as heading of an issue may be indicated in the first page. Authors are encouraged to choose one of the following: physiology, cell biology, molecular

194

biology, genetics, immunology, biochemistry, de- velopmental biology, reproductive biology, endo- crinology, behavior biology, ecology, phylogeny, taxonomy, or others (specify). However, the Editors are responsible for the choice and arrange- ment of headings. 2. Introduction

This section should clearly describe the objec- tives of the study, and provide enough background information to make it clear why the study was undertaken. Lengthy reviews of past literature are discouraged. 3. Materials and Methods

This section should provide the reader with all the information that will make it possible to repeat

the work. For modification of published methodol- .

ogy, only the modification needs to be described with reference to the source of the method. 4. Results

Results should be presented referring to tables and figures, without discussion. 5. Discussion

The Discussion should include a concise state- ment of the principal findings, a discussion of the validity of the observations, a discussion of the findings in the light of other published works deal- ing with the same subject, and a discussion of the significance of the work. Redundant repetition of material in Introduction and Results, and extensive discussion of the literature are discouraged. 6. Statistical analysis

Statistical analysis of the data using appropriate methods is mandatory and the method(s) used must be cited. 7. References

References should be cited in the text at the appropriate places by numbers in square parenthe- ses. All references cited in the text should be listed at the end of the paper on a separate page, arranged in alphabetical order and numbered consecutively. For example:

1 Campbell RC (1974) Statistics for Biologists. Cambridge Univ. Press, London, 2nd ed, pp 59-61

2 Shima A, Ikenaga M, Nikaido O, Takabe H,

Egami N (1981) Photoreactivation of ultra- .

violet light-induced damage in cultured fish cells as revealed by increased colony forming

ability and decreased content of pyrimidine dimers. Photochem Photobiol 33: 313-316 3. Takewaki K (1931) Oestrous cycle of female rat in parabiotic union with male. J Fac Sci Imp Univ Tokyo, Sec IV, 2: 353-356 4 Wiersma CAG (1961) Reflexes and the central nervous system. In “The Physiology of Crustacea Vol2” Ed by TH Waterman, Academic Press, New York, pp 241-279 Titles of cited papers may be omitted in Reviews and Rapid Communications. The source of refer- ence should be given following the commonly accepted abbreviations for journal titles (e,g., refer to ‘International List of Periodical Title Abbreviations’). The use of “in preparation”, “submitted for publication” or “personal com- munication” is not allowed in the reference list. “Unpublished data” and “Personal communica- tion” should appear parenthetically following the name(s) in the text. Text citations to references with three or more authors should be styled as, e.g., Everett et al. [7].

E. ABBREVIATIONS

Abbreviations of measurement units, quantity units, chemical names and other technical terms in the body of the paper should be used after they are defined clearly in the place they first appear in the text. However, abbreviations that would be rec- ognized by scientists outside the author’s field may be used without definition, such as P, SD, SE, DNA, RNA, ATP, ADP, AMP, EDTA, UV, and CoA. The metric system should be used for all measurements, and metric abbreviations (Table)

Table. Abbreviations for units of measure which may be used without definition

Length: km, m, cm, mm, um, nm, pm, etc.

Area: km?, m*, cm?, mm? ym”, nm?, pm”, etc.

Volume: km?, m?, cm®, mm?, ~m°, nm*, pm’, kl, liter (always spellout), ml, sl, nl, etc.

Weight: kg, g, mg, wg, ng, pg, etc.

Concentration: M, mM, uM, nM, %, g/l, mg/l, ug/l, ete.

Time: hr, min, sec, msec, psec, etc.

Other units: A, W, C, atm, cal, kcal, R, Ci, cpm,

dB, v, Hz, lx, Xg, rpm, S, J, IU, etc.

should, in general, be expressed in lower case without periods.

F. PREPARATIONS OF TABLES

Tables should only include essential data needed to show important points in the text. Each table should be typed on a separate sheet of paper and must have an explanatory title and sufficient explanatory material. All tables should be referred to in the text, and their approximate position indicated in the margin of manuscript.

G. PREPARATION OF ILLUSTRATIONS

All figures should be appropriately lettered and labelled with letters and numbers that will be at least 1.5 mm high in the final reproduction. Note the conventions for abbreviations used in the journals so that usage in illustrations and text is consistent. All figures should be referred to in the text and numbered consecutively (Fig. 1, Fig. 2, etc.). The figures must be identified on the reverse side with the author’s name, the figure number and the orientation of the figure (top and bottom). The preferred location of the figures should be indi- cated in the margin of the manuscipt. Illustrations that are substandard will be returned, delaying publication. Illustrations in color may be pub- lished at the author’s expense.

1. Line drawings and graphs

Original artwork of high quality, glossy prints mounted on appropriate mounting card (less than 25 X38 cm) should be submitted for reproduction. Author(s) may indicate size preference by making on the back of figures, such as “Do not reduce”, “Two-column width” (no wider than 14cm), or “One-column width” (no wider than 7 cm). Lines must be dark and sharply drawn. Solid black, white, or bold designs should be used for histo- grams. Xerox or any other copying mean may be used for the two review copies.

2. Continuous-tone prints

Three sets of continuous-tone prints (photo- micrographs, etc.) must be submitted. One set for reproduction should be mounted on appropriate mounting card, and the other two for reviewers may be unmounted prints. Xerox or similar copies of photomicrographs are not acceptable for review purposes. The continuous-tone prints should be

195

submitted preferably at the exact magnification which is to be used in the published papers and trimmed to conform to the page size (in no case should it exceed 1420cm). Press-on numbers should be applied to the lower right corner of individual prints. Letters (a, b, c, etc.) should be used for multiple parts of a single figure. If important structures will be covered by use of the lower right corner, identification may be applied in the lower left corner.

Reproduction of color photographs will have to be approved by the Editors. The extra costs of color reproduction will be charged to the au- thor(s).

3. Figure legends

Each figure should be accompanied by a title and an explanatory legend. The legends for several figures may be typed on the same sheet of paper. Sufficient detail should be given in the legend to make it intelligible without reference to the text.

H. PROOF AND REPRINTS

A galley proof and reprint order will be sent to the submitting author. The first proofreading is the author’s responsibility, and the proof should be returned within 72 hours from the date of receipt (by air mail from outside Japan). The minimum quantity for a reprint order is fifty. Manuscript, tables and illustrations will be discarded after the editorial use unless their return is requested when the manuscript is accepted for publication.

I. WORD-PROCESSOR DISKS

The Zoological Science can use your word- processor disks (51/4 or 312 inch). If available, please send a copy of the disk with your final revised manuscript (two copies of the printout). This gives us a substantial decrease in typesetting time and assures you that there are no re-keying errors in the article. Moreover, using word- processor disks helps us to shorten publication times. If you would be willing to lend us the disk, please obey the following instructions.

1. Label the disk with the author’s name; the word processor/computer used (e.g., NEC 9801, IBM-PC, Macintosh); the name of program (e.g., WordPerfect, WordStar, MS-Word, MacWrite).

196

If you have an editor’s manuscript number, include it on the label.

2. Send the manuscript as three files; if possi- ble keep the text, figure legends, and tables in separate files. File names must clearly indicate the content of each file.

3. Send separate disks if you submit more than one article at a time.

4. Ahard copy printout of the manuscript that exactly matches the disk files must be supplied. Special characters and symbols which are not

included in the printer used, should be left as blank spaces in the files and these should be correctly written in the printout in red.

5. Do not use footnote.

6. Do not justify the right-hand margin.

7. For further information on preparing the disk, please contact the publisher (Daigaku Letter- press Co., Ltd., 809-5, Asakita-ku Kamifuka- wacho, Hiroshima 739-17, Japan; telephone: (082) 844-7500; fax:(082) 844-7800).

D eve | O m e Nn t Published Bimonthly by the Japanese Society of Pp Developmental Biologists Distributed by Business Center for Academic

Growth & Differentiation Societies Japan, Academic Press, Inc.

Papers in Vol. 35, No. 1. (February 1993)

1. REVIEW: S. Yasugi: Role of Epithelial-Mesenchymal Interactions in Differentiation of Epithelium of Vertebrate Digestive Organs

2. M. Fukuzawa and H. Ochiai: Spatiotemporal Patterning of Discoidin I and II during Development of Dictyostelium discoideum

3. H. Grunz: The Dorsalization of Spemann’s Organizer Takes Place during Gastrulation in Xenopus laevis embryos

4. K. Urase and S. Yasugi: Induction and Inhibition of Epithelial Differentiation by the Mixed Cell Aggregates of the Mesenchymes from the Chicken Embryonic Digestive Tract

5. R.G. Summers, J. B. Morrill, A. Leith, M. Marko, D. W. Piston and A. T. Stonebraker: A Stereometric Analysis of Karyokinesis, Cytokinesis and Cell Arrangements during and following Fourth Cleavage Period in the Sea Urchin, Lytechinus variegatus

6. T. Itoh, K. Ohsumi and C. Katagiri: Remodeling of Human Sperm Chromatin Mediated by Nucleoplasmin from Amphibian Eggs

7. T. Kitamura, S. Kobayashi and M. Okada: Developmentally Regulated Splicing of the Third Intron of P Element in Somatic Tissues in Drosophila Embryos

8. A. Yanagi: Role of Germ Nuclei in Conjugation of Paramecium caudatum

9. T. Atsumi, Y. Miwa, Y. Eto, H. Sugino, M. Kusakabe, H. Kitani and Y. Ikawa: The Activin A-dependent Proliferation of PCC3/A/1 Embryonal Carcinoma Cells in Serum-free Medium

10. Y.Kamata, A. Fujiwara, S. Furuya and I. Yasumasu: Does ADP-ribosylation of Proteins in Nuclei Contribute to Ectoderm Cell Differentiation in Sea Urchin Embryos?

11. M. Kiyomoto and H. Shirai: The Determinant for Archenteron Formation in Starfish: Co-culture of an Animal Egg Fragment-derived Cell Cluster and a Selected Blastomere

12. M. Kiyomoto and H. Shirai: Reconstruction of Starfish Eggs by Electric Cell Fusion: A New Method to Detect the Cytoplasmic Determinant for Archenteron Formation

13. N. Shibata, M. Yoshikuni and Y. Nagahama: Vitellogenin Incorporation into Oocytes of Rainbow Trout, Oncorhynchus mykiss, in Vitro: Effect of Hormones on Denuded Oocytes

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.

The Ultimate Name in Micromanipulation Ease of operation, and the most advanced

Can be combined with New WR & MX series

Model MX -2 3-D Micromanipulator

Model MX-1 3-D Micromanipulator

Model WR-87

One Dimensional Aqua Purificate Micromanipulator

r J Model SR-6 Model PP -83 Stereotaxic Instrument for Rat Glass Microelectrode Puller

* * %& Send enquiries for request for MODIFICATIONS or IMPROVEMENTS * * * Physiological, Pharmacological, Zoological & Neurosciences Research Equipments

TWARISHIGE ) NARISHIGE SCIENTIFIC INSTRUMENT LAB.

LLY, 9-28 KASUYA 4-CHOME SETAGAYA-KU, TOKYO 157, JAPAN PHONE (INT-L) 81-3-3308-8233, FAX (INT-L)81-3-3308-2005

CABLE: NARISHIGE LABO, TELEX, NARISHIGE J27781

(Contents continued from back cover)

diacea, Tunicata) Kobayashi, T., Y. Asakawa, H. Iwasawa: Kinetics of first spermatogenesis in young toads of Xenopus laevis (RAPID COM- NMAWINICATION)) stcecscrecteneras teenies 189 Hamlin, G. P., S. D. Kholkute, W. R. Duke- low: Toxicology of maternally ingested car- bon tetrachloride (ccl4) on embryonal and fetal development and in vitro fertilization in TMCS odogowenreeceEeto ts mran ae eee nna 111 Nakamura, M., F. Tsuchiya, M. Iwahashi, Y. Nagahama: Reproductive characteristics of precociously mature triploid male masu salm- on, Oncorhynchus masou

Morphology

Miyazaki, K., T. Makioka: A case of inter- sexuality in the sea spider, Cilunculus arma- tus (Pycnogonida; ammotheidae) Zizavy, J.,O. Nedved, R. Socha: Metameric color pattern in the bugs (Heteroptera: Lygaeidae, Largidae, Pyrrhocoridae, Rho- palidae): A morphological marker of insect body compartmentalization

Systematics and Taxonomy Suzuki, T., T. Takagi, S. Ohta: N-terminal amino acid sequences of 440 kDa hemoglo- bins of the deep-sea tube worms, Lamellib- rachia sp. 1, Lamellibrachia sp. 2 and slender vestimentifera gen. sp. 1. Evolutionary rela- tionship with annelid hemoglobins Sawada, I., K. Koyasu, K. C. Shrestha: Two new species of the genus Staphylocystis (Ces- toda: Hymenolepididae) from the house shrew, Suncus murinus, in Nepal Tsurusaki, N., D. Song: Two new species of Sabacon from Sichuan province, China (Arachnida: Opiliones: Sabaconidae) ....155 Mashiko, K., K. Numachi: Genetic evidence for the presence of distinct fresh-water prawn (Macrobrachium nipponense) popula- tions in a single river system Miura, T., J. Hashimoto: Mytilidiphila, a new genus of nautiliniellid polychaetes living in the mantle cavity of deep-sea mytilid bivalves collected from the Okinawa trough

ZOOLOGICAL SCIENCE

VOLUME 10 NUMBER 1

FEBRUARY 1993

CONTENTS

REVIEWS yilenks, BuG 7 ewe Leenders, G. J. M. Mar- tens, E. W. Roubos: Adaptation physiolo- gy: The functioning of pituitary melanotrope cells during background adaptation of the amphibian Xenopus laevis ................. Fingerman, M., R. Nagabhushanam, R. Saro- jini: Vertebrate-type hormones in crusta- ceans: Localization, identification and func- tional significance Lambert, C. C., D. E. Battaglia: Loss of the paternal mitochondrion during fertilization

ORIGINAL PAPERS

Physiology Yasuyama, K., B. Chen, T. Yamaguchi: Im- munocytochemical evidence for the involve- ment of RFamide-like peptide in the neural

control of cricket accessory gland

Kamiya, Y., S. Harada, S. Okoyama, H. Yamamoto, R. Hosono: Developmental and pharmacological studies of acetylcho- linesterase-defective mutants of Caenorhab- ditis elegans

Cell Biology Takagi, Y., K. Nimura, Y. Tokusumi, H. Fujisawa, K. Kaji, H. Tanabe: Secretion of mitogenic factor(s) from stocks of Para- mecium tetraurelia, P. caudatum and P. multi- micronucleatum Kelly, K. L., E. L. Cooper, D. A. Raftos: Cytokine-like activities of a humoral opsonin from the solitary urochordate Styela clava

Biochemistry Sasaki, T., M. Sasaki, J. Enami: Mouse y- casein CDNA: PCR cloning and sequence analysis Mita, M., M. Nakamura: Phosphatidylcho- line is an endogenous substrate for energy metabolism in spermatozoa of sea urchins of the order echinoidea

Developmental Biology Itow, T.: Grafting of center cells of horseshoe crab embryos into host embryos at different developmental stages Suzuki, A., S. Nishimatsu, K. Murakami, N. Ueno: Differential expression of Xenopus BMPs in early embryos and tissues (RAPID COMMUNICATION) Koya, Y., K. Takano, H. Takahashi: Ultra- structural observations on sperm penetration in the egg of elkhorn sculpin, Alcichthys alcicornis, showing internal gametic associa- tion : Ishijima, S., Y. Hamaguchi, T. Iwamatsu: Sperm behavior in the micropyle of the medaka egg (RAPID COMMUNICATION)

Yoshizato, K., A. Nishikawa, Y. Izutsu, M. Kaiho: Epidermal cells of the tail of an anuran larva are competent to transform into the adult-type cells (RAPID COMMUNI- CATION)

Reproductive Biology Fukumoto, M., T. Numakunai: The acro- some and its differentiation during sper-

miogenesis in Halocynthia roretzi (Asci-

(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 Printed by Daigaku Letterpress Co., Ltd., Hiroshima, Japan

ZOOLOGICAL SC] ie

An Intern national Jou

ZOOLOGICAL SCIENCE The Official Journal of the Zoological Society of Japan

Editors-in-Chiet: The Zoological Society of Japan:

Seiichiro Kawashima (Tokyo) Toshin-building, Hongo 2—27-2, Bunkyo-ku,

Hideshi Kobayashi (Tokyo) Tokyo 113, Japan. Phone (03) 3814-5675 Managing Editor: Officae

Tsuneo Yamaguchi (Okayama) President: Hideo Mohri (Chiba) Assistant Editors:

Yoshihisa Kamishima (Okayama)

Masaki Sakai (Okayama)

Akiyoshi Niida (Okayama)

Secretary: Hideo Namiki (Tokyo) Treasurer: Makoto Okuno (Tokyo) Librarian: Masatsune Takeda (Tokyo)

Editorial Board:

Howard A. Bern (Berkeley) Walter Bock (New York) Aubrey Gorbman (Seattle) Horst Grunz (Essen) Robert B. Hill (Kingston) Yukio Hiramoto (Chiba) Susumu Ishii (Tokyo) Yukiaki Kuroda (Tokyo) John M. Lawrence (Tampa) Koscak Maruyama (Chiba) Roger Milkman (Iowa) Kazuo Moriwaki (Mishima) Richard S. Nishioka (Berkeley) Chitaru Oguro (Toyama) Tokindo S. Okada (Okazaki) Andreas Oksche (Giessen) Hidemi Sato (Nagano) Mayumi Yamada (Sapporo)

Ryuzo Yanagimachi (Honolulu) Hiroshi Watanabe (Tokyo)

ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and communications in the broad field of Zoology. The journal appears bimonthly. An annual volume consists of six numbers of more than 1200 pages including an issue containing abstracts of papers presented at the annual meeting of the Zoological Society of Japan.

MANUSCRIPTS OFFERED FOR CONSIDERATION AND CORRESPONDENCE CONCERN- ING EDITORIAL MATTERS should be sent to: Dr. 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 1992 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 10: 197-222 (1993)

REVIEW

Behavior of Exogenously-introduced DNAg. in’ Early Embryos of Xenopus laevis

MISAKI ASANO AND KOICHIRO SHIOKAWA

Laboratory of Molecular Embryology, Zoological Institute, Faculty of Science, University of Tokyo, Bunkyo-ku, 113 Tokyo, Japan

ABSTRACT—Studies of the behavior of circular and linearized DNAs, injected into fertilized eggs of Xenopus laevis, have been reviewed. The review consists mainly of three parts: Formation of nucleus-like structures in the cytoplasm of the DNA-injected eggs, preservation and replication of injected DNAs within embryos, and expression of injected DNAs in the course of early embryonic development. The last part of the review is concerned with the control of the expression of circular and linearized DNAs, and will be discussed in some details, particularly in relation to the timing of the expression in early embryogenesis.

Previously, it has been reported that DNAs injected into fertilized Xenopus eggs are activated first at the 12th cleavage (the midblastula stage) along with the changes which are collectively called midblastula transition (MBT). However, this review will focus on our own data, in which it was shown that the nature of the promoter, and not the changes associated with the MBT, determines the

© 1993 Zoological Society of Japan

timing of the expression of injected DNAs.

INTRODUCTION

Molecular biology of the gene expression in early amphibian embryogenesis started in 1962, when Brown and Caston [9] succeeded for the first time in extracting the RNA from Rana pipiens embryos of various developmental stages using phenol methods. In 1964, Brown and Littna [11] extended their study by _ injecting ~*’P- orthophosphate into female Xenopus laevis and subsequently following the accumulation of the *°P_labeled RNAs in the fertilized eggs during their development. Subsequently, Shiokawa and Yamana [70, 71], and Woodland and Gurdon [87] joined the studies of RNA synthesis in Xenopus embryonic cells. Consequently, a general view that in early embryogenesis of Xenopus laevis there are at least two characteristically distinct phases with respect to the pattern of RNA syn- thesis has been obtained [27]: The first phase being

Received September 21, 1992

that of very active 4S RNA (mainly tRNA) syn- thesis with no apparent rRNA synthesis (blastula stage) and the second phase being that of the active rRNA synthesis (gastrula and later stages).

In 1981, we surveyed newly labeled 2’ -O-methyl groups in high-molecular-weight RNA fractions using a large number of (*H-methyl)methionine- labeled embryos at stages earlier than the blastula stage, and found that rRNA synthesis starts at the late blastula stage and not before that stage [65, 68], approximately 4hr earlier than the gastrula stage. Subsequent to our studies, it has been proposed that large changes in cellular activities which are collectively termed as “midblastula tran- sition (MBT)” take place in the embryonic cells at the 12th cleavage [52, 53]: The changes include the appearance of G1 phase in cell cycle, acquisition of cell motility, and initiation of the transcription from the nuclear genome.

More recently, however, we have found that expression of heterogeneous mRNA-like RNA takes place during the cleavage stage [51, 61, 64].

198 M. ASANO AND K. SHIOKAWA

Thus, we proposed that RNA synthesis in early Xenopus embryogenesis consists of three different phases: The phase of a low but distinct level of mRNA-like RNA synthesis (cleavage stage), the phase of active tRNA synthesis (the blastula stage), and the phase of active rRNA synthesis (at and after the late blastula stage).

While the studies on gene expression from en- dogenous DNAs were in progress, Gurdon and Brown [28] went on to study the gene expression from exogenous DNA. They microinjected exog- enous genes (5S DNA and rDNA) into fertilized Xenopus laevis eggs and followed the expression of these genes throughout the course of their early embryonic development. Following their study, many laborotaries started analysis of the control mechanism of the expression of cloned DNAs that had been introduced into fertilized Xenopus eggs (for review of the initial step of the studies, see 26).

Since then, numerous papers concerning the behavior of exogenously-injected DNAs using fer- tilized eggs of Xenopus laevis have been published. In reviewing these studies, we will first describe briefly about the technical aspects of the DNA injection in Xenopus system, and then go on to describe the morphological changes which take place in the cytoplasm of DNAs-injected eggs. Then, we will describe the general aspects of preservation and replication of injected DNAs, and compare the results of the experiments carried out to study the expression of the various injected DNAs in the course of embryonic development. We will touch briefly on the experiments which were carried out using coenocytic egg cells, or fertilized eggs whose cleavage was arrested by a low speed centrifugation, then finally compare the behavior of exogenous DNAs introduced into oocytes, unfertilized eggs, and fertilized eggs.

UNIQUE ASPECTS OF DNA-INJECTION STUDIES WHICH UTILIZE FERTILIZED XENOPUS EGGS

It may be of an advantage to start with very brief description of the technical aspects of DNA injec- tion experiments which utilize fertilized Xenopus eggs. Unlike the microinjection in mouse embryos

where exogenous DNAs are injected into the female pronucleus directly, microinjection in Xenopus eggs has at least two unique aspects.

Firstly, it is unpractical to inject DNA into the nucleus of Xenopus egg, because the egg is not transparent, and therefore cannot visualize its nu- cleus. Consequently, DNAs must be injected into the cytoplasm of Xenopus egg. Secondly, Xenopus eggs cleave rapidly (every 30 min), and during the cleavage its nuclear envelope disappears and re- appears quickly in each division cycle. Therefore, even if it were possible to inject exogenous DNAs into the egg nucleus, the injected DNAs will be scattered promptly into the cytoplasm. However, this also means that exogenously-injected DNAs could be assembled rapidly in the nuclei together with chromosomal DNAs at each cleavage cycle, even though they were injected into the cytoplasm. Consequently, the cytoplasmic injection of exog- enous DNAs in a fertilized Xenopus laevis egg may be same in effect as the injection into the nucleus.

Thus, the DNA injection in fertilized Xenopus egg clearly differs from those in the mouse, where the division of egg nucleus is very slow (every 24 hr), and in the medaka (Oryzias latipes), where DNA is injected into the oocyte nucleus, and the injected oocytes were cultured for 8 hr till matura- tion and then fertilized in vitro [55]. However, similarity can be seen between the DNA injection in Xenopus and those in sea urchin [20, 48, 83] (except for the special case of unusually trans- parent eggs of Lytechinus variegatus, where DNA could directly be injected into zygotic nucleus under the microscope; [22]), in zebrafish [77] and in rainbow trout [88].

CYTOLOGICAL CHANGES OBSERVED IN FERTILIZED XENOPUS EGGS INJECTED WITH EXOGENOUS DNAS

In Xenopus laevis, fertilized eggs undergo rapid and multiple cell divisions in order to increase cell number quickly for future morphogenesis. The first step of the process is the protrusion of second polar body, which is induced by sperm entry, and is accompanied by the assemblage of the female pronucleus. It appears that the formation of the pronucleus depends on the maternal stockpiles

DNAs Introduced into Xenopus Embryos 199

within the egg cytoplasm, since formation of nu- cleus-like structures is induced even in an Jn vitro system which consists of demembranated sperm nuclei (or naked DNAs) and egg homogenates [35, 44, 54].

It has been shown that when a relatively large amount of exogenous DNA is injected into the cytoplasm of either unfertilized [21] or fertilized [17, 66, 69, 82] eggs of Xenopus laevis, large nucleus-like structures are formed in the cyto- plasm. Therefore, we shall describe the cytological features of such nucleus-like structures first.

Assemblage of the nucleus-like structures in the animal hemisphere

Unfertilized eggs of Xenopus laevis contain metaphase-arrested spindle in the animal-most re- gion of the egg (Fig. 1A). When fertilized, the arrested nucleus completes its nuclear division, and forms a female pronucleus. Uninjected con- trol eggs which were fixed 1.5 hr after insemination contained normal zygotic nucleus of ca. 25 ~m in diameter usually in the upper part of the animal hemisphere. When a large amount (24ng) of bacteriophage lambda DNA was injected into the

Fic. 1. Light microscopic appearance of nucleus-like structures. (A) Animal pole region of the control uninjected unfertilized Xenopus laevis egg. The arrow indicates the metaphase-arrested nucleus. An arrow head indicates the position of equatorial plate. (B) A fertilized Xenopus laevis egg was injected with 24 ng of bacteriophage lambda DNA using Narishige microinjector and the was fixed in glutaraldehyde about 1.5 hr after the injection. Arrows indicate cluster of nucleus-like structures. 330. From Shiokawa et al. [69].

200 M. ASANO AND K. SHIOKAWA

fertilized eggs, large nucleus-like structures (ca. 150 ~m in diameter) were formed. Figure 1B shows typical nucleus-like structures. Since these nucleus-like structures were not visible in the cytoplasm of the eggs which were injected with a low dose (1 ng/egg) of lambda DNA, the size of the nucleus-like structures appears to depend on the amount of DNAs injected.

Unlike unfertilized eggs which have no peri- vitelline fluid to permit free egg rotation, fertilized eggs rotate freely within the vitelline membrane, and therefore, their animal-vegetal axis is oriented along the gravity. In such eggs the site of the formation of nucleus-like structures is mostly in the upper part of the animal hemisphere (see, Fig. 1B). The preferential formation of nucleus-like structures in the animal hemisphere also takes place even when DNAs are injected in the vegetal hemisphere. In such cases, it was frequently observed that the shape of the nucleus-like struc- tures resembled a balloon with a tail coming up from the vegetal part of the egg. Therefore, we assume that in fertilized eggs, there are some mechanisms to confine both zygotic nuclei and artificially formed nucleus-like structures in the animal-most region.

The preferencial formation of nucleus-like struc- tures in the animal hemisphere may at least partly explain why telolecithal eggs of amphibia undergo unequal holoblastic cleavage. Recently, it has been reported that the formation of nuclear en- velope has important effect on the initiation of DNA replication [7, 43]. Therefore, the constant formation of nucleus-like structures in the animal hemisphere may explain the reason for the recent finding of Hofman et al. [33] that exogenously- injected DNA does not replicate in the vegetal hemisphere.

Forbes et al. [21], who first observed the forma- tion of large nucleus-like structures in the cyto- plasm of Xenopus unfertilized egg, demonstrated histochemically that injected lambda DNA itself is assembled into the nucleus-like structures. The nucleus-like structures formed in the cytoplasm of the fertilized egg also appeared to contain the injected DNA, since Hoechst 33258 specifically stained the structures, and furthermore, both biotin-conjugated and *H-thymidine-labeled

DNAs were localized on the nucleus-like struc- tures [69, 74].

In these experiments, especially in the experi- ment with biotin-conjugated DNA, we detected little DNA signal left in the cytoplasm. Therefore, majority of the injected DNA appears to be assem- bled within the nucleus-like structures. If so, it follows that Xenopus egg cytoplasm contains maternal substances necessary for assemblage of the DNA corresponding to about 4,000 somatic nuclei (see also, [21]), since the DNA injected was 24 ng and a Xenopus diploid cell contains 6 pg of nuclear DNA [14].

It has been shown also by Forbes et al. [21] that the nucleus-like structures formed in the unfertil- ized egg contain lamin underneath the “nuclear envelope”. By using monoclonal antibody raised against calf thymus histones, we showed that nu- cleus-like structures contain histones, which must be maternal in nature [74]. Then, it appears that the injected DNAs form nucleosomes before they are assembled in the nucleus-like structures.

Ultrastructural features of the nucleus-like struc- tures

The nucleus-like structures formed by a large amount of exogenously-injected DNA contain homogeneous “nucleoplasm” (Fig. 2A). Inside the “nucleoplasm”, however, patches of matrix structures whose electron density is similar to that of the cytoplasm were frequently observed. Further- more, such matrix structures were seen also at the periphery of the nucleus-like structures, some- times adjacent to the “nuclear envelope” as shown in Figure 2. These matrix structures consist mainly of particulate components which seem to be gly- cogen granules and/or ribosomes. Sometimes, lipid droplets were also found in the nucleus-like structures. Therefore, it appears that a part of the cytoplasm is trapped within the large nucleus-like structures during their rapid assemblage.

It has been shown that the giant nucleus-like structures formed in the cytoplasm of unfertilized eggs are surrounded by double membranes equip- ped with normal-looking nuclear pore complexes [21]. However, more recently, we have shown that the “nuclear envelope” of the nucleus-like struc- tures formed in the cytoplasm of fertilized eggs

Fic.

DNAs Introduced into Xenopus Embryos 201

2 Pie ae Se Nee Be Eek are oe ee s

2. Electron microscopic pictures of the large nucleus-like structures. fertilized egg at 24 ng/egg and nuclear envelope was examined 1 hr after the injection. Normal (arrow heads) and unusual (small arrows) nuclear pore complexes are seen. M, Mitochondrion; L, lipid droplet. The fibrous structure in the “nucleoplasm” (larger arrow in the nucleus-side) is similar to the one seen on a yolk platelet (Y) in the cytoplasm (larger arrow in the cytoplasm-side). (B) Enlarged picture of the central portion of (A). Arrows show unusual pore complexes which are seen only when a relatively large amount of DNA was injected. The fibrous structure serves as a marker to correlate (B) to (A). Bars, 1 wm (A) and 0.1 ~m (B). From Shiokawa et al. [74].

202 M. ASANO AND K. SHIOKAWA

carry not only normal-looking pore complexes but also pore complexes of unusual appearance [74]. Such unusual pore complexes are characterized by the occurrence of the blebbing of the inner leaflet, which protrudes into the “perinuclear space” (Fig. 2).

Szollosi and Szollosi [78] observed a similar blebbing in the nuclear envelope of mouse embryonic cells. They concluded that the blebbing represents active nucleo-cytoplasmic transport of nuclear materials through the pore complexes. In the “pore complexes” of the Xenopus egg, neither enlargement of the blebbing nor its transformation into cytoplasmic vesicles was observed. Therefore, we assume that the blebbing may simply represent the formation of incomplete nuclear pore com- plexes, probably induced by the rapid assemblage of abnormally large nucleus-like structures.

So far, it has not yet been clarified what kind of interrelation exists between the normal nucleus and the large nucleus-like structures formed by the injected DNA. When we observed the formation of large nucleus-like structures, we were unable to find the normal nucleus in the cytoplasm. Since both normal nucleus and the large nucleus-like structures are formed in the animal-most region, we assume that the incorporation of the normal nucleus into the large nucleus-like structures may have taken place during their rapid assemblage stage.

Partition of the nucleus-like structures into daughter cells

Nucleus-like structures formed in fertilized eggs are partitioned into descendant cells during cleav- age [17, 63, 66, 74, 82]. When fertilized eggs are injected with a relatively large amount (24 ng) of either lambda DNA, circular pBR322, or circular pXlIrl01A (a plasmid that contains a Xenopus rDNA single repeat) [2], the pattern of the cleay- age became abnormal. Therefore, the DNA at the dose used is toxic on embryonic development, although there is an experiment in which as much as 50 ng/egg of plasmid DNA was injected, yet no description was made about the toxicity [45].

The abnormal cleavage pattern caused by the exogenous DNA is usually observed in the animal half region. This is probably because the injected

DNAs tend to localize in the animal-most region of the egg. Embryos derived from such DNA- injected fertilized eggs often contain blastomeres of abnormally large size especially in the animal hemisphere. Large blastomeres in such DNA- injected embryos often contain large nuclei whose nucleoplasm shows apparent similarity to that of the large nucleus-like structures formed in the fertilized eggs. When the amount of the injected DNA was 2 ng/egg or less, the damage of cleavage was not observed (see also, [85]). It has been reported that when radioactively-labeled DNA is injected, some of the injected DNA becomes associated with chromosomes, and partitioned into daughter nuclei [17].

PERSISTENCE AND REPLICATION OF INJECTED DNAS WITHIN THE DEVELOPING EMBRYO

To understand the basis for the formation of the nucleus-like structures described above, and also to understand the mechanism of the expression of genetic information within the injected DNAs in the developing embryos, it is necessary to know how the injected DNAs are preserved and repli- cated within the embryonic cells. The behavior of the injected DNAs differs depending upon their molecular forms (circular or linear). We review here mainly the results obtained by Southern blot analysis of the DNAs recovered from the embryos which had been injected with either circular or linearized DNAs.

Changes in the amount and size of circular DNAs injected into the embryo

After the injection of closed circular (c.c.) plas- mid DNAs into fertilized eggs, DNAs were ex- tracted from developing embryos and analyzed using Southern blot hybridization techniques with- out prior treatment with restriction enzymes. Generally, it can be seen that c.c. DNA is im- mediately converted into open circular (0.c.) form, and soon after that c.c. DNA is reformed. The change usually takes place within an hour and after that the levels of c.c. and o.c. DNAs are kept more or less constant throughout later stages, until sometime during the tailbud to tadpole stages,

DNAs Introduced into Xenopus Embryos 203

where they disappear.

These changes can be seen in our experiments in which we injected a relatively small amount (0.2 ng/egg) of circular forms of pBR322 that carries either an rDNA single repeat (pXIr101A) (15.7 Kb), a Drosophila copia gene (cDm2055) (10.3 Kb), a P-element (p725.1) (9.1 Kb), or nothing (pBR322; 4.3 Kb) (Fig. 3). In this experiment, both c.c. and o.c. DNAs which were present in embryos an hour after injection were maintained

(<b) w 5 io] 12} Qa = — %) ee S =>) Sts dein OG, Se =o =. oO ro OPS iss a F aaw) eaebo ee oe = Rom On ae = ret 20PS2 145, 5:16) i ic8 A Wy OC- et Om ee. 4... 2 C C- jm = iy 9.4— vin pXIrlOlA

Des O C— ae seme aioe sem 94 =e exes CC— o a ow 8 ee oe ge 4.4 —

pir 25.1

Fic. 3.

at constant levels, and eliminated at the tailbud stage (in the case of larger plasmids, pXIr101A and cDm2055) or slightly later at the muscular re- sponse stage (in the case of smaller plasmids, p725.1 and pBR322) [73] (Fig. 3). These results show that the injected circular DNAs did not replicate appreciablly throughout the stages. It was also consistently shown among the four plas- mids studied that when these circular DNAs dis- appear from the embryos, o.c. DNAs do so faster

3) wo c a S 2 o

3 = = re o OD 3 - SoS ©) & @ oS 3 = v= =e eae tice pa -~-+-_- PH & wo A O YO fT SS SS] @ Pa Sh OPC) (ype ~pnouwZztre 3 14 15 16

109) Te) fe) XS)

9.4 —

CC MP Se me ee cree te ce

cDm2055

D 25 26 27 28 29 3031 32

9.4—

3,6 =

44—

CC 4 Gn Gb an ap an Gm ~~ (Ace ae

pBR322

Southern blot analysis of DNAs from plasmid-injected embryos. Fertilized eggs were injected with 0.2 ng/

egg of circular pXlr101A (A), cDm2055 (B), pz25.1 (C), and pBR322 (D). Embryos were harvested at the indicated stages and subjected to Southern blot analysis without prior enzyme digestion. The positions of open

circular (o.c.) DNA, closed circular (c.c.) DNA, and size markers are given.

From Shiokawa et al. [73].

204 M. ASANO AND K. SHIOKAWA

than c.c. DNAs (Fig. 3). In Figure 3, it is also apparent that most of the injected circular genes are converted to o.c. DNA before they disappear. In the case of the injection of circular form of vitellogenin genes, however, it has been reported that some of the injected DNA are preserved in the nuclei probably as minigenes even after meta- morphosis [1].

When circular plasmid DNAs contain viral re- plication origin in their promoters, they replicate actively after introduction into the embryos. Thus, circular pSV2CAT, which is the bacterial chloram- phenicol acetyltransferase (CAT) gene connected to SV40 early promoter (containing viral replica- tion origin), and circular pAd12E1aCAT, which is a circular CAT gene connected to the promoter (that also contains the replication origin) of type 12 adenovirus Ela protein were replicated in the injected embryos quite actively, especially in later stages of development (Fig. 4). In this case, it is apparently shown that only c.c. DNA disappeared at 20 min after injection (Fig. 4, lane 2), then c.c.

o ge = Sa oO o o Ww % ie) co fe SoC ones = = 2 fy FR ® _— > ss = = fe) on SoS eee yas Oo & io) (oe) S iS nN O © — O c et ao) i _ 12) — — -0.c.DNA E \ -c.c.DNA | 2.43 4.5 <6 circular pSV2CAT Fic. 4. Southern blot analysis of DNAs from embryos

injected with circular pSV2CAT. Fertilized eggs were injected with 1 ng/egg of circular pSV2CAT and DNAs were extracted at the indicated stages and processed as in Figure 3. Zero time and cleav- age samples were obtained actually at 20 min and at 3 hr, respectively, after the injection. From Fu et al. [23].

DNA reappeared 3 hr after injection (lane 3), then both c.c. and o.c. DNAs increased in their amounts through early blastula to gastrula stages.

In the experiments shown in Figure 3, there were only very small amounts of DNAs which migrated slower than o.c. DNA. However, in the experiment shown in Figure 4, a considerable amount of large-sized DNAs which migrated slower than o.c. DNA appeared in later stages.

We treated these large-sized DNAs with type II topoisomerase, under the conditions in which kinetoplast DNA, as an authentic internal control for catenated DNA, was completely decatenated to the circular DNA of monomer size. However, the amount and size of the large-sized DNA did not change to any appreciable extent. These results show that the large-sized DNAs formed are not catenated DNA.

When we treated the large-sized DNAs with various restriction enzymes, linear DNA of appro- ximately monomer size was obtained in all cases. Then, the large-sized DNA may be either circular form of oligomer of injected DNA or concatemers formed by ligation of DNAs which were linearized after being introduced into the egg.

Factors that affect replication of circular DNAs in developing embryos

Figures 3 and 4 show that the extent of the replication of circular DNA varies considerably depending on the constitution of the plasmid. The general rule may be that the plasmids like pSV2CAT and pAd12ElaCAT which carry viral promoter with replication origin on it replicate actively, whereas other plasmids like pBR322, pXIrl01A, cDm2055, and pz25.1 which do not contain such promoter replicate poorly. Here, it must be noted that, sometimes, especially when the analysis was carried out in early stages of development, even the circular pSV2CAT that carries viral replication origin replicates only poorly (2 to 3-fold). Also, plasmids without eukaryotic replication origin which in most case do not repli- cate can replicate considerably (at most ca. 2-fold) sometimes in embryos at later stages [18, 73]. Thus, it is still not known why injected circular DNAs sometimes replicate and sometimes not.

It has been reported that supercoiled plasmids

DNAs Introduced into Xenopus Embryos 205

replicate but nicked circular plasmids do not [15]. One report proposes that only a certain fraction of circular DNA can replicate probably by being localized in a special compartment in the cell which is particularly suited for replication [46, 47]. In another paper, it is reported that plasmid DNAs which are injected into the animal half of the fertilized egg, 60-65 min after fertilization, repli- cate preferentially [33].

In unfertilized egg it has been well established that eukaryotic replication origins are unnecessary for the efficient replication of exogenously- injected DNAs [30, 49]. However, as stated above, the replication is significantly greater in the presence of such origin (pSV2CAT) than in their absence (pBR322, pSVOCAT and others) [23, 73]. Therefore, in developing embryos, the presence of replication origin within the injected circular DNAs may have significant stimulating effects on their replication.

We assume that the presence of the replication origin within the circular DNAs makes little differ- ence in their replication in the cytoplasm of unfer- tilized egg, where the relative amount of DNA polymerase per DNA is extremely high. By con- trast, the presence of the replication origin may have significant stimulating effects on the replica- tion of the circular DNAs in the cytoplasm of embryos, especially in later stages, where the amount of DNA polymerase per cell is very similar to that of the adult-type cells.

Changes in the amount and size of linearized DNAs injected into the embryo

In developing embryos, a great contrast was seen between the behavior of circular DNAs and linearized DNAs: While circular DNAs sometimes replicated but most of time not as already shown in the above section, all the linearized DNAs so far tested (lambda DNA, linearized pSV2CAT, pSVOCAT, and actin-CAT fusion gene) replicated quite actively (by about 50 to 100-fold). Furth- ermore, the size of the injected linear DNA was increased by 10 to 20-fold. Results in Figure 5 provides a typical example showing such differ- ences. We injected here circular and linearized plasmid DNA known as actin-CAT fusion gene #254 (containing the CAT gene and the promoter

a = 6 2 oye av avo Tn ©) an 9 Gh ce ONS ® = ED pe = 2 Ss Aish ei = Ein ae - 2Fa5a5 € Ss 25 5 i = a= mH = Ps “ oo 2 ar ~ oD —_ o— o5 $s Oo. o) Ss NO©O2ZS Ww N ©O222%0 | 2.8 48 Caco lO - " + ——-S for —oc — —lin

SSS Circular Linear

Fic. 5. Southern blot analysis of DNAs extracted from embryos injected with circular or linearized actin- CAT fusion gene. Fertilized eggs were injected with 2 ng of circular or linearized actin-CAT fusion gene #254, and 5 embryos each were harvested at indi- cated stage and processed as in Figure 3. Zero time (15 min after injection). o.c. (open circular), lin (linear), and c.c. (closed circular) DNA. In lanes 7- 9 faint signals are seen at about the position of c.c. DNA, and in lane 6 another signal is seen at the position close to o.c. DNA. However, the nature of

these have not been examined. From Shiokawa et al. [62].

of Xenopus a-actin gene) into fertilized eggs. When linear DNA was injected, a large amount of large-sized DNAs was formed at the blastula and neurula stages of development (Fig. 5, lanes 7 and 8). It is seen here that some DNA is stuck on the gel slots (lanes 7 and 8), probably because of the largeness in the size. However, later at the muscu- lar response stage (lane 9), the majority of the large-sized DNA disappeared [62]. These results are in a remarkable contrast with the practically constant preservation of the circular form of the

206 M. ASANO AND K. SHIOKAWA

fusion gene in the control experiment (Fig. 5, lanes 2 to 5) (note that, c.c. DNA disappeared shortly after injection as in lane 1, but reappeared soon as in lane 2).

We treated the large-sized DNAs obtained by injection of linearized DNAs with type II topoisomerase, under the conditions in which kinetoplast DNA as an internal reference for catenated DNA was completely decatenated to the circular DNA of monomer size. However, we found here again practically no change in the amount and size of the large-sized DNAs before and after the treatment. Therefore, the large-sized DNAs produced by the injection of linearized DNAs are not catenated DNAs.

It was found that such large-sized DNAs were not sensitive to the restriction enzymes which were used to linearize the plasmid DNAs before they were injected. Thus, the large-sized DNA ex- tracted from embryos that had been injected with Acc I-digest of pSV2CAT was not digested with Acc I. Likewise, large-sized DNAs extracted from embryos that had been injected with Hpa I-digest of pSV2CAT, were not digested by Hpa I. However, almost complete conversion of these large-sized DNAs to the DNA of approximately monomer size was seen when they were digested with restriction enzymes that were not used to linearize the injected DNA.

These results strongly suggest that the large- sized DNAs obtained by the injection of linearized DNA are concatemers, formed by ligation of prob- ably all the possible combinations (head to head, head to tail, and tail to tail), as reported previously [3, 4, 60]. The resistance of the large-sized DNAs to digestion with the enzyme which has been used to linearize the injected DNA suggests that distal parts of the injected DNAs are modified probably by exonucleolytic digestion prior to ligation.

Possibility of the integration of the injected linea- rized DNAs into the host genome

At the beginning of our DNA injection experi- ment, we injected circular pXIr101A into the fertil- ized egg of wild-type Xenopus laevis, as one of the preliminary experiments to try to rescue anuc- leolate mutant Xenopus embryos which do not synthesize rRNA throughout development. In this

pSV2CAT wo circular 5s 2OMNOWO a Oo- —-N WM

1234 56

linear

Days SOD

CAT gene fragment

o linear >mNmnNMnnMOM © OC oO 53 oO _ AM TH CO YE =] Mw a

16 17 18 19 2021 22 23

*

Fic. 6. Southern blot analysis of DNAs obtained from embryos which were cultured for a long period after injection of circular or linearized DNAs. Fertilized eggs were injected with lng each of circular pSV2CAT (top), Bal I digest of pSV2CAT (mid- dle), or CAT gene fragment (bottom), and embryos were harvested at the indicated dates and processed as in Figure 3. From Fu ef al. [23].

DNAs Introduced into Xenopus Embryos 207

experiment, from the 2 week-old tadpoles, we obtained the evidence that the injected pXIrl01A was preserved as chromosome-integrated DNA [80].

However, later, it became clear that linearized DNAs are not only replicated more efficiently [23, 46] but also integrated more frequently into the host genome [17, 19, 23]. Then, we started to inject instead of circular DNA, linearized DNAs, and found in the DNA-injected embryos the formation of a large amount of large-sized DNA that co-migrated with host DNA. In most of such experiments, majority of the large-sized DNA formed at early stages (blastula to neurula stages) always disappeared after about 2 days (during early swimming tadpole stage). However, a small but distinct amount of large-sized DNA was pre- served even in 1-month-old tadpoles.

Interestingly, the amount of such long-lasting large-sized DNA varied from preparation to pre- paration. Figure 6 gives typical results showing the presence of long-lasting DNAs in 1 month-old swimming tadpole, which developed from fertil- ized eggs injected with either Bal I-digest of pSV2CAT (middle panel) or CAT gene fragment (bottom panel) [23]. The results at the top indicate the DNAs which were recovered from embryos injected with circular pSV2CAT in which we see practically no large-sized DNA. Thus, the long- lasting DNA shown in these 1 month-old embryos (middle and bottom panels of Fig. 6) may be the strong candidates for the chromosome-integrated DNAs.

In Xenopus laevis, however, studies which unequivocally demonstrated the integration of the injected gene in host genome are so far very scanty. Only Etkin and Pearman [17] are excep- tions, who injected linearized pSV2CAT into fertil- ized eggs and found that injected DNA persisted in the tissues of 60% of 6- to 8-month-old frogs, exhibiting a mosaic pattern of distribution (2 to 30 copies/cell) among different tissues. Furthermore, they showed that such exogenous DNA sequences were transmitted to the genome of various tissues of offspring through the male germ line.

EXPRESSION OF EXOGENOUSLY-INJECTED DNAS IN THE DEVELOPING EMBRYOS

Expression of exogenous genes in developing embryos has been extensively studied. However, majority of the papers published so far are con- cerned with the functional characteristics of spe- cific gene cloned by each of the research groups. In addition, experiments are relatively scanty that aimed to elucidate the general control mechanism which regulates the expression of exogenously- injected, as opposed to endogenous (or chromo- somal), genes. Therefore, we emphasize here mainly the general aspect of the control mechan- isms applicable to various genes introduced into the developing embryo.

Expression of exogenously-introduced ribosomal DNA

One of the most abundant gene products in the early amphibian embryonic cells at RNA level is no doubt rRNA, and naturally, one of the earliest data on the genes introduced into Xenopus embryos was concerned with the expression of Xenopus tDNA-containing plasmid, pXIrl01A. As described above, we have shown that rRNA synthesis from endogenous genes starts at the late blastula stage [65, 68], and furthermore, presented evidence that the initiation of the rRNA synthesis may be regulated at the transcriptional level by decrease at the blastula stage of the intracellular ammonium ion or pH [61, 64].

Shortly after our findings [65, 68], Busby and Reeder [12] injected pXIr101A, a plasmid which contains Xenopus laevis rDNA single repeat, into fertilized eggs of Xenopus borealis. This “hybrid” combination of X. laevis rDNA and X. borealis embryo was constructed, because it has been pre- viously shown that the expression of the X. laevis rDNA is dominant over that of the X. borealis rDNA [34], and it is much easier to detect X. laevis new transcript in the presence of a large amount of X. borealis maternal rRNA than in the presence of a large amount of X. /aevis maternal rRNA (4,000 ng/egg). They then tested the effect of the pre- sence of varying lengths of 5’-upstream sequences on the efficiency of the expression of the injected X. laevis rDNA, using X. laevis specific S1-

208 M. ASANO AND K. SHIOKAWA

protection analysis.

Busby and Reeder [12] have shown that the injected Xenopus laevis rDNA is expressed first at the late blastula stage and the efficiency of the expression depends on the lengths of the S’- upstream sequence [12]. Then, it follows that in developing embryos, the expression of exogenous- ly-injected rDNA begins at the late blastula stage [12], concurrently with the expression of endog- enous rDNA [61, 64, 68]. These results suggest that the control mechanism which is intrinsic for the expression of endogenous (or chromosomal) tDNA may also operate on the exogenously- introduced rDNA.

Expression of circular CAT genes which carry viral promoters

In order to obtain the basic data about how

100

50

Acetylated chloramphenicol (%)

protein-coding exogenous genes are expressed in Xenopus embryos, we first tested the expression of five different circular plasmid DNAs that carry the CAT gene connected to different viral promoters. Plasmids used here were pSV2CAT containing a CAT gene linked to a relatively strong promoter of SV40 early gene [25], pAIOCAT2 including a part of the SV40 early promoter (the GC-rich 2]l-bp repeats and TATA box) but without enhancer sequences (two tandemly repeated 72-bp ele- ments) [59], pAdl2.ElaCAT carrying a relatively strong promoter of Ela protein of type 12 adeno- virus [40], pSVOCAT containing no viral promoter [25], and pAIOCAT3m which is a polylinker- containing derivative of pSVOCAT.

We injected a reportedly non-toxic dosage (1 ng/egg) [85] of these plasmid DNAs into the fertilized eggs, and tested the CAT enzyme activity

Sm -_-—- -_—- = =

O pSv2caT A pAdl2.ElaCAT @ pAIOCAT2 A pSVOCAT

O pAIOCAT3m

— = = — ee

20 25

O 5 10 IS

Hours after fertilization

Fic. 7. Kinetics of the accumulation of CAT enzyme activity during development. Following injection of 1 ng of circular plasmids into fertilized eggs, embryos were collected at the blastula, gastrula, and neurula stages. The relative strength of the CAT signal was determined by densitometry and plotted as a function of the time after fertilization. From Fu et al. [23].

DNAs Introduced into Xenopus Embryos 209

by autoradiographic exposure of CAT assay sam- ples for 2 weeks. As shown in the summary of the densitometric calibration of these data (Fig. 7), pSV2CAT which contains the strong promoter of SV40 early gene was expressed at as early as the blastula stage. In this case, the CAT enzyme level showed a great increase during the gastrula stage, and then reached a plateau at the late neurula stage. Very interestingly, an extrapolation of the level of the CAT enzyme activity from the gastrula to blastula stages gives us an impression that there might be some CAT enzyme expression even before the blastula stage (we shall discuss this point in detail later on). pAdl2.ElaCAT which also contains a relatively strong promoter was not expressed at a detectable level at the blastula stage, but was expressed at a high level at the gastrula stage and then reached a plateau at the late neurula stage. By contrast, pSVOCAT and pAIOCAT3m which contain no promoter, and in addition, pAIOCAT2 which contains the enhancer- lacking SV40 promoter were expressed only at a background level throughout the stage.

By Northern blot analysis using antisense RNA as a probe, we confirmed that CAT enzyme ex- pression was accompanied by the appearance of 1.6Kb CAT mRNA. Therefore, in the Xenopus embryo system, CAT enzyme level appeared to reflect the transcriptional activity of CAT gene. Therefore, we concluded that the efficiency of the expression of circular DNAs depends mainly on the relative strength of the promoter. For the small amount of background level expression of CAT enzyme activity from promoter-less plas- mids, we assume that the cryptic promoters within the plasmid sequence may be responsible for it (cf.

[42]).

Expression of linearized CAT genes which carry or do not carry viral promoters

Having selected pSV2CAT and pSVOCAT from the five plasmids used above, we produced variously-linearized forms of these plasmids as shown in Figure 8. We then injected these linea- rized plasmids at 1 ng/egg into the fertilized eggs, and tested the CAT enzyme expression at various stages of development.

Figure 9 indicate that all of the linearized DNA

pSV2CAT -= Bad | im Acc | ——§ ff) [ca] Bam HI Sea el Hpa | -——_ i iia _}—_—- Pst a }—)— Bal | —$§— a a—__+— stul ——_ i _}+—__ Bam Hi+Stu |. -————#2="} - —[_}¥—- Bam Hi+Ace| -————1-#4: }- —_})——— pSVOCAT Hae Il SSS SL Hpa | a __y Pst | ——_—_)——__{iiéa# {_} -+1- Isolated CAT fragment BamHl+ Hind Ill mz Fic. 8. Plasmids used in the experiment. pSV2CAT

and pSVOCAT were digested with restriction en- zymes as listed. In circular plasmids, the straight and bent arrows represent gene region for ampicillin and replication origin of pBR322, respectively. Heavily dotted box, enhancer region; lightly-dotted box, promoter region; closed box, CAT coding sequence; open box, poly(A) addition site. Arrow- heads show sites of digestion by restriction enzymes indicated. From Fu et al.[23].

preparations of pSV2CAT (lanes 6 to 14) were expressed at the levels which were more or less similar to that of circular pSV2CAT (lanes 15 to 17). Furthermore, unexpectedly, all the linearized DNAs of pSVOCAT which did not carry any promoter (Fig. 8) were expressed (not shown in Fig. 9) at levels which were comparable to that of circular pSV2CAT. In addition, a linear CAT coding sequence, which had been prepared from pSVOCAT by double enzyme digestions (Bam HI and Hind III) and freed from the vector sequence prior to the injection (Fig. 8), was expressed at a considerably high level at the gastrula and neurula stages (lanes 4 and 5), albeit the level is lower than that of circular pSV2CAT at the blastula stage

210 M. ASANO AND K. SHIOKAWA

Circular

pSVOCAT CAT gene fragment

Linearized pSV2CAT

Circular pSV2CAT

[ell || II |

ae - Acs

AC3- wt - AC2 @- ACI AC2 - <= - “secs -ee@ ACI - —

C -@2@@aeo - @@eae-

12345 67 8 9 IO ll 12 13 14 15 1617 BGBGNBGNBGNBGNBGN Let Seal) jl ail pg) line meres] Bal | Stul BamHI, Stu |

Fic. 9.

CAT enzyme assay in embryos following injection of CAT genes. Circular pSVOCAT (lanes 1 and 2), CAT

gene fragment (lanes 3-5), linearized pSV2CAT (lanes 6-14), and circular pSV2CAT (15-17) were injected at 1 ng/egg into fertilized eggs. Embryos were collected at the blastula (B), gastrula (G) and neurula (N) stages and used for CAT enzyme assay. The restriction enzymes used were Bal I (lanes 6-8), Stu I (lanes 9-11), Bam HI

plus Stu I (lanes 12-14).

C, AC1, AC2, and AC3 are chloramphenicol, l-acetylated chloramphenicol,

3-acetylated chloramphenicol, 1,3-diacetylated chloramphenicol, respectively. From Fu ef al. [23].

(lanes 15 to 17). Thus, the expression of linearized pSVOCAT (not shown) and CAT gene fragment isolated from pSVOCAT (lanes 4 to 5) makes a great contrast to that of circular pSVOCAT where the expression was almost negligible (lanes 1 and 2).

These results led us to conclude that linearized CAT genes are expressed in embryos by some mechanism which is independent from the promo- ter. For a DNA to be active as a template, it is necessary for it to have supercoiled domain, and to generate the torsional stress [31, 56]. A possible explanation for the expression of linearized DNA in embryos may be the attachment of the large- sized DNA to some cellular structures such as

nuclear matrix.

It may be worth noting here that pSVOCAT digested by Hpa I at the poly(A) addition site and pSV2CAT digested by Bal I within the CAT coding sequence (Fig. 8) were expressed at con- siderably high levels. A possible explanation for these may be that these broken genes have been re-circularized and thus expressed from the recon- stituted coding sequence. However, this explana- tion is only applicable to the case of the injection of Bal I-digest of pSV2CAT, but not to that of the Hpa I digest of pSVOCAT, since circular pSVOCAT is expressed only at a background level (cf. Fig. 7). Therefore, more likely explanation may be that the genes have been ligated to form

DNAs Introduced into Xenopus Embryos 211

concatemers in every possible combination, and some of the concatemers which happened to re- constitute functionally active CAT coding sequ- ence were expressed.

Expression of the injected circular CAT genes with viral promoters before the midblastula transition (MBT)

Newport and Kirschner [52, 53] proposed that the new gene expression from zygotic genome in Xenopus embryogenesis takes place only at and after the 12th cleavage (at and after the timing called MBT). In order to obtain the evidence to support this idea, these authors injected a plasmid containing an yeast leucine tRNA gene (pYLT20) into the Xenopus coenocytic egg cell, a fertilized egg whose cleavage has been inhibited by low speed centrifugation. They found that although the tRNA genes were transiently expressed shortly after the injection, the genes became inactive during the cleavage stage, and then re-activated at the 12th cleavage.

Three years later, Etkin and Balcells [16] in- jected 1 ng/egg (10’ copies/egg) of pSV2CAT into Xenopus fertilized eggs to further confirm the idea of MBT-associated activation of exogenously- injected genes. They reported that injected CAT genes are expressed only during and after the 12th cleavage (MBT). Thus, the MBT theory which maintains that there is no expression of nuclear and exogenously-injected DNAs during the cleav- age stage has been much strengthened.

On the contrary, we have shown that even in cleavage stage embryos there is the synthesis of mRNA-like heterogeneous RNA, which is not due to mitochondrial DNA and is sensitive to acti- nomycin D and a-amanitin [51, 61, 64]. Therefore, we expected that the exogenously-injected DNAs will be actually expressed, and therefore, their expression may be detected during the cleavage stage if sufficiently sensitive methods were applied.

It may be worth pointing out here the difficulty involved in the determination of the gene activity in early Xenopus embryos, in which a rapid in- crease in cell number (or nuclear number) is taking place. For instance, in the experiment of Etkin and Balcells [16] described above, CAT enzyme activity was compared using the samples obtained

from the same number of embryos (15 embryos) at all the stages tested. As shown in the above section, DNAs injected into the fertilized eggs are preserved throughout the development, and in this sense, the amount of the injected DNA-templates per embryo is approximately constant throughout their development.

However, under the conditions used, it is not known what percentage of the injected DNA is actually available as templates for transcriptional machinery at each stage of the development. In fact, there is a possibility that only a limited portion of the injected DNA is available as the actual templates at the beginning and the amount of the effective template increases as development proceeds. Then, it would be difficult to detect the expression of the gene in the embryos at early stages using the methods which are sensitive enough to detect the expression of the gene in the embryo at the later stages.

As a method to increase the sensitivity in detect- ing the gene expression in embryos at the early stages of development, we modified the experi- ment of Etkin and Balcells [16] by increasing the amount of circular pSV2CAT by 10-fold (10 ng/ egg; 10° copies/egg) (the number of embryo used was still 10 embryos at all stages) [72]. Under these conditions, development of the DNA- injected eggs was slightly abnormal, especially when embryos reached the neurula stage, although nevertheless, they continued development beyond the neurula stage.

When CAT enzyme assay was carried out using these embryos and autoradiography for the detec- tion of acetylated chloramphenicol was performed for 2 weeks, data similar to those of Etkin and Balcells [16] were obtained; the CAT enzyme signal started to appear at the MBT, and increased thereafter until the late neurula stage. However, when we elongated the autoradiographic exposure time to 2 months, we detected CAT enzyme activity not only at the MBT stage but also at 3 hr (early cleavage stage) and Shr (late cleavage stage) after the DNA injection (Fig. 10, left). When measured on a densitometer, the increase in the CAT enzyme signal roughly paralleled to that in cell number per embryo, suggesting the necessi- ty of the increase in the number of nucleus for the

212 M. ASANO AND K. SHIOKAWA

|2h(early gastrula)

3 i=) io) => =) @® oO > TS ie) cS) & ™

5h(late cleavage) 8h(midblastula)

[vous wnkgonne4 pSV2CAT

I2h(early gastrula)

—_= D) S) S) > re) @ ) Pa) = iS) oo) =—_ i oP)

2 S — =) © e re Oo N

Sh(midblastula)

-AC3

-AC2

-AC|

ee: DL WE mwesyine US actin-CAT #254

lOng/eagg, |Oeggs/lane

Fic. 10. CAT enzyme assays using embryos injected with circular CAT genes. Fertilized eggs were injected with 10 ng of either circular pSV2CAT (lanes 1 to 4) or actin-CAT fusion gene +254 (lanes 5 to 8), and 10 embryos each were harvested at the indicated stages for CAT enzyme assay as in Figure 9. Autoradiography was carried out for 2 months. A faint signal which appeared between ACI and AC2 is not due to CAT enzyme expression from the injected genes because it was obtained even in uninjected embryo after long autoradiographic exposure. From

Shiokawa et al. [62, 72].

increased CAT gene expression.

In the above experiment, we carried out a control experiment using the actin-CAT fusion gene +254 which was also used in the experiment in Figure 5 [50]. Since this CAT gene is connected to the promoter of Xenopus a-actin gene and is

known to be expressed only after the embryo reached the neurula stage [50], we thought it would be interesting to see how this developmen- tally-controlled gene is expressed when the gene is injected into the embryo at a high dose.

Then, we injected the fusion gene #254 either

DNAs Introduced into Xenopus Embryos 213

at 1ng/egg or at 10ng/egg, and assayed CAT enzyme activity using the samples extracted from 10 injected embryos at four different stages. The results obtained after 2 months of autoradio- graphic exposure showed that CAT enzyme signal was detected not at the cleavage but at the neurula stage at the two different dosage (Fig. 10, right; only the results with 10 ng/egg are shown). These results indicate that injected fusion gene was ex- pressed in a temporally controlled manner just as its endogenous counterpart. Therefore, it appears that the results obtained using 10 ng/egg of pSV2CAT are not just artifact produced by some toxic effects of the excessive amount (10 ng) of the DNA injected.

However, we thought it may still be possible to argue that the expression of CAT gene during the cleavage stage (pre-MBT stage) (Fig. 10, left) might be due to “aberrant expression” which took place only for the high dose (10ng/egg) of pSV2CAT for some unknown pSV2CAT-specific reason. Therefore, we repeated the experiment using the dosage (1 ng/egg; 10’ copies/egg) which is exactly the same as that used by Etkin and Balcells [16]. In doing this experiment, however, we altered the other parameters: We used 10-fold larger number of embryos (100 embryos), and actually, instead of preparing one sample (consist- ing of 10 embryos), we prepared 5 samples (each consisting of 20 embryos) for two time-points (4 hr; cleavage stage, and 8 hr; blastula stage). Then, the 20 embryo samples were processed separately and at the final step of chromatography, all the samples (100 embryos) were analyzed on one spot.

Even by conducting such a large scale experi- ment, we were unable to detect CAT enzyme activity during the cleavage stage, when auto- radiographic exposure was carried out for 2 weeks. Therefore, again, we extended the exposure time to 2 months. Then, as shown in Figure 11, we were able to detect the presence of the clear signal of acetylated chloramphenicol for the cleavage stage embryos (4 hr after fertilization). In this experi- ment, plasmid constitutions without viral promoter (pSVOCAT: pSVOOCAT) were not active, and pAd12ElaCAT gave a signal significantly weaker than that of pSV2CAT. Therefore, also during the cleavage stage, CAT genes with viral promoters

8h(midbiastula) ~ enzyme control (I unit)

wo m 5 > oO () 1S) — <= t+

| 2 5) Ing/egg, lOOeggs/lane

Fic. 11. CAT enzyme assay in embryos from the cleav- age stage embryos. Fertilized eggs were injected with 1 ng of circular pSV2CAT and harvested at the indicated stages. Twenty embryos were pooled as a group, and five such groups of embryos were pre- pared for each sample. Samples were processed separatedly as in Figure 9, but finally pooled and assayed together as one sample at the step of thin-layer chromatography. Thus, each spot is for the sample of 100 injected-embryos. From Shioka- wa et al. (72).

can be expressed according to the strength of the promoter.

Controlled expression of the exogenous DNAs which carry promoters of Xenopus mRNA-coding genes

It has been shown that the circular DNAs in- jected into Xenopus embryos are transcribed at

214

correct initiation sites [3, 5, 60]. It has also been reported that the plasmids which carry the promo- ters for house-keeping functions such as Xenopus histone H1 [85] and mouse histone H1° [75] are expressed in almost all cell types. Whereas many genes of so-called luxury function, such as a-actin genes [50, 76, 85], cytoskeletal actin gene [8], and epidermal keratin gene [36] have been reported to be expressed during development not only at cor- rect timing but also at correct embryonic regions (an example for the correct temporal expression for CAT gene connected to a-actin gene promoter is already given in the right panel of Fig. 10). There have been papers which reported the presence of a promoter whose function is to deter- mine the timing of the expression of the gene

— Drosophila ™ Uninjected neurula * Cleavage(3hr)

Ww Zero time o Blastula

® Gastrula

[ ~ Neurula

pRRamy

Fic. 12.

3% polyacrylamide gel.

M. ASANO AND K. SHIOKAWA

during development. One typical example is the promoter of the temporally-controlled gene, GS17 (gastrula-specific gene 17), whose essential part consists of only 74 base pairs localted at 682 bp upstream of its transcription initiation site [39]. Another paper of a special interest is the one which suggested the presence of two different regulatory elements in the promoter of a-actin gene of Xenopus laevis skeletal muscle: One for cell-type specific and the other for stage-specific expression of the gene within the embryo [76]. The elucidation of the structure and function of these control elements in developmentally impor- tant promoters would help us understand the mechanism of the gene actions in embryogenesis.

@ Zero time © Cleavage © Blastula Gastrula Neurula

Amy 3.8

Expression of Drosophila amylase in Xenopus early embryos. injected with 1 ng/egg of circular pRRamy 1.00 (lanes 3-7) or 2 ng/egg of circular Amy 3.8 (lanes 8-12). Five embryos were harvested at the indicated stages, and homogenized and assayed for amylase by electrophoresis on

Fertilized eggs of Xenopus laevis were

Lane 1 shows the authentic amylase extracted from a fly (Drosophila melanogaster,

Oregon R). Lane 2 shows the sample from control 5 neurulae which were not injected with the plasmid. From

Shiokawa et al. [73].

DNAs Introduced into Xenopus Embryos 215

Controlled expression in Xenopus embryos of ex- ogenously-injected protein-coding genes from other sources

There have been several papers which report the results of the injection into fertilized Xenopus eggs of protein-coding genes derived from materials other than Xenopus. For example, we injected Drosophila amylase genes (pRRamy 1.00 and Amy 3.8; both code for the same Drosophila enzyme amylase, but differ only in the franking sequences) into fertilized Xenopus eggs and found that they are expressed at the gastrula and neurula stages (Fig. 12) [73]. The amount of the Amy 3.8 injected was twice that of pRRamy, and the extent of the expression was approximately proportionate to the amount of the injected genes. The forma- tion of functional enzyme whose electrophoretic mobility is exactly the same as that of the fly indicates the correct transcription, processing, and translation of the injected genes as in the Dro- sophila cells. Thus, genes from other sources can also be expressed in a regulated way in Xenopus embryos.

In some cases, it has been shown that such regulation takes place in a promoter-dependent manner. The chimeric CAT gene carrying the promoter of Drosophila hsp (heat-shock-protein) gene has been shown to be heat-inducible after its introduction into Xenopus embryos [41]. Also, the gene for the rat glucose-regulated-protein which is known to be inducible by tunicamycin has been shown to be induced by tunicamycin treatment after its introduction into Xenopus embryos [86]. These experiments indicate that the promoters of genes from sources other than Xenopus can be recognized in Xenopus embryonic cells in specific ways.

Temporally-uncontrolled expression of linearized genes with the promoter which was originally under developmental control

In spite of the considerable amount of data available on promoter-dependent control of gene expression, little attention has been paid to the difference in the mode of expression of injected DNAs in relation to their structures (circular or linear). For instance, in the DNA injection experi-

ments by Etkin and Balcells [16] described above, no clear difference has been detected in CAT enzyme expression between the embryos injected with circular pSV2CAT and the embryos injected with linearized pSV2CAT, even though they found that both forms are expressed. In the experiments by Mohun ef al. [50] and Wilson et al. [85] which studied expression of the genes that contained a-actin promoters, the different modes of the expression of circular and linearized DNAs have not yet been recognized.

We have shown above that the expression of exogenously-introduced CAT genes greatly varies depending on whether they are circular or linear.

(<b) oO S. i.e) oO st 2Zzoss 5 wh -~¥ Oo W = ORO a Or ma tora" TAGS an “= Gp -C2 ~ mm -Cl . = SO GP Gp wae = 7, * © © ¢ 9 @ oe e a

Pale 2 eS hae 52 765 Circular Linear

actin-CAT gene

Fic. 13. Earlier expression of linearized actin-CAT fu- sion gene than circular gene during Xenopus de- velopment. Ten fertilized eggs were injected with 2 ng of either circular or linearized actin-CAT fusion gene and harvested at the indicated stages. CAT enzyme assay was as in Figure 9. From Shiokawa et al. [62].

216 M. ASANO AND K. SHIOKAWA

Therefore, we examined if the expression of CAT gene connected to the developmentally-controlled Xenopus cardiac a-actin promoter (actin-CAT fu- sion gene #254) [50] also varies depending on whether the gene is circular or linearized.

As shown in Figure 10 (right) and in Figure 13 (lanes 1 to 3), the circular form of the fusion gene +254 is expressed only after embryos reached the early neurula stage. However, when we linearized the gene in the vector sequence and injected it into fertilized eggs at 1 ng/egg, the expression of the injected DNA took place as early as at the blastula stage (Fig. 13, lane 5) [62]. These results imply that when linearized, the expression of the fusion gene deviates from the normal temporal control even though the DNA possesses the intact pro- moter.

The changes in the amount and size of the linearized actin-CAT fusion gene injected here is as shown in Figure 5: while circular gene showed little change in its amount and size, linearized gene was ligated and replicated by about 10-fold at the blastula stage, and about 100-fold at the neurula stage. However, the increase in the amount of the injected DNA template at the blastula stage (by about 10-fold) does not by itself explain why the linearized DNA was expressed at the blastula stage, because the injection of 10 ng/egg of circu- lar fusion gene did not induce the earlier express- ion of the gene (Fig. 10, right). Therefore, we assume that the temporally-uncontrolled expres- sion of linearized actin-CAT fusion gene in the early embryos must be due to some structural features of the linearized DNA. Probably, the formation of large-sized concatemers would be the reason for their uncontrolled expression, as has been expected from the experiments with linear- ized pSVOCAT and isolated CAT gene fragment shown in Figure 9 [23].

EFFECTS OF DNA INJECTION ON GENE EXPRESSION IN COENOCYTIC EGG CELLS AND DEVELOPING EMBRYOS

Active cell division takes place in a fertilized Xenopus laevis egg and when the cell number is increased to about 4,000 (i.e. blastula stage), the formative movement starts within the embryo.

During the process, the content of DNA per embryo increases very rapidly. We have previous- ly carried out experiments to alter the ratio of the DNA to the cytoplasm either by arresting the cell division using colchicine and cytochalasin B [79], or by extracting the cytoplasm from the fertilized egg [67]. In both cases, we found that the change in the ratio (DNA/cytoplasm) did not exert any measurable effect on the timing of the activation of endogenous rDNA.

However, the DNA injection experiment pro- vides more direct and powerful methods to in- crease the amount of DNA within the embryos. It is of an interest to see whether there is a change in gene expression in fertilized eggs or in developing embryos, when the relative ratio of DNA to cyto- plasm, which could be a possible factor to deter- mine the timing of gene expression [53], is changed artificially by the injection of a large amount of exogenous DNA. And so, we will briefly summa- rize here the experiments carried out on this line using coenocytic egg cells and cleaving embryos.

Experiments using coenocytic egg cells

Coenocytic egg cells which are viable for a relatively long period of time without cell division are prepared by centrifuging fertilized Xenopus egg at a low speed for a short time and then removing the spindle apparatus from its normal site. Although these egg cells (or coenocytic egg cells) do not divide, they synthesize DNA to some extent, and produce uncleaved egg cells with in- creased DNA content.

Newport and Kirschner [53] injected yeast tRNA gene (1 ng/egg; pYLT20) and then after 2 hr a large amount (24ng/egg) of plasmid (pBR322) DNA into the coenocytic egg cell. They found that the latter injection of DNA (pBR322) which corresponds to the total DNA of an embryo at the MBT stage (cell number; 4,000) caused earlier expression of both previously-injected yeast tRNA genes and endogenous (nuclear) genes. From these observations, Newport and Kirschner [53] proposed a “titration” model, in which they predicted the presence in the egg of a limited amount of DNA-binding inhibitor of transcription. Furthermore, these authors postulated that the decrease in the amount of the inhibitor to a certain

DNAs Introduced into Xenopus Embryos 217

level, which could be caused by the increase in the amount of DNA, triggers the initiation of the transcription from both endogenous and previous- ly-injected DNAs.

Although this titration model has been modified later by Kimelman et al. [37] who proposed that events at the MBT are regulated by some changes related to cell cycle, we were still interested in the experiment by Newport and Kirschner [53] and so reexamined whether the second injection of a large amount of DNA is able to modify the timing and the level of the expression of the previously- injected plasmid DNA in coenocytic egg cells. When we first injected pSV2CAT (1 ng/egg) and then after 2hr rDNA-containing plasmid pXlr101A (12 ng/egg) [2] into the coenocytic egg cells, we observed the faster and enhanced ex- pression of CAT genes by the second injection of pXIr101A [63]. Therefore, in coenocytic egg cells the second injection of a large amount of DNA appears to have stimulating effect on the transcrip- tion of the previously-injected DNA.

More recently, however, Lund and Dahlberg [45] have commented on the previous studies made by Newport and Kirschner [53]. They reexamined the macromolecular synthesis in coenocytic egg cells, and found that the increase in the amount of DNA and RNA in these cells is reduced heavily when compared to that of the normal cleaving embryos. They also showed that RNA metabolism including the synthesis of snRNA and tRNA is quite aberrant in such coenocytic egg cells due to the poor replication of these genes. Therefore, Lund and Dahlberg [45] proposed that the pre- vious experiments on transcriptional activation in coenocytic egg cells at the “MBT” must be re- evaluated (in fact, MBT stage or 12th cleavage stage does not exist in coenocytic egg cells).

Experiments in cleaving embryos

We carried out experiments consisting of two consecutive DNA injections using the normally developing embryos in order to compare the sequ- ence of events in the above-described coenocytic egg cells with those in normally developing embryos [63]. We first injected 1ng/egg of pSV2CAT and immediately after that injected 12 ng/egg of DNAs from different sources, including

pXIrl101A, total DNA (sonicated) of Xenopus tadpoles, and pBR322. As expected, these DNA injections interfered with cleavage in the animal hemisphere to varying extents. However, the injection of the second DNA neither modified the level nor affected the timing of the expression of the CAT genes.

It was confirmed using *H-labeled lambda DNA that nuclei of most of the blastomeres of the injected embryos contained the label, indicating the wide spreading of the injected DNA during cleavage. Then, the embryos injected with 12 ng/ egg of the above DNAs (pXIr101A, lambda DNA, or sonicated Xenopus tadpole DNA) were dissoci- ated into cells at the blastula stage and their RNA synthesis was examined by labelling cells with 3H-uridine. We found here on agarose-poly- acrylamide gels no appreciable change in the pat- tern of RNA synthesis between the control and the DNA-injected embryos at later stages (gastrula and early neurula stages).

In developing embryonic cells, 1t was also shown that 12 ng of secondarily-injected DNAs did not affect the level and the timing of the expression of previously-injected pSV2CAT (1 ng/egg). There- fore, the program of gene expression in the nuclei of developing embryos appears to be constructed in such a way that it is not easily altered by introduction of excess amounts of exogenous DNAs.

EXPRESSION OF EXOGENOUS GENES IN OOCYTES AND UNFERTILIZED EGGS

At first the DNA injection experiments in Xeno- pus system started by introducing the genes into the nucleus (germinal vesicle) of Xenopus oocytes. Initially injected were those that did not produce proteins, for example, rDNA [81], 5S DNA (genes for ribosomal 5S RNA) [10, 29], and tRNA genes [13, 38], but later, genes that code for proteins were also injected [56, 84]. The next step seen was to inject DNAs into unfertilized eggs. However, these experiments were mostly to study the mechanism of replication of exogenous DNAs in the egg cytoplasm [30, 32]. It was only after these experiments that the researches to inject exog- enous DNAs into developing embryos started.

218 M. ASANO AND K. SHIOKAWA

When considering these historical advance- ments, it may be important to compare the be- havior of DNAs injected into the oocyte nucleus with the behavior of those injected into the cyto- plasm of the unfertilized egg, and the cytoplasm of the fertilized egg. Therefore, we summarize here the results of the studies of gene expression from exogenously-introduced DNAs first in oocytes, then in unfertilized eggs.

Gene expression in the oocyte nucleus

We investigated the expression of various CAT gene-containing DNAs after injecting them into the oocyte nucleus. We observed that while circu- lar DNAs were efficiently expressed, linearized DNAs were not expressed appreciably (Table 1) [23, 24], as has been reported previously by Har- land et al. [31] and Bendig and Williams [5]. The inactiveness of linearized DNAs to serve as the templates in oocyte nucleus may be due to the lack of supercoiled domain and hence the absence of torsional stress, which is assumed to be necessary for efficient transcription [31, 56, 57].

A unique feature of the transcription in oocyte nucleus is that when circular CAT genes are in- jected, CAT enzyme expression is independent of the presence or absence of the promoter within the

TABLE 1. and oocytes

plasmid. This feature may be due to the presence of a large amount of DNA dependent RNA polymerase II in the oocyte nucleus (ca. 10* times the amount in a differentiated cell) [58], which creates enough chance of the binding of polymerase to the DNA templates. In this connec- tion, it is said that the site of the initiation of transcription on the template DNA in the oocyte nucleus is not necessarily correct (or faithful), even if the circular plasmid contains the appropriate promoter [5].

In relation to this unique feature of the trans- cription in oocyte nucleus, a very interesting ex- periment was carried out by Wickens ef al [84] using circular forms of cDNA and genomic DNA of Xenopus vitellogenin. They showed that both DNAs were transcribed into RNA within the oocyte nucleus, a result which is consistent with the above observation that pSVOCAT which does not contain promoter is expressed as strongly as the pSV2CAT which contains SV40 promoter.

In the studies of Wickens et al [84], while the transcript from genomic DNA was processed and transported into the cytoplasm as a functional mRNA for translation, the RNA from cDNA, which is by itself completed mRNA, remained in the nucleus and was not transported to the cyto-

Expression and molecular structures of CAT genes in Xenopus laevis embryos, eggs

Stages CAT gene in circular form Cat gene in linear form Embryos Actively expressed, depending on the Actively expressed, irrespective of the strength of the promoter. presence or absence of the promoter. DNAs replicate moderately or only DNAs replicate actively and form slightly mostly as DNAs of monomer concatemers, and integrated in the size. host DNA. Nucleus-like Nucleus-like structures formed. structures formed. Eggs Very weakly expressed, depending on _ Very weakly expressed, irrespective of the promoter. the presence or absence of the prom- oter. DNAs replicate moderately mostly as DNAs are ligated, and form con- circular DNAs of monomer size. catemers and replicated actively. Nucleus-like structures formed. Oocytes Actively expressed, even when the Expression undetectable even when a

promoter is missing. DNA stably preserved.

DNAs injected into the cytoplasm are degraded.

strong promoter is present. DNAs stably preserved.

DNAs nested into the cytoplasm are degraded.

DNAs Introduced into Xenopus Embryos 219

plasm to be translated into protein. Since the cDNA has neither promoter for correct initiation of transcription nor introns to be spliced out during the processing, it was assumed that nuclear proces- sing may be coupled with nucleo-cytoplasmic transport of the mRNA. Therefore, in conclusion, there may be a limit in using the oocyte nucleus system for the study of the correct transcription of DNA templates.

Gene expression in unfertilized eggs

In unfertilized eggs, changes seen in the amount and size of the injected DNAs are essentially the same as those described above for the DNAs injected into the cytoplasm of fertilized eggs. Thus, in the cytoplasm of unfertilized eggs, c.c. DNA is rapidly converted into 0.c. DNA, but c.c. DNA reappears shortly after, and the amount of these circular DNAs increases only to a limited extent during the incubation. Linearized DNAs, on the other hand, actively form concatemers and are replicated extensively [6, 30, 32].

It has been reported that in unfertilized eggs the transcription from circular DNAs starts from the correct initiation point [5], just as in the fertilized eggs [3, 60]. In the unfertilized eggs injected with CAT-containing circular plasmids, such as pSV2CAT and pAd12E1laCAT, CAT enzyme was expressed in a promoter-dependent way, although the extent of the expression was relatively quite low [23, 24]. The expression was observed also for linearized DNAs, but the extent of the expression was extremely low as compared with that observed with circular DNAs.

In the cytoplasm of an unfertilized egg, there is no nucleus but instead metaphase arrested chromosomes are present as has been shown above in Figure 1A. However, the pricking of the egg with a glass needle activates the egg, and then probably induces assemblage of nucleus-like struc- tures [21]. Then, it may be assumed that the low efficiency of the expression of the injected DNA in unfertilized eggs may be due to the assemblage of most injected DNA in the form of chromatin probably in an inactive form. Or, it may be that the scattermmg of DNA-dependent RNA poly- merase II in the egg cytoplasm at the germinal vesicle breakdown (GVBD) has resulted in the

presence of only a limited amount of the polymerase in the site of the DNA transcription.

Much higher efficiency of the expression of injected DNAs in embryos as compared with that in unfertilized eggs would be due to the assem- blage of a larger amount of injected DNA and RNA polymerase II in nuclei in embryonic cells than that in the unfertilized egg. If so, it follows that the injected DNAs are expressed after they are incorporated into the nuclei, as suggested above in relation to the parallel increase in the CAT enzyme expression and cell number in the experiment in Figure 10.

CONCLUSIONS

Studies of the control mechanism of exogenous- ly-introduced DNAs in Xenopus early embryos provide a special opportunity to obtain clues to the understanding in molecular terms of many impor- tant aspects of the regulation of zygotic gene expression in animal development, especially in its early phases. When relatively large amount (10 ng/egg or more) of circular or linearized DNAs are injected into the cytoplasm of fertilized eggs, injected DNAs are assembled in the form of giant nucleus-like structures mostly in the animal region of the egg, probably after being complexed with maternal histones. Such giant nucleus-like struc- tures are partitioned into descendant blastomeres, and embryonic development becomes more or less abnormal, especially in the animal half region. However, when relatively small amount (1-2 ng/ egg or less) of such DNAs are injected, nucleus- like structures are not seen, and injected embryos show apparently normal development.

In the developing embryos, injected DNAs are replicated and expressed to different extent de- pending upon their structure. When circular plas- mid DNAs are injected, they replicate only to a limited extent mostly as circular DNAs of mono- mer size, and the expression of the genes included is dependent upon the promoter. Thus, expression of circular form of bacterial CAT genes connected to viral promoter is initiated from the early cleav- age stage, and the expression increases as the embryos develop. Also, CAT genes connected to the developmentally-controlled promoter, such as

220

cardiac a-actin gene, are expressed in both tempo- rally and spatially controlled manner.

By contrast, when linearized DNAs are injected into the fertilized eggs, the injected DNAs are ligated to form large-sized concatemers of prob- ably every possible conbinations (head-to-head, head-to-tail, and tail-to-tail orientation), and repli- cated extensively to form a large amount of large- sized DNA. Under these conditions, injected linearized DNAs are actively expressed, but the expression is independent on the presence of the promoter within the DNA.

When the linearized DNAs contain the develop- mentally-controlled promoter such as the promo- ter of cardiac a-actin gene, their expression takes place in cell-type specific manner in the correct region of the embryo. However, the temporal control of their expression is abolished in such linearized DNAs. Thus, while circular CAT genes with the above Xenopus a-actin promoter are expressed at the correct timing at the neurula Stage, these genes start to be expressed already from the blastula or even from the early blastula stage if the DNAs were injected in linearized form.

It has been previously proposed that in Xenopus embryos the earliest stage of the expression of both endogenous (chromosomal) and exogenous- ly-introduced genes is at the 12th cleavage, or at the stage when so-called midblastula transition (MBT) takes place. However, our data strongly Suggest that it is the nature of the promoter contained in the DNA templates, and not the changes associated with the MBT, that determines the timing of the expression of the genes which are exogenously-introduced in circular form into Xenopus embryos.

ACKNOWLEDGMENTS

We wish to thank Professor Seiichiro Kawashima for his kind help in editing the manuscript. We are grateful to Professors Kiyotaka Yamana, Keiichi Hosokawa, and Tsuneyuki Yamazaki for their warm encouragements, and also to Drs. Kosuke Tashiro and Yuchang Fu, for their technical helps in doing the studies cited above. Special thanks are also due to Ms. Reiri Kurashima in our laboratory for her kind help in preparing our manu- script. The present work is supported in part by Narishige Zoological Science Award (1992) given to M. Asano.

M. ASANO AND K.

29

SHIOKAWA

REFERENCES

Andres A-C, Muellner DB, Ryffel GU (1984) Nucl Acids Res 12: 2283-2302

Bakken A, Morgan G, Sollner-Webb B, Roan J, Busby S, Reeder RH (1982) Proc Natl Acad Sci USA 79: 56-60

Bendig MM (1981) Nature (London) 292: 65-67 Bendig MM, Williams JG (1984) Mol Cell Biol, 4: 567-570

Bendig MM, Williams JG (1984) Mol Cell Biol 4: 2109-2119

Berg CA, Gall JG (1986) J Cell Biol 103: 691-698. Blow JJ, Laskey RA (1988) Nature 332: 546-548 Brennan SM (1990) Differentiation 44: 111-121 Brown DD, Caston JD (1962) Dev Biol 5: 412-434 Brown DD, Gurdon JB (1977) Proc Natl Acad Sci 74: 2064-2068

Brown DD, Littna E (1964) J Mol Biol 8: 669-687 Busby SJ, Reeder RH (1983) Cell 34: 989-996 Cortese R, Harland RM, Melton DA (1980) Proc Natl Acad Sci USA 77: 4147-4151

Dawid IB (1965) J Mol Biol 12: 581-599

Endean DJ, Smithies O, (1989) Chromosoma 97: 307-314

Etkin LD, Balcells S. (1985) Dev Biol 108: 173-178 Etkin LD, Pearman B (1987) Development 99: 15— 23

Etkin LD, Pearman B, Ansah-Yiadom R, (1987) Exptl Cell Res 169: 468-477

Etkin LD, Pearman B, Roberts M, Bektesh S (1984) Differentiation 26: 194-202

Flytzanis CN, McMahon AP, Hough-Evans BR, Katula KS, Britten RJ, Davidson EH (1985) Dev Biol 108: 431-442

Forbes DJ, Kirschner MW, Newport JW (1983) Cell 34: 13-23

Franks RR, Hough-Evans BR, Britten RJ, David- son EH (1988) Development 102: 287-299

Fu Y, Hosokawa K, Shiokawa K (1989) Roux’s Arch Dev Biol 198: 148-156

Fu Y, Sato K, Hosokawa K, Shiokawa K (1990) Zool Sci 7: 195-200

Gorman CM, Moffat LF, Howard BH (1982) Mol Cell Biol 2: 1044-1051

Gurdon JB, Melton DA (1981) Ann Rev Genet 15: 189-218

Gurdon JB (1974) The control of Gene Expression in Animal Development Clarendon Press Oxford Gurdon JB, Brown DD (1977) In “The Molecular Biology of the Genetic Apparatus vol 2” Ed by P T’so, North-Holland Publ Co, Amsterdam, pp 111- 123

Gurdon JB, Brown DD (1978) Dev Biol 67: 346-

30 31

32 33

34

3)

36

37

38

39

40

4]

42

43

44 45

46

47

48

49

50

51

52

53}

54

25)

56

57

58

59

60

DNAs Introduced into Xenopus Embryos

356 Harland RM, Laskey RA (1980) Cell 21: 761-771

Harland RM, Weintraub H, McKnight SL (1983) Nature 302: 38-43

Hines PJ, Benbow RM (1982) Cell 30: 459-468

Hofmann A, Montag M, Steinbeisser H, Tren- delenburg F, (1990) Cell Differ Dev 30: 77-85 Honjo T, Reeder RH (1973) J Mol Biol 80: 217-228 Iwao Y, Katagiri C (1984) J Exp Zool 230: 115-124 Jonas EA, Snape AM, Sargent TD (1989) Develop- ment 106: 399-405

Kimelman D, Kirschner M, Scherson T (1987) Cell 48: 399-407

Kressmann A, Clarkson SG, Pirrotta V, Birnstiel ML (1978) Proc Natl Acad Sci USA 75: 1176-1180

Krieg PA, Melton DA (1987) Proc Natl Acad Sci USA 84: 2331-2335

Kruczek I, Doerfler W (1983) Proc Natl Acad Sci USA 80, 7586-7590

Krone P H, Heikkila JJ (1989) Development 106: 271-281

Langner KD, Weyer U, Doerfler W (1986) Proc Natl Acad Sci USA 83: 1598-1602

Leno GH, and Laskey RA (1991) J Cell Biol 112: 557-566

Lohka MJ, Masui Y (1983) Science 220: 719-721

Lund E, Dahlberg JE (1992) Genes Develop 6: 1097-1106

Marini NJ, Etkin LD, Benbow RM (1988) Dev Biol 127: 421-434

Marini NJ, Hiriyanna KT, Benbow RM (1989) Nuc Acids Res 17: 5793-5808

McMahon AP, Flytzanis CN, Hough-Evans BR, Katula KS, Britten RJ, Davidson EH. (1985) Dev. Biol., 108: 420-430

Mechali M, Kearsey (1984) Cell 38: 55-64

Mohun TJ, Garrett N, Gurdon JB (1986) EMBO J 5: 3185-3193

Nakakura N, Miura T, Yamana K, Ito A, Shiokawa K (1987) Dev. Biol., 123: 421-429

Newport J, Kirschner M (1982) Cell, 30: 675-686 Newport J, Kirschner M (1982) Cell, 30: 687-696 Newport J (1987) Cell 48: 205-217

Ozato K, Kondoh H, Inohara H, Iwamatsu T, Wakamatsu Y, Okada TS (1986) Cell Differ 19: 237-244

Probst E, Kressmann A, Birnstiel ML (1979) J Mol Biol 135: 709-732

Pruitt S, Reeder RH (1984) J Mol Biol 174, 121-139 Roeder RG (1974) J Biol Chem 249: 249-256 Rosenthal N, Kress M, Gruss P, Khoury G (1983) Science 222: 749-755

Rusconi S, Schaffner W (1981) Proc Natl Acad Sci USA 78: 5051-5055

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78 79

80

81

82

83

84

85

86

221

Shiokawa K (1991) Develop Growth Differ 33: 1-8 Shiokawa K, Fu Y, Hosokawa K, Yamana K (1990) Roux’s Arch Dev Biol 199: 174-180

Shiokawa K, Fu Y, Nakakura N, Tashiro K, Sameshima M, Hosokawa K (1989) Roux’s Arch Dev Biol 198: 78-84

Shiokawa K, Misumi Y, Tashiro K, Nakakura N, Yamana K, Oh-uchida M (1989) Cell Differ Dev 28: 17-26

Shiokawa K, Misumi Y, Yamana K (1981) Dev Growth Differ 23: 579-587

Shiokawa K, Sameshima M, Tashiro K, Miura T, Nakakura N, Yamana K (1986) Dev Biol 116: 539- 542

Shiokawa K, Takeichi T, Miyata S, Tashiro K, Matsuda K (1985) Cytobios 43: 319-334

Shiokawa K, Tashiro K, Misumi Y, Yamana K (1981) Dev Growth Differ 23: 589-597

Shiokawa K, Tashiro K, Yamana K, Sameshima M (1987) Cell Differ 20: 253-261

Shiokawa K, Yamana K (1965) Exptl Cell Res 38: 180-186

Shiokawa K, Yamana K (1967) Dev Biol, 16: 368- 388

Shiokawa K, Yamana K, Fu Y, Atsuchi Y, Hosoka- wa K (1990) Roux’s Arch Dev Biol 198: 322-329 Shiokawa K, Yamazaki T, Fu Y, Tashiro K, Tsurugi K, Motizuki M, Ikegami Y, Araki E, Andoh T, Hosokawa K (1989b) Cell Struct Func 14: 261-269 Shiokawa K, Yoshida M, Fukamachi H, Fu Y, Tashiro K, Sameshima M (1992) Develop Growth and Differ 34: 79-90

Steinbeisser H, Alonso A, Epperlein HH, Tren- delenburg MF (1989) Int J Dev Biol 33: 361-368 Steinbeisser H, Hofmann A, Stutz F, Trendelenburg MF (1988) Nuc Acids Res 16: 3223-3238

Stuart GW, McMurray JV, Westerfield M (1988) Development 103: 403-412

Szollosi MS, Szollosi D (1988) J Cell Sci 91: 257-267 Takeichi T, Satoh N, Tashiro K, Shiokawa K (1985) Dev Biol 112: 443-450

Tashiro K, Inoue M, Sakaki Y, Shiokawa K (1986) Cell Struct Func 11: 109-114

Trendelenburg MF, Gurdon JB (1978) Nature 276: 292-294

Trendelenburg MF, Oudet P, Spring H, Montag M (1986) J Embryol exp Morph 97: Supplement 243- 255

Vitelli L, Kemler I, Lauber B, Birnstiel ML, Bus- slinger M (1988) Dev Biol 127: 54-63

Wickens MP, Woo S, O’Malley BW, Gurdon JB (1980) Nature 285: 628-634

Wilson C, Cross GS, Woodland HR (1986) Cell 47: 589-599

Winning RS, Bols NC, Wooden SK, Lee AS, Heik-

227) M. ASANO AND K. SHIOKAWA

kila JJ (1992) Differentiation 49: 1-6 88 Yoshizaki G, Oshiro T, Takashima F (1991) Nippon

87 Woodland HR, Gurdon JB (1968) J Embryol Exp Suisan Gakkaishi 57: 819-824 Morph 19: 363-385

ZOOLOGICAL SCIENCE 10: 223-234 (1993)

© 1993 Zoological Society of Japan

REVIEW

Comparative Aspects of Intestinal Calcium Transport in Fish and Amphibians

DouGLas R. ROBERTSON

Department of Anatomy & Cell Biology, State University of New York Health Science Center Syracuse, New York, 13210, USA

INTRODUCTION

One of the features in the transition of verte- brates from an aquatic to a terrestrial environment is the maintenance of a relatively constant cellular ionic calcium concentration of <1 “M. Such a level is easily achieved by passive diffusion alone since a gradient exists between the ionic calcium concentration in plasma, usually greater than 1 mM, and the ionic calcium within the cytosol of the cell.

In a marine environment, a gradient also exists between seawater (10-15mM) and the plasma compartment where the plasma ionic calcium con- centration in marine fish is about 1.7-1.9 mM [6, 8]. Thus, for marine animals, calcium can be acquired by passive diffusion across epithelia; i.e. gills (20, 60]. For marine forms calcium homeosta- sis is concerned primarily with calcium exclusion, rather than acquisition; thus, excess calcium is secreted into the gut lumen and expelled with the feces.

For terrestrial animals, lacking gills, calcium acquisition is obtained by dietary sources, and plasma ionic levels are maintained in a range from 0.9-1.4 mM in frogs [54]; 1.2-1.4 mM in chickens [46] and 1.5 mM in man [31]. With the intestine serving as a primary source for calcium acquisition, it is apparent that movement into the plasma compartment cannot be achieved by passive diffu- sion alone if gut lumen concentration is below the plasma ionic concentration.

Received October 14, 1992

For fish in freshwater with low environmental calcium (<1 mM) and terrestrial animals, an addi- tional mechanism has been added to allow the gut to “up-regulate”, and actively absorb calcium aganist a concentration gradient. This additional capacity is achieved with a vitamin D-dependent calcium binding protein (CaBP) localized in the primary gut absorptive cell, the enterocyte.

MORPHOLOGY

The intestinal epithelium, composed of entero- cytes, arise from stem cells found within crypts at the base of villus projections in birds and mam- mals. These cells are attached to one another via tight junctions situated near the apical region of the cell, with a narrow intermediate junction and wider intervening paracellular space near the basal region of the cell (Fig. 1). Enterocytes have the capacity to move calcium from mucosa to the underlying serosa (Cainfux, Jm—s), OF aS secretion from serosa to mucosa (Caougux, Js+m). Both Cainfux by passive diffusion and Caoutqux occur within the paracellular space, and is largely depen- dent upon the calcium concentration gradient be- tween the two compartments.

On the other hand, transcellular movement of calcium through the enterocyte is energy depen- dent, and enters via the apical microvillus border, and is extruded across the basolateral membrane to the underlying vascular compartment. Calcium entry into the cell is by passive diffusion, or possibly through specific calcium channels. Once into the cell, calcium movement may be facilitated

224 D. R. ROBERTSON

ACa’

2LQYG Prey SAS 2604 Meratetans Neletaheietal

Fic. 1.

we

Schematic representation of calcium movement across gut epithelium.

A & A Aca

A526 SAS 269625 FAIS talaterarateta tan tetalotels Ulett? Sriseteetaeten etre cern

Spahr pons nnonsy

oe sy Seabee

LPL L TELE LD SS

sa

Columnar enterocytes have

microvillus apical surface while lateral membranes are attached by apical tight junctions (TJ) with a basal-lateral intervening paracellular space. Caingux by passive diffusion or Caouraux (secretion) occurs in paracellular space while active calcium transport is transcellular. Calcium ions entering the enterocytes cross apical border to bind to organelles; cytosolic vesicles or mitochondria, or proteins; e.g. CaBP, to reduce cytosolic ionic calcium and increase diffusion constant. Calcium is extruded across basal membranes via Ca** /Mg**-ATPase and Na*/

Ca**-exchange mechanisms.

by enzymes such as alkaline phosphatase, be se- questered into specific organelles; i.e., mitochon- dria, microtubules, lysosomes, vesicles, or be bound to specific proteins; i.e. calmodulin or a vitamin D-dependent calcium binding protein (CaBP). With movement of calcium into the basal cytosol, calcium is extruded across the basolateral membranes via a Ca”*/Mg**-dependent ATP ase

or by a Na‘ /Ca**-exchange mechanism.

As enterocytes arise from the crypts between intestinal villi, they mature and acquire the capac- ity to transport calcium with maximal transport in cells isolated near the villus tip [70, 72]. In the rat model, isolation of various populations of cells along the villus axis show high alkaline phospha- tase and calcium binding protein (CaBP) acitivity

Calcium Absorption in Heterotherms 225

in those cells at the villus tip, with decreasing amounts in those cells at the villus base and within the crypts. ATP-dependent Ca** transport is highest just below the villus tip while calmodulin is equally distributed from tip to crypt. In the rat, a Na*/Ca** exchanger has been found at all villus axis levels, but is of minor importance.

PATTERNS OF CALCIUM ABSORPTION

For marine fish, with a high environmental calcium, acquisition is primarily through the gills [20, 60]; although Sundell and Bjornsson [62] concluded that about 30% of whole body influx was contributed by intestinal absorption of calcium in the Atlantic cod, Gadus. When placed in 50% seawater the river lamprey, Lampetra fluviatilis [48], exhibited both Cajngux, and Caoutqux with net calcium secretion. However, in a freshwater en- vironment or during periods of increased need for calcium; e.g. gonadal maturation, intestinal cal- cium absorption may assume a greater role in calcium acquisition [4]. Intestinal calcium absorp- tion has been identified in freshwater forms such as the trout, Salmo gairdneri [33, 58], and goldfish, Carassius auratus, [16]; while other studies with the carp, Cyprinus carpio [35] have shown no calcium absorption. .

In terrestrial animals, the intestine serves as the organ for calcium acquisition; the proximal small intestine (duodenum) has the highest calcium up- take capacity, followed by small intestine with the lowest capacity in the colon. Amphibians as a group have been little studied except for the anurans. Using the everted gut sac technique, higher calcium transport capacity is seen in the proximal duodenum than in distal small intestine of the leopard frog, Rana pipiens [52]. Uptake is enhanced 50% in the duodenum after vitamin D; administration, indicating the presence of an ac- tive, vitamin D-dependent calcium transport mechanism [50]. In R. esculenta, uptake of *Ca injected into the intestinal lumen confirmed that the proximal intestinal segment has a higher up- take capacity than the distal segment [13].

For birds [73] and mammals [14], highest vita- min D-dependent active transport capacity also occurs in the duodenum, and decreases distally

along the length of the small intestine as: duo- denum>jejunum>ileum. However in some mammals; specifically the rat, segments of the large intestine; i.e. cecum, show a higher uptake in normally fed animals. Vitamin D> or D3; increases calcium uptake in the avian duodenum [67], while vitamin D3 can also increase uptake in distal ileum [26]. In the rat, vitamin D3 not only enhances duodenal calcium absorption, but cecum as well [14]. However, in humans no detectable levels of CaBP is found in cecum or colon [61].

MECHANISMS OF CALCIUM ABSORPTION

Passive diffusion.

Passive diffusion, or the nonsaturable uptake of calcium in the gut from mucosal to serosal surface (Jm—s), 1s linear with luminal calcium concentra- tion. Voltage clamp studies on rat ileum [39] indicate that passive diffusion is paracellular and is significant when luminal calcium concentrations are at 10 mM; but negligible at physiological con- centrations of 0.05mM. Nellans [37] also has proposed a translocation pathway that includes negative charged endocytoic vesicles that isolates the bound calcium from cytosolic calcium, mimick- ing a paracellular route.

Using mannitol fluxes to measure paracellular routes, regional differences in paracellular calcium fluxes have been seen between rat cecum and ileum [38, 39]. Pansu and coworkers [41-43], applying a graded series of calcium concentrations with an in situ loop preparation, determined that the diffusion constant (Kg) was relatively constant along the length of the adult rat intestine, where Kg=0.20-0.25/hour (passive diffusion was limited to 20-25% of available luminal calcium).

Sundell and Bjornsson [62] using an in situ intestinal loop procedure found in the Atlantic cod, Gadus morhua, a passive transport component that amounted to about 40% of total calcium absorption at a luminal concentration of about 15mM. In the frog Rana pipiens, the passive diffusion component (Kg) in the total small intes- tine is 0.12-0.15/0.5hr of available calcium. Thus, at a luminal concentration of 15 mM, 50% of total calcium absorption is derived from passive

226

diffusion [Robertson, unpublished observations]. A value of Kg=0.16/hr obtained in the adult rat [7] would suggest that the higher diffusional con- stants in the frog and marine fish may reflect a more permeable mucosa than in the adult rat.

Secretion.

In contrast to passive diffusion of calcium from lumen to plasma, Caourfux, Or secretion, represents the movement of calcium from serosa to mucosa (Jsm), and can effectively reduce net calcium absorption as,

Canet = Cainfux ae Caoutfiux

(1)

where Cay; represents the sum of the two uni- directional calcium fluxes. For example, in vitamin D-depleted rats, with a luminal calcium concentra- tion of 0.54mM, 56% of Caijngux is secreted back into the lumen, with net calcium absorbed repre- senting only 44% of Caoutfux. Administration of vitamin D increases net calctum absorption, and reduces secretion (Caourfux) to 31% of Cainfux [76]. While vitamin D up-regulates the duodenum to increase Caingux, the secretory component can diminish the efficiency of active calcium transport.

60 50 40 30 20

10

Calcium flux (% of lumen calcium)

0.8 mM (-D)

Histogram depicting calclum movement in duodenum of adult frog (Rana pipiens) of control and vitamin D, (500 yg) using in situ loop preparation with 0.8 and 40 mM calcium in incubation buffer. At 0.8 mM, net Cajaqux is reduced by high Cagurnux, While at 40 mM Caguiux is relatively low with resultant high net Caj,pux. At both luminal concentrations, vitamin D3 increases *Cajnqy, and therefore net Cajjnus-

Fic. 2.

0.8 mM (+D)

D. R. ROBERTSON

Serosal-to-mucosal calcium and mannitol per- meability studies indicate that secretion is primari- ly paracellular [14]. The process is not energy dependent since it is not abolished or reduced by metabolic inhibitors or temperature sensitive [39]. Two routes have been identified; solvent-drag that is dependent on paracellular water flow and diffu- sional flux, dependent on paracellular permeability [37]. Thus, secretion can be driven by a calcium gradient between plasma and lumen ionic calcium when luminal calcium falls below 1.5 mM.

Some fish such as carp, Cyprinus carpio, do not absorb or secrete calcium at any point along the intestine [35]; whereas the Japanese eel, Anguilla japonica, kept in freshwater, exhibits secretion of both calcium and magnesium ions into the intestin- al lumen. Neither hypophysectomy, stanniectomy or calcitonin administration had any effect on net calcium secretion [36]. In lamprey, in addition to calcium J,_., along the length of the intestine, there is a greater calcium J,., at the posterior intestine, resulting in a positive net Caourqux- Rain- bow trout, Salmo, kept in either fresh water or seawater, exhibit a general increase in calcium concentration of intestinal contents progressing

45Ca_ Influx 40Ca Outflux

Net Flux

40 mM (-D) 40 mM (+D)

Calcium Absorption in Heterotherms 227

toward the distal segments, indicative of calcium secretion into the feces [12].

Caoutfux Or secretion occurs in the frog (Rana pipiens) duodenum and is influenced by luminal calcium concentration. At a luminal calcium of <1mM, Cainfux is about 35% but a high secretion rate reduces the net Cajnqu, to about 6%; while vitamin D increases Cajnay, to elevate net Cainaux to 12% (Fig. 2). At a high luminal calcium (40 mM) with or without vitamin D, net Caingux is elevated, a result of a decrease in secretion [Robertson, unpublished observations]. For whole animal balance studies, the secretory component should be included, especially at low luminal con- centrations when Ca,., absorption will be lower than Caijnfux, despite the presence of an active calcium uptake mechanism.

Transcellular calcium transport.

In contrast to diffusion mechanisms that are paracellular, active calcium transport occurs within the enterocyte and can be represented as a three step process: 1) transfer across the apical mem- brane to the cytosol, 2) transfer through the cyto- sol to the basal region of the cell, and 3) transfer across the basolateral membranes. Since the cell interior has a negative voltage compared to the lumen, calcium movement can cross the apical membrane down the charge gradient. Also, move- ment is favored by simple diffusion since a concen- tration gradient can exist from the lumen (1-5 mM) to less than 1 “M within the cytosol [69]. Use of calcium channel blockers, such as verapamil, appears to reduce gut calcium transport, but its effectivness only at high concentrations (1-2 mM) in contrast to concentrations effective at lower concentrations (1-10 4M) in excitable tissues, sug- gests that the effect may be nonspecific [15, 21, 47, 75).

Transfer within the cytosol 1s affected by a variety of individual processes. First, calcium may be sequestered into various organelles; i.e., mito- chondria, microtubules, lysosomes or vesicles, or bound to specific proteins; i.e. calmodulin. Early studies in the chick recognized that active calctum transport was vitamin D-dependent and was sig- nificantly correlated with the presence of a CaBP along the length of small intestine [66] and with the

amount of CaBP in the transport tissue [32]; i.e. that CaBP is preferentially localized in duodenum, with a decreasing content at more distal gut seg- ments. Since the degree of active calcium trans- port, as expressed by the value of Jmax, varies linearly with the amount of CaBP [7], and the presence of CaBP is vitamin D-dependent, CaBP is viewed as the primary cytosolic “carrier” for movement of calcium within the cytosol. In mam- mals CaBP is a9 Kd molecule or a 28 Kd protein in birds (Calbindin). While the specific role of this protein is not completely defined, its capacity to bind calcium lowers the cytosolic free calcium, effectively increasing transcellular calcium diffu- sion by 60-fold [7]. Thus, it appears as a rate- limited “carrier”; its quantitative presence in intes- tinal mucosa is linearly proportional to calcium absorption.

For the final transfer across the basolateral membrane, there are two mechanisms of calcium extrusion that are recognized; an ATP-dependent Ca**/Mg?* transport mechanism [23, 25], and a Nat /Ca’* exchanger [5], both in the same plasma membrane. In mammals, the ATP-dependent calcium transport pump is the dominant mode of calcium extrusion [23, 25, 34], and does not appear to be a rate-limiting step under normal nutritional conditions [7]; while the Na*/Ca?*-exchanger plays a minor role [71].

In fish, where calcium extrusion mechanisms have been examined, there is evidence for both a ATP-dependent calcium transport pump and a Na*/Ca*t-exchanger. In vitamin D-treated goldfish, Carassius, chlorpromazine, a Ca** ATP- ase inhibitor, reduced calcium absorption by 50%, but only 30% in nontreated animals, suggesting that the vitamin D-stimulated transport mechanism was chlorpromazine-sensitive [17]. A similar re- sponse was seen in the Atlantic cod, Gadus, where chlorpromazine reduced intestinal calcium influx by 45%; however, specificity of the site of action or its role in active transport could not be con- firmed [62]. Using isolated basolateral membrane vesicles from the freshwater fish, Oreochromis (tilapia), Flik and co-workers [19] identifed both the Nat/Ca’* exchanger and an ATP-ase extru- sion mechanism, and noted that calcium extrusion is sodium dependent with the ATP-dependent

228 D. R. ROBERTSON

Ca**+ /Mg** pump only one-sixth as effective as the Na‘t/Ca’* exchanger in extruding calcium from the enterocyte.

Viewing all three steps in intracellular calcium transport, evidence indicates that the rate-limting step is cytosolic transfer; e.g., CaBP-dependent. Since CaBP effectively reduces the cytosolic cal- cium concentration, transfer across the microvilli is facilitated by diffusion; while transfer across the basolateral membranes are by mechanisms that have a higher rate constant than the CaBP- dependent cytosolic transfer.

KINETICS OF CALCIUM ABSORPTION

Active calcium transport is transcellular and Operates against an electrochemical gradient. For kinetic studies, uptake can be assessed using either the everted gut sac assay, or the in situ intestinal loop preparation. Since active transport (A) is a saturable process that follows Michaelis-Menten kinetics; it follows the equation proposed by Taylor and Wasserman [66]:

dred [C2*]

Kn [Ce] i

where Jmax iS the maximum saturable flux from lumen to plasma, K,, is the apparent half- saturation constant for Jmax, and Ca’! is the luminal calcium concentration.

However, since Cajngux also includes a nonsatu- rable component from passive diffusion (Kg); total Cainfux OF Jm—s is represented by:

Cainfux=A+Kg[Ca**] (3)

and requires the use of the in situ loop technique. An idealized model for calcium absorption in a tissue displaying both a saturable and nonsaturable component is illustrated in Figure 3. Following a procedure of Pansu et al., [41], the passive or nonsaturable component (Kq) is represented by a linear uptake profile whose slope is constant; and derived from the uptake data at higher calcium concentrations whose slope is linear, represented by the diagonal line. The remaining component, representing saturable calcium uptake, is used to The resulting curvilinear function (equation 3), expresses the

derive values for J,,,, and K,.

= = Net Calcium s Uptake ® a2 ro) 7) Passive a transport £ = =) = Active transport Oo Calcium concentration (mM) Fic. 3. Idealized model of calcium absorption in situ

intestinal loop preparation with increasing concen- trations of calcium in incubation buffer. Diagonal line, derived from linear uptake at high calcium concentrations, represents uptake by passive diffu- sion whose slope is Kg. Active calcium transport depicted by lower curvilinear line, calculated for values of Ji,ax and K,,, follows Michaelis-Menten kinetics and is derived from uptake data minus contribution from passive diffusion. Upper curvi- linear function, connecting data points, is best-fit of uptake data by active transport and passive diffu- sion.

uptake profile of the original data.

Using such a procedure, Sundell and Bj6rnsson [62] identified both a saturable and nonsaturable component in the small intestine of Atlantic cod, Gadus morhua, where Jmax Was 0.604 “«mol/Kg/hr with a K,,=8.41mM at 10°C. With a luminal calcium concentration of about 15 mM, 60% of uptake was from the saturable component; with the remainder from passive diffusion. In similar studies in the adult male frog, Rana pipiens, assayed at 20-22°C, duodenal active uptake (satu- rable) had a Jmax=7-0 ~mol/0.5 hr/gm wt. tissue compared to the jejunoileum where Jmax=3.3 yvmol/ 0.5 hr/gm. With a luminal concentration of 15 mM, 50% of duodenal uptake was saturable, and 50% by passive diffusion [Robertson, unpub- lished observations]. In the 30-35 day Wistar rat (assayed at 37°C) active saturable calcium trans-

Calcium Absorption in Heterotherms 229

port in the duodenum had a Jmax=31 pmol/hr; while jejunum was 12 “mol/hr with no detectable saturable component in more distal ileum [41]. With a duodenal lumen concentration of 15 mM, the saturable component in the rat accounts for 70% of the total calcium absorbed. Although there were differences in assay temperature, the magnitudes of the active, saturable uptake compo- nent were similar, with active transport observed in more proximal intestinal segments.

Localization of vitamin D-dependent calcium bind- ing protein.

Assessment of the kinetics of active calcium influx that follow Michaelis-Menten kinetics, spe- cifically the derived value of Jmax, Show a sig- nificant correlation with mucosal CaBP. In the growing rat, Jax Varies with age [42, 68]. Young 3-day old rats that lack CaBP exhibit only calcium uptake by passive diffusion; whereas with older animals, there is a progressive increase in the active, saturable uptake mechanism that is corre- lated with the presence of CaBP [43]. Evidence presented indicates that CaBP concentration is directly correlated with the quantity of saturable transport (Jmax) [41] and therefore is related to the total quantity of an available “carrier”. Bronner [7] takes this in support of the concept that the rate-limiting movement of calcium within the cyto- sol is dependent upon CaBP, that acts as a carrier to move calcium from the apical to basal cytosol.

Since intestine of fish and amphibians exhibit an increase in calcium transport in response to vita- min D and its metabolites, there has been interest in localization of vitamin D-dependent CaBP in these forms. Using a “Ca binding procedure, a heat stable intestinal CaBP was identified in the carp (Cyprinus cario) [9]. In another study, CaBP from carp intestine co-migrated with a pig intestin- al CaBP; while a frog CaBP co-migrated with CaBP from rat and chick [40]. It is now recognized that there are two different vitamin D-dependent intestinal calcium-binding proteins present in other vertebrates; a 9 Kd protein found exclusively in mammalian intestine, and a more widely distri- buted 28 Kd molecule found in birds, reptiles and some amphibians. Wesserman and Corradino [74] indicated that frog, toad and turtle intestine

showed cross-reactivity to antisera to chick CaBP; however, Parmentier et al., [45] employing anti- bodies to a chick CaBP (Calbindin 28 Kd), only was able to demonstrate immunocytochemical cross-reactivity in the small intestine of three spe- cies of reptiles, but was unable to demonstrate cross-reactivity in the small intestine of four spe- cies of amphibian or three species of fish. A slight cross-reaction was apparent in Western blots of gels of gut homogenates of Xenopus and Triturus. Subsequently, using immunocytochemical and Western blot techniques, Calbindin 28 Kd was localized in the large intestine of the toad (Bufo bufo), but was not apparent in the proximal duodenal segment [44]. Thus, the distribution pattern of a 28 Kd CaBP seen in the toad differs from that seen in birds [73], rat [43, 65] or humans [61], that show a direct correspondence to vitamin D-dependent active calcium transport. Recent studies [49] have shown Calbindin 28 Kd to coexist in the intestine of pig and jerboa (a leaping rodent) with the 9 Kd intestinal CaBP; although the pre- sence of both types of CaBP is not observed in rat, mouse or goat. These findings bring to question whether the 28 Kd CaBP is always responsive to vitamin D. The presence of a 28 Kd CaBP in the colon of toads [44], but not in the duodenum that exhibits a vitamin D-dependent increase in calcium transport [50, 52] would suggest that further stu- dies are needed to clarify the molecular species of intestinal vitamin D-dependent CaBP in fish and amphibians.

Endocrine relationships.

An increase in intestinal calcium absorption in response to variations in dietary calcium, growth requirements or reproductive demands, are medi- ated by changes in enterocyte vitamin D- dependent CaBP. In birds [30] and mammals [29], 1,25-(OH),-D; is the most biologically active form of vitamin D that is a requirement for CaBP synthesis. To become effective, vitamin D must undergo sequential hydroxylations at the 1- and 25-carbon positions. The first conversion is in the liver to form 25-(OH)-D3; with the second hyd- roxylation in the kidney by 25-hydroxyvitamin D3 (25-D3)-1 alpha-hydroxylase to form 1,25-(OH)>- D; [22]. Feedback mechanisms exist to modulate

230 D. R. RoBERTSON

this conversion. Evidence in rats suggests that 1,25-(OH)>-D; inhibits the conversion in the liver by increasing liver cytosolic calcium [2]; an inhibi- tion also seen in man by increasing the dietary calcium [3]. This observation might suggest that an elevation in venous portal blood draining the small intestine, after a high calc1um load that flows directly to hepatocytes, can modulate 25-(OH)-D3 production. In the kidney an additional metabo- lite, 24,25-(OH)>-Ds, also is formed [28], and may represent a step in a degradative pathway [27]. The renal conversion to 1,25-(OH),-D3 is modu- lated indirectly by plasma ionic calcium mediated by parathyroid hormone (PTH) [22].

Of interest is the phyletic relationship of the metabolites of vitamin D with the appearance of active calcium transport in fish and amphibians and the relationship to parathyroid hormone (PTH). Early studies in the eel (Anguilla anguilla) sug- gested that vitamin D3 was more effective in stimulating intestinal calcium absorption than the more polar 1,25-(OH)2-D3 [10]. A somewhat similar response was seen in the marine Atlantic cod, Gadus [63] where 25-(OH)-D3 increases in- testinal calcium uptake by 65%; whereas the more polar 1,25-(OH)2-D3 was without effect; while 24 ,25-(OH)2-D3 depressed uptake by 36%. The response to 25-(OH)-D3 may be direct since the more polar 1,25-(OH) -D3, showed no response.

On the other hand, in freshwater fish; e.g. eel and talapia, intestinal calcium uptake is stimulated by 1,25-(OH)2-D3 [17, 18, 63], or in goldfish, exposed to vitamin D alone [16]. Recently, Takeuchi et al., [64] found 25-hydroxyvitamin D3-1 alpha-hydroxylase in the liver of carp and halibut. The capacity of some fish to convert vitamin D3 to the more polar metabolites within the liver, in contrast to birds and mammals, would suggest that the kidney in some fish may not be an essential organ in vitamin D metabolism.

Parathyroid glands are lacking in fish, but with their appearance in the amphibians, an additional mechanism is present that may modulate the formation of 1,25-(OH)2-D3, and therefore intes- tinal calcium absorption. Parathyroid removal depresses plasma ionic calcium in the frog Rana pipiens, by 20% [55] and duodenal calcium uptake [51]. While vitamin D administration increases

duodenal active calcium transport [50, 51]), it is depressed in PTH-depleted frogs after vitamin D3 administration [51, 53]; suggestive the the PTH- dependent conversion to the active 1,25-(OH)>-D3 metabolite was inhibited, resulting in depression of the active, saturable component of calcium absorp- tion.

Other factors in calcium absorption.

A variety of factors come into play that can alter intestinal calcium absorption; a major factor is developmental age. Various studies [42, 68] recog- nized that intestinal calclum uptake in newborn mammals is by passive diffusion alone, with a complete lack of active calcium transport. Pansu and co-workers [42] found in newborn rats that the passive diffusion constant was high (Kg=0.83) with a constant decrease to about Kyg=0.33 at 40 days of age, that remained constant at Kg=0.47 thereafter to 160 days of age. During the period of decline, active transport, as expressed by the value of Jmax, increased to peak at 28-40 days with a gradual decline to about 13% of maximum after 150 days. The appearance of active transport was paralleled by a concomitant increase in the mucos- al CaBP. The sequential appearance of 1,25- (OH)>-D3 receptors on crypt cells by 10 days of age [11] and increasing circulating levels of 1,25- (OH)>-D; during the third week [24, 68] indicates the maturation of enterocytes capable of support- ing active calcium transport.

Little is known of gut calcium transport at early stages of development in heterotherms. Calcium balance studies with pre-metamorphic anurans (bullfrog tadpoles), indicate that calcium is ac- quired primarily (70%) by the gills with only 5% attributable to intestinal calcium uptake [1]. However, direct infusion of high calctum (50 mM) into the intestine of bullfrog tadpoles elevates plasma calcium [57], providing a more direct in- dication that calcium can be absorbed by the gut of pre-adult anurans; although specific mechanisms of absorption are unknown.

Other factors that affect calium absorption in- clude availability of calcium; i.e. complexed or ionically “free”, the ionic calcium concentration, and transit time of intestinal contents. A consis- tent relationship in most terrestrial animals is the

Calcium Absorption in Heterotherms 2

high active calcium transport mechanism in the proximal (duodenal) segment of the small intes- tine. Stomach contents are typically at low pH (pH <2) resulting in an increased availability of ionic calcium immediately distal to the pylorus. Studies in humans show that pH in the duodenum remains between pH 3-6 and gradually rises toward neut- rality near the proximal jejunum [59]. Thus, the kinetics of saturable calcium uptake at low luminal calcium concentration in the duodenum result in high efficiency of Ca influx. On the other hand, calcium food sources are diluted by digestive fluids as they pass along the tract, further reducing the calcium concentration, and passive diffusion be- comes less effective. Finally, the lower uptake capacity of the saturable component is compen- sated by the relatively longer length of the je- junoileum, resulting in almost complete absorp- tion of calcium at the terminus. In mammals with defined segments of large intestine with an active transport mechanism, the longer transit time in this segment, may insure that all available calcium is absorbed.

Active -

Active- Normal

Passive Diffusion

Relative Absorption

Duodenum

Fic. 4.

Jejunoileum

Diagrammatic representation of topology of calcium absorption in small and large intestine of adult frog.

Ww —

TOPOLOGY OF CALCIUM ABSORPTION

Of the heterotherms studied, a basic pattern of intestinal calctum absorption is apparent; e.g. pas- sive diffusion occurs along the length of the small (and large) intestine that may be modified by the addition of 1) an active saturable uptake mechan- ism, and 2) net calcium secretion. These modifica- tions exhibit polarity along the length of the gut; e.g. active uptake is higher in proximal segments; whereas net calcium secretion is found in distal segments. Thus, it is recognized [14] that the intestine exhibits a topology of calcium transport mechanisms with segment-specific net calcium absorption capacity. The presence of an active uptake process, mediated by vitamin D-dependent CaBP, provides a means to up-regulate the intes- tine when there is a physiological demand. Under conditions of high dietary calcium, passive diffu- sion will dominate, with down-regulation of active transport. Conversely, with low dietary calcium, passive diffusion is less effective, with up- regulation of active transport. The relative capac- ity of each intestinal segment is illustrated in figure 4 in normal and vitamin D-treated frogs with a

Vitamin D-stimulated

Colon

Passive diffusion is relatively constant along the duodenum, jejunoileum and colon; whereas active transport is highest in duodenum and reduced in jejunoileum and absent in colon in normal frogs. After vitamin D3 (500 yg) adminstation, active transport is elevated in duodenum and jejunoileum but unchanged in colon.

232 D. R. ROBERTSON

luminal concentration of 15 mM. Paracellular cal- cium absorption can account for about 50% of all calcium absorbed in duodenum, about 70% in jejunoileum and 100% in large intestine. After vitamin D, calcium uptake is increased in duode- num and jejunoileum, while uptake in colon re- mains unchanged after 500 “g vitamin D [Robert- son, unpublished data]. On the other hand, Ros- toff et al., [56] noted that uptake in colon could be increased only with supraphysiological doses of vitamin D.

CONCLUSIONS

Of the heterotherms studied, paracellular cal- cium transport in the intestine appears to play as important a role as active, transcellular calcium transport; reflecting a more permeable epithelium to passive diffusion, similar to that in newborn mammals. In adult homeotherms, active calcium transport is well established to respond to the demands of increased body and skeletal size, and episodic demand; e.g. egg shell deposition or lactation. The basic patterns and mechanisms identified in birds and mammals are similar to that in fish and amphibians; however, there are some differences that also may reflect basic nutritional needs.

First, is the role of the various vitamin D metabolites in active calcium transport in fish and amphibians. Current evidence indicates that the less polar metabolites of vitamin D; i.e. 25-(OH)- D3, are effective in stimulating active calcium transport, and may not require the kidneys for further metabolism to the more polar 1,25-(OH)>- D3, as seen in avian and mammalian systems. On the other hand, since some fish possess the enzyme necessary in the liver to form 1,25-(OH)>-D3, a tight feed-back loop may exist between the fish proximal intestine and liver via the hepatic portal system.

The appearance of a well defined intestinal topology of active transport in anuran amphibians, and evidence of PTH-sensitive vitamin D- dependent active gut transport indicates that an avian-mammalian pattern is established in the amphibians. What is lacking is characterization of an intestinal CaBP, and its sensitivity to specific

vitamin D metabolites. Basic information on gut calcium transport in the more aquatic urodele amphibians and appearance of these mechanisms during metamorphosis of anurans is needed.

The elegant studies of Flik and co-workers [19] on calcium extrusion mechanisms in teleost entero- cytes, provides insight into mechanisms that may be more common in the enterocytes of hetero- therms than in homeotherms. While fish gills possess an active Ca** /Mg**-ATPase extrusion mechanism, the presence of a prominent entero- cyte Na*/Ca’*-exchange mechanism hints that intracellular calctum transport in some hetero- therms may not be entirely comparable to the avian/mammalian models.

REFERENCES

1 Baldwin GF and Bentley PJ (1980) J exp Biol 88: 357-365

2 Baran DT and Milne ML (1986) J Clin Invest 77: 1622-1626

3 Bell NH, Shaw S and Turner RT (1987) J Bone Min Res 2: 211-214

4 Berg A (1970) Mem. Ist. Ital. Idrobiol. Dott Marco de March 26: 241-255

5 Birge SJ, Gilbert HR and Avioli LV (1972) Science 176: 168-170

6 Bj6érnsson BTh, and Nilsson S (1985) Am J Physiol 248: R18—R22

7 Bronner F, Pansu D and Stein WD (1986) Am J Physiol 250: G561-569

8 Chan DKO, Chester Jones I and Smith RN (1968) Gen Comp Endocrinology 11: 243-245

9 Chartier Baraduc MM (1973) C R Acad Sc Paris, Ser D, 276: 785-788

10 Chartier MM, Millet C. Martelly E, Lopez E and Warrot S (1979) J Physiol, (Paris) 75: 275-282

11 Clark SA, Stumpf WE, Hamstra A and DeLuca HF (1983) J Histochem Cytochem, 31: 1063

12 Dabrowski K, Leray C, Nonnotte G and Colon DA (1986) Comp Biochem Physiol 83A: 27-39

13. El Maraghi-Ater H, Hourdry J, Mesnard J, and Dupuis Y (1987) Reprod Nutr Dévelop 27: 407-412

14 Favus MJ (1985) Am J Physiol 248: G147—G157

15 Favus MJ and Angeid-Backman E (1985) Am J Physiol 248: G676-G681

16 Fenwick JC (1984) Can J Zool 62: 34-36

17. Fenwick JC, Smith K, Smith J and Flik G (1984) Gen Comp Endocrinology 55: 398-404

18 Flik G, Reijntjens FMJ, Stikkelbroeck J and Fen- wick JC (1982) J Endocrinol 94: 40

28

29

30

31 32

33

34

35

36

3H

38

39

40

41

42

43

44

Calcium Absorption in Heterotherms

Flik G, Schoemnakers Th JM, Groot JA, van Os CH and Wendelaar Bonga SE (1990) J Memb Biol 113; 13-22

Flik G, van Rijs JH and Weendelaar Bonga SE (1985) J exp Biol 119: 335-347

Fox J and Green DT (1986) Europ J Pharm 129: 159-164

Fraser DR and Kodicek E (1970) Nature, New Biol 241: 163-166

Ghijsen WEJM, DeJong MD, and van Os CH (1983) Biochem Biophys Acta 730: 85-94 Halloran BP, Barthell EN and DeLuca HF (1979) Proc Natl Sci USA 76: 5549-5553

Hildmann B, Schmidt A and Murer H (1983) J Memb Biol 65: 55-62

Holdsworth ES, Jordan JE and Keenan E (1975) Bio J 152: 181-190

Holick MF, Baxter LA, Schraufrogel PK, Tavela TE and DeLuca HF (1976) J Biol Chem 251: 397- 402

Holick MF, Schonoes HK, DeLuca HF, Gray RW, Boyle IT and Suda T (1972) Biochemistry 11: 4251- 4255

Kanis JA, Guilland-Cumming DR and Russell RGG (1982) In:Endocrinology of Calcium Metabolism, Ed. JA Parsons, Raven Press, New York pp 321- 362

Liang CT, Barnes J, Balakir RA and Sacktor B (1976) J Memb Biol 90: 145-156

Moore EW (1970) J Clin Inverst 49: 318-334 Morrissey RL and Wasserman RH (1971) Am J Physiol 220: 1509-1515 Mugiya Y and Ichi T (1981) Comp Biochem Physiol 70A: 97-101

Murer H and Hildmann B (1981) Am J Physiol 240: G409-G416

Nakamura Y (1985) Comp Biochem Physiol 80A: 17-20

Nakamura Y and Hirano T (1986) Comp Biochem Physiol 84A: 595-599

Nellans HN (1990) Miner Electrolyte Metab 16: 101-108

Nellans HN and Goldsmith RS (1981) Am J Physiol 240: G424-G431

Nellans HN and Kimberg DV (1978) Am J Physiol 235: E726-E737

Ooisumi K Moriuchi S and Hosoya N (1970) Vita- mins Jap 42: 171-175

Pansu D, Bellaton C and Bronner F (1981) Am J Physiol 240: G32-—G37

Pansu D, Bellaton C and Bronner F (1983) Am J Physiol 244: G20-G26

Pansu D, Bellaton C, Roche C and Bronner F (1983) Am J Physiol 244: G695-G700

Parmentier M (1990) Biol Cell 68: 43-49

45

46

47

48

49

50

51 ov

53 54

55

56

S7/

58

59

60

61 62

63

64

65

66

67

68

69

70

71

72

233

Parmentier M, Ghysens M, Rypens F, Lawson DEM, Pasteels JL and Pochet R (1987) Gen Comp Endocrinology 65: 399-407

Parson AH and Combs GF (1980) Poultry Sceince 60: 1520-1524

Pento JT and Johnson ME (1983) Pharmacol 27: 343-349

Pickering AD and Morris R (1973) J exp Biol 58: 165-176

Pochet R, Blachier F, Gangji V, Kielbaska V, Duée P-H and Résibois A (1990) Biol Cell 70; 91-99 Robertson DR (1975) Comp Biochem Physiol 51A: 705-710

Robertson DR (1975) Endocrinology 96: 934-940 Robertson DR (1976) Comp Biochem Physiol 54A: 225-231

Robertson DR (1978) J Endocrinol 79: 167-178 Robertson DR (1985) Am J Physiol 249: R290- R295

Robertson DR (1986) Comp Biochem Physiol 85A: 359-364

Rostoff CJ, Baldwin GF and Bentley PJ (1983) Am J Physiol 245: R91—-R94

Sasayama Y and Oguro C (1985) In: Current Trends in Comparative Endocrinology. Ed. B Lofts and WN Holmes, Hong Kong Univ. Press pp 838-838 Shehadeh ZH and Gordon MS (1969) Comp Biochem Physiol 30: 397-418

Sheikh MS, Schiller LR and Fordtran JS (1990) Miner Electrolyte Metab 16: 130-146

Simkiss K (1974) In: Ageing of Fish. Ed. by TB Begenal, Univ. Old Working, England, pp 1-12 Staun M (1987) Gut 28: 878-882

Sundell K and Bjérnsson BT (1988) J exp Biol 140: 171-186

Sundell K and Bjérnsson BT (1990) Gen Comp Endocrinol 78: 74-79

Takeuchi A, Okano T and Kobayashi T (1991) Life Sci 48: 275-282

Taylor AN, Gleason WA and Lankford GL (1984) J Histochem Cytochem 32: 153-158

Taylor AN and Wasserman RH (1967) Biochem Biophys 119: 536-540

Taylor AN and Wasserman RH (1969) Fed Proc 28: 1824-1838

Toverud SU and Dostal LA (1986) J Ped Gastroent Nutr 5: 688-695

Tsien RY, Possan T and Rink TJ (1982) J Cell Biol 94: 325-334

van Corven EJJM, Roche C and van Os CH (1985) Biochem Biophys Acta 820: 274-282

Van Os CH (1987) Biochem Biophys Acta 906: 195-222

Walters JRF and Weiser MM (1987) Am J Physiol 252: G170-G177

Arch.

234 D. R. ROBERTSON

73 Wasserman RH (1962) J Nutr 77: 69-80 45: 385-387 74 Wasserman RH and Corradino RA (1973) Vitamins 76 Yeh JK and Aloia JF (1984) Endocrinology 114: and Hormones 31: 43-103 1711-1717

75 Wrobel J and Michalska L (1977) Europ J Pharm

ZOOLOGICAL SCIENCE 10: 235-244 (1993)

Physiological Evaluation of the Role of the Liver as a Mediator of the Growth-Promoting Action of Somatotrophin

NIcoLe L. SCHLECHTER, Larry S. Katz', SHARON M. RUSSELL

and CHARLES S. NICOLL”

Department of Integrative Biology, The Cancer Research Laboratory and Group in Endocrinology, University of California, Berkeley, CA 94720, USA

ABSTRACT—The role of the liver as a mediator of the somatotrophic actions of GH was investigated in three experiments in hypophysectomized rats. The GH was infused via osmotic minipumps into the hepatic portal vein (HPV) or the external jugular vein (EJV) for 7 days at a constant rate or in 8 pulses of 1 hour each/day. Infusion of rat (r) GH at 1 ~g/rat/day into the two vessels in rats hypophysecto- mized one day after catheterization failed to affect body growth, but it increased their tibial epiphysial plate width (TEPW), and serum IGF-I concentration to the same degree regardless of the site of delivery or the schedule of infusion. When bovine (b) GH was infused at (50 “g/rat/day) starting 12-14 days after hypophysectomy, the EJV route was more effective at stimulating growth and elevating serum IGF-I than was HPV infusion by both modes of delivery, and constant infusion was more effective than pulsed delivery. In the third experiment human (h) GH infusion was started 30 days after hypophysectomy at 50 ug/rat/day. This treatment caused much more striking growth responses than did either of the other two GH preparations, and the EJV route of delivery was again more effective than was infusion into the HPV by both schedules. In addition, the constant mode of delivery was again more effective than was pulsed infusion. These results indicate that effects of GH on peripheral tissues may be more important for growth than effects mediated by the liver. Thus, the hepatic effects of GH

© 1993 Zoological Society of Japan

may involve maintaining the responsiveness of the liver to other regulators of IGF-I secretion.

INTRODUCTION

The somatomedin hypothesis [1, 2] proposed that growth hormone (GH) acts indirectly to promote growth by stimulatng hepatic production of a mediator, which is now known as insulin-like growth factor-I (IGF-I). Numerous observations support this hypothesis. For example, several groups have reported that GH increases the secre- tion of somatomedin/IGF-I by the perfused rat liver and by hepatic explants or cells in culture [see reviews in 3]. In addition, in vivo studies involving partial hepatectomy [4], cross-hepatic blood sam- pling [5] and studies of patients with liver disorders

Accepted September 18, 1992 Received June 16, 1992

" Present address: Department of Animal Sciences, Cook College, Rutgers University, New Brunswick, NJ 08903-0231

* To whom reprint requests should be addressed.

[6] demonstrated that the liver produces IGF-I, and that this production is stimulated by GH. Furthermore, Schwander et al. [7] estimated that hepatic production of IGF-I in response to GH can account for all of the IGF-I in serum.

Other data indicate that GH can also act peripherally [8-14]. These results raise questions about the relative significance of peripheral vs hepatic actions of GH in promoting somatic growth. Mick and Nicoll [15] developed a techni- que for testing the direct effects of hormones on hepatic functions in vivo. Their system involves chronic infusion into the hepatic portal vein, which delivers hormone directly to the liver via one of its two blood supplies. Thus, the direct effect of substances on hepatic functions in vivo can be studied in a physiologically meaningful way. In the present study we used their procedure [15] to reevaluate the basic tenet of the somatomedin hypothesis [1, 2].

236 N. L. ScHLecuter, L. S. Katz et al.

MATERIALS AND METHODS

Animals

Male rats weighing 120-150 gm were obtained from our own breeding colony or from Simonsen Laboratories (Gilroy, CA, U.S.A.) and main- tained as described previously [10]. They were hypophysectomized (Hx) by a transauricular approach [16] either before or after catheteriza- tion, depending upon the schedule of the experi- ment. All procedures used on the rats were described in detail in a protocol that was approved by our Institutional Animal Care and Use commit- tee, and all experiments conformed to the regula- tions described in the N.I.H. Guide to the Care and Use of Laboratory Animals.

Catheterization

The procedures used to construct the catheters and insert them into either the hepatic portal vein (HPV) or the external jugular vein (EJV) and connecting them to Osmotic minipumps were as described [10, 13]. In brief, for constant infusion, Alzet minipumps (2001: Alza Corp, Palo Alto, CA, U.S.A.) were filled with solvent [17] with or without GH, and a 7-cm long catheter filled with solvent containing heparin was attached. For pulsatile delivery, the polyethylene 50 catheter was extended to 70 cm and 1-mm segments of solvent or GH solution were drawn into it. The segments were separted by 2-mm-long air bubbles. By this means the GH solution or solvent was infused in 8 pulses/day of about 1 hour duration with approx- imately a 2-hour interpulse interval. This mode of delivery should mimic the endogenous GH secre- tion pattern in adult male rats [18].

Measurements

Changes in body weight were recorded in all three experiments and changes in tail length were measured in two of them. These measurements were made only at the beginning and the end of each experiment. The rats were anesthetized and blood was drawn by cardiac puncture and the tibiae were removed and processed for measure- ment of the width of the epiphysial cartilage plate [TEPW; 19]. The sellar region was examined

carefully using a binocular dissecting microscope for the presence of pituitary remnants. Animals with detectable remnants and/or lack of obvious testicular regression were judged to be incomplete- ly hypophysectomized and were excluded.

Hormones and Assays

Serum from the rats used in the first two experi- ments was processed for measurement of IGF-I by RIA and the concentration of GH was measured by RIA in the serum samples from the rats that received constant infusion of the (r)GH. Details of these RIA procedures have been published [20, 21]. However, because of the breakdown of a freezer, some of the serum samples from the three experiments were lost and so they could not be assayed for IGF-I, human (h)GH or bovine (b)GH levels. The three preparations of GH used were obtained from the National Hormone and Pituit- ary Program of the NIH (USA). The rGH (NIH B-9), bGH (NIH B-18), and hGH (NIDDK B-1) had potencies of 1.9, 3.2 and 2.4 1.U./mg, respec- tively.

Experiments

Three experiments were conducted—one each with rat (r), bovine (b), and human (h) GH. In the first study the rats were Hx one day after catheterization and were killed 6 days later. Thus, this experiment was designed to test the effective- ness of the rGH at maintaining growth after pituit- ary removal. The dose of rGH of 1 ug/rat/day was selected on the basis of results of our previous study [10] in which rGH was infused into the arterial supply of one hindlimb of Hx rats. A dose of 2 ug/rat/day caused a striking increase in the TEPW in the infused limb but it caused a signi- ficant, though lesser growth effect in the contra- lateral hindlimb. Thus, the 1 ug/day dose given via the HPV should have a significant effect on hepatic secretion of IGF-I and growth while the same dose given into the EJV should have a lesser effect.

In the second experiment, the rats were Hx 12- 14 days prior to catheterization and bGH was infused for 7 days at a dose of 50 ug/rat/day to obtain larger responses. In the third experiment, the rats were Hx 30 days before catheterization

GH, the Liver and Growth 237

and bGH was infused at a dose of 50 ug/rat/day for 7 days. In these experiments with rGH and hGH the rats were killed on the 8th day after the placement of the catheters and minipumps. In the experiments with rGH and hGH control rats re- ceived infusion of the solvent that was used to dissolve the GH preparations. In the experiment with bGH the abdomen of the controls was opened to expose the intestines, but no catheter was inserted into an intestinal vein. In all three experi- ments the rats were killed on the 8th day after the start of infusion.

Significance of differences was determined by the Student’s ¢ test or by analysis of variance, as described [10].

RESULTS

In the experiment with rat GH, all of the experimental groups showed an equivalent loss in

275 Constant

250

225

200

175

Tibial Epiphyseal Width (um)

PZ)

150 ae

solv

Fic. 1.

oS

solv rGH L_pyy—1 L—pHpy—

Effect of infusing solvent (solv) or rGH (1 yg/rat/day) in solvent into the external jugular vein (EJV) or

body weight and reduction in tail growth relative to intact or sham Hx controls. Infusion of the rGH did not affect these changes in body weight or tail length in any of the experimental groups (data not shown). Serum GH was not detectable by RIA (sensitivity of 0.7 ng/ml) in any of the rats that received solvent infusions. In those given constant infusion of the hormone via the EJV or the HPV, the serum GH concentrations were 3.0+0.4 ng/ml (n=7) and 3.2+0.1 ng/ml (n=5), respectively. Thus, the route of administration did not affect the peripheral concentration of GH. Serum GH levels were not measured in the rats that received pulse infusion because meanigful data from these anim- als could be obtained only by taking multiple samples over an extended period.

Intact rats of the same age as those used in these experiments had a mean TEPW of 314 um. Thus, the data in Figure 1 show that the TEPW in the rats infused with the solvent had regressed by more

Pulsed

ICSQGQv;

ww)

aD) Cod ZZ soly rGH solv rGH

L_pyy—1 L—ypy—

&

hepatic portal vein (HPV) on the tibial epiphysial plate width of hypophysectomized rats for 7 days. The rats were Hx one day after placement of the catheters and the infusions were either constant or pulsatile (8 pulses of 1 hour each). The number of rats in each group is given within the base of each column. Columns with the same letter superscript are not significantly different from each other. Those with different letters are significant at P< 0.001. The columns represent the mean+SEM values for each group.

238 N. L. ScHLECHTER, L. S. Katz et al.

than 100m during the 6 days following hypophysectomy. The TEPW of the rats that received the solvent in pulses was slightly lower than those measured in the animals given constant infusions. This difference is probably due to the fact that the two groups of experiments (constant and pulsed infusions) were done at different times.

The data in Figure 1 show that constant delivery of rGH into the EJV or the HPV increased the mean TEPW over that in the controls given solvent to the same extent [i.e. by 46 wm (P<0.01), and 43 ym (P<0.001), respectively]. When the rGH was infused in pulses into the EJV or the HPV, the degree of growth stimulation (+48 ~m in both groups, P<0.01) was similar to that produced by constant infusion. Thus, intrahepatic delivery of the hormone either constantly or in pulses was no more effective than infusing it into the peripheral vessel.

The rats infused with solvent at a constant rate had mean serum IGF-I levels of 55-57 ug/ml, regardless of the site of infusion (Fig. 2). This concentration is less tha 10% of the level measured in intact rats of the same age (i.e. 615 ~g/ml).

100

& N i) —) Se —)

Serum IGF-I (ng/ml) =

SOLV

Intrajugular

Serum IGF-I levels in the rats that received constant infusions as described in Fig. 1. Otherwise as in the

Fic. 2.

rGH

Infusion of rGH into either the HPV or the EJV elevated serum IGF-I concentrations to the same degree over those in the rats that were infused with solvent. Thus, the serum IGF-I levels in these animals are consistent with their equivalent TEPW measurements (Fig. 1) and serum concentrations of rGH.

The effects of infusing bGH at 50 ug/rat/day on changes in body weight and tail length are shown in Fig. 3. The solvent-infused rats lost about 10 gm in body weight during the 7-day experimental period but constant infusion of bGH into either vein prevented this loss. However, constant deliv- ery into the EJV was significantly more effective than was infusion into the HPV. Pulsed delivery of bGH into the HPV did not prevent the loss in body weight. Although pulsed infusion of GH into the EJV was effective in this regard, it was significantly less effective than was constant infusion into that vein. Tail growth was increased significantly only in the groups that received bGH constantly via the EJV.

Only the group that received constant infusion of bGH into the EJV showed a singificant TEPW

rGH

SOLV Intrahepatic

legend for Fig. 1, but columns with different letters as significantly different at P<0.01.

GH, the Liver and Growth 239

5, Body Weight

Change in Body Weight (gm)

47 Tail Length

Change in Tail length (mm)

EJV HPV EJV HPV

Sham Constant Pulsed

—bGH-50ug/day —

Fic. 3. Effects of infusing bGH at 50 ~g/rat/day for 7 days on body weight gain and tail growth of Hx rats. Infusions were started 12-14 days after Hx and the bGH was given constantly or in pulses, as described in Fig. 1. The sham operated rats were subjected to the same surgical procedures as those that were catherized for intrahepatic infusions but without placement of either a catheter or minipumps. Columns with different letters are

signicficantly different at P< 0.05 or better.

response to the hormone (Fig.4). However, were not completely consistent with the other serum IGF-I levels were significantly elevated in growth parameters measured. In the third experi- the rats given constant infusion of bGH via either ment, the rats that received solvent infusion into the EJV or the HPV (2.5-fold and 2.1-fold, respec- either the HPV or the EJV starting 30 days after tivley). Thus, in this study, serum IGF-I levels pituitary ablation lost about 5 gm in body weight

240 N. L. ScHLEcHTER, L. S. Katz et al.

300

TEPW

250

200

150

100

Tibial Epiphyseal Width (um)

50

200

100

Serum IGF-I (ng/ml)

Ss

WM

Y

SSI

0 EJV HPV EJV HPV Sham Constant Pulsed —bGH-50y g/day ~

Fic. 4. Tibial epiphysial plate widths and serum IGF-I levels in the rats described in the legend of Fig. 3. Columns with different letter superscripts are significantly different from each other at P<0.001.

during the 7-day experimental period (Fig. 5). Constant infusion of hGH into either vein of such Hx rats caused significant increases in body weight gain, but the EJV route was almost 3 times more effective than was the HPV (Fig. 5). Pulsed deliv- ery stimulated significant weight gain only when the hGH was given into the EJV, but it was only about half as effective as was constant infusion into

that vessel.

The data on TEPW measurements with hGH infusion are consistent with the body-weight changes (Fig. 5). Infusion into the EJV was more effective than delivery into the HPV by either mode of delivery, and constant infusion into the EJV or the HPV was significantly more effective than pulsed delivery into either vein.

GH, the Liver and Growth 241

30) Body weight

20

10

Change in Body Weight (gm)

400

300

Tibial Epiphyseal Width (um) >) —) —)

EJV HPV Solvent

Fic. 5.

YN \O

EJV HPV

EJV HPV

Constant Pulsed

—hGH-50ug/day Effects of infusing hGH (50 ~g/rat/day) for 7 days on body weight gain and tibial epiphysial plate width of

rats that were Hx 30 days previously. Otherwise as in the legend in Fig. 1. Columns with different letters are

significantly different at P<0.05 or better.

DISCUSSION

The results presented in this paper show that delivery of a low dose of rat GH directly to the liver via its portal blood supply was no more effective at maintaining growth or elevating serum IGF-I concentration than was infusion at a distant

site, regardless of whether the hormone was given constantly or in pulses (Figs. 1 and 2). These findings were unexpected because the liver is generally thought to be the major source of GH- stimulated IGF-I secretion [2, 3], and intrahepatic infusion should have exposed the organ to a con- centration of GH that was at least 4 times higher

242 N. L. SCHLECHTER, L. S. Katz et al.

than that achieved by infusion into the EJV [see 22]. Administration of a higher daily dose of either bGH or hGH gave results that were even more surprising. Infusion of these hormones into the EJV was significantly more effective at stimulating all of the growth parameters measured than was delivery directly into the liver, regardless of whether they were infused constantly or in pulses. Thus, none of these experiments provide data to support the idea that stimulation of hepatic IGF-I secretion by GH is the major means by which the hormone promotes growth.

Another surprising aspect of our studies was the finding that constant infusion of GH was either as effective (with rGH) or more effective (bGH, hGH) than was pulsatile delivery. Several groups have reported that pulsed infusion [23-25] or frequent injections [26-28] of GH are more effec- tive at restoring growth in Hx rats than is constant infusion. However, in one of these studies [23], pulsatile delivery was not more effective than constant infusion during the first week of the treatment; in other experiments [27, 28], the fre- quency of injections was not related to the magni- tude of growth responses until after S—7 days of treatment. The schedule that we used corresponds to the period when these investigators [23, 27, 28] found that intermittent delivery was not more effective than continuous treatment. Neverthe- less, this does not explain our finding that constant infusion of bGH and hGH was more effective than giving pulses of 1-hr duration, while none of the other studies found this effect during the first few days of treatment, and one of them [25] found the opposite result.

By measuring the flow of air bubbles and solvent segments in the catheters, we determined that the minipumps did perform, on average, as expected in terms of delivering the pulses according to the schedule selected [29]. However, there was con- siderable variation in pulse duration and the inter- pulse interval. We were unsuccessful in determin- ing the pattern of GH in the blood of the infused animals because of difficulties in collecting a suf- ficient number of samples serially. Nevertheless, a major difference between our study and those of the others who infused the GH in pulses was in the duration of the pulses. While our system delivered

the hormone in square-wave pulses of about 1 hr duration, the other groups used pulse intervals of only a few minutes.

In separate experiments we have investigated the effectiveness of constant vs. pulsed delivery of ovine PRL at restoring lactation in rats in which endogenous PRL secretion was suppressed by bro- mocriptine [29]. Constant infusion was again found to be consistently more effective than any of several pulse infusion schedules [29], but the dif- ference was not as striking as we observed here with hGH and bGH. In that experiment [29] and in another study [30] we also found that intrahepa- tic infusion of the PRL was not more effective at restoring lactation than was infusion into the EJV. It is unlikely that downregulation or desensitiza- tion of the hepatic GH receptors can account for the lesser effectiveness of intrahepatic vs intra- jugular delivery of the hormone as constant deliv- ery should be more effective than pulsed infusion at reducing hepatic responsiveness to GH.

When PRL was infused into the HPV of pigeons [15] or rats [31], it was more effective at stimulat- ing growth of target organs of PRL than was infusion of the same dose of the hormone into the EJV. Likewise, infusing insulin into the HPV of diabetic rats was more effective at promoting body growth and normalizing their metabolic status than was delivery into the EJV [32]. It has also been shown that direct delivery of PRL to the liver of bullfrog tadpoles (Rana catesbeiana) was more effective at promoting growth and inhibiting meta- morphosis than was delivery to the kidney or subcutaneously [33]. These results with PRL [15, 31, 33] and insulin [32] indicate that if GH did have a preferential effect on the liver to promote IGF-I secretion and growth, our infusion system should have allowed it to be manifested. Accordingly, it seems likely that direct hepatic mediation of the growth-promoting effects of GH may be of less significance than is generally considered [1, 2]. Thus, the direct growth-promoting effect of GH on peripheral tissues, which has been well established by recent studies [8-14], may be of greater import for growth than are the hepatic effects of the hormone. This conclusion is supported by the data which show that although intrahepatic delivery of bGH was almost as effective as infusion into the

GH, the Liver and Growth

EJV at elevating serum IGF-I levels (Fig. 4), the HPV route was much less effective at stimulating growth responses (Figs. 4 and 5).

This conclusion does not mean that an effect of GH on the liver is unimportant for growth. In- deed, it has been well established that GH can stimulate IGF-I secretion by liver cells both in vivo and in vitro [see 3]. However, the significance of the hepatic receptors for GH and the effects that GH may have on liver functions related to growth via these receptors may be in maintaining hepatic responsiveness to other factors that can stimulte IGF-I secretion, such as insulin [32] and thyroid hormones [34].

ACKNOWLEDGMENTS

We are indebted to the National Hormone and Pitui- tary program of the NIH for the GH preparations and for the antiserum to IGF-I, and to Mr. Richard Lin for preparing the figures. Dr. E.M. Spencer, Children’s Hospital (San Francisco, CA) kindly supplied us with ®5T-labeled IGF-I. This work was supported by NIH Grant HD-14661.

REFERENCES

1 Salmon WD, Daughaday WH (1957) A hormonally controlled serum factor which stimulates sulfate incorporation by cartilage in vitro. J Lab Clin Med 49: 825-834

2 Daughaday WH (1989) A personal history of the origin of the somatomedin hypothesis and recent challenges to its validity. Persp Biol Med 32: 194— Aili

3 Spencer EM (1983) Insulin-like growth factors/ somatomedins. de Gruyter, Berlin

4 Uthne K, Uthne T (1972) Influence of liver resec- tion and regeneration on somatomedin (sulphation factor) activity in sera from normal and hypophysectomized rats. Acta Endocrinol (Copenh) 71: 255-264

5 Schimpff RM, Donnadieu M, Glasinovic JC, War- net JM, Girard F (1976) The liver as source of somatomedin (an in vivo study in the dog). Acta Endocrinol (Copenh) 83: 365-372

6 Schimpff RM, Lebrec D, Donnadieu M (1977) Somatomedin production in normal adults and cir- rhotic patients. Acta Endocrinol (Copenh) 86: 355— 362

7 Schwander JC, Hauri C, Zapf J, Froesch ER (1983) Synthesis and secretion of insulin-like growth factor and its binding protein by the perfused rat liver:

10

11

12

14

15

16

17

18

19

20

21

243

dependence on growth hormone status. Endocrinol- ogy 113: 297-303

Isaksson OGP, Jansson J-O, Gause IAM (1982) Growth hormone stimulates longitudinal bone growth directly. Science 216: 1237-1239

Russell SM, Spencer EM (1985) Local injections of human or rat growth hormone or of purified soma- tomedin-C stimulate unilateral tibial epiphyseal growth in hypophysectomized rats. Endocrinology 116: 2563-2567

Schlechter NL, Russell SM, Greenberg S, Spencer EM, Nicoll CS (1986) A direct effect of growth hormone in rat hindlimb shown by arterial infusion. Am J Physiol 250: E231-E235

Isgaard J, Nilsson A, Lindahl A, Jansson J-O, Isaksson OGP (1986) Effects of local administration of GH and IGF-I on longitudinal bone gowth in rats. Am J Physiol 250: E367-E327

Nilsson A, Isgaard J, Lindahl A, Dahlstrom A, Skottner A, Isaksson OGP (1986) Regulation by growth hormone of number of chondrocytes con- taining IGF-I in rat growth plate. Science 233: 571- 573

Schlechter NL, Russell SM, Spencer EM, Nicoll, CS (1986) Evidence suggesting that the direct growth- promoting effect of growth hormone on cartilage in vivo is mediated by local production of soma- tomedin. Proc Nat’l Acad Sci (USA) 83: 7932-7934 Nilsson A, Isgaard J, Lindahl A, Peterson L and Isaksson O (1987) Effects of unilateral arterial infusion of GH and IGF-I on tibial longitudinal bone growth in hypophysectomized rats. Calcif Tissue Int 40: 91-95

Mick CCW, Nicoll CS (1985) Prolactin directly stimulates the liver in vivo to secrete a factor (syn- lactin) which acts synergistically with the hormone. Endocrinology 116: 2049-2053

Gay VLA (1967) A stereotaxic approach to transau- ticular hypophysectomy in the rat. Endocrinology 81: 1177-1179

Patel DG (1983) Rate of infusion with a minipump required to maintain a normoglycemia in diabetic rats. Proc Soc Exp Biol Med 172: 74-78

Eden S (1979) Age and sex related differences in episodic growth hormone secretion in the rat. En- derinology 105: 555-560

Geschwind II, Li CH (1955) The tibia test for growth hormone. In “The Hypophyseal Growth Hormone, Nature and Actions” Ed by RW Smith, OH Gaebler, and CNH Long, McGraw Hill, New York, pp 28-53

Chiang M, Russell SM, Nicoll CS (1990) Growth- promoting properties on the internal milieu of pre- gnant and lactating rats. Am J Physiol 258: E98- E102

Russell SM (1981) Differential effects of somatosta-

22

23

24

25

26

27

28

244

tin on the release of bioactive and immunoactive rat growth hormone in vitro. Neuroendocrinology 33: 67-72

Goldman H (1966) Catecholamine-induced redis- tribution of blood flow in the unanesthetized rat. Am J Physiol 210: 1419-1424

Clark RG, Jansson J-O, Isaksson O, Robinson CAF (1985) Intravenous growth hormone: growth re- sponses to patterned infusion in hypophysectomized rats. J Endocrinol 104: 53-61

Isgaard J, Carlsson L, Isaksson OGP, Jansson J-O (1988) Pulsatile intravenous growth hormone (GH) infusion to hypophysectomized rats increases in- sulin-like growth hormone messenger ribonucleic acid in skeletal tissues more effectively than con- tinuous GH infusion. Endocrinology 123: 2605-2610 Pampori NA, Agrawal AK, Shapiro BH (1991) Renaturalizing the sexully dimorphic profiles of cir- culating growth hormone in hypophysectomized rats. Acta Endocrinol (Copenh) 124: 283-289 Maiter D, Underwood LE, Maes M, Davenport ML, Ketelslegers JM (1988) Different effects of intermittent and continuous growth hormone admi- nistration on serum somatomedin insulin-like growth factor I and liver GH receptors in hypo- physectomized rats. Endocrinology 123: 1053-1059 Jansson J-O, Albertsson-Wickland K, Eden S, Thorngren KO, Isaksson O (1982) Circumstantial evidence for a role of the secretory pattern of growth hormone in control of body growth. Acta Endocir- nol (Copenh) 99: 24-30

Thorngren K-G, Hansson LI (1977) Effect of admi-

29

30

31

32

33

34

N. L. ScHLecHTER, L. S. Katz et al.

nistration frequency of growth hormone on longitu- dinal bone growth in the hypophysectomized rat. Acta Endocrinol (Copenh) 84: 497-511

Hebert NJ, Kim JM, Lin RJ, Nicoll CS (1992) Restoration of lactation in bromocriptine-treated rats by prolactin replacement: Comparison of con- stant vs pulsatile infusion and intrahepatic vs intra- jugular routes of delivery. J Endocrinol Invest 16: 29-35

Hebert NJ, Kim JH and Nicoll CS (1991) On the possible role of the liver in the galacto-poretic action of prolactin in the rat. Endocirnology 128: 1505— 1510

English DE, Russell SM, Katz LS, Nicoll CS (1990) Evidence for a role of the liver in the mammatrophic action of prolactin. Endocrinology 126: 2252-2256

Griffen SC, Russell SM, Katz LS, Nicoll CS (1987) Insulin exerts metabolic and growth-promoting effects by a direct action on the liver in vivo: clarification of the functional significance of the portal vascular link between the B cells of the pancreatic islets and the liver. Proc Nat’! Acad Sci (USA) 84: 7300-7304

Delidow BC, Baldocchi RA, Nicoll CS (1988) Evi- dence for hepatic involvement in the regulation of amphibian development by prolactin. Gen Comp Endocrinol 70: 418-425

Ikida T, Fujiyama K, Hoshiro T, Tanaka Y, Takeuchi T, Mashiba M, Tominaga M (1991) Sti- mulating effect of thyroid hormone on insulin-like growth factor-I release and synthesis by perfused rat liver. Growth Reg 1: 39-45

ZOOLOGICAL SCIENCE 10: 245-256 (1993)

Isolation and Some Characterization of Vitellogenin and Its Related Egg Yolk Proteins from Coho Salmon (Oncorhynchus kisutch)

AxktHiko Hara!, Craic V. SULLIVAN? and WALTON W. DicKHoFF**

'Nanae Fish Culture Experimental Station, Faculty of Fisheries, Hokkaido University, Nanae, Kameda, Hokkaido 041-11, Japan, *Department of Zoology, North Carolina State University, Raleigh, North Carolina 27695-7617, *School of Fisheries, University of Washington, Seattle, Washington 95195, ‘National Marine Fisheries Service, Northwest Fisheries Center 2725 Montlake Blvd. East, Seattle,

Washington 98112, USA

ABSTRACT— Vitellogenin (Vg) and its related egg protein 1 (E1) and egg protein 2 (E2) were isolated from serum or eggs of mature female coho salmon by precipitation in distilled water followed by chromatography on Sepharose 6B (Vg) or Sephadex G-200 (yolk proteins). The coho salmon proteins reacted specifically with respective antisera raised against Vg (a-Vg), El (a-E1) or E2 (a-E2) purified from chum salmon. Two female-specific proteins were identified in serum from mature coho salmon. Coho salmon Vg had an aparent molecular weight of 540 kDa after chromatography on Superose 6, appeared as a major 240 kDa band and a minor 165 kDa band in SDS-PAGE, and resolved into a major 165 kDa band and several minor 70-150 kDa bands after SDS-PAGE under reducing conditions. It reacted immunologically with a-E1 and a-E2. The other female-specific serum protein, designated coho salmon E2, reacted with a-E2 but not with a-E1. The apparent molecular weight of purified E1 and E2 were 230 and 35 kDa, respectively, after chromatography on Superose 6. El appeared as two main bands of 150 and 92 kDa in SDS-PAGE which resolved into two smaller bands (92 kDa and 20 kDa) after reduction. E2 appeared as a 30 kDa band in SDS-PAGE and as a 15 kDa band after reduction. The above immunological and biochemical characteristics and subunit structure of coho salmon Vg, E1, and E2 were found to be nearly identical to the corresponding proteins in several other salmonid species

© 1993 Zoological Society of Japan

of diverse genera. These properties of Vg have been highly conserved during salmonid evolution.

INTRODUCTION

Vitellogenin (Vg) has been well-characterized in avian and amphibian species as a precursor for egg yolk. Hepatic synthesis of Vg is induced by estrogen in maturing females. The protein is released into the bloodstream from where it is taken up by developing oocytes and chemically modified in the process of yolk formation (for review, [24, 34, 36, 38]). For example, in the amphibian, Xenopus, Vg is a lipoglycophospho- protein complex of precursors to several individual egg yolk proteins including lipovitellin, phosvitin

Accepted November 13, 1992 Received September 7, 1992 " To whom all correspondence should be addressed.

and phosvettes [39].

In teleost fish, there are many reports of the identification of Vg or Vg-like proteins using im- munological, electrophoretical and chromatog- raphical methods [1, 3-5, 7-12, 25, 26, 32, 33, 35]. However, there are few reports on the specific biochemical relationship between serum Vg and the egg yolk proteins derived from it. We reported previously that salmonid Vg (female-specific serum protein) from rainbow trout, Oncorhynchus mykiss [16], chum salmon, O. keta [15] and white- spotted char, Salivelinus leucomaenis [17], is a precursor to at least two egg yolk proteins, egg protein one (E1) and egg protein two (E2). El was identified as a lipovitellin similar to that of Xeno- pus based on its physicochemical properties such as molecular weight, amino acid composition, sub-

246 A. Hara, C. V. SULLIVAN AND W. W. DICKHOFF

unit structure and lipid content. Though E2 elutes in size exclusion chromatography in the fraction (mol. wt. 25-30kDa) where Markert and Van- stone [22] reported two soluble egg yolk proteins that they termed phosvitin and the £’-component, we could identify only one antigenic protein in this fraction (E2). In addition, we identified two distinct female-specific proteins in serum from mature females or estrogen-treated males and im- mature fish using immunoelectrophoresis with rab- bit antiserum raised against female-specific serum proteins from the species mentioned above [15, 17]. One of the mature female-specific serum proteins was identified as a Vg containing both El and E2 antigenicity. The other appeared to be a Vg fragment, since it was deficient in E1 antigenic- ity but had E2 antigenicity. We concluded that the two female-specific proteins found in the serum of maturing female salmonids represent a complex of E1+E2 (Vg) and a free E2.

The present paper describes the identification of coho salmon (Oncorhynchus kisutch) Vg by im- munological methods using antisera raised in rab- bits against chum salmon Vg, E1, and E2, and the isolation and partial biochemical characterization of coho salmon Vg and its related egg yolk pro- teins.

MATERIALS AND METHODS

Experimental animals, blood and tissue samples

The experimental animals used this study were adult female coho salmon that returned on their spawning migration to the University of Washing- ton experimental fish hatchery where they were sampled for blood and eggs. They were fully mature but had not yet ovulated. Blood samples were collected from the caudal blood vessels using a syringe fitted with a 21 ga needle. The blood samples were held on ice and immediately trans- ported to the laboratory where they were allowed to clot at 0-4°C overnight. The serum was col- lected after the blood samples were centrifuged at 1000 g for 15 min, and it was stored at —20°C until use.

One year-old immature coho salmon of both sexes were used for experiments in which Vg was

induced by estrogen treatment. They were anes-

thetized in a solution (50 mg/1) of tricaine methane sulfonate (Argent Chemical Lab.) buffered to pH 7.0 with sodium bicarbonate and injected intra- muscularly with 1 mg of estradiol-178 per kg of body weight dissolved in 100% propylene glycol. Two weeks after hormone treatment, blood sam- ples were collected from these fish and processed as described above. Ovulated eggs were collected from female coho salmon during their spawning season (November). The eggs were washed twice with 0.9% NaCl and then kept frozen at —20°C until use.

Purification of Vg and egg yolk proteins

Coho salmon Vg and its related egg yolk pro- teins were isolated using our previously reported procedures [16]. Briefly, 2.5 ml of a sample of serum pooled from several mature females was added to 25 ml of ice-cold distilled-deionized water and the mixture was allowed to stand for 30 min at 0°C. A precipitate formed and was centrifuged at 2,500 g for 15 min at 4°C. The pellet was resus- pended in water (25 ml), recentrifuged and dis- solved in 0.5 ml of 0.02 M Tris-HCl buffer, pH 8.0, containing 2% NaCl and 0.1% NaN3. The solution was then applied to a 1.6x60cm gel filtration column of Sepharose 6B (Pharmacia Inc.) equili- brated with the Tris-HCl buffer. The column flow rate was adjusted to 15 ml/hr and 1.9 ml fractions were collected. The protein concentration in the fractions was estimated by absorbance at 280 nm and the one major peak was collected as the purified Vg. The presense of Vg in the fractions was assessed by single radial immunodiffusion [21] using an antiserum raised in rabbits against chum salmon E1 [15].

Egg yolk was collected from ovulated eggs using a syringe fitted with a 22 ga needle. The yolk was centrifuged at 1000Xg for 5 min and the super- natant was collected and added dropwise to a 10 X volume of ice-cold distilled-deionized water. The precipitate that formed was sedimented by centri- fugation at 1000 g for 30 min and the pellet was redissolved in Tris-HCl buffer. This procedure was repeated twice. The final clear solution in Tris-HCl buffer was applied to a 1.660 cm gel filtration column of Sephadex G-200 (Pharmacia

Vitellogenin in Coho Salmon 247

Inc.). The elution buffer, flow rate, fraction volume and assay of protein in the fractions during Sephadex G-200 chromatography were the same as described above for Sepharose 6B chromatogra- phy. The elution pattern yielded two peaks, designated as El and E2 according to our pre- viously defined criteria [15, 17].

Antisera

Rabbit antisera raised against mature female chum salmon serum proteins, Vg and egg yolk proteins (El and E2) were prepared as described previously [15]. An antiserum was also raised in rabbits against the coho salmon egg yolk protein, El, by intradermal injection with purified El (approximately 0.5 mg) emulsified in an equal volume of complete Freund’s adjuvant. Injections were done 4 times at 7-10 day intervals.

Electrophoresis and immunological procedures

Immunoelectrophoresis and single radial im- munodiffusion [21] and double immunodiffusion [28] were performed using 1.2% agarose gels as described previously [15]. The concentration of serum Vg was determined by single radial im- munodiffusion using the antiserum to coho salmon E1 and the purified coho salmon Vg as the refer- ence standard protein. Discontinuous polyacryla- mide gradient (3.75-18.75%) gel electrophoresis (PAGE) was carried out with (SDS-PAGE) or without (Disc PAGE) addition of SDS to the gel and samples as described by O'Farrell [27]. Prepa- ration of samples for SDS-PAGE was done as described previously [16]. The marker proteins used for molecular weight determination were lactate dehydrogenase (subunit, 36kDa), egg albumin (45kDa), catalase (subunit, 60 kDa), bovine serum albumin (67 kDa), ferritin (subunits, 18.5 and 220 kDa) and thyroglobulin (subunit, 330 kDa).

FPLC chromatography

Chromatography was performed on a Pharmacia FPLC system using a prepacked Superose 6 HR 10/30 column (Pharmacia Inc.). The column was used for determination of the molecular weights of the purified proteins and for observation of the elution pattern of serum proteins. Serum samples

were diluted 10 xin Tris-HCl buffer and 200 pl of the solution, or about 6 mg of purified protein in 200 41 buffer, was applied to the column and eluted with the same buffer with a flow rate of 0.5 ml/min. The column was calibrated with the marker proteins: trypsinogen (24 kDa), egg albu- min (45 kDa), bovine serum albumin (67 kDa), a-amylase (200 kDa), apoferritin (443 kDa) and thyroglobulin (669 kDa).

Comparative studies on salmonid Vg, El and E2

Blood serum and eggs were collected as de- scribed above from mature females of 10 species of salmonids from the genera Salmo, Oncorhynchus, Salvelinus (SI.), and Hucho, which were obtained during annual spawning operations at various national, prefectural, university and commercial fish hatcheries in Hokkaido, Japan. These species included sockeye salmon (O. nerka), masu salmon (O. masou), amago salmon (O. rhodurus), rain- bow trout, brown trout (S. trutta), Japanese huchen (H. perryi), brook trout (S/. fontinalis), dolly varden (S/. malma), white-spotted char (SI. leucomaenis), and lake trout (S/. namaycush). The egg yolk proteins, El and E2, were isolated as described above from the eggs of each of these species. The mature female serum, E1 and E2 of each species was reacted against the a-Vg, a-El and a-E2 antisera generated against the chum salmon proteins, and two additional a-Vg antisera generated against rainbow trout or spotted char vitellogenin [15, 16], in double immonodiffusion assays and immunoelectrophoresis as described above.

RESULTS

Identification of coho salmon Vg

Immunological methods were used for iden- tification of coho salmon Vg. Figure 1 shows immunoelectrophoresis of serum from vitellogenic female coho salmon using the four antisera against chum salmon proteins as follows: antiserum against mature female chum salmon serum pro- teins (a-S), specific antiserum to chum salmon Vg (a-Vg), and specific antisera to chum salmon egg yolk proteins El (a-E1) and E2 (a-E2). With the

248 A. Hara, C. V. SULLIVAN AND W. W. DiICKHOFF

see. a9

2 @eVg

_ a-E1

Fic. 1. salmon.

aE 2 =

Immunoelectrophoresis of mature female coho salmon serum using four different antisera against chum a-S: antiserum to mature female chum salmon serum proteins, a-Vg: antiserum to chum salmon

vitellogenin, a-E1: antiserum to egg yolk protein one (lipovitellin) of chum salmon, a-E2: antiserum to egg yolk protein two (phosvitin/ 8 -component fraction) of chum salmon.

exception of a-S, all the antisera reacted only with serum from vitellogenic females but not with male serum (data not shown). The a-Vg antiserum produced two distinct precipitin lines and the two antisera against egg yolk proteins (a-E1 and a-E2) each produced one precipitin line when reacted against serum from mature female coho salmon in immunoelectrophoresis. Based on our previous studies of salmonid Vg from rainbow trout and chum salmon [13] and white-spotted char [17], the two precipitin lines formed against mature coho salmon female serum by the a-Vg antiserum were considered to be Vg (the inner line formed near antigen well) and a fragment of Vg (the outer line formed near the antiserum trough), respectively. The Vg concentration in serum from mature but unovulated female coho salmon, assayed by single radial immunodiffusion, ranged between 4.68 mg/ ml and 15.91 mg/ml (average 13.0 mg/ml).

Purification of Vg and its related egg yolk proteins

The elution pattern in Sepharose 6B chroma- tography of protein from the water-insoluble frac- tion of coho salmon serum is shown in Fig. 2. The main peak was found to consist of Vg when assayed by single radial immunodiffusion using the a-El antiserum. A single and symmetrical peak

was obtained and collected as the purified coho salmon Vg. After immunoelectrophoresis, this preparation of Vg gave rise to only one precipitin line when reacted against the a-S and a-Vg anti- sera, as shown in Fig. 3. As shown in Fig. 4, the elution pattern of the water-insoluble fraction of coho salmon egg extracts consisted of two distinct peaks, designated E1 and E2, after gel filtration on Sephadex G-200. The relative protein concentra- tions of El and E2, determined planimetrically from charts of gel filtration monitored by absorp- tion at 280 nm, were found to be 13.3 and 1.0 respectively. When adjusted for the estimated molecular weights of El and E2 (discussed below), the apparent molar ratio of El to E2 is 2 to 1. The peak fractions were tested for their reactivity to the a-Vg, a-El and a-E2 antisera by immunodiffu- sion. The fraction designated coho salmon E1 reacted with the a-Vg antisera and a-E1 antisera, but not with antiserum a-E2. The fraction desig- nated coho salmon E2 reacted with the a-Vg and a-E2 antisera but not with antiserum a-El1 (data not shown). After Disc-PAGE, the preparation of coho salmon Vg displayed a major band and a secondary band which migrated faster than the major band as shown in Fig. 5 (lane e). Some faint bands that migrated more slowly than the major

Vitellogenin in Coho Salmon

1.5

1.0

A2gonm

0.5

0 Nop Goin O

O 40 50 Tube number

Fic. 2. Elution pattern of the water-insoluble fraction of coho salmon serum chromatographed on Sepharose 6B. The fractions pooled from tube number 43-48 were collected and concentrated as coho salmon vitellogenin for

further analysis.

== Fic. 3. Immunoelectrophoresis of purified coho salmon vitellogenin. S: mature female coho salmon serum, Vg: purified coho salmon vitellogenin, a-Vg and a-S: same as in Fig. 1.

250 A. Hara, C. V. SULLIVAN AND W. W. DICKHOFF

30 20 1.0 EAL £ Cc oO (oe) WN < 05 0 30.40

50 6 Tube number

E2

70

Fic. 4. Chromatography on Sephadex G-200 of water-insoluble proteins isolated from coho salmon eggs. Two major

peaks were designated as El and E2.

band were also observed. The major band of purified Vg migrated to the position corresponding to the strong band observed only in serum of mature females (lane b) and estradiol-treated fish (lane d). While coho salmon E1 gave a single distinct band in Disc-PAGE (Fig. 5, lane g), coho salmon E2 gave a rather broad band composed of several sharp lines (Fig. 5, lane h). These results are essentially identical to those we reported for chum salmon [15].

In order to compare antigenicity between the three purified coho salmon proteins, double im- munodiffusion was performed in 1.2% agarose gels using the a-Vg antiserum. The precipitin line from coho salmon Vg formed a spur over the precipitin lines of the two egg yolk proteins, El and E2, and the precipitin lines of each egg yolk

protein crossed each other (data not shown), con- firming that the coho salmon Vg molecule contains both E1 and E2 antigenicity.

Molecular weight of purified proteins

As shown in Fig. 6, the molecular weights of the purified coho salmon proteins, estimated by FPLC chromatography on the Superose 6 column, were 540 kDa for Vg, 230 kDa for El, and 35 kDa for E2. The electrophoretic patterns of coho salmon Vg, El and E2 after SDS-PAGE + reduction with 2-mercaptoethanol (2-ME) are shown in Fig. 7. Vg displayed one main band and a minor band coresponding to molecular weights of 240 kDa and 165 kDa, respectively (lane C). Vg reduced with 2-ME showed a main band with a molecular weight of 165 kDa as well as some minor bands of lower

Vitellogenin in Coho Salmon 251

h

+

Fic. 5.

Gradient Disc-PAGE of purified vitellogenin, E1 and E2 and egg yolk extracts from coho salmon. a: serum

proteins from mature male, b: serum proteins from mature female, c: serum from immature male (control fish from estrogen-treatment experiment), d: serum from estrogen-treated immature fish, e: vitellogenin, f: egg yolk

extracts, g: El, h: E2.

(70 to 150kDa) molecular weight (lane c). El showed a main band corresponding to a molecular weight of about 150 kDa and minor 92 kDa band, as well as a faint band migrating near the dye front (lane D). After reduction with 2-ME, El appeared as two main 92 kDa and 20 kDa bands, as well as some faint bands of intermediate (70 kDa to 90 kDa) molecular weight. E2 appeared in SDS-PAGE as a single 30 kDa band before reduc- tion, and as a single 15 kDa band after reduction with 2-ME. These results were very similar to those observed Vg, El and E2 of rainbow trout and chum salmon [15].

Serum elution profiles in FPLC

Typical patterns of coho salmon serum after

chromatography on Superose 6 are shown in Fig. 8. Distinct differences between the sexes were seen in the patterns. The major peak from female serum eluting at effluent volume 13.9 ml (panel b) was not seen in male serum (panel a). This peak was also induced in immature fish by treatment with estradiol-17 (panel c). Purified Vg (see Fig. 6) eluted the same position (arrow, Fig. 8) as the mature female specific and estradiol-inducible peak.

Cross-reactivity of antisera to chum salmon with other salmonid fish The elution patterns on Sephadex G-200 of El

and E2 from the 10 species of salmonids were nearly identical to that seen for coho salmon (data

252 A. Hara, C. V. SULLIVAN AND W. W. DIcCKHOFF

Fic. 6. Estimation of the molecular weight of coho salmon vitellogenin by Supurose 6 gel filtration in the FPLC system. The open circles correspond to the elution volume of purified vitellogenin (Vg), E1 and E2. The marker proteins used for calibration and indicated by the dark circles were, 1: trypsi- nogen, 2: egg albumin, 3: bovine serum albumin, 4: a-amylase, 5: apoferritin, and 6: thyroglobulin.

104 102 OE Mol. wt.

ABCD E abed ie

+—165K 150K

— 92K

230K ~— 20K K

—— 1

Fic. 7. SDS-PAGE on 3.75-18.75% gradient slab gels of mature male serum (A, a), female serum (B, b),

vitellogenin (C, c), El (D, d), and E2 (E, e) in coho salmon. Lane A-E represent non-reduced samples and lanes a-€ represent samples reduced with 2-mercaptoethanol.

Vitellogenin in Coho Salmon

A280nm

C

5

10 15 20

iw) Nn Oo

; | = ae

25

Effluent volume (ml)

Fic. 8. Elution profiles of mature male (a) and female serum (b), and estrogen-treated male serum (c) in coho salmon on Supurose 6 in the FPLC system. The arrow indicates the elution position of vitellogenin.

not shown). As noted above, the a-Vg, a-E1 and a-E2 antisera raised against chum salmon Vg, El and E2, respectively, reacted strongly and speci- fically to the respective coho salmon proteins. In addition, we found that these antisera, as well as two other a-Vg antisera raised against Vg from rainbow trout and white-spotted char [15, 16], yielded a pattern of immunoreactivity identical to that reported above for coho salmon when they were reacted in double immunodiffusion assays and immunoelectrophoresis against the mature female serum, E1 and E2, from the 10 salmonid species. For each species, two mature female- specific serum proteins were seen in immunoelec- trophoresis. One reacted with all of the antisera tested and the other reacted only with the a-Vg and a-E2 antisera. E1’s from the various species

reacted with all three a-Vg’s and with a-E1, but not with a-E2. E2’s from the various species reacted with all three a-Vg’s and with a-E2, but not with a-E1.

DISCUSSION

In the present study, coho salmon Vg was purified within one day from the serum of mature females by precipitation with distilled water fol- lowed by gel filtration on Sepharose 6B. When the gel filtration step was done on a Supurose 6 column using the FPLC system, complete purification of Vg was accomplished within 3 hours, but yields were much reduced (data not shown). The pre- cipitation step with distilled water is a very simple, easy and effective method for purification of sal-

254 A. Hara, C. V. SULLIVAN AND W. W. DICKHOFF

monid Vgs, though its success depends on the starting conditions of the serum and its concentra- tion of Vg. For example, precipitation of coho salmon Vg was not possible using serum containing Vg concentrations below about 5 mg/ml or using serum that was not previously frozen. We have used hydroxylapatite column chromatography to purify Vg from salmonid serum with low Vg con- centrations that cannot be precipitated with distil- led water, and the purified protein is similar in all respects to that obtained using the precipitation method [14-17].

So far, the detection of teleost Vg has been performed primarily by immunological, elec- trophoretical and chromatographical procedures. In the present study, in addition to immunological methods, we identified Vg in serum of mature females and estrogen-treated males and immature fish using a Supurose 6 column in an FPLC chro- matography system. This method required only one drop of serum for assay and the procedure is simple and rapid. Silversand and Haux [31] recent- ly reported a simiar method for purification of turbot Vg, using a Mono Q column in an FPLC system. Though further studies will be needed to confirm the relation between the concentration of Vg and its peak height or area in FPLC or HPLC chromatography, the method should prove useful for rapid routine assay of fish Vgs.

Vertebrate Vg has been generally confirmed to be the precursor of the major yolk proteins, lipo- vitellin and phosvitin. Rainbow trout egg yolk proteins can be chromatographically separated into three major components [20]. Phosvitin is a phosphoprotein rich in serine [5, 19, 30], soluble in trichloroacetic acid [5, 30, 37], and poorly stained with Coomassie Blue on electrophoretic gels [18]. Lipovitellin is the main yolk protein of oviparous vertebrates, and it contains a high concentration of lipids and little or no protein phosphorus. In rainbow trout, the lipovitellin has been reported to consist of two subunits of mol. wt. 90 kDa and 15 kDa [16] or 95 and 24 kDa [6] or 92 and 20 kDa [1], or to consist of four subunits from 145-160 kDa (n=2) and from 14-19 kDa (n=2) which can easily dimerize [29]. A third yolk protein, the 8’-component [20], is characteristic of the yolk of salmonid eggs [22, 23]. Its amino acid composition

is distinct from those of lipovitellin and phosvitin [5]. It can escape precipitation when the yolk is diluted with distilled water [20, 22]. Though the origin of the 8’-component remains to be conclu- sively verified, it has been suggested that it is derived from Vg like lipovitellin and phosvitin [38]. In the present study, the water-insoluble egg yolk proteins of coho salmon were separated into two components, namely, El and E2. The molecular weight of E1 and E2 were estimated by gel filtration to be 230 kDa and 35 kDa, respec- tively. Using SDS-PAGE without reduction by 2-ME, the molecular weights of El and E2 were estimated to be 150 kDa and 30 kDa, respectively. Upon reduction with 2-ME, the 150 kDa compo- nent of E1 further split two major polypeptides with molecular weights of 92 and 20 kDa. The 30 kDa E2 split into a single component with a molecular weight of 15 kDa, suggesting that it may be a homodimer. These results are essen- tially identical to those that we previously reported for rainbow trout egg yolk proteins [16].

Immunological methods using antisera to chum salmon Vg, El and E2 revealed two female- specific proteins in the serum of mature coho salmon. One (Vg) was, in part, a complex of El+ E2. The observation that coho salmon Vg con- tained both the E1 and E2 components of egg yolk proteins is similar to our previously reported re- sults on other salmonid species [15]. The second female-specific serum protein contained only E2 antigenicity. It did not react with antiserum to El. A relation of these two serum proteins to the coho salmon £’-component and/or phosvitin [23] re- mains to be established. Our results suggest that an antiserum raised against purified El would be preferable to an antiserum to Vg for measurement of Vg concentrations in serum.

Chen et al. [7] developed a rapid sensitive assay for rainbow trout Vg using an antiserum against lipovitellin in rocket immunoelectrophoresis. They reported that their antibody against rainbow trout lipovitellin showed a continuous precipita- tion arc when reacted against coho salmon lipo- vitellin in Ouchterlony immunodiffusion and Western immunoblotting assays. These results support our observation that antisera to chum salmon Vg, El and E2 strongly react to the

into

Vitellogenin in Coho Salmon

respective coho salmon proteins. We also found that the antiserum to chum salmon Vg, E1 and E2, react fully and specifically with the corresponding proteins of various other species from diverse salmonid genera, including sockeye salmon, masu salmon, amago salmon, rainbow trout, brown trout, Japanese huchen, brook trout, dolly varden, white-spotted char, and lake trout. Benfey et al. [2] also showed that plasma from various salmonid species exhibited parallelism in a coho salmon Vg radioimmunoassay. These results suggest that affinity chromotography on a column containing antibodies against the proteins purified from one salmonid species should be a viable approach to purifying Vg from a variety of other salmonid fishes. They also indicate that Vg and its related egg proteins have been highly conserved during salmonid evolution.

ACKNOWLEDGMENTS

We would express our gratitude to C. V. W. Mahnken, National Marine Fisheries Service, Northwest Fisheries Center, Seattle, Washington, for facilities and finacial support, and to Drs. E. Brannon and W. K. Hershber- ger, School of Fisheries, University of Washington, Seattle, Washington, for access to fish and samples. Appreciation is extended to Drs. E. Plisetskaya and P. Swanson, School of Fisheries, University of Washington, for their kind advice and valuable suggestions, to M. G. Bernard and Dr. L. Yan for help with fish sampling and care, and to Dr. K. Takano for help collecting samples from the various species of salmonids. This investigation was supported by grant from the National Science Foundation (DCB 8615521) to W.W.Dickhoff, by a grant from the International Exchange Programs Com- mittee of Hokkaido University, Sapporo, Japan, to C.V.Sullivan, and by a Grant-in-Aid for Scientific Re- search from the Japanese Ministry of Education, Science and Culture (01440014) to A. Hara.

REFERENCES

1 Babin PJ (1987) Apolipoproteins and the associa- tion of egg yolk proteins with plasma high density lipoproteins after ovulation and follicular atresia in the rainbow trout (Salmo gairdneri). J Biol Chem 262: 4290-4296

2 Benfey TJ, Donaldson EM, Owen TG (1989) An homologus radioimmunoassay for coho salmon (Oncorhynchus kisutch) vitellogenin, with general applicability to other Pacific salmonids. Gen Comp Endocrinol 75: 78-82

3

11

12

13

14

15

16

255

Bradley JT, Grizzle JM (1989) Vitellogenin induc- tion by estradiol in channel catfish, Ictalurus punc- tatus. Gen Comp Endocrinol 73: 28-39

Campbell CM, Idler DR (1976) Hormonal control of vitellogenesis hypophysectomized winter flounder (Pseudopleuronectes americanus Walbaum). Gen Comp Endocrinol 28: 143-150

Campbell CM, Idler DR (1980) Characterization of an estradiol-induced protein from rainbow trout as vitellogenin by the cross reactivity to ovarian yolk fractions. Biol Reprod 22: 605-617

Chen TT (1983) Identification and characterization of estrogen-responsive gene products in the liver of rainbow trout. Can J Biochem Cell Biol 61: 802-810 Chen TT, Reid PC, Van Beneden R, Sonstegard RA (1986) Effect of Aroclor 1254 and Mirex on estradiol-induced vitellogenin production in juvenile rainbow trout (Salmo gairdneri). Can J Fish Aquat Sei 43:169-173

Copeland PA, Sumpter JP, Walker TK, Croft M (1986) Vitellogenin levels in male and female rain- bow trout (Salmo gairdneri Richardson) at various stages of the reproductive cycle. Comp Biochem Physiol 83B: 487-493

Covens M, Covens L, Ollevier F, de Loof A (1987) A comparative study of some properties of vitel- logenin (Vg) and yolk proteins in a number of freshwater and marine teleosts. Comp Biochem Physiol 88B: 75-80

de Vlaming VL, Wiley HS, Delahuntry G, Wallace RA (1980) Goldfish (Carassius auratus) vitel- logenin: Induction, isolation, properties and rela- tionship to yolk proteins. Comp Biochem Physiol 67B: 613-623

Emmersen BK, Petersen IM (1976) Natural occurr- ence and experimental induction by estradiol 17-2, of a lipophophoprotein in flounder (Platichthys flesus L.). Comp Biochem Physiol 54B: 443-446 Fletcher PE, Fletcher GL (1980) Zinc- and copper- binding proteins in the plasma of winter flounder (Pseudopleuronectes americanus). Can J Zool 58: 609-613

Hara A (1976) Iron-binding activity of female- specific serum proteins of rainbow trout (Salmo gairdneri) and chum salmon (Oncorhyncus keta). Biochim Biophys Acta 427: 549-557

Hara A (1978) Sexual differences in serum proteins of chum salmon and the purification of female- specific serum protein. Bull Jap Soc Sci Fish 44: 689-693

Hara A (1987) Studies on female-specific serum proteins (vitellogenin) and egg yolk proteins in teleosts:Immunochemical, physicochemical and structural studies. Mem Fac Fish Hokkaido Univ 34: 1-59

Hara A, Hirai H (1978) Comparative studies on

17

18

19

20

21

22

23

24

25

26

27

28

256

immunochemical properties of female-specific serum protein and egg yolk proteins in rainbow trout (Salmo gairdneri). Comp Biochem Physiol 48B: 389-399

Hara A, Matsubara T, Saneyashi M, Takano K (1984) Vitellogenin and its derivatives in egg yolk proteins of white-spotted char (Salivelinus leuco- maenis). Bull Fac Fish Hokkaido Univ 35: 144-153 Hegenauer J, Ripley L, Nace G (1977) Staining acidic phosphoproteins (phosvitin) in electrophore- tic gels. Anal Biochem 78: 308-311

Ito Y, Fujii T, Kanamori M, Hattori T, Yoshioka R (1966) Phosvitin of the trout roe. J Biol Chem 60: 726-728

Jared DW, Wallace RA (1968) Comparative chro- matography of the yolk proteins of teleosts. Comp Biochem Physiol 24: 437-443

Mancini G, Carbonara AO, Heremans JF (1965) Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry 2: 235- 254

Markert JR, Vanstone WE (1968) Egg proteins of Pacific salmon, genus Oncorhynchus : Proteins in the soluble fraction. Can J Biochem 46: 1334-1337 Markert JR, Vanstone WE (1971) Egg proteins of coho salmon (Oncorhynchus kisutch): chromatog- raphic separation and molecular weights of the major proteins in the high density raction and their presence in salmon plasma. J Fish Res Board Can 28: 1853-1856

Mommsen TP, Walsh PL (1988) Vitellogenesis and oocyte assembly. In “Fish Physiology Vol XIA” Eds by WS Hoar, DJ Randall, Academic Press, New York, pp 347-406

Nath P, Sundararaj BI (1981) Isolation and iden- tification of female-specific serum lipophosphopro- tein (vitellogenin) in the catfish, Heteropneustes fossilis. Gen Comp Endocrinol 43: 184-190

Ng TB, Idler DR (1983) Yolk formation and differentiation in teleost fishes. In “Fish Physiology Vol IXA” Eds by WS Hoar, DJ Randall, Academic Press, New York, pp 373-404

O'Farrell PH (1975) High resolution two dimen- sional electrophoresis of proteins. J Biol Chem 250: 4007-4021

Ouchterlony O (1953) Antibody-antigen reactions in gels. IV. Types of reactions in coordinated sys- tems of diffusion. Acta Path Microbiol Scand 32: 231-240

29

30

31

32

33

34

35

36

37

38

39

A. Hara, C. V. SULLIVAN AND W. W. DICKHOFF

Riazi A, Fremont L, Gozzelino MT (1988) Charac- terization of egg yolk proteins from rainbow trout, Salmo gairdneri (Rich). Comp Biochem Physiol 89B: 399-407

Schmidit G, Bartsch G, Kitagawa T, Fujisawa K, Knolle J, Joseph J, De Marco P, Liss M, Has- chemeyer R (1965) Isolation of a protein of high phosphorus content from the eggs of brown brook trout. Biochem Biophys Res Commun 18: 60-65 Silversand C, Haux C (1989) Isolation of turbot (Scophthalmus maximus) vitellogenin by high- performance anion-exchange chromatography. J Chromatography 478: 387-397

So YP, Idler DR, Hwang SJ (1985) Plasma vitel- logenin in landlocked Atlantic salmon (Salmo salar Quananche): Isolation, homologous radioimmu- noassay and immunological cross-reactivity with vitellogenin from other teleosts. Comp Biochem Physiol 81B: 63-71

Sullivan CV, Tao Y, Hodson, R.G, Hara A, Ben- nett RO, Woods III LC (1991) Vitellogenin and vitellogenesis in striped bass (Morone saxatilis) broodstock. In “Proceedings of the Fourth Interna- tional Symposium on the Reproductive Physiology of Fish”. Eds by AP Scott, JP Sumpter, DE Kime, MS Rolfe, Sheffield, U.K. pp 315-317

Tyler C (1991) Vitellogenesis in salmonids. In “Proceedings of the Fourth International Sympo- sium on the Reproductive Physiology of Fish”. Eds by AP Scott, JP Sumpter, DE Kime, MS Rolfe, Sheffield, U.K. pp 295-299

Wallace RA (1978) Oocyte growth in nonmamma- lan vertebrates. In “The Vertebrate Ovary”. Ed by RE Jones, Plenum Press, New York, pp 469-502 Wallace RA (1985) Vitellogenesis and oocyte growth in nonmammalian vertebrate. In “Develop- ment Biology Vol 1” Ed by LW Browder, Plenum Press, New York, pp 127-177

Wallace RA, Jared DW, Eisen AZ (1966) A general method for the isolation and purification of phosvitin from vertebrate eggs. Can J Biochem 44: 1647-1655

Wiegard MD (1982) Vitellogenesis in fishes. In “Reproductive Physiology of Fish”. Eds by CJJ. Richter, HJT Goos, Wageningen, pp 136-146 Wiley HS, Wallace RA (1981) The structure of vitellogenin: Multiple vitellogenins in Xenopus laevis give rise to multiple forms of the yolk pro- teins. J Biol Chem 256: 8626-8634

ZOOLOGICAL SCIENCE 10: 257-265 (1993)

Structure and Function of the Molluscan Myoactive Tetradecapeptides

ARATA Harapa!, MASAYUKI YosHIDA!, HiroyukKI MINAKATA~ KyosukE Nomoto*, Yourro MuNEoKA! and Makoto KosayasuHt *

'Physiological Laboratory, Faculty of Integrated Arts and Sciences, Hiroshima University, Hiroshima 730, and Suntory Institute for Bioorganic Research, Osaka 618, Japan

ABSTRACT—Effects of myoactive tetradecapeptides, Achatina excitatory peptide 2 and 3 (AEP2 and AEP3) and Fusinus excitatory peptide 4 (FEP4), on several molluscan muscles and neurons were investigated. In the penis retractor and radula retractor muscles of Achatina fulica (pulmonate), the three peptides enhanced the tetanic contraction elicited by nerve stimulations. The order of potency was AEP2>AEP3>FEP4, although the effects of AEP3 and FEP4 on the radula retractor were somewhat irregular. AEP2 also induced rhythmic bursts of activity in the buccal ganglionic neuron B4 known as a cholinergic motoneuron of the radula retractor. In the radula protractor and retractor muscles of Fusinus ferrugineus (prosobranch), FEP4 was most potent in enhancing the contraction. The

© 1993 Zoological Society of Japan

enhancement was greater in the protractor than in the retractor.

It was suggested that myoactive

tetradecapeptides modulate mainly the cholinergic transmission in molluscan muscles.

INTRODUCTION

It has been known that the contraction of mol- luscan muscles is regulated by multiple bioactive substances including neuropeptides [11]. In the African giant snail Achatina fulica (pulmonate, mollusc), more than twenty neuropeptides have been purified from the ganglia and hearts [8], and the function of some of them has already been investigated [1, 5]. Two bioactive tetradecapepti- des which are highly homologous in structure were isolated from the ganglia of this snail [10]. Fur- ther, they appeared to be very close to another tetradecapeptide purified from the ganglia of a prosobranch Fusinus ferrugineus. During the pro- cess of purification of these peptides, they were found to be active in enhancing the contraction of a few muscles of A. fulica [10]. They were named myoactive tetradecapeptides (MATPs) [10].

In the present study, actions of MATPs on

Accepted November 19, 1992 Received October 15, 1992 " To whom reprint requests should be addressed.

several molluscan muscles and neurons were ex- amined and the mechanisms of their action on the neuromuscular systems were studied.

MATERIALS AND METHODS

Two Achatina tetradecapeptides and a Fusinus tetradecapeptide were isolated from the ganglia of each animal by using the penis retractor muscle of A. fulica and the radula retractor muscle of F. ferrugineus as a bioassay system, respectively. These isolation procedures have been described previously [10].

Peptides used in the present experiments were synthesized by a solid-phase peptide synthesizer and purified by HPLC. Their structures were confirmed by amino acid sequence analysis and fast atom bombardment mass spectrometric (FAB- MS) measurement.

Bioactivities of the synthesized peptides were examined on the penis retractor muscle, the radula retractor muscle and the buccal ganglionic neurons of A. fulica, and the radula protractor and retrac- tor muscles of F. ferrugineus and of Rapana thoma-

258

siana (prosobranch). Physiological saline used for the preparation from A. fulica consisted of (mM): NaCl 61.0, KCl 3.3, CaCl, 10.7, MgCl 13.0, glucose 5.0 and Hepes-NaOH 10.0 (pH7.5). In some experiments, saline containing 65 mM MgCl, (high-Mg* saline) was used, which was prepared by merely adding MgCl, to normal saline. Com- position of artificial seawater used for the prepara- tion from F. ferrugineus and R. thomasiana was as follows (mM): NaCl 475.0, KCl 10.0, CaCl, 10.0, MgCl, 20.0 and Tris-HCl 10.0 (pH 7.8).

Stimulation of these muscles and recording of the tension were made according to the method previously described by Muneoka and Twarog [9]. Electrical activities of the neurons were recorded by the conventional intracellular microelectrode method [13].

A. HarapbA, M. YosHIDA et al.

RESULTS

Two Achatina peptides were provisionally termed Achatina excitatory peptide 2 (AEP2) and 3 (AEP3), and a Fusinus peptide was termed Fusinus excitatory peptide 4 (FEP4) [10]. The structures of these peptides are as follows:

AEP2: Gly-Phe-Arg-Gln-Asp-Ala-Ala-Ser- Arg-Val-Ala-His-Gly-Tyr-NH>

AEP3: Gly-Phe-Arg-Gly-Asp-Ala-Ala-Ser- Arg-Val-Ala-His-Gly-Phe-NH>

FEP4: Gly-Phe-Arg-Met-Asn-Ser-Ser-Asn-

Arg-Val-Ala-His-Gly-Phe-NH>

In the present paper these terms will also be employed.

Figure 1A demonstrates effects of three kinds of peptides on the tetanic contraction of the penis retractor muscle of A. fulica. The contraction was

@® i = Ay N = J‘ e

10°’M AEP2 10°’M AEP3

ty : ik SM = a

10°7M FEP4 in tt er Oe 4 < o-----0 AEP3 = BMverccese, A FEP4 6 200 , ”) Cc 2 x 150 oO Ae 100

10°8 10:5 Concentration (M)

106

Fic. 1. Effects of three peptides on the teta- nic contraction of the penis retractor muscle of A. fulica. A. The muscle was stimulated with a train of electrical pulses (0.8 msec, 10 V, 40 Hz) for 1 sec at 10 min intervals. Peptide was applied 8 min before applying the stimulation (upward arrow), and washed out immediately af- ter the recording (downward arrow). B. Dose-response relationships of three pep- tides. Peak tension was expressed as percent of the control tension.

Molluscan Myoactive Tetradecapeptides 259

evoked by applying a train of electrical pulses (0.8 msec, 10 V, 40 Hz) for 1 sec to the muscle. AEP2 remarkably enhanced the contraction with a threshold concentration of 10~? M, and at concen- trations more than 10-’M it directly induced a contraction. AEP3 and FEP4 produced similar effects on the penis retractor muscle, although they were slightly less potent than AEP2. Figure 1B illustrates the dose-response relations of the three peptides. As shown in this graph the order of potency in enhancing the muscle contraction was AEP2>AEP3>FEP4.

To test the site of action of the peptides the following experiments were performed (Fig. 2). After confirming that a train of pulses of short duration (0.8 msec, 10 V, 40 Hz) produced a teta- nic contraction in normal saline, it was replaced by high-Mg*" saline. In the latter the contraction by electrical pulses of short duration (0.8 msec) was completely blocked, supporting the assumption that short-duration pulses stimulated the muscle indirectly through intramuscular nerve fibers. When pulse duration was increased to 8.0 msec, however, the tetanic contraction of considerable size could be elicited even in high-Mg*" saline. Such long-duration pulses probably stimulated the muscle fibers directly. AEP2 of 10° M, which potentiated the tetanic contraction by short-

AEP2 108m

KA ae

0.8ms 0.8 08

aot

high-Mg" saline

duration pulses in normal saline, could also po- tentiate the contraction in response to long- duration pulses in high-Mg** saline. Similarly, AEP3 and FEP4 also enhanced the contraction by long-duration pulses in high-Mg** saline and the efficiency of the three peptides was about the same.

In experiments shown in Fig. 3, effects of the three peptides on the tetanic contraction of the radula retractor muscle elicited by a train of pulses of short duration (0.8 msec, 10 V, 40 Hz) were examined. AEP2 enhanced the contraction in a dose-dependent manner with a threshold concen- tration around 10-?-10-*M, and at concentra- tions of 10°-°M or more it directly induced a contraction. However, the effects of AEP3 and FEP4 were quite variable with different prepara- tions, and in some preparations inhibition of con- traction during application and potentiation after washing of the peptide were observed (Fig. 3). These results may imply that AEP3 and FEP4 act on multiple nerve fibers of inhibitory and excita- tory nature.

In applying several kinds of bioactive amine substances such as acetylcholine (ACh), catechol- amines, octopamine, serotonin and histamine on the radula retractor muscle, we have found that only ACh at comparatively lower concentrations

AEP3 FEP4 10°°m 10° 8m 8.0 8.0 8.0 8.0 ]o59 — 20sec

Fic. 2. Effects of three peptides on the tetanic contraction of the penis retractor muscle of A. fulica in normal saline and high-Mg’* saline. The muscle was stimulated with a train of short-duration pulses or long-duration pulses at

10 min intervals.

Soon after recording the second contraction, high-Mg”*

saline was applied (upward arrow).

Procedures for applying peptides were the same as in Fig. 1.

260

HERES

108M 107M 10°°M 20sec AEP2 ieasece fovedios rheafaet f | 1M 10°’M 10°°M fivin fron didn} f | 19M 10°7M 10°°M

elicited a contraction. Further, an identified motoneuron (B4) of the radula retractor muscle has been shown to be cholinergic [13]. Thus, the modulatory effects of peptides on the contraction in response to ACh were investigated. Figure 4A shows contractions evoked by applying ACh dur- ing the period indicated by a horizontal line under each record. AEP2 potentiated the ACh-evoked contraction in a dose-dependent manner, and AEP3 and FEP4 also showed similar effects. The order of efficiency was AEP2 >AEP3 >FEP4 as shown by dose-response curves (Fig. 4B).

In molluscan neuromuscular junctions, prop- antheline has been shown to be a cholinergic blocking agent [12, 13]. In the experiments shown in Fig. 5A and B effects of propantheline on the contraction of radula retractor muscle elicited by ACh and by electrical stimulation were compared. ACh-evoked contraction was extremely depressed by propantheline at a concentration of 10-° M and was completely blocked at 10 *M (Fig. 5A). However, the contraction in response to electrical

A. Harapba, M.

ie

YOSHIDA et al.

.5g

Fic. 3. Effects of three peptides on the teta- nic contraction of the radula retractor muscle of A. fulica. Procedures for sti- mulating the muscle and applying pep- tides were the same as in Fig. 1.

stimulation was inhibited by only 30% of the control during application of 10~*M propanthe- line. To demonstrate that short-duration pulses (0.6 msec) selectively stimulated the intramuscular nerve fibers the preparation was immersed in high-Mg’~ saline. Stimulation of short-duration pulses almost failed to evoke the contraction, but that of long-duration pulses (6.0 msec) elicited a contraction of considerable size, supporting the foregoing assumption. Propantheline of 10°*M did not depress the contraction by long-duration pulses in high-Mg’* saline. From these experi- ments, the contraction elicited by short-duration pulses during application of propantheline in nor- mal saline is considered to be evoked by the stimulation of nerve fibers other than cholinergic ones. Will this contraction be modulated by pep- tides? Experiments shown in Fig. 5C were per- formed to answer this question. AEP2 of 10~’M enhanced the contraction elicited during applica- tion of propantheline by only about 14%, in con- trast to the contraction without propantheline

Molluscan Myoactive Tetradecapeptides

3x10°°M t | : ACh 10°7M 10°°M AEP AEP2 SS - {s ee} 10-5M AEP3 350 <300- o-- ln 2 250 D F 200 x oO 2 150 100

10°" Concentration (M)

40°8

which was enhanced by about 46% by applying the same concentration of AEP2. This seems to imply that AEP2 enhances the contraction of the radula retractor muscle by acting mainly on the cho- linergic transmission.

Among the identified neurons in the buccal ganglia of A. fulica, six pairs have been known as excitatory motoneurons for the radula retractor muscle [13, 14]. Figure 6 shows effects of the peptides on such identified neurons B1 and B4. AEP2 induced rhythmic bursts of activity in both neurons (Fig. 6A), and AEP3 also elicited several bursts, although they were rather irregular (Fig. 6B). However, FEP4 had no significant effects on these neurons.

In experiments shown in Fig. 7, effects of the three peptides on the twitch contraction of the radula protractor and retractor muscles of Fusinus ferrugineus were tested. In these muscles FEP4 showed stronger effects than AEP2 or AEP3 in contrast to their effects on Achatina muscles and

261

Fic. 4. Effects of three peptides on the con- traction of the radula retractor muscle of A. fulica induced by ACh. A. ACh was applied during the period shown by a horizontal line under each record. Proce- dures for applying peptides were the same as in Fig.1. B. Dose-response relationships of three peptides. Peak tension was expressed as percent of the control tension.

neurons. FEP4 enhanced the contraction in a dose-dependent manner with a threshold concen- tration of 107 '° M, and at concentration of 10-° M or more it directly induced a contraction. Similar tendency of effects was observed in AEP2 and AEP3, with AEP3 having weaker effects than the others. However, the efficiency of the peptides was different between these two muscles; in the radula protractor 10” ’ M FEP4 enhanced the con- traction by more than 150% of the control, while in the retractor the enhancement was less than 30% of the control (Fig. 7). AEP2 and AEP3 also produced stronger enhancing effects on the con- traction of the protractor than the retractor. Similar experiments were performed by using the radula protractor and retractor muscles of Rapana thomasiana, since the mode of excitatory transmission has been shown to be different be- tween these muscles, the protractor being mainly cholinergic and the retractor glutaminergic [6, 7]. As demonstrated in Fig. 8, the three peptides

262 A. HarabA, M. Yosuipa et al.

10°°M 10°*M 3X10°&M ACh J 05g 30sec 6) Prop. © 10°*M 10°*M Jt

0.6msec 06 06 , 06 60 60 £60 13V f Joss

40Hz high-Mg*saline 30sec © Prop. 10°*M

t 4 t 4 1 |osa AEP2 AEP2 AEP2

10-7M 10-7M 10°-"M 30sec @)

B1

3X10-°M AEP2 wash

B1

f 3X10-°M AEP3 wash

Fic. 5. Effects of propantheline (Prop) on

the contraction of the radula retractor muscle of A. fulica elicited by ACh (A) and by electrical stimulation (B). Prop- antheline was applied 8 min before giving ACh or stimulation and washed out im- mediately after the record. In A, ACh was applied during the period shown by a horizontal line under each record. In B, the muscle was stimulated with a train of short-duration pulses or long-duration pulses at 10 min intervals. Soon after the third record, high-Mg** saline was ap- plied (upward arrow). C. Effects of AEP2 on the tetanic contraction of the radula retractor muscle of A. fulica dur- ing and without application of propanthe- line. The muscle was stimulated with a train of short-duration pulses (0.6 msec, 13 V, 40 Hz) for 1 sec at 10 min intervals. Propantheline was applied 8 min before the fourth record and washed out after the sixth record. Procedures for applying peptide were the same as in Fig. 1.

Fic. 6. Effects of AEP2 (A) and AEP3 (B)

on the spontaneous activity of identified neurons B1 or B4 in the buccal ganglion.

Molluscan Myoactive Tetradecapeptides 263

@) wulll oN A a all ae

FEP4 AEP2 AEP3 10°7M 10°"M 10°"M

il Mtl ps

10°7M 10°7M 10°7M Fic. 7. Effects of three peptides on twitch contractions of the radula protractor muscle (A) and the retractor muscle (B) of F. ferrugineus. The muscle was stimulated with a train of electrical pulses (2 msec, 15 V, 0.2 Hz, 5 pulses) at 10 min intervals. Procedures for applying peptides were the same as in Fig. 1.

‘ita ada :

AEP3 FEP4 AEP2 10°’M 10°’M 10°’M

UL aah Jose { | fru 4 { } AEP3 FEP4 AEP2. . 1 10°7M 10°"M 10°7M 1 min

Fic. 8. Effects of three peptides on twitch contractions of the radula protractor muscle (A) and the retractor muscle (B) of R. thomasiana. The muscle was stimulated with a train of pulses (1 msec, 15 V, 0.2 Hz, 5 pulses) at 10 min intervals. Procedures for applying peptides were the same as in Fig. 1.

strongly increased twitch contractions of the radula_ _ protractor 10~’ M AEP3 enhanced ACh-induced protractor, but they did not show any enhancing contraction prominently, whereas in the retractor effects on the contraction of the retractor. Finally, it showed no significant effects on ACh-induced as effects of the peptide AEP3 on the contraction well as glutamate-induced contractions (data not evoked by application of ACh or glutamate were shown).

compared between these muscles. In the radula

264 A. Harapba, M. YOSHIDA et al.

DISCUSSION

From results obtained in the present experi- ments the functional mechanisms of myoactive tetradecapeptides (MATPs: AEP2, AEP3 and FEP4) were studied. In the penis retractor muscle of A. fulica, the three peptides enhanced the tetanic contraction elicited by nerve stimulations and the order of potency was AEP2 >AEP3> FEP4. On the other hand, the enhancing effect of the three peptides on the contraction evoked by muscle stimulations was about the same. These results may well be explained by assuming that actions of the three peptides on the presynaptic sites are different. Molluscan muscles generally receive multiple innervations [2, 11], and it is possible that the three kinds of peptides showed different modulatory actions on the nerve fibers.

In the radula retractor muscle of A. fulica, AEP2 enhanced the tetanic contraction in a dose- dependent manner with a threshold concentration as low as 10° M, suggesting the possibility that this peptide acts physiologically in the regulation of contraction of this muscle. AEP2 also induced rhythmic bursts of activity in the buccal ganglionic neuron B4 which is known as a cholinergic motoneuron of the radula retractor [13]. Fur- thermore, from the pharmacological examination using propantheline, AEP2 was suggested to act mainly on the cholinergic transmission. Therefore, it is considered that AEP2 regulates contraction of the radula retractor muscle by acting excitatorily on the ganglionic neuron(s) as well as on the muscle.

There are some instances that a kind of peptide regulates the muscle contraction by acting both on the central nervous system and the periphery. Achatin-I isolated from A. fulica increased the heart activity by acting excitatorily on the cardio- excitatory neuron PON and on the ventricle of this snail [1]. On the other hand, FMRFamide was shown to act inhibitorily on the central neurons and excitatorily on the myocardium to regulate the heart beat [3].

In the radula muscles of F. ferrugineus, on the other hand, FEP4 isolated from this animal showed the most potent activity of the three pep- tides. FEP4 showed strong enhancing effects on

the twitch contraction of the protractor with a threshold concentration of 10~'M, suggesting that FEP4 physiologically regulates the contraction of the radula protractor muscle. It is noteworthy that in F. ferrugineus and also in R. thomasiana the three peptides showed strong enhancing effects on the contraction of the radula protractor but slight or no effects on the radula retractor. Our previous studies have shown that in the radula protractor of R. thomasiana the principal excitatory neurotrans- mitter is ACh, whereas in the retractor it is gluta- mate [6, 7, 12]. The same appears to be true in the radula muscles of F. ferrugineus (Muneoka, un- published data). The present results suggest that the three peptides enhanced the contraction of the radula protractor by modulating the transmission mediated by ACh. It is probable that in general MATPs act mainly on the cholinergic transmission in molluscan muscles. Several results obtained in the present experiments appear to support this notion.

We found that the amino acid sequence of a peptide allatotropin purified from the tobacco hornworm Manduca sexta [4] is significantly ho- mologous to that of MATPs as shown below.

AEP2 GF RQDAAS RV AHGY amide AEP3 GF RGDAAS RVAHGEF amide FEP4 GFRMNS S NRVAHGEF amide Allatotropin

GF K* NVEMMT AR GF amide

Although allatotropin is not a 14-residue peptide but a 13-residue peptide, it shares 36% of residues with AEP3 and 43% with FEP4. Particularly, the N- and C-terminal dipeptide moieties of allatotro- pin are the same with those of the three molluscan peptides, except the C-terminus of AEP2. Since we have not yet examined the effects of allatotro- pin on molluscan muscles, the structure-action relationships are not known. However, these results are considered to suggest that MATP- related peptides are distributed not only in mol- luscs but also in other phyla including arthropods.

ACKNOWLEDGMENTS

This research was supported in part by the Fujisawa Foundation and by Grants-in-Aid for Scientific Research

Molluscan Myoactive Tetradecapeptides 265

7 Muneoka Y, Kobayashi M (1980) Modulatory ac- tions of octopamine and serotonin on the contrac-

from The Ministry of Education, Science and Culture of Japan.

tion of buccal muscles in Rapana thomasiana. II. Inhibition of contraction in radula retractor. Comp

REFERENCES Biochem Physiol 65C: 81-86 3 2 F 8 Muneoka Y, Kobayashi M (1992) Comparative EROILESOND TSS ee pen ciie gee war aspects of structure and action of molluscan “aie! El, IN@teKte uae Coot 2 mee a E neuropeptides. Experientia 48: 448-456 Muneoka Y, Kobayashi M (1991) Purification of : ; : F 3 9 Muneoka Y, Twarog BM (1977) Lanthanum block achatin-I from the atria of the African giant snail, af connection anil of nékeaiion in mearone to eatna falicapandhitstpessible“iunctions Bioehen serotonin and dopamine in molluscan catch muscle. = a “ eae 3 ee J Pharmacol Exp Therap 202: 601-609 eee if ae Re eran, = 10 OhtaN, Kanda T, Kuroki Y, Kubota I, Muneoka Y, oe te a Pit 1. Bd b ‘ AOD Hee Kobayashi M (1990) Three novel tetradecapeptides Wo dlemic Press. New York Pay ; isolated from the ganglia of the molluscs, Achatina ’ y aR fulica and Fusinus ferrugineus. In “Peptide Chemis- AIDE US, WIBELSavia) i sSclpayerson RM (IED) See try 1989” Ed by N Yanaihara, Protein Research tory actions of 5-hydroxytryptamine and some Foundation, Osaka, pp 57-62 tid the heart of the Afri iant , : en. feb ceca ie Sci Te a 11 Walker RJ (1986) Transmitters and modulators. In : : q “The Mollusca Vol. 9” Neurobiology and Behavior Kataoka H, Toschi A, Li JP, Carney RL, Schooley Part 2. Ed by AOD Will GERNOT N DA, Kramer SJ (1989) Identification of an allatotro- vat ie ef ae euitaas the leew in fi dult Mand ta. Sci 243: 1481- ; ies ail Baten nr pare 12 Yanagawa M, Takabatake I, Muneoka Y, ; iy Kobayashi M (1987) Further pharmacological study Seey eon Ms REIPEETCNO) S, HSU! I Coy) SUS on the cholinergic and glutaminergic properties of ture and action of novel neuropeptides isolated from | he Afri : 1. In “Moll the radula muscles of a mollusc, Rapana thoma- lam eMM ee yee siana. Comp Biochem Physiol 88C: 301-306 1 > > : : Rati ad Ee Gees SIN GtStaaTA 13 Yoshida M, Kobayashi M (1991) Neural control of 60-264 2 es : the buccal muscle movement in the African giant ae hi M. M ka Y (1980) Modulat snail Achatina fulica. J exp Biol 155: 415-433 ape peruano Mea SD)nModulatony aC! 1a! “Yeenida'M. Kobayashi'M (1992) ‘Cerebral’ and

tion of octopamine and serotonin on the contraction of buccal muscles in Rapana thomasiana. |. En- hancement of contraction in radula protractor. Comp Biochem Physiol 65C: 73-79

buccal neurons involved in buccal motor pattern generation in Achatina fulica. Acta Biol Hung 43: 361-367

Po abnor Teagan cal

i i ’ =

0 4 ic an vyotelUbOM (ORCL) WU daueedod WV aliata F ), ‘i shear d Ate ovine oclt 0G ni t38 av SLOTS i ern ” fereeetindid. none itavi\eabaessa) hh Nr A la A n io FE gh

: rua P29 ast yi fl On Lei eey iV) (ava) hiniictngd PF ah a a nS iw ws vahiay er Ti a if iy Pandit Patty

f ‘ ; ; hoe PAN Rae | dei ) ann! , * pul y en vor é fe Figg ar - 7 - ‘ei aire het) eee i Pata 4 pores ne ronerniewa ides atl ie Ts = oe) Jibs hia SCP Abb (Ph aie wri emer eey , ‘ r 4 aid ey “ » (hat iS Soild iets MLD 1H eased GF sateen m N hd Bier ar

Pe)

}

1 Ee qt tr a sos a enh sua

te MENT vt Geert Tp) Parts Hine 4 *

¢ 5 j use ' fe cL p | | it A eenaam ; reese ahs 2

Cary ho Ve i arate i reat asd MTR sacinete b

4 loti sia it he ‘ biusr fe ey f (AAS Wh tie ail on Fy ; meee - V8 A ees a <anuliee HOM, is ; heart 4 ds, eublecgnary Soot Wei ie ae Ds en’ ti of st she Ue : . PALDEN CUES arias t a » oes PRE de gree ose pee: en ny ig LOL bak jul Te hay Sopa GRC bates ane pie ih Se Shaliha yeaa AEN iam 7 ac. Might Fed gina, el thy srt On Hine ALAS" conor? (esa ry

st aan nuts Or paca appro WALES tot WOR, Aeneid

fia eit OS fo wt est Pes i aig =? Ppa fSh | - a) it ‘Ss ‘9 £ «tid 4 * f Ta eri) oe © t € ‘ i fi) oy” ri iva? EAPO ihe Ak W260) bese A one ies ! j Tots { - ' ya aad oo ica ell irpeeves sere oa ) s 4 * Iie 2) * Y = bi use nL vrei Ai ies peal enone ole dimemtanet

a (ie hide) a } te icra hi ithe Ah ee Di ale ait sap ‘4 i sbrratetinA RPS ae heb

a prtilpbeiyt Mat) X ndeor ay ) & agains pate iveiqray Uy ’ : . “4 WS. etvhneie malt Leraiet vet” i : of

; : soleito oe hire! Ligtiea ar huttes® a) wens a

pa FAT= f FT, + +h i are dope 0-68 eat Oa

t a MAY a \) Bl ~- ee Ue Garett “ul ny 3 Fi i . Es > & if . - ee 1 YT i? ’ ‘ h L r\fvortend Je ‘lore avr’ I inane he

LP ary CATE Plbs

gegen dr | > Ad Ga 7

i) ie

ZOOLOGICAL SCIENCE 10: 267-273 (1993)

© 1993 Zoological Society of Japan

Effects of Proteolytic Digestion on the Control Mechanism of Ciliary Orientation in Ciliated Sheets from Paramecium

Munenort Nocucut', KEN-ICHI OKAMOTO

and SHOGO NAKAMURA

Department of Biology, Faculty of Science, Toyama University, Toyama 930, Japan

ABSTRACT—Effects of proteolytic digestion on the ciliary response to Ca** and cyclic nucleotides in ciliated cortical sheets from Paramecium were examined. Ciliary orientation toward 5 o’clock (with the anterior of the cell defined as 12 o’clock) in response not only to lowering Ca** concentration, but also to raising cyclic nucleotides concentrations, was lost by trypsin digestion. Differential sensitivity of cilia, depending on the location of the cell surface to cyclic nucleotides, was also lost by trypsin digestion. The cilia reoriented toward 12 o’clock, which was the ciliary orientation in response to Ca** above | uM before trypsin digestion. Ciliary axonemes lost a large number of outer dynein arms after trypsin digestion. This suggests that the outer dynein arm may contribute to controlling the ciliary beating direction in response to intracellular Ca** and cyclic nucleotides.

INTRODUCTION

The locomotor behavior of Paramecium de- pends on the ciliary beating direction which is controlled by intracellular Ca’* concentration [3, 7-9] and presumably by cyclic nucleotides concen- tration [1, 4, 6, 11]. We observed the top view of ciliary orientation controlled by Ca** and cyclic nucleotides in ciliated cortical sheets from demem- branated cell model of Paramecium, and found that cyclic nucleotides competed with Ca** in determining the ciliary orientation [14]. Also, it was revealed that the ciliary sensitivity to cyclic nucleotides depended on the location of the cell surface [14].

We reported that both the change of ciliary orientation and the stable position in response to lowering the Ca** concentration disappeared by trypsin digestion using Triton-glycerol-extracted Paramecium [13]. This seemed to occur as a result of selective digestion of the Ca?*-dependent con-

Accepted December 10, 1992 Received August 31, 1992

" To whom correspondence should be addressed.

> Present address: Department of Biophysical Engineer- ing, Faculty of Engineering Science, Osaka University, Osaka 560, Japan

trolling mechanism of ciliary beating. It is impor- tant to examine the changes that proteolytic diges- tion induces in ciliary orientation on the cortial sheets of Paramecium, and the structural changes in ciliary axonemes after proteolytic digestion.

We report here observations of ciliary orienta- tions on cortical sheets of Triton-glycerol- extracted Paramecium undergoing proteolytic digestion. Tryptic digestion removed the cyclic nucleotide sensitivity, as well as Ca** sensitivity, in the ciliary reorientation response on the cortical sheets. After trypsin digestion, the cilia pointed toward 12 o’clock irrespective of cyclic nucleotides and Ca** concentrations. Electron micrographs showed that a large number of outer dynein arms disappeared after trypsin digestion. The results suggest that the mechanism which controls ciliary orientation, depending on intracellular Ca’t and cyclic nuceotides, may be closely related to the outer dynein arms.

MATERIALS AND METHODS

Paramecium caudatum (stock G3) was cultured in a hay infusion. Cells were grown to late- logarithmic phase at 25°C.

Preparation of ciliated sheets of permeabilized

268 M. Nocucui, K. Okamoto AND S. NAKAMURA

cells and reactivation of cilia were essentially the same as the preceding paper [14]. Concentrated and washed cells were extracted with an extraction medium containing 0.1% Triton X-100, 20mM KCl, 10 mM EDTA and 10 mM Tris-maleate buf- fer (pH 7.0) for 5 min in an ice bath. The extracted specimens were then washed three times with an ice-cold washing medium which consisted of 50 mM KCl, 2mM EDTA and 10 mM Tris-maleate buffer (pH 7.0). At that time they were pipetted several times through a glass pipette with a small inside tip diameter to tear or to give a nick to the cell cortex. Then the models were suspended and equilibrated in an ice-cold glycerol-KCI solution which consisted of 30% glycerol, 50 mM KCI and 10 mM Tris-maleate buffer (pH 7.0) prior to ex- perimentation.

A simple perfusion chamber was prepared by placing the sample between a slide and a coverslip. The slide and the coverslip were separated by a thin layer of vaseline applied to two opposite edges of the coveslip. To observe the reactivation of cilia on the sheet of cell cortex, 20 pl of the sample was placed on a slide glass, and a coverslip with vaseline was placed gently on the sample. Solu- tions were then perfused through the narrow open- ing at one of the edges of the coverslip while the excess fluid was drained from the opposite end with small pieces of filter paper. During the first perfusion using a reference glycerol-KCl solution, some torn cell cortex adhered flat to the glass surface. The cortical sheets were then perfused successively with reactivation solutions. All the reactivation solutions contained 1 mM ATP, 1 mM MgCl, 30% glycerol, 50mM KCI and 10mM Tris-maleate bufer (pH 7.0) as well as the compo- nent(s) noted in the results. Free Ca** concentra- tion in the reactivation solution was controlled by a Ca-EGTA bufer with 1 mM EGTA [12, 15]. The reactivation was carried out at 22-25°C. The movements of the cilia were observed under a dark field microscope, equipped with a 100 W mercury light source, a heat filter and a green filter, and recorded on video tape using a National WV-1550 TV camera. Pointing directions of cilia which were observed from above were indicated by using a clock face with a definition of the anterior end of the cells as 12 o’clock. The anterioposterior axis of

cortical sheets were determined by anterioposter- ior lines of cilia. Ciliary orientation of cilia were measured as angles to an anerioposterior axis and then converted to an analogue clock face expres- sion using each cilium on the cell surface except for the oral groove. With the definition that the surface area of the anatomical left-hand side is the left-hand field of the cortical sheets, we used the cilia on the left third and the right third for measuring the ciliary orientation of the left-hand field and the right-hand field, respectivley. In the measurement of ciliary orientation after trypsin digestion, we used the cilia on both the left- and right-hand field.

Prepared Triton-glycerol-extracted whole cell models were suspended in a reactivation solution containing 100 ~g/ml trypsin (from bovine pan- creas, Boehringer, 110 units/mg) or elastase (from porcine pancreas, Sigma type IV, 120 units/mg) in a test tube to make the cellular protein concentra- tion approximately 1 mg/ml. A minute amount of the suspension was placed on a slide glass and the ciliary response to digestion was observed under the dark field microscope. Five minutes after mixing, soybean trypsin inhibitor was added to make the final concentration of 1.0 mg/ml. The samples were fixed with 2% glutaraldehyde in 50 mM sodium cacodylate (pH 7.4) for 1 hr then post-fixed for 1 hr with 1% OsQOy, in the same buffer. They were dehydrated in an ascending series of ethanol to 100% and embedded in epoxy resin following the method of Nakamura [10]. Sections were cut on a Porter-Blum MT-1 ultra- microtome and were stained with uranyl acetate and Reynold’s lead citrate. A JEOL 100C electron microscope was used for the observations.

RESULTS

The cilia on the cortical sheets reoriented toward 12 o’clock after trypsin digestion irrespec- tive of Ca** concentration (Fig. 1). When the Ca** concentration was low enough to produce ciliary orientation toward 5 o’clock, which corre- sponded to normal ciliary beat, and consequent forward swimming [14], almost quiescent cilia pointing 5 o’clock began to rotate counterclock- wise within a minute after perfusing a reactivation

Effects of Proteolytic Digestion on Cilia

————or =]

_. 10} Le

6 44 be

oy | ait

oad “ell i pare

& ee / iu

ci} A {

em fi

oO

Sl /

a) a?

3S 5} 7 j | —————— = =H

0) 10” 10° 10

Ca** conc. (M)

Fic. 1. Ciliary orientation in response to Ca** concen- tration after trypsin digestion. Circles and broken lines indicate the ciliary orientations before trypsin digestion. Triangles and solid lines indicate ciliary orientations 3-5 minutes after trypsin digestion. Ciliary orientation measured as angles to an an- terioposterior axis and then converted to an ana- logue clock face expression as indicated in the ordinate. Measurements were performed about cilia on the dorsal area. The data was obtained from four individual preparations. The bars indicate standard deviations.

solution containing 100 “g/ml trypsin. Finally the cilia pointed toward 12 o’clock and stopped within a few minutes (Fig. 1). When the Ca** concentra- tion was above 10 °M, cilia which oriented toward 10-12 o’clock moving counterclockwise in a circular pattern stopped moving and finally pointed toward 12 o’clock within a few minutes after the perfusion of reactivation solution contain- ing 100 “g/ml trypsin (Fig. 1).

As reported in the preceding paper [8], cAMP induces ciliary orientation toward 5 o’clock and cGMP induces ciliary orientation toward 3 o’clock, even in the presence of Ca** above 10~° M which induces ciliary orientation toward 11-12 o’clock if cyclic nucleotides are not present. The cilia on the left-hand field of the cell are more sensitive to cyclic nucleotides than those on the right-hand field with the definition that the surface area of the anatomical left-hand side is the left-hand field of the cortical sheets. As shown in Figures 2 and 3, after perfusing a reactivation solution containing 100 ug/ml trypsin, the cilia oriented uniformly toward 12 o’clock irrespective of cyclic nucleotide concentrations and location of the cilia on the cell surface. Also in the presence of cyclic nucleotides

269

without Ca**, trypsin digestion induces the ciliary orientation toward 12 o’clock (data not shown). Differential response of cilia to cyclic nucleotides in the left- and right-hand field of the cells dis- appeared by trypsin digestion (middle part of Figs. 2 and 3, and Fig. 4). Elastase which was thought to digest nexin [2], and V8 protease which produced a proteolytic effect different from trypsin on the 21S dynein from sea urchin sperm flagella [5], never produced such a change in ciliary orientation that was found to be induced by trypsin digestion in all cases previously described at concentrations of 100 pee/ml.

The cahnge of ciliary orientation after trypsin digestion in the absence of ATP was also tested to examine whether the changes of ciliary orientation by trypsin digestion were induced by active sliding or induced passively by structural changes of some

Sle

10 oo wane 8 41 rh = 2 \ NO O \ © 12 hd ae ety \ ‘, A a! \ ‘ (= \ S © Sat \ © 3 ? — > = Pe el ot & 5

Gl

1 —t {_j O 10 100 1000

CAMP conc. (uM)

Fic. 2. Ciliary orientation in response to cAMP concen- tration after trypsin digestion. Ca** concentration is 5 ~M throughout. Circles and broken lines indi- cate ciliary orientations before trypsin digestion. Open circles and closed circles indicate the ciliary orientations in the left-hand field and in the right- hand field, respectively. Triangles and solid lines indicate the ciliary orientations 3—5 minutes after trypsin digestion. Ciliary orientation of cilia were measured as angles to an anterioposterior axis and then converted to an analogue clock face expression as indicated in the ordinate. In the measurement of ciliary orientation before trypsin digestion, we used the left third and the right third of each ciltum for measuring the ciliary orientation of the left-hand field and the right-hand field, respectively. In the measurement of ciliary orientation after trypsin digestion, we used the cilia on both the left- and right-hand field. The data was obtained from five individual preparations. The bars indicate standard deviations.

270 M. Nocucui, K. OkAMoTO AND S. NAKAMURA

tt =| Silt cae ell oO 42 PrIko Oa Ses S (Se) Sie ee Sa iS “iain ~ A 5 Uy; mSr 7S % 2+ "aR ‘se d 3 ae vy g Gee ° o) mall = 5, oO 6 Rap—il 4 4 O 10 100 1000 CGMP conc. (uM) Fic. 3. Ciliary orientation in response to cGMP concen-

tration after trypsin digestion. Ca** concentration is 1.8 4M throughout. Circles and broken lines indicate ciliary orientations before trypsin digestion. Open circles and closed circles indicate the ciliary orientations in the left-hand field and in the right- hand field, respectively. Triangles and solid lines indicate the ciliary orientations 3-5 minutes after trypsin digestion. Measurements of ciliary orienta-

axonemal components. As shown in Figure 5, the change of ciliary orientation from 5 to 12 o’clock induced by trypsin digestion required ATP. When the ciliated sheets were perfused with a reactiva- tion solution containing EGTA to keep Ca** concentration under 10~’M, the cilia pointed toward 5 o’clock (Fig. 5a). Following successive perfusions with a solution without ATP and a solution containing trypsin without ATP did not affect the ciliary orientation toward 5 o’clock. Eighty seconds after the trypsin perfusion, a solu- tion containing a trypsin inhibitor without ATP was perfused. Then the reactivation solution con-

tion were performed as explained in the legend of Figure 2. The data was obtained from four indi- vidual preparations. deviations.

The bars indicate standard

Fic. 4.

Ciliary orientations before and after trypsin digestion. The photograph shows a dorsal view of the ciliated

sheet. The cortex of a cell model tore along an oral groove. The top of each photograph is the anterior direction of the cell. The sheet was perfused successively with reactivation solutions containing 10 ~M cAMP and 5 uM Ca?* (a), and 100 g/ml trypsin as well as 10 ~M cAMP and 5 uM Ca** (b). Arrows indicate the pointing directions of cilia. Pointing direction of cilia on the left-hand field differed from that on the right-hand field in (a). In (a), as indicated by arrows, ciliary orientation on the left hand field was 5 o’lock, and on the right-hand field was 11 o’clock. On the other hand, ciliary orientation was 12 o’clock uniformly in the whole area of the cortical

sheet after trypsin digestion (b). Bar, 10 um.

Effects of Proteolytic Digestion on Cilia 271

Ciliary orientation (o'clock)

RICK x

Fic. 5.

The effect of trypsin digestion on the ciliary orientation in the absence of ATP. Cortical sheets were perfused

successively with solutions fundamentally contained 1 mM MgCl, 30% glycerol, 50 mM KCl, Tris-maleate buffer (pH 7.0) as well as 1 mM EGTA (a) or 1 ~M Ca** +100 ~M cAMP (b). +ATP indicated under abscissa means the solution contained 1 mM ATP. Trypsin concentration was 50 ~g/ml. Cortical sheets were first perfused with

solutions containing ATP.

In both cases, ciliary orientations on the cortical sheets were not changed by

subsequent trypsin digestion in the absence of ATP. After stopping the digestion by the addition of trypsin inhibitor, the solution containing ATP was perfused. The perfusion elicited the change of ciliary orientation from 5 toward 12 o’clock. Trypsin digestion time was 80 sec in (a) and 20sec in (b). The data presented is a typical example of one of the many series of successive perfusions examined. The bars indicate standard deviations.

taining ATP was perfused. Immediately after the perfusion, the cilia changed their orientation from 5 o’clock to 12 o’clock and never returned toward 5 o'clock. The ciliary orientation toward 5 o’clock induced by cAMP also did not change the orienta- tion by trypsin digestion in the absence of ATP (Fig. 5b). Thus, the change of ciliary orientation from 5 to 12 o’clock after trypsin digestion re-

Fic. 6. Cross-sections of the Triton-glycerol-extracted cil

quired ATP, and the trypsin digestion removed the ability to change and to keep the ciliary orientation toward 5 o’clock either in response to cyclic nu- cleotides or to lowering Ca?* concentration.

We examined the effects of trypsin digestion on the axonemal structures by electron microscopy. The structure which underwent serious break down by trypsin digestion was outer dynein arms

iary axonemes before proteolytic digestion (a) and after

digestion by 100 g/ml trypsin (b) and by 100 “g/ml elastase (c). bar, 100 nm.

272 M. Nocucui, K. OkAMoTO AND S. NAKAMURA

(Fig. 6b). On the contrary, axonemes treated by elastase in the same condition as that of trypsin digestion kept their outer dynein arms (Fig. 6c).

DISCUSSION

Ciliary orientation on the cortical sheets corres- ponds to the ditection of the effective stroke of cilia [14]. The ciliary orientation toward 5 o’clock and 12 o’clock correspond to normal beating of cilia and ciliary reversal, respectively. The direc- tion of the effective power stroke is essentially controlled by intracelular Ca** concentration [3, 7-9]. However, as we reported previously, “ciliary reversal” was induced by trypsin digestion [13] in Triton-glycerol-extracted whole cell model of Pa- ramecium even if Ca** concentration was low enough to produce normal beating. In this paper, we confirmed that the ciliary orientation induced by trypsin digestion was in actual fact 12 o’clock using cortical sheets of Triton-glycerol-extracted Paramecium (Fig. 1). This indicates that the Ca’*- dependent controlling mechanism of ciliary beat- ing direction is digested at least in part by trypsin. The part affected by trypsin may be the component which is responsible for switching off the ciliary reversal mechanism when Ca** concentration is lowered.

The beating direction of cilia is also modulated by cyclic nucleotides [1, 6, 11, 14]. In the cortical sheets, ciliary orientation is controlled by cyclic nucleotides in the competing mode with Ca**. The cilia on the left-hand field of the cell surface were more sensitive to cyclic nucleotides than those on the right-hand field [14]. Tryptic diges- tion removed all these ciliary responses to cyclic nucleotides. As indicated in Figures 2-4, ciliary orientations towad 3-5 o’clock in response to cyclic nucleotides together with the differential sensitivity disappeared by trypsin digestion. This indicates that trypsin digestion destroys the com- ponent responsible for cyclic nucleotides depen- dent responses, as well as the Ca**-dependent component in the controlling mechanism of ciliary beating direction.

Ciliary orientation toward 5 o’clock, induced in the conditions in which Ca** concentration was low enough, or cAMP concentration was high

enough to suppress the Ca** effect, remained the same after perfusing the solution without ATP. Subsequent perfusions with a solution containing trypsin without ATP, did not induce any change in ciliary orientation. However, the change of ciliary orientation from 5 to 12 o’clock was elicited by the addition of ATP (Fig. 5). This indicates that the change of ciliary orientation induced by trypsin digestion requires ATP and suggests that the change of ciliary orientation requires dynein- microtubul interaction at least to some extent. Thus, the trypsin digestion induced the inabilities to keep the ciliary orientation toward 5 o’clock and to change the ciliary orientation from 12 to 5 o’clock in response to either lowering Ca?* con- centration or to cyclic nucleotides.

It has been clarified that ciliary beating is con- trolled by Ca** [3, 7-9, 14] and cyclic nucleotides [1, 6, 11, 14]. However, the molecular mecha- nisms which transmit the signal of the second messengers to dynein-microtubule interaction as a motor system, as well as its localization and structural bases are scarcely known. After trypsin digestion, outer dynein arms disappeared conspi- cuously in ciliary axonemes (Fig. 6). On the con- trary, outer dynein arms remained intact in the axonemes after the digestion by elastase which affected neither Ca** nor the cyclic nucleotides dependent mechanism as stated previously, where- as axonemes which detached from cell cortex and attached to the glass surface of the perfusion chamber exhibited sliding disintegration within a minute not only by trypsin but also by elastase. Trypsin has been used to produce limited digestion of isolated dynein [5], but how the trypsin affect outer dynein arms within axonemes is not yet known. Although we could not known the exact protein ratio of trypsin to ciliated sheets in perfu- sion chambers, high concentration of trypsin (above 20 «g/ml) was required to induce the cili- ary response by trypsin digestion. This might imply that the quick attack by trypsin on the outside of axonemes is essential to remove selec- tively the Ca?*-dependent and cyclic nucleotides- dependent controlling mechanism of ciliary orientation. The disappearance of the outer arms after tyrpsin digestion might suggests that outer dynein arms play a key role in Ca**-dependent

Effects of Proteolytic Digestion on Cilia

and also in the cyclic nucleotides dependent con- trolling mechanism of ciliary beating direction in Paramecium.

ACKNOWLEDGMENTS

This work was supported in part by a Grant-in-Aid from the Ministry of Education, Science and Culture of Japan. We thank Mr. Takehiko Taneyama and Takumi Shimura for technical assistance. We also thank Miss Sandy Peever and Mercedes Zanon for reading the manuscript.

REFERENCES

1 Bonini NM, Nelson DL (1988) Differential regula- tion of Paramecium ciliary motility by cAMP and cGMP. J Cell Biol 106: 1615-1623

2 Brokaw CJ (1980) Elastase digestion of demem- branated sperm flagella. Science 207: 1365-1367

3 Eckert R, Naitoh Y, Machemer H (1976) Calcium in the bioelectric and motor functions of Para- mecium. Symp Soc Exp Biol 30: 233-255

4 Hamasaki T, Barkalow K, Richmond J, Satir P (1991) cAMP-stimulated phosphorylation of an ax- onemal polypeptide that copurifies with the 22S dynein arm regulates microtubule translocation velocity and swimming speed in Paramecium. Proc Natl Acad Sci USA 88: 7918-7922

5 Inaba K, Mohri H (1989) Dynamic conformational changes of 21S dynein ATPase coupled with ATP hydrolysis revealed by proteolytic digestion. J Biol Chem 264: 8384-8388

6 Majima T, Hamasaki T, Arai T (1986) Increase in cellular cyclic GMP level by potassium stimulation

15

273

and its relation to ciliary orientation in Paramecium. Experientia 42: 62-64

Naitoh Y, Kaneko H (1972) Reactivated Triton- extracted models of Paramecium: modification of ciliary movement by calcium ions. Science 176: 523- 524

Naitoh Y, Kaneko H (1973) Control of ciliary activities by adenosinetriphosphate and divalent ca- tions in Triton-extracted models of Paramecium caudatum. J Exp Biol 58: 657-676

Naitoh Y, Sugino K (1984) Ciliary movement and its control in Paramecium. J Protozool 31: 31-40 Nakamura S (1981) Two different backward- swimming mutants of Chlamydomonas reinhardtii. Cell Struct Funct 6: 385-393

Nakaoka Y, Ooi H (1985) Regulation of ciliary reversal in Triton-extracted Paramecium by calcium and cyclic adenosine monophosphate. J Cell Sci 77: 185-195

Noguchi M, Inoué H, Kubo K (1986) Control of the orientation of cilia by ATP and divalent cations in Triton-glycerol-extracted Paramecium caudatum. J Exp Biol 120: 105-117

Noguchi M (1987) Ciliary reorientation induced by trypsin digestion in Triton-glycerol-extracted Para- mecium caudatum. Cell Struct Funct 12: 503-506 Noguchi M, Nakamura Y, Okamoto K (1991) Con- trol of ciliary orientation in ciliated sheets from Paramecium— Differential distribution of sensitivity to cyclic nucleotides. Cell Motil Cytoskeleton 20: 38-46

Portzehl H, Caldwell PC, Riiegg JC (1964) The dependence of contraction and relaxation of muscle fibers from the crab Maia squinado on the internal concentration of free calcium ions. Biochim Biophys Acta 79: 581-591

Ne f tive it ty | tt i iy i t ji ‘4 t if i rie 1 wt) v i i i 7 4

4 itn oa trad Bs

hazed Woke te nl | ‘hoy r souagh a a va a ita bite ile $s en arid ees ve JF iin

“inet Yo ait batt Zoe pore A er ste

«

"

di tis ;

eae pane] ih omceGal Pika

; ocr lip ecienta Pyle Wa ie

oe r 1 : “s sit 5 ia pe! Nh ude ¥ donot # Ary tien tone a a whe 9 pera aay ey! er haf: ARV) 4 ie (a bn Gn

OAT Dan oe Mane eee a wT FT | reap vienna wee iT toric Gare 4 Aa rt Ysa NP anh eal ; > bila A Thdens We en Po mlicians i) i 807) Fie Re node tor aot a Wei w tn. niet qpimameetty . oT hadkhe Repeater alle esi trative an aaa PUPA A 6G nm © Oba" oF + ia sla ie via Pas) prea pefy iesi i . Rel, : 7 Demy, TA Ay 1 reliog A Seines . 2 ’ Te, rat} Mion yale aay tet pnd) . sls) nae ES jawed WOOL 14 r - : ve hot agaiyet eral | eae Neri 2 i 4 i fs . $B, O40) Wiis sith N D Wee Lo fcayane 1a ie 3 ‘ ; ty lo D tye! rooegrorie, Lali eh IN? : He Oa Ue Paige ae 4 ie ; wher CdR By ae ne ' wiey) Naar aa corte Hie a te Sate i ieatexcs' i stamtnhy ’ Sob Poot vebio> a with ont a ee ret ¢ ation ane 7 ; h & Bu are dh Sant phi Pe ET cae a a wend <-4arrRaieak HOP eet - ae iy iellioe atria 2 - ' i ee re as ij 1 t ’ fi Ly f

Ta ae

ZOOLOGICAL SCIENCE 10: 275-279 (1993)

The Nucleotide Content of the Octopus Photoreceptor Cells: No Changes in the Octopus Retina Immediately Following an Intense Light Flash

Masatsucu SEipou!, KoHzoH OutTsu~, ZINPEI YAMASITA’,

Kinya Narita‘ and Yust Kitro*

‘Aichi Prefectural University of Fine Arts, Nagakute, Aichi 480-11, 7 Ushimado Marine Laboratory, Okayama University, Ushimado Oku, Okayama 701-43, 3Tsotope Research Center, Osaka University, Suita, Osaka 565, “Department of Biology, Faculty of Science, Osaka University, Toyonaka, Osaka 560, Japan

ABSTRACT—The amounts of cyclic-guanosine monophosphate (cGMP) and cyclic-adenosine monophosphate (cAMP) in octopus retina were measured by competitive radioimmunoassay. The amounts of the respective di- and tri-phosphonucleotides, and of visual pigment (as retinal isomers), were determined by high-performance-liquid chromatography (HPLC). Unirradiated octopus retina contained 2.4+0.5 pmol of cGMP and 36+7.0 pmol of cAMP. When 27.6% of the visual pigment in the retina was converted into the meta-form by an intense light flash, the rapidly-frozen retina contained 2.3+0.9 pmol of cGMP and 32+4.0 pmol of cAMP: no significant changes from the unirradiated state. The molecular ratios of CGMP and cAMP to total visual pigment were estimated to be 1/1800 and 1/ 120, respectively. Assuming that all the cGMP and cAMP were contained only in the rhabdome, it was estimated that there are 1.6 molecules of cGMP and 24 molecules of cAMP per microvillus. The low content of cGMP in the microvillus and the absence of significant change in cGMP content on exposure to light lead us to conclude that cGMP is not the ‘intracellular messenger’ of phototransduction in

© 1993 Zoological Society of Japan

octopus photoreceptor cells.

INTRODUCTION

It is known than, in the rhabdomeric photore- ceptor cells of cephalopods and arthropods, the cation channels in the plasma membrane are opened by light stimulation, resulting in depolar- ization of the photoreceptor cell [17]. However, details of the events between photoisomerization of rhodopsin and channel opening are unknown. Cyclic-GMP (cGMP) [12, 15] and inositol tris- phosphate (IP3) [3, 4, 8, 19] are possible candi- dates for the intracellular messenger, and it is well established that cGMP fulfils this role in vertebrate rod cells [9, 13].

In the ventral photoreceptor cell of Limulus, injection of solutions containing cGMP or cAMP

Accepted January 25, 1993 Received September 9, 1992

does not increase the influx of Na* ions [16], but cGMP can depolarize the cell in a similar manner to light stimulation [12]. In the squid retina, it has been reported that (1) light increases the activity of guanylate cyclase [15], and (41) according to John- son et al. [12], the amount of cGMP in the light- irradiated retina of Loligo pealei is up to 1.7 times that in the dark-adapted state. This reported cGMP change in the squid retina would probably be enough to open a cGMP-gated cation channel (cf. the light-stimulated decrease in cGMP to 80 % of that of dark-adapted rods in the frog, Rana catesbiana, retina [6]). However, Brown et al. [2] reported that light induced no significant changes in the cGMP content of the photoreceptor cells of the squids Loligo pealei and L. opalescens.

We consider that Johnson’s data did not provide sufficient evidence to support the candidacy of cGMP as an intracellular messenger of phototrans-

276 M. Semou AND K. Ontsu et al.

duction in cephalopods, and also that Brown’s data did not reject a role for cGMP, because both studies were unable to express the content of cGMP as a proportion of the initial content of a single microvillus. In the present study, we com- pare the amounts of cGMP, cAMP, and their di- and tri-phosphorylated derivatives in octopus re- tina, before and after light irradiation. Visual pigment content (proportional to the number of microvilli) was determined by HPLC analysis of the chromophore retinal, which was used as an internal standard to normalize the size of the retina and the nucleotide content for each experiment. The cephalopod retina is composed almost entirely of photoreceptor cells (and their axons), which are assumed to be the site of most of the cGMP and cAMP measured in whole retina.

MATERIALS AND METHODS

Small octopuses (30-100 g adult Octopus fang- siao d’Orbigny;=O. ocellatus Gray [10]), were collected near the Ushimado Marine Laboratory. The animals were kept in the dark for more than 24 hours prior to dissection. All procedures were carried out under infrared light using the noctovi- sion equipment (NEC, Tokyo, Japan). ‘Eye-cups’ were prepared from enucleated eyes (attached fragments of optic lobe and white body were carefully removed) and the retina was incised radially, flattened sclera-down on filter paper, and mounted on a polystyrene block. The retina was irradiated with white light from an electronic flash unit (5 cm from the retina; 2.5 ms duration, Sun- pak Co. Ltd., Tokyo, Japan) and rapidly (within 300 ms) frozen by slamming it against a copper block previously cooled in liquid nitrogen. The frozen retinae with scleral materials were kept below—80°C. Subsequent procedures were car- ried out under dim red light. Each frozen retina was homogenized in 1.5 ml of ice-cold 6% trichlor- oacetic acid (TCA) on ice, and centrifuged at 12000 rpm for 30 min. Homogenization and ex- traction were repeated twice (When a known amount of cGMP was added to an octopus retina, 89.7% of the added amount was recovered by the first extraction). The precipitate was retained for chromophore extraction (see below). The pooled

supernatants were washed 3 times with 10 ml diethylether to remove TCA, freeze-dried and dissolved in 2 ml distilled water. A 100 ul of this solution was succinylated and cGMP and cAMP contents were measured by radioimmunoassay (according to the instructions provided with com- mercial kits; Yamasa Shoyu Co. Ltd., Chiba, Japan).

The amounts of GTP, ATP and ADP were determined by HPLC analysis of the remaining solution, using a SUMIPAX ODS column (6 x 150 mm, Sumitomo Chem. Co. Ltd., Tokyo, Japan) with a solvent system of 22.5% (v/v) methanol, 4 mM tetra-butyl-ammonium phosphate and 15.5 mM potassium phosphate (pH 5.6) at 30°C. The flow rate was 0.5 ml/min for initial 30 min after sample injection and then 1.0 ml/min. Absorb- ance at 260 nm was monitored and the amount of each nucleotide estimated using respective molecular extinction coefficients at this wavelength [1].

Visual pigment chromophore was extracted into hexane [11, 18] from the TCA precipitate, applied to a YMC silica gel column (6 x 150 mm, YMC Co. Ltd., Japan) and eluted with a solvent system of 10% (v/v) ethylacetate and 0.4% (v/v) ethanol in hexane, at a flow rate of 1 ml/min, monitoring absorbance at 360 nm.

The late receptor potential (LRP) of the octopus retina was measured as described in a previous paper [14], with the retina set in an experimental chamber perfused with aerated sea water.

RESULTS AND DISCUSSION Table 1 shows the amounts of nucleotides and

retinal in the octopus retina. The octopus retina

TaBLeE 1. Contents of chromophore retinal and nucleotides in one retina (X10~° mol)

dark (n=4) light (n=S) retinal 4.4+0.7 4.2+0.4 cGMP 0.0024 + 0.0005 0.0023 + 0.0009 cAMP 0.036 +0.007 0.032 + 0.004 GTP ile 3} 11+4 ATP 17+8 17+6 ADP 4.0+1.3 Sp ejae! lsi/

Nucleotide Content in Octopus Photoreceptor Cells UE

used in this experiment contained about 4 nmol of retinal. The 27.6 % of 11-cis retinal was photo- isomerized to all-trans form by the flash. Signi- ficant differences in the amounts of cGMP, cAMP, GTP, ATP and ADP were not found between the irradiated retina and the unirradiated one. It seemed unlikely that the amounts of the nuc- leotides in the irradiated retina returned to the level in the unirradiated retina before freezing the retina, because the magnitude of the (negative- going) LRP sustained 81.4% of its maximum even at 500 ms after the flash, as shown in Figure 1.

S & a ‘Oo S & £ (0) 200 400 600 Time (ms)

Fic. 1. The late receptor potential of the octopus re- tina. Flash intensity was attenuated to half that for the determination of cGMP content. The vitreous- side positive signal is upward. The downward spike-shaped signal at zero time is an electrical stimulus artifact.

Since the ratio of GTP to cGMP was 4600:1 and that of ATP to cAMP 470:1, there were ample amounts of both GTP and ATP in the retina as source materials for both cGMP and cAMP to be produced. The amounts of AMP, GMP and GDP could not be determined accurately due to con- tamination in the TCA extract, which had some absorbance at the monitoring wavelength 260 nm, as shown in Figure 2.

It has been considered that cGMP may be the intracellular messenger of phototransduction in invertebrate photoreceptor cells as well as in verte- brate rod cells. This was supported by the findings of GTP-binding protein (G-protein), phosphodies- terase and guanylate cyclase in cephalopod photo- receptor cells [5, 7, 20], and by the fact that cGMP injection into the Limulus ventral photoreceptor cell could induce a response in a similar manner to

A260=0.003

ot \ cre \ \ cGMP \ GTP ATP GDP ADP cAMP l peal | 0 0 45 Time (min ) Fic. 2. HPLC analysis of nucleotides in the octopus

retina. Trace 1 is a chromatogram of the TCA extract from an octopus retina. Trace 2 is a mixture of the authentic compounds. Flow rate was 0.5 ml/ min for first 30 min, then 1 ml/min. A569, absorbance at 260 nm. Other abbreviations as given in Table 1.

the light response [12]. However, our results demonstrated that the amount of cGMP in the octopus retina was not significantly changed by the flash, in agreement with Brown et al. [2]. Inability to detect a change in cGMP content is not in itself proof that cGMP is not the intracellular messenger of phototransduction: it is possible that light- induced variation of cGMP occurred but was masked by processes independent of phototrans- duction. However, this possibility is discounted as follows.

Assuming that all cGMP molecules were local- ized only in microvilli, the number of cGMP molecules in one microvillus was estimated as follows. The dimensions of Octopus fangsiao photoreceptor cells and microvilli were estimated from electron micrographs taken by Prof. M. Yamamoto (Ushimado Marine Laboratory). On average, each photoreceptor cell has a cross- sectional area of 3.910 ~m*. From inspection of

278 M. SEIDOU AND K. Ontsu et al.

longitudinal sections of photoreceptors, each cell with rhabdomeres of length 150 um has 3.5 x 10° microvilli. The mean area of the retina used in these experiments was 1.0 cm”: an estimated total 9.110"! microvilli. Assuming that all the cGMP exists in the microvilli, then there are about 1.6 molecules of cGMP per microvillus. Of course, if it is assumed that most of the cGMP in the microvillus is concerned with other functions with- in the cell, less than one molecule of cGMP per microvillus would be available to participate in phototransduction. There is therefore no reason to propose that changes in cGMP have gone undetected because of the presence in the retina of large amounts of cGMP independent of photo- transduction.

It is considered that the intensity of the flash used in this experiment was sufficient to activate substantial numbers of visual pigment molecules in all the microvilli, by the following reasoning. (1) About 3000 molecules of visual pigment were contained in each of 9.1X<10'! microvilli. (2) Assuming that visual pigment was uniformly distri- buted, only in the rhabdomes, and with mean cross-section of 2.2 x 10 um? and length of 150 um, the concentration of visual pigment in the rhab- dome is estimated to be 0.5 mM. (3) According to Lambert-Beer’s law, the intensity of light passing through the 150 ~m rhabdome is reduced to 52% of the incident light. Therefore, with the flash irradiation used in the present study, which con- verts 28% of the total visual pigment to the meta-form, 36% would be converted in the most distal region of the rhabdome and 18% proximally: i.e. phototransduction occurred in all microvilli.

It is estimated that if phototransduction led to an increase or decrease of only one molecule of cGMP per microvillus, the sum of changes in cGMP over the whole retina would be +60%. This would be sufficient for detection by the methods used in the present study, but no such changes were detected. This suggests that there is no change in cGMP content during phototransduc- tion in the octopus retina.

Brown et al. [2] discussed the magnitude of the light-induced change in cGMP concentration, esti- mating that if the dissociation constant of the cGMP-gated channels for cGMP is similar to that

reported for vertebrate rods, then the minimum cGMP concentration jump sufficient to saturate all the channels in the microvillar membrane would be 25-225 uM. Assuming that all the measured cGMP was distributed only in the outer segments of the octopus photoreceptor cells, the cGMP concentration estimated from the present study was 0.17 M in the unirradiated retina. This is 150 to 1300 times less than the magnitude of the concentration jump required, and if this jump happened it would be easy to detect by the methods used here. Since no changes were de- tected, as shown in Table 1, it is concluded that cGMP is not the intracellular messenger of photo- transduction in the octopus retina.

The content of cAMP was 15 times more than cGMP in the retina but no significant change was detected in cAMP content (Table 1). Using argu- ments similar to those used above with regard to cGMP, it is concluded also that cAMP is not an intracellular messenger during phototransduction in this system.

ACKNOWLEDGMENTS

We wish to express our thanks to Prof. M. Yamamoto of Ushimado Marine Laboratory for the use of his unpublished electron micrographs of the photoreceptor of Octopus fangsiao, which were used for calculation of the mean size of the photoreceptor cell and microvillus. We thank Dr. Ian Gleadall for critical reading of this manuscript.

REFERENCES

1 Beaven GH, Holiday ER, Johnson EA (1955) Optical properties of nucleic acids and their compo- nents. In “The nucleic acids” Ed By E Chargaff and JN Davidson, Academic press, New York, Vol 1, pp 493-553

2 Brown JE, Faddis M, Combs A (1992) Light does not induce an increase in cyclic-GMP content of squid or Limulus photoreceptors. Exp Eye Res 54: 403-410

3 Brown JE, Rubin LJ, Ghalayini AJ, Tarver AP, Irvine RF, Berridge MJ, Anderson RE (1984) myo- inositol polyphosphate may be a messenger for visual excitation in Limulus photoreceptors. Nature, 311: 160-163

4 Brown JE, Watkins DC, Malbon CC (1987) Light- induced changes in the content of inositol phos- phates in squid (Loligo pealei) retina. Biochem J

10

11

12

Nucleotide Content in Octopus Photoreceptor Cells

247: 293-297

Calhoon R, Tsuda M, Ebrey TG (1980) A light- activated GTPase from octopus photoreceptors. Biochem Biophys Res Commun 94: 1452-1457 Cote RH, Biernbaum MS, Nicol GD, Bownds MD (1984) Light-induced decreases in cGMP concentra- tion precede changes in membrane permeability in frog rod photoreceptors. J Biol Chem 259: 9635- 9641

Ebrey T, Tsuda M, Sassenrath G, West JL, Waddell WH (1980) Light activation of bovine rod phospho- diesterase by nonphysiological visual pigments. FEBS lett 116: 217-219

Fein A, Payne R, Corson DW, Berridge MJ, Irvine RF (1984) Photoreceptor excitation and adaptation by inositol 1,4,5-trisphosphate. Nature 311: 157-160 Fesenko EE, Kolesnikov SS, Lyubarsky AL (1985) Induction by cyclic GMP of cationic conductance in plasma membrane of retinal rod outer segment. Nature 313: 310-313

Gleadall IG, Naggs FC (1991) The Asian ocellate octopuses. II. The validity of Octopus fangsiao d’Orbigny. Ann Appl Info Sci, Sendai, Japan 16: 173-180

Groenendijk GWT, De Grip WJ, Daemen FJM (1980) Quantitative determination of retinals with complete retention of their geometric configuration. Biochim Biophys Acta 617: 430-438

Johnson EC, Robinson PR Lisman JE (1986) Cyclic GMP is involved in the excitation of invertebrate photoreceptors. Nature 324: 468-470

13

14

16

17

279

Miki N, Keirns JJ, Marcus FR, Freeman J, Bitensky MW (1973) Regulation of cyclic nucleotide concen- trations in photoreceptors:an ATP-dependent sti- mulation of cyclic nucleotide phosphodiesterase by light. Proc Nat Acad Sci USA 70: 3820-3824 Ohtsu K, Kito Y (1985) A photoproduct with 13-cis retinal generated by irradiation with violet light in the octopus retina. Vision Res 25: 775-779

Saibil HR (1984) A light-stimulated increase of cyclic GMP in squid photoreceptors. FEBS Lett 168: 213-216

Stern JH, Lisman JE (1982) Internal dialysis of Limulus ventral photoreceptors. Proc Natl Acad Sci USA 79: 7580-7584

Stieve H (1985) Bumps, the elementary excitatory responses of invertebrates. In “The molecular mechanism of photoreception” Ed By H Stieve, Springer-Verlag, Berlin, pp 199-230

Suzuki T, Makino-Tasaka M (1983) Analysis of retinal and 3-dehydroretinal in the retina by high- pressure liquid chromatography. Anal Biochem 129: 111-119

Szuts EZ, Wood SF, Reid MS, Fein A (1986) Light stimulates the rapid formation of inositol trisphos- phate in squid retinas. Biochem J 240: 929-932 Vandenberg CA, Montal M (1984) Light-regulated biochemical events in invertebrate photoreceptors. 1. Light-activated guanosinetriphosphatase, guanine nucleotide binding, and cholera toxin catalyzed labeling of squid photoreceptor membranes.

Biochem 23: 2339-2347

— Petes |) thee OR RUN Sa ‘onsite * 3 Wi (oe cS Na apne ah ee any OR, hha; peyelay pert Ay oatng OL oa | aonniiea, * Lat eRe yal 1% eviry nt ibe 4 fain ad f. RABIN) 1 a $5ih Avie Ae ais A Re titi + beige Mia ye), ei, ee it cn i leitag ek WH ‘pas sty

» ARO Yrs

y 1% a4 2 DP awOneee Oot ep Dad se n hw ser-sennp Se cia Thberiti Mme Hed (eet Sie Hsien wy

f re PRONE La atinitictac thine is he

ih ) ai : : > evi SEE ag,

in. dae rH. See L Martel Ait res, al ; ; ; . Hee" 1 A GAG a! ‘y isha? ‘talghi i 7 a “on Re ee , rth A Tas Se Site vanities Sri +} fe wt argh “y qir +f e yaieeg Shean iy y $ibh Se cT lie ipso de Stee as + -VOi oy bot calle Ve AS ivf aed fipzse, t pol pL, nae : 4m. E i tonsibyden— ice Mn ‘ . és i ‘ tt a esa. Vara ahem eT te “ wire” f rv f bie enh ov bop ogra Gh tet Mvoa), Wile oe beieh eal lveaeiaener aS ih Yotam, cag Sage spats ‘e et y \ 4k 1 t as “3 sai 1s a) y PURI? | 7 Pa , >| a INE A tare =) age : ieeenet G) t fo DUR GHORE SOL Avabedliy 1 | Ps fi ; \ ’ Z t 1 1 :

Mis iS seanee

ai Me ponbeers tention SOME

Fie

ee ib

feu. revane Lenk anneal Tey Oar patie oH - heey tae ager ache Ww als spb sf ay in paras ee hap rs LPP GTiAer ane . cers PPL ET

woh ecw; beulioMen

CAMP eden

ralleeyanad vil, hs

ask ani hay 3 th yr a ah si Y tees reat) ' ; 1": i r Wie mv ier f (el Woes AW at ma a : ‘dor Mester le eonwinnsialS Ogres Te ThE ACL GOR Ue

Le + LOH): at sand 9 Mate

ey tel deh

oy ert abun AE = ut be c — y " ay Zhi x“ we! I Wists) 1 if ae, Trp sep by PY, oy ‘ =i dan tot ea A +. ere, A Ke) CHENG 4 a 4 cs | Fount hy ‘ j re" nit] Anne ocihe hi OAL. Pd rare 7 Ah P ‘ N , » ee ‘ at le hp hy ae t s es 4 a % ; yt 7 it. Abia eS jes TAy rere, iy Neo” eaters “y r \ if

ZOOLOGICAL SCIENCE 10: 281-285 (1993)

© 1993 Zoological Society of Japan

Effect of Osmolality on the Motility of Sperm from the Lamprey, Lampetra japonica

WATARU KOBAYASHI

Zoological Institute, Faculty of Science, Hokkaido University, Sapporo 060, Japan

ABSTRACT—Lamprey (Lampetra japonica) sperm were immotile not only in the seminal plasma but when immersed in an isotonic lamprey Ringer solution. Sperm were also found to be immotile when the semen was diluted with NaCl, KCl, or glucose solution isotonic to the Ringer solution. Sperm became motile, however, after dilution of semen with an hypotonic solution. These results suggest that motility of the lamprey sperm is probably suppressed by the osmolality of the seminal plasma and is initiated by a decrease in osmolality upon dilution into the spawning media (river water).

INTRODUCTION

Sperm of many animal species are quiescent in male reproductive organ and become motile after discharge into spawning media. It has been known that sperm of many species that show external fertilization recognize the difference between their seminal plasma and environmental media as a signal for the initiation of sperm motility [12]. Morisawa and his collaborators reported that hypotonicity induces the initiation of sperm motil- ity in some cyprinid fishes (freshwater spawners) while hypertonicity induces sperm motility in marine teleosts [6, 7]. A change in osmotic press- ure is not the sole factor for the initiation of sperm motility; the sperm of some freshwater fish show motility in isotonic media [12]. For example, salmonid sperm move after a decrease in the concentration of surrounding potassium ions even when the seminal plasma is diluted with salt solu- tions isotonic to it [7, 9]. Knowledge of the external factors that induce sperm motility may offer a basis for understanding the adaptation of gametes to the spawning environments.

I am interested in the fertilization of lampreys, because these animals represent one of the most ancient group of vertebrates. Information obta- ined from lamprey reproduction may be useful in

Accepted February 2, 1993 Received November 12, 1992

helping us to understand the evolution of repro- ductive strategy in vertebrates. However, few studies have focused on the reproductive physiolo- gy of lamprey gametes. At spawning, female and male lampreys release their gametes into hypo- tonic river water (external fertilization). Yamamo- to [15] has stated that “artificial activation of the lamprey eggs is possible in an isotonic salt solution (137 mM NaCl, 2.7 mM KCl, 2 mM CaCh, pH 7.3 with NaHCO; ) but fertilization is impossible in the solution”. He succeeded in induction of artificial fertilization by immersing the eggs, which had been inseminated with “dry” sperm, into fresh water. These facts suggest that lamprey sperm are immotile in the Ringer’s solution and initiate movement after dilution of the seminal plasma with freshwater. The present study was carried out to decide whether decrease in osmolality induces motility in lamprey sperm.

MATERIALS AND METHODS

Adult males of the river lamprey, Lampetra japo- nica, were taken in May (their spawning season) in the Ishikari river at Ebetu, Hokkaido. They were kept without feeding in constant darkness in plastic aquaria containing well water (12°C). In these conditions, they matured within 10 days. Semen was obtained by manual stripping of the animals anesthetized in 0.05% tricaine methanesulfonate (MS-222). The semen from different individuals

282

were stored in separate petri dishes at 1°C until use. Care was taken to avoid dehydration during storage. Although the semen retained a capacity for fertilizing the eggs for more than a week, the samples were used within two days after stripping. Lamprey Ringer (LR) consisted of 137mM NaCl, 2.9mM KCl, and 2.1mM CaCl (pH7.4 with 10mM Hepes-NaOH) [16]. Test solutions used for the assessment of the initiation of sperm motility were 1/10 strength LR and solutions con- taining various concentrations of NaCl, KCl, or glucose; they were buffered with 10 mM Hepes- NaOH (pH7.4). To examine whether external Ca** ions participate in the initiation of sperm motility, we used 1/10 strength Ca** free LR containing 13.7mM NaCl, 2,9mM KCI, and 10 mM EGTA (pH 7.4 with 10 mM Hepes-NaOH). For assessment of sperm motility, two wl of the semen was quickly mixed with 198 yl of test solu- tion in a test tube. Ten wl of the mixture was placed on a glass slide without a cover slip within 5 seconds after mixing and examined under a micro- scope. Sperm movement was observed under a light microscope (200) and recorded by a video recorder. Sperm represented to be motile when

100

E

(<b)

[ok

7?)

£&

=

° eRe

€

—

fo)

se

gl

(aee< ates ail 1 0

glucose

Fic. 1.

W. KoBayASHI

the sperm head showed forward movement. Per- centage of moving sperm (more than 50 cells) and swimming speed of the moving sperm (n=10) were monitored by tracing the location of sperm heads on a hard copy of the frame by a video printer. Assessment of sperm motility was carried out in samples from three males. All experiments were performed in triplicate for each sample and performed at room temperature. No significant difference was recognized among these males. Data (mean +SE, n=3) from a representative fish are therefore presented in the figures. Student’s t-test was used to determine statistically significant differences, which were considered significant when P<0.05.

RESULTS

Lamprey sperm were immotile in the seminal plasma. Upon dilution of semen with full strength LR, less than 1% of sperm began to move within the first few seconds. In the 1/10 strength LR, nearly all sperm immediately showed active for- ward movement; most of the sperm continued the movement for more than 5 minutes. Sperm initi-

Oo Na eK 4 glucose O buffer

(mmol/kg) =

4 400

(mmol/kg) (mosmol/ kg)

Effect of NaCl, KCl or glucose on the motility of lamprey sperm. Percentage (mean + SE) of motile sperm 10

seconds after dilution of semen with the test solutions are presented.

Sperm Motility in Lamprey 283

ated similar movement in the 1/10 strength Ca** Figure 1 clearly shows that sperm were com- free LR. It is clear that the initiation of sperm pletely immotile in NaCl and KCI solutions at motility does not require the presence of Ca** ions _ concentrations from 100 mmol/kg (200 mOsmol/

in external medium. kg) to 200 mmol/kg (400 mOsmol/kg). Percen- o Na

Se eK

g A glucose

& Oo buffer

Baga

be

oe)

fe)

Om

>

E

i

Dh eit:

Q

id)

0 e (evry oe seerreer be) 2 ! l Ith ETAT 0 50 100 Na, K (mmol/kg) PORES TEINS Oe POL IS 1 =i 0 100 200 glucose (mmol /kg)(mosmol/kg)

Fic. 2. Effect of NaCl, KCl or glucose on the velocity (mean+SE) of forward movement of sperm 10 seconds after dilution of semen with test solutions.

100 OoNa 25 eK

E 4 glucose 50

® 0 buffer

a

(7p)

(<b)

Y

°

£

—

ie)

=

0

pS a as as as sss nnnnsnsnsnsssssss————————)

010 60 300 600 Time after dilution (sec)

Fic. 3. Effect of the NaCl, KCI or glucose on the duration of the sperm motility. Changes in percentage of motile sperm in the test solutions (50 mOsmol/kg) and in buffer solution alone.

284 W. KOBAYASHI

tage of motile cells increased as the concentration of the electrolytes decreased; a maximum value was obtained at a concentration between 0 and 25 mmol/kg (50 mOsmol/kg). A similar lack in the sperm motility in glucose solutions was observed at concentrations between 300 and 400 mmol/kg (mOsmol/kg). The percentage of sperm that showed forward movement increased as the con- centration of glucose decreased to less than 150 mOsmol/kg. The value at 150mOsmol/kg is apparently higher in KCl solution than that in NaCl or glucose solution (P<0.01). These results indicate that an isosmotic solution of electrolytes (NaCl, KCl) and nonelectrolytes do not induce forward movement in lamprey sperm, and that a decrease in the osmotic pressure of the extracellu- lar medium is essential for the initiation of lamprey sperm motility.

Velocity of the forward movement was affected by osmotic concentration of the solution; a max- imum speed was obtained at 50 mOsmol/kg. At this osmolality, sperm showed faster movement in KCI solution than that in NaCl (P<0.001) or glucose (P<0.005) solution (Fig. 2). At 100 and 150 mOsmol/kg, sperm velocity in NaCl or KCl solution was higher than that in glucose solution.

The percentage of motile sperm was estimated at 50 mOsmol/kg. Figure 3 shows that about 40% of total sperm were still motile in NaCl or KCl solution 5 minutes after dilution. In these solu- tions, the percentage of moving cells reached zero at 10 minutes. The life span of sperm was, how- ever, short in a glucose or buffer alone; in these solutions, motile sperm was nearly zero at 6 mi- nutes.

DISCUSSION

This study demonstrates that, although lamprey sperm are immotile in either electrolytes or nonelectrolyte solutions isotonic to blood plasma, they showed active forward movement when the semen is diluted with hypotonic media. These facts suggest that the external signal for the initia- tion of sperm motility in the lamprey is hyposmo- lality. Although no data has been reported on the osmolality of lamprey seminal plasma, it is known that in many species the osmolality of seminal

plasma is nearly the same as that of blood plasma or isotonic saline solution (Ringer solution) [4, 8]. Osmolality of the seminal plasma is probably in- volved in the quiescence of lamprey sperm in the semen.

The percentage of motile sperm (at 150 mOsmol/kg) and the velocity of forward move- ment (at 50 mOsmol/kg) were higher in the solu- tion of KCl in the solution of NaCl or glucose. Similar action of potassium on sperm motility is reported in cyprinid fish [10]. Several investigators have reported that in many vertebrates the seminal plasma contains a high concentration of potassium ions [4, 8, 10, 14]. These facts suggest that the concentration of potassium may be also related to the motility of lamprey sperm when the seminal plasma is diluted in the natural spawning.

Morisawa and Suzuki [7] have reported that some species of freshwater (cyprinids) and marine fish adopt a change in osmolality of the external solution as a signal for the initiation of sperm motility: an external signal for the freshwater species is a decrease in osmolality, while a signal for the marine fish is an increase in osmotic pressure. Amphibian sperm (newt and Xenopus) seem to possess a mechanism similar to the fresh- water fish [3, 4, 10, 14]. In contrast, the external signal for the initiation of sperm motility in salmo- nids (primitive type of fish) is the extracellular concentration of potassium ion [9]: motility of salmonid sperm is suppressed by a high concentra- tion of potassium in the seminal plasma and is induced when a decrease in the potassium concen- tration is caused after dilution of the semen with freshwater at spawning. It is well known that the life cycle of some lamprey species (“parasitic” species) resembles that of some salmonid species [2]. Both groups of animals invariably spawn their gametes in freshwater (river). The embryos hatch there and the young animals migrate to the sea. They grow in the seawater environment and return to freshwater for spawning (anadromous migra- tion). In spite of such a similarity of the life cycle, the external factors that regulate the sperm motil- ity in the two primitive vertebrates, lampreys and salmonids, are quite different. On the other hand, the pattern of the early stages of the embryonic development (cleavage) in the lamprey is very

Sperm Motility in Lamprey

similar to that of amphibians [13]. Electrophysi- ological studies of fertilization have further re- vealed that, unlike the case of fish eggs [11], lamprey eggs as well as anuran amphibians possess a Cl -dependent fertilization potential that par- ticipates in a fast block to polyspermy [1, 5]. These facts, together with the present results, indicate that the physiological properties of lamprey ga- metes are very similar to those of anuran amphi- bians. This situation is of interest in relation to the evolution of lower vertebrates and the conver- gence of reproductive strategy.

ACKNOWLEDGMENTS

I wish to thank Professor T. S. Yamamoto, Hokkaido university, for critical reading of the manuscript.

REFERENCES

1 Elinson RP (1986) Fertilization in amphibians: The ancestry of the block to polyspermy. Int Rev Cytol 101: 59-100

2 Hardisty MW (1979) Biology of the cyclostomes. Chapman and Hall, London.

3 Hardy MP, Dent JN (1986) Regulation of motility in sperm of the red-spotted newt. J Exp Zool 240: 385-396

4 Inoda T, Morisawa M (1987) Effect of osmolality on the initiation of sperm motility in Xenopus laevis. Comp Biochem Physiol 88A: 539-542

5 Kobayashi W, Baba Y, Shimozawa T (1991) Ferti- lization potential in the lamprey, Lampetra japonica eggs. Zool Sci 8: 1097

6

14

15

16

285

Morisawa M (1985) Initiation mechanism of sperm motility at spawning in teleosts. Zool Sci 2: 605-615 Morisawa M, Suzuki K (1980) Osmolality and potassium ion: Their roles in initiation of sperm motility in teleosts. Science 210: 1145-1147 Morisawa M, Hirano T, Suzuki K (1979) Changes in blood and seminal plasma composition of the ma- ture salmon (Oncorhynchus keta) during adaptation to freshwater. Comp Biochem Physiol 64A: 325-329 Morisawa M, Suzuki K, Morisawa S (1983) Effects of potassium and osmolality on spermatozoan motil- ity of salmonid fishes. J exp Biol 107: 105-113 Morisawa M, Suzuki K, Shimizu H, Morisawa S, Yasuda K (1983) Effects of osmolality and potas- sium on motility of spermatozoa from freshwater cyprinid fishes. J exp Biol 107: 95-103

Nuccitelli R (1980) The electrical changes accom- panyingfertilization and cortical vesicle secretion in the medaka egg. Dev Biol 76: 483-498

Stoss J (1983) Fish gamete preservation and sper- matozoan physiology. In “Fish physiology Vol 9B” Ed by WS Hoar, DJ Randall, EM Donaldson Academic Press, New York, pp 305-350

Tahara Y (1988) Normal stages of development in the lamprey, Lampetra reissneri (Dybowski). Zool Sci 5:109-118

Viljoen BCS, Van Vuren JHJ (1991) Physical composition of the semen of Labeo ruddi and Labeo rosae (Pisces: Cyprinidae). Comp Biochem Physiol 98A: 459-462

Yamamoto T (1961) Physiology of fertilization in fish eggs. Int Rev Cytol 12: 361-405

Yamamoto TS (1956) Digestion of egg envelopes and the chemical properties of the lamprey’s egg, Lampetra japonica. J Fac Sci Hokkaido Univ Ser VI Zool 12: 273-281

, avacegt ti Peskin gai wr eranericenst aubastiaaiteadte a APS he gale a RAR amare RRM TE WIC Tyr Sempre INTE frie TEAS » e at a dl " fod AUDA A . ' 7

we “SUBNET PORE Sibyl di ative aii venoent: ' tila \ \ 1 ' ay bo etl «i “Bey Ee pny a P dra Ah 4 wae ease | ~omeyuads falieenad = | iy of te mOLIoy hla lee hag abt otuET Peiskfotametquil og “AOR, SUE (hes elon) ial hth won hephenty” ee aed ree OPE et lame phate eee) ma ae: oR dentegitteel Chee mailing ‘

- \

. raerhis) iF nets Arya, hl emepeeay ¢ | Per dye G7 leh bine b: re Vs Maiti Fabien Tee spobtt on wil me 5 | i717 } “ 4 Mronmwna 1 [seed 7 ay : ’ . yee I a ie a oar ortcieseutt peotmnpiiite : i ASH tsrvtuie 3) TY es tii Da ea aria tt Fin ORTES seth tyne) Ba . ee

aX “nhc Mia) eT DP wee = reperiond 2a lryaasnaeia

—

; iedh) aca 4 Hinton Ten qi mary peed " sive CONtuie who cone Bint : 7 ‘Fs roth ¥ ' bh. A Jie 44 SD Ce hicaey ni nel} Lid yh ui ieee 3 bed Rope NG» Hse OD OA" ve thts Nie Wee Mer Snr et io, BORON SUAS Cds Ba oat “abetted enstnie ae , hal Pay tort i a artsy Ga ee ae : se ey ne. iiney Wak I i Huey His : aril ' it kaes 4 <i ¥ aan fs wat! 4 i Mae i) as Ty AA Bee 5. vt 0 YS tis at ; Lvigiirge ity: ae eieeenal | eae

MEL Ae Wy 4-4i¢ th ph; stich onthe naeglinaes ta Mad ite FASE De west VE al! ee ee eee gros “bet bee iy } r+ "yt read 10 ‘tpatt t yl pring t) Reohina: [ SPE ab Eo, 2S loos, nad 1: Ms G4 { L SOPOCUIMG tbe. Al i ss +} ea x lan? vou : Hall. { ‘ealpon ie te : yyy That : tr, MELT VAdie i epithe Wie Be sitioer tie, 8 rah | th? ANB ARH ; VED {Po staat och) CAVE Nei ebrquannl he ae : , ae: pve he some otea anya | get anepay al AEN e

Sp RT) Ue BK yi

= — wn ee oe hep hat

tas

te: Whalley

\ t t * = , pice i, ld S nuwey F : Zz rivl NS - May : we i ta I i ur Bh tly

{ i Tae Hie wee Veet aint ;

wir etefant, On La 7 J ad i yt ute

ZOOLOGICAL SCIENCE 10: 287-294 (1993)

Identification and Immunological Properties of an Olfactory System-Specific Protein in Kokanee Salmon (Oncorhynchus nerka)

MunetaKA Suimizu!*, HipeAKt Kupo!”, Hirosut Uepa!*,

AxtHiko Hara*, Kens SHIMAZAKI’ and KoHEI YAMAUCHI’

'Toya Lake Station for Environmental Biology, Faculty of Fisheries, Hokkaido University, Abuta 049-57, *Research Institute of North Pacific Fisheries, Faculty of Fisheries, Hokkaido University, Hakodate 041, *Department of Biology, Faculty of Fisheries, Hokkaido University, Hakodate 041, and 4Nanae Fish Culture Experimental Station, Faculty of Fisheries, Hokkaido University, Nanae 041-11, Hokkaido, Japan

ABSTRACT— An olfactory system-specific protein of 24 kDa (N24) was identified in kokanee salmon (Oncorhynchus nerka) by the comparison of proteins restricted to the olfactory system (olfactory epithelium, olfactory nerve and olfactory bulb) with those found in other parts of the brain (telencephalon, optic tectum, cerebellum, hypothalamus) and hypophysis. A polyclonal antibody to N24 was raised in a rabbit, and the specificity of the antiserum was examined by Western blotting analysis; the antiserum recognized only one 24 kDa band in the olfactory system but not in other parts of the brain in kokanee salmon, sockeye salmon (O. nerka), masu salmon (O. masou) and chum salmon (O. keta). Immunocytochemical analysis revealed that positive immunoreactivity occurred exclusively in the olfactory nerve and in some olfactory neuroepithelial cells. Both at the time of imprinting of maternal stream odorants and at the time of homing to the maternal stream, the immunoreactivity of N24 in fish in the maternal stream was stronger than that in seawater fish. The present study indicates that an olfactory system-specific protein may be importantly related to both imprinting and homing

© 1993 Zoological Society of Japan

mechanisms in salmonids.

INTRODUCTION

The imprinting and homing behaviors of salmon in relation to their maternal stream constitute one of the most interesting phenomena in salmon biology. Since Hasler and Wisby [10] introduced the olfactory hypothesis for anadromous salmo- nids, olfactory imprinting and homing mechanisms have been studied in many behavioral experiments (e.g., [9, 24]). Several neurophysiological ap- proaches using electrophysiological techniques have also pointed to the olfactory recognition of maternal stream water during spawning migration in several salmonids [8, 26]. It is now widely accepted that in juvenile salmon the olfactory

Accepted February 2, 1993 Received December 14, 1992

* To whom correspondence and reprint requests should be addressed.

system plays important roles in imprinting of the maternal stream odorants during downstream migration, and that in adult salmon the early olfactory exposure allows recognition of the mater- nal stream odorants during upstream spawning migration (homing). However, few attempts have been made to investigate biochemical aspects of the olfactory system in any salmonid species. There are numerous literatures indicating a close relation between olfactory-specific proteins and recognition and discrimination of odorants in higher vertebrates[1, 13, 16]. Several olfactory- specific proteins have been identified and used as molecular markers to study olfactory function, such as the olfactory marker protein in the mammalian olfactory bulb [11, 14], carnosin in tetrapod olfactory systems [3, 15], and olfac- tomedin in frog olfactory neuroepithelium [23]. The present study was desinged to compare

288 M. Suimizu, H. Kubo et al.

proteins restricted to the olfactory system (olfac- tory epithelium, olfactory nerve and olfactory bulb) with those found in other parts of the brain (telencephalon, optic tectum, cerebellum, hypo- thalamus) and hypophysis of several salmonid spe- cies. In addition, two immunological analyses (Western blotting and immunocytochemistry) were conducted to investigate the possible role of an olfactory system-specific protein in olfactory functions and to evaluate its possible usage as a molecular marker in the study of olfactory mechanisms.

MATERIALS AND METHODS

Fish

Male and female mature kokanee salmon (Oncorhynchus nerka; land-locked sockeye sal- mon), 3 to 5 years old, were caught in Lake Toya from October 22 to November 19, 1991. Mature sockeye salmon (QO. nerka) of both sexes, 3 to 5 years old, were collected from the Chitose Branch of the Hokkaido Salmon Hatchery on October 29, 1991. Adult chum salmon (O. keta) of both sexes, 3 to 6 years old, were captured at sea and in fresh water at two different stages of sexual maturation (pre-spawning fish about 10-20 days before final maturation and spawning fish) on September 25, 1991. Wild one-year-old juvenile masu salmon (O. masou) were caught once a month in the Shakotan River from November 21, 1991 to May 25, 1992.

Fish were anesthetized with 10% ethyl p- aminobenzoate, and their heads were removed by decapitation and placed on ice. Under a dissecting microscope, olfactory organs and brains were dis- sected into olfactory epithelium, olfactory nerve, olfactory bulb, telencephalon, optic tectum, cere- bellum, hypothalamus and hypophysis. Each tis- sue was washed quickly in ice-cold salmon Ringer (0.15 M NaCl, 0.31mM KCl, 0.03 mM MgSO,- 7H20, 0.34 mM CaCl,:2H,0O, 0.4 mM glucose, 0.4 mM Hepes buffer, pH 7.5) containing aprotinin (1 pvg/ml), phenylmethylsulfonyl fluoride (PMSF; 1 mM) and ethylenediaminetetraacetic acid (EDTA; 1 mM). The tissue was homogenized in about 1.5 volumes of salmon Ringer and centrifuged at 10,000 g for 10min. The supernatant fluid was

stored at —30°C and used for gel electrophoresis.

Gel electrophoresis

Sodium dodecyl sulfate polyacrylamide gel elec- trophoresis (SDS-PAGE) [12] was performed us- ing a Protean II slab gel system (Bio-Rad, South Richmond, CA, USA). Samples containing 50 ug protein, as determined by Lowry’s method, were heated at 100°C for 2 min in double volumes of SDS sample buffer containing 2-mercaptoethanol, and applied to slab gels consisting of 7.5-17.5% logarithmic gradient of polyacrylamide and 5% stacking gel. Electrophoresis was carried out at 12.5 mA for stacking gel followed at 25 mA for gradient gel. Molecular weight markers (Pharma- cia Fine Chemicals, Uppsala, Sweden) for deter- mination of low molecular weight were run in a separate lane. Gels were fixed and stained for 2 hr in 0.1% Coomassie Brilliant Blue in 5% ethanol- 5% acetic acid solution, destained overnight in 10% acetic acid, soaked in 20% ethanol-5% acetic acid-2.5% glycerin, and dried with gel dryer.

Two-dimmensional (isoelectric-SDS) PAGE (2D-PAGE) [19] was also carried out to obtain the evidence whether an olfactory system-specific pro- tein appeared single spot or not.

Antiserum production

A protein band corresponding to an olfactory system-specific protein, designated as N24, was prepared from olfactory bulbs of mature kokanee salmon of both sexes. The band was excised carefully with scissors, and eluted from gel slices using an Electro-Eluter (Bio-Rad, South Rich- mond, CA, USA) following the manufacturer’s instructions. An elution buffer was constituted with 25 mM Tris, 192 mM glycine and 0.1% SDS, and elution was done at 10 mA for 2 hr.

A polyclonal antiserum was raised in a rabbit using lymph node injections. Initial injection of 35 pg eluted N24 was made into inguinal lymph nodes in Freund’s complete adjuvant. Subsequent boost- er injections of 90 ug N24 emulsified with Freund’s incomplete adjuvant were made intradermally into the back of rabbit at four weekly intervals. One week after the final booster injection, a test bleed- ing was taken from the ear vein, and the serum was assayed by Western blotting analysis.

Olfactory System-Specific Protein in Salmonids 289

Western blotting

Protein samples were resolved by SDS-PAGE and transferred to Immobilon PVDF membranes (Millipore, Bedford, MA, USA) by the method of Towbin et al. [25]. The PVDF membranes were incubated for 1 hr with 1% skim milk in Tris- buffered saline (TBS, pH7.4) to reduce non- specific protein absorption, and allowed to react for 2hr with polyclonal antiserum against N24 (working dilution from 1:100 to 1:5000) in a solution of 3% bovine serum albumin and 1% normal goat serum in TBS. After rinsing with TBS, the membranes were incubated for 1 hr with horseradish-peroxidase-labeled goat anti-rabbit IgG (Cappel Laboratories, West Chester, PA, USA) at a dilution of 1:1000 in TBS, and de- veloped in 4-chloro-1-naphthol solution (60 mg in 100 ml TBS containing 0.01% H>O>). The reac- tion was stopped after 3-5 min by washing in distilled water, and the membranes were air-dried.

Immunocytochemistry

The olfactory system and other parts of the brain of mature kokanee salmon were fixed with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.3) for 18hr at 4°C. After rinsing with 0.1M phosphate buffer containing 10% sucrose, tissues were embedded in Tissue-Tek OCT compound (Miles Laboratories, Elkhart, IN, USA) and stored at —30°C, to be sectioned later at 10-~m thickness in a cryostat. Tissue sections were im- mersed in 0.3% hydrogen peroxide in methanol for 30 min to eliminate endogenous peroxidase activity. After rinsing with phosphate-buffered saline (PBS), sections were incubated with 1% normal goat serum in PBS for 15 min, and im- munoreacted with anti-N24 serum (working dilu- tion 1:2500 or 1:5000) for 2 hr at room tempera- ture. The strepto-avidin-biotin complex (sABC) method (Histofine, Nichirei, Tokyo, Japan) was used for the immunocytochemistry. In order to confirm specificity, normal rabbit serum or PBS was substituted for anti-N24 serum.

RESULTS

Soluble extracts of the olfactory system and

other parts of the brain from mature kokanee salmon were resolved by SDS-PAGE (Fig. 1A). Several protein bands were observed in extracts of the olfactory system but not in other parts of the brain or the hypophysis. Among these proteins, the most clearly visible protein band (arrowhead in Figure 1A), which had a molecular weight of 24 kDa, was designated N24, and appeared single spot by the analysis of 2D-PAGE (manuscript in preparation). This protein was eluted from gel slices for antibody generation.

In Western blotting analysis, a polyclonal anti- serum to N24 at dilutions of 1:500 and 1: 1000 recognized only the band whose position was iden- tical to N24 in extracts of kokanee salmon olfac- tory bulb. At a dilution of 1:100, presumably weak non-specific reactions were observed in bands of molecular weight around 40kDa. At a dilution of 1:5000, a faint reaction only was observed in the 24kDa band. At a dilution of 1:2500, the antiserum reacted with the 24kDa protein in the olfactory system, which was absent from other parts of the brain and hypophysis in kokanee salmon (Fig. 1B).

Electrophoretic patterns of soluble extracts of the olfactory system and telencephalon of chum salmon, sockeye salmon and masu salmon were similar to those of kokanee salmon, and a similar olfactory system-specific protein of molecular weight 24 kDa was recognized in all these salmo- nid species (Fig. 2).

During parr-smolt transformation in masu salm- on, although there were no distinct changes in electrophoretic patterns of the olfactory system and telencephalon between parr and smolt, the immunoreactivity of N24 to the antiserum was more intense in parr than in smolt (Fig. 3). Simi- larly, during spawning migration of chum salmon from coastal sea to maternal river, electrophoretic patterns of olfactory nerve, olfactory bulb and telencephalon showed no prominent changes. However, the cross-reactivity of anti-N24 serum was different in fish from the sea and the river; the antigenicity of N24 was more intense in the mater- nal river fish sample (Fig. 4).

The olfactory rosette of mature kokanee salmon stained with Nissl solution showed olfactory epithelium and several branches of the olfactory

290 M. Suimizu, H. Kupo et al.

sn Me Joi 3 TF (Di Stage so 67% Bf eilte2ye Sie 4. SiGe

Fic. 1. SDS-PAGE (A) and Western blotting (B) of soluble extracts of the olfactory system (1, olfactory epithelium;

Fic.

2, olfactory nerve; 3, olfactory bulb), other parts of the brain (4, telencephalon; 5, optic tectum; 6, cerebellum; 7, hypothalamus) and hypophysis (8) from kokanee salmon. Arrowhead in A identifies an olfactory system-specific protein of molecular weight 24 kDa (N24). Anti-N24 serum (1 :2500) recognizes a band of molecular weight 24 kDa in the olfactory system but not in other parts of the brain. Numbers at left are Mr standards x 10°.

94- ri

67-

ae x :

30- ETL

20- i id ikon 1k 20 JOvt4s, Se GN iN/AsSi,SRlO 46:2); Seon 4d nS 6cu Teele Chum Sockeye Masu Chum Sockeye Masu Salmon Salmon Salmon Salmon Salmon Salmon

2. SDS-PAGE (A) and Western blotting (B) of soluble extracts of the olfactory system (1, 4, 8, olfactory nerve; 2, 5, 9, olfactory bulb; 7, olfactory epithelium) and telencephalon (3, 6, 10) from chum salmon, sockeye salmon and masu salmon. Anti-N24 serum (1 :2500) recognizes a band of molecular weight 24 kDa only in the olfactory system. Numbers at left are Mr standards x 10°.

Olfactory System-Specific Protein in Salmonids 291

% i k = b% = % £ & t =

B eno Seep ar D8e Baie. 2S) 4 i POD sh 3h iA,

Parr Smolt Parr Smolt

Fic. 3. SDS-PAGE (A) and Western blotting (B) of soluble extracts of the olfactory nerve system (1, olfactory epithelium; 2, olfactory nerve; 3, olfactory bulb) and telencephalon (4) in masu salmon during parr-smolt

transformation. The immunoreactivity of anti-N24 serum (1:2500) is more intense in parr than in smolt. Numbers at left are Mr standards x 10°.

Menta SAbeatileh tic Site tlieic leno ip 2.3 Wp S gl 2.3 On Ob a On Ob ab

Fic. 4. SDS-PAGE (A) and Western blotting (B) of soluble extracts of the olfactory system (On, olfactory nerve; Ob, olfactory bulb) and telencephalon (T) in chum salmon during spawning migration (1, coastal sea; 2,

pre-spawning fish in the maternal river; 3, spawning fish in the maternal river). The immunoreactivity of anti-N24

is more intense in fish from the maternal river than in fish from the coastal sea. Numbers at left are Mr standards < 10°.

292 M. Suimizu, H. Kupo et al.

Fic. 5.

G

Adjacent 10-m thick sections of the olfactory rosette of mature kokanee salmon stained with Nissl solution

(A), sABC method with anti-N24 at a dilution of 1:2500 (B), and sABC method with normal rabbit serum at a

dilution of 1: 2500 (C).

the olfactory nerve. MP, melanophore; OE, olfactory epithelium; ON, olfactory nerve.

nerve (Fig. 5A). The sABC method demonstrated that N24 immunoreactivity was localized exclusive- ly in almost all branches of olfactory nerve and in some olfactory epithelial cells, presumably neuroepithelial cells (Fig. 5B). The immunoreac- tivity of apical portions of the olfactory epithelium was non-specific because the same portions were also stained in control sections, in which anti-N24 serum was replaced with normal rabbit serum (Fig. 5C). Specific immunoreactivity was never observ- ed in other nervous tissues studied.

DISCUSSION

Electrophoretic comparison of proteins re- stricted to the olfactory system with those in other parts of the brain in kokanee salmon demonstrated the existence of several olfactory system-specific proteins. Among them, the most clearly visible protein designated as N24 based on its 24kDa molecular weight was used to raise a specific antibody in a rabbit. Western boltting analysis revealed that the antiserum recognized only this 24 kDa band in the olfactory system, which was absent from other parts of the brain not only in kokanee salmon but also in sockeye salmon, masu

N24 immunoreactivity occurs in some olfactory neuroepithelial cells and in branches of

x 250.

salmon and chum salmon. However, no cross- reactivity of anti-N24 serum was observed in the olfactory system of carp (unpublished data). Therefore, all salmonids examined to date possess a 24kDa protein (N24) specific to the olfactory system.

The olfactory sensory cells (olfactory neuro- epithelial cells) are bipolar neurons, whose den- drites project to the external environment and axons to the olfactory bulb, which form the olfac- tory nerve. Preliminary immunocytochemical study showed that positive N24 immunoreactivity occurred exclusively in the olfactory nerve and in some olfactory neuroepithelial cells. The specific immunoreactivity was also found in the olfactory bulb where the olfactory nerve was penetrated (unpublished data). We could not identify types of neuroepithelial cells at the light microscopic level. These observations suggest that N24 has possible roles in neurotransduction from neuroepithelial cells to olfactory bulb; however, further detailed immunocytochemical studies at the electron mi- croscopic level are needed to clarify this point.

A mammalian olfactory marker protein of molecular weight 20 kDa is localized in olfactory sensory neuron cell bodies and axons; its appear-

Olfactory System-Specific Protein in Salmonids 293

ance coincides with the establishment of sensory synapses in the olfactory bulb [7]. Although it seems unlikely that the mammalian olfactroy mark- er protein is identical to N24 because of the difference in their molecular weight, further im- munological approaches are necessary. Carnosine is the predominant dipeptide present in the olfac- tory neurons and nerve in mammals, birds, repti- lians and amphibians, but has not been reported in fishes [3]. Olfactomedin, a 57 kDa glycoprotein, is localized exclusively in frog olfactory neuroepithe- lium [23]. These olfactory-specific proteins may play important roles in chemoreception, however, their precise functions remain uncertain.

Another biochemical approach in higher verte- brates concerns the involvement of olfactory re- ceptor proteins which activate adenylate cyclase [21, 22] via a GTP-binding protein [2] leading to the opening of cyclic nucleotide-gated cation chan- nels [5, 18]. A multigene family of GTP-binding protein-coupled receptors expressed uniquely in olfactory tissues has recently been reported [4]. Immunocytochemical studies using monoclonal antibodies against rabbit olfactory bulb have de- monstrated subclasses of olfactory nerve fibers and their projections to the olfactory bulb [6, 17]. These aspects need investigation in teleosts.

During downstream migration of masu salmon (at the time of imprinting) and upstream spawning migration of chum salmon (at the time of homing), the immunoreactivity of N24 in fish in the maternal stream was stronger than that in seawater fish. Either qualitative changes (including immunolo- gical properties) or quantitative changes of N24 may occur during migration. We found no evident changes in N24 or in other protein bands as judged by their staining properties with Coomassie blue. None the less, the change observed indicates that N24 may have a role in both imprinting and homing mechanisms in salmonids. Inhibitors of RNA and protein synthesis have reported to in- hibit olfactory bulbar discrimination in maternal stream water of chinook salmon, as judged by electrophysiological methods [20]. So it is reason- able to consider that some specific proteins must be involved in the homing behavior of salmonids.

In conclusion, the present study identifies an olfactory system-specific protein (N24) in four

salmonid species and describes some of its im- munological properties. N24 may prove to be good molecular marker for the study of salmonid olfactory functions. Detailed biochemical and cytological analyses of N24 as well as the other olfactory system-specific proteins are now in prog- ress in our laboratory.

ACKNOWLEDGMENTS

We thank Professor Howard A. Bern, University of California, for his interest in the present study and for critical reading of the manuscript. We also thank Dr Osamu Hiroi, Dr Masahide Kaertyama, Hokkaido Sal- mon Hatchery, and Mr Hiroshi Kawamura, Hokkaido Fish Hatchery, for providing the chum salmon, the sockeye salmon and the masu salmon, respectively, and Mr Hiroyuki Haruna and Mrs Rirako Yamada, Toya Lake Station for Environmental Biology, for their tech- nical assistance. This study was supported in part by the Special Grant-in-Aid for Promotion of Education and Science in Hokkaido University Provided by the Ministry of Education, Science and Culture of Japan, and by a reseach grant from the Akiyama Foundation to H.U.

REFERENCES

1 Anholt RRH (1989) Molecular physiology of olfac- tion. Am J Physiol 257: C1043—C1054

2 Anholt RRH, Mumby SM, Stoffers DA, Girard PR, Kuo JF, Snyder SH (1987) Transductory proteins of olfactory receptor cells: identification of guanine

’ nucleotide binding proteins and protein kinase C.

Biochemistry 26: 788-795

3 Artero C, Marti E, Biffo S$, Mulatero B, Andreone C, Margolis FL, Fasolo A (1991) Carnosine in the brain and olfactory system of amphibia and reptilia: a comparative study using immnocytochemical and biochemical methods. Neurosci Lett 130: 182-186

4 Buck L, Axel R (1991) A novel multigene family may encode odorant receptors: a molecular basis for odor recognition. Cell 65: 175-187

5 Dhallan RS, Yau KW, Schrader KA, Reed RR (1990) Primary structure and functional expression of cyclic nucleotide-activated channel from olfactory neurons. Nature 347: 184-187

6 Fujita SC, Mori K, Imamura K, Obata K (1985) Subclasses of olfactory receptor cells and their segre- gated central projections demonstrated by a mono- clonal antibody. Brain Res 326: 192-196

7 Graziadei GAM, Stanley RS, Graziadei PPC (1980) The olfactory marker protein in the olfactory system of the mouse during development. Neurosci 5: 1239-1252

10

11

12

13

14

294

Hara, TJ (1970) An electrophysiological basis for olfactory discrimination in homing salmon: a review. J Fish Res Board Can 27: 565-586

Hasler AD, Scholz AT (1983) Olfactory imprinting and homing in salmon. Springer-Verlag, Berlin, pp 1-134

Hasler AD, Wisby WJ (1951) Discrimination of stream odors by fishes and relation to parent stream behavior. Am Naturalist 85: 223-238

Kott JN, Vickland H, Dong XM, Westrum LE (1992) Development of olfactory marker protein and tyrosine hydroxylase immunoreactivity in the transplanted rat olfactory bulb. Exp Neurol 115: 132-136

Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T,. Nature 227: 680-685

Lancet D (1986) Vertebrate olfactory reception. Annu Rev Neurosci 9: 329-355

Margolis FL (1972) A brain protein unique to the olfactory bulb. Proc Nat Acad Sci USA 69: 1221- 1224

Margolis FL (1974) Carnosine in the primary olfac- tory pathway. Science 184: 909-911

Margolis FL, Getchell TV (1991) Receptors: cur- rent status and future directions. In “Perfumers: Art, Science and Technology” Ed by P Muller, D Lamparsky, Elsevier Applied Science Publishers, Essex, pp 481-498

Mori K, Fujita SC, Imamura K, Obata K (1985) Immunohistochemical study of subclasses of olfac- tory nerve fibers and their projections to the olfac-

M. Suimizu, H.

Kupo et al.

18

19

20

21

22

23

24

25

26

tory bulb in the rabbit. J Comp Neurol 242: 214-229 Nakamura T, Gold GH (1987) A cyclic nucleotide- gated conductance in olfactory receptor cilia. Nature 325: 442-444

O'Farrell PH (1975) High resolution two- dimmensional electrophoresis of proteins. J Biol Chem 250: 4007-4021

Oshima K, Gorbman A, Shimada H (1969) Mem- ory-blocking agents: effects on olfactory discrimina- tion in homing salmon. Science 165: 86-88

Pace U, Hanski E, Salomon Y, Lancet D (1985) Odorant-sensitive adenylate cyclase may mediate olfactory reception. Nature 316: 255-258

Sklar PB, Anholt RRH, Snyder SH (1986) The odorant-sensitive adenylate cyclase of olfactory re- ceptor cells: differential stimulation by distinct clas- ses of odorants. J Biol Chem 261: 15538-15543 Snyder DA, Rivers AM, Yokoe H, Menco BPhM, Anholt RRH (1991) Olfactomedin: purification, characterization, and localization of a novel olfac- tory glycoprotein. Biochemistry 30: 9143-9153 Stabell OB (1984) Homing and olfaction in salmo- nids: a critical review with special reference to the Atlantic salmon. Biol Rev 59: 333-388

Towbin H, Staehelin T, Gordon J (1979) Elec- trophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Nat Acad Sci USA 76: 4350-4354 Ueda K (1985) An electrophysiological approach to the olfactory recognition of homestream waters in chum salmon. NOAA Tech Rep NMEFS 27: 97-102

ZOOLOGICAL SCIENCE 10: 295-306 (1993)

© 1993 Zoological Society of Japan

Development of the Brittle Star, Ophioplocus japonicus H.L. Clark. I

MieKo Komatsu and ToMoKo SHOSAKU

Department of Biology, Faculty of Science, Toyama University, Toyama 930, Japan

ABSTRACT—The entire process of development from eggs to juveniles in the brittle star, Ophioplocus japonicus, is described with special attention to external morphology. The breeding of this brittle star in Toyama Bay, Sea of Japan, occurs in August. The eggs are spherical, orange, and 300 “m in average diameter. Cleavage is total and radial. Gastrulation takes place by invagination. Larvae possessing tube-foot buds and ophiuroid rudiments on the posterior part of their bodies, develop into the vitellaria stage. Vitellariae of the present species have transverse ciliary bands, but no larval skeletons. Six days after fertilization at 25°C, metamorphosis is completed and the resulting juvenile is pentagonal in outline and about 400 “m in diameter. The present study shows that ciliary bands are formed in an

earlier larval stage than was reported previously.

INTRODUCTION

A considerable number of feeding (plankto- trophic) larvae, the ophiopluteus, has been re- ported in Ophiuroidea [5]. Development of ophio- pluteus larvae is generally called an indirect or planktotrophic type [5, 9]. However, ophiuroids with yolky eggs have diverse developmental courses as will be shown below. The developmen- tal pattern was classified into 7 types based on the breeding and brooding styles, the nature of the larvae, and other features [20]. On the other hand, Hendler [5] designated 3 types of development (planktotrophic, direct and abbreviated) based on the egg size, life span, larval size and some other factors. From his view, development of the vitel- laria belongs to the abbreviated type. However, the abbreviated type has some heterogeneous lar- val forms, whereas the planktotrophic and direct types are distinctive. Thus, agreement has not yet been achieved on what are the developmental types and on their nomenclature in Ophiuroidea. This is partly because of the comparative paucity of knowledge available on the development pas- sing through other than typical ophiopluteus larval

Accepted February 17, 1993 Received November 13, 1992

stages.

At present, 8 species are known to possess vitellaria larvae. They are Ophiocoma pumila, Ophioderma brevispina, O. cinereum, O. longi- cauda, Ophiolepis cincta, O. elegans, Ophionereis annulata and O. squamulosa [1, 3, 5, 7, 8, 12, 14, 15, 19]. To date, however, detailed studies on development through vitellaria have been done only in 2 species; O. brevispina [1] and O. squamu- losa [14]. In addition, descriptions on the gastrula- tion, including formation of the archenteron, are very limited. Indeed, no regular and convenient method, such as KCI and 1-methyladenine which induces spawning in echinoids and asteroids, re- spectively, has been reported to induce spawning in ophiuroids. However, as the result of several chances of spawning the authors have gotten embryos of this species. The present paper de- scribes the larval development of the brittle star, Ophioplocus japonicus through metamorphosis.

MATERIALS AND METHODS

Specimen collection

Ophioplocus japonicus H. L. Clark have their habitat under stones on a sandy bottom. Speci- mens were collected from the seashores of Uozu,

296 M. KomaTsu AND T. SHOSAKU

Toyama Prefecture, of Tsukumo Bay, Ishikawa Prefecture, and of Sado Island, Niigata Prefecture, all of them facing the Sea of Japan, for nearly every summer between 1980 and 1990. This brittle star is gonochoric. In both sexes, the gonad was arranged in pairs in each interradius. In the breeding season, the body cavity is almost com- pletely occupied with gonads. Ovaries in this season are orange or brown in color and the testes are milky white.

Breeding season

In August 15, 1980, natural spawning occurred in a glass container with small amounts of seawater in which about 20 individuals were kept. They had been en route to a laboratory at Toyama Universi- ty after they had been collected at Uozu. Natural spawning also occurred in a tank with running seawater at Noto Marine Laboratory, Ishikawa Prefecture, on August 13, 1985, and at Sado Marine Laboratory, Niigata Prefecture, on August 9, 1990, a few days after collection. Specimens, which were held in a glass container containing as much seawater as needed to immerse the dorsal surface of their disk, spawned spontaneously at Noto Marine Laboratory, on August 25, 1983, August 21, 1984, August 6 and 9, 1986, and July 31, 1991. These ova were fertilized using a dilute suspension of spermatozoa from the dissected testes of homologous male specimens.

Culture of embryos and larvae

Temperature in the laboratory culture was approximately 25°C, which is close to the water temperature at the natural habitat in August. The embryos and larvae were taken from the culture dish at appropriate intervals and observed under dissecting and light microscopes. They were mea- sured live with an ocular micrometer. The skeletal system, from fresh squash preparations, was ex- amined under standard and Nomarski differential interference microscopes, and from samples pre- served by the following method. Embryos and larvae were fixed in 70% alcohol and then macer- ated in a 10% aqueous solution of potassium hydroxide.

For histological observation, the embryos and larvae were fixed with Bouin’s solution. The fixed

material was embedded in paraffin, serially sec- tioned at a thickness of 6 um, and stained with Delafield’s hematoxylin and eosin. Embryos and larvae were fixed for scanning electron microscopy in 2% OsO, in 50 mM Na-cacodylate buffer (pH 7.4); the osmolarity of the fixative was adjusted by the addition of sucrose (final conc. 0.6M). The fixed materials were dehydrated in an ethanol series and dried with a critical point dryer (Hitachi, HCP-2). They were observed with a scanning microscope (Hitachi, S-510) after being coated with gold-palladium (Hitachi, E101 Ion Sputter).

RESULTS

Annual reproductive cycle

In order to study the annual reproductive cycle, specimens were collected at a fixed locality of the coast of Himi along Toyama Bay, Toyama Prefec- ture. Sampling of the animals was made once a month during the period from March, 1979 through October, 1981. At each collection, 5 individuals of both sexes and those which were more than 10 mm in disk diameter were selected. Soon after collection, the disk was cut off from the arms and the fresh weight was determined. Gonads were then dissected out and weighed. The gonad index for the gross estimation of the annual reproductive cycle was calculated by the following formula:

Gonad Index=(Fresh Weight of Gonads/ Fresh Weight of Total Disk) x 100

The annual reproductive cycle was determined in the population of O. japonicus at Himi as shown in Figure 1. Although there were some differences throughout the years, the gonad indices of males and females were highest in June, 1981 and July, 1980, while they were lower in August of these 2 years. After September, the value for both sexes increased gradually. According to the transition of the gonad index, and the natural and inductive spawnings, the spawning of this species is pre- sumed to occur in August.

Embryos of cleavage stage

In the present study, the entire process of de-

Development of Ophioplocus japonicus 297

%% 0--0 OVARY e—e TESTIS

MA MJ JA S ON DJ FMAM J JA S O 1980: \I98l MONTH Fic. 1. Monthly variations of the gonad index in Ophioplocus japonicus, from March, 1980 through October, 1980. Each point shows an average of 5 individuals and standard error (vertical bar).

TaBLe 1. Chronology of development of Ophioplocus japonicus (25°C)

Time

(after fertilization) Stage 1 hr Ist cleavage 2 hrs 2nd cleavage 3 hrs 8-cell stage 4 hrs 16-cell stage 5 hrs 32-cell stage 6 hrs 64-cell stage 7.5 hrs Blastula with blastocoel 12 ‘hrs Hatching 16 hrs Gastrulation 20 hrs Gastrula, 350 um long and 250 um wide 30 hrs Larva with 5 visible hydrolobes 36 —shhrs Vitellaria bearing tube-foot buds in hydrolobe, ciliary bands appear 2 days Terminal tentacles and oral tube-feet present 2.5 days Vitellaria being rectangular in outline on the posterior potion 3 days Rudiments of adult skeleton appear as spicules 3.5 days Vitellaria with 4 distinct transverse ciliary bands on dorsal side, 530 ym long and 350 ~m wide 4 days Preoral lobe begins to be resorbed 5 days Oral plates are apparent 6 days Metamorphosis completed, 500 ~m diameter

14. days Juvenile, 600 ~m diameter

298

velopment from fertilization to complete meta- morphosis was observed in the laboratrory at 25°C. Naturally spawned eggs lacked the germinal vesi- cle and were spherical, with an average diameter of 300 wm. They are opaque, grayish brown and heavier than seawater. A transparent jelly coat with a thickness of about 25 ~m surrounded the egg surface. The head of the sperm was spherical, about 4 ~m in diameter, and the tail was about 50 ym in length.

A chronology of larval development for Ophio- plocus japonica is presented in Table1. The fertilized egg had a fertilization membrane with a perivitelline space approximately 10 um in height. Polar bodies, which direct the animal pole, were visible in the perivitelline space. The Ist cleavage occurred through the animal-vegetal axis 1 hr after fertilization and the 2nd followed at about 1 hr later, also through the egg axis but perpendicular to the first plane. Successive cleavages were equal, radial, and total, and occurrd at approximately 1 hr intervals. The resulting blastula was furnished

M. Komatsu AnD T. SHOSAKU

with a large blastocoel in its center at 7.5 hr after fertilization (Fig. 3-1). At this stage, the entire surface of the blastula was furnished uniformly with cilia about 15 ~m long. The embryo then began to rotate. Hatching from the fertilization membrane took place 12 hr after fertilization. The hatched larvae were almost spherical in shape, and they swam actively.

Gastrula

Gastrulation occurred by invagination at the vegetal pole about 16hr after fertilization. The early gastrula was shaped like a pyramid with corners, with a small blastopore at the broad vegetal end. The blastopore became larger as gastrulation proceeded. The 20 hr-old gastrulae were oval, being now 350 um long and 250 um wide (Figs. 2-1, 3-2). Their anterior end was round and posterior end was truncate with a blastopore at the center. At this stage, mesen- chyme cells were observed in the histological pre- parations. They were set free from the top of the

Fic.

Fic.

2. Development of Ophioplocus japonicus. Each drawing was made from a living specimen. 1. Twenty hr, gastrula. 2. Thirty hr, early vitellaria larva bearing 5 hydrolobes. 2A; dorsal side, 2B; ventral side. 3. Two and a half days, vitellaria larva with tube-feet and oral tube-feet in a ventral view. 4. Three days, spicules of the rudimental skeletal plate of the adult. 5. More developed spicules than those shown in Fig. 2-4. 6. Three and a half days, vitellaria larva, arrows show transverse ciliary bands. 6A; dorsal side, 6B; ventral side, 6C; spicules of the rudimental skeletal plate of the adult on the dorsal surface of the larva. Dotted lines represent 5 hydrolobes with 2 pairs of tube-feet each. 7. Five days, metamorphosing larva with the absorbed preoral lobe. 7A; dorsal side, 7B; ventral side. 8. Six days, metamorphosed juvenile with a pen-tamerous shape. 8A; aboral side, 8B; oral side. 9. Eight days, juvenile bearing 5 pairs of lateral arm plates in the radius. 9A; aboral side, 9B; oral side. 10. Three months, juvenile with a long arm, 10A; aboral side, 10B; oral side. Symbols: dep; dorsocentral plate, hl; hydrolobe, lap; lateral arm plate, m; mouth, mpp; madreporic plate, op; oral plate, otf; oral tube-foot, pl; preoral lobe, rp; radial plate, tf; tube-foot, tfp; tube-foot pore, to; tooth, tp; terminal plate, tt; terminal tentacle, vp; ventral plate. Magnification: ca. 65x in all drawings except for 4, 5 and 6C (ca. 200).

3. Development of Ophioplocus japonicus. 1 and 2 are micrographs of sections (6 ~m) of specimens and others are scanning electron micrographs. 1. Seven and a half hr, blastula with blastocoel (blc). 2. Thirty hr, gastrula showing outer ectodermal layer and inner endodermal layer, same stage as shown in Fig. 2-1. 3. Twenty-four hr, late gastrula. Note a rise (arrows) corresponding to the area of rudiments of a ciliary band. 4, 5 and 6. Thirty-six hr, early vitellaria larva with primary ciliary bands. 4A; dorsal side. Flaglet and arrow show ciliary bands on the preoral lobe and in the posterior portion, respectively. 4B; magnified view of the anterior left lateral part, 5; ventrolateral (left) side showing hydrolobes (hl). Note the ciliary band on the preoral lobe (flaglet), in the middle of the posterior portion (arrow), and below the hydrolobe (arrowhead). 6; ventral side. Note rudiments (©) of tube-foot on hydrolobe. Flaglet shows ciliary band on the preoral lobe. 7. Two days. 7A; ventral side of a vitellaria larva bearing apparent tube-foot (tf), oral tube-foot (otf) and terminal tentacle (tt) on the hydrolobe, and spiral ciliary band (flaglets) on the preoral lobe, 7B; Magnified view of spiral ciliary band (flaglets) on the preoral lobe. 8 and 9. Two and a half days. 8; dorsal side of vitellaria larva showing 3 parts (arrows) of the upper ciliary band on the posterior portion and the ciliary bands (arrowheads) at the posterior end, 9; magnified view of a posterior part of a specimen. Note ciliary bands (arrowheads) in the future interradius. Each of the hydrolobes in the future radius has a terminal tentacle (%), | pair of tube-feet (©) and a pair of oral tube-feet (@). Scale; 100 ym in 1-4A, 5-7A and 8, and 20 «~m in 4B, 7B and 9.

Development of Ophioplocus japonicus 299

M. Komatsu AND T. SHOSAKU Fic. 3.

300

Development of Ophioplocus japonicus 301

archenteron into the blastocoel. Around 25 hrs after fertilization, the area of the middle portion of the body, where the ciliary band will be formed, became apparent as a ridge (Fig. 3-3). The larvae had a distinct constriction which separated the anterior portion from the posterior portion by 30 hr after fertilization (Fig.2-2A, 2-2B). The anterior portion was seen as tapering and corres- ponded to the preoral lobe. It disappeared during the metamorphosis into the adult form. The posterior portion, which forms the juvenile body, was more or less pentaradiate in this stage. Projec- tions of the rudiments of the hydrolobes appeared in the future oral side of the larva. In the 1.5 day-old larva, each of the 5 hydrolobes bore one pair of tube-foot buds (Fig. 3-6). The hydrolobes encircled a concave portion, which becomes the oral area of the adult after metamorphosis. At that stage, a ciliary band surrounded the middle part of the preoral lobe (Fig. 3-4A, 3—4B, 3-5, 3-6). In addition to the transverse ciliary band on the preoral lobe, a band encircling the middle of the posterior portion became discontinuous at the ventral and dorsal sides of the body, and could be observed from both lateral sides (Fig. 3-5). On the ventral side, the lateral parts of the ciliary band extended across the hydrolobe (Fig. 3-5), which was situated in the posterior end of the body. There was a transverse ciliary band below the hydrolobes on the ventral side (Fig. 3-5), although it had not yet been formed on the dorsal side. Thus, the larva at this stage should be called a vitellaria, because a transverse ciliary band was now apparent.

Vitellaria

Two days after fertilization, an oral tube-foot appeared in the aboral side of each tube foot (Fig. 3-7A). A terminal tentacle was distinguishable on the peripheral portion of each hydrolobe. At this stage, the ventral parts of the ciliary band on the preoral lobe assumed a spiral form (Fig. 3-7B). The ciliary band below the hydrolobe was dis- placed to the posterior end of the body and extended to the dorsal side. Two and a half days after fertilization, the posterior portion of the vitellaria larva became clearly rectangular in out- line (Fig. 2-3). Because of the change of the

vitellaria larval form, the ciliary band of the post- erior end became situated in the future interradii of the ventral side (Fig. 3-9). On the dorsal side, each of 2 ciliary bands in the posterior portion were almost continuous (Fig. 3-8). The upper ciliary band consisted of 3 parts: 2 short larval parts and a long median one between them. Three day-old vitellaria larvae had spicules in the post- erior portion, as shown in Figure 2-4. These spicules are the rudiments of the adult skeletal plates, some of which later become tetraradiate or hexaradiate, measuring 20-40 4m in diameter (Fig. 2-5).

Three and a half days after fertilization, the vitellaria larvae became dorso-ventrally flattened (Fig. 2-6A, 2-6B). The vitellaria was now 530 ~m long and 350 ~m wide. By this stage, the juvenile mouth opened. Four transverse ciliary bands were distinct on the dorsal side (Fig. 4-1, 4—2A) and were completely continuous except for the 2nd anterior band, which consisted of small lateral parts (Fig. 4-2B). SEM observation showed that 2 (the Ist and 2nd) anterior bands, which were situated in the preoral lobe, made contact on the ventral side (Fig. 4-3). At this stage, the tube-feet were functional and the vitellariae could either swim freely or creep on the substrata. Figure 2—6C shows spicules viewed from the future aboral side of the juvenile specimen. However, there is no trace of the larval skeleton, in contrast to the findings in the vitellaria of Ophionereis annulata by Hendler [8].

Metamorphosis

Four days after fertilization, the anterior portion of the vitellaria, the preoral lobe, began to de- generate (Fig. 4-4A). By this stage, the cilia had disappeared on the oral side and only a trace of ridges of ciliary bands remained (Fig. 4-SA, 4- 5B). On the dorsal side, the cilia on the incom- plete ciliary bands decreased in number and be- came short (Fig. 4-4C). On the other hand, 3 transverse ciliary bands were retained (Fig. 4—4B). Absorption of the preoral lobe proceeded rapidly after the commencement of the metamorphosis. The metamorphic stage shown in Figure 4—6 is at approximately half a day after the commencement of metamorphosis. Five days after fertilization,

M. KoMaTsu AND T. SHOSAKU

302

Development of Ophioplocus japonicus 303

only a trace of the preoral lobe remained (Figs. 2- 7A, 2-7B, 4-7). By this stage, the skeletal system of the adult started to grow and the larval body became more rigid except for the tube-feet. A pair of the oral plate was now apparent in each radius. Six days after fertilization, the anterior portion of the larva disappeared and thus metamorphosis was completed (Fig. 2-8A, 2—8B). Metamorphosed juveniles, 500 ~m in diameter, bore 5 rudimental arms, each with a terminal plate at its tip. They were benthonic and moved on the substratum by means of their tube-feet. On the aboral side, rudiments of the skeletal plate were apparent. The aboral side of the disk became convex 1 day later. One dorsocentral plate and 5 radial plates around the former became firm at this stage.

Juvenile

Eight days after fertilization (2 days after the completion of metamorphosis), 1 pair of lateral arm plates each of which bore 1 spine, were formed in both sides of the arm (Fig. 2-9A). The madreporic plate was now recognizable on 1 inter- radius. Skeletal plates of the oral side of the disk were indistinct except for the oral and terminal ones (Fig. 2—9B). One pair of tube-foot pores were seen to exist in each radius. Ten days after fertilization, the 1st ventral arm plates appeared. Figure 4—8 shows the aboral view of a juvenile at 2 weeks after fertilization. The skeletal plates of the aboral side had many pore-like sieves. The juve- nile seen here is 600 4m in diameter. One month after fertilization, juveniles grew to 700 um in diameter. Although a number of juveniles were kept alive in the laboratory for several months, no further differentiation could be observed. Three

month-old juveniles measured 500 ym in disk dia- meter and 250 “m in arm length (Figs. 2-10A, 2— 10B, 4-9). They bore 2 arm segments from the development of the 2nd lateral arm plates. Each arm was furnished with 3 pairs of tube-feet.

DISCUSSION

Detailed studies on the development through metamorphosis are very limited in ophiuroids, although larvae of ophiuroid have been recognized in about 80 species [5].

It was reported that gastrulation takes place by invagination at the vegetal pole [6, 10, 17, 21]. Little is known about the formation of the archen- teron in the species with the vitellaria larva. It was reported that the archenteron is formed by split- ting of the central core in Ophioderma brevispina [1]. On the other hand, judging from the figure by Mortensen ([14], Plate XX XI, Fig. 5) the archen- teron seems to be formed by invagination in a 26 hr-old larva of Ophionereis squamulosa. In the present species, gastrulation is apparently achieved by an invagination. Thus, formation of the archenteron seems to be diverse in different species in ophiuroids with vitellaria larvae. However, further studies are needed to promote a deeper understanding of the archenteron forma- tion.

Ophioplocus japonicus is the 9th ophiuroid spe- cies that has been shown to have vitellaria larvae. The larva of O. brevispina was called worm-like larva by Miller [16]. Brooks and Grave [1] described in some detail the larva of O. brevispina. Later the term, vitellaria, was proposed by Fell [2] for a distinctive larval type of echinoderm which

Fic. 4. Development of Ophioplocus japonicus. 2-8 views are scanning electron micrographs. 1 and 9 are light (regular) micrographs. 1, 2 and 3. Three and a half days, same stage as that shown in Figs. 2-6 and 4-2A. 1; vitellaria larva from the dorsal side with 4 transverse ciliary bands (arrows), 2A; dorsal side, 2B; magnified view of the anterior part, 3; ventral side. 4 and 5. Four days. 4A; dorsal side, vitellaria larva showing the disappearance of cilia on the ciliary band (short arrow). The long arrow indicates a ciliary band on the preoral lobe. 4B; magnified view of the ciliary band (the long arrow shown in Fig. 4—4A) on the preoral lobe, 4C; magnified view of the ciliary band (the short arrow shown in Fig. 4—4A) showing the disappearance of cilia, 5A; ventral side of the vitellaria, 5B; magnified view of the anterior part. Arrow indicates the junction of two ciliary bands on the preoral lobe. 6. Four and a half days, metamorphosing vitellaria with reduced preoral lobe (arrow), view from dorsal side. 7. Five days, metamorphosing vitellaria showing absorbed preoral lobe (arrow), view from ventral side, same stage as that shown in Fig. 2-7. 8. Two weeks, juvenile with a dorsal central plate and 5 radial plates on the dorsal side. 9. Three months, juvenile with arm segments, view from aboral side, same stage as shown in Fig. 2-10. Scale: 100 um in 1-4A, 5A, and 7-9, 10 um in 4B and 5B, and 1 zm in 4C.

304 M. Komatsu AND T. SHOSAKU

corresponds to the larva of O. brevispina. Occur- rence of ciliary rings is one of the characteristics of the vitellaria: the existence of 4 ciliary bands has been reported in the vitellariae of Ophioderma longicauda [3] and O. brevispina [1]. In fact, Hendler [8] noted that ophiuroid vitellariae gener- ally have 4 transverse ciliary bands. However, the number of the ciliary bands varies in different species. There are 3 ciliary bands in Ophionereis annulata because of the lack of the 4th transverse band [8]. Mortensen [15] described that vitellariae of Ophiolepis cincta are furnished with a ciliated tuft in addition to 4 transverse ciliary bands. Judging from his figure (Fig. 26, p. 48), the actual number of ciliary bands seems to be 3; the 3rd and the 4th bands are continuous and should be counted as 1. The number of ciliary bands was not described in Ophioderma cinereum [7], Ophiolepis elegans [19] or O. squamulosa [14]. However, it can be determined in the latter 2 species from available figures (Fig. lc, p. 9; Figs. 2 and 3, Pl. XXXI). Each of these vitellariae is furnished with 3 ciliary bands and 1 ciliated tuft. In the vitellaria of the present species, no ciliated tuft was observed at the anterior portion of the body by both LM and SEM observations. Ciliation of O. Japonica is, therefore, different from that of O. cincta, O. squamulosa and O. elegans. Thus, ciliation of the ophiuroid vitellaria differs in diffe- rent species. The vitellaria larva of O. japonica seems to have 4 transverse ciliary bands when observed from the dorsal view. However, the actual number of ciliary bands 1s 3, since 2 anterior ciliary bands on the preoral lobe are continuous on the ventral side. Therefore, detailed observation of ciliation of the vitellaria should be one of the points to be investigated in the future.

As noted before, the classification of the de- velopmental pattern in ophiuroids seems to be different judging from the literature. Hendler [5] classified the development of ophiuroids into 3 types; planktonic, direct, and abbreviated. According to his view, development that includes vitellariae belongs to the abbreviated develop- ment. The abbreviated development category includes Gorgonocephalus caryi [18], which de- velops through only a pear-shaped larva without ciliary bands. Thus, the abbreviated type compris-

es heterogeneous larval forms. Hendler’s proposal was supported by Mladenov [11]. Strathmann and Rumnill [20] later called the abbreviated develop- ment a pelagic or demersal development with lecithotrophic larvae.

According to the original proposals by Fell [2], the vitellaria larva is a larval form which bears transverse ciliary bands. Therefore, in terms of abbreviated development, it is reasonable to dis- tinguish development with vitellariae from that with other types of larvae.

The relation between developmental type and egg size in ophiuroids was shown by Hendler [5] and then revised by Mladenov [11]. According to the latter, the ranges of the egg’s diameter in the direct and planktotrophic developments are 100- 1000 ~m and 70-200 um, respectively. Thus, there is a considerable spread of egg size in these 2 developmental types. On the other hand, the diameter of an egg of the species undergoing the abbreviated development type is greater than 130 ym and less than 350 «zm.

It is likely that ophiuroid eggs, being from 200 to 350 um in diameter, develop through the vitellaria larva. As a matter of fact, egg of O. japonicus is approximately 300 um in diameter. The egg of Ophiocoma pumila is very small (73 “m) among the ophiuroids with vitellaria larvae. However, this species is exceptional since O. pumila passes through both the vitellaria and the planktotrophic pluteus stages.

In the present species, larva completes meta- morphosis 6 days after fertilization at about 25°C. Stancyk [19] reported that in O. annulata, meta- morphosis is completed 3 days after fertilization at 24°C. The majority of species which have vitellaria larvae were reported to complete metamorphosis within 4 or 5 days, although the temperatures were not indicated [4, 13, 15]. On the other hand, in Amphipholis kochii which has planktotrophic de- velopment with eight-armed ophiopluteus, meta- morphosis is incipient 12 days after fertilization at 23°C [21]. This term seems to be the shortest among planktotrophic developers according to the table by Mladenov ([11], p. 60). Thus, it may be concluded that developments through the vitellaria larva proceed faster than those through the ophio- pluteus larva.

Development of Ophioplocus japonicus 305

In O. japonicus, no larval skeleton is present in the vitellariae. No trace of any larval skeleton was observed in either the vitellaria larvae of O. squamulosa or O. brevispina [1, 14]. On the other hand, occurrence of a larval skeleton was reported in O. annulata [8]. Furthermore, an irregular calcareous structure was found in the vitellaria of O. cincta [15]. It is, therefore, interesting that there are 2 types of vitellariae, one with a pluteus- like skeleton and the other with no skeleton at all.

O. japonicus is one of the most common ophiuroids in Japan, with its habitat being under the stones of shallow water. In the present study, the larva of this species is demonstrated as being vitellaria, not ophiopluteus, which is thought to be a typical planktotrophic larva of ophiuroids [5, 9, 21]. Twenty-four species of ophiuroids have been known to pass through ophiopluteus larvae. They belong to 5 families of 2 suborders, Gnathophiuri- na and Chilophiurina. On the other hand, vitellar- iae have been found in 9 species including the present species, and all belong to 4 families of 1 suborder, Chilophiurina. These families are Ophiodermatidae, Ophiocomidae, Ophiuridae and Ophionereidae. Thus, the occurrence of vitel- laria is very limited and this may be phylogeneti- cally significant.

ACKNOWLEDGMENTS

The authors thank Dr. Chitaru Oguro, President of Toyama University, for his valuable advise on this study and critical reading of the manuscript. The present study was supported in part by a Grant-in-Aid from the Ministry of Education of Japan to MK (No. 57740392).

REFERENCES

1 Brooks WK Grave C (1899) Ophiura brevispina. Mem natn Acad Sci 8: 79-100

2 Fell HB (1945) A revision of the current theory of echinoderm embryology. Trans Roy Soc New Zea- land 75: 73-101

3 Fenaux L (1969) Le développmen larvaire chez Ophioderma longicauda (Retzius). Cah Biol Mar 15: 59-62

4 Grave C (1916) Ophiura brevispina Il. An embryological con tribution and a study of the effect of yolk substance upon development and develop- mental processes. J Morph 27: 413-453

5

10

11

14

15

16

17

18

20

Hendler G (1975) Adaptational significance of the patterns of ophiuroid development. Amer Zool 15: 691-715

Hendler G (1977) Development of Amphioplus abditus (Verrill) (Echinodermata: Ophiuroidea) I. Larval biology. Biol Bull 152: 51-63

Hendler G (1979) Reproductive periodicity of ophiuroids (Echinodermata: Ophiuroidea) on the Atlantic and Pacific coasts of Panama In “Reproduc- tive Ecology of Marine Invertebrates” Ed by SE Stancyk, Univ of South Carolina Press, USA, pp 145-156

Hendler G (1982) An echinoderm vitellaria with a bilateral larval skeleton: evidence for the evolution of ophiuroid vitellariae from ophioplutei. Biol Bull 163: 431-437

Hyman L (1955) The Invertebrates VolIV Echi- nodermata. McGraw-Hill, New York, pp 763 MacBride EW (1914) Invertebrate In “Text-Book of Embryology” Ed by W Heape, Macmillan, Lon- don, Vol IV, pp 692

Mladenov PV (1979) Unusual lecithotrophic de- velopment of the Caribbean brittle star Ophiothrix oerstedi. Mar Biol 55:55-62

Miadenov PV (1985) Development and meta- morphosis of the brittle star Ophiocoma pumila: evolutionary and ecological implication. Biol Bull 168: 285-295

Mortensen Th (1898) Die Echinodermenlarven der Plankton-Expedition nebst neiner systematischen Revision der bisher bekannten Echinodermenlar- ven. Ergebnis Plankton-Exped Humboldt-Stift 2J: 1-118

Mortensen Th (1921) Studies of the Development and Larval Forms of Echinoderms. GEC Gad, Copenhagen, pp 261

Mortensen Th (1938) Contributions to the study of the development and larval forms of Echinoderms IV. Kgl Dan Vidensk Selsk Skr, Naturvid Math Afd, Ser 9, 7(3): 1-45

Muller J (1850) Ueber die Larven und die Meta- morphose der Echinodermen. Dritte Abhandlung Abhand! Konigl Preuss Akad Wiss Berlin, 1849: 35— 12

Oguro C, Shosaku T, Komatsu M (1982) Develop- ment of the brittle-star, Amphiolis japonica Matsu- moto In “Echinoderms” Ed by JM Lawrence, AA Balkema, Rotterdam, pp 491-496

Patent DH (1970) The early embryology of the basket star Gorgonocephalus caryi (Echinodermata, Ophiuroidea). Mar Biol 6: 262-267

Stancyk SE (1973) Development of Ophiolepis elegans (Echinodermata: Ophiuroidea) and its im- plications in the estuarine environment. Mar Biol 21: 7-12

Strathmann MF, Rumrill SS (1987) Phylum Echi-

306 M. Komatsu AND T. SHOSAKU

nodermata, Class Ophiuroidea In “Reproduction 21 Yamashita M (1985) Embryonic development of the and Development of Marine Invertebrates of the brittle-star Amphipholis kochii in laboratory cul- North Pacific Coast” Ed by MF Strathmann, The ture. Biol Bull 169: 131-142

Univ Was Press, USA, pp 556-573

ZOOLOGICAL SCIENCE 10: 307-312 (1993)

© 1993 Zoological Society of Japan

Maitotoxin Induces Acrosome Reaction and Histone Degradation of Starfish Asterina pectinifera Sperm

TosikKAZU AMANO, YISHIHITO OKITA’, TAKESHI YASUMOTO27

AND MotTonori Hosur

Department of Life Science, Faculty of Bioscience and Biotechnology, Tokyo Institute of Technology, Nagatsuda, Midori-ku, Yokohama 227, Japan *Department of Biological Chemistry, Faculty of Agriculture, Tohoku University, Tsutsumidori-Amemiyamachi,

Aoba-ku, Sendai 981, Japan

ABSTRACT—Maitotoxin (MTX) is a marine toxin which presumably activates Ca**-channels and stimulates phosphoinositide breakdown in various mammalian cells. We report here that MTX induces the acrosome reaction of starfish spermatozoa by a cooperative action with a diffusible fraction of the egg jelly. MTX alone does induce the degradation of sperm histones. Both reactions induced by MTX depend on extracellular Ca**. Verapamil, a Ca?‘ -channel antagonist, inhibits MT'X-induced acrosome reaction but not affects MTX-induced histone degradation. MTX bypasses the blockage by ARIS of the jelly-induced acrosome reaction and histone degradation.

INTRODUCTION

Changes in ion permeability of plasma mem- brane regulate sperm reactions to extracellular signals [6, 11, 27-29]. Egg jelly, with which spermatozoa interact first in the echinoderms, trig- gers Na‘- and Ca*'-influx, and K*- and H*- efflux in spermatozoa, and eventually induces the acrosome reaction [19, 27, 28]. We have reported that the egg jelly also induces the degradation of sperm histones in the starfish, Asterina pectinifera [1]. Both reactions are induced by a cooperative action of homologous ARIS, a fucose sulfate rich glycoprotein having an extremely large molecular weight, and a diffusible fraction of the egg jelly (M8) containing Co-ARIS and sperm-activating peptides (SAP) [2, 14, 15, 17, 20, 22, 24, 26]. The acrosome reaction and the histone degradation induced by the egg jelly depend upon extracellular Ca** and are inhibited by Ca?*-channel antagon-

Accepted October 10, 1992 Received September 21, 1992

' Present address: Suntory Pharmatech Center, Chiyo- damachi, Gunma 370-05, Japan

3 To whom all correspondence should be addressed.

ists, suggesting that Ca**-influx via Ca**-channels plays important roles in these reactions [1]. However, A23187, a Ca?* ionophore which in- duces the acrosome reaction of sea urchin sperm is not effective to induce the acrosome reaction and the histone degradation of Asterina pectinifera sperm [1]. It induces these reactions only in combinations with monensin or M8 (1; unpub- lished results). These results suggest that an increase in the intracellular Ca** concentration ([Ca**]i) is essential but not sufficient to induce the acrosome reaction and histone degradation. Although these reactions of Asterina pectinifera sperm are induced by a cooperative action of ARIS and Co-ARIS [3], several lines of evidence indicate that two reactions are regulated by distinct pathways at least partly [1, 2].

Maitotoxin (MTX) has been isolated from a poisonous marine dinoflagellate, Gambierdiscus toxicus [23]. MTX stimulates the uptake of Ca’* in various cells [4, 10, 16, 18, 30, 32] and elicits phosphoinositide breakdown in several cell lines [5, 12, 13]. Effects of MTX absolutely depend upon extracellular Ca*t. In most cells, the sti- mulation of Ca**-uptake by MTX is blocked by Ca?*-channel antagonists [16, 32], and therefore it

308 T. Amano, Y. Oxira et al.

is proposed that MTX is a direct activator of Ca**-channels [32]. In contrast to this, the sti- mulation of phosphoinositide breakdown by MTX is not affected by a variety of organic and inorganic Ca**-channel blockers [5, 12, 13].

In the starfish, Asterias amurensis, it is reported that, in alkaline seawater, MTX triggers only a part of acrosome reaction: it induces acrosomal exocytosis but not the formation of perfect acro- somal process [25].

In this paper, we show that MTX is effective to induce normal acrosome reaction in Asterina pecti- nifera and the effect of MTX on inducing the acrosome reaction is greatly enhanced by M8 and is susceptible to verapamil. MTX also initiates the degradation of sperm histones in a verapamil- insensitive manner.

MATERIALS AND METHODS

Artificial seawater, jelly components and drugs

Normal artificial seawater (ASW) consisted of 450 mM NaCl, 10 mM KCl, 10 mM CaCl, 30 mM MgCl, 20 mM MgSO,, 15 mM N-2-hydroxyethyl- piperazine-N -propanesulfonic acid-NaOH, pH 8.2. Ca’*-free seawater (CFSW) was prepared as above except for omitting CaCh.

Egg jelly, ARIS and M8 were prepared as previously described [1, 3]. Verapamil was dis- solved in DMSO to make a 200 mM stock solution. For the control, an equal concentration of DMSO MTX was dissolved in distilled water at a concentration of 300 ng/ml (87.6 ~M), and diluted in ASW or CFSW.

(1%) was run.

Assays of the acrosome reaction and the histone degradation

Spermatozoa of the starfish, Asterina pectinifera were collected “dry” and kept on ice until use. Assays of the acrosome reaction and the histone degradation were performed as previously de- scribed [1, 3]. Spermatozoa with an acrosomal process were scored as “reacted”. Each experi- ment was repeated at least two different batches of sperm.

RESULTS

Effects of MTX on starfish spermatozoa: It has been reported that MTX (0.06—-0.73 ~M) does not induce the acrosome reaction of Asterias amurensis spermatozoa in normal seawater, but that it in- duces only the acrosomal exocytosis in alkaline seawater (pH 9.5) [25]. We examined first the effects of MTX on the acrosome reaction and the histone degradation in Asterina pectinifera. MTX (0.6 ~M) induced the acrosome reaction consider- ably, although at the concentration of 0.3 uM its effect was not appreciable (Table 1). Induction of the acrosome reaction by MTX was greatly en- hanced by M8 but not by ARIS (Table 1). MTX- induced acrosome reaction was quite normal in the morphology (Data not shown). It required exter- nal Ca?* and was completely blocked by 200 ~M verapamil (Table 1).

TaBLeE 1. Induction of the acrosome reaction by MTX Treatments Ae Conehe Jelly 100 ARIS+M8 82 ARIS 9 M8 0 MTX (0.6 uM) 78 + verapamil** 3 MTX (0.3 uM) 12 + ARIS 19 +M8 56 +M8 in CFSW +M8-+ verapamil** 4

Sperm were treated with egg jelly (100 ug fucose/ ml), ARIS (50 «g/ml), M8 (62.5 ug/ml) or com- binations of them for 3 min and fixed for microscopic observation. “The percentage of jelly-induced acrosome reaction was referred as 100%. Values are means for two experiments. *“*Sperm were pretreated with verapamil (200 ~M) for 3 min and then MTX plus M8, or MTX, was added to the sperm suspension.

In contrast to the acrosome reaction, the histone degradation was much sensitive to MTX: 0.3 uM of MTX induced the histone degradation as well as

Maitotoxin Induces Sperm Reactions 309

TaBLE2. Induction of the histone degradation by MTX Treatments Histone A icy ait le Jelly 52 ARIS 9 M8 2 MTX (0.6 “M) 64 +M8 73 +verapamil* 61 MTX (0.3 “M) 42 +ARIS 61 +M8 56

Sperm were treated with egg jelly (SO ug fucose/ ml), ARIS (50 “g/ml) or M8 (25 ~g/ml) and incu- bated for 60 min. Reactions were stopped by the addition of an equal volume of sample buffer for SDS-polyacrylamide gel electrophoresis.

* Sperm were pretreated with verapamil (200 ~M) for 3 min and then MTX was added to the sperm suspension.

the egg jelly did (Table 2). Neither ARIS nor M8 significantly enhanced the effect of MTX on the histone degradation. MTxX-induced histone de- gradation also required external Ca**, but it was not blocked by 200 “M verapamil (Table 2).

Sequential treatments of MTX and egg jelly compo- nents: It is known that spermatozoa treated with either ARIS or M8 become unresponsive to the egg jelly [2, 14, 21]. We examined whether MTX also showed a similar “pretreatment effect”. As shown in table 3, when sperm were pretreated for 3 min with an insufficient concentration of MTX (0.3 «M) to induce the acrosome reaction, they did not undergo the acrosome reaction by the addition of M8. In response to the egg jelly they underwent the acrosome reaction, but much less than intact spermatozoa did (Table 3). ARIS-pretreated sper- matozoa, which were unresponsive to the egg jelly, underwent the acrosome reaction in response to a mixture of MTX and M8 (Table 3).

Similarly, ARIS-pretreated spermatozoa did not undergo the histone degradation in response to the egg jelly [2] but they did so in response to MTX (Table 4). Previous study has shown that concana- valin A (Con A) specifically inhibits the jelly-

TABLE 3. Effects of sequential treatments of MTX and jelly components on the acrosome reaction

Treatments Acrosome Reaction Ist 2nd (% Control)* a Jelly 100 — MTX+M8 56 ARIS Jelly 23 ARIS MTX+M8 70 MTX Jelly 38 MTX M8 6

Sperm were preincubated with ARIS (50 g/ml) or MTX (0.3 “M) for 3 min and then egg jelly (100 ng fucose/ml), a mixture of MTX (0.3 ~M) and M8 (62.5 ug/ml), or M8 (62.5 ~g/ml) was added to the sperm suspension. The mixtures were incubated for another 3 min and fixed.

* The percentage of jelly-induced acrosome reac- tion was referred as 100%. Values are means for two experiments.

TaBLE4. Effects of sequential treatments of MTX and jelly components on the histone degradation

Treatments Histone H1 Degradation 1st 2nd (%) — Jelly 52 ARIS Jelly 11 ARIS MTX 70 Con A Jelly 18 Con A MTX 83

Sperm were treated with ARIS (50 ug/ml) or Con A (0.2 mg/ml) for 3 min, and then egg jelly (50 ug fucose/ml) or MTX (0.3 4M) was added to the sperm suspension. The mixtures were incubated for another 60 min and incubations were stopped by addition of sample buffer for SDS-polyacrylamide gel electrophoresis.

induced histone degradation [1]. Thus, we ex- amined the effects of Con A on the MTX-induced histone degradation. As shown in table 4, Con A (0.2 mg/ml) blocked the jelly-induced histone de- gradation but did not affect the MTX-induced one.

DISCUSSION

External Ca’* is essential for inducing the acro- some reaction and histone degradation by the egg jelly in starfish spermatozoa [1, 21]. Pharmacolo-

310 T. Amano, Y.

gical studies suggest the involvement of voltage- dependent Ca**-channels in these reactions [1]. We have reported here that MTX (0.6 ~M) in- duces morphologically normal acrosome reaction in Asterina pectinifera. This concentration is much higher than effective doses for mammalian cells, but it is comparable to that for the acrosomal exocytosis in Asterias amurensis [25]. Echinoderm sperm are generally much less sensitive to various drugs than mammalian cells. The effect of MTX on the acrosome reaction was greatly enhanced by simultaneous addition of M8 but not by ARIS. Similar to the stimulation of Ca**-uptake by MTX in various cells [16, 32], the acrosome reaction induced by MTX with or without M8 required extracellular Ca”* and was completely blocked by verapamil. This suggests that MTX induces the acrosome reaction by activating Ca** -channels. From the present data, we do not know exact roles of M8 in stimulation of MTX-induced acro- some reaction. It seems reasonable to assume that MB facilitates the opening of Ca**-channels and/ or other intracellular changes that are essential for complete acrosome reaction. The acrosome reac- tion consists of the acrosomal exocytosis and pro- cess formation [7, 8, 33]. Like many other ex- ocytotic events, acrosomal exocytosis probably re- quires an increase in [Ca’*]i. The process forma- tion is thought to be regulated by an increase in intracellular pH (pHi) [31, 34]. Actually, in sea urchin sperm, both Ca*t-uptake and a pHi in- crease can not be dissociated from the acrosome reaction. It is reported that MTX induces the acrosomal exocytosis significantly, but the process formation slightly, in mussel sperm at normal pH (8.0) and in Asterias amurensis sperm at pH 9.5 [25]. A Ca’*-ionophore A23187 did not initiate the acrosome reaction in Asterina pectinifera [1]. Thus we think that an increase in [Ca**]i is essential but not sufficient to induce the acrosome reaction of starfish spermatozoa. It is known in Asterias amurensis that ARIS and Co-ARIS in- duce the acrosome reaction without any detectable pHi increase, and SAP stimulates the reaction by increasing the pHi [14, 20, 22]. Stimulation by M8 of MTX-induced acrosome reaction in Asterina pectinifera suggests that M8 induces some other intracellular changes also required for the acro-

Oxira et al.

some reaction, such as a pHi increase.

Pretreatment of sperm with MTX remarkably decreased the egg jelly-induced acrosome reaction suggesting that MTX induces irreversible changes in sperm, which may participate in the jelly- induced acrosome reaction. MTX plus M8 bypas- sed the blockage of the jelly-induced acrosome reaction in ARIS-pretreated spermatozoa suggest- ing that ARIS-pretreatment did not irreversibly inactivate MTX-sensitive Ca?*-channels. Two mutually non-exclusive working hypotheses are being considered: (I) ARIS directly makes Ca?" - channels unresponsive to the jelly, but MTX can activate ARIS-modified Ca**-channels. (II) ARIS acts on a site upstream to the MTX-site(s), which indirectly regulates Ca?*-channels.

MTX was also effective to induce the histone degradation. This effect of MTX was not en- hanced by M8 nor affected by verapamil. These results suggest that the action site of MTX for the histone degradation is different from that for the acrosome reaction. This is consistent with the hypothesis that two reactions are controlled by distinct pathways at least partly [1, 2]. It has been proposed that MTX is a potent activator of phos- phoinositide breakdown and that Ca**-channel blockers hardly inhibit this effect [5, 12, 13]. It has been also reported that egg jelly promotes the formation of inositol 1,4,5-triphosphate in sea urchin spermatozoa [9]. Taking these into account, there is a possibility that MTX-induced phosphoinositide-breakdown in sperm cells trig- gers the histone degradation. Therefore, impor- tant questions to be answered are whether phos- phoinositide breakdown is induced by the egg jelly also in starfish spermatozoa and, if so, whether it is involved in the induction of histone degradation.

Preincubation of sperm with Con A inhibits jelly-induced histone degradation [1] but it did not affect the MTX-induced one. MTX bypassed the blockage of the jelly-induced histone degradation in ARIS-pretreated sperm. These results suggest that the MTX-sensitive and verapamil-insensitive step in the reaction pathway leading to the histone degradation is later than the action sites of ARIS and Con A, or that MTX induces the histone degradation by a quite different way from the egg jelly does.

Maitotoxin Induces Sperm Reactions

It is to be answered how and at which step, the

two sperm reactions are linked together.

ACKNOWLEDGMENTS

We thank Dr. Ikegami of Hiroshima University for generous supply of the animals. This work was sup- ported by the Grant-in-Aid for Scientific Research on Priority Areas No. 03207103 from the Ministry of Educa- tion, Science and Culture Japan.

10

11

12

REFERENCES

Amano T, Okita Y, Hoshi M (1992) Treatment of starfish sperm with egg jelly induces the degradation of histones. Develop Growth and Differ 34: 99-106 Amano T, Okita Y, Matsui T, Hoshi M (1992) Pretreatment effects of jelly components on the sperm acrosome reaction and histone degradation in the starfish, Asterina pectinifera. Biophys Biochem Res Commun 187: 268-273

Amano T, Okita Y, Okinaga T, Matsui T, Nishiyama I, Hoshi M (1992) Egg jelly components responsible for histone degradation and acrosome reaction in the starfish, Asterina pectinifera. Biophys Biochem Res Commun 187: 274-278

Anderson JM, Yasumoto T, Cronin MJ (1987) Intracellular free calcium in rat anterior pituitary cells monitored by fura-2. Life Sci 41: 519-526 Berta P, Sladeczek F, Derancourt J, Durand M, Travo P, Haiech J (1986) Maitotoxin stimulates the formation of inositol phosphates in rat aortic myocytes. FEBS Lett 197: 349-352

Collins F, Epel D (1977) The role of calcium ions in the acrosome reaction of sea urchin sperm. Exp Cell Res 106: 211-222

Dan J C (1952) Studies on the acrosome. I. Reac- tion to egg-water and other stimuli. Biol Bull 103: 54-66

Dan JC (1954) Studies on the acrosome. III. Effect of calcium deficiency. Biol Bull 107: 335-349 Domino SE, Garbers DL (1988) The fucose-sulfate glycoconjugate that induces an acrosome reaction in spermatozoa stimulates inositol 1,4,5-trisphosphate accumulation. J Biol Chem 263: 690-695 Freedman SB, Miller RJ, Miller DM, Tindall DR (1984) Interactions of maitotoxin with voltage- sensitive calcium channels in cultured neuronal cells. Proc Natl Accad Sci USA 81: 4582-4585 Gonzalez-Martinez M, Darszon A (1987) A fast transient hyperpolarization occurs during the sea urchin sperm acrosome reaction induced by egg jelly. FEBS Lett 218: 247-250

Gusovsky F, Yasumoto T, Daly JW (1987) Maitoto- xin stimulates phosphoinositide breakdown in

14

15

16

17

18

20

21

22

23

24

25

311

neuroblastoma hybrid NCB-20 cells. Neurobiol 7: 317-322

Gusovsky F, Yasumoto T, Daly JW (1989) Maitoto- xin, a potent, general activator of phosphoiositide breakdown. FEBS Lett 243: 307-312

Hoshi M, Matsui T, Nishiyama I, Amano T, Okita Y (1988) Physiological inducers of the acrosome reaction. Cell Differ and Develop 25(suppl.): 19-24 Ikadai H, Hoshi M (1981) Biochemical studies on the acrosome reaction of starfish Asterias amurensis. I. Factors participating in the acrosome reaction. Develop Growth and Differ 23: 73-80

Ikadai H, Hoshi M (1981) Biochemical studies on the acrosome reaction of starfish Asterias amurensis. II. Purification and characterization of acrosome reaction-inducing substance. Develop Growth and Differ 23: 81-88

Kobayashi M, Kondo S, Yasumoto T, Ohizumi Y (1986) Cardiotoxic effects of maitotoxin, a principal toxin of seafood poisoning, on guinea pig and rat cardiac muscle. J Pharmacol Exp Ther 238: 1077- 1083

Lebrun P, Hermann M, Yasumoto T, Herchisely A (1987) Effects of maitotoxin on ionic and secretory events in rat pancreatic islets. Biophys Biochem Res Commun 144: 172-177

Lee HC, Johnson C, Epel D (1983) Changes in internal pH associated with initiation of motility and acrosome reaction of sea urchin sperm. Develop Biol 95: 31-45

Matsui T, Nishiyama I, Hino A, Hoshi M (1986) Induction of the acrosome reaction in starfish. De- velop Growth and Differ 28: 339-348

Matsui T, Nishiyama I, Hino A, Hoshi M (1986) Acrosome reaction-inducing substance purified from the egg jelly inhibits the jelly-induced acrosome reaction in starfish: An apparent contradiction. De- velop Growth and Differ 28: 349-357

Matsui T, Nishiyama I, Hino A, Hoshi M (1986) Intracellular pH changes of starfish sperm upon the acrosome reaction. Develop Growth and Differ 28: 359-368

Murata M, Iwashita T, Yokoyama A, Sasaki M, Yasumoto T (1992) Partial structures of maitotoxin, the most potent marine toxin from dinoflagellate Gambierdiscus toxicus. J Am Chem Soc 114: 6594- 6596

Nishiyama I, Matsui T, Hoshi M (1987) Purification of Co-ARIS, a cofactor for acrosome reaction- inducing substance from the egg jelly of starfish. Develop Growth and Differ 29: 161-169 Nishiyama I, Matsui T, Yasumoto T, Oshio S, Hoshi M (1986) Maitotoxin, a presumed calcium channel activator, induces the acrosome reaction in mussel spermatozoa. Develop Growth and Differ 28: 443- 448

Cell Mol

26

Di

28

29

30

312

Okinaga T, Ohashi Y, Hoshi M (1992) A novel saccharide structure, Xyl-3Gall-(SO3 )3,4Fuc-, is present in acrosome reaction-inducing substance (ARIS) of the starfish, Asterias amurensis. Biophys Biochem Res Commun 186: 405-410

Schackmann RW, Christen R, Shapiro BM (1981) Membrane potential depolarization and increased intracellular pH accompanying the acrosome reac- tion of sea urchin sperm. Proc Natl Accad Sci USA 78: 6066-6076

Schackmann RM, Eddy EM, Shapiro BM (1978) The acrosome reaction of Strongylocentrotus pur- puratus sperm: Ion movements and requirements. Develop Biol 65: 483-495

Schackmann RW, Shapiro BM (1981) A partial sequence of ionic changes associated with the acro- some reaction of Strongylocentrotus purpuratus. De- velop Biol 81: 145-154

Schettini G, Koike K, Login IS, Judd AM, Cronin

31

32

33

34

T. Amano, Y. Okita et al.

MJ, Yasumoto T, MacLeod RM (1984) Maitotoxin stimulates hormonal release and calcium flux in rat anterior pituitary cells in vitro. Am J Physiol 247: E520-E525

Schroeder TE, Christen R (1982) Polymerization of actin without acrosomal exocytosis in starfish sperm. Visualization with NBD-phallacidin. Exp Cell Res 140: 363-371

Takahashi M, Ohizumi Y, Yasumoto T (1982) Maitotoxin, a Ca** channel activator candidate. J Biol Chem 257: 7287-7289

Tilney LG (1985) The acrosome reaction. In “Biolo- gy of Fertilization Vol 2” Ed by Metz CB, Monroy A Academic Press, Orlando, pp 157-213

Tilney LG, Kiehart DP, Sardet C, Tilney M (1978) Polimerization of actin. IV. The role of Ca** and H* in the assembly of actin and in membrane fusion in the acrosome reaction of echinoderm sperm. J Cell Biol 77: 536-550

ZOOLOGICAL SCIENCE 10: 313-319 (1993)

© 1993 Zoological Society of Japan

Reproduction by Overwintering Adults of Water Strider, Aquarius paludum (Fabricius)

Tetsuo Harapa!

Department of Biology, Faculty of Science, Osaka City University, Osaka 558, Japan

ABSTRACT— Overwintering adults of a water strider, Aquarius paludum having the macropterous and brachypterous wing forms continued to lay eggs for about 35 days in spring. The oviposition occurred at a high and constant rate under quasi-natural conditions. There was no significant difference between the two wing forms with respect to the number of eggs throughout their most reproductive period. Some of the second generation adults which had overwintered after reproduction in the previous autumn also became reproductively active in the following spring together with the third and diapause

generation adults.

In early spring, macropterous adults appeared on the water surface later than

brachypterous adults, demonstrating the longer distance between diapause sites on land and breeding sites on the water surface in macropterous adults than brachypterous adults.

INTRODUCTION

Among insects, there are many species that are in diapause and overwinter as adults, and the diapause in the adult stage (adult diapause) means that post-emergence development in ovaries does not occur in the female and the ovaries remain small [2]. In most cases of adult diapause, di- apause Occurs in young and nonparous adults, and the adults became reproductively active after over- wintering [2]. Even in reproductive adults, di- apause can be induced by suitable cues (e.g., a short-day photoperiod) in a few species of bugs and beetles [6, 9, 11, 16]. However, reproduction after overwintering has never been reported to be induced twice in the same individuals which nor- mally live for less than one year. In some coleop- teran species living for several years, adults over- winter in diapause and reproduce repeatedly in successive years in nature, e.g., Agonum assimile [10].

Aquarius paludum (Fabricius) is trivoltine and shows alary dimorphism in Kochi, Japan [4, 5]. Although reproductive activity in the first genera-

Accepted January 18, 1993 Received October 21, 1992

" Department of Biology, Faculty of Science, Kochi University, Kochi 780, Japan

tion continued until death in summer, the females of the second generation entered diapause after a reproductive period in September and October [4]. It is unclear whether the adults of the second generation became reproductively active again in the following spring together with the adults of the third generation.

The present work aims, first, at making this point clear and, second, at studying the appear- ance on the water surface and the oviposition process by overwintering brachypterous and macropterous adults of A. paludum under quasi-natural conditions in spring.

MATERIALS AND METHODS

Overwintering adults of Aquarius paludum were collected from a pond in Kochi City (33.3°N, 133.3°E), Japan on the Ist, 4th, 16th, 17th, and 24th of April, and in late September, 1988. The collected water striders were released back into the pond after measurement of wing lengths. Wing length was expressed in terms of the index as per Harada and Taneda [5]. When it was difficult to distinguish macropterous from brachypterous adults by fore-wing lengths, the distinction was made by reference to hind-wing lengths because the two wing forms showed more clear differences

314 10 9 8 x é o 7 67h as 8 6 Number of individuals a 5 Sones fh Ss I 3 ao) 1S AG

Qo =

OF M1 O82) BBN Ze wre Gye:

& 9 i@ il

T. HARADA

TST Ce ce O12 I OOF Ole

o —

0 ie a A yOu ts\u ct) ”1(0) . (I

Fore-wing index Fic. 1. Fore- and hind-wing lengths of the adults collected from a water way in Kochi in late September, 1988.

in the hind-wing lengths than in the fore-wing lengths (Fig. 1).

Overwintering adults of A. paludum were col- lected in early April, 1991, from the pond. In 1990, fifth instar nymphs of the second generation were collected just before adult emergence from two small canals (A- and B-canal). The water temperature in the A- and B-canal had been higher than 27°C and lower than 22°C, respectively, in the daytime in July and August. Fifth instar nymphs belonging to the third generation were collected from A-canal in mid-October, 1991. Pairs consist- ing of one male and one female with the same wing form were put just after collection or adult emerg- ence in individual plastic pots, 14 cm in diameter and 5 cm in depth, and reared under quasi-natural conditions in Osaka City (35.0°N, 135.5°E), Japan. Longevity was examined in each pair and the number of eggs was counted every 1-2 days in September, 1990 to May, 1991, and in October, 1991 to May, 1992. The plastic pots were filled with water to a depth of 5-10mm. A wooden stick, about 10cm in length and about 1 cm in diameter, was put in each pot for oviposition and resting sites. Water in the pots was replaced daily. Water striders were fed on adult specimens of Fannia canicularis or Lucilia illustris which were

supplied daily at a rate of one fly per two adults.

RESULTS

Overwintering females (mainly the third genera- tion) with either of the two wing forms continued to lay eggs for an average of 36.0 days in spring and then died. Eggs were laid at a constant and high rate (about 60 eggs per 5 days per female) through- out the reproductive period for both wing forms (Fig. 2). The average total number of eggs per female was 386.0 in brachypterous females and 396.7 in macropterous females. Difference in numbers of eggs laid per five days was not signi- ficant between the two wing forms (t-test, P>0.05) except a period between 16th and 21st May (P< 0.05). Furthermore, no significant difference was observed between the two wing forms with respect to longevity (t-test, P>0.05).

In the second generation, all the reproductive females laid the fertilized eggs in fall. The repro- ductive females stopped laying eggs between the 21st September and the 31st October, and entered diapause (Fig. 3). None of the pairs in the third generation reproduced in autumn, and all entered diapause (Fig.3). In early December, the di- apause adults of the second and third generations

Reproduction by Overwintering Aquarius 315

Number of eggs / female, 5days

survival

Percent

10 20

April

30 10 20 31

May

Fic. 2. Survival curves and rates of oviposition for overwintering females (mainly the third generation). Open and solid symbols show brachypterous and macropterous groups, respectively. Arrows show the day on which adults were collected from the pond in Kochi, Japan, 1991. n=34 for brachypterous pairs; n=29 for macropterous

pairs.

were transferred onto fallen leaves as overwinter- ing sites. At the beginning of the following March, surviving adults were moved again onto the water surface. For specimens derived from the A-canal, the survival percentage just prior to the move in March was higher in the third generation (females, 84.6%; males, 76.9%) than that in the second generation (females, 21.4%; males, 42.9%) (Fig. 3). In the second generation, when the surviving adults were transferred to the water surface, three pairs were made and the four remaining males were paired with virgin females which had been reared during the nymphal stage under a long-day (14.5L-9.5D) photoperiod at 20+2°C. In the third generation, ten pairs were made in early March. The three females in the second generation and surviving five females in the third generation began to lay eggs on average about 2 weeks after the move to the water surface [second generation, 13.3 +1.5 (S.D.) days; third generation, 15.6+9.0 (S.D.) days] (Fig. 3). The number of eggs laid

after diapause was thought to correlate positively with the number of eggs before diapause, although the correlation could not be inferred statistically because of the insufficient number of samples (Fig. 4A). Both the three overwintering females and the reproductive four females that had been paired with the remaining overwintering males in the second generation laid fertilized eggs in addition to 4 out of 5 females of the third generation.

The proportion of brachypterous adults on the water surface was higher at the beginning than the end of April following overwintering (77-test be- tween 4th and 24th April, P<0.05; Fig. 5).

DISCUSSION

Diapause and migration are most closely associ- ated when diapause sites and breeding sites are separated in insects [7]. Water striders are typical examples of this close association. Landin and Vepsaladinen [8] caught many flying long-winged

316 T. HARADA

Sep. Oct. Nov. Dec.

individuals

of

Number

Temperature

@—e—e— @—_ ©

individuals

Number of

Sep. Oct. Nov. Dec. Fic. 3.

Jan. Feb.

| oe Aa stage MET” YF

A--~-A----A----A----& @——_ © © ‘a----A-----A---'1@—@—@—@——@

SS An-=- Ae --=-A-----A-----A ®

. a ~~ 4 Jan. Feb. Mar. Apr.

Survival curves for adults in the second (A, B) and third (A’) generations derived from A- (A, A’) and

B-canal (B). Allindividuals: macropterous. Circles, females; triangles, males. Shaded area shows the number of females which were laying eggs. Arrows indicate the days on which the water striders were transferred from fallen leaves to the water surface in 1991 (A, B) or 1992 (A’). The mean date of adult emergence was 29 August and 15 October in the second and third generations, respectively. C, Seasonal variations in air temperature in Kochi (circles) and Osaka (triangles, squares). The triangles and squares show the means for the first, middle, or last ten days of each month in 1990-1991 and 1991-1992, respectively. The circles indicate the means of the values for 30 years (1961-1990). The data of air temperature were gained by Osaka and Kochi Meteorological observatories.

adults of Gerris argentatus on the way from their diapause sites to water surfaces in a meadow on the shore of a lake in Finland. Their capture suggests that the macropterous adults of G. argen- tatus overwintered at sites far from water surfaces. In early spring, no late-awakening from diapause but the long distance between diapause and breed- ing sites is demonstrated in the case of macropter- ous adults of A. paludum by the following: mac- ropterous adults appear on water surface later than

brachypterous adults (Fig. 5), macropterous adults show similar reproductive and survival processes to brachypterous adults (Fig. 2). Moreover, the two facts, the earlier appearance of brachypterous adults (Fig. 5) and the lack of flight ability in the brachyperious having very short hind-wing (Fig. 1), suggest that their diapause sites are near water surface, e.g. under logs or stones on the banks as is the case with Aquarius remigis [13].

A high rate of oviposition of more than 60 eggs

Reproduction by Overwintering Aquarius 317

hibernation

400

300

eggs after

200

100

Number of

0 100 200 300 400

500

400

300

200

100

500 600 700

Number of eggs before hibernation

Fic. 4. The total number of eggs laid by each adult before or after overwintering in the second (A) and third (B) generations. The circle and bar in B indicate the mean and S.D., respectively.

per female per 5 days was recorded only during a part of the reproductive period of the first and second generations [4]. However, overwintering adults laid more than 60 eggs per female per 5 days throughout almost their entire reproductive period (Fig. 2). Overwintering adults of A. paludum also continue to show strong positive phototaxis for about 50 days in spring [3]. Harada showed experimentally that the experience of overwinter- ing made the adults easy to show the strong positive phototaxis. The overwintering adults of A. paludum may be, physiologically, in a specific phase after overwintering. Spence [12] also re- ported that in some Canadian water striders, Ger- ris pingreensis, G. comatus, G. buenoi, and Lim- noporus dissortis, diapause breeders showed high- er reproductive rates in spring than direct breeders in summer. The experience of overwintering in- cludes the experience of diapause as an internal condition, as well as exposure to low temperature and other winter-related environmental condi- tions. It is unclear which of these factors is the major one which induces a constant and high rate of oviposition in A. paludum.

Spence [12] reported that overwintered mac- ropterous females of G. pingreensis had longer pre-oviposition periods than overwintered apter- ous females, and that their reproductive rate was lower than that of apterous females in early spring.

By contrast, there was no difference between the two wing forms in oviposition rates throughout the reproductive period after overwintering in A. palu- dum (Fig. 2). Why do macropterous adults of A. paludum continue to lay eggs at the same rate as brachypterous adults, even though macropterous adults seem to expend so much energy during their spring dispersal flight? They may histolyze their wing muscles and their resulting nutrients may be used for the maturation of eggs, as shown in G. odontogaster [1], and in G. lacustris, G. lateralis, and G. argentatus [14].

A high fecundity during the early adult stage was shown in the second generation and that is advan- tageous for many nymphs of the third generation who can develop into adults before winter comes [4]. The overwintered females of the second generation participated in the reproduction again in the following spring together with third genera- tion females (Fig. 3). The fecundity of the second generation adults was not smaller than that in the third generation (Fig. 4), and moreover the surviv- ing males in the second generation fertilized the eggs in the following spring. Therefore, it seems that diapause after reproduction in autumn and the resulting second reproduction during the subse- quent spring are adaptive tactics for the second generation adults to bequeath a larger number of offspring.

318 T. HARADA

= é

tba 4hG7 Bem ee e4aby? BY ton

Che) Go

(ead obey OowWmM eae 6 YH 8 iw

Wing index Fic. 5. The appearance of Aquarius paludum on the water surface of a pond in Kochi in early spring, 1988. Open bars, brachypterous adults; shaded bars, macropterous adults.

Reproduction by Overwintering Aquarius

The overwintered (macropterous) adults in both generations laid eggs shortly after the move to water surfaces, although, in our study, the adults were moved to water surfaces earlier than mac- ropterous adults would normally do so in nature (Figs. 3, 5). This earlier reproduction suggests that the water surface as habitat becomes a cue that induces the maturing of the reproductive system. Wilcox and Maier [15] reported that adults of Aquarius remigis estivated facultatively during summer in damp areas underneath stream-bed rocks when a pool in a small stream went dry. However, they showed no data on whether the reproductive organs matured or not in the esti- vated adults. The causal role of the existance of water surfaces in reproductive maturation should be investigated further experimentally on water striders in the future.

ACKNOWLEDGMENTS

I would like to thank Assistant Professor Dr. Hideharu Numata, Osaka City University for his valuable advice.

REFERENCES

1 Andersen NM (1973) Seasonal polymorphism and developmental changes in organs of flight and repro- duction in bivoltine pondskaters (Hem. Gerridae). Ent Scand 4: 1-20

2 Danks HV (1987) Insect Dormancy: an Ecological Perspective. Biological Survey of Canada, Ottawa, pp 71-74

3 Harada T (1991) Effects of photoperiod and temperature on phototaxis in a water strider, Gerris paludu minsularis (Motschulsky). J Insect Physiol 37: 27-34

4 Harada T (1992) The oviposition process in two direct breeding generations in a water strider, Aquarius paludum (Fabricius). J Insect Physiol 38: 687-692

5

11

12

13

14

16

319

Harada T, Taneda K (1989) Seasonal changes in alary dimorphism of a water strider, Gerris paludum insularis (Motschulsky). J Insect Physiol 35: 919-924 Hodek I (1977) Photoperiodic response in spring in three Pentatomoidea (Heteroptera). Acta Ent Bohemoslov 74: 209-218

Kennedy JS (1975) Insect dispersal In “Insects, Science, and Society” Ed by D Pimentel, Academic Press, New York, pp 103-119

Landin J, Vepsalainen K (1977) Spring dispersal flights of pond-skaters Gerris spp. (Heteroptera). Oikos 29: 156-160

Muraji M, Nakasuji F (1990) Effect of photo- periodic shifts on egg production in a semiaquatic bug, Microvelia douglasi (Heteroptera: Veliidae). Appl Ent Zool 25: 405-407

Neudecker C, Thiele HU (1974) Die jahreszeitliche Synchronisation der Gonadenreifung bei Agonum assimile Payk. (Coleopt. Carab.) durch Temperatur und Photoperiode. Oecologia (Berl.) 17: 141-157

Numata H, Hidaka T (1982) Photoperiodic control of adult diapause in the bean bug, Riptortus clavatus Thunberg (Heteroptera: Coreidae). I. Reversible induction and termination of diapause. Appl Ent Zool 17: 530-538

Spence JR (1989) The habitat templet and life history strategies of pond skaters (Heteroptera: Gerridae): reproductive potential, phenology, and wing dimorphism. Can J Zool 67: 2432-2447 Torre-Bueno JR de la (1917) Life-history and habits of the larger waterstrider, Gerris remigis Say. En- tomol New 28: 201-208

Vepsdlainen K, Patama T (1983) Allocation of reproductive energy in relation to the pattern of environment in five Gerris species In “Diapause and Life Cycle Strateries in Insects” Ed by VK Brown and I Hodek, Dr W. Junk Publishers, The Hague, pp 189-207

Wilcox RS, Maier HA (1991) Facultative estivation in the water strider Gerris remigis. Can J Zool 69: 1412-1413

Zaslavsky VA (1975) [The distribution of stepwise photoperiod reactions among Insecta and Acari.] Ent Obozr 54: 291-304 (In Russian) [Translation in Ent Rev 54: 36-45]

ry

; sehen », “yr vos

A ive ‘whunett. 2 = ¢ Weel an eae dvbanP scgeah 5 invader a aun Teoh tpi! - ih Lictesidinet 5 TL) ie ‘yA Wein stall t i ie @ ; ir. : 1 a ma) 4 yf mh mtd By o t ¥ 4 1 “f ¥

eee stub “Say aim oft fin

it | flube 7h) be shat fel

bs lamba he dinsuapb nd

aun

a Mashigs we “ea

pred fai ‘

* Quen fi 62 ol yhioseacty: os

Oily ate SggUe noioukwagat litt ony sonore tate tai te av ataalsoryeyt ig 1a And

Ae rf b gloviustouy| 1) ese es fi pm

_ 75 Tit sores tebe re

ly? sbhgiests wit ire ous iw = " i 7 a Laois Whew teeny by pilpwbene

‘Mh talindw ao piah sar 4 nt

i i a osinatineny

tio oo wenden toll

t* neodcihOl anes sania :

ZOOLOGICAL SCIENCE 10: 321-328 (1993)

Prostaglandins and Sex Steroids from Reptilian (Podarcis sicula sicula) Ovarian Follicles at Different Developmental Stages

ANNA GosBETTI, MASSIMO ZERANI, MARIA MADDALENA Di Fiore!

and VirciILio Botte!

Department of Molecular, Cellular, and Animal Biology, University of Camerino, 62032 Camerino MC, Italy, and ‘Department of Zoology, University of Naples, 80134 Naples, Italy

ABSTRACT—Prostaglandin F,, (PGF2,), prostaglandin E, (PGE>), progesterone, androgens, and estradiol-17£ in vitro basal release by follicles of the oviparous lizard, Podarcis s. sicula was studied; in addition, the in vitro effect of PGF,, and PGE, on sex steroid release was evaluated. Follicles were divided according to the different vitellogenic developmental stages: pre-vitellogenic, early-vitellogenic, mid-vitellogenic and fully-grown. PGF>, and progesterone basal release was highest in fully-grown follicles; PGE, and estradiol basal release was highest in early-vitellogenic follicles; androgens basal release was detectable in mid-vitellogenic and fully-grown vitellogenic follicles only. PGF>, increased progesterone release by fully-grown follicles; PGE, increased estradiol release by all follicle types, except by early-vitellogenic ones. The present data suggest that PGF, and PGE, exert different roles on follicles: PGF, seems to induce ovulation through the mediation of progesterone, while PGE, seems to be implied in the start and the sustaining of oocyte vitellogenic development through the mediation of

© 1993 Zoological Society of Japan

estradiol.

INTRODUCTION

In mammals, prostaglandins (PGs) of both F and E series modulate the ovary function, including steroidogenesis [6, 9, 18, 29]. PGs intervene in ovulation and oviduct relax in birds [1, 30]. As regards oviparous reptiles, oviposition is corre- lated with a dramatic elevation of prostaglandin F (PGF) and prostaglandin E> (PGE?) in the logger- head turtle, Caretta caretta [20] and the tuatara, Sphenodon punctatus [19]; PGF, exhibits luteo- lytic effects in the lizard, Anolis carolinensis [21], and the snapping turtle, Chelydra serpentina [24], determining a decline in plasma progesterone. Little is known about the involvement of PGs in the vitellogenic development of follicles in reptiles, and, in particular, nothing is known about the role of PGs on the reproductive processes of the ovi- parous lizard, Podarcis sicula sicula, even if the

Accepted November 9, 1992 Received September 18, 1992

reproductive cycle of this lizard has been well studied [2, 3, 10].

In this study we have compared the PGF,,, PGE, progesterone, androgens, and estradiol-17 in vitro basal releases by follicles, at different vitellogenic developmental stages, of the lizard, Podarcis s. sicula; in addition, the in vitro effect of PGF), and PGE, on progesterone, androgens, and estradiol-178 releases by the same follicles has been examined.

MATERIALS AND METHODS

Animals The reproductive cycle of the female lizard, Podarcis s. sicula, living in the sourround- ings of Naples (Campania, Italy, 25-75 m above sea level) is here briefly described. From the end of summer to the beginning of the next spring (stasis), gonads and genital apparatus are quies- cent; then, soon after emerging, the ovary and oviduct undergo recovery and reach their full

growth from March to mid-May (recovery period);

322 A. GosBetti, M. ZERANI et al.

ovulation and egg deposition occur from mid-May to July (ovulatory period); after deposition, gonads and oviduct regress (postovulatory period) [25 3, 10).

Adult female lizards, Podarcis s. sicula, of this population were captured in May and transferred to laboratory terraria, maintained in ambient photothermal conditions (L/D: 15/9; 18°C), and fed with meal worms and fresh vegetables ad libitum.

In vitro incubations of follicles The lizards were sacrificed by decapitation, the developmental stage of the reproductive organs was observed and the ovaries rapidly removed. Follicles were freed under dissection microscope and placed in cold Dulbecco’s modified Eagle medium (DME, Sig- ma, USA) containing 10mM Hepes, 1mg Penicillin/ml and 2 mg Streptomycin/ml. Follicles were divided according to the different vitellogenic developmental stages: pre-vitellogenic (500-1000 ym ¢), early-vitellogenic (1500-3000 um ¢), mid- vitellogenic (4000-6000 ~m ¢) and fully-grown (+8000 um ¢). The follicles were placed in multi- ple tissue culture plates (Beckton Dickinson, USA); each well contained one follicle. Six folli- cles of each type were divided into 3 experimental groups (each consisting of 2 wells). The ex- perimental groups were: a) follicle with DME alone; b) follicle with DME plus PGF,, (150 ng); c) follicle with DME plus PGE; (150 ng). The final volume of each well was 1 ml. Culture plates were wrapped with aluminium foil and incubated in a shaking water bath (32°C), set at 30 rpm. One well of each experimental group was removed, respec- tively, after 3 and 6 hr of incubation. The incuba- tion medium samples were immediately stored at —20°C for later hormone determination. In addi- tion, the experiment was repeated without folli- cles. Tests on 5 parallel incubation sets were carried out. Preliminary evidence led our choosing to the incubation conditions and the PGF>, or PGE, minimum effective dose utilized in the pre- sent in vitro study (data not shown).

PGF),, PGE, progesterone, androgens, and estra- diol-178 determination PGF>, and PGE> were determined in incubation medium samples by

radioimmunoassay (RIA) [12, 13]. The deter- minations were carried out on duplicate incubation medium samples (500 1) that were extracted with 5ml of diethyl ether for 4min. The organic fractions were transferred into glass tubes and evaporated to dryness under a nitrogen stream. The extracts were resuspended with 1 ml of assay buffer and assayed. The recovery of labelled PGF,, and PGE> were 83.3+1.3% and 88.8+ 1.7%, respectively. Parallelism among the stan- dard curve in buffer, a standard curve in incuba- tion medium (then extracted) and a serial dilution of a single medium sample (extracted) were con- stant.

Concentrations of progesterone, androgens and estradiol-17/ in incubation media were determined by RIA in accordance with methods previously reported [7, 13].

The following sensitivities were recorded: PGF>, 13 pg (intraassay variability: 7.5%; interassay variability: 14.5%); PGE, 15.5 pg (intraassay variability: 7.0%; interassay variability: 12.5%); progesterone, 9 pg (intraassay variability: 8.0%; interassay variability: 10.0%); androgens, 7.5 pg (intraassay variability: 6.0%; interassay variabil- ity: 11.5%); estradiol-172, 7pg (intraassay variability: 8.5%; interassay variability: 13.0%). PGF,,, progesterone, testosterone, and estradiol- 178 antisera were provided by Dr. G. F. Bolelli and Dr. F. Franceschetti (CNR-Physiopathology of Reproduction Service, University of Bologna, Italy). The PGE, antiserum was purchased from Cayman Chemical (USA). Testosterone was not separated from 5a-dihydrotestosterone; therefore, the antiserum used is not specific and the data are expressed as androgens. Tritiated PGF>,, PGE», progesterone, testosterone, and estradiol-17 were purchased from Amersham International (UK) and non-radioactive PGF>,, PGE2, progesterone, testosterone, and estradiol-17? from Sigma.

Statistics Data relative to each hormone were submitted to analysis of variance (ANOVA) fol- lowed by Duncan’s multiple range test [8, 31]. Correlation coefficients followed the procedures of Scossiroli and Palenzona [28].

PGs and Steroids from Podarcis Follicles 323

RESULTS

PGF>, basal release was higher (P<0.01) in fully-grown follicles with respect to the other folli- cle types; PGF>, was lower (P<0.01) in pre- vitellogenic with respect to early-vitellogenic and mid-vitellogenic follicles (Fig. 1). PGE, basal re- lease was higher (P<0.01) in early-vitellogenic with respect to the other follile types; PGE, was

pg/follicle

1000

750

500

lower in pre-vitellogenic with respect to mid- vitellogenic and fully-grown follicles (Fig. 2). Progesterone basal release was higher (P<(0.01) in fully-grown with respect to the other follicle types; progesterone was higher (P<0.01) in mid- vitellogenic with respect to the pre-vitellogenic and early-vitellogenic follicles; progesterone was high- er (P<0.01) in early-vitellogenic with respect to pre-vitellogenic follicles (Fig.3, upper panel).

Incubation times (hours)

Fic. 1.

PGF,, basal release by follicles, at different vitellogenic developmental stages (see Materials and Methods),

of the oviparous lizard, Podarcis s. sicula. Follicles: pre-vitellogenic (black bars), early-vitellogenic (white bars), mid-vitellogenic (grey bars), fully-grown (hatched bars). Each mean refers to 5 values+SD. a, P<0.01 vs pre-vitellogenic, early-vitellogenic and mid-vitellogenic follicles; b, P<0.01 vs pre-vitellogenic follicles (Duncan’s multiple range test).

pg2/follicle

4500

3000

1500

Incubation times (hours) PGE; basal release by follicles, at different vitellogenic developmental stages (see Materials and Methods), of the oviparous lizard, Podarcis s. sicula. Follicles: pre-vitellogenic (black bars), early-vitellogenic (white bars),

Fic. 2.

mid-vitellogenic (grey bars), fully-grown (hatched bars). Each mean refers to 5 values +SD. a, P<0.01 vs pre-vitellogenic, mid-vitellogenic and fully-grown follicles; b, P<0.01 vs pre-vitellogenic follicles (Duncan’s multiple range test).

324 A. Gossetti, M. ZERANI et al. pg2/follicle 6000

4000

2000

pg/follicle 6000

4000

2000

Incubation times (hours)

Fic. 3. Progesterone basal release (upper panel) and PGF>, effects (lower panel) on progesterone release by follicles, at different vitellogenic developmental stages (see Materials and Methods), of the oviparous lizard, Podarcis s. sicula. Follicles: pre-vitellogenic (black bars), early-vitellogenic (white bars), mid-vitellogenic (grey bars), fully-grown (hatched bars). Each mean refers to 5 values+SD. a, P<0.01 vs pre-vitellogenic, early-vitellogenic and mid-vitellogenic follicles; b, P<0.01 vs pre-vitellogenic and early-vitellogenic follices; c, P <0.01 vs pre-vitellogenic follicles; *~P<0.01 vs the same follicle type incubated with PGF>, (Duncan’s multiple range test).

Androgens basal release was not detectable in early-vitellogenic with respect to the other follicle pre-vitellogenic and early-vitellogenic follicles; types; estradiol was lower (P<0.01) in pre- androgens were higher (P<0.01) in fully-grown vitellogenic with respect to mid-vitellogenic and with respect to mid-vitellogenic follicles (Fig. 4). fully-grown follicles (Fig. 5, upper panel). PGF, Estradiol-17 basal release was higher (P<0.01)in basal release values were positively correlated

Fic. 5. Estradiol-17 basal release (upper panel) and PGE; effects (lower panel) on estradiol-17/ release by follicles, at different vitellogenic developmental stages (see Materials and Methods), of the oviparous lizard, Podarcis s. sicula. Follicles: pre-vitellogenic (black bars), early-vitellogenic (white bars), mid-vitellogenic (grey bars), fully-grown (hatched bars). Each mean refers to 5 values+SD. a, P<0.01 vs pre-vitellogenic, mid-vitellogenic and fully-grown follicles; b, P<0.01 vs pre-vitellogenic follicles; *P<0.01 vs the same follicle type incubated with PGE, (Duncan’s multiple range test).

PGs and Steroids from Podarcis Follicles 325 pg/follicle a 120

80

40

Incubation times (hours)

Fic. 4. Androgens basal release by follicles, at different vitellogenic developmental stages (see Materials and Methods), of the oviparous lizard, Podarcis s. sicula. Follicles: pre-vitellogenic (black bars), early-vitellogenic (white bars), mid-vitellogenic (grey bars), fully-grown (hatched bars). Each mean refers to 5 values +SD. a, P< 0.01 vs mid-vitellogenic follicles (Duncan’s multiple range test); nd, not detectable.

pg2/follicle

750

500

250

pg2/follicle

750

500

250

Incubation times (hours)

326 A. GossettiI, M. ZERANI et al.

TaBLE 1. Correlation coefficents among PGF>,, PGE>, progesterone and estradiol-177 levels re- leased by follicles, at different vitellogenic de- velopmental stages, of Podarcis s. sicula

Incubation times

3 hr 6 hr PGF,, vs progesterone 0.804 0.814 PGE, vs estradiol-17£ 0.862 0.911

All correlations show the same level of significance (P<0.001); df=18.

(P<0.001) to those of progesterone (Table 1); PGE, basal release values were positively corre- lated (P<0.001) to those of estradiol-17£ (Table Ys

PGF;, induced an increase (P<0.01) in proge- sterone release by fully-grown follicles (Fig. 3, lower panel). PGE, induced an increase (P<0.01) in estradiol release by pre-vitellogenic, mid- vitellogenic, and fully-grown follicles (Fig. 5, low- er panel).

DISCUSSION

This is the first study regarding the in vitro basal release of PGs and sex steroids by the follicles, kept at different vitellogenic developmental stages, of the oviparous lizard Podarcis s. sicula.

PGF,, showed the highest amounts in fully- grown follicle incubation samples; this finding sup- ports the idea that PGF), could be involved in the stimulation of ovulation; on the other hand, Goetz [15] reported that in several fish species, PGs can induce ovulation; in particular, plasma and ovarian tissue PGF,, levels are higher during ovulation in the yellow perk, Perca flavescens [17], in the brook trout, Salvelinus fontinalis [16], in the goldfish, Carassius auratus [4], and in the pond loach, Misgurnus anguillicaudatus [25]. Also in amphi- bians, PGF, seems to be implied in ovulation processes, seeing that PGF, in vitro induces ovulation in Rana pipiens [27]; more recently Gobbetti and Zerani [14] suggested that PGF, is implied in the control of ovulation in Rana esculenta. The involvement of PGF>, in the ovula- tory process of this lizard is also supported by the stimulatory role exerted by this prostaglandin on

progesterone increase by fully-grown follicles. Progesterone basal release was higher in fully- grown follicles, if compared with the other follicle types, so suggesting that this steroid is involved in the ovulation of Podarcis, as many authors re- ported for amphibian species [23]; in fact, prog- esterone, produced by follicle cells, induces ovula- tion in Xenopus laevis [32] and Rana pipiens [26].

The early-vitellogenic follicles released the high- est values of PGE, and estradiol, which both, however, remained still high in mid-vitellogenic and fully-grown follicles. The positive correlation between the basal release values of PGE> and estradiol suggests a causal relationship and this hypothesis is validated by the PGE, effect in increasing estradiol release. This finding might indicate a possible trigger role for this prostaglan- din in the start of the vitellogenic development of the oocytes and also a role in sustaining the following vitellogenic growth of oocytes. In this context we recall that estradiol induces the hepatic vitellogenin synthesis for the development of the follicle, as widely reported for reptiles [5, 22]. The PGE, stimulatory effect on estradiol release was detected in all follicle types, except for early- vitellogenic ones. This phenomenon could be due to the refractoriness of this follicle type to an additional stimulation of PGE); in fact the early- vitellogenetic follicles released the highest amounts of PGE; and estradiol.

Androgens were detected only in the incubation media of mid-vitellogenic and fully-grown follicles in accordance with Fortune [11], who found that only large follicles released these hormones in vitro in the amphibian Xenopus laevis. The meaning of this last result is unclear; it could be due to the catabolism of progesterone which exhibits the highest values in incubation media of mid- vitellogenic and: fully-grown follicles; however, a role for androgens produced by these two follicle types cannot be excluded.

Summarizing, this work suggests different roles for PGF>, and PGE; released by Podarcis s. sicula follicles: PGF>, could induce ovulation through progesterone mediation, while PGE, could be implied in the start and the sustaining of vitel- logenic development of oocytes through estradiol mediation.

PGs and Steroids from Podarcis Follicles

ACKNOWLEDGMENTS

The authors would like to thank James Burge, of the

Camerino University Institute of Linguistics, for help

wi

M

th English. This work was supported by grant from URST.

REFERENCES

Asem EK, Todd H, Hertelendy F (1987) In vitro effect of prostaglandins on the accumulation of cyclic AMP in the avian oviduct. Gen Comp Endo- crinol 66: 244-247

Botte V, Angelini F, Picariello O, Molino R (1976) The regulation of the reproductive cycle of the female lizard, Lacerta sicula sicula Raf. Monitore Zool Ital 10: 119-133

Botte V, Delrio G (1965) Ricerche istochimiche sulla distribuzione dei 3- e 17-chetosteroidi e di alcuni enzimi della steroidogenesi nell’ ovario di Lacerta sicula. Boll Zool 32: 191-198

Bouffard RE (1979) The role of prostaglandins during sexual maturation, ovulation and spermiation in the goldfish, Carassius auratus. M Sc Thesis, University of British Columbia, Vancouver, BC, Canada

Carnevali O, Mosconi G, Angelini F, Limatola E, Ciarcia G, Polzonetti-Magni A (1991) Plasma vitel- logenin and 17(-estradiol levels during the annual reproductive cycle of Podarcis s. sicula Raf. Gen Comp Endocrinol 84: 337-343

Dharmarajan AM, Sueoka K, Miyazaki S, Atlas SJ, Ghodgaonkar RB, Dubin NH, Zirkin BR, Wallach EE (1989) Prostaglandins and progesterone secre- tion in the in vitro perfused pseudopregnant rabbit ovary. Endocrinolgy 124: 1198-1203

D'Istria M, Delrio G, Botte V, Chieffi G (1974) Radioimmunoassay of testosterone, 17f-oestradiol and estrone in the male and female plasma of Rana esculenta during sexual cycle. Ster Lip Res 5: 42-48 Duncan DB (1955) Multiple range and multiple F test. Biometrics 11: 1-42

Espey LL, Norris C, Saphire D (1986) Effect of time and dose of indomethacin on follicular prosta- glandins and ovulation in the rabbit. Endocrinology 119: 746-754

Filosa S (1973) Biological and cytological aspects of the ovarian cycle in Lacerta sicula Raf. Monitore Zool Ital 7: 151-165

Fortune JE (1983) Steroid production by Xenopus ovarian follicles at different developmental stages. Dev Biol 99: 502-509

Gobbetti A, Zerani M (1991) Gonadotropin- releasing hormone stimulates biosynthesis of pro- staglandin F>, by the interrenal gland of the water

14

15

16

17

18

19

20

21

22

23

24

327

frog, Rana esculenta, in vitro. sen Comp Endocri- nol 84: 434-439

Gobbetti, A, Zerani M (1992) PGF3,, PGE, and sex steroids from the abdominal gland of the male crested newt Triturus carnifex (Laur.). Prostaglan- dins 43: 101-109

Gobbetti A, Zerani M (1992) A possible involve- ment of prostaglandin F3, (PGF>,) in Rana esculenta ovulation: effects of mammalian gonadotropin- releasing hormone on in vitro PGF, and 17/- estradiol production from ovary and oviduct. Gen Comp Endocrinol 87: 163-170

Goetz FW (1983) Hormonal control of oocyte final maturation and ovulation in fishes. In “Fish Physiol- ogy VolIXB” Ed by WS Hoar, DJ Randall, EM Donaldson, Academic Press, New York, pp 117- 170

Goetz FW, Duman P, Ranjan M, Herman CA (1989) Prostaglandin F and E synthesis by specific tissue components of the brook trout (Salvelinus fontinalis) ovary. J Exp Zool 250: 196-205

Goetz FW, Theofan G (1979) Jn vitro stimulation of germinal vesicle breakdown and ovulation of yellow perch (Perca flavescens) oocytes: effects of 17a- hydroxy-20-dihydroprogesterone and prostaglan- dins. Gen Comp Endocrinol 37: 273-285 Greenhalgh EA (1990) Luteal steroidogenesis and regression in the rat: progesterone secretion and lipid peroxidation induced in luteal cells by human chorionic gonadotrophin, phospholipase A, and prostaglandin F5,. J Endocrinol 125: 397-402 Guillette LJ, Jr, Bjordal KA, Bolten AB, Gross TS, Palmer BD, Witherington BE, Matter JM (1991) Plasma estradiol-17@, progesterone, prostaglandin F, and prostaglandin E, concentrations during natu- ral oviposition in the loggerhead turtle (Caretta caretta). Gen Comp Endocrinol 82: 121-130 Guillette LJ, Jr, Cree A, Gross TS (1990) Endocri- nology of oviposition in the tuatara (Sphenodon punctatus): I. Plasma steroids and prostaglandins during natural nesting. Biol Reprod 43: 285-289 Guillette LJ, Jr, Lavia LA, Walker NJ, Roberts DK (1984) Luteolysis induced by prostaglandin F>, in the lizard, Anolis carolinensis. Gen Comp Endocri- nol 56: 271-277

Ho SM (1987) Endocrinology of vitellogenesis. In “Hormones and Reproduction in Fishes, Amphi- bians, and Reptiles” Ed by DO Norris, RE Jones, Plenum Press, New York, pp 145-170

Jones RE (1987) Ovulation: insights about the mechanisms based on a comparative approach. In “Hormones and Reproduction in Fishes, Amphi- bians, and Reptiles” Ed by DO Norris, RE Jones, Plenum Press, New York, pp 203-240

Mahmoud IY, Cyrus RV, McAsey ME, Cady D, Woller MJ (1988) The role of arginine vasotocin and

25

26

Dif

28

328

prostaglandin F>, on oviposition and luteolysis in the common snapping turtle Chelydra serpentina Gen Comp Endocrinol 69: 56-64

Ogata H, Nomura T, Hata M (1979) Prostaglandin PGF,, changes induced by ovulatory stimuli in the pond loach, Misgurnus anguillicaudatus. Bull Jpn Soc Sci Fish 45: 929-931

Schuetz AW (1967) Effect of steroids on germinal vesicle of oocytes of the frog (Rana pipiens) in vitro. Proc Soc Ex Biol Me 124: 1307-1310

Schuetz AW, Lessman AL (1981) Role of the surface epithelium of the ovarian follicle in the ovulation of frog (Rana pipiens) oocytes. J Cell Biol 91: 187a

Scossiroli RE, Palenzona DL (1979) Manuale di biometria. Zanichelli Editore, Bologna

29

30

31

32

A. Gossetti, M. ZERANI et al.

Sedrani SH, El-Banna AA (1987) Effect of in- domethacin on plasma level of vitamin D metabo- lites, oestradiol and progesterone in rabbit during early pregnancy. Comp Biochem Physiol 87: 635— 639

Shimada K, Olson DM, Etches RJ (1984) Follicular and uterine prostaglandin levels in relation to uter- ine contraction and the first ovulation of a sequence in the hen. Biol Reprod 31: 76-82

Sokal RR, Rohlf FJ (1981) Biometry. Freeman, San Francisco, 2nd ed

Thibier-Fouchet C, Mulner O, Ozon R (1976) Pro- gesterone biosynthesis and metabolism by ovarian follicles and isolated oocytes of Xenopus laevis. Biol Reprod 14: 317-326

ZOOLOGICAL SCIENCE 10: 329-336 (1993)

Plasma Steroid Hormone Profiles in HCG-Injected Male Walking Catfish Clarias batrachus

MUHAMMAD ZAIRIN, JR., KryosHi ASAHINA!,

KIYOSHI FURUKAWA and KAtTsuMI AIDA

Department of Fisheries, Faculty of Agriculture, The University of Tokyo, Bunkyo-ku, Tokyo 113, Japan ‘College of Agriculture and Veterinary Medicine, Nihon University, Setagaya-ku, Tokyo 154, Japan

ABSTRACT—The present experiment was performed in order to investigate the response of matured male walking catfish to human chorionic gonadotropin (HCG).

Short-term experiment. Mature male walking catfish received a single intramuscular injection of HCG (0.2 and 11U/g BW) or saline. Blood samples were taken at 0, 6, 12, 18, 24, 36, 48 and 72 hr. Long-term experiment. Mature male walking catfish received a single intramuscular injection of HCG (1 IU/g BW) or saline. Blood samples were taken at 2-day intervals until day 10. In both experiments, plasma testosterone, 11-ketotestosterone (11-KT), 17a-hydroxyprogesterone, 17a,20a-dihydroxy-4- pregnen-3-one and 17a,20-dihydroxy-4-pregnen-3-one levels were measured by radioimmunoassay, and 17a@,20,21-trihydroxy-4-pregnen-3-one levels were measured by enzyme immunoassay.

Results can be summarized as follows. 1) HCG firstly induces an increase in plasma testosterone and 11-KT levels. 2) Thereafter, 11-KT levels decrease and progestin levels increase. 3) Finally, progestin levels decrease, and 11-KT levels increase again. These results clearly indicate that a shift in the

© 1993 Zoological Society of Japan

steroidogenic pathway is induced in males in response to HCG.

INTRODUCTION

Walking catfish (Clarias batrachus) is a freshwa- ter teleost which is widely distributed throughout Southeast Asia. Although this fish has been com- monly cultured for a long period, little information is available on its reproductive physiology, espe- cially regarding males. The reproductive organ of the male tropical walking catfish differs from other teleosts such as cyprinids and salmonids. Walking catfish testes are very small; therefore, milt cannot be obtained by applying pressure to the abdomen in the manner that is commonly employed with other fishes. Male walking catfish reared under 23-25°C start to mature at an age of 9 months, and thereafter maintain conditions of maturity. Plasma testosterone and 11-ketotestosterone (11-KT) in males were always maintained at high levels throughout the year under warm water conditions,

Accepted December 3, 1992 Received August 24, 1992

but progestin levels were low [26].

It has been proposed that during spermiation and spawning, a shift in the steroidogenic pathway from androgen to progestin production occurs in response to increased gonadotropin secretion in salmonids [6, 12, 24], in cyprinids [5], and in a perciform [9]. Progestins, such as 17a,20/-dihy- droxy-4-pregnen-3-one (17a,208-P), and 17a,20a- dihydroxy-4-pregnen-3-one (17a,20a-P), have been reported to exist in males, and to increase during spermiation and spawning in response to increased gonadotropin secretion. In addition, another progestin, 17a,208,21-trihydroxy-4-preg- nen-3-one (20-S) has been reported to increase in the female Atlantic croacker [23], in the spotted seatrout [22] and in the walking catfish [25] during final oocyte maturation. However, the presence of 206-S in male teleosts has not been reported yet.

There still remains a paucity of information on hormonal changes during natural spawning in the male walking catfish; it is generally difficult to follow hormonal changes in individual fish due to

330 M. ZarrRIN Jr., K. ASAHINA et al.

extreme sensitivity to handling stress. Therefore, in this investigation changes in plasma levels of testosterone, 11-KT, 17a-hydroxyprogesterone (17a-P), 17a,20e-P, 17a,208-P and 208-S were monitored in the HCG-treated male walking catfish.

MATERIALS AND METHODS

Fish Walking catfish fingerlings were originally obtained from The Faculty of Fisheries, Bogor Agricultural University, Bogor, Indonesia in 1988. These fish were reared at 23-25°C under 12L12D photoperiod until reaching maturity. Fish began to mature at an age of 9 months and conditions of maturity were maintained thereafter; but under the above stocking conditions, fish were observed not to undergo natural spawning [26].

Short-term experiment Twenty-four 2-year old matured male walking catfish (400-650 g in weight and 35-40 cm in total length) were selected from the stock, and then randomly assigned to three groups, injected with 0.2 TU HCG/g BW, 11U HCG/g BW and saline. Each group was kept separately in a 1X3X1m indoor concrete tank supplied with running fresh water of deep-well origin. Water temperature and photoperiod were maintained at 24+0.5°C, and 12L12D, respective- ly. Fish did not receive feed during the experi- ment.

Long-term experiment. Fourteen 16-month old matured male walking catfish (200-275 g in weight and 30-33 cm in total length) were divided into two groups. Eight fish received 1 IU HCG/g BW, whereas the remaining six fish received saline solution. Each group was kept separately in the same manner as in the short term monitoring experiment.

Treatment HCG was purchased from Teikoku Zoki Pharmaceutical Company, Japan. In order to meet the required dosage, the original hormone was diluted with 0.6% saline solution into 0.2 IU/ wl and 1 [U/l solution.

HCG treated groups received a single intra- muscular injection of HCG, whereas the saline

group received an equal volume of saline solution. Injection was done just below the beginning of the dorsal fin at 6 a.m.

Blood sampling Initial blood samples were taken from all fish before administering hormone solution or saline. Sampling was carried out at 6, 12, 18, 24, 36, 48 and 72hr in the short-term experiment. In the long term experiment, sam- pling was carried out 2, 4, 6, 8, and 10 days after HCG injection (at 10a.m.). Fish were sacrificed for the measurement of gonadosomatic index (GSI) at the termination of the experiments.

Approximately 0.8 ml of blood was drawn from the caudal vasculature with a heparinized syringe fitted with a 24-gauge, 1 inch needle after anesthe- tizing fish with 600 ppm of 2-phenoxyethanol (Wako, Japan). Blood samples were centrifuged at 3000 rpm, and plasma was stored in 1.5 ml polypropylene tubes at —20°C until analysis by radioimmunoassay (RIA).

RIA In this experiment, plasma testosterone, 11-KT, 17a-P, 17a,20a-P and 17a,208-P were de- termined by RIA. Details of RIA for each steroid have been described previously [1, 13, 25]. Valida- tion of the assay system for testosterone, 17a-P, 17a,20a-P and 17a,208-P has also been previously reported [25].

Validation of the 11-KT assay system for walk- ing catfish plasma was carried out in the same manner as has been reported previously [8]. In- traassay coefficients of variation at binding rates around 25%, 50% and 75% were 6.3, 4.0, and 4.5, respectively. Interassay coefficients of variation at binding rates around 25%, 50% and 75% were 18.6, 9.1, and 19.8, respectively.

EIA Plasma 208-S levels were measured using a specific enzyme immunoassay system (EIA). Validation for this EIA was reported previously [25].

Statistics The Student-t test, and the Duncan’s multiple range test and the Kruskal-Wallis test were used for statistical analysis.

Steroid Profiles in HCG-Injected Male Catfish 331

RESULTS

Short-term experiment

Changes in plasma testosterone levels following HCG injection are presented in Figure 1. Plasma testosterone levels in 0.2 IU HCG-treated group became elevated at 6hr (P<0.01; Ohr vs 6hr), reaching a peak at 18 hr (207.4 ng/ml), and there- after remained higher than initial levels (5.2 ng/ ml) until the end of the experiment (P<0.01; 0 hr vs 72hr). Testosterone levels in 11U HCG- treated group showed changes similar to those of the 0.2I1U HCG-treated group. No significant changes occurred in the control group.

225 | . — 11U HCG/g BW 200 Y, 0.2 1U HCG/g BW ——-— Control E @ 150 = Ww z fe} & 100 = a (e) = i] w 50 0 -—-e-—@-—-@-—_¢ ___§__¢_____¢____-----—-— © 0 6 12 18 24 36 48 72 HOURS AFTER TREATMENT Fic. 1. Plasma testosterone levels of male walking

catfish following HCG injection in short-term ex- periment. Each point represented as mean+SEM (n=8 for each treatment).

30 ,

—— 11U HCG/g BW i d omnee 0.2 1U HCG/g BW i aN ——— Control! nL

11-KETOTESTOSTERONE (ng/ml)

HOURS AFTER TREATMENT

Fic. 2. Plasma 11-KT levels of male walking catfish following HCG injection in short-term experiment. Each point represented as mean+SEM (n=8 for each treatment).

Changes in plasma 11-KT levels are shown in Figure 2. Plasma 11-KT levels in 0.2 IU HCG- treated group peaked at 6 hr (25.6 ng/ml; P<0.01; O hr vs 6 hr), and then decreased until 36 hr (7.6 ng/ml; P<0.01, 18hr vs 36hr), and thereafter increased again until the end of the experiment (28.6 ng/ml). Plasma 11-KT levels in the 11U HCG-treated group showed a similar pattern of changes. 11-KT levels in the control group were maintained at low levels during the experiment.

Changes in plasma 17a-P levels are presented in Figure 3. Plasma 17a-P levels increased from 36 hr until 72 hr in both 0.2 and 11U HCG injected groups (P<0.01; 24 hr vs 72 hr). Levels at 72 hr for 0.2 and 1 IU HCG/g BW were 13.6 ng/ml and 26.2 ng/ml, respectively. On the other hand, no significant changes were observed in the control

group.

27

—— 11U HCG/g BW +—-—- 0.2 IU HCG/g BW ——— Control

17a-P (ng/ml)

12 18 24 36 48 72 HOURS AFTER TREATMENT Fic. 3. Plasma 17a-P of male walking catfish following HCG injection in short-term experiment. Each point represented as mean+SEM (n=8 for each treatment).

0 6

— 11U HCG/g BW -—.—. 0.2 IU HCG/g BW ——— Control

20£8-S (ng/ml)

0 6 12 18 24 36 48 72 HOURS AFTER TREATMENT Fic. 4. Plasma 20-S levels of male walking catfish following HCG injection in short-term experiment. Each point represented as mean+SEM (n=8 for

each treatment).

332 M. ZAIRIN JR., K. ASAHINA et al.

Changes in plasma 20£-S levels are presented in Figure 4. Plasma 20-S levels in the 1 IU HCG/g BW-treated group started to increase at 24 hr (P< 0.01; 18 hr vs 24 hr), and reached 10.5 ng/ml at 72 hr. Plasma 206-S levels in the 0.2 TU HCG/g BW-treated group increased in the same manner as those of 1 IU HCG/g BW group, but levels for 48 and 72 hr were lower. Plasma 208-S levels in the control group slightly increased during 24—48 hr (P<0.01; Ohr vs 24hr), and then decreased (P<0.01; 48 hr vs 72 hr).

Changes in plasma 17a,20a-P levels are pre- sented in Figure 5. Injection of 1 or 0.2 IU HCG/g BW caused a parallel increase in 17a,20a-P levels, starting from 6hr (P<0.01; Ohr vs 6hr) con- tinuing until 72 hr (levels for 0.2 and 1 IU HCG/g BW treated groups were 1.7 and 3.0 ng/ml, re- spectively). Levels in the control group were always under the detectable limit throughout the experiment.

—— 11U HCG/g BW P= b 0.2 IU HCG/g BW ——-— Control

17a,20a-P (ng/ml)

0 6 12 18 24 36 48 72 HOURS AFTER TREATMENT

Fic.5. Plasma 17a,20a-P of male walking catfish fol- lowing HCG injection in short-term experiment. Each point represented as mean+SEM (n=8 for each treatment).

Changes in plasma 17a,208-P levels are pre- sented in Figure 6. Injection of 1 and 0.21U HCG/g BW caused a similar increase (P<0.01; 6 hr vs 72 hr) in 17a,20-P levels, starting from 6 hr. This trend continued until 72 hr (2.8 and 4.6 ng/ml for 0.2 and 1 IU HCG/g BW, respectively). Ex- cept for values at 24 and 36 hr, plasma levels were higher in the 1 IU HCG/g BW treated group than in the 0.2 IU treated group. Levels in the control group were always under the detectable limit.

—— 11U HCG/g BW —-—: 0.2 IU HCG/g BW 4 ——— Control

17a,208-P (ng/ml)

0 6 12 #18 24 36 48 72 HOURS AFTER TREATMENT

Fic. 6. Plasma 17a,208-P levels of male walking catfish following HCG injection in short-term experiment. Each point represented as mean+SEM (n=8 for each treatment).

GSI values at the end of the short-term experi- ment are presented in Figure 7a. GSI increased significantly from 0.15 % to 0.22% in 0.21U HCG/g BW treated group, and to 0.20% in 11U HCG/g BW treated group.

Long-term experiment

Changes in plasma testosterone, 11-KT, and 17a-P levels are presented in Figure 8.

Plasma testosterone levels increased sharply from 9.9 ng/ml to 77.6 ng/ml on day 2 (P<0.01; day 0 vs day 2), and then decreased but remained high until day 10 (34.1 ng/ml). Testosterone levels on day 10 were still significantly higher than initial and control levels. No significant changes occurred in the control group.

Plasma 11-KT levels decreased from 5.7 ng/ml

0.20 0.205 < 0.15 a 0.15 Fe g o 0.10 = 0.10 © a © a 0.05 0 asa 0 0 0.2 1 0 Y HCG (IU/g BW) HCG (IU/g BW) (a) (b) Fic. 7. Changes in GSI values following HCG injection

in short-term (a) and long-term (b) experiments. Each bar represented as mean+SEM (n=8 for each treatment in short-term experiment, and n=8 and 6 for HCG and control group, respectively in long- term experiment).

Steroid Profiles in HCG-Injected Male Catfish 333

1 IU HCG/g BW

-—--- Control

T (e), 11-KT (0), 17a-P (4) (ng/ml) » o

DAYS AFTER TREATMENT

Fic. 8. Plasma testosterone, 11-KT, and 17a-P levels following HCG injection in long-term experiment. Each point represented as mean+SEM (n=8 and 6 for HCG and control group, respectively). Solid line is HCG treatment, dotted line is control.

to 1.8 ng/ml on day 2 (P<0.01; day 0 vs day 2). Thereafter levels increased, peaking on day 8 (17.6 ng/ml; P<0.01; day 2 vs day 8) and decreasing on day 10. No significant changes occurred in the control group.

Plasma 17a-P levels peaked on day 2 (25.4 ng/ ml; P<0.01; day 0 vs day 2), and then decreased to initial levels on day 10 (2.2 ng/ml). No significant changes occurred in the control group.

Changes in plasma 208-S, 17a,20e-P, and 17a,20B-P levels are presented in Figure 9. Plasma 208-S levels started to increase on day 2 (7.4 ng/ ml), peaking on day 4 (9.1 ng/ml; P<0.01; day 0 vs day 4) and returning to initial levels on day 10.

—— 11U HCG/g BW Control

DAYS AFTER TREATMENT Plasma 28-S, 17a,20a-P, and 17a,20£-P levels following HCG injection in long-term experiment. Each point represented as mean+SEM (n=8 and 6 for HCG and control group, respectively). Solid line is HCG treatment, dotted line is control.

6 20f-S (A), 170,200-P (m), 170,20f-P (0) (ng/ml)

Fic.

Meanwhile, levels in the control group decreased gradually (P<0.01; day 0 vs day 10).

Plasma 17a,20a-P levels in the HCG-treated group peaked on day 2 (4.5 ng/ml; P<0.01; day 0 vs day 2) and then returned to basal levels on day 10. No significant changes were observed in the control group.

Plasma 17a,208-P levels in the HCG-treated group also peaked on day 2 (3.8 ng/ml; P<0.01; day 0 vs day 2), and remained high until day 6, returning to the initial levels on day 10. No significant changes occurred in the control group.

Changes in GSI values following HCG injection in the long-term experiment are presented in Fig- ure 7b. HCG injection caused significant increase in GSI from 0.14% to 0.18%.

Summary of the changes in levels of the six plasma steroids following HCG injection, both in short-term and long-term experiments are pre- sented in Figure 10. 1) HCG firstly induces an increase in plasma testosterone and 11-KT levels. 2) Thereafter, 11-KT levels decrease and progestin levels increase. 3) Finally, progestin levels de- crease, and 11-KT levels increase again.

STEROID LEVELS

DAYS

Fic. 10. Changes in levels of the six plasma steroids following HCG injection shown schematically for both the long- and short-term experiments. Lines represent trends, not absolute values. I: Androgen increase; II: Androgen decrease and progestin in- crease; and III: Progestin decrease and 11- ketotestosterone increase.

DISCUSSION

A single injection of HCG in mature male tropical walking catfish caused significant changes

334 M. ZArRIN JR., K. ASAHINA et al.

in all of the steroids monitored. HCG injection also caused enlargement of the testes.

HCG injection in both the 0.2 and 1 IU HCG- treated groups caused rapid increases in testoster- one levels peaking at 18hr. Thereafter, levels decreased but remained high. This suggests that the enzymes which are involved in testosterone synthesis are activated by HCG injection. In rainbow trout, gonadotropin has been suggested to stimulate the activity of two enzymes: C7_ 39 lyase which converts 17a-P to androstenedione, and 178-HSD which converts androstenedione to tes- tosterone [11, 15].

The function of testosterone in teleosts is still unclear, but is present in appreciable amounts during maturation in both males and females. Testosterone is regarded as an intermediate androgen product in the biosynthesis of 11-KT. However, based on replacement therapy in hypophysectomized catfish [17], it has been found that testosterone is able to maintain all stages of spermatogenesis, except spermatogonia mitosis. In addition, it has been shown also that testoster- one isobutyrate crystals induce in vitro completion of spermatogenesis in undeveloped testes of goldfish [18]. In the present experiment, HCG injection elevated levels of testosterone which were much higher than those of 11-KT. This is in contrast to eel, in which the testosterone peak was lower than that of 11-KT [16]. These results suggest that the function of testosterone in male walking catfish is not merely to serve as a precursor to 11-KT synthesis.

The fluctuation pattern of 11-KT following HCG- injection in the present experiment is parti- cularly interesting, since it showed two peaks. 11-KT levels peaked at 6hr, decreased to initial levels at 36 hr, and then peaked again on day 8. The first peak was perhaps induced by the stimula- tory effect of HCG on the conversion from testos- terone to 11-KT, which is also observed in other fishes. The decreases, as will be discussed later, may be caused by the inhibitory effect of the progestin, as the decrease coincided with the in- crease in progestin levels. The fact that second peak occurred when progestin levels had already decreased, and that its magnitude was smaller than that of the first peak, suggests that the increase in

the second peak was probably due to the elimina- tion of the inhibitory effects of progestin. On the other hand, injection of HCG in dab [8] and goldfish [14], or salmon gonadotropin in brown bullhead [19] did not induce increases in 11-KT. One possible explanation is that these fish are annual spawners in which the sensitivity of testes to HCG or gonadotropin varies with time. How- ever, under our stocking conditions, since plasma testosterone and 11-KT fluctuated at high levels throughout the year [26], male tropical walking catfish may possesses all developmental stages of the testes, so that fish can readily respond to HCG injection.

Plasma 17a-P levels increased slowly after HCG injection and peaked at 72 hr, then decreased and returned to initial levels on day 10. This indicates that HCG treatment stimulates the testes to pro- duce 17a-P. Increased 17a-P is converted into androstenedione by Cj7_29 lyase, and testosterone production is increased as a result. However, it is considered that when lyase activity has been used to its maximum capacity, the conversion to prog- estin formation occurs, resulting the decrease in androgen production and the increase in 17a,20a- P, 17a,208-P, and 208-S. This is in contrast to dab [8] in which HCG did not increase 17a-P levels.

One of the interesting findings in the present experiment is the simultaneous increase in the levels of three progestins, 208-S, 17a,20a-P and 17a,208-P, following HCG injection. Levels of 206-S were highest, followed by 17a,208-P and 17a,20a-P. The occurrence of more than one progestin for example, 17a,20a-P and 3,17,20a-P- 58 in dab [8], and 17a@,20a-P and 17a,208-P in carp [4] during spermiation in males has been reported. Progestin types seem to vary with species.

Little is known, about the function of 208-S in male fish. Our results show that HCG injection can elevate plasma 20(-S levels. Levels of this progestin were higher than those of 17a,20a-P, and 17a,206-P. The role of this hormone in the repro- ductive system of male walking catfish is still unknown. In HCG-injected female walking catfish as well, this hormone increases prior to ovulation [25].

17a,20a-P level increased due to HCG injection, and peaked at the day 2. High levels were main-

Steroid Profiles in HCG-Injected Male Catfish 335

tained until the day 8. This increase suggests the presence of 20a-HSD in the male walking catfish. In the other teleosts, activity of this enzyme has been detected in spermatozoa of carp [2], dab and plaice [7]. The real function of this hormone in male walking catfish is yet unknown. In carp, however, this hormone has been suggested to be a spermiation regulator [2] as this hormone posses- sed an inhibitory effect on androgen production as did its isomer [3]. The fact that 17a,20a-P is produced at high levels by mature spermatozoa suggest that it is released into the environment at the time of spawning, and could therefore play an important behavioral/pheromonal role [2].

In this investigation, HCG injection possibly enhanced the activity of 208-HSD which converts 17a-P to 17a,208-P production. In the short-term experiment, 17a,208-P levels started to increase after HCG injection, peaking at 72 hr. However, in the long-term experiment, this hormone peaked on day 2. Several roles have been proposed for 17a,208-P. Based on an in vitro study in carp, it has been suggested that this hormone inhibits androgen production [4]. The hormone was re- ported to induced spermiation in amago salmon, Oncorhynchus rhodurus [24], and change the K */ Nat ratio of rainbow trout seminal fluid [20]. Furthermore, pheromonal effects of this hormone have been reported to occur in the goldfish, Caras- sius auratus [21]. Recently, 17a,208-P has been suggested to be a steroidal mediator in the sper- miation of eel [16].

In the present experiment, HCG injection caused long-term changes in elevation of levels in all of the steroids monitored. These results were in contrast to those corresponding to females of the same species, in which HCG effects lasted only 24 hr [25]. The duration of steroid elevation follow- ing HCG injection in the present experiment is longer than the reported value during natural spawning or GtH treatment in other fishes. These differences may be due to the high dose and/or slow clearance of HCG in the male walking catfish. In goldfish, small doses of GtH (0.2 pg/g body weight) are cleared within 1 hr [10]. This has also been shown in Ictalurus nebulosus [19] injected with sGtH 5 to 500 ng/g body weight. The reason may be that the dosage used here is relatively high

and that HCG is not an endogenous GtH.

In the short-term experiment, androgen levels were higher in the low dose group, whereas prog- estin levels were higher in high dose group. Lower levels in androgen and higher levels in progestin in the high dose group were probably caused by the conversion shift from androgen to progestin, since it is reported that high dose of gonadotropin induces the shift in salmonid [15]. Inhibition of androgen synthesis by increased progestin, which was found in carp [3], may also decrease androgen levels in catfish.

In summary, HCG injection increases testoster- one production which is further converted to 11- KT. As a result, 11-KT production increases several hours after injection. Meanwhile, ster- oidogenic pathways favoring progestin production gradually become active. Increased progestin pro- duction possesses an inhibitory effect on 11-KT production. At this point, GSI increases and probably more sperm is produced. Consequently, 11-KT production decreases during progestin peaks, and increases again when progestin levels decrease.

ACKNOWLEDGMENTS

We express our thanks to colleagues in the Fisheries Laboratory, The University of Tokyo, Maisaka, Shi- zuoka Prefecture,for allowing use of facilities for field experiments. We thank Marcy N. Wilder for correcting the English. This study was partially funded by a Grant-in Aid for Scientific Research from the Ministry of Education, Science and Culture.

REFERENCES

1 Aida K, Kato T, Awaji M (1984) Effects of castra- tion on the smoltification of precocious male masu salmon Oncorhynchus masou. Bull Japan Soc Sci Fish 50: 565-571

2 Asahina K, Barry TP, Aida K, Fusetani N, Hanyu I (1990) Biosynthesis of 17a,20a-dihydroxy-4-pre- gnen-3-one from 17a-hydroxyprogesterone by sper- matozoa of the common carp, Cyprinus carpio. J Exp Zool 255: 244-249

3 Barry TP, Aida K, Hanyu I (1989) Effect of 17a,208-dihydroxy-4-pregnen-3-one on the in vitro production of 11-ketotestosterone by testicular frag- ments of the common carp, Cyprinus carpio. J Exp Zool 251: 117-120

12

13

15

336

Barry TP, Aida K, Okumura T, Hanyu I (1990) The shift from C-19 and C-21 steroid synthesis in spawn- ing male common carp, Cyprinus carpio, is reg- ulated by the inhibition of androgen production by progestogen produced by spermatozoa. Biol Reprod 43: 105-112

Barry TP, Santos AJG, Furukawa K, Aida K, Hanyu I (1990) Steroids profiles during spawning in male common carp. Gen Comp Endocrinol 80: 223- 231

Baynes SM, Scott AP (1985) Seasonal variations in parameters of milt production and in plasma concen- tration of sex steroids of male rainbow trout (Salmo gairdneri). Gen Comp Endocrinol 57: 150-160 Canario AVM, Scott AP (1989) Synthesis of 20a- hydroxylated steroids by ovaries of the dab (Liman- da limanda). Gen Comp Endocrinol 76: 147-158 Canario AVM, Scott AP (1991) Levels of 17a,20a- dihydroxy-4-pregnen-3-one, and 3/,17a,20a-trihy- droxy-5f-pregnane, and other sex steroids, in blood plasma of male dab, Limanda limanda (marine flatfish) injected with human chorionic gonadotro- pin. Gen Comp Endocrinol 83: 258-264

Colombo L, Belvedere PC, Simontacchi C, Lazzari M (1987) Shift from androgen to progesterone biosynthesis by the testes of the northern pike, Esox lucius L., during transition from spermatogenesis to spermiation. Gen Comp Endocrinol 66: 18 [Abstract 48]

Cook AF, Peter RE (1980) Plasma clearance of gonadotropin in goldfish, Carassius auratus, during the annual reproductive cycle. Gen Comp Endocri- nol 42: 76-90

Depeche J, Sire O (1982) In vitro metabolism of progesterone in the testis of the rainbow trout, Salmo gairdneri Rich., at different stages of sperma- togenesis. Reprod Nutr Develop 22: 427-438

Hunt SMV, Simpson TH, Wright RS (1982) Season- al changes in the levels of 11-oxotestosterone and testosterone in the serum of male salmon, Salmo salar L., and their relationship to growth and maturation cycle. J Fish Biol 20: 105-119 Kobayashi M, Aida K, Hanyu I (1985) Radioimmu- noassay for silvercarp gonadotropin. Bull Japan Soc Sci Fish 51: 1085-1091

Kobayashi M, Aida K, Hanyu I (1986) Effects of HCG on milt amount and plasma levels of steroid hormones in male goldfish. Bull Jap Soc Sci Fish 52: 755

Le Gac F, Loir M (1988) Control of testis function in fish: in vitro studies of gonadotropic regulation in the trout (Salmo gairdneri). Reprod Nutr Develop

16

17

18

19

20

21

22

23

24

25

26

M. ZaArRIN JR., K. ASAHINA et al.

28: 1031-1046

Miura T, Yamauchi K, Takahashi H, Nagahama Y (1991) Involvement of steroid hormones in gonado- tropin-induced testicular maturation in male Japanese eel (Anguilla japonica). Biomed Res. 12: 241-248

Nayyar SK, Keshavanath P, Sundararaj BI, Donald- son EM (1976) Maintenance of spermatogenesis and seminal vesicles in hypophysectomized catfish, Heteropneustes fossili (Bloch): effects of ovine and salmon gonadotropin and testosterone. Can J Zool 54: 285-292

Remacle C (1976) Actions hormonales sur les cellules germinales males de Carassius auratus L. en culture oranotypique. Gen Comp Endocrinol 29: 480-491

Rosenblum PM, Callard IP (1987) Response of male brown bullhead catfish, Jctalurus nebulosus Lesueur, to gonadotropin-releasing hormone and gonadotropin. J Exp Zool 243: 189-199

Scott AP, Baynes SM (1982) Plasma levels of sex steroids in relation to ovulation and spermiation in rainbow trout (Salmo gairdneri). Proc Second Int Symp on the Reprod Physiol of Fish, Wageningen, The Netherlands, pp 103-106

Stacey NE, Sorensen PW, Dulka JG, Van der Kraak GJ, Hara TJ (1987) Teleost sex pheromone: Recent studies on identity and function. Proc Third Int Symp on the Reprod Physiol of Fish, St John’s, Newfoundland, pp 150-153

Trant JM, Thomas P (1989a) Evidence that 17a,208,21-trihydroxy-4-pregnen-3-one is a matura- tion inducing steroid in spotted seatrout. Fish Phy- siol Biochem 7: 185-191

Trant JM, Thomas P (1989b) Isolation of a novel maturation-inducing steroid produced in vitro by ovaries of Atlantic croaker. Gen Comp Endocrinol 75: 397-404

Ueda H, Young G, Crim LW, Kambegawa A, and Nagahama Y (1983) 17a,208-dihydroxy-4-pregnen- 3-one: Plasma levels during sexual maturation and in vitro production by the testes of amago salmon (Oncorhynchus rhodurus) and rainbow trout (Salmo gairdneri). Gen Comp Endocrinol 51: 106-112 Zairin M Jr; Asahina K, Furukawa K, Aida K (1992) Plasma steroid hormone profiles during HCG induced ovulation in female walking catfish Clarias batrachus. Zool Sci 9: 607-617

Zairin M Jr (1993) Endocrinological studies on maturation and spawning in the walking Catfish, Clarias batrachus. Doctoral thesis, The University of Tokyo, pp 1-90

ZOOLOGICAL SCIENCE 10: 337-351 (1993)

Annual Patterns of Testicular Development and Activity in the Chinese Bullfrog (Rana rugulosa Wiegmann)

Yuno-Hs1 Kao!”, Pau S. ALEXANDER’, VIVIAN V. CHENG YANG

and JoHN YuH-LiIn Yu!*

‘Endocrinology Laboratory, Institute of Zoology, Academia Sinica, Nankang, Taipei, Taiwan 115, R.O.C. and 7Department of Biology, Tunghai University, Taichung, Taiwan 407, R.O.C.

ABSTRACT— Annual patterns of testicular development and of testicular and plasma androgen levels in the Chinese bullfrog, Rana rugulosa Wiegmann, are reported. The animals were collected from a frog culture farm from November of 1986 through November of 1987, at monthly intervals or more frequently during the breeding season. Proliferation of primary spermatogonia was observed in November-December. In January-February, the number of primary spermatogonia decreased while secondary spermatogonia increased. Primary and secondary spermatocytes showed little changes until February when their number rose sharply. Although every stage of spermatogenesis was present year around, spermatozoa were massively formed in April-June. Thus, the spermatogenetic cycle of R. rugulosa, inhabiting subtropical Taiwan, is of continuous type. Leydig cells began to proliferate in late March and were abundant from April through July. Testicular androgen content showed basal values in January-February and August-December, began to increase in late March, peaked in April, and dropped markedly in May. The pattern of plasma androgen levels was similar to that of testicular androgen content, except peaking in middle April-early May and dropped in late June. The unimodal pattern of seasonal changes in plasma androgen levels parallels Leydig cell proliferation, spermatogene- sis, testicular androgen content, and reproductive activities. In addition, the testicular weight was negatively correlated with liver weight, but not with fat body weight during an annual reproductive cycle

© 1993 Zoological Society of Japan

of R. rugulosa.

INTRODUCTION

The patterns of reproductive cycle have been investigated in many species of anurans. The annual spermatogenetic pattern in anurans varies from discontinuous spermatogenesis in the temperate zone to continuous spermatogenesis in tropical species [22]. Earlier studies have shown that ranids, one of the major groups of anurans, display a diverse type of spermatogenesis: such as Rana temporaria, R. arvalis, and R. dalmatina exhibiting the discontinuous spermatogenesis; R. tigerina, R. esculenta, and R. perezi, the potencial-

Accepted December 3, 1992 Received August 18, 1992

> Reprint requests should be sent to: Dr. J.Y.L. Yu, Endocrinology Laboratory, Institute of Zoology, Academia Sinica, Nankang, Taipei, Taiwan 115, R.O.C.

ly continuous spermatogenesis; and R. hexa- dactyla, R. cyanophlyctis, and R. catesbeiana, the continuous spermatogenesis [6, 22, 43, 44, 58]. The secretion patterns of the pituitary gonado- tropins in relation to testicular development and activity during a reproductive cycle of anurans have been relatively less studied [13, 19]. It was demonstrated that the circulating luteinizing hor- mone (LH) and _ follicle-stimulating hormone (FSH) were highly correlated, respectively, with the changes in plasma androgen levels and in testicular weight during an annual cycle of male toad, Bufo japonicus [13]. On the other hand, seasonal changes in testicular and/or circulating androgen levels in relation to spermatogenetic and/or reproductive activity have been established in several anurans, such as R. esculenta [7, 38, 43, 56], R. perezi [6], R. catesbeiana [19, 58], B. japonicus [13], B. mauritanicus [48], Pachymedusa

338 Y. Kao, P. S. ALEXANDER, et al.

dacnicolor |41], R. ridibunda [55], and Rana occi- pitalis [17]. However, the annual patterns of circulating androgen levels in anurans exhibit a great variation of species specificity.

The Indian bullfrog (R. tigerina Daudin) was reported to show a potentially continuous sperma- togenetic cycle [1, 45]. Correlations of the androgen levels in testis and in blood with sperma- togenesis and reproductive activity during an annual cycle of this species were, however, not investigated.

The Chinese bullfrog, R. rugulosa, is phylogeni- cally closely related to the Indian bullfrog, R. tigerina [9, 20, 39]. Little information has been available in male R. rugulosa, inhabiting subtropi- cal Taiwan, and central and southern China, with respect to its patterns of reproductive cycle: sper- matogenetic cycle, androgen levels, breeding activity, secondary sexual characters, and changes in various organs in association with reproduction, such as liver and fat body. We have conducted a series of studies investigating the annual reproduc- tive patterns of R. rugulosa in both males and females. We report here the annual patterns of male R. rugulosa with respect to: 1) the changes in the sizes of body, testis, liver, and fat body; 2) the spermatogenetic cycle, and the changes in the proliferation of Leydig cells; 3) the androgen levels in testis and plasma; and 4) the correlation of plasma and testicular androgen levels with sperma- togenesis and other reproductive activity.

MATERIALS AND METHODS

Study site

Rana rugulosa were collected from a frog culture farm located in Tungshih (23°40'N, 120°15’E), Yunlin County, Central Taiwan. Briefly, the frogs were raised in an open pond which was surrounded with fishing nets. They were fed with “frog food” manufactured by Fu-Shou Food Co., Taichung, and the natural preys as well. During hibernation, the frogs were dormant under digged earth caves or the hayricks of rice and woody box.

Chorussing of male frogs was heard from March through July at the frog culture farm, and was occasionally heard in August and September.

Young tadpoles with external gills were first observed in late May, and metamorphosing frog- lets were first observed in early June.

Data on rainfall, temperature, and photoperiod at Tungshih, are summarized in Fig. 1. The period between March and September, 1987 was the rainy season. The lowest monthly mean air temperature was 15.4°C in January and the highest was 28.5°C in August. Monthly minimum air temperature averaged about 5°C in January and February. The photoperiod was shortest in December and longest in June. Shortest and longest days had about three hours difference in photoperiod.

Collection of blood and tissues

Monthly samples of 12—25 sexually mature male frogs (hatched in April, 1986) were collected from November, 1986 through November, 1987, except for sampling at weekly or biweekly intervals during the breeding season from March through early July. Frogs, after collection, were imediately placed into a wet fishing net and were anesthetized with ether. Blood (0.8-2.0 ml) was drawn from the conus arteriosus with a 2-ml heparinized syringe, and centrifuged at 1000 G for 10 min at 8°C; and plasma (0.4-1.0 ml) was collected and stored at —20°C until androgen assay. Both right and left testes of each frog were removed and their lengths and weights were measured. One testis was fixed in Bouin’s fluid for histological observa- tions, and the other was stored at —20°C for androgen assay. Liver and fat bodies were re- moved and weighed as well.

Histological examination

Bouin’s-fixed testes were wax-embedded, sec- tioned at 7 um, and stained with Harris’ hemato- xylin and eosin. We determined spermatogenetic cycle by two phases: spermatogenesis, and sper- miogenesis. The stages of spermatogenesis were identified according to the method used by Rastogi et al. [43]: stage 1, primary spermatogonia (ISPG); stage 2, secondary spermatogonia (IISPG); stage 3, primary spermatocytes (ISPC); stage 4, secon- dary spermatocytes (IISPC); and stage 5, sperma- tids (SPT). Spermiogenesis was recorded by the extent of spermatozoa present in the seminiferous Both sper-

tubules. spermatogenesis and

Annual Testicular Activity in Bullfrog 339

300

250

200

150

c Rainfall (mm)

100

50

1986

Fic. 1. Taiwan.

miogenesis of each animal were determined from twenty cross-sectioned seminiferous tubules repre- senting four different cross-sections of each testis. The proliferation of Leydig cells in the interstitial tissues was recorded.

Radioimmunoassay of androgen in plasma and testis

Androgen concentrations in blood plasma and testis were measured by the radioimmunoassay described previously by Chen et al. [4]. The steroids in samples were extracted with diethyl ether. The average recoveries of the added ster- oids were 70.2% and 83.3% for plasma and tes- ticular androgen, respectively, and individuals re- coveries were run with each sample. The cross reactivities of the antiserum with various

40

e—e Air temperature (°C)

o---o Day length (hr)

Months Monthly cahgnes in the rainfall, photoperiod (day length), and air temperature in Tungshih, Yunlin County,

androgens, in relative to testosterone (100%), were: 5a-dihydrotestosterone (74%), andro- stenedione (1.23%), and androstenediol (0.59%) [59]. Thus, the data are expressed as “androgen” which mainly representing testosterone and dihy- drotestosterone.

The coefficient of variation (CV) of the interas- say was 11.0% (N=5) and that of intraassay was 6% (N=6) for plasma androgen. The CV of interassay was 9.2% (N=8) and that of intraassay was 2.4% (N=6) for testicular androgen.

Statistical analysis

Analysis of variance (ANOVA) and Duncan’s multiple range tests were used to examine differ- ences among the numerous means. Linear correla- tions were performed for all variables. Probability

340

levels of 0.05 and 0.01 were used to indicate significance in comparison of means and correla- tions [49].

RESULTS

Annual changes in the sizes of body, testis, liver and fat body

Changes in body weight and body length are shown in Table 1. During the annual reproductive cycle of R. rugulosa, body weights ranged from 30-82 g; and body lengths from 70-92 mm. The body weights were greater during May—-November

TABLE 1. cycle of R. rugulosa

Y. Kao, P. S. ALEXANDER, et al.

than those during December—April. The weight of paired testes increased during Dcember—Febru- ary, peaked in March, sharply declined in April May, and remained at low levels in June—October (Table 1). Differential changes in the weights and lengths of left and right testes were observed (Fig. 2). The mean weight of the left testis was 20% higher than that of the right testis; while, the mean length of the left testis was 40% higher than that of the right testis. The shape of left testis was less uniform than that of right testis. In addition, the annual changes in lengths of both testes were not as great as those of testicular weight although the lengths of both testes rose in February—April.

Changes in the sizes of body, testis, liver, and fat body during an annual reproductive

Body* Organ weights* Months’ N? length weight testes liver fat bodies (cm) (g) (mg) (g) (g) 1986 Nov 21 21 78.5+1.0 46.9+1.4 86.4+5.9 1.33+0.06 2.40+0.20 Dec 28 12 80.1+1.4 40.9+2.0 88.5+8.7 0.89 +0.05 1.63+0.18 1987 Jan 23 20 78.8+0.9 42.5+1.3 HN Aha D2 1.05 +0.06 1.38+0.15 Feb 26 20 80.4+0.7 43.0+1.3 132.8+10 0.97+0.05 1.52+0.14 Mar 12 16 80.2+£0.8 42.7+1.0 38) 229) 7/ 0.78 +0.03 1.10+0.17 26 16 82.8+0.6 47.9+1.2 131.3+4.8 0.85 +0.03 1.28+0.16 Apr 2 15 79.2+0.6 38.5+0.9 115.7+8.2 0.66 +0.02 0.82+0.11 9 19 79.2+0.6 389), 22(0).8) 107.6+4.8 0.59+0.03 0.61+0.09 16 14 79.4+0.7 39.4+1.2 122.1+10 0.62 +0.03 0.51+0.01 23 17 77.6+0.7 Mo EDLY 102.5+5.3 0.62 £0.03 0.34+0.10 30 16 77.5+0.7 36.2+1.0 110.6+5.2 0.58+0.03 0.29 +0.06 May 7 11 78.3+1.0 3H) ae IS) 109.8+6.5 0.65 +0.04 0.29 +0.06 21 17 80.5+1.3 44.8+2.0 85.0+4.6 0.88 +0.06 0.14+0.07 Jun 4 19 79.9+0.6 49.2+1.5 WAL Sse D7) 1.36+0.09 0.14+0.03 28 21 81.3+0.7 3)9),Il ae Il) 72.6+4.2 2.07 +0.14 0.19+0.03 Jul 29 25 81.3+0.8 43.8+1.2 79.4+7.1 1.28+0.07 1.07+0.22 Aug 29 22 84.8+0.7 56.6+2.5 60.8+6.3 3.14+0.28 1.35+0.17 Sep 24 17 81.6+0.9 Sasi? 65.0+7.4 3.40 +0.28 2.96+0.25 Oct 30 12 80.6+0.6 Siloilse tS) 67.3+9.4 3.00 +0.27 2.42+0.14 Nov 23 17 81.3+0.7 50.7 +2.0 99.0+6.4 1.68+0.11 2.39 +0.26

' The frogs were collected monthly, or more frequently during the active breeding season from march

through June. * Number of animals * The data are expressed as the means+SE.

Annual Testicular Activity in Bullfrog 341

A—“ left 14

12

Testicular length (mm) © =

90

70

50

30

Testicular weight (mg)

4—aA Right

10 wrals=aleralaciredinaltylaamicalc cee Tcl

Months

Fic. 2. Changes in the weights and lengths of the left and right testes during an annual reproductive cycle of R. rugulosa. The data are expressed as the means+SE, and the number of animlas as indicated in Table 1.

Annual pattern in the weights of liver and fat body are shown in Table 1. The liver and fat body weights decreased in November—February, reached the minimal sizes in early May and late June, repsectively, and increase in early June and late July, repsectively. In September, both organs attained the maximal sizes, and then began to decrease in late October. Highly significant cor- relation was observed between liver and fat body weights (r=0.60, p<0.01). As shown in Fig. 3, an inverse relationship existed between the hepato- somatic index and the gonadosomatic index (r= —(0.828, p<0.01). The correlation between fat body-somatic index and the gonadosomatic index was, however, not significant.

Testicular development and spermatogenesis

Proliferation of Leydig cells changed with sea- son (Fig.4). In November—December and January-February, Leydig cells were sparse and difficult to distinguish from other connective tissue (Fig. 4-A). The Leydig cells began to proliferate in late March and were abundant in the broad intertubule space in April—July (Fig. 4-B) after which they rapidly decreasd between the decreas- ing intertubular space (Fig. 4-C).

Seasonal changes in histological observations of the spermatogenetic cycle in the testis of R. tiger- ina rugulosa are also representatively shown in Fig. 4. The spermatogenetic cycle showed every

342 Y. Kao, P. S. ALEXANDER, et al.

HSI x 10 -2

> (op) {ee}

FBSI x 10 -2 ine)

GSI x 10 -3 no oo RO

—

0 aN D JteIula miglyla slo nN | 1986 1987 Months

Fic. 3. The relations of liver and fat body weights to the testicular weights during an annual reproductive cycle of R.

rugulosa. The data are expressed as mean+SE, and the number of animals as indicated in Table 1. GSI, testes weight/body weight; HSI, liver weight/body weight; FBSI, fat body weight/body weight.

Fic.

4. Seasonal changes in the spermatogenetic cycle and the proliferation of Leydig cells of R. rugulosa. a, the nest of primary spermatogonia; b, the nest of secondary spermatogonia; ¢c, the nest of primary spermatocytes; d, the nest of secondary spermatocytes; e, the nest of spermatids; f, spermatozoa; g, Leydig cells. A. Hibernation (December). Primary and secondary spermatogonia were the most numerous spermatogenetic cells in this period. A few clusters of spermatozoa were produced. Leydig cells existed in sparse between the seminiferous tubules. B. Breeding Season (May). Massive spermatozoa were formed in this period. Leydig cells were abundant in the intertubular space. C. Postbreeding Season (August). Cell nests of all stages of spermatogenetic cycle were existent. Few spermatozoa occurred in this period. Leydig cells were greatly reduced in number in the intertubules.

343

Annual Testicular Activity in Bullfrog

ses

SONY gay

]

eae aa

ras es WHO es 2° : tas te

4. ee hy

hay "

3 fe be Ss . . Cs

344 Y. Kao, P. S. ALEXANDER, et al.

stage of spermatogenesis year around. Only a few clusters of spermatozoa were found in the semi- niferous tubules during hibernation (November-— February) (Fig. 4-A). Massive formation of sper- matozoa occurred during the active breeding sea- son mainly in April-June (Fig. 4-B). Spermatozoa were greatly reduced in number during postbreed-

2 B

0

Ze O ae Yn | =|

I SPC x 2

Il SPG x 2 nN

I SPG x 2 (S) —- >) Ww

ND J FM 1986

ing in August—October (Fig. 4-C).

Significant annual changes occured in the sper- matogenetic cycle (ANOVA, p<0.01) based on the estimate of the number of various stages of cell nests from primary spermatogonia to spermatids (Fig. 5). In November—December of 1986, ISPG were most numerous, accounting for 48% of the

1987

Months

Fic. 5. rugulosa.

Quantitative changes in different stages of spermatogenesis during an annual reproductive cycle of R. Values are expressed as the average number of cell nests of various spermatogenetic stages per

seminiferous tubule. ISPG and IISPG, primary and secondary spermatogonia; ISPC and IISPC, primary and

secondary spermatocytes; SPT, spermatids.

Annual Testicular Activity in Bullfrog 345

hibernation breeding [ES] early breeding (Ml] post-breeding

5 12 12 4 4 Ze) ro) pes) N ey) q Sep A/h3 q ) et) fe) as S 2 q 9

5 3 od 12 rf 12 i

0

20

Plasma androgen (ng/ml) S)

1986 1987

Months Fic. 6. The unimodal patterns of changes in testicular androgen content and plasma androgen levels during an annual reproductive cycle of R. rugulosa. The annual reproductive cycle was divided into four periods: hibernation, early breeding, breeding, and post-breeding. The data are expressed as the means+SE, and the number of animlas as indicated with Arabic numerals.

346 Y. Kao, P. S. ALEXANDER, et al.

total germinal cell nests; other stages of germ cells began to multiply in December. In January and February, the number of ISPG and SPT declined, but those of IISPG, ISPC, and IISPC increased with the rise in ISPC and IISPC lagging behind that of IISPG. In March there were increases in the ISPG and IISPG followed by sharp decreases in April. ISPG then exhibited a slight increase with some fluctuations in May—August; it de- creased again in September, followed by sharp increases in October and November. IISPG re- mained virtually constant from May through September, and then increased in October. ISPC decreased in March—May, and increased again in June-July; by contrast, a reverse pattern existed for IISPC in March—July. At this period (March— July), the cell nests of SPC (the sum of ISPC and IISPC) became the most numerous. During the breeding season (March—July), spermatozoa were massively formed from SPT. ISPC, IISPC, SPT, and spermatozoa all reduced in August, and in- creased in September—October.

Plasma and testicular androgen levels

Plasma and testicular androgen levels showed significant seasonal changes (ANOVA, p<0.05, Fig. 6). Plasma androgen levels were very low

(<1ng/ml) during hibernation (November- February), increased in late March, peaked in early May (14.50+1.05 ng/ml), decreased sharply in late May and June, and reached basal levels afterwards (from July through October). Testicu- lar androgen content was also low during hiberna- tion period (<1 ng/testis), increased rapidly in early April (3.28+0.28 ng/testis) and peaked in middle (3.88+0.45 ng/testis) and late April (3.90 +0.44 ng/testis); the androgen levels were de- creased markedly in May, followed by a gradual drop afterwards, and reached basal values in September (0.7+0.07 ng/testis). | During the annual reproductive cycle of R. rugulosa, the difference between maximal and minimal values in testicular androgen content was 10 fold, while that in plasma androgen was S50 fold.

Correlations of plasma androgen concentration with testicular development, liver, and fat body are summarized in Table 2. Highly significant were observed between plasma androgen levels and testicular androgen content. There were significant negative correlations of plasma and testicular androgen levels with liver weight and fat body weight respectively, as well as with two stages of spermatogonia (the number of nests of primary spermatogonia, ISPG, and of

correlations

TABLE 2. Correlation coefficients (r) of plasma and testicular androgen levels with sperma- togenetic cell types, and liver and fat body weights of male R. rugulosa during an annual reproductive cycle

Plasma androgen Testicular androgen

Testicular androgen 0.734** 1.0 Testicular weight 0.087 0.188 GSI 0.288 0.426 Liver weight —0.438" —0.460* HSI —0.448° —0.466" Fat body weight —0.702** —0.696** FBSI =), 719" —0.707** Primary spermatogonia —0.515" —0.628* Secondary spermatogonia —0.499* —0.403 Primary spermatocytes 0.025 0.139 Secondary spermatocytes 0.389 0.141 Spermatids —0.074 —0.019

Correlation coefficients (r) were analyzed from the monthly means of those parameters. GSI, gonadosomatic index=testis weight/body weight; FBSI, fat body weight/body weight; HSI, liver weight/body weight; ', p<0.10; *, p<0.05; **, p<0.01.

Annual Testicular Activity in Bullfrog 347

secondary spermatogonia, IISPG). The rising phases of both plasma and testicular androgen levels were synchronized with active sper- miogenesis (Figs. 4-B and 6).

DISCUSSION

Annual patterns of testis, liver, and fat body

A unimodal pattern of annual variation exists in the testicular weight of R. rugulosa. There are significant differences between the maximal and minimal mean values of both testicular weight (2 fold) and GSI (3 fold) during the annual cycle. Such annual changes in GSI of this species are similar to those of R. tigerina [44], B. marinus [44], R. esculenta (21, 43], and B. japonicus [13, 30], but are different from those of R. perezi [6] and R. catesbeiana [19, 58], showing little variation of GSI. The annual change in the testicular weight of R. rugulosa also shows a positive correlation with secondary spermatogonia (r=0.625; p<0.05). Thus, the rising phase of testicular weight in January—March is presumably due to the increases in the secondary spermatogonia even though other factors, such as water reabsorption and cellular hypertrophy, may be involved. On the other hand, the observations that the testicular weight of R. rugulosa exhibited some fluctuations during the entire breeding season (March—July) and sharp decreases in the late breeding season (June—early July) may be attributed to the evacuation of sper- matozoa. We observed that both lengths and weights of left testis were greater than those of right testis during an annual cycle of R. rugulosa. Such findings are similar to those of B. bufo and R. nigromaculata reported by other investigators [53].

In male anurans, the roles of liver in the repro- ductive cycle include the formation of sex steroid- binding proteins, the formation and storage of lipids and carbohydrates for the metabolic activity of the gonads, the degradation of gonadal steroid hormones and others [23, 26, 29, 35, 47]. we demonstrated that a unimodal pattern of annual change in liver weight exists in male R. rugulosa. This finding is similar to those of A. crepitans [23], B. woodhousei [23], B. canorus [29], and R. esculenta [47], but is different from that of R.

nigromaculata [27], a temperate species, which showing no significant difference in the liver weight during the annual cycle. We also found that the annual change in liver weight of R. rugulosa positively correlates with that of the body weight (r =0.877, p<0.01), but negatively with that of the testicular weight. Such findings are in disagree- ment with those reported for R. ridibunda [24], which displaying high positive correlations among the weights of body, liver, and testes.

The functions of fat body on reproduction in male anurans are likely the supply of metabolic energy for the maintenance of testicular activity during hibernation, and for androgen production during the breeding season, as demonstrated in R. esculenta [28, 43], and in Acris crepitans and B. woodhousei [23]. Male R. rugulosa, as observed in the present study, exhibits a unimodal pattern of annual change in the fat body weight, which being positively correlated with liver weight, but not with the body weight (r=0.421, p>0.05). The annual change in the fat body weight is not correlated with that in the testicular weight, but is negatively correlated with androgen levels in testis and plas- ma (Table 2). These results are similar to those of R. perezi [6] and B. japonicus [30] thus support the proposals that the fat bodies serve as a major lipid nutrient for metabolic energy in general and in the testis in particular [5, 8].

Spermatogenetic cycle

The annual spermatogenetic cycles in anurans have been classified into three types: discon- tinuous, potentially continuous, and continuous [22]. Discontinuous spermatogenesis is commonly found in temperate and cold zone species such as R. temporaria [22]; while continuous sperma- togenesis 1s characteristic of the spermatogenetic cycle of tropical and subtropical anurans [22]. Some anuran species such as R. esculenta and R. tigerina are classified as “potentially continuous cycle” where the primary spermatogonia never become refractory to gonadotropin and continue to undergo multiplication until the lowering temperature of autumn and winter causes a re- tardation [1, 22, 34, 44, 45]. In the present study, we observed that R. rugulosa inhabiting subtropic- al Taiwan exhibits a continuous spermatogenetic

348 Y. Kao, P. S. ALEXANDER, et al.

cycle with germ cells of all stages found throughout the year, although massive spermatozoa being existed in April—May.

Androgen patterns

Annual changes in plasma androgen levels have been studied in many species of male anurans: seven temperate species, R. perezi [6], R. cates- beiana [19], B. japonicus [13], R. esculenta [7], B. mauritanicus [48], R. ridibunda [55], and R. nigro- maculata [51]; three subtropical species, P. dacni- color [41] and B. bufo gargarizans and B. melanos- tictus [12]; a tropical species, D. occipitalis [17]. These male anurans exhibit either unimodal or bimodal types of annual patterns in plasma androgen characterized by a single peak or two peaks, respectively, of annual circulating androgen levels [13]. Studies have demonstrated that sub- tropical male anurans show an annual unimodal pattern of plasma androgen levels [12, 41], while temperate male anurans show both unimodal and bimodal patterns of androgen levels [6, 7, 13, 19, 48, 51, 55]. However, D. occipitalis [17], a tropical anura, displays a bimodal pattern of annual plasma androgen level. We have demonstrated in the present study that male R. rugulosa, inhabiting subtropical Taiwan, displays a unimodal pattern of circulating androgen levels during an annual repro- ductive cycle. This pattern is comparable with those of subtropical anurans reported by other investigators [12, 41].

In vitro studies of testicular androgen biosynthe- sis in Amphibia indicate that dihydrotestosterone (DHT) is the major androgen in Anura and that testosterone (T) is the major androgen in Urodele [15]. Such a conclusion is supported by in vivo investigations of the circulating androgens in Anurans, R. catesbeiana [19], R. nigromaculata [51], R. pipiens [57], Eleutherodactylus coqui [54], and in Urodeles, Necturus maculosus [2], Salaman- dra salamandra [18], Pleurodeles waltl [10], and Ambystoma tigrinum [33]. However, in several anuran species, such as B. japonicus [13], R. esculenta [38], B. mauritanicus [48], and P. dacni- color [41], T is more predominant than DHT in circulation. Whether T or DHT is the predomi- nant androgen in the male Chinese bullfrog, R. rugulosa, requires further investigations. In mam-

mals, DTH is identified to be primarily a metabo- lite of T in peripheral target tissues. DTH is considered to be directly formed by testis of anu- ran amphibians [11, 15, 16, 31, 32, 36, 37], while it has not been identified in the testis of urodel amphibians [15, 25, 29, 50, 52]. The physiological role of DHT in amphibians is not yet clearly defined, and thus needs further studies.

Androgen secretion is regulated by the interac- tions of hormonal and environmental factors [43, 44]. Rastogi et al. [43] reported that androgen in R. esculenta shows a negative feedback interaction with gonadotropins by a study of hypophysectomic implantation with homoplastic pituitary extract; on the other hand, in R. catesbeiana, plasma gonadot- ropins and steroids are simultaneously elevated throughout most of the reproductive cycle, as well as LH levels exhibit a weak correlation with androgen levels [19]. In B. japonicus, Itoh et al. [13] reported that plasma LH levels, but not FSH levels, were highly correlated with circulating androgen levels.

Correlation of androgen patterns with spermato- genetic cycle

The hormonal control of spermatogenesis in anurans has been investigated [6, 40, 41, 43, 44, 58]. It was demonstrated that FSH stimulates the spermatogenesis on the basis of the studies with mammalian hormones [40, 44]. On the other hand, the controversial results of the effect of androgen on spermatogenesis have been reported in anurans: such as in R. hexadactyla, R. esculenta and R. tigerina, where androgen exhibiting sup- pression of spermatogenesis (mainly at the se- condary spermatogonia stage); and in B. fowleri and B. arenarum, where androgen showing the stimulation of spermatogenesis [1, 40, 44]. In male R. rugulosa, as observed in this study, androgen levels in both testis and plasma show a weakly reverse correlation with spermatogenetic stages of primary and secondary spermatogonia, but pa- rallel with changes in the number of spermatozoa. The role which androgen plays on different stages of spermatogenesis in R. rugulosa needs further investigation.

Annual Testicular Activity in Bullfrog 349

Relations of androgen patterns to breeding activity

In male R. rugulosa, we observed that the pattern of androgen secretion parallels with breed- ing activity; the circulating androgen peak occur- ring in early May was found to be coincident with the active breeding period in April-June. Such patterns are similar to those of P. dacnicolor [41], but different from those of R. perezi [6], B. japonicus [13], R. esculenta [38], and R. ridibunda [55] which showing a peak of plasma androgen level prior to breeding season. The patterns of testicular activity and androgen secretion observed in male R. rugulosa were synchronized with the patterns of ovarian maturation and estradiol secre- tion of the females, which were simultaneously investigated in a parallel study at the same frog culture farm [3].

The role of androgen in induction and subse- quent maintenance of male reproductive behavior, has been established in several anurans [14, 48, 54, 57]. For example, B. Mauitanicus [48] and E. coqui [54] both show a high androgen level during the amplexing behavior; while, B. japonicus [14] shows a high androgen level during the migration behavior. We also observed in the present study that high androgen levels in plasma of R. rugulosa are paralleled with its reproductive activity such as chorussing and spermiation, as well as with nuptial pad development in relation to amplexing be- havior. These observations support the proposal that androgen plays a role in maintenance of the male reproductive behavior in this species.

Association of environmental factors with repro- ductive cycle

Effects of environmental cues on male anuran reproductive cycle have been studied [6, 22, 42, 43, 44, 46]. It was reported that, in most subtropical and tropical species, rainfall is the most important factor associated with initiation of breeding activ- ity since temperature changes little with seasons in these areas [11]. However, the mechanisms of rainfall in the regulation of anuran reproductive cycle are not clear [46]. On the other hand, temperature is the primary factor in temperate species, which exhibiting a cyclic reproductive pattern with season. The role of temperature in

the regulation of anuran spermatogenetic cycle has been investigated in many speices [6, 22, 43, 46]. In R. esculenta, the proliferation of primary sper- matogonia is available in low temperature; in addition, the formation of secondary spermatogo- nia, and primary and secondary spermatocytes occurs with increasing temperature [42, 43]. It was subsequently demonstrated in this species that the temperature plays a major role in the sperma- togenesis via stimulation of the release of pituitary gonadotropins and testicular androgen secretion [43]. Rastogi et al. [43] also demonstrated that photoperiod serves as a permissive role on temper- ature influence upon spermatogenesis of R. esculenta. In the present study, we found that the rising phases of climatic factors (rainfall, tempera- ture, and photoperiod) parallel temporally with active testicular development and spermatogene- sis, the rising androgen levels in plasma and testis, and breeding activity during the annual reproduc- tive cycle of R. rugulosa. Regulatory mechanisms of environmental factors on a reproductive cycle of anurans are highly complex, and are presumably mediated by a pathway of optic or other transmis- sions to the hypothalamo-hypophysial-gonadal axis.

ACKNOWLEDGMENTS

We are grateful to the Department of Biology, Tun- ghai University, Taichung, and Institute of Zoology, Academia Sinica, Taipei, Taiwan, R. O. C., for the grants supporting this study. Thanks are also due to Mr. Shen San-tai, Miss Chen Chao-yin, Miss Chern Jyun-rur, and Miss Su Huey-min for assistance during various periods of this study.

REFERENCES

1 Basu SL (1962) A study of the spermatogenesis of common Indian frog (Rana tigerina) treated with testosterone. Naturwissenschaften 49: 188

2 Bolaffi JL and Callard IP (1979) Plasma steroid profiles in male and female mudpuppies, Necturus maculosus (Rafinesque). Gen Comp Endocrinol 37: 443-450

3 Chen CY (1988) Seasonal changes in ovarian activ- ity and plasma sex steroids levels in the Chinese bullfrog, Rana tigerina rugulosa Wiegmann. Mas- ter’s Thesis, Tunghai University, Taichung, Taiwan, ROC

4

17

350

Chen RH, Lin JY, Yu YL and Cheng HY (1987) Annual changes in plasma and testicular androgen in relation to reproductive cycle in a Japalura Lizard in Taiwan. Zool Sci 4: 323-329

Chieffi G, Rastogi RK, Iela L and Milone M (1975) The function of fat bodies in relation to the hypotha- lamohypophyseal-gonadal axis in the frog, Rana esculenta. Cell Tiss Res 161: 157-165

Delgado MJ, Gutierrez P and Alonso-Bedate M (1989) Seasonal cycle in testicular activity in the frog, Rana perezi. Gen Comp Endocrinol 73: 1-11 D’Istria M, Delrio G, Botte V and Chieffi G (1974) Radioimmunoassay of testosterone, 17-oestradiol and oestrone in the male and female plasma of Rana esculenta during sexual cycle. Steroids Lipids Res 5: 42-48

Fitzpatrick LC (1976) Life history patterns of stor- age and utilization of lipids for energy in amphibians Amer Zool 16: 725-732

Frost DR (1985) Amphibian Species of the World. Lawrence, Kansas: Assoc Syst Coll pp 512-518 Garnier DH (1985) Androgen and estrogen levels in the palsma of Pleurodeles waltl, Michah, during the annual cycle. I. Male cycle. Gen Comp Endocrinol 58: 376-385

Gavaud J (1975) Etude experimentale du role des facteur externes sur la spermatogenese et la ster- oidogenese des male de Nectophrynoides occidenta- lis Angel J Physiol (Paris) 70: 549-559

Huang, WS (1991) Male reproductive cycles of two sympatric toads, Bufo bufo gargarizas and Bufo melanostictus. Master’s Thesis, Tunghai University, Taichung, Taiwan, ROC

Itoh M, Inoue M and Ishii S (1990) Annual cycle of pituitary and plasma gonadotropins and plasma sex steroids in a wild population of the toad, Bufo japonicus. Gen Comp Endocrinol 78: 242-253

Itoh M and Ishii S (1990) Changes in plasma levels of gonadotropins and sex steroids in the toad, Bufo Japonicus, in association with behavior during the breeding season. Gen Comp Endcrinol 80: 451-464 Kime DE (1980) Comparative aspects of testicular androgen biosynthesis in nonmammalian verteb- rates. In “Steroids and Their Mechanism of Action in Nonmammalian Vertebrates”. Eds by G Delrio and J Brachet, Raven Press, New York pp 17-32 Kime DE and Hews EA (1978) Androgen biosyn- thesis in vitro by testes from amphibia. Gen Comp Endocrinol 35: 280-288

Kuhn ER, Gevaerts H, Vandorpe G and Jacobs GFM (1987) Plasma concentrations of testosterone and thyroxine in males of the giant swamp frog, Dicroglossus occipitalis at the equator. Gen Com Endocrinol 68: 492-493

Lecouteux A, Garnier DH, Bassez T and Joly J (1985) Seasonal variations of androgens, estrogens,

19

20

21

22

23

24

25

26

Di

28

29

30

32

Y. Kao, P. S. ALEXANDER, et al.

and progestorone in the different lobules of the tesits and in the plasma of Salamandra salamandra. Gen Comp Endcrinol 58: 211-221

Licht P, McCreery BR, Barnes R and Pang R (1983) Seasonal and stress related changes in plasma gona- dotropins, sex steroids, and corticosterone in the bullfrog, Rana catesbeiana. Gen Comp Endocrinol 50: 124-145

Liu CC (1950) Amphibians of Western China. Fieldiana: Zoology Memoirs 2: 1-400

Lofts B. (1964) Seasonal changes in the functional activity of the interstitital and spermatogenetic tis- sues of the green frog, Rana esculenta. Gen Comp Endocrinol 4: 550-562

Lofts B (1974) Reproduction. In “Physiology of the Amphibia”. Ed by B Lofts, Academic Press, New York Vol 2 pp 107-218

Long DR (1987) A comparison of energy substrates and reproductive patterns of two anurans, Acris crepitans and Bufo woodhousei. Comp Biochem Physiol 87A (1): 81-91

Loumbourdis NS and Kyriakopoulou-Sklavounous P (1991) Reproductive and lipid cycles in the male frog Rana ridibunda in northern Greece. Comp Biochem Physiol 99A: 577-583

Lupo di Prisco C, Basile C, Delrio G and Chieffi G (1972) In vitro metabolism of cholesterol-4-C'* and testosterone-4-C" in testes and fat bodies of Tritu- rus cristatus carnifex. Comp Biochem Physiol 41B: 245-249

Martin B (1980) Steroid-protein interactions in nonmammalian vertebrates: distribution, origin, regulation, and physiological significance of plasma steroid binding proteins. In “Steroids and Their Mechanisms of Action in nonmammalian Verte- brates”. Eds by G Delrio and J Brachet, Raven Press New York pp 63-73

Maruyama K (1979) Seasonal cycles in organ weights and lipid levels of the frog, Rana nigromacu- lata. Annot Zool Japon 52(1): 18-27

Milone M, Chieffi G, Caliendo MF and Pierantoni R (1984) Effect of the removal fat bodies on lipid biosynthesis in frog testis. Boll Zool 51: 75 (Ab- stract)

Morton MC (1981) Seasonal changes in total body lipid and liver weight in the Yosemite toad. Copeia 1981(1): 234-238

Moriguchi Y and Iwasawa H (1987) Annual changes in male reproductive organs in Bufo japonicus formosus: Histological observation. General Endo- crinol 6: 115-120

Muller CH (1975) Jn vitro steroidogenesis by bull- frog testis. Soc Study Reprod 8th Ann Meeting pp 95-96 (abstract 104)

Muller CH (1976) Secretion of 5a-dihydrotestoste- rone by bullfrog testis. Prog Abstr 58th Ann Meet-

33

34

35

36

37

38

39

40

41

42

43

44

45

46

Annual Testicular Activity in Bullfrog

ing Endocrine Soc p 286 (abstract 423)

Norris DO, Norman MF, Pancak MK and Duvall D (1985) Seasonal variations in spermatogenesis, tes- ticular weights, vasa deferentia, and androgen levels in neotenic male tiger salamanders, Ambystoma tigrinum. Gen Comp Endocrinol 60: 51-57

Oordt PGWJ van (1960) The influence of internal and external factors in the regulation of the sperma- togenetic cycle in amphibia. Symp Zool Soc London 2: 29-52

Ozon R (1972) Androgens in fishes, amphibians, reptiles, and birds. In “Steroids in Nonmammalian Vertebrates”. Ed by DR Idler, Academic Press, New York pp 329-389

Ozon R, Breuer H and Hisboa BP (1964) Etude du metabolisme des hormones steroides chez les ver- tebre inferieurs. III Metabolisme in ivtro de la testosterone par l’ovaire et la testicule de la gre- nouille Rana temporaria. Gen Comp Endocrinol 4: 577-583

Ozon R and Stocker C (1974) Formation in vitro de 5a-dihydrotestosterone par le testicular de Discog- lossue pictus. Gen Comp Endocrinol 23: 224-236 Pierantoni R, Iela L d’'Istria M, Fasano S, Rastogi RK and Delrio G (1984) Seasonal testosterone profile and testicular responsiveness to pituitary factors and gonadotrophin-releasing hormone dur- ing two different phases of the sexual cycle of the frog (Rana esculenta) J Endocrinol 102: 387-392 Pope CH and Boring AM (1940) A survey of Chinese Amphibia. Peking Nat Hist Bull 15: 49-50 Rastogi RK and Iela L (1980) Steroidogenesis and spermatogenesis in anuran amphibia: A brief sur- vey. In “Steroids and Their Mechanism of Action in Nonmammalian Vertebrates”. Eds by G Delrio and J Brachet, Raven Press, New York pp 131-146 Rastogi RK, Iela, L, Delrio G and Bagnara JT (1986) Reproduction in Mexican leaf frog, Pachymedusa dacnicolor Il. male Gen Comp En- docrinol 62: 23-35

Rastogi RK, Iela L, Delrio G, Dimeglio M, Rosso A and Chieffi G (1978) Environmental influence on testicuiar activity in the green frog, Rana esculenta. J Exp Zool 206: 49-64

Rastogi RK, lela L, Saxena PK and Chieffi G (1976) The control of spermatogenesis in the green frog, Rana esculenta. J Exp Zool 196: 151-166

Saidapur SK (1983) Patterns of testicular activity in Indian amphibians. Indian Rev Life Sci 3: 157-184 Saidapur SK and Nadkarni VB (1975) Seasonal variation in the structure and function of testis and thumb pad in the frog Rana tigerina (Daud.). Indian J Exp Biol 13: 432-438

Salthe SN and Mecham JS (1974) Reproductive and

47

48

49

50

51

a2

53

54

55

56

Syi/

58

59

351

courtship patterns. In “Physiology of the Amphi- bia”. Ed by B Lofts, Academic Press, New York Vol 2 pp 309-521

Schlaghecke R and Blum V (1978) Seasonal varia- tions in liver metabolism of the green frog Rana esculenta (L). Experientia 34: 456-457

Siboulet R (1981) Variations saisonnieres de la teneur plasmatique en testosterone et dihydrotestos- terone chez le crapaud de mauritanie (Bufo maurita- nicus). Gen Comp Endcrinol 43: 71-75

Steel RGD and Torrie JH (1980) In “Principles and Procedures of Statistics”, 2nd ed McGraw-Hill Book Company

Tajima H, Arai R, Tomaoki B-I and Hanaoka K-I (1969) In vitro steroidogenesis in testicular homogenates of the Japanese newt, Cynops Pyrrho- gaster (Boie). Gen Comp Endcrinol 12: 549-555 Tanaka S, Iwasawa H and Wakabayashi K (1988) Plasma levels of androgens in growing frogs of Rana nigromaculata. Zool Sci 5: 1007-1012

Tanaka S and Takikawa H (1983) Seasonal changes in plasma testosterone and 5a-dihydrotestosterone levels in adult male newt, Cynops pyrrhogaster pyrrhogaster. Endocrinol Japon 30(1): 1-6

Ting HP and Boring AM (1939) The seasonal cycle in the reproductive organs of the Chinese toad Bufo bufo and the pond frog Rana nigromaculata. Peking Nat Hist Bull 14: 49-80

Townsend DS and Moger WH (1987) Plasma androgen levels during parental care in a tropical frog (Eleutherodactylus). Horm Behav 21: 93-99 Vandorpe G, Jacobs GFM and Kuhn ER (1987) Seasonal changes of the 5’-monodeiodination activ- ity in kidney and skin homogenates of male Rana ridibunda: relation to plasma thyroxine (14) and testosterone. Gen Comp Endocrinol 68: 163-169 Varriale B, Pierantoni R, Matteo LD, Minsucci S, Fasano S, D’Antonio M and Chieffi G (1986) Pla- sma and testicular estradiol and plasma androgen profile in the male frog Rana esculenta during the annual cycle. Gen Comp Endocrinol 64: 401-404 Wada M, Wingfield JC and Gorbman A (1976) Correlation between blood level of androgens and sexual behavior in male leopard frogs, Rana pipiens. Gen Comp Endocrinol 29: 72-77

Yoneyama H and Iwasawa H (1985) Annual changes in the testis and accessory sex organs of the bullfrog, Rana catesbeiana. Zool Sci 2: 229-237 Yu JYL, Liaw JJ, Chang CL, Lee SN and Chen MC (1988) Seasonal changes in circulating levels of gonadal steroids during an estrous cycle of holstein- friesian cows Bull Inst Zool Academia Sinica (ROC) 27(2): 133-143

a) Aa | enz lity; sidan a re Lanning inl ah gilts cx

ee: ae

ory ¥ dui bon & showtyetied .

oes a ea td paid

+ Wve Sorta Tp yaoi ea ar ( _ QR peter et ? scm evty satel

“ah er | i J (hoaa

ir3 4 i¢ ) a F t ; j 0 < Raia f i r * ' . a } ri s i ! it } am Oe

“?

se Ueermaatarcens earl abearvlite’y

VP cbwt? (Menthe AL

ae ib ocasdowd eine 1 Srp #list gle yaks Katha ZA 1a 4

retry? ite lel iroinitlargans lh ik en tied VOR Maen a angel nA / colopy Momide

elitr ts viata tie ‘ bas ‘

m4 .eichthe x bh ate e Gla ny

ok ,

rer woh date, 2olimaars Beri

h Teameeoberre. ane ripeles “oqy : mule 4 ewhd ih ptkerady AO R yy: Pr ere a ues ME 450 st oruerethnd UH Benes im whi ; Hohe V acdayuab? 1 qb) i hie. oO) | pee vie el my seedy rife | garmaghyt 7 ewe hee OY SGA e eng sae jecpetif. kaw of

-_. Wert) MA. one

ye Ljiet) gti eirc” qurdaffe ~aeyiukieale (try athe wrth eT) eles Cosine A Letty lade? eel iy’ c

rts vO acdc Waeeeciaey : - J ay a lAl ere ' Meat 4.441 Ghisll ol mea Ole able syetile yte ete VL afta

hes rea

j it reivt Lote he ei ih vere 7iawet f FD, hel 4 ’ ia A’ weeret ' 4 Pe ok aa, f 4

lege 1 yeren few Ge ei! saat qed ae TT ae hog eal Pum sat) ervealt palo 2a Obed Te ‘ev ye fre TP V0) ee teohe) ; ec Td

ad ” . a

ZOOLOGICAL SCIENCE 10: 353-360 (1993)

© 1993 Zoological Society of Japan

Prolactin Binding Sites in Normal Uterus and the Uterus with Adenomyosis in Mice

TIPPAWAN SINGTRIPOP, TAKAO Mori, MIN Kyun Park,

Kolcui SHIRAISHI, TOSHIO HARIGAYA*

AND SEIICHIRO KAWASHIMA

Zoological Institute, Faculty of Science, University of Tokyo, Tokyo 113, and *Laboratory of Functional Anatomy, Faculty of Agriculture, Meiji University, Kawasaki 214, Japan

ABSTRACT— Radioreceptor assay for mouse prolactin (mPRL) was developed in the mouse uterus. The specific binding of ['*°I]-mPRL to the uterine preparations was demonstrated and the binding was found to have a similar affinity to that in the liver preparations. The specific PRL-binding to the uterine preparations tended to increase after estrogen treatment, while the differences among the three groups, diestrous intact, ovariectomized, and ovariectomized and estrogen-treated mice, were statistically not significant. Immunohistochemical staining of PRL revealed that PRL immunoreactivity was preferen- tially localized in the myometrial cells. Pituitary grafting into the uterus induced a high incidence of adenomyosis associated with the increase in plasma PRL levels. Furthermore, the PRL-binding activity

in the uterus with adenomyosis was significantly higher than that in normal uterus.

These results

indicate that the elevation of PRL binding in the uterus, hyperprolactinemia, and the development of

adenomyosis are interrelated.

INTRODUCTION

Prolactin (PRL) is known to act on various organs, such as the brain, mammary gland, liver, ovary, testis and some reproductive tracts in mam- mals [1, 6, 20, 21, 26]. Evidence of PRL effects on these target organs involves the presence of spe- cific PRL-binding sites. Estrogens have been found to increase the specific PRL-binding sites in the liver of rats in the presence of intact pituitary glands [22, 24]. In mice, however, Marshall et al. [9] reported that ovariectomy increased the num- ber of liver PRL-binding sites, and exogenous estrogen administration resulted in a decrease in the number. By contrast, estrogen administration increases the number of PRL-binding sites in the mouse mammary gland [29]. Therefore, the reg- ulation of the number of PRL-binding sites is variable among different organs.

Previous studies have revealed that the specific

Accepted November 20, 1992 Received December 11, 1992

PRL-binding sites are detected in the uteri of rat [34], rabbit [4], pig [36], sheep and cow [23], mink [27] and human [5] by means of radioreceptor assay. In mice, PRL-binding sites have been found in the brain, mammary gland, liver and kidney [3, 25]. However, the presence of PRL receptors in the uterus has not yet been reported in this species, in spite of evident synergistic effect of PRL with estrogen on the mouse uterus.

While long-term exposure to estrogen is known to induce uterine adenomyosis, a pathological dis- order of endometrial tissues defined as the out- growth of endometrial glands and stroma into the myometrium in some animal species [14, 15, 17, 32], our experiments have demonstrated that in mice isologous anterior pituitary transplantation into the uterine lumen induces an early and a high incidence of uterine adenomyosis associated with the elevation in blood PRL levels [11, 13]. Fur- thermore, we have found the accelerative effect of dopamine antagonist which stimulates PRL release from the pituitary [30], and the inhibitory effect of danazol which suppresses pituitary PRL release

354 T. Sinctrivop, T. Mort et al.

[31] on the development of adenomyosis. From these findings, we have proposed that PRL plays an important role for the development of ade- nomyosis, and have claimed that the uterus is one of the target organs of PRL, suggesting the pre- sence of PRL receptors in the mouse uterus.

In the present study, radioreceptor assay of PRL binding to the mouse uterus and liver was de- veloped. In addition, the relationship between the development of adenomyosis and the PRL-binding ability of the mouse uterus was examined.

MATERIALS AND METHODS

Animals and treatment

Experiment I Five-week-old female mice of the ICR/Jcl strain were purchased from Japan CLEA Inc. (Tokyo, Japan) and kept in light(12 hr/day)- and temperature(25+0.5°C)-controlled animal room. Forty-five mice were ovariectomized at 8 weeks of age. One day after ovariectomy, 20 out of the 45 mice were given a subcutaneous injec- tions of 5 4g estradiol-17 «f in 0.1 ml sesame oil daily for 7 days at 9:00 AM and killed by decapita- tion one hour after the last injection. Remaining 25 ovariectomized mice given no steroid injections were also killed at 9 weeks of age. In addition, 25 intact mice at estrus were killed in the early morning at 9 weeks of age.

Experiment IT Fifty adult female mice of the SHWN strain were used in this experiment. The experimental group of 24 mice was given trans- plantation of the anterior pituitary gland (AP) into the uterine lumen at 6 weeks of age and killed by decapitation at 20 weeks of age. The AP donors

were age-matched male littermates. The AP- grafted mice which were at diestrus and marked subserosal nodules visible on the outer surface of the uterus, an advanced state of adenomyosis [13] were chosen for further analysis, and the AP- grafted mice which showed no sign of the develop- ment of nodules were discarded. Twenty six intact mice which were killed by decapitation at diestrus at 30 days or 20 weeks of age and showed no sign of the development of nodules were used as con- trols.

Experiment III Female mice of the ICR /Jcl

strain purchased from Japan CLEA Inc. were used in this experiment. Three intact mice at diestrus were killed by cervical dislocation at 50 days of age and used for PRL immunohistochemical staining.

Animal handlings were carried out in accord- ance with the NIH Guide to the Care and Use of Laboratory Animals.

Receptor preparations

Receptor samples were prepared essentially fol- lowing the method of Kelly et al. [7, 8] and Young et al. [37]. Immediately after autopsy, the uteri and liver from about 10-15 mice were separately pooled on ice. Pooled tissues were weighed and rinsed in 5 ml of ice-cold homogenizing buffer (pH 9.0) composed of 100 mM Tris, 150 mM NaCl, 50 mM _ ethyleneglycol-bis-(B-aminoethyl ether)-N, N-tetraacetic acid (EGTA), 50mM_ ethylene- diaminetetraacetic acid (EDTA), 300 mM sucrose, 1 mM phenyl-methylsulfonyl fluoride (PMSF) and aprotinin (400 kallikrin inhibitory units/ml), and were chopped by a razor blade for 5 min on an ice-cold teflon plate. The chopped-tissues were homogenized in cold homogenizing buffer (4 ml/g tissue fresh weight) by using a Polytron PT 3000 homogenizer for three times at 20,000 rpm, 20 sec each, in a glass tube in ice bath. After centrifuga- tion of the tissue homogenates at 3,750 rpm for 15 min at 4°C, the precipitate was discarded and the supernatant was recentrifuged at 33,000 rpm for 120 min at 4°C. The second precipitate was used as the receptor preparation and was resuspended in 500 ul of cold assay buffer (pH 7.4), containing 25 mM Tris, 10 mM MgCh, 0.1% bovine serum albu- min and 0.02% sodium azide. Protein measure- ment was performed in 10 ul of the resuspension by micro BCA protein assay reagent kit (Pierce Co., U.S.A). Remaining receptor preparations were stored at —80°C until radioreceptor assay for PRL.

Todination of mPRL

Mouse PRL (mPRL) (AFP6476C, donated by Dr. A. F. Parlow) was radioiodinated with '™I (Na'I, Radiochemical Centre, Amersham, UK) in the presence of lactoperoxidase (Boehringer Mannheim GmbH, Germany) and hydrogen per- oxide according to the method of Miyachi ef al.

Prolactin Receptors in Mouse Uterus 355

[10] with minor modifications. After the reaction, the radio-labeled hormone was separated from free radioiodine by gel filtration using a Sephadex G-50 column (disposable plastic column; Whale Sci. Inc. U.S.A). The specific activity of labeled mPRL was about 50 “Ci/ yg.

Procedures for PRL-binding assay

The procedures were basically the same as the methods of Kelly et al. [7, 8] and Young et al. [37]. For routine PRL-binding assay of liver samples of ICR mice, 500 “ug of protein of the second precipi- tate of liver homogenates in 100 ul of assay buffer was incubated with ['*I]-mPRL (ca 55,000 cpm, 0.5 ng/100 wl) at 25°C in an assay tube (1275 barosilicate glass tube) in the presence or absence of an excess amount (5 ug) of unlabeled ovine PRL (OPRL, Sigma) for 18 hrs. Incubation was terminated by addition of 3 ml of cold assay buffer. Unbound hormone was separated by centrifuga- tion at 4,000 rpm for 30 min at 4°C. After aspira- tion, the radioactivity of precipitates was counted by an autowell gamma counter (Aloka ARC-300). Each assay was composed of triplicate determina- tions. In some experiments, recombinant mPRL (5 wg/tube) raised by Yamamoto ef al. [35] was used as the competitor instead of oPRL.

The specificity of the binding was analyzed by competitive displacement experiments. Specific binding of ['*I]-mPRL was calculated as the dif- ference between the binding in the absence and presence of various concentrations of unlabeled OPRL (5, 50, 500, 5,000, 50,000 ng). All the reaction tubes contained 500 yg of protein in 300 yl assay buffer and 55,000 cpm of ['°I]-mPRL (100 1). The affinity of binding (Ka) and the dissociation constant (Kd) were determined from the Scatchard plots constructed from the binding displacement experiments [28].

In some experiments, the receptor samples were pretreated with 4M MgCl to dissociate endoge- nously bound PRL from its binding sites before assay following the method of Kelly et al. [8]. However, routine assays were carried out without MgCl, pretreatment, because MgCl. treatment was not effective in increasing PRL binding in either uterus or liver preparations (data not shown). PRL-binding assays at each point were

repeated twice in pooled samples of two groups consisting of 10-15 mice each.

Radioimmunoassay (RIA ) of serum PRL

In Experiment II in SHN mice, blood was collected at autopsy by decapitation and allowed to clot at room temperature for one hour. After centrifugation at 3,000 rpm for 20 min, the serum was stored at —80°C. Serum PRL levels were determined by homologous RIA using mPRL RIA kit (reference preparation AFP6476C; rabbit anti- mPRL serum, AFP131078; and mPRL, AFP6476C, ['*°I]-labeled) donated by Dr. A. F. Parlow. Assay reaction mixture was incubated overnight at room temperature. PRL-antibody complexes were separated from unbound PRL by centrifugation after the incubation for 3 hr at room temperature with goat anti-rabbit immunoglobulin serum, donated by Dr. K. Wakabayashi. The PRL concentrations assayed in 100 yl serum were ex- pressed in terms of ng AFP6476C/ml.

Immunohistochemistry of PRL

In Experiment III, the uterus and kidney were dissected out and fixed in 4% paraformaldehyde in phosphate buffer(pH 7.3) at 20°C overnight. Tis- sues were dehydrated in alcohol, embedded in paraffin, sectioned at 5 ~m, and mounted on glass slides previously coated with gelatin. The sections were deparaffinized and rinsed with 0.1 M phos- phate-buffered saline (PBS). Immunohistochem- ical demonstration of PRL was carried out by Histofine SAB-PO (Nichirei Co. Tokyo) method. The sections were preincubated with goat normal serum diluted with PBS 10 times for 10 min and washed briefly with a fresh PBS solution. Rabbit anti-mPRL (AFP6476C) at various dilutions (1:2,500, 1: 10,000 and 1:50,000) in PBS was used as the primary antiserum. After incubation with the primary antiserum overnight at room tempera- ture, the sections were washed with 0.1M PBS containing triton X-100 (PBST), and then incu- bated with biotinylated anti-rabbit IgG for 15 min, followed by incubation with peroxidase- conjugated streptavidin for 20 min. Following 15 min washing with PBST, the sections were incu- bated for 15 min with 0.05 % 3,3’-diaminobenzi- dine (DAB) tetrahydrochloride solution contain-

356 T. Sinctrirpop, T. Mori et al.

ing 0.01% hydrogen peroxide and lightly counter- stained with hematoxylin. All these steps were performed at room temperature. The specificity of the method was examined by the use of diluted normal rabbit serum 1:50,000 or PBS instead of primary antiserum. For another test for the anti- gen specificity, primary antiserum (1 : 500 dilution) was added with mPRL (400 ng/ml antiserum), stirred for one hour at room temperature, and incubated at 4°C overnight. To eliminate the resolubilization of the antibody-antigen com- plexes, the mixture was centrifuged at 4,000 rpm for 15 min, and the supernatant was used. The sections incubated without primary antiserum or with adsorbed antiserum showed no immuno- staining.

Statistical analysis

The significance of differences between any two groups was evaluated by Student’s t-test, where P <0.05 was considered as statistically significant.

RESULTS

In order to determine the optimum incubation time, 500 ug of liver protein and 55,000 cpm of ['*I]-mPRL were incubated with or without 5 ug of unlabeled hormone (OPRL) for 3, 6, 12, 18 and

0.05 0.04 ® ® ic 0.03 ao} 5 © 0.02 aa) 0.01 0 LS iS 60 120 180 240 300 oPRL Bound (pg/ml) —e— uterine preparation «— liver preparation Fic. 1. Scatchard plots of binding of ['*°I]-mPRL to

ICR mouse liver and uterine preparations, based on the data of the binding of ['*°I]-mPRL to the liver and uterine preparations of ICR mouse under var- ious concentrations of unlabeled oPRL.

24 hrs. Incubation for 18 hrs resulted in the max- imum specific binding of ['7°I]-mPRL to 500 yg of protein of liver preparations. When various amounts of liver protein (125, 250, 500, 1,000, 2,000 wg) and 55,000 cpm of ['*I|-mPRL were incubated in the presence or absence of 5 ug of unlabeled of oPRL, the specific binding increased as a function protein concentration.

Figure 1 shows Scatchard plots drawn from the data of displacement experiments in liver and uterine preparations. From the plots, the equilib- rium constant of dissociation (Kd) and the affinity constant (Ka) were calculated: for liver Kd=0.048 x10~-§M, Ka=2.08x 10° M~!; for uterus Kd= 0.054 10~*M, Ka=1.8x10°M ?.

Thus, in this study the binding assays in both liver and uterine preparations were routinely car- ried out under the following conditions; 500 yg of protein for a one assay sample; 55,000 cpm of ['*I]-mPRL, with or without 5 ~g unlabeled oPRL to obtain specific binding and 18 hrs of incubation als D(C.

Ovariectomy increased [!*°I]-mPRL binding to the mouse liver preparations compared to that in intact control mice at estrus (Fig. 2). Estrogen treatment to ovariectomized mice tended to de- crease the binding to the liver preparations (Fig. 2), while the differences among the three groups

Specific binding/mg protein (%) aS

estrus OX OX+E,

G.2. The specific binding of ['*I]-mPRL (% of total radioactivity added) to liver preparations 500 ug of ICR mice at estrus, 7 days after ovariectomy (OX) and 7 days of estrogen treatment following ovariec- tomy (OX+E;). Five ug unlabeled oPRL was used to assess non-specific binding. The bars and vertical lines denote mean of two pooled samples.

F

—

Prolactin Receptors in Mouse Uterus 357

Specific binding/mg protein (%)

OX+E,,

Fic. 3. The specific binding of ['*°I]-mPRL (% of total radioactivity added) to uterine preparations (500 ug protein) of ICR mouse at estrus , OX and OX+E). Five vg unlabeled oPRL was used to assess non- specific binding. The bars and vertical lines denote mean of two pooled samples. For explanation of abbreviations, refer to legend of Fig. 2.

estrus Ox

were not statistically significant.

In the uteri of ovariectomized mice, the specific binding of ['*I]-mPRL was lower than that in intact mice at estrus (Fig. 3). Daily injections of 5 yg estradiol for 7 days induced a slight increase in the binding activity as compared to the prepara- tions from ovariectomized mice. However, the

ine} (¢°) -

—

Specific binding/mg protein (%)

young adult adenomyosis

Fic. 5. The specific binding of ['*°I]-mPRL (% of total radioactivity added) to uterine preparations of intact control 30-day-old or 20-week-old SHN mice and the 20-week-old mice with adenomyosis. The re- combinant mPRL was used as competitor to assess non-specific binding. The bars and vertical lines denote mean of two pooled samples. *P<0.05 (vs. young & adult controls).

ine) (ee)

—

Specific binding/mg protein (%)

0 control adenomyosis

Fic. 4. The specific binding of ['°I]-mPRL (% of total radioactivity added) to uterine preparations of intact control SHN mice and the 20-week-old mice with adenomyosis. Five ug unlabeled oPRL was used to assess non-specific binding. The bars and vertical lines denote mean of two pooled samples.

differences among the groups were again statisti- cally not significant (Fig. 3).

The specific binding of [!*°I]-mPRL to the uter- ine preparations from 20-week-old SHN mice at diestrus with adenomyosis tended to be higher than that in age-matched control mice at diestrus, where 5 ug oPRL were used as the competitor to assess nonspecific binding. However, the differ- ence was not statistically significant (Fig. 4). When 5 wg of unlabeled recombinant mPRL was used in

100

PRL concentration (ng/ml) - (op) (oo) [o) (o) (o)

Ie) [o)

EES Eee] young adult adenomyosis Fic. 6. Plasma PRL levels (ng AFP6476C) in intact

control 30-day-old or 20-week-old SHN mice at diestrus and 20 week-old SHN mice with adenomy- osis in their uteri. The bars and vertical lines denote means+S.E.M. (n=4-9). Significantly different from both intact controls: **P<0.01.

358 T. Sinctripop, T. Mori et al.

Sey

en

Fic. 7.

Hacnoneechemicnl localization of mPRL in paraffin sections of the mouse mane PRL is detected in the

proximal convoluted tubules (P). AntirmPRL (AFP6476C) at 2,500 dilution. Bar represents 100 um.

Fic. 8.

Immunohistochemical localization of mPRL in paraffin sections of the mouse uterus. PRL is detected in the

muscle layer (M). Anti-rmPRL (AFP6476C) at X2,500 dilution. Bar represents 100 ~m.

place of oPRL, the specific binding of [!7°I]-mPRL to uterine preparations in mice with advanced state of adenomyosis (3.90+0.13%) was significantly higher than that in 30-day-old or 20-week-old normal mice (P<0.05, in either comparison) (Fig. 5)))o

The serum PRL level is significantly higher in pituitary-grafted 20-week-old mice bearing ade- nomyosis in their uteri than those in normal 30- day-old or 20-week-old mice, (P<0.01, in either comparison) (Fig. 6).

The specificity of antibody and immunohis- tochemical reaction was tested by the sections of mouse and rat kidneys, since the influence of PRL on the kidney was well established [13, 16, 32]. Immunoreactivity to mPRL was localized in some proximal convoluted tubules and the medulla of mouse kidneys (Fig.7). In the uterus, im- munoreactivity to anticmPRL serum (in either dilutions, < 2,500, 10,000 or x 50,000) was pre- ferentially localized in the muscle layers, and the reactivity of other areas of the uterus was negligi- ble (Fig. 8).

DISCUSSION

In the present experiments, optimum conditions for the specific PRL-binding assay were deter- mined by using mouse liver, because the assay methods in the liver have been well established. Assay data from the present experiments revealed

that maximum binding of [!*°I]-mPRL to its recep- tors occurred at 18 hrs of incubation at 25°C. In addition, 500 ug of sample protein and 55,000 cpm of [!*I]-mPRL with or without of 5 ug unlabeled oPRL were found to be the optimum condition for the binding assay. These conditions were basically similar to those reported by previous investigators using rat liver preparations [4, 5, 7, 23, 27, 34, 36]. According to Marshall et al. [9], ovariectomy in- creased [!*°I]-mPRL binding to the liver prepara- tions of female mice, and estrogen treatment de- creased the binding, indicating the manifestation of down-regulation in the binding of PRL to the liver by estrogen. In the present study, the same tendency was observed in the PRL-binding to the liver samples.

The present results indicate that specific PRL receptors were present in the uterus of mice. Scatchard plot analysis revealed that the affinity of binding of mPRL to mouse uterine preparations was similar to that to liver preparations.

The level of specific PRL-binding in the uterus was low in ovariectomized mice and tended to increase after estrogen treatment, although these differences were not statistically significant. The present results that the changes in the PRL- binding activity were very slight under various hormonal conditions may be due to the insufficien- cy in estrogen stimulation. Treatment by greater doses of estrogen for longer period might be needed. However, the present immunohistochem-

Prolactin Receptors in Mouse Uterus 359

ical observations revealed that the target cells of PRL in the uterus were preferentially myometrial cells, confirming our previous suggestion [11, 16]. Because the amount of myometrial tissue in the uterus constitutes only a small portion of the organ, any hormonal influence on the number of PRL receptors in the myometrium might be too meager to get a significant difference.

In these experiments, oPRL was routinely used to obtain non-specific binding. When recombinant mPRL was used, significant increase in PRL recep- tors was observed in the uterine preparations of mice with adenomyotic changes. Therefore, re- combinant mPRL might be a better competitor for future study.

PRL plays a key role in the development of adenomyosis [11—13, 15, 16, 30, 31]. The present results showed that the specific binding of PRL in the uterus and plasma PRL level were significantly higher in pituitary-grafted mice with adenomyosis than in normal age-matched controls. Previous reports indicated that changes in the endometrial PRL receptors in the pig are regulated by exoge- nous estrogen administration and circulating levels of PRL during gestation and lactation [36, 37]. The specific binding of PRL to the mouse mam- mary gland was increased by treatment with var- ious doses of estrogen associated with elevated serum PRL levels in ovariectomized mice [19, 29]. These results suggest a possible direct effect of estrogen to increase the specific binding of PRL and/or indirect effect of estrogen through its ac- tion on the pituitary prolactin release. Thus, it is most likely that statistically significant increase in the specific binding of PRL may be due to the longterm hyperprolactinemia in the pituitary- grafted mice bearing uterine adenomyosis.

In conclusion, this is the first report to demon- strate the presence of specific PRL-binding sites in the mouse uterus. Immunohistochemical findings suggest that PRL binds to the myometrial cells. In addition, the present results strongly support our hypothesis that the development of adenomyosis in mice with pituitary grafting is related to the increase in both blood PRL level and PRL-binding activity in the uterus.

ACKNOWLEDGMENTS

This study was supported by Grants-in-Aid for Fun- damental Scientific Research from the Ministry of Education, Science and Culture, Japan. to T.M. and S.K., and Research Grants from Takeda Foundation to T.M. and from Zenyaku Kogyo LTD. to S.K.. T. Singtripop is a recipient of the fellowship from Hitachi Scholarship Foundation.

The authors thank Dr. K. Kohmoto and Dr. S. Sakai (Department of Animal Breeding, Faculty of Agricul- ture, University of Tokyo) for their technical advice. We are grateful to Dr. A. F. Parlow (Pituitary Hormones and Antisera Center Harbor-UCLA Medical Center, California) for his generous gift of mouse PRL RIA kit, and Dr. K. Wakabayashi (Institute of Endocrinology, Gunma University, Gunma, Japan) for providing us goat anti-rabbit IgG serum.

REFERENCES

1 Bern HA, Nicoll CS (1968) The comparative endo- crinology of prolactin. Rec Prog Hormone Res 24: 681-720

2 Dube D, Kelly PA, Pelletier G (1980) Comparative localization of prolactin-binding sites in different rat tissues by immunohistochemistry, radioautography and radioreceptor assay. Mol Cell Endocrinol 18: 109-122

3. Frantz WL, Macindole JH, Turkington RW (1974) Prolactin receptors: Characteristics of the particu- late fraction binding activity. J Endocrinol 60: 485- 497

4 Grissom FE, Litteton GK (1988) Evolution of lactogenic receptors in selected rabbit tissue during pregnancy. Endocrinol Res 14: 1-19

5 Healy DL (1984) The clinical significance of en- dometrial prolactin. Aust NZJ Obstet Gynecol 24: 111-116

6 Hermanns U, Hafez ESE (1981) Prolactin and male reproduction. Arch Androl 6: 95-125

7 Kelly PA, Posner BI, Friesen HG (1975) Effects of hypophysectomy, ovariectomy, and cycloheximide on specific binding sites for lactogenic hormone in rat liver. Endocrinology 97: 1408-1415

8 Kelly PA, Leblanc G, Dyjiane J (1979) Estimation of total prolactin-binding sites after in vitro desatura- tion. Endocrinology 104: 1631-1637

9 Marshall S, Bruni JF, Meites J (1978) Prolactin receptors in mouse liver: Species differences in response to estrogenic stimulation. Proc Soc Exp Biol Med 159: 256-259

10 Miyachi Y, Vaitukitis JL, Nieschlag, Lipset MP (1972) Enzymatic radioiodination of gonadotropins. J Clin Endocrinol Metab 34: 23-28

11 Mori T, Nagasawa H, Takahashi S (1981) The

12

13

14

15

16

17

18

19

20

21

22

23

24

360

induction of adenomyosis in mice by intrauterine pituitary isografts. Life Sci 29: 1277-1282

Mori T, Nagasawa H (1983) Mechanisms of de- velopment of prolactin-induced adenomyosis in mice. Acta Anat 116: 46-56

Mori T, Nagasawa H (1984) Alteration of the development of mammary hyperplastic alveolar nodules and uterine adenomyosis in SHN mice by different schedules of treatment with CB-154. Acta Endocrinol 107: 245-249

Mori T, Nagasawa H, Ohta Y (1989) Prolactin and uterine adenomyosis in mice In “Prolactin and Le- sions in Breast, Uterus and Prostate” Ed by H Nagasawa, CRC Press, Florida, pp 123-129

Mori T, Nagasawa H (1989a) Multiple endocrine syndrome in SHN mice: Mammary tumors and uterine adenomyosis In “Comparative Aspects of Tumor Development” Ed by HE Kaiser, Kluwer Academic Pub, Netherlands, pp 121-130

Mori T, Nagasawa H (1989b) Causal analysis of the development of uterine adenomyosis. Zool Sci 6: 103-112

Mori T, Kawashima S, Nagasawa H (1991a) Induc- tion of uterine adenomyosis by pituitary grafting and retardation of its development by bromocriptine- mesilate (CB-154) in Balb/c mice. In vivo 5: 107— 110

Mori T, Singtripop T, Kawashima S (1991b) Animal model of uterine adenomyosis: Is prolactin a potent inducer of adenomyosis in mice. Am J Obstet Gyne- col 165: 232-234

Muldoon TG (1981) Interplay between estradiol and prolactin in the regulation of steroid hormone receptor levels, nature, and functionality in normal mouse mammary tissue. Endocrinology 109: 1339- 1346

Nagasawa H (1979) Prolactin: Its role in the de- velopment of mammary tumors. Med Hypoth 5: 1117-1121

Nicoll CS, Bern HA (1972) On the action of prolactin among the vertebrates: Is there a common denominator? In “Lactogenic Hormones” Ed by GEW Wolstenholme and J Knight, Churchill Livingstone, London, pp 299-324

Posner BI, Kelly PA, Friesen HG (1974a) Induction of lactogenic receptor in rat liver: Influence of estrogen and the pituitary. Proc Natl Acad Sci USA 71: 2407-2410

Posner BI, Kelly PA, Shiu, RPC, Friesen HG (1974b) Studies on insulin, growth hormone and prolactin binding: ontogenesis effects of sex and pregnancy. Endocrinology 96: 521-531

Posner BI (1976a) Regulation of lactogen specific

5)

26

27

28

29

30

31

32

33

34

35

36

37

T. Sinctripop, T. Mori et al.

binding sites in rat liver: Studies on the role of lactogen and estrogen. Endocrinology 99: 1168- 1177

Posner BI (1976b) Characterization and modulation of growth hormone and prolactin binding in mouse liver. Endocrinology 98: 645-654

Reddi AH (1969) Role of prolactin in the growth and secretory activity of the prostates and other accessory glands of mammals. Gen Comp Endocri- nol 2, Suppl. 81-96

Rose J, Stormshak F, Adair J, Oldfield JE (1983) Prolactin binding sites in the uterus of the mink. Mol Cell Endocrinol 31: 131-139

Scatchard G (1949) The attraction of proteins for small molecules and ions. Ann NY Acad Sci 51: 660-672

Sheth NA, Tikekar SS, Ranadive KJ, Sheth AI (1978) Influence of bromocriptine on estrogen- modulated prolactin receptors of mouse mammary gland. Mol Cell Endocrinol 12: 167-176

Singtripop T, Mori T, Park MK, Sakamoto S, Kawashima S (1991) Development of uterine ade- nomyosis after treatment with dopamine antagonists in mice. Life Sci 49: 201-206

Singtripop T, Mori T, Sakamoto S, Sassa S, Park MK, Kawashima S (1992) Suppression of the de- velopment of uterine adenomyosis by danazol treat- ment in mice. Life Sci 51: 1119-1125

Squire RA, Goodman DG, Valerio MG, Fredrick- son T, Strandberg JD, Levitt MH, Lingeman CH, Harschbarger JC, Dawe CJ (1978) Female Repro- ductive System In “Pathology of Laboratory Anim- als Vol2” Ed by K Benirschk, FM Garner, TC Jones, Springer-Verlag, New York, p 1172 Sternberger LA, Pertrali JP, Joseph SA, Meyer HG, Mill KR (1978) Specificity of the immuno- cytochemical-luteinizing hormone-releasing hor- mone receptor reaction. Endocrinology 102: 63-73 William GH, Hammond JM, Weisz J, Mortel R (1978) Binding sites for lactogenic hormone in the rat uterus. Biol Reprod 18: 697-706

Yamamoto M, Harigaya T, Ichikawa T, Hoshino K, Nakashima K (1992) Recombinant mouse prolactin: Expression in Escherichia coli, purification and biological activity. J Mol Endocrinol 8: 165-172 Young KH, Bazer FW (1989) Porcine endometrial prolactin receptors detected by homologous radioreceptor assay. Mol Cell Endocrinol 64: 145- 154

Young KH, Kraeling RR, Bazer FW (1990) Effect of pregnancy and exogenous ovarian steroids on endometrial prolactin receptor ontogeny and uterine secretory response in pig. Biol Reprod 43: 592-599

ZOOLOGICAL SCIENCE 10: 361-366 (1993)

© 1993 Zoological Society of Japan

Juvenile Hormone Analogue, S-31183, Causes a High Level Induction of Presoldier Differentiation in the Japanese Damp-Wood Termite

KimiHtro Ocino, YosHtyuk Hirono!, TapAo Matsumoto!

and HasImMe ISHIKAWA?

Zoological Institute, Faculty of Science, University of Tokyo, Bunkyo-ku, Tokyo 113, and ‘Department of Biology, College of Arts and Sciences, University of Tokyo, Meguro-ku, Tokyo 153, Japan

ABSTRACT—The effect of a juvenile hormone analogue, 2-[1-methyl-2(4-phenoxyphenoxy)etho- xy]pyridine (S-31183), on differentiation of the presoldier was tested with the Japanese damp-wood termite, Hodotermopsis japonica. Sixth and 7th instar larvae were reared on filter papers impregnated with various amounts (10-1000 ug/filter paper) of the analogue for 3 weeks. While presoldiers were induced, with some mortality, under all the conditions tested, the rate of induction was significantly higher (>90%) with lower mortality (<10%) than under the other conditions when 10 yg or 30 yg of

S-31183 were administered to the 6th instar larvae.

The length of the premolt interval of larval-

presoldier molt was about 5.5 days, equal to that of larval-larval molt.

INTRODUCTION

The society of termites is characterized by poly- morphism and polyethism, which, like other social insects, are based on the caste system. Termites usually consist of several castes: primary reproduc- tives, supplementary reproductives, soldiers, and workers or pseudergates. In addition, replacement reproductives appear after the death of primary reproductives. The set of castes and scheme of postembryonic pathway producing castes vary with species.

The caste of the termite is not genetically, but epigenetically determined: the first instar larva can become any caste. It is well established that juvenile hormone (JH) is a key factor controlling the caste differentiation and regulation [12]. Yin and Gillot [23] proposed the following relationship between the titer of JH and the outcome of molts in a primitive termite, Zootermopsis angusticollis (Termopsidae). A high titer of JH in hemolymph

Accepted November 6, 1992 Received September 4, 1992 2 To whom offprint request should be addressed.

at a specific stage leads to the formation of soldier via presoldier, and a low titer to the production of alata, which will shed the wings and become primary reproductive. An intermediate titer in- duces the molts without differentiation, or station- ary molts.

While many studies have shown that the arti- ficial increase of JH titer in larvae of termites by exposing them to exogenous JHs or their ana- logues (JHAs) causes an increase in the number of presoldier [5, 14], its production at the rate of 100% has never been observed. Luescher [9] explained this by the concept of competence in asserting that termite can respond to stimulus only at a specific period (“sensitive” or “competence” period) within the intermolt. This explanation was confirmed for some termites in several studies [11, 15, 18]. In the meantime, Lenz [8] emphasized the influence of nutrition and caste composition on the differentiation of soldiers. For example, the pre- sence of soldiers in an experimental colony will tend to inhibit the differentiation of larvae to presoldiers. The actual JH titer in a termite is supposed to be influenced by these environmental factors. Lenz suggested that continuous contact

362 K. Oaino, Y. Hirono et al.

and exchange of food among termites bring about the transfer of pheromones. Although Luescher [10] hypothesized the existence of pheromone which is involved in the termite caste regulation, no actual substance has been identified yet.

In spite of its decisive role, little is known about the mechanism of JH action in the caste regulation [17]. This is mainly due to the difficulty in per- forming biochemical and molecular biological analyses on termites. In order to overcome the difficulty, it is desirable to control the caste dif- ferentiation artificially. If larvae are successfully induced into presoldiers synchronously at a high rate, we will be able to obtain much important information about the physiological, biochemical, and molecular biological mechanism underlying the caste determination in termites.

The Japanese damp-wood termite, Hodo- termopsis japonica Holmgren is a lower termite species whose differentiation of caste occurs at a late stage of development. Even the last (7th) instar larvae are capable of differentiation into either of nymphs, neotenics and soldiers. Soldiers also derive from either the 6th instar larvae or nymphs [13]. Since 2-[1-methyl-2(4-phenoxy- phenoxy)ethoxy]pyridine (S-31183) displayed higher hormonal activity than that of a well-known JHA, methoprene on some insects [1, 22], it was expected that the JHA would also induce the differentiation of presoldiers of termite at a high rate. To determine the conditions that induce differentiation of larvae into presoldiers at the highest rate, we have examined the effects of the amount of S-31183, group size, instar of larvae, and locality of original colony on the presoldier production in H. japonica.

MATERIALS AND METHODS

Termites

Hodotermopsis have been known from Sichu- ang, Guizhou, Hunan, Guangxi, Guangdong, Zhejiang, Fujian, Jiangxi, Vietnam on the conti- nent, and also from the Hainan and Taiwan is- iands. In Japan, H. japonica have been found from the Nansei archipelago and the Capes of Seta and Ashizuri [21]. Colonies of H. japonica were

collected in the natural forest at Onoaida in the Yaku-shima island and at Takamoto in the Naka- no-shima island of the Nansei archipelago, Japan. They were brought back to the laboratory and kept in plastic cases in an incubator at 25°C in the constant darkness.

Juvenile hormone analogue

2 -[1- Methyl - (4 - phenoxyphenoxy)ethoxy|pyri- dine, S-31183 was kindly offered by Sumitomo Chemical Company, Osaka, Japan. Technical grade (98.1%) of S-31183 was used in this study.

Treatments

Dark-colored 6th and 7th instar larvae were picked up from stock colonies, and groups of 10 or 20 individuals were placed in glass petri dishes (70 mm diam.) containing filter paper disks (approx. 33 cm’). Generally, newly-molted termite larvae are light-colored and gradually become darker. Preliminary studies suggested that the light- colored larvae do not have the competence to differetiate into presoldiers under the influence of JHA. Six experimental groups, each consisting of either 10 or 20 6th instar larvae, were kept with filter paper disks containing 0 (control), 10, 30, 100, 300, and 1000 ug of S-31183, respectively. The disks had been uniformly impregnated with 150 ul of the acetone solutions of S-31183 and then air-dried, on which insects fed instead of wood. The same procedures were used for the experi- ments with the 7th instar larvae. Equal numbers of males and females were included in each ex- perimental group. Insects were supplied with water daily and the treated filters were renewed weekly. All experimental groups were kept at 25°C in the constant darkness.

Experimental, groups were observed daily for three weeks. Each experimental series was tripli- cated. Larvae from the stock colony collected in the Yaku-shima island were used in one series, and larvae from the Nakano-shima island were used in two other series. The data were evaluated by multiple-way of variance and Tukey’s multiple comparison.

Determination of the length of premolt interval

At the premolt stage, larvae empty the hindgut

Induction of Presoldier of Termite

and become white, which is called “whitening”. We observed the larvae every about 12 hours and labeled individuals which began to whiten with the date and time by attaching sticker on the back. In the following days, the number of newly-molted larvae and presoldiers were counted. We esti- mated the length of premolt interval of each individual based on the duration between the whitening and molting. Larvae which did not molt and died after whitening were counted in mortal- ity, and not subjected to post-mortem examina- tions. Larvae which died a few days after molting were counted in both mortality and those molted.

RESULTS

It was obvious that S-31183 impregnated in filter paper disks caused superfluous production of pre- soldiers and some mortality of larvae (Table 1). In control group with no JHA, no presoldier was produced while a few larvae underwent larval- larval molt with little mortality. Significantly more presoldiers were produced with all the amounts tested (10-1000 yg) compared with the controls (P

TABLE 1.

6th-instar larvae

363

<0.01). Regardless of other factors, with 10 or 30 ug of S-31183, about 90% of larvae differentiated into presoldiers within 3 weeks from the start of experiment. The three higher doses (100, 300, 1,000 xg) led to significantly fewer (P<0.01) pro- duction of presoldiers than the above two. As shown in Fig. 1, with higher amounts of S-31183, the level of presoldier production attained the maximum on earlier days.

As the amount of S-31183 was increased, the mortality increased. The mortality of the larvae treated with 300 or 1,000 ug of S-31183 was signi- ficantly higher than with the lower dosages (P< 0.01), which was incurred mainly as a result of incomplete ecdysis (Table 1). The dosage of JHA also effected the larval-larval molting (f=9.37, P <0.01). Throughout all the experimental groups except control, only a few larvae underwent larval- larval molt.

With the 6th and 7th instar larvae, results were significantly different on the rate of presoldier production (f=15.34, P<0.01) and mortality (f= 11.03, P<0.01). In the 7th instar larvae, the rate of presoldier production was lower and mortality

Differentiation and mortality of H. japonica exposed to various doses of S-31183 for 3 weeks

7th-instar larvae

JHA % molted to

dose

% mortality

Larvae? Presoldiers?””

Groups of 10 larvae

0 10.0+5.8 00+ 0.0 6.7+ 3.3 ( 0.0+0.0) 10 0.0+0.0 96.74 3.3 3.3+ 3.3 ( 0.0+£0.0) 30 ©6©60.0+40.0 96.74 3.3 3.34 3.3 ( 0.0+0.0) 100 00.040.0 83.34 8.8 10.04 5.8 ( 0.0+0.0) 300 §=63.343.3 76.7+ 3.3 30.04 5.8 ( 6.76.7)

1000 0.0+0.0 56.74 6.7 23.34 8.8 (16.7+8.8) Groups of 20 larvae

O 11.741.7 00+ 0.0 5.0+ 2.9 ( 0.0+£0.0) 10 3.343.3 93.34 1.7 3.3+ 1.7 ( 0.0+0.0) 30 6©60.0+0.0 91.74 1.7 3.34 3.3 ( 0.0+£0.0) 100 1.74£1.7 85.04 2.9 16.74 3.3 ( 0.0+0.0) 300 =0.0+0.0 70.0+10.0 30.0410.4 (11.7+6.7)

1000 0.0+0.0 55.0+415.0 33.3+12.0 (25.047.6)

Values are means+SE of %

JHA % molted to

dose

% mortality

Larvae* Presoldiers*”

Groups of 10 larvae

0 33433 00400 6743.3 (000.0) 10 6.7433 90.0+ 0.0 3.3+3.3 ( 0.0+0.0) 30 0.0+0.0 93.34 3.3 6.7+3.3 ( 0.0+0.0) 100 0.0+0.0 70.0+10.0 13.3+8.8 ( 0.0+0.0) 300 0.0+0.0 63.3+ 88 43.3+8.8 (13.3+3.3)

1000 0.0+0.0 46.7+ 8.8 40.0+0.0 (30.0+0.0)

Groups of 20 larvae

0 8.3444 00+ 0.0 3.3+1.7 ( 0.0+0.0) 10 0.0+0.0 90.0+ 2.9 16.7+1.7 ( 0.0+0.0) 30 0.0+0.0 93.34 1.7 3.3+1.7 ( 0.0+0.0) 100 0.0+0.0 63.3+ 6.0 20.0+2.9 ( 0.0+0.0) 300 3.341.7 48.3+ 6.0 35.0+2.9 (25.0+2.9)

1000 0.040.0 40.0+ 2.9 50.0+2.9 (35.0+7.6)

initial 10 or 20 larvae of three replicates of experimental group. Values in

parentheses are means+SE of mortality due to incomplete molt.

a

b

: including individuals died after complete molting.

: not including individuals died due to incomplete molts into presoldiers.

364 K. Oaino, Y. Hirono et al.

6th-instar larvae 100- 10 individuals

50

6th-instar larvae 100- 20 individuals

Mean °/. of induced presoldier

7th-instar larvae 400 10 individuals

50

7th-instar larvae 100- 20 individuals

10 15 20

Days exposed

Fic. 1. Induced presoldiers in experimental groups exposed to various amounts of S-31183. Cumulative rates of induced presoldiers against initial number of larvae were plotted. Values represent means of triplicated experimental groups. Amount of S-31183 impregnated into each filter disk: O, 10 ug; @, 30 wg; A, 100 ug; a,

300 wg; O, 1,000 pg.

TaBLE 2. Timing of premolt interval for various types of molt

6th-instar larvae to

Types of molt

7th-instar larvae to

Larvae Presoldiers Larvae Presoldiers

Length oPprenol ; stage (days) Molting % total Molting % total Molting % total Molting % total 4.5 2 4.7 8 6.9 3 5.9) 5 4.6 5 9 20.9 25 21.6 7 1337) 14 13.0 51.5) 25 58.1 71 61.1 35 68.6 66 61.1 6 6 14.0 10 8.6 4 7.8 21 19.4 7 0 0.0 1 0.9 uy 2.0 0 0.0 Total 43 100.0 116 100.0 51 100.0 108 100.0 Mean (SD) 5.44 (0.39) 5.39 (0.40) 5.46 (0.42) 5.50 (0.38)

was higher than in the 6th instar larvae (Table 1).

Group size and locality of original colony did not influence the rate of induction of presoldiers or mortality (Table 1). As shown in Table 2, there was no significant difference between the length of premolt interval of larval-larval molt and that of

larval-presoldier molt with either the 6th or 7th instar larvae (Mann-Whitney’s U-test, P>0.05). In both cases, the length of premolt interval was about 5.5 days with little variance irrespective of instar. However, the 6th instar larvae became presoldier earlier than the 7th instar larvae (Mann-

Induction of Presoldier of Termite 365

Whitney’s U-test, P<0.05).

DISCUSSION

In a preliminary experiment, it was shown that the lower amounts of S-31183 (1 and 3 ug) caused just a few presoldier production accompanied with intercaste production. As shown in Table 1, with 10 and 30 ug of S-31183, the rate of presoldier production was over 90%. Moreover, the inter- caste production, which has inevitably accompa- nied the presoldier production even under the most ideal conditions so far, was not observed in this study.

The successful induction of presoldier in this study will be due to several factors. Among others, one important factor is strong JH activity of S-31183. While S-31183 is structurally different from other well-known JHAs such as methoprene and hydroprene, it exerts a higher activity than others in some insects [1, 22]. In Musca domestica, its activity in inducing supernumerary larvae is 50 times higher than methoprene [1]. High chemical stability of S-31183 [22] probably accounts, in part, for the strong activity of this JHA.

In the meantime, it is likely that the successful induction of presoldier by S-31183 in H. japonica is not only due to the strong JH activity of the analogue, but also due to the high sensitivity of this species of termite to S-31183. Actually, while in Coptotermes formosanus S-31183 is not as effective as methoprene in inducing the presoldier dif- ferentiation [2], our preliminary experiments have suggested that in H. japonica the effects of the two JHAs are just the reverse. Su and Scheffrahn demonstrated also that C. formosanus and Reticu- litermes flavipes show different sensitivity to S- 31183 [20].

High rate of the presoldier induction in this study may be also due to our experimental proce- dures employed, in which we examined ex- perimental groups composed entirely of larvae instead of natural colonies including soldiers, con- sidering that some termite species with lower population density of soldier are more sensitive to JHAs [8]. In addition, relatively low population density of soldier in the natural colony of H. Japonica (<8%) [13] may also account for its high

sensitivity to S-31183.

As shown in Fig. 1, presoldier production was saturated earlier in the experimental groups tre- ated with higher amounts (100, 300, 1,000 4g) of S-31183. This may indicate that when the concen- tration of JHA in the filter paper is higher, the level of JHA in the insect reaches the threshold earlier. Alternatively, this is a reflection of the difference of amount received by termites. Although, with these concentrations of S-31183, the rate of overall presoldier production was re- latively low, this was not because of low induction to presoldier but because of high mortality under these conditions. Indeed, many of the termites that had survived under these conditions diffe- rentiated to presoldiers (Table 1). Toxic effects of excess amount of JHA were reported by many investigators [2, 4-7, 14, 15, 19, 20].

In the 6th instar larvae, the rate of presoldier production was higher and mortality was lower than in the 7th instar larvae. This suggests that the 6th instar larvae are more resistant to the toxic effect of S-31183 and have more potentials for differentiation to presoldiers.

Presoldiers induced by S-31183 appeared 11-21 days after the first treatment (Fig. 1). It has been known that the timing of molting depends on not only the titer of JH during the intermolt period, but also many other factors such as the titer of ecdysteroid [16]. We demonstrated that both the premolt interval of larval-larval molt and larval- presoldier molt are fixed on about 5.5 days (Table 2). In addition, when the larvae treated with S-31183 underwent larval-larval molt, it always took place during the first 9 days from the first treatment. In other words, all the larvae that whitened on day 6 or later did not become larvae but presoldiers after molting 5.5 days later. This enables us to predict the fate of larvae at intermolt stage, and analyze the events underlying molting to presoldier.

ACKNOWLEDGMENTS

We are grateful to Sumitomo Chemical Company for supplying the JHA S-31183. This work was supported by a Grant-in-Aid for Scientific Research on Priority Areas (“Symbiotic Biosphere: Ecological Interaction Network Promoting the Coexistence of Many Species”

366

[03269102]) from the Ministry of Education, Science and Culture of Japan.

12

REFERENCES

Hatakoshi M, Agui N, Nakayama I (1986) 2-[1- methyl-2-(4-phenoxyphenoxy)ethoxy]pyridine as a new insect juvenile hormone analogue: Induction of supernumerary larvae in Spodoptera litura (Lepi- doptera; Noctuidae). Appl Entomol Zool 21: 351- 353

Haverty MI, Su NY, Tamashiro M, Yamamoto R (1989) Concentration-dependent presoldier induc- tion and feeding deterrency: Potential of two insect growth regulators for remedial control of the For- mosan subterranean termite (Isoptera; Rhinotermi- tidae). J Econ Entomol 82: 1370-1374

Hirono Y (1990) Histological and ecological studies on the caste differentiation of the Japanese damp- wood termite Hodotermopsis japonica. Ph.D. thesis, University of Tokyo

Howard RW (1983) Effects of methoprene on binary caste groups of Reticulitermes flavipes (Kol- lar) (Isoptera; Rhinotermitidae). Environ Entomol 12: 1059-1063

Howard RW, Haverty MI (1979) Termites and juvenile hormone analogues: A review of methodol- ogy and observed effects. Sociobiol 4: 269-278 Jones SC (1984) Evaluation of two insect growth regulators for the bait-block method of subterranean termite (Isoptera; Rhinotermitidae) control. J Econ Entomol 77: 1089-1091

Lebrun D (1990) Termite control: Biological basis. Sociobiol 17: 115-127

Lenz M (1976) The dependence of hormone effects in termite caste determination on external factors In “Phase and caste determination in insects: Endoc- rine aspects” Ed by M Luescher, Pergamon Press Inc. New York, pp 73-89

Luescher M (1952) Die Produktion und Elimination von Ersatzgeschlechtstieren bei der Termite Kalotermes flavicollis Fabr. Z Vergleich Physiol 34: 123-141

Luescher M (1961) Social control of polymorphism in termites In “Insect polymorphism” Ed by JS Kennedy, Symp R ent Soc, London pp 57-67 Luescher M (1974) Die Kompetenz zur Soldatenbil- dung bei Larven (Pseudergaten) der Termite Zootermopsis angusticollis Rev Suisse Zool 3: 710- 714

Luescher M (1976) Evidence for an endocrine control of caste determination in higher termites In

14

15

17

18

20

21

22

23

K. Oaino, Y. Hirono et al.

“Phase and caste determination in insects: Endoc- rine aspects” Ed by M Luescher, Pergamon Press Inc. New York, pp 91-103

Matsumoto T, Hirono Y (1985) On the caste composition of a primitive termite Hodotermopsis japonica Holmgren (Isoptera; Termopsidae). Sci Pap Arts and Sci univ Tokyo, 35: 211-216 Okot-Kotber BM (1980) The influence of juvenile hormone analogue (ZR515, methoprene) on soldier differentiation in Macrotermes michaelseni (Isop- tera; Macrotermitinae) Physiol Entomol 5: 407-416 Okot-Kotber BM (1980) Competence of Mac- rotermes michaelseni (Isoptera; Macrotermitinae) larvae to differentiate into soldiers under the in- fluence of juvenile hormone analogue (ZR-S515, methoprene). J Insect Physiol 26: 655-659 Okot-Kotber BM (1985) Mechanisms of caste deter- mination in a higher termite, Macrotermes michaelseni (Isoptera; Macrotermitinae). In “Caste differentiation in social insects”. Ed by JAL Wat- son, BM Okot-Kotber, Ch Noirot, Pergamon Press, Oxford, pp 267-306

Okot-Kotber BM, Prestwich GD (1991) Juvenile hormone binding proteins of termites detected by photoaffinity labelling: Comparison of Zootermopsis nevadensis with two Rhinotermitids. Arch Insect Biochem Physiol 17: 119-128

Springhetti A (1972) The competence of Kalotermes flavicollis pseudergates to differentiate into soldier. Monitore Zool Ital (NS) 6: 97-111

Su NY, Tamashiro M, Haverty MI (1985) Effects of three insect growth regulators, feeding substrates, and colony origin on survival and presoldier produc- tion of the Formosan subterranean termite (Isop- tera; Rhinotermitidae). J Econ Entomol 78: 1259- 1263

Su NY, Scheffrahn RH (1989) Comparative effects of an insect growth regulators S-31183 against the Formosan subterranean termite (Isoptera; Rhi- notermitidae). J Econ Entomol 82: 1125-1129 Wang J, Masuko K, Kitade O, Matsumoto T (1992) Allozyme variation and genetic differentiation of colonies in the damp wood termite Hodotemopsis of Japan and China. Sci Pap Arts and Sci Univ Tokyo, 42: 95-109

Wimmer Z, Romanuk M (1989) Insect juvenile hormones and their bioanalogues. Collect. Czech Chem Commun 54: 2302-2329

Yin CM, Gillot C (1975) Endocrine control of caste differentiation in Zootermopsis angusticollis Hagen (Isoptera). Can J Zool 53: 1701-1708

ZOOLOGICAL SCIENCE 10: 367-373 (1993)

© 1993 Zoological Society of Japan

A New Species of the Genus Pasiphaea (Crustacea, Decapoda, Pasiphaeidae) from the North Pacific

TOMOYUKI KoOMAI and KUNIO AMAOKA

Laboratory of Marine Zoology, Faculty of Fisheries, Hokkaido University, Hokkaido, Japan

ABSTRACT—A new pasiphaeid shrimp, Pasiphaea oshoroae sp. nov., is described based on material collected from scattered localities in the northern North Pacific Ocean. It differs from the closely allied congeners in having the short rostrum with regularly concave anterior margin, the telson being shorter than the sixth abdominal somite, the merus of the first pereopod armed with 0-3 (mostly 1) spines, the basis of the first pereopod rounded ventrodistally, and the dactylus of the second pereopod being shorter than the palm. The new species seems to be a mesopelagic inhabitant and to be widely

distributed in the northern North Pacific Ocean.

INTRODUCTION

The caridean genus Pasiphaea now contains about 50 species from the world oceans, and most of the members are known as pelagic inhabitants. In 1988, T/S Oshoro-Maru of Hokkaido Universi- ty made a collection of the pelagic organisms using a 2.0X2.5m non-closing rectangular midwater beam trawl net [7] in the eastern part of the northern North Pacific Ocean during the “Cruise 23”; six specimens of a caridean shrimp belonging to the genus Pasiphaea were captured from four stations (Table 1). Detailed examination proved that these specimens represent an undescribed species. In addition, a single specimen collected from east of the Cape Erimo, Hokkaido, Japan, was showed to be identical with the eastern Pacific specimens. In this paper we describe these speci- mens as a new species.

The specimens were preserved in 75% ethyl alcohol. Drawings were made with a camera lucida mounted on a WILD 308700 stereomicro- scope. The holotype was deposited in the National Science Museum, Tokyo (NSMT), and the para- types were deposited at the Laboratory of Marine Zoology, Faculty of Fisheries, Hokkaido Universi- ty (HUMZ). Postorbital carapace length (CL)

Accetped December 15, 1992 Received September 28, 1992

(measured from the orbital margin to the pos- teromedian margin of the carapace) is used for indication of measurements.

DESCRIPTION

Family Pasiphaeidae Genus Pasiphaea Savigny, 1816

Pasiphaea oshoroae sp. nov.

(Figs. 1-3) Type series. Holotype: NSMT-Cr 11117, female, 12.3mm CL, off Aleutian Islands,

OBT8822. Paratypes: HUMZ-C 750, 1 female, 12.4 mm CL, Gulf of Alska, OBT8811; HUMZ-C 751, 1 female, 13.8 mm CL, Gulf of Alaska, OBT 8816; HUMZ-C 752, 1 female, 11.1 mm CL, Gulf of Alaska, OBT8821; HUMZ-C 753, 2 youngs, 8.6 and 8.7mm CL, OBT8822; HUMZ-C 935, 1 female, 12.0 mm CL, east of Cape Erimo, eastern Hokkaido, Japan, OST8901. Sampling data are summarized in Table 1.

Diagnosis. Rostrum directed upward, falling short of anterior margin of carapace, with regular- ly concave anterior margin. Carapace bluntly carinate dorsally in anterior three-quarters; bran- chiostegal sinus obscure. Abdomen rounded, un- armed dorsally. Telson slightly shorter than sixth abdominal somite, with posterior margin deeply notched. Merus of first pereopod armed with 0-3

368 T. Komal AND K. AMAOKA

TaBLeE1. Sampling data of Pasiphaea oshoroae sp. nov.

Station Position Towed depth Time Date

OBT8811 54°32.1°N 0-400 m 20 : 36-21 :31 9 July 1988 151°54.5°W

OBT8816 54°58.7N 0-400 m 20 :53-21:50 16 July 1988 139°55.6 W

OBT8821 52°01.2'N 0-400 m 22 :02-23 : 06 31 July 1988 154°34.6'N

OBT8822 49°59.9'N 0-400 m 21 : 08-22 :12 4 Aug. 1988 176°55.4 W

OST8901 42°24.0'N 0-720 m 11:00-13:31 5 Sept. 1988 144°02.3'E

scl wy eZ at sinha

Fic. 1. Pasiphaea oshoroae sp. nov., holotype, female (NSMT-Cr 11117): A, entire animal in lateral view (thoracic appendages removed); B, anterior part of carapace and cephalic appendanges; C, posterior part of telson in dorsal view; D, left mandible; E, left maxillule; F, left maxilla; G, left first maxilliped; H, left second maxilliped. ANT1=antennule; ANT2=antenna.

A New Species of Pasiphaea 369

ABEHIK J (sae

Fic. 2. Pasiphaea oshoroae sp. nov., holotype, female (NSMT-Cr 11117): A, left third maxilliped; B, left first pereopod; C, same, basis and ischtum; D, same, chela; E, right second pereopod; F, same, basis and ischium; G, same, chela; H, left third pereopod; I, left fourth pereopod; J, same, dactylus; K, left fifth pereopod; L, same, dactylus; M, endopod of left first pleopod.

A B

A-F 2 mm

Fic. 3. Pasiphaea oshoroae sp. nov., paratypes, variations of rostral shape: A, HUMZ-C 750; B, HUMZ-C 751; C, HUMZ-C 752; D, E, HUMZ-C 753; F, HUMZ-C 935.

370 T. Komal AND K. AMAOKA

spines; basis with rounded ventrodistal angle. Merus of second pereopod armed with 9-11 spines or spinules; ischium with 0 or 1 spine; basis with 0- 3 spines; dactylus slightly shorter than palm.

Description of holotype. Body (Fig. 1A) strong- ly compressed. Integument thin, glabrous.

Rostrum (Fig. 1B) slightly falling short of ante- rior margin of carapace, directed obliquely up- ward, anteriorly narrowed into acute tip, anterior margin regularly concave. Carapace (Fig. 1A) bluntly carinate dorsally in anterior three-quarters, posterior quarter rounded; hepatic groove shal- low; lower branchial ridge blunt, falling far short of posterior margin of carapace; branchiostegal tooth arising just posterior to anterolateral margin of carapace, supported by short buttress; bran- chiostegal sinus obscure.

Abdomen (Fig. 1A) dorsally rounded over en- tire length, all somites lacking posterodorsal tooth. Pleuron of first somite straight ventrally, those of second to fifth somites rounded ventrally. Sixth somite about 0.5 times as long as carapace; pos- teroventral angle pointed; posterolateral process obliquely truncate. Telson (Fig. 1A, C) 0.9 times as long as sixth somite, dorsally grooved; posterior margin deeply notched, with 9 pairs of spines.

Cornea of eye (Fig. 1B) darkly pigmented, almost as wide as eyestalk.

Antennule (Fig. 1B) with peduncle reaching three-fifths of scaphocerite; stylocerite twisted, distally acute, falling considerably short of proxi- mal segment of peduncle; intermediate segment less than half length of proximal segment; outer flagellum somewhat thickened proximally, ventral excavation filled with aesthetascs.

Antenna (Fig. 1B) with basicerite bearing mod- erate ventrolateral tooth; scaphocerite with lateral margin slightly convex, distolateral tooth distinctly reaching beyond blade; carpocerite _ slightly reaching beyond distal end of proximal segment of anternnular peduncle.

Mouthparts typical of genus. Mandible (Fig. 1D) with 10 sharply pointed teeth on mesial mar- gin of incisor process; palp absent. Maxillule (Fig. 1E) with palp bearing 2 subapical setae on mesial margin, distal portion broadly rounded; distal en- dite armed with 10 stout spines on cutting edge; proximal endite truncate, with 2 setae at posterior

angle. Maxilla (Fig. 1F) with endites vestigial, without marginal setae; palp with apical seta. First maxilliped (Fig. 1G) reduced to large elongate lamina, with palp rudimentary; central part of mesial margin with sparse setae; distal lobe ovate, not articulated, fringed with setae; posterior half of lateral margin somewhat thickened, separated into two lobes by shallow, but distinct notch. Second maxilliped (Fig. 1H) five-segmented, pedi- form, lacking both of epipod and exopod; prop- odus with sparse setae but no spinules on extensor margin; dactylus armed with 2 spines distally. Third maxilliped (Fig.2A) overreaching scaphocerite by distal part of ultimate segment; ultimate segment 1.6 times as long as penultimate segment.

All pereopods with well-developed expod (Fig. 2B, E, H, I, K). First pereopod (Fig. 2B) over- reaching scaphocerite by length of fingers and one-third of palm; basis (Fig. 2C) with rounded ventrodistal angle; merus bearing 1 or 3 spines on ventral margin; carpus unarmed; dactylus (Fig. 2D) 0.6 times as long as palm, cutting edges of both fingers pectinated with minute spinules. Second pereopod (Fig. 2E) exceeding scaphocerite by length of fingers and half of palm; basis (Fig. 2F) armed with 3 spines on ventral margin; ischium (Fig. 2F) with 1 or 2 spines on ventral margin; merus bearing 11 spines including some minute ones on ventral margin; carpus with ventrodistal spine; dactylus somewhat curved inward (Fig. 2G), 0.8 times as long as palm, cutting edges with minute numerous spinules. Third pereopod (Fig. 2H) very thin, overreaching anterolateral margin of carapace by length of dactylus and distal two- thirds of propodus. Fourth pereopod (Fig. 21) much shorter than other pereopods, reaching level of slightly anterior to middlength of carapace by tip of dactylus; propodus with flexor face densely setose; dactylus (Fig. 2J) ovate, with setae on distal and flexor margins, 0.25 times as long as propodus. Fifth pereopod (Fig. 2K) falling slightly short of level of branchiostegal tooth of carapace; dactylus (Fig. 2L) ovate, 0.3 times as long as propodus, fringed with setae increasing in length distally on flexor to distal margin.

Branchial formula as shown below. Epipods absent from pereopods; arthrobranchs on first to

A New Species of Pasiphaea 371

third pereopods well developed.

Maxillipeds Pereopods

Me VS Balan Sra dans Pleurobranch paar Ss pila) A odiey ay ih Arthrobranch SS SS lend Sele Podobranch 8 — — — — — — — — Epipods — — — — — — — — Exopods ll PeS2) chomiid ait all Slee

Endopod of first pleopod (Fig. 2M) roughly ovate, with short, stout appendix interna. En- dopod of uropod (Fig. 1A) exceeding apex of telson by about distal one-fifth; exopod over- reaching apex of telson by about distal one-third, distolateral tooth reaching beyond blade.

Coloration. In fresh, entire animal reddish translucent, densely with fine red dots over body; fingers of chelae deep red. Cornea of eye brown.

Variations. The paratypes agree well with the holotype, but show some variations. The anterior margin of the rostrum varies from weakly concave to strongly concave, as depicted in Fig. 3.

In two small specimens with CL 8.6 and 8.7 mm CL, the dorsal carina on the carapace is rather obscure.

The merus of the first pereopod bears 0 or 1 spine. The merus of the second pereopod is armed with 7-12 spines on ventral margin; ischium with 0 or | spine; basis with 0-3 spines. It is found that

TABLE 2.

the number of spines on the ventral margin of the merus and ischium of the second pereopod in- creases with age in the present new species, as well as in other congeners [4, 5].

Determination of sex. In all of specimens ex- amined, the second pleopod lacks the appendix masculina. Four specimens with CL more than 11.1mm (holotype, paratypes: HUMZ-C 750; 751; 752) had a small, but distinct ovary containing ova, clearly visible posterodorsally through the transparent cuticle of the carapace. A specimen with CL 12.0mm (HUMZ-C 935) is seriously damaged in the internal structure of the cepha- lothorax, but this specimen is probably female judging from its size. In two small specimens with CL 8.6 and 8.7mm (HUMZ-C 753), the gonad was too poorly developed to determine the sex.

Etymology. Named after T/S Oshoro-Maru of Hokkaido University, which collected the present new species.

Distribution. Known with certainty only by the type series: Gulf of Alaska; Pacific off Aleutian Islands; Pacific off eastern Hokkaido.

DISCUSSION

The carinated carapace, rounded abdomen, and deeply notched posterior margin of the telson distinguish the new species from all but four spe- cies of the approximately 50 described and recog-

Comparison of characters among Pasiphaea oshoroae sp. nov. and four allied species

Items P. oshoroae sp. nov. P. alcocki P. corteziana P. liocerca P. rathbunae Anterior margin concave vertical inclined forward, concave convex and of rostrum and sinuous without concavity sinuous Dorsal carina blunt sharp blunt sharp sharp

of carapace

Branchiostegal obscure distinct distinct distinct dustinct sinus

Ventrodistal _ blunt acute blunt blunt blunt corner of ischium

of first pereopod

Number of meral 0-3, mostly 1 — 3 or 4 0 9

spines of first pereopod

Dactylus of shorter than palm equal to palm

second pereopod

Distribution northern North Bay of Bengal Pacific Indian Ocean

Reference present study [1]

shorter than palm

shorter than palm

longer than palm

California Bermuda Antarctic

Eastern Pacific Western Atlantic

[9] [2] [8]

—!))/_oO_00°cc——————————_

B72 T. Komal AND K. AMAOKA

nized species of the genus Pasiphaea. These four species are as follows: P. alcocki (Wood-Mason, 1891) from the Bay of Bengal, Indian Ocean; P. corteziana Rathbun, 1902 from off California, eastern Pacific; P. liocerca Chace, 1940 from Ber- muda, Western Atlantic; P. rathbunae (Stebbing, 1914) from the Antarctic.

The differences that help to separate the new species from the four allied species are given in Table 2. The anterior margin of the rostrum is regularly concave in the new species and P. liocer- ca, though the rostrum of the latter is considerably smaller than that of P. oshoroae [2]. In P. alcocki, it is vertical and sinuous [1]. In P. corteziana, it is inclined forward and upward without concavity or convexity [9]. In P. rathbunae, it is convex and sinuous [8].

The dorsal carina on the carapace of the P. oshoroae is blunt as in P. corteziana [9], while sharply ridged in the other three species [1, 2, 8].

The branchiostegal sinus is very obscure in P. oshoroae, while it is more distinct in the other four species [1, 2, 8, 9].

The number of spines on the merus of the first pereopod is useful in separating these species: the spines are 0 to 3 in the new species, 3 to 4 in P. corteziana [9], 9 in P. rathbunae [8], and 0 in P. liocerca [2]. Unfortunately, this character has been remained uncertain in P. alcocki.

P. alcocki appears to be unique among these five species in having the basis of the first pereopod armed with ventrodistal spine [1]. In the other species, the ventrodistal angle of basis of the first pereopod is rounded [2, 8, 9].

In P. oshoroae the dactylus of the second pereopod is slightly shorter than the palm as in P. liocerca and P. rathbunae [2, 8]. On the other hand, the dactylus is equal to the palm in P. alcocki [1], and it is distinctly longer than that in P. corteziana [9].

Recently, Hayashi [6] enumerated five Japanese species of Pasiphaea. Thus, P. oshoroae repre- sents the sixth species of the genus from the Japanese waters.

The present specimens strongly suggest that P. oshoroae is widely distributed in the northern North Pacific Ocean. Like the other members of this genus as well as related genera [3], the present

new species seems to be a pelagic inhabitant. Unfortunately, the bathymetric range could not be determined satisfactorily, since all of the speci- mens were collected by oblique tow of open net. All but one of the type series were collected from mesopelagic zone shallower than 400 m where the bottom depth greater than 2000 m. One specimen from the eastern Hokkaido (HUMZ-C 935) was captured by the bottom trawl, but it was entangled in the wing net together with other pelagic organ- isms such as Bentheogennema borealis (Rathbun, 1902) (Dendrobranchiata, Benthesicymidae) or myctophiid fishes.

ACKNOWLEDGMENTS

We are grateful to Dr. Y. Sakurai and Mr. O. Yama- mura of Research Institute of North Pacific Fisheries, Hokkaido University, for the opportunity to examine and report upon this interesting shrimp. We thank Dr. B. Kensley of the National Museum of Natural History, Smithsonian Institution for reading the manuscript and helpful advice.

REFERENCES

1 Alcock A (1901) A descriptive catalogue of the Indian deep-sea crustacea decapoda macrura and anomala in the Indian Muserum. Being a revised account of the deep-sea species collected by the Royal Indian Marine Survey Ship “Investigator”. Indian Museum, Culcatta, pp 286, pls 3

2 Chace FA Jr (1940) Plankton of the Bermuda Oceanographic Expedition. IX The bathypelagic cari- dean Crustacea. Zoologica 25: 117-209

3 Crosnier A Forest J (1973) Les crevettes profondes de l’Atlanticque oriental tropical. Faune Tropicale 19: 1-409

4 Figueira AJG (1957) Madeiran decapod crustaceans in the collection of the Museu Municipal do Funchal. I On some interesting deep-sea prawns of the family Pasiphaeidae, Oplophoridae and Pandalidae. Bol Mus Municipal Funchal 10: 22-51, pls 1-4

5 Iwasaki N (1989) Pasiphaeid shrimps from the east- ern North Atlantic and the Caribbean Sea with the description of a new species of Pasiphaea (Crustacea: Decapoda: Pasiphaeidae). Zool Meded 63: 187-203

6 Hayashi K (1990) Prawns, shrimps and lobsters from Japan (54). Family Pasiphaeidae-Genera Para- pasiphae and Pasiphaea. Aquabiol 69: 304-307 (in Japanese)

7 Nakatani T (1987) Sampling techniques for fish eggs, larvae and juveniles and their food organisms.

A New Species of Pasiphaea 373

Aquabiol 49: 108-110 (in Japanese) 9 Rathbun MJ (1904) Decapod crustaceans of the 8 Stebbing TRR (1914) Stalk-eyed Crustacea Malacos- northwest coast of North America. Harriman Alaska

traca of the Scottish National Antarctic Expedition. Exped 10: 1-190, pls 1-10

Trans R Soc Edinburgh 50: 253-307, pls 23-32

i! epaiewing (baquomth Pha ¢ oliitnese fs aot iy Rese ae

sey e as Moles as j 0 pated

i trig Mer AEP drat ane: if ihe tipe sd we vt ig is “a iG trrie, ois pelagie-zkene, shih By fay sonia greatex Mike 28 ee ea 1 rum the ts ety Ho ca s SR give be ~eaptaecay ate: pattonicty , wor hhtel, way Hut fopeth san ; wick F Oates : sen StH oh is Aervad red j cn | voi his kala is enc NON Primaries il

be Fee ievaepiiid tate» |

: 3 At KNOWL EE vii Aide ! cher Dy y Sue : . ; : _». * swore! Reseatdh InSite oF

Mens 21rd vate. for re 4, = : : ma ‘ port gnewl The witer> A 4 ¥ 7 ‘ revi dt wes I attra fae wel) i. ‘ ; and

7 i RE ERE

1 AlWock 5 (1000 hvtleami 7, =r) rile i*

- im le hie Maka ‘a ; = Par ie) ee * F(Caapt fy i heet Fitted Ppenlive eh. 2ACHipgks & - ' 4 ri " 4 a = ] - > J ~ \ i i Wy y

ZOOLOGICAL SCIENCE 10: 375-379 (1993)

[RAPID COMMUNICATION]

© 1993 Zoological Society of Japan

Immunocytochemical Localization of Troponin I and C in the Muscles of Caenorhabditis elegans

Hiroki Nakage! and TaKASHI OBINATA~

"Advanced Research Laboratory, Research and Development Center, Toshiba Corporation 1, Komukai Toshiba-cho, Saiwai-ku, Kawasaki 210, and *Department of Biology, Faculty of Science, Chiba University, 1-33 Yayoi-cho,

Inage-ku, Chiba 263, Japan

ABSTRACT—Troponin-I (TNI) and troponin-C (TNC) in muscle tissues of a nematode, C. elegans, were investigated using antibodies against the respective tro- ponin-components of Ascaris. Immunoblot analysis demonstrated that these antibodies recognize the tropo- nin components of C. elegans that have the same molecular masses as the corresponding troponin compo- nents in Ascaris body-wall muscle. In the immuno- cytochemical staining of the serial cryosections of C. elegans, the anti-ITNI antibody stained both the pharyn- geal and the body-wall muscles. On the other hand, the anit-TNC antibody reacted with the body-wall muscle but not with the pharyngeal muscle. These results suggest that variants of troponin components exist between the pharyngeal and body-wall muscles, and troponin is involved in the actin-linked regulatory system of C. elegans.

INTRODUCTION

Caenorhabditis elegans is an eligible system for studying morphogenesis of muscle. Various pro- tein components in the C. elegans muscles have been identified and investigated genetically and biochemically. Especially, contractile proteins, for example, actin, myosin, and paramyosin, have been studied intensively (for review, [19]). The existence of actin- and myosin-linked calcium regulatory systems for muscle contraction was

Accepted December 21, 1992 Received November 20, 1992 " To whom correspondence should be addressed.

suggested in C. elegans [8]. The myosin regulatory light chains of 18 kDa which may be responsible for the myosin-linked regulation have been char- acterized (for review, [1]). However, only limited attempts have so far been made to clarify the actin-linked Ca*t-regulatory proteins in C. ele- gans.

Troponin, a Ca*‘-dependent regulatory pro- tein, is localized on actin filaments and plays a significant role together with tropomyosin in the Ca?*-dependent regulation of muscle contraction [3, 4]. In the previous study, troponin composed of three components, namely troponin-T (ITNT), troponin-I (TNI), and troponin-C (TNC), was detected in the muscle of a nematode, Ascaris lumbricoides, and the components were isolated and characterized [9]. Although it is known that some invertebrate troponin, for example in asci- dian [5] and molluscan [16] muscles, activates actin-myosin interaction in a Ca*t-dependent manner, the Ascaris troponin inhibited actomyosin ATPase in the absence of Ca’*, but did not activate it in the presence of Ca?*. Moreover, the polyclonal antibodies against TNI and TNC were prepared by immunizing rabbits with each tropo- nin component and location of troponin compo- nents along this filaments at constant periodicity was demonstrated by immunoelectron microscopy [9]. In this study, the existence of the troponin components in C. elegans has been ascertained,

376 H. NAKAE AND T. OBINATA

and their immunocytochemical localization has been investigated, using the cross-reactivity of the antibodies against Ascaris troponin components.

MATERIALS AND METHODS

Caenorhabditis elegans ver. Brisol (strain N2) was grown at 20°C on NG agar plates with Escherichia coli OPS0 as feed, described by Bren- ner [2].

The affinity-purified polyclonal antibodies against Ascaris TNI and TNC were prepared as described [9]. The antibodies were absorbed with acetone-dried liver powder. Fluorescein (FITC)- labeled goat anti-rabbit IgG antibody (GAR) was purchased from Cappel Laboratories.

For all studies with nematodes, worms on agar plates were washed in M9 buffer (6.0 g NasHPO,, 3.0g KH>PO,, 5.0g NaCl and 0.25¢ MgSO,- 7H20 in 11 H,O), and collected by incubating at 4°C for 1 hr. For SDS-polyacrylamide gel electro- phoresis (PAGE), an SDS-sample buffer (2% SDS, 2mM £-mercaptoethanol, 40 mM sodium- phosphate buffer, pH 7.0) was added to the pellet of worms, and then the mixture was frozen in liquid nitrogen. The frozen mixture was dissolved immediately by incubating at 95°C for 3 min, and applied for SDS-PAGE. SDS-PAGE was carried out using 13.5% polyacrylamide gel according to Laemmli [10]. Peptides were transferred elec- trophoretically from electrophoretic gel to nitro- cellulose paper according to Towbin et al. [18], and the paper was reacted with the antibodies against Ascaris TNI or TNC. Immunoglobulin bound to the paper was detected with 'I-labelled protein A.

Squash preparation was performed basically according to Miller et al. [13]. A small droplet of the collected adult worms was placed between two 3% gelatin-coated glass slides, squashed by gentle pressure, and then immersed directly in liquid nitrogen until frozen. The slides were pried on dry ice with a razor blade. The slides were treated with ethanol cooled at —20°C for 10 min, and air-dried.

To prepare specimens for cryosections, the pellet of adult worms in M9 buffer was mixed with Tissue Tek-II O.C.T. compound (Miles Lab. Inc.), then immersed in liquid nitrogen-cooled

isopentane. Serial cryosections were cut at 8 um, and rapidly air-dired.

Immunostaining was performed as follows. The squashed specimens and the sections were fixed with phosphate buffered saline (PBS) containing 3.5% formalin for 5 min. After washing with PBS, the specimens were treated with PBS containing 1% bovine serum albumin. They were then exposed to the antibodies against TNI or TNC, and the antibody binding was detected with FITC- GAR. Each antibody incubation was performed for 1 hr at room temperature followed by thorough wshing in PBS. The antibody-treated specimens were mounted in 90% glycerol-10% 0.2 M carbon- ate buffer (pH 8.6), and the fluorescence was observed and photographed by using a fluores- cence and phase contrast microscope.

RESULTS

At first, immunoblotting was performed to

12345

kDa 67 > & 43 >

30 > 20 >

14>

FR

Fic. 1. Cross-reactivity of the antibodies against Ascar- is TNI and TNC with a whole protein extract of C. elegans. The protein extracted with an SDS-sample buffer from C. elegans (1, 2, 3), purified Ascaris TNC (4), TNC (5) were electrophoresed on 13.5% polyacrylamide gels. The extract was transferred on nitrocellulose paper, and then reacted with the anti-T'NI (2) or the anti-TNC (3) antibodies. Lane 1 shows the Amide black stainign of the extract on the nitrocellulose paper. The anti-TNI and anti-TNC antibodies reacted with the protein bands having approximately same molecular masses to Ascaris TNI (36 kDa) and TNC (20 kDa), respectively.

Immunocytochemical Localization of Troponin I and C 377

Fic. 2.

examine whether the antibodies prepared by immunizing Ascaris TNI and TNC cross-react with the corresponding proteins in C. elegans. Fig- ure 1 shows the results of the immunoblotting. Each antibody reacted with a single peptide band in the whole extract of C. elegans. Moreover, the apparent molecular masses of the proteins of C. elegans which were recognized by the anti-Ascaris TNI and TNC were almost equivalent to those of the respective troponin components of Ascaris which were described previously [9]. Therefore, it is reasonable to conclude that the antibodies against Ascaris TNI and TNC also recognize TNI and TNC in C. elegans, respectively.

Using the antibodies, the immunocytochemical localizations of TNI and TNC in C. elegans were examined by staining the squashed specimens (Fig. 2) and the serial cross-sections in the head region (Fig. 3). As shown in Figure 2, antibodies stained the squashed specimens of the animal to give striated patterns. The positively stained regions probably correspond to I-band of myogibrils in body-wall muscle, because the spots of reduced staining, which seemed to correspond to the positions of dense-bodies, were observed in the striation of the staining. Further, anti-TNI and TNC exhibit different reactivity to the muscles of nematode. The sections shown in Figure 3 con- tained two kinds of muscles namely body-wall and pharyngeal muscles. Body-wall muscles were stained positively with both antibodies. On the other hand, pharyngeal muscle exhibited the differ-

Indirect immunofluorescent stainig of squashed specimens of C. elegans. specimens were treated with anti-TNI (A) and the anti-TNC (B) antibodies.

Bar: 10 um. The squashed

ent reactivity to anti-TNI or anti-TNC antibodies. As shown in Figure 3 anti-TNI antibody stained the pharyngeal muscle clearly, while anti- TNC antibody failed to react with this muscle.

DISCUSSION

The immunoblot analyses showed that the anti- bodies which were prepared against the Ascaris TNI and TNC, recognize the protein bands of C. elegans, whose molecular masses are equivalent to TNI and TNC in Ascaris muscles. It has been also ascertained by immunocytochemical methods that the respective antigens are localized in the I-bands of the body-wall muscles of C. elegans. Consider- ing these results together with the close phylogene- tic relationships between Ascaris and C. elegans, we have concluded that the troponin regulatory system functions in C. elegans muscles as in Ascaris, and both animals contain troponin com- ponents of identical size. In this study, we could not examine the other troponin component, TNT, but it is likely that TNT also exists in the muscle of C. elegans.

We further observed that the pharyngeal muscle and the body-wall muscle exhibit different immunoreactivity to the anti-troponin antibodies. The anti-TNI antibody exhibited positive reaction with both body-wall and pharyngeal muscles, while anti-TNC antibody reacted only with body-wall muscle. These results may potentiate three inter- pretations. First, pharyngeal muscle may not have

378 H. NAKAE AND T. OBINATA

Fic. 3.

Immunocytochemical localization of the TNI and TNC in the muscles of C. elegans head region. Bar: 10 um.

Serial cross sections of worms were treated with anti-TNI (A, B) and the anti-TNC (C, D) antibodies. Florescent images (B, D) were corresponded to the phase-contrast micrographs (A, C).

TNC. Secondly, the TNC in pharyngeal muscle may be too small in amount to be detected. Thirdly, pharyngeal muscle may have the other TNC isoform(s) with different antigenicity. Pre- vious comparative biochemical studies have dem- onstrated that three troponin components mostly exist roughly in equimolar ratio in various inverte- brate animals [5, 12, 16], although they vary in size and sometimes exist in multiple isoforms [15]. Therefore, it is most likely that body-wall and

pharyngeal muscles of nematodes contain different TNC isoforms, although further investigation is needed to reach a definit conclusion. We can not eliminate the possibility that TNIs in the two muscles share common antigenic site(s) but differ somehow in moecular structure.

Besides the troponin conponents, it is konwn that pharyngeal muscle differs from body-wall muscle in several points. Differences in myosin isoforms between these muscles are particularly

Immunocytochemical Localization of Troponin I and C 379

remarkable. It has been demonstrated that four myosin heavy chain isoforms, named myoA, myoB, myoC and myoD, are distinguishable electrophoretically in the wild-type C. elegans, and the pharyngeal muscle possesses myoC and myoD, while the boyd-wall muscle contains myoA and myoD [6, 14, 17, 20]. Obliquely striated myofibrils have been observed in the body-wall muscle [7], but the pharyngeal muscle exhibits different structural organization [for reivew, 1]. The pharyn- geal muscle seems to be functionally distinct somewhat from the body-wall muscle. Previous studies suggested that nematode muscle may be controlled dually by actin-linked and myosin- linked regulatory systems [8, 11]. Contractile activity of the two muscles of the nematode could be altered by the difference in myosin and tropo- nin isoforms.

In conclusion, we have demonstrated that the body-wall muscle contain the troponin compo- nents, INI and TNC, of approximately the same size as the counterparts of Ascaris. Thus, it is obvious that the contraction of body-wall muscle is under the control of the actin-linked troponin regulatory system. In addition, it is very likely that pharyngeal and body-wall muscles possess distinct troponin components. It is a matter of interest for furture studies how these protein variants are derived, whether encodered by different genes or generated from a single gene.

ACKNOWLEDGMENTS

Caenorhabditis elegans N2 was provided by the Caenorhabditis Genetic Center, which is funded by the NIH National Center for Research Resources (NCRR), through Dr. Joji Miwa of NEC Fundamental Research Laboratories. The authors thank Dr. H. Takano- Ohmuro for technical advice about the immunob- lot analyses. One of the authors (H.N.) also thanks Drs. K. Ando, H. Nakanishi, and Y. Ishimori for advice in many phases of this work, and Ms. M. Sugano and Mrs. K. Nakae for helpful comments on the manuscript. This work was partially supported by research grants from the

Ministry of Education, Science and Culture of Japan.

REFERENCES

Anderson P (1989) Annu Rev Genet 23: 507-525 2 Brenner S (1974) Genetics 77: 71-94 3 Ebashi S, Ebashi F (1964) J Biochem Tokyo 55: 604-613 4 Ebashi S, Endo M (1968) In “Progress in biophysics and molecular biology” Ed by JAV Butler and D Noble, Vol 18, Pergamon Press, Oxford, New York, pp 123-183 5 Endo T, Obinata T (1981) J Biochem Tokyo 89: 1599-1608 6 Epstein HF, Waterston RH, Brenner S (1974) J Mol Biol 90: 291-300 7 Francis GR, Waterston RH (1985) J Cell Biol 101: 1532-1549 8 Harris HE, Tso M-YW, Epstein HF (1977) Bioche- mistry 16: 859-865 9 Kimura K, Tanaka T, Nakae H, Obinata (1987) Comp Biochem Physiol 88B: 399-407 10 Laemmli UK (1970) Nature, Lond 227: 680-685 11 Lehman W, Szent-Gyérgyi AG (1975) J Gen Physiol 66: 1-30 12 Lehman W, Regenstein JM, Ransom AL (1976) Biochim Biophys Acta 434: 215-222 13 Miller DM III, Ortiz I, Berliner GC, Epstein HF (1983) Cell 34: 477-490 14 Miller DM III, Stockdale FE, Karn J (1986) Proc Natl Acad Sci USA 83: 2305-2309 15 Ohshima S, Komiya T, Takeuchi K, Endo T, Obinata T (1988) Comp Biochem Physiol 90B: 779- 784 16 Ojima T, Nishita K (1986) J Biol Chem 261: 16749- 16754 17 Schachat FH, Harris HE, Epstein HF (1977) Cell 10: 721-728 18 Towbin H, Staehelin T, Gordon J (1979) Proc Natl Acad Sci USA 76: 4350-4354 19 Waterston RH (1988) In “The nematode Caenorhabditis elegans” Ed by WB Wood and the community of C. elegans Researchers, Cold Spring Harbor Lab., New York, pp 281-335 20 Waterston RH, Moerman DG, Baillie DL, Lane TR (1982) In “Disorders of the motor unit”, Ed by DM Schotland, John Wiley, New York, pp 747-760

AVE SY

Ad

> bow 1 vores

CF SANG tbat ermal

Wie. | Te EERE ‘ eae pe eae * i ‘ : ; ie)

oS

vy i : | ae iigetarh ) t

case tara ae —

ZOOLOGICAL SCIENCE 10: 381-383 (1993)

[RAPID COMMUNICATION]

© 1993 Zoological Society of Japan

Biosynthesis of 17a,20a-dihydroxy-4-pregnen-3-one from 17a-hydroxyprogesterone by Goldfish (Carassius auratus) Spermatozoa

Kryosu1 AsAHina’, Katsumi Arpa? and Tr1zo Hicasut!

‘College of Agriculture and Veterinary Medicine, Nihon University, Setagaya-ku, Tokyo 154, and *Department of Fisheries, Faculty of Agriculture, The University of Tokyo, Bunkyo-ku, Tokyo 113, Japan

ABSTRACT — Goldfish (Carassius auratus) spermatozoa were incubated with 4-'*C-labeled 17a-hydroxyproge- sterone at 22°C for 60min. After extraction of the steroids from the incubation medium, they were sepa- rated by thin-layer chromatography (TLC). Radioactive steroids were localized on the TLC-plates by auto- radiography. The major metabolite showed an Rf value lower than that of 17a,208-dihydroxy-4-pregnen-3-one (17a,206-P) but co-migrated with 17a,20a-P, the isomer of 17a,208-P. The TLC behaviour of the metabolite was identical with authentic 17a,20a-P after both acetylation and oxidation. The identity of the metabolite with 17a,20a-P was finally confirmed by repeated crystalliza- tion to constant specific activity. These results indicate that goldfish sperm cells show 20a-hydroxysteroid dehy- drogenase activity.

INTRODUCTION

The mature sperm cells of the rainbow trout (Oncorhynchus mykiss) produce very high levels of the steroid 17a,208-dihydroxy-4-pregnen-3-one (17a,20B-P) when incubated with 17a- hydroxyprogesterone (17a-P) in vitro [11, 14]. In contrast, it was recently found that spermatozoa of the common carp (Cyprinus carpio) produce 17a,20a-dihydroxy-4-pregnen-3-one (17a,20a-P), the isomer of 17a,208-P, when incubated with 17a-P [1]. In the present study we investigated which 17a-P metabolites were formed by sperma- tozoa of another cyprinid fish, the goldfish (Caras-

Accepted February 12, 1992 Received December 21, 1992

sius auratus).

MATERIALS AND METHODS

Chemicals

4-'4C-Labeled 17a-hydroxyprogesterone (*17a- P; specific activity, 1.85 GBq/mmol) was obtained from Du Pont (UK). All organic solvents were of special grade and were obtained from the Tokyo Kasei Chemical Co. (Tokyo, Japan). 17a,20a- Dihydroxy-4-pregnen-3-one (17a,20a-P), and other non-radioactive steroids and chemicals were obtained from the Sigma Chemical Company (St. Louis, U.S.A.).

Fish

Male goldfish were kept with females in 5x0.8 <0.6m tanks under natural temperature and photoperiod conditions at the University of Tokyo; they were fed with commercial trout pellets. In April, 1991, five spermiating males were selected from the tank. Milt samples were collected in glass tubes from their genital pores by applying gentle pressure to the abdomen. Milt from all five fish was pooled and transported on ice to the department of fisheries, faculty of agricul- ture and veterinary medicine, Nihon University.

Incubation

Sperm cells were washed twice with carp Ringer

382 K. ASAHINA, K. AIDA AND T. HIGASHI

and collected with centrifugation at 2000 rpm for 15 min. The incubation was carried out in 5 ml of carp culture medium [8] containing 240 mM NADPH, “17a-P (1X 10° cpm), and approximately 100 mg wet weight of washed spermatozoa. They were incubated in a 50 ml round-bottomed flasks for 1 hr at 22°C. The incubation was stopped by adding 15 ml dichloromethane and shaking vigor- ously for one minute.

Extraction and TLC

The procedures were the same as described previously [2]. In short, the incubation medium containing the spermatozoa was extracted twice with dichloromethane. The extracts were pooled and dried with anhydrous Na,SO,. After evapora- tion of the solvent, the residues were subjected to thin-layer chromatography (TLC) with six un- labeled steroids for standards; progesterone, 17a- P, androstenedione, testosterone, 17a,208-P and 17a,20a-P in a benzene: acetone (4:1, v/v) solvent system. After development, the standard steroids were detected under UV light and radioactive spots on the chromatograms were detected auto- radiographically by exposing the TLC plate to a sheet of medical X-ray film for one week. Each spot was then separately scraped off the plate. Radioactive compounds were eluted from the silica gel with a mixture of chloroform and ethanol (1:1, v/v), and aliquots were used to determine the radioactivity by liquid scintillation counting. The metabolites were identified on the basis of the following criteria:

(1) Isopolarity with authentic steroids on TLC developed in two systems of the following solvent mixture: benzene: acetone, 4:1 (v/v) and dichlo- romethane : diethyl ether, 5:2 (v/v).

(2) Identical chemical behaviour of the radioactive metabolites with the authentic prepa- rations after acetylation [15] and oxidation [10].

(3) Constant specific radioactivity of crystals after repeated crystallization of the radioactive metabolite with the corresponding authentic pre- paration.

RESULTS

In the initial thin-layer chromatogram de-

veloped in a benzene: acetone (4:1, v/v) system, only one clear band (Rf. =0.24) corresponding to authentic 17a,20a-P was detected, in addition to the unchanged radioactive substrate, *17a-P. Fol- lowing scintillation counting it was found that the conversion of radioactivity to 17a,20e-P was approximately 34 times greater than that obtained from the area of authentic 17a,208-P (Table 1). The mobility of the metabolite was also identical with authentic 17a@,20a-P on TLC using dichloro- methane : diethyl ether (5:2, v/v) as solvent sys- tem. Furthermore, it behaved identically with authentic 17a,20a-P after both acetylation and oxidation. The identity of the metabolite was finally confirmed by repeated crystallization to constant specific activity with authentic 17a,20a-P (Table 2).

TaBLeE1. Jn vitro conversion of 17a-hydroxy- progesterone by goldfish (Carassius auratus) sper- matozoa

Metabolite Yield

Unchanged substrate (oS ole

(17a-hydroxyprogesterone)

17a,20a-dihydroxy-4-pregnen-3-one 64.6

17a,208-dihydroxy-4-pregnen-3-one** 1.9

* pmol of steroid/100 mg tissue/60 min ** Tentatively identified. TABLE 2. Recrystallization of the radioactive

metabolite with the authentic preparation of 17a,20a-dihydroxy-4-pregnen-3-one for identifica-

tion Specific radio activiry (cpm/mg)

Before recrystallization 371 After recrystallization

Ist (dichloromethane-n-hexane) 333 2nd (dichloromethane-n-heptane) 361 3rd (chloroform-n-hexane) 350

DISCUSSION

Our results indicate that mature sperm cells of the goldfish contain the enzyme 20a-hydroxy- steroid dehydrogenase (20a-HSD) and produce

Steroid Metabolism by Fish Sperm 383

significant levels of 17a,20@-P when incubated with 17a-P for one hour in vitro. In addition, low levels of radioactivity co-migrated with authentic andro- stenedione. However, there was too little radio- activity for further analysis.

17a,20a-P is reported to be synthesized in vitro in the ovarian tissues of a marine teleost, the dab [5]. In the dab, high levels of this steroid were detected after hCG treatment of both female [4] and male fish [3]. Because the 20a-HSD activity was suggested to reside in dab spermatozoa [5], the sperm cells may be responsible for the synthesis of plasma 17a@,20a-P. Recently, 17a,20a-P was also shown to be synthesized by goldfish ovary in vitro [6].

We did not detect a clear 208-HSD activity in goldfish spermatozoa after 60 min of incubation. However, increased plasma levels of 17a,208-P were detected with radioimmunoassay in male goldfish in relation to spawning behaviour [7] and after the administration of GTH [12]. Since a low 206-HSD activity was detectable after a long term (18 hr) incubation of common carp spermatozoa [1], it cannot be ruled out that goldfish sperm cells also show low 206-HSD activity. This, however, was almost undetectable after our short term incubation.

17a,208-P has been reported to be effective by inducing spermiation in amago-salmon and goldfish [13], and the acquisition of sperm motility in salmonid fish [9]. From our results, however, it is reasonable to hypothesize that 17a,20a-P also has some role in regulating spermiation in the goldfish. Recently we have found that during spawning the 17a,20a-P plasma levels increase to almost the same levels as those of 17a,208-P in both male and female goldfish (Asahina er al., unpublished results). The reason(s) for the discre-

pancy between the in vitro production and the plasma levels of these two steroids is currently under investigation.

ACKNOWLEDGMENTS

We are grateful to Dr. R. Schulz at University of Utrecht for critical reading of the manuscript.

REFERENCES

1 Asahina K, Barry TP, Aida K, Fusetani N, Hanyu I (1990) J Exp Zool 255: 244-249 2 Asahina K, Suzuki K, Aida K, Hibiya T, Tamaoki B (1985) Gen Comp Endocrinol 57: 281-282 3 Canario AVM, Scott AP (1991) Gen Comp Endo- crinol 83: 258-264 4 Canario AVM, Scott AP (1990) Gen Comp Endo- crinol 77: 177-191 5 Canario AVM, Scott AP (1989) Gen Comp Endo- crinol 76: 147-158 6 Kime DK, Scott AP, Canario, AVM (1992) Gen Comp Endocrinol 83: 258-264 7 Kobayashi M, Aida K, Hanyu 1(1986) Gen Comp Endocrinol 62: 72-79 8 Kobayashi M, Aida K, Hanyu I (1986) Nippon Suisan Gakkaishi 52: 1153-1158 9 Miura T, Yamauchi K, Takahashi H, Nagahama Y (1992) J Exp Zool 261: 359-363 10 Poos GI, Arth GE, Beyler RE, Sarett LH (1953) J Amer Chem Soc 75: 422-429 11 Sakai N, Ueda H, Suzuki N, Nagahama Y (1989) Biomed Res 10: 131-138 12 Suzuki Y, Kobayashi M, Aida K, Hanyu I (1988) J Comp Physiol B 157: 753-758 13 Ueda H, Kambegawa A, Nagahama Y.(1985) Gen Comp Endocrinol 59: 24-30 14 Ueda H, Kambegawa A, Nagahama Y (1984) J Exp Zool 321: 435-439

15 Zaffaroni A, Burton RB (1951) J Biol Chem 193: 749-767

Utes ea Wee a)

7 etd 1 SSL Me ai oa ip ;

crane ‘ echanalabead mast! :

Akar cy. Qvtved a elt as \eetig ¢ REA

pa abirwtae. tae arts Beat lien Ssandedel cet ae s : Sathiioke itive — <orthnentasrithuttlddran

t

‘ 4 2 : 4 ANISM. ols 274, ie hel Vignal? y ‘ , 4 = 4 atiidiey set’ se ; 3: 4 eetrrdyi a i isi: Ae Dereeetact? im uN tail te eetT i ‘ 1c.) i NTS «4 l 4 si) Fare i — Ox eat eee ak Swett tel i i ity Lita 3 he i ry fy ie Thm ico ; 3

Gtk cer oharel gorctt ¢ rascal. si iowina & sphibfalaen | cou wit or ci berms inattrongs dbae a “epanae iekae hen we Sane. blonoleoaiy! louise ile “Tih olpensehortl toed amenntandite

Hiv: Gebel adie it fo hoosteentntiogetriatn Get to man Geyderaieste “aritiigaws TAME ormahione yal ‘ef afen Web rennt

fiety: confirmed. fe. te nat “piegt ihuith 5 CAs Tito a tiatidGial to aja he wile | FA0CeTT 4p. elavel wren, | #2, at (aie: ORS OTE ery, ihe ih) mune vsathely syneiag ant

at cme 121) HO Io ee Wal 4aol a Tae Sia AIGA je Pad Henne

prene TESTA ay iyo

owon wut nivion Oe

; aT ' crre! BAP nS ware Aria

fA _

MAO] biotleah Soo ges, malice demeanor qn

& NOC PIG GE noted balled ke: gaya

ain “MAGL OUT “igeli tee actignlts PM ce eth) “ef

att iinet sved sae

“uO Ries

weston! dovel amenig Tok

we lo Sait) SB ake ire) deities shee

(t.20 {wiauesy1 af

pe >

iO Rs fied ates

We SAT VR

f : Lh ti

D eve | O m e nt Published Bimonthly by the Japanese Society of Pp Developmental Biologists Distributed by Business Center for Academic

Growth & Differentiation Societies Japan, Academic Press, Inc.

Papers in Vol. 35, No. 2. (April 1993)

14. N. Moriya, H. Uchiyama and M. Asashima: Induction of Pronephric Tubules by Activin and Retinoic Acid in Presumptive Ectoderm of Xenopus laevis

15. Y. Sadakane and Y. Iwao: A Factor That Interrupts the Cell Cycle at G» or Prophase is Present in Full-Grown Frog Oocytes

16. Q. Yang, P. D. Kingsley, D. J. Kozlowski, R. C. Angerer and L. M. Angerer: Immunoche- mical Analysis of Arylsulfatase Accumulation in Sea Urchin Embryos

17. N. D. Holland, L. Z. Holland, Y. Honma and T. Fujii: Engrailed Expression during Development of a Lamprey, Lampetra japonica: A Possible Clue to Homologies between Agnathan and Gnathostome Muscles of the Mandibular Arch

18. C. Tatone, R. Carotenuto, R. Colonna, C. Chaponnier, G. Gabbiani, M. Giorgi and C. Campanella: Spectrin and Ankyrin-like Proteins in the Egg of Discoglossus pictus (Anura): Their Identification and Localization in the Site of Sperm Entrance versus the Rest of the Egg

19. K. Yakiguchi-Hayashi and K. Kitamura: The Timing of Appearance and Pathway of Migration of Cranial Neural Crest-derived Precursors of Melanocytes in Chick Embryos

20. T. Saiki and Y. Hamaguchi: Difference between Maturation Division and Cleavage in Starfish Oocytes: Dependency of Induced Cytokinesis on the Size of the Aster as Revealed by Transplantation of the Centrosome

21. E. Koyama, S. Noji, T. Nohno, F. Myokai, K. Ono, K. Nishijima, A. Kuroiwa, H. Ide, S. Taniguchi and T. Saito: Cooperative Activation of HoxD Homeobox Genes by Factors from the Polarizing Region and the Apical Ridge in Chick Limb Morphogenesis

22. G. Peaucellier, K. Shartzer, W. Jiang, K. Maggio and W. H. Kinsey: Anti-peptide Antibody Identifies a 57 kDa Protein Tyrosine Kinase in the Sea Urchin Egg Cortex

23. H. Yoshida, S. Nishikawa, H. Okamura, T. Sakakura and M. Kusakabe: The Role of c-kit Proto-oncogene during Melanocyte Development in Mouse. In vivo Approach by the In utero Microinjection of Anti-c-kit Antibody

24. S. Fujiwara: Temporal and Spatial Expression of a Gene for the Nuclear Protein Hgv2 in Embryos and Adults of the Ascidian Halocynthia roretzi

25. Akira Yanagi: Nuclear Differentiation in Paramecium caudatum: Analysis by the Monoclonal Antibody against a Micronuclear Antigen

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.

-shape _ : Y ergonomic design.

OLYMPUS!

Este

YE:S is the answer.

The BX Series was developed with your needs and our ideals in mind.

Y-shaped ergonomic design ensures maximum comfort. Easy operation. Easy recognition. And fatigue-free use. To the question of what SYSTEM MICROSCOPE constitutes the ideal microscope from the user’s point of view, Our answer is

a unique Y-shaped design. We invite you to discover unprecedented operational ease and comfort.

Excellent optics provide superior image quality. Developed through the application of more than 70 years of experience and what the latest technology has to offer, Olympus’ own UIS (Universal Infinity System)

infinity-corrected optical system assures exceptional image quality. You benefit BX40/BX50 from both higher contrast and outstanding flatness throughout the field of view. System versatility answers diverse research requirements. N E W

Frame rigidity and stability have been increased dramatically through use of computer simulation (FEM analysis). You'll find that reliability has been enhanced, too, when the time comes for system expansion, via devices for image processing, microscopic light metering, photography and so on.

OLYMPUS OPTICAL CO.,LTD. OLYMPUS (JAPAN) CO.,LTD. For free catalog, please contact: OLYMPUS (JAPAN) CO.,LTD. Ryumeikan Bldg., 3-4, Kanda-surugadai, Chiyoda-ku, Tokyo Tel. (03) 3251-8971

The Ultimate Name in Micromanipulation Ease of operation, and the most advanced

Can be combined with New WR & MX series cma

Model MxX-2 3-D Micromanipulator

Model MX-1 3-D Micromanipulator

Model WR-87

One Dimensional Aqua Purificate Micromanipulator

Model SR-6 Model PP-83 Stereotaxic Instrument for Rat Glass Microelectrode Puller

* * + Send enquiries for request for MODIFICATIONS or IMPROVEMENTS * * * Physiological, Pharmacological, Zoological & Neurosciences Research Equipments

(NARISHIGE) NARISHIGE SCIENTIFIC INSTRUMENT LAB.

\ PGRRIN 7, 9-28 KASUYA 4-CHOME SETAGAYA-KU, TOKYO 157, JAPAN PHONE (INT-L) 81-3-3308-8233, FAX (INT-L)81-3-3308-2005

CABLE:NARISHIGE LABO, TELEX, NARISHIGE J27781

- a <i ne a ae :

qr otis <a. NON sve gasenee ni ampere bsonsvbs t2om sapbas note

(Contents continued from back cover)

Kao, Y. H., P.S. Alexander, V. V. C. Yang, J. Y. L. Yu: Annual patterns of testicular development and activity in the chinese bull- frog (Rana rugulosa Wiegmann)

Singtripop, T., T. Mori, M. K. Park, K. Shira- ishi, T. Harigaya, S. Kawashima: Prolactin binding sites in normal uterus and the uterus with adenomyosis in mice ................ 353

Environmental Biology and Ecology Ogino, K., Y. Hirono, T. Matsumoto, H.

Ishikawa: Juvenile hormone analogue, S- 31183, causes a high level induction of pre- soldier differentiation in the Japanese damp- wood termite

Systematics and Taxonomy Komai,T., K. Amaoka: A new species of the genus Pasiphaea (Crustacea, Decapoda, Pasiphaeidae) from the North Pacific ....367

ZOOLOGICAL SCIENCE

VOLUME 10 NUMBER 2

APRIL 1993

CONTENTS

REVIEWS Asano, M., K. Shiokawa: Behavior of ex- ogenously-introduced DNAs in_ early embryos of Xenopus laevis Robertson, D. R.: Comparative aspects of intestinal calcium transport in fish and amphibians

ORIGINAL PAPERS

Physiology

Schlechter, N. L., L. S. Katz, S. M. Russell, C. S. Nicoll: Physiological evaluation of the role of the liver as a mediator of the growth- promoting action of somatotrophin

Hara, A., C. V. Sullivan, W. W. Dickhoff: Isolation and some characterization of vitel- logenin and its related egg yolk proteins from Coho Salmon (Oncorhynchus kisutch) ...245

Harada, A., M. Yoshida, H. Minakata, K.

Nomoto, Y. Muneoka, M. Kobayashi: Structure and function of the molluscan myoactive tetradecapeptides ............. 257 Noguchi, M., K. Okamoto, S. Nakamura: Effects of proteolytic digestion on the con- trol mechanism of ciliary orientation in cili- ated sheets from Paramecium ...........- 267

Seidou, M., K. Ohtsu, Z. Yamashita, K. Nari- ta, Y. Kito: The nucleotide content of the octopus photoreceptor cells: No changes in the octopus retina immediately following an intense light flash

Kobayashi, W.: Effect of osmolality on the motility of sperm from the lamprey, Lampet- ra Japonica

Biochemistry Nakae, H., T. Obinata: Immunocytochemical

localization of troponin I and C in the mus- cles of Caenorhabditis elegans (RAPID

COMMUNICATION) ................... 375 Shimizu, M., H. Kudo, H. Ueda, A. Hara, K. Shimazaki, K. Yamauchi: Identification

and immunological properties of an olfactory system-specific protein in Kokanee Salmon (Oncorhynchus nerka)

Developmental Biology Komatsu, M., T. Shosaku: Development of the brittle star, Ophioplocus japonicus H. L. Clark. I

Reproductive Biology Amano, T., Y. Okita, T. Yasumoto, M. Hoshi: Maitotoxin induces acrosome reaction and histone degradation of starfish Asterina pecti- nifera sperm Harada, T.: Reproduction by overwintering adults of water strider, Aquarius paludum (Fabricius) Asahina, K., K. Aida, T. Higashi: Biosynthe- sis Of 17a,20a-dihydroxy-4-pregnen-3-one from 17a-hydroxyprogesterone by goldfish (Carassius auratus) spermatozoa (RAPID COMMUNICATION)

Endocrinology

S/Govbetti, A., M. Zerani, M. M. Di Fiore., V. Botte: Prostaglandins and sex steroids from reptilian (Podarcis sicula sicula) ovarian follicles at different developmental stages 321

Zairin, M. Jr., K. Asahina, K. Furukawa, K. Aida: HCG-injected male walking catfish Clarias batrachus

Plasma steroid hormone profiles in

(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 April 15 Printed by Daigaku Letterpress Co., Ltd., Hiroshima, Japan

oH

ZOOLOGICAL SCI ENCE

An International Jou

ZOOLOGICAL SCIENCE

The Official Journal of the Zoological Society of Japan

Editors-in-Chief:

seule : The Zoological Society of Japan: Seiichiro Kawashima (Tokyo)

: Toshin-building, Hongo 2—27-2, Bunkyo-ku, Tsuneo Yamaguchi (Okayama) Tokyo 113, Japan. Phone (03) 3814-5675 Assistant Editors: Olean Akiyoshi Niida_ (Okayama) President: Hideo Mohri (Chiba) Meal Sakai (Okayama) Secretary: Hideo Namiki (Tokyo) Sumio Takahashi (Okayama) Treasurer: Makoto Okuno (Tokyo) Librarian: Masatsune Takeda (Tokyo)

Editorial Board:

Howard A. Bern (Berkeley) Walter Bock (New York) Aubrey Gorbman (Seattle) Horst Grunz (Essen) Robert B. Hill (Kingston) Yukio Hiramoto (Chiba) Susumu Ishii (Tokyo) Yukiaki Kuroda (Tokyo) John M. Lawrence (Tampa) Koscak Maruyama (Chiba) Roger Milkman (Iowa) Kazuo Moriwaki (Mishima) Richard S. Nishioka (Berkeley) Chitaru Oguro (Toyama) Tokindo S. Okada (Okazaki) Andreas Oksche (Giessen) Hidemi Sato (Nagano) Mayumi Yamada (Sapporo)

Ryuzo Yanagimachi (Honolulu) Hiroshi Watanabe (Tokyo)

ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and communications in the broad field of Zoology. The journal appears bimonthly. An annual volume consists of six numbers of more than 1200 pages including an issue containing abstracts of papers presented at the annual meeting of the Zoological Society of Japan.

MANUSCRIPTS OFFERED FOR CONSIDERATION AND CORRESPONDENCE CONCERN- ING EDITORIAL MATTERS should be sent to: Dr. 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 1993 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 10: 385-392 (1993)

REVIEW

Proliferation of Anterior Pituitary Cells in

© 1993 Zoological Society of Japan

Aging and Longevity in the Rat

HirRosHI OoOKA

Department of Cell Biology, Tokyo Metropolitan Institute of Gerontology, 35-2 Sakaecho, Itabashi-ku, Tokyo 173, Japan

ABSTRACT—Some of the hormones secreted from the anterior pituitary are related to the potency of survival during the aging process. The proliferation of thyrotrophs affects the lifespan of the rat through the activity of the thyroid gland. Adrenocorticotropic hormone promotes the deterioration of brain function in aged rats. Average prolactin levels and the incidence of prolactinoma increase in aged rats. A longitudinal study showed that the increase in prolactin levels represents the precancer-

ous state for a prolactinoma.

Culture experiments indicate that the proliferation of acidophils is

regulated by hypothalamic growth hormone-releasing factor and other hormones, which are possibly responsible for the age-related change in acidophil populations.

ANTERIOR PITUITARY CELLS AND THE RATE OF AGING IN THE RAT

Theories on the biological mechanisms of aging include the alteration of the humoral regulatory system in which the hypothalamo-pituitary axis has central importance. In this chapter I will deal only with the role of the anterior pituitary in the aging of structures and functions related to the survival of individuals. The anterior pituitary is deeply concerned in age-related changes in the reproduc- tive system. With regard to these aspects, other review articles can be consulted [2, 20, 28, 77].

In his series of research on surgically hypo- physectomized rats, Everitt [19] has shown that several cirteria for aging (hardening of tail tendon collagen, increase in protein excretion in urine, renal and cardiac hypertrophy, etc.) are prevented by hypophysectomy. Despite the retarded rate of aging in these criteria, hypophysectomy shortens the duration of life. Later work showed, however, that replacement therapy with cortisone acetate prolongs the lifespan of hypophysectomized rats to even longer than the lifespan of intact rats, thereby

Received December 25, 1992

suggesting that adrenocorticotropic hormone (ACTH) is an essential life-maintaining factor in survival [21]. From these results he concluded that the pituitary secretes “aging hormone(s)” which accelerate the rate of aging processes.

Chronic food restriction has been the most effec- tive means for prolonging the lifespan of rats [30, 46]. The onset of various kinds of age-associated disease is also delayed by food restriction. Some studies have shown that food restriction inhibits the secretion of pituitary hormones [13, 48], sug- gesting that the anti-aging effect of food restriction may partly be ascribed to the reduced secretion of “aging hormone(s)”.

One candidate for such an aging factor is the pituitary-thyorid axis. A very early study showed that mice fed desiccated thyroid throughout their liver had shorter lifespans than controls [67]. Exp- osure to low temperature (9°C), which was ex- pected to induce the secretion of thyroxine (T4), shortened the lifespan compared to control rats maintained at 28°C [35]. The periodic acid— Schiff (PAS) stainability of the anterior pituitary decreased in tryptophan-deficient rats, which also showed retardation of aging. The concentrations of T4, 3,3 ,5-triiodothyronine (T3), and thyrotropic

386 H. Ooka

hormone were lower in these tryptophan-deficient animals than in control rats [57].

More direct and precise experiments were car- ried out using the neonatally Ty4-treated hypo- thyroid rats of both sexes. Thyrotrophs in the anterior pituitary of neonatal rats proliferate actively and increase rapidly in number [55]. When rats were injected intraperitoneally with 1-2 pg Ts per g body weight 5 times during the first 10 days of life, the division of thyrotrophs was sup- pressed by the exogenous T,. There was almost no increase in the number of thyrotrophs in rats at 10 days of age. The capacity to increase the number of thyrotrophs was restored after discontinuation of T, injection; the pituitaries in adult rats treated neonatally with T, contained almost the same density of thyrotrophs as the controls [55]. The levels of T, in the treated rats were lower than in control rats, as reported in previous studies [3, 5], throughout almost the whole lifespan, indicating that the function of the thyroid gland was impaired by the inhibition of thyrotroph proliferation during neonatal T,-treatment. The levels of T3 were also lower than controls in young rats. These artifitial hypothyroid rats of both sexes lived longer than control rats [56]. On the other hand hyperthyroid rats, produced by continuous administration of Ty in the drinking water, had shorter lifespans than controls [58]. The life-shortening effect of T, was observed even when the rats were treated with T, during only the first or second half of life. Ty, treatment was not effective, however, when treat- ments were confined to the senescent period (more than 26 months old). These results indicate that T, does not directly induce the senescent diseases that are the causes of death, but accelerates the aging rate during the young and middle-aged periods [58].

The aging factors suggested by the experiments of Everitt seem not to include ACTH, because he supplied glucocorticoid to hypophysectomized rats that showed aging retardation. However, it has been reported that the ACTH-adrenal cortex axis accelerates some aspects of the aging process. Wexler [91] has reported the premature aging of the cardiovascular system and many other tissues in hyperadrenocorticoid rats produced by repeated breeding. Recent evidence indicates that some

aging phenomena in the hippocampus and other brain regions are modulated by glucocorticoids [23]. Sapolsky [71] emphasized that the hippocam- pus is a primary CNS target for glucocorticoids which induce neuronal loss in the aged hippocam- pus. Minimal damage in the hippocampus was observed in adrenalectomized rats after microin- jection of toxins (kainic acid or 3-acetylpyridine). Enhanced damage was induced in ticosterone-treated rats. Adrenalectomy appears to protect the brain against age-related reductions in neuronal density and age-related increases in glial density [44]. However, part of the effect of adrenalectomy seems to be due to prolonged neu- ral stimulation resulting from elevated ACTH, rather than to a reduction in steroids per se, since the morphologic correlates of brain aging were reduced and maze learning was improved by treat- ment with a non-steroidogenic ACTH analog.

COor-

AGE-ASSOCIATED DEVELOPMENT OF HYPERPROLACTINEMIA AND PROLACTINOMA

The concentration of prolactin (PRL) in the blood of female rats increases with age [78, 84]. The incidence of spontaneous prolactinoma also increases in the pituitary of aged rats, but there is a wide variation among strains [93]. High tumor incidence in one strain is generally associated with age-related hyperprolactinemia, not only in the individuals which have pituitary hyperplasia, but also in the rats with histologically normal pituitar- ies [89]. Although aged male rats of many strains also have high PRL levels, the magnitude differs depending on the strain [7, 65, 70, 90]. In mice, CS57BL/6J females over 14 months of age have a high tumor incidence [22], but PRL levels are not elevated in male mice at ages up to 28 months [17].

Even in strains in which aged female rats have very high average PRL levels, the individual varia- tions are very large with some rats showing low PRL values at advanced ages [89]. Therefore it is impossible to determine any chronological pattern for the development of hyperprolactinemia or pro- lactinoma in individual rats from the cross- sectional data. A longitudinal study on PRL levels in female Wistar rats by serial sampling of blood

Proliferation of Pituitary Cells and Aging 387

from the same individuals was useful for such purpose [59]. A prolactinoma inevitably develops within 3 months in rats with blood PRL levels higher than 70 ng/ml. On the other hand, rats with no tumors at the time of death showed very low PRL levels throughout life (Fig. 1). These facts indicate that moderate hyperprolactinemia is the precancerous state prior to the development of prolactinoma. The duration of tumor growth is 2- 3 months (Fig. 1). The highest generation of new tumors is observed at around 27 months of age. Some rats with a large tumor survive for more than 10 months, indicating that prolactinoma is not always a direct cause of death for rats [59]. This finding may have some relation to a report that increased immunoreactive PRL molecules show low biological activity when assayed for proli- feration-inducing potency on Nb, lymphoma cells [12]. Moreover, we can expect anti-aging effects of PRL since the administration of PRL prevents an age-related decrease in dopamine receptors in the rat striatum [68]. PRL-treated rats recovered from motor dysfunctions.

As with many other aging criteria, the age- associated increase in PRL levels is suppressed in

i=

Ao)

©

iS

®

1S)

i=

jo}

oO

=

x3)

1c

2

a

Age

Fic. 1. Schematic representation of the longitudinal

changes in the PRL levels of female Wistar rats with large tumors (A), small tumors (B), and no tumors (C) at the time of death. The time of tumor development was accompanied by a sudden rise in PRL level (T). Prolactinomas always developed in hyperprolactinemic rats within 3 months. When the period between tumorigenesis and examination of the pituitary was less than 2 months, the pituitary had a small tumor (less than 1 mm), indicating that the duration of tumor growth is about 2 months.

rats by food restriction [81]. Synthesis of PRL messenger RNA is also reduced in energy- restricted C3H/SHN F,-hybrid mice [41]. These suppressions might be due to direct inhibition of the production or secretion of pituitary hormones by underfeeding, since the level of growth hor- mone (GH), which decreases with age, is remark- ably reduced by food restriction [62]. Measure- ments of PRL levels in rats refed after 24 months of food restriction by an intermittent food supply, however, showed that low PRL levels are main- tained for more than 9 weeks, indicating that the prevention of PRL increase in food-restricted rats is not due to a direct effect of underfeeding, but represents a retardation of aging in the hypotha- lamo-pituitary axis (Fig. 2).

70

60

(é) (o)

Prolactin (ng/ml)

0 2 5 i) Time of refeeding (weeks)

Fic. 2. PRL levels in food-restricted and refed male Fischer 344 rats. Experimental rats (open circles) were fed every other day from 1 to 24 months of age, with a total food intake approximately 60% that of controls (closed circles). Radioimmunoassay for PRL was performed on serially sampled blood from the same individuals. Protection against hyperpro- lactinemia continued for more than 9 weeks of ad libitum refeeding.

The close association of hyperprolactinemia with prolactinoma suggests the possibility that the number of mammotrophs increases in the pituitary of hyperprolactinemic rats. The volume fraction occupied by the mammotrophs in the pituitary increases with age [89]. The ratio of mammotroph number to total secretory cells also increases in the

388 H. Ooxa

pituitary of hyperprolactinemic rats [15], although no increase was observed in aged individuals with low PRL levels [59]. Moreover, many previous studies have shown that the synthesis and secretion rate of PRL in each mammotroph do not increase, or rather decrease in the pituitary of aged rats [86]. These results indicate that changes in the regula- tion of cell number, besides the regulation of hormone secretion [6, 18, 27, 33, 37, 40, 53, 54, 82], are important in the mechanism for the de- velopment of hyperprolactinemia.

REGULATION OF ACIDOPHIL PROLIFERATION IN AGING AND TUMORIGENESIS

Mammotrophs in the pituitary of adult animals can increase through cell proliferation, or by con- version from somatotrophs [38]. The possibility of transdifferentiation of mammotrophs and somatot- rophs is suggested by the presence of cells which secrete both PRL and GH (mammosomatotrophs) [29, 75]. Estrogen induces the production of PRL in somatotrophs, and increases the number of mammosomatotrophs in primary pituitary cell cul- ture obtained from adult male rats [10] and in a clonal pituitary cell line, GH; [9]. Insulin and IGF-1 stimulate the secretion of PRL, and inhibit the secretion of GH, in GH; cells [49, 61]. On the other hand, cortisol antagonizes insulin in the secretion of PRL by GH; cells [61]. Cortisol also increases the number of both somatotrophs and mammosomatotrophs in cultured bovine pituitary cells by transdifferentiation [39]. More detailed informations about mammosomatotrophs are available in a recent review [86].

The increase in mammotrophs by cell prolifera- tion, however, is most likely in age-related hyper- prolactinemia, since tissue hyperplasia is a general feature of the precancerous state. The prolifera- tive activity of mammotrophs is stimulated by estrogen, which also enhances the secretion of PRL [32, 47, 78]. Both the mitotic index and the number of cells in S-phase increase after in vivo estradiol treatment of rats [34, 60, 87]. Prolonged administration of natural or synthetic estrogen induces prolactinomas in rats and mice [11, 92]. The sustained plasma estradiol levels associated

with persistent vaginal cornification in middle-aged rodents are suspected by some authors to be a factor responsible for both the elevated PRL levels and the high incidence of prolactinoma in females [23]. In constant estrous Long-Evans rats older than 25 months, moderately high estrogen levels are accompanied by high PRL concentration in the blood [45]. However, plasma estrogen levels are rather lower in constant estrous C57BL/6J mice than in the estrous and diestrous stages of cyclic young in which estrogen levels are very high in the proestrous stage [22]. Moreover, only a small portion of rats older than 25 months are in con- stant estrus [1, 45], while many rats show an increase in PRL levels and prolactinoma develop- ment between 25 and 30 months [59]. In some strains of rats, for example, Wistar and Fischer 344 rats, age-related hyperprolactinemia and prolacti- noma are observed even in males [7]. These facts suggest the presence of other factors involved in the increase in mammotroph number.

Hypothalamic dopamine is a well-known inhibi- tor of PRL secretion. Sulpiride, a dopamine antagonist, activates both DNA synthesis and the mitosis of mammotrophs in vivo, and bromocrip- tine, a stable dopamine agonist, suppresses them [34, 87]. The content of dopamine in the hypotha- lamus decreases in aged rodents [16, 80]. Old female rats with spontaneous prolactinomas, as well as young females with prolactinomas induced by prolonged estrogen treatment, show damaged tuberoinfundibular dopaminergic neurons [72-74]. These reports suggest that dopamine plays an important role in the regulation of mammotroph proliferation. However, the sensitivity of mam- motroph DNA synthesis to amantadine, a com- pound that causes an increase in dopamine synth- esis, seems to be lower than that ot PRL secretion [42]. Dopamine concentration in the hypophyseal portal blood has been reported to decrease with age in both male [26] and female [64] rats, but a more recent study has shown that the rate of dopamine secretion is markedly elevated in aged rats [31]. Dopamine content in the anterior pituit- ary is also higher in aged rats than in young rats [63].

Among secretory cells in the anterior pituitary of adult rats, mammotrophs are the only cells to

Proliferation of Pituitary Cells and Aging 389

proliferate actively in culture with ordinary basal medium containing no hypothalamic factors [4]. Recent studies on cultured anterior pituitary cells have revealed the direct influence of hormones, especially hypothalamic peptides, on the prolifera- tion of these cells. Billestrup et al. [8] reported that DNA synthesis in cultured somatotrophs is stimulated by the addition of growth-hormone releasing factor (GRF), and suppressed by soma- tostatin. Frawley and Hoeffler [24] studied the effects of various hypothalamic hormones on the number of mammotrophs and somatotrophs in 5-day-old rats by a reverse-plaque assay. They reported that the relative number of mammo- trophs to somatotrophs increases in cultures con- taining GRF or luteinizing hormone-releasing hor- mone (LHRH). The proliferation of both cell types is stimulated by corticotropin-releasing fac- tor (CRF), but thyrotropin-releasing hormone (TRH) stimulates the proliferation of other cell types. Treatment of pituitary reaggregate cell culture from immature (14-day-old) rats with LHRH or neuropeptide Y increased *H-thymidine incorporation into mammotrophs and cortico- trophs, whereas these peptides decreased the num- ber of labeled somatotrophs [88]. The actions of these peptides were suggested to be mediated by specific paracrine growth factors released from gonadotrophs.

The effects of various hypothalamic hormones on the proliferation of pituitary cells from adult rats were examined in culture using 8-month-old female Wistar rats [79]. In a basal medium consist- ing of Eagle’s MEM and 10% fetal bovine serum, mammotrophs proliferated actively, and no in- crease in the number of somatotrophs was observed. The proliferative situation was com- pletely reversed by the addition of GRF (Fig. 3). This result suggests that the relative number of acidophils is regulated by hypothalamic GRF through inverse responses of mammotrophs and somatotrophs to GRF. GH and PRL belong to the same gene family, and both secretory cells diffe- rentiate from the same precursor cells [38]. The mechanism by which the expression of one gene changes the characteristics of cell proliferation is worth investigating. Another possibility is that GRF suppresses the proliferation of mammot-

200

180

160

140

120

Relative cell number

100

80 ojo" WO oO we we Concentration of GRF (M)

.3. Changes in the numbers of mammotrophs (open circles) and somatotrophs (closed circles) from adult rats during a 6 days in culture with varying concen- tration of GRF. Relative cell number represents the percentage of cells at the start of culture period.

F

a)

rophs indirectly through an unknown inhibitor secreted by somatotrophs or other cell types under the stimulation of GRF. Failure to detect GRF binding sites on the mammotroph surface suggests this possibility [51]. If this assumption is correct, the postulated inhibitor should be specific to mam- motrophs since the proliferation of somatotrophs is not suppressed. In contrast with the effect of dopamine, GRF does not inhibit the secretion of PRL [14, 76, 83].

The content of GRF in the median eminence decreases markedly with aging in the rat [52]. This is not conclusive evidence that GRF secretion into the protal blood is reduced, since a decreased content of a substance in the median eminence could occur with either increased or decreased secretion. However, the fact that the secretion of GH [62] decreases with aging, while total volume of GH cells does not change [86], suggests a decrease in GRF secretion in aged rats. It is possible that a decrease in GRF secretion plays a role in the change in the relative number of acidophils in aged rats.

The addition of LHRH into the culture acceler- ates the rate of mammotroph proliferation (sub- mitted for publication). It has been reported that the content [69] and secretion rate [85] of LHRH increase gradually in middle-aged and old rats. High luteinizing hormone (LH) levels in the blood

390 H. Ooxa

of aged mice also suggest an increase in LHRH secretion in these animals [25]. On the other hand, no increase is observed in the blood LH levels of aged female rats [45]. However, a decrease in the responsiveness to LHRH may prevent the increase in LH release in aged rats, since the density of LHRH receptor is reduced in the pituitary cells of aged rats [50]. From these data it can be assumed that LHRH also contributes to the age-related increase in the number of mammotrophs.

Although the proliferation of mammotrophs in adult rats is activated by LHRH, the stimulation of PRL secretion by physiological doses of LHRH seems to be restricted to very young animals [66]. In older rats, the PRL-secreting response to high, nonphysiological doses of LHRH has been re- ported, and this phenomenon is suggested to be mediated by angiotensin II produced by paracrine secretion [36, 43]. But angiotensin II had no distinct effect on the proliferation of mammo- trophs in culture experiments.

The suppressive effect of bromocriptine on mammotroph proliferation has been confirmed in vitro (unpublished data). TRH stimulates the secretion of PRL [27, 33], but showed no effect on the proliferation of mammotrophs.

The effects of hypothalamic and ovarian hor- mones on the proliferation of acidophils, as shown by in vitro studies, are summarized in Fig. 4. Age-related changes in the production and secre- tion of these hormones probably cause the changes in the acidophil populations of the anterior pituit- ary in aged rats, the increase in mammotrophs and the decrease in somatotrophs.

GRF

GRF mon is GH Series FRM Palsy SRIF E

Fic. 4. Regulation of the proliferation of mammot- rophs (PRL) and somatotrophs (GH) by GRF, LHRH, dopamine (DA), somatostatin (SRIF), and estrogen (EB).

ACKNOWLEDGMENTS

The author is grateful to Dr. T. Shinkai, Department of Cell Biology, Tokyo Metropolitan Institute of Geron- tology, for his cooperation in the preparation of this review. Hormones and antibodies for radioimmunoassay and immunostaining were kindly supplied by the Nation- al Institute of Diabetes and Digestive and Kidney Dis- eases, USA, and Gunma University, Japan.

REFERENCES

1 Aschheim P (1976) In “Hypothalamus, Pituitary, and Aging” Ed by AV Everitt, JA Burgess, Charles C. Thomas, Springfield, pp 376-418

2 Aschheim P (1983) In “Neuroendocrinology of Aging” Ed by J Meites, Plenum Press, New York, pp 73-101

3. Azizi F, Vagenakis AG, Bollinger J, Reichelin S, Braverman LE, Ingbar SH (1974) Endocrinology 94: 1681-1688

4 Baker BL, Reel JR, Van Dewark SD, Yu YY (1974) Anat Rec 179: 93-106

5 Bakke JL, Lawrence N, Wilber JF (1974) Endocri- nology 95: 406-411

6 Bernton EW, Beach JE, Holaday JW, Smallridge RC, Fein HC (1987) Science 238: 519-521 Bethea CL, Walker RF (1979) J Geront 34: 21-27

8 Billestrup N, Swanson LW, Vale W (1986) Proc Natl Acad Sci USA 83: 6854-6857

9 Boockfor FR, Hoeffler JP, Frawley LS (1985) En- docrinology 117: 418-420

10 Boockfor FR, Hoeffler JP, Frawley LS (1986) Am J Physiol 250: E103-105

11 Brawer JR, Sommerstein C (1975) Am J Anat 144: 57-87

12 Briski KP, Sylvester PW (1990) Neuroendocrinol 51: 625-631

13 Campbell GA, Kurcz M, Marshall S, Meites J (1977) Endocrinology 100: 580-587

14 Casanueva FF, Brorras CG, Bruguera B, Lima L, Muruais C, Tresguerres JAF, Devesa J (1987) En- docrinology 27: 517-523

15 Chuknyska RS, Biackman MR, Hymer WC, Roth GS (1986) Endocrinology 118: 1856-1862

16 Damarest KT, Riegle GD, Moore KE (1980) Neuroendocrinol 31: 222-227

17 De vellis JS, Swerdloff RS (1977) Endocrinology 101: 1310-1318

18 Dymshitz J, Laudon M, Ben-Jonathan N (1992) Neuroendocrinol 55: 724-729

19 Everitt AV (1976) In “Hypothalamus, Pituitary, and Aging” Ed by AV Everitt, JA Burgess, Charles C. Thomas, Springfield, pp 68-85

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

45

Proliferation of Pituitary Cells and Aging

Everitt AV, Meites J (1989) J Geront 44: B139-147 Everitt, AV, Seedman NJ, Jones F (1980) Mech Ageing Dev 12: 161-172

Felicio LS, Nelson JF, Finch CE (1980) Exp Geront 15: 139-143

Finch CE, Landfield PW (1988) In “Handbook of the Biology of Aging” Ed by CE Finch, EL Schneider, Van Nostrand Reinhold, New York, pp 567-594

Frawley LS, Hoeffler JP (1988) Peptide 9: 825-828 Gee DM, Flurkey K, Finch CE (1983) Biol Reprod 28: 598-607

Gudelsky GA, Mansel DD, Porter JC (1981) Brain Res 204: 446-450

Hanna S, Shin SH (1992) Neuroendocrinol 55: 591- 599

Harman SM, Talbert GB (1985) In “Handbook of the Biology of Aging” Ed by CE Finch, EL Schneider, Van Nostrand Reinhold, New York, pp 457-510

Hashimoto S, Fumagalli G, Zanini A, Meldolesi J (1987) J Cell Biol 105: 1579-1586

Holeman AM, Merry BJ (1986) Biol Rev 61: 329- 368

Hotta H, Ito H, Matsuda K, Sato A, Tohgi H (1991) Jap J Physiol 41: 317-325

Huang HH, Marshal S, Meites J (1976) Biol Re- prod, 14: 536-543

Hyde JF, Murai I, Ben-Honathan N (1987) Endoc- rinology 121: 1531-1539

Jahn GA, Machiavelli GA, Kalbermann LE, Szijan I, Alonso GE, Burdman JA (1982) J Endocrinol 94: 1-10

Johnson HD, Kintner LD, Kilber HH (1963) J. Geront 18: 29-36

Jones TH, Brown BL, Dobson RPM (1988) J. Endocrinol 116: 367-371

Karanth S, McCann SM (1991) Proc Natl Acad Sci USA 88: 2961-2965

Karin M, Castrillo JL, Theill LE (1990) Trends Genet 6: 92-96

Kineman RD, Faught WJ, Frawley LS (1992) En- docrinology 130: 3289-3294

Kochman K, Kowatch MA, Roth GS, Blackman MR (1987) Gerontologist 27: 147A

Koizumi A, Tsukada M, Masuda H, Kamiyama S, Walford RL (1992) Mech. Ageing Dev 64: 21-35 Krawczyk Z, Lyson K, Stawowy A, Pisarek H, Stepien H (1990) Neuroendocrinol 51: 632-636 Kubota T, Judd AM, McLeod RM (1990) J Endoc- rinol 125: 225-232

Landfield PW, Baskin RK, Pitler TA (1981) Science 214: 581-584

Lu KH, Hopper BR, Vargo TM, Yen SSC (1979) Biol Reprod 21: 193-203

46 47 48

73

74

391

Masoro EJ (1988) J Geront 43: B59-64

Meites J (1980) Neurobiol Aging 1: 141-143 Meites J, Goya R, Takahashi S (1987) Exp Geront 22: 1-15

Melmed S (1984) J Clin Invest 73: 1425-1433 Miyamoto A, Maki T, Blackman MR, Roth GS (1992) Exp Geront 27: 211-219

Morel G (1991) Endocrinology 129: 1497-1504 Morimoto N, Kawakami F, Makino S, Chihara K, Hasegawa M, Ibata Y (1988) Neuroendocrinol 47: 459-464

Nagy GM, Frawley LS (1990) Endocrinology 127: 2079-2084

Nikolics K, Masoro AJ, Szonyi E, Ramachandran J, Seeburg PH (1985) Nature 316: 511-517

Ooka H (1986) In “Pars Distaris of the Pituitary Gland—Structure, Function and Regulation” Ed by F Yoshimura, A Gorbman, Elsevier Science Pub- lishers, Amsterdam, pp 29-35

Ooka H, Fujita S, Yoshimoto Y (1983) Mech Ageing Dev 22: 113-120

Ooka H, Segall PE, Timiras PS (1978) Mech Ageing Dev 7: 19-24

Ooka H, Shinkai T (1986) Mech Ageing Dev 33: 275-282

Ooka H, Shinkai T (1992) Arch Geront Geriat Suppl 3: 287-293

Perez RL, Machiavelli GA, Romano MI, Burdman JA (1986) J Endocrinol 108: 399-403

Prager D, Yamashita S, Melmed S (1988) Endocri- nology 122: 2946-2952

Quigley K, Goya R, Nachreiner R, Meites J (1990) Exp Geront 25: 447-457

Reymond MJ, Donda A, Lemarchand-Beraud J (1989) Hormone Res 31: 32-38

Reymond MJ, Porter JC (1981) Brain Res Bull 7: 69-73

Riegle GD, Meites J (1976) Proc Soc Exp Biol Med 151: 507-511

Robberecht W, Andries M, Denef C (1992) Neuroendocrinol 56: 185-194

Robertson TB (1928) Aust J Exp Biol Sci 5: 69-88 Roth GS, Joseph JA (1988) Gerontol 34: 22-28 Rubin BS, Elkind-Hirsh K, Bridges RS (1985) Neurobiol Aging 6: 309-315

Saksena SK, Lau IF (1979) Exp Aging Res 5: 179- 194

Sapolsky RM (1987) Trends Neurosci 10: 346-349 Sarkar DK, Gottschall PE, Meites J (1982) Science 218: 684-686

Sarkar DK, Gottschall PE, Meites J (1983) In “Neuroendocrinology of Aging” Ed by J Meites, Plenum Press, New York, pp 353-376

Sarkar DK, Gottschall PE, Meites J (1984) Endocri- nology 115: 1269-1274

75

76 77

78

79

80

81

82

83

84 85

392

Sasaki F, Iwama Y (1988) Endocrinology 123: 905- 912

Schally AV (1978) Science 202: 18-28

Schneider EL (1978) “The Aging Reproductive System” Raven Press, New York

Shaar CJ, Euker JS, Riegle GD, Meites J (1975) J Endocrinol 66: 45-51

Shinkai T, Ooka H, Noumura T (1991) Neurosci Lett 123: 13-16

Simpkins JW, Mueller GP, Huang HH, Meites J (1977) Endocrinology 100: 1672-1678

Snyder DL, Wostmann BS, Pollard M (1988) J Geront 43: B168-173

Spangelo BL, Judd AM, Isakson PC, McLeod RM (1989) Endocrinology 125: 575-577

Stachura ME, Tyler JM, Farmer PK (1988) Endoc- rinology 123: 1476-1482

Steger RW (1981) Neurobiol Aging 2: 119-123 Steger RW, De Paolo LV, Shepherd AM (1985) Neurobiol Aging 6: 113-116

H. Ookxa

86 87

88

89

90

91

92

93

Takahashi S (1992) Zool Sci 9: 901-924 Takahashi S, Kawashima S (1987) Zool Sci 4: 855- 860

Tilemans D, Andries M, Denef C (1992) Endocri- nology 130: 882-894

Van Putten LJA, Van Zwieten MJ, Matheij JAM, Van Kemenade JAM (1988) Mech Ageing Dev 42: 75-90

Van Putten LJA, Bonga SEW, Van Kemenade JAM (1988) Mech Ageing Dev 45: 253-276 Wexler BC (1976) In “Hypothalamus, Pituitary and Aging” Ed by AV Everitt, JA Burgess, Charles C. Thomas, Springfield, pp 333-361

Yokoro K, Furth J, Haran-Ghera N (1961) Cancer Res 21: 178-186

Yokoro K, Ito A (1981) In “Hormone Related Tumors” Ed by H Nagasawa, K Abe, Japan Scien- tific Society Press, Tokyo/Springer Verlag, Berlin, pp 3-21

ZOOLOGICAL SCIENCE 10: 393-410 (1993)

© 1993 Zoological Society of Japan

REVIEW

The Peculiar Case of the Giants of Oxytricha bifaria (Ciliata, Hypotrichida): a Paradigmatic Example of Cell Differentiation and Adaptive Strategy

NICOLA Ricci and ROSALBA BANCHETTI

Dipartimento di Scienze dell’Ambiente e del Territorio Via A Volta, 6-56100 PISA, Italy

INTRODUCTION

In a stage of biological research such as today’s, characterized by the extreme specialization of each scientist, in turn due to the need of obtaining ever new information about Life, the quantity of new reports available daily, even within a relatively “restricted” area such as protozoology is so large and evergrowing, that most of us cannot study and master much of the recent knowledge, even in the field of their own specific interest. It is our opinion that every day it becomes more and more practi- cally urgent and culturally necessary for all of us to make all possible attempts to reconsider each of our results in a double perspective, if we want to avoid the danger of creating isolated, limitedly useful pieces of Science. On one hand, we must place them correctly in the wider context of Na- ture, but try to interpret them in the general light of evolutionary biology on the other. This is the rational of the present review article, in which the whole story of the giants of Oxytricha bifaria will be described overall and then discussed (and the relative implications examined) in its double-faced valence, namely both as an example of cell dif- ferentiation and of reversible carnivorism.

The idea of a review article about nature and biology of the giants of Oxytricha bifaria (Ciliata, Hypotrichida) arose when the overall picture of the process became sufficiently exhaustive and selfstanding as a consequence of a complex round

Received January 5, 1993

of experiments recently carried out [59, 60, 65, 66, 69, 71]: it now lends itself to be rationalized in terms of a general model which, in turn, can be used fruitfully to penetrate further into the general adaptive strategies which shaped the history of life.

Protozoa are as unique and peculiar as they are precious, for approaching many of the most fun- damental problems of biology with ever new inves- tigation tools and a fresh mind. What makes them unique among the other organisms is their double sided nature: each single protozoon is indeed a perfect eukaryotic cell (“a physiological unit”) and a complete organism (“a selective unit”) at the same time. This character of theirs enables us to pass from the cellular to the organismic-adaptive level directly, without approximations. Protozoa, moreover, were the first eukaryotes to appear in the primeval Ocean: they are very ancient organ- isms, 2.2 billion years old [47]. This trait of Protozoa is a highly relevant one, once we consider that they reached first, the eukaryotic organiza- tion, exploiting then all its potentialities and realiz- ing the widest range of variations on the theme “eukaryotic cell”, adapting their morphology and physiology to match very different environmental challenges. We could say that Protozoa reach the highest peaks of complexity at the cellular level, in much the same way as metazoa do at that of the organismic level [3]. This consideration, then, should urge us to make a sort of “Copernican revolution” in Biology, leading us to recognize their primigenial nature and to use them more correctly than to date. From this point of view,

394 N. Ricci AND R. BANCHETTI

indeed, it was Protozoa which first found the correct solutions of most general adaptive prob- lems (regulation of differentiative processes, ex- change of information through cellular interac- tions, cell locomotion as behaviour capable of reaching optimal conditions of environmental fac- tors), so that we should learn to study and to understand them according to the characters of their own biology, instead of continuing to consid- er them as just “simple metazoa” or even as sort of “simple” free-living lymphocytes, neurons, etc. This thought will prove to be quite a relevant one in section III, when the cell differentiation aspects of the process object of this article will be dis- cussed.

In our opinion Protozoa are among the best biological material for a wide interdisciplinary investigation enabling us to collect information which can be compared and integrated directly and immediately. The clearest tendency illustrating this theory of ours is the eco-ethological approach recently proposed for protozoa by Ricci [57]: he tends to link the study of their behaviour [54], with that of their ecology [16] in an attempt to compre- hend their general adaptive philosophy. The first line of research has its roots in the studies of cell locomotion [33, 34], but draws new strength and perspectives when the electrophysiological studies [40, 41] are brought in [56]. The second half of this study, namely the ecological one was suggested by Fenchel in his masterpiece [16], and it led to a new understanding of the importance of the protozoa’s microbial loop in aquatic environments, of their totally antiintuitive world [53, 62, 63, 64], of their usefulness in monitoring the environmental condi- tions [7, 17, 55].

This example shows how powerfully these rather disregarded organisms can be investigated by means of widely different technical approaches to achieve very useful results in widely different fields of Biology: what we would like to do in the present paper is to describe the complete picture of the morpho-functional steps leading a normal Oxy- tricha bifaria to differentiate into a gigantic organ- ism, to discuss then this story in the more general context of “cell differentiation” problematics, on one hand, and in that of the different secondary consumer adaptive strategies, on the other.

I The giant of O. bifaria: a multi-step adaptation

The data henceforth described as a single body of information, were published in six papers [59, 60, 65, 66, 69, 71] as the results of partial rounds of experiments carried out from 1982 to 1990, in which all the technical and methodological details can be found concerning both the culturing of O. bifaria and the various experimental approaches. The steps in the story of the giant in O. bifaria will be discussed in series, putting together all the available information, each indicated by one of the above numbers, to facilitate bibliographic con- sultation.

O. bifaria is a freshwater hypotrich ciliate [8], with a typical ellipsoidal body (~110X60 um) differentiated into a convex dorsal surface and into a more or less planar ventral one, where all the composed ciliary organelles are placed: they can be distinguished into (a) somatic locomotory organelles (namely the fronto-ventral, transverse and marginal cirri) and (b) peristomial organelles (namely the anterior-left Adoral Zone Mem- branelles, AZM, and the mid-ventral Undulating Membranelles (UM) (Fig. 1). The nuclear appar- atus is formed by one macronucleus divided into two pieces and by two micronuclei, each close to either macronuclear envelope. O. bifaria lives in freshwater canals, springs, creeks, rivers, ponds, lakes, typically feeding on bacteria: its cell cycle [12, 58] usually lasts about 7-8 hr at about 20°C. As to its life cycle it is characterized by quite a long immaturity period (~ 180 binary fissions, Ricci & Cetera, unpublished results), following a sexual phenomenon (conjugation), by a maturity period (which lasts no longer than 2-3 years, under laboratory conditions, Ricci & Banchetti, unpub- lished results) and by a rather short senescence, (progressive loss of the mating competence) un- failingly leading to the death of the clone. The general description of O. bifaria’s biology cannot be considered complete, without taking into account the nature of its habitat; in the natural environment where we find this species, indeed, we measured very different conditions at different times as to temperature (—2°C to 36°C), to pH (5.7-7.5) and dissolved oxygen (3.5 mg solved oxygen/ml), not to speak of the water itself,

Giants as Adaptive Differentiations 395

Fic. 1. The normal Oxytricha bifaria, in dorsal (1) and ventral (3) views; the giant of the same species, in dorsal (2) and ventral (4) views. AZM: Adoral Zone Membranelles; UM: Undulating Membranelles; MC: Marginal Cirri; FVC: Fronto-Ventral Cirri; TC: Transverse Cirri; DB: dorsal bristles. The bar represents 50 um.

396 N. Ricci AND R. BANCHETTI

which, quite frequently in the summer, can dis- appear completely (Ricci & Banchetti, unpub- lished results). We could say that the environment of O. bifaria (that of inland, often impermanent waters) is characterized by the widest (and wildest) variations of the physical, chemical and biological parameters describing it: in agreement with such an extreme environmental variability, the species we have been studying since 1972, shows a similar- ly extreme adaptability, being capable of changing dramatically both its shape and its physiology as it is required to withstand the challenges periodically met in such an environment. Although more than 95% of its life span is spent in the normal morpho- physiological state, three other states are possible and actually found in our samples collected in nature in different periods of the year (Fig. 2). The conjugating pair is a peculiar morpho- functional state formed by two oxytrichas united side by side, which carry on their sexual processes,

g Conjugating fair

Fic. 2. mal organisms, capable of differentiating conjugat- ing pairs, cysts and giants, according to the internal/ external conditions of the system.

O. bifaria spends ~95% of its lifetime as nor-

through a complex cascade reaction: (a) two-step cell interactions leading to (b) cell membrane fusion, (c) trigger of micronuclear meiosis, actual (d) meiotic divisions of micronuclei, (e) exchange of pronuclei (=cross fertilization), (f) separation of the partners [52]. Such a complex sexual process mediates the rejuvenation of the popula- tion: it can be adaptively interpreted as the periodical solution (triggered by proper environ- mental conditions) found by the species to the problem of overcoming its main biological bot- tleneck, namely the absolute need of producing genetic recombinants and of fighting progressive ageing.

Cysts represent the second possibility O. bifaria evolved to match the challenges periodically met in its widely changing environment: they represent a sort of life-boat for the species when severe, prolonged environmental stresses (drought in sum- mer, for instance) affect natural populations. The dramatic morpho-physiological changes a cell undergoes to produce the sophisticated structure that a cyst is, were studied mainly from ultras- tructural [75], cytochemical [68] and adaptive s./. points of view [67].

On the basis of the above integrated results concerning the nature and story of both pairs and cysts, a thorough, specific giant formation was then undertaken.

A Normal populations

This is the phase preceding the onset of any specific induction of the phenomenon “giant- formation”.

O. bifaria is a species whose natural populations can produce gigantic organisms: only 1-2% of the strains collected in nature, however, show such a capability. A certain strain, which proves to be capable of differentiating giants, under proper conditions, tends to maintain such a trait steadily in time.

Many strains have been collected and stocked in Lab cultures and each clone has its own giant producing potentiality, which can be measured by two parameters: (i) number of giants produced and (ii) time lag before the appearance of the first giant. According to these two traits, the clones we used in the last 8 years can be ordered as it follows:

Giants as Adaptive Differentiations 397

C9>S9>S6, C9 being the strain producing the largest number of giants in the shortest time. No exhaustive theory has been proposed so far to account for the nature of the character “giant formation”: no clear-cut genetic inheritance of this trait has yet demonstrated, for any clone, isolated and cultured in our Lab. Only very few cells can produce giants ex-novo even in the richest popula- tion of O. bifaria belonging to a giant-producing strain: these few cells have such a capability only for a relatively short time, being then substituted by other individuals, in turn only temporarily predisposed for such an adaptive task (Ricci and Pelamatti, unpublished results). It has been found, indeed, that although hundreds of giants can be observed in the experimental populations, by far the largest part of their number is produced through binary fission of preexisting giants, while only very few (2-3 per day) are differentiated ex-novo from normal individuals, even in cultures as dense as 1600 cells/ml. While the problem of the nature of this labile predisposition to produce giants seems today far from being accounted for, both at clonal and at individual level, it might be simpler to answer to the question “why can so few, and so frequently changing oxytrichas differentiate a giant?” In our opinion, a certain degree of loss of fitness s./. is to be expected for the “ready-to- differentiate” organisms, while a clear advantage is evident for the gigantic forms. If this is true, the labile determination of only few oxytrichas to differentiate giants might be considered as the best solution to limit the reduction of fitness within the population (due to the acquisition of such a poten- tial state which only very rarely actually leads to the advantageous giantic form) without missing the adaptive convenience brought to the population by the occurrence of such a differentiated gigantic state. The temporal lability of the determination to produce giants could have been acquired to avoid the risks of a progressive reduction of gene- tic variability in time, if only a few cells (and only they) should be destined to become giants: all the possible genomic combinations of a certain genetic pool, indeed, have the same, although small, prob- ability of undergoing the gigantic adventure.

This general hypothesis seems to us supported and strengthened by the finding that, as the indi-

viduals of a certain population are periodically determined to undergo gigantic differentiation, there is no convenience for any population to have a large number of organisms predisposed to cell differentiation: the need of producing a large population of giants, whenever a proper rich pabu- lum should exist, is matched by the cell-binary- fission-mechanism, a fairly convenient and fast asexual reproductive process.

Terminological foreward: (i) the time lag be- tween the onset of the conditions suitable for the formation of giants and the actual formation of the first giant has been defined “induction period” and indicated “At FG”; (ii) the induction period is formed by an “activation period”, followed by the successive “predation”, namely the actual feeding of activated cells on the potential preys: (iii) within the activation, two successive periods have been distinguished: the “Early-Activation Window” (= EAW, roughly corresponding to the first third of At FG) and the “Late-Activation Window” (= LAW, occurring during the second third of the At FG): the last third of the At FG corresponds to predation. All these concepts are illustrated in Fig. 3.

B_ The Early Activation Window (EAW)

Giants of O. bifaria form whenever favourable conditions are created [1], namely whenever an overcrowded population is obtained. Different strains have different thresholds for the occurrence of the phenomenon. “Threshold” can be defined as the lowest cell density capable of producing a First Giant in that strain within 12 hr from the onset of the overcrowding conditions: as an exam- ple +C9 had a threshold of 340 cells/ml, while +S9 and #S6 had thresholds of 390 and 500 cells/ ml, respectively (Ricci & Cifarelli, unpublished results).

Starvation, on the contrary, has been shown to play only a minor, disturbing role in giant forma- tion. A clearcut, statistically significant, positive linear correlation has been found to occur between the cell density of an overthreshold population and the number of giants formed in time: only beyond extreme values of cell densities (>40,000 cells/ml for #(C9) this tendency shows a clear reversal, very likely due to the strong disturbance of the

398 N. Ricci AND R. BANCHETTI

Favourable conditions

1/3 AtFG

EARLY WINDOW

Fic. 3.

= INDUCTION (=AtFG) >

1/3 AtFG

LATE ACTIVATION | ACTIVATION WINDOW

FIRST GIANT

PREDATION

1/3 AtFG

The temporal succession and the various relative correspondances of the phases of the process leading on O.

bifaria from its normal state to the production of the first giant.

locomotory behaviour of the potential giants, in turn induced by the continuous, unavoidable cell- against-cell bumps [66].

It has been found also that there is a clear, significant, negative, linear correlation between the number of cells/ml and the Induction Period (4t FG): optimizing the number of cells/ml, their biological state and growing conditions, it has been found that a minimum, non reducible, At FG exists and that it lasts approximately 45 min [65]: thus, we can say that the higher the cell density of the overthreshold population, the shorter the induc- tion period, the larger the total number of the

giants produced. The biological phenomenon controlling the be-

ginning of the story of the differentiation of giants in O. bifaria (and guiding its development later on) is a really interesting contact-dependent cellular interaction, demonstrated by using Con-A as a specific inhibitor of the process [59]: no soluble factor plays any role, as shown by the lack of the slightest effects of both the Cell Free Fluid of giant-producing populations and the perfusion microchamber experiments, using top quality cul- tures [59]. The adaptive reasons of this choice (the choice of using short range signals to trigger the differentiation of giants rather than long range soluble factors, as happens for the conjugation of the same O. bifaria, [52]) seems to us to reside in the nature itself of the giants of O. bifaria, which are true, opportunistic, all-devouring carnivores, rather than cannibals: direct contact seems to us more effective than soluble factors in informing a potential giant about the number of the preys

living in the water volume surrounding it, namely about the possible convenience of differentiating into an actual giant.

Several things must be said to characterize the kind of cell contacts required by this differentia- tion. First of all they must occur between normally behaving cells: neither frozen-thawed nor mildly K5Cr,07 fixed ciliates ever induce any kind of activation [65]: according to these results we can expect a mechanism very likely relying on a certain activation energy which enables O. bifaria to dis- tinguish living (=food) from inert (=no food) objects and, therefore, to avoid senseless, quite expensive differentiations of giants [65]. Thus, it seems to us more appropriate to speak of active contacts, or bumps, which must occur among different cells, to trigger the process, rather than of simple, cellular contacts. By the way, quite a similar recognition mechanism seems to occur also in the story of Litonotus, a specific, a very efficient predator feeding on Euplotes (Ricci & Verni in prep.).

A second trait characterizes the bumps activat- ing induction: they are not species-specific bumps. It has been demonstrated that many different species of ciliated protozoa can bump fruitfully against an oxytricha, activating it to differentiate into a giant, provided that it has such a potential- ity. In our experiments Paramecium aurelia per- fectly succeeded in activating O. bifaria, the only difference being that the threshold for giant forma- tion induction was a little higher than with conspe- cific individuals: ~1,000 paramecia/ml are re- quired to trigger the process, while only 340

Giants as Adaptive Differentiations 399

oxytrichas/ml are sufficient for the same effect [59]. The adaptive meaning of the giant of O. bifaria is clearly demonstrated by such a finding: we cannot speak of cannibals [66] any more, while we have to look at this heteromorphic state as a true carnivorous gigantic form [59]. The basic nature of the giant of O. bifaria is well described by several elements characterizing the differentia- tion: (a) the choice of direct interactions: (b) their non species-specificity; (c) the prompt answer (At FG ~40’) induced by favourable conditions. This way of understanding the biological significance of the giants of O. bifaria is of the highest relevance to interpret the giant itself properly in the context of this species’ adaptive biology: not any more a form by which the species escapes environmental stress conditions feeding the few at the expenses of the many conspecific organisms, but a form through which it opportunistically shifts from its normal diet (bacteria) to a new, temporarily richer pabulum (other ciliates). In so doing, O. bifaria changes its trophic niche, shifting from its normal one (primary consumer) to a new one, that of a secondary consumer. The general interpretation is very likely to be found in the attempt of this bacterivorous species to prolong opportunistically its existence in a certain spatial spot growing at first at the expenses of bacteria (“minimum”, ever present food), and later feeding on the ciliated population supported by the bacteria themselves: by producing the giants, O. bifaria reaches at least four very convenient goals: (a) it reduces the intraspecific competition for bacterial food; (b) it reproduces at the expenses of different species; (c) by predating the bacterivorous species it reduces the interspecific competition for bacterial food; (d) it survives in the first favourable micropatch for a longer period, the peak of primary consumers being reached later than that of bacteria [16]. One surprising thing must be added with regard to the nature of the bumps triggering the giant formation: they can occur even with organisms not lending themselves as preys [65]! Blepharisma japonicum (~300 ym long) induces giants in O. bifaria populations, whose cells are activated, although having no possibility of predating those large organisms! This kind of experiment is the only one so far enabling us to separate the inducing

stimuli (cell-cell bumps) from the presence of preys: this possibility is a truly important one for further investigations about the nature of the molecular aspects of this part of the story. The finding that organisms not suitable as preys induce the giant formation demonstrated that “activation” truly exists as a biological step in the process leading to giant formation. The occurrence of these peculiar bumps in the first third of the At FG is a prerequisite for the process to continue [60]: thus, the objective existence of an Early Activa- tion Window (=namely a short period during which, and only during which, something con- ditioning the occurrence itself of the whole process has to happen!) has been shown clearly by simple, successive dilutions of experimental populations. The conclusion to draw is therefore that, if proper bumps occur at the right moment, they cannot but trigger the physiological steps of the next phase, the Late Activation Window [60].

Before concluding this chapter about the nature of the phenomena leading to giant formation in O. bifaria, the so called “Labile Memory Counter” (LMC) working hypothesis [65] must be briefly described and discussed: the LMC is supposed to be a cellular device somehow capable of counting the cell bumps activating the cortical keys, of recording them for a while (labile memory) and of adding the new contacts to the total: when the sum reaches a certain threshold value (within a certain period of time?) the initial processes of activation period are triggered. Although nothing can be proposed about the basic nature of the LMC, it accounts for several of the observations so far made about the giant differentiation story: the different, strain specific “thresholds” for instance, may be explained in terms of (a) different strain specific number of cortical activating keys, (b) different strain specific number of bumps neces- sary for the activation, (c) different forgetting velocities, (d) different key activation energy.

C_ The Late Activation Window (LAW)

The second half of the Activation (roughly cor- responding to the second third of the induction period) has been found to be characterized by the occurrence of a specific protein synthesis triggered by the LMC when the threshold is reached. While

400 N. Ricci AND R. BANCHETTI

the use of Act-D had demonstrated the relevance of the role played by some protein synthesis, specific low dosage-short term pulses of cyclohex- imide demonstrated the temporal occurrence of that synthesis [60]. It was also found that higher dosages of Chx induce longer At FG and smaller number of giants formed: this seems to indicate that the Chx action is a fairly specific one, inducing just quantitative, absolutely physiological effects. It will be of great interest to ascertain the nature of the specific protein synthesized during the LAW, also in attempting to understand the nature of its action, which, in turn, very likely represents the basic meaning of LAW.

D_ Predation

Both the EAW and the LAW proved to be necessary, but not sufficient steps for the dif- ferentiation of giants of O. bifaria: the actual engulfment of several/many preys is necessary (as shown by the induction by Blepharisma [65] to make the potentialities triggered by the activation real, namely to change the gross cell morphology, thus producing the First Giant. Activated, mor- phologically normal oxytrichas feed successfully on casually encountered preys because of their pecul- iar behaviour. Although specific ethological stu- dies of the problem are not yet available (being in progress in our Lab at the present moment): however, we can say that in a creeping oxytricha activation induces periodical, violent, forward jerks, which enable it to engulf any suitable (=of the right size) prey (Ricci & Riggio, unpublished results). Although only very few things can be said about this step in the induction of giants of O. bifaria, a sort of working hypothesis can be prop- osed, to orientate our future research: does activa- tion affect the bioelectrical state of the temporarily predisposed cells? What is the role played by the specific protein(s) synthetized during the LAW in this change of the electrical properties of the membrane? Answers to these two guide-questions will help us in further penetrating the nature of the biological processes constituting that complex phe- nomenon called induction period.

E The First Giant (FG)

The end of the induction period is represented

by the appearance of the First Giant, namely of that organism easily and unfailingly distinguishable from the other cells for its clearly altered morphol- ogy: it must be noted that it is exactly the same as that of the Steady State Giants that will appear later on in the population. The most typical traits of the FG are the strikingly irregular morphology, which is due to the large number of preys engulfed within roundish food vacuoles, together with dark- er cytoplasm, larger dimensions of the body and of macronuclear pieces as well. Only very recently has it been possible to measure both micro- and macronucleus DNA content. As far as these two very important parameters are concerned the FG definitely presents no significant difference at all from normal cells. This finding cannot but suggest that the first morphological alterations of the giants are induced by the relatively extraordinary diet, the true, nuclear regulation of the process occurring only later.

We must recall here that only a very small percentage (1-2%) of O. bifaria can differentiate directly into FG, while most of the giants of a population (>98%) are produced by transverse binary fissions of preexisting giants [59]. An important, still unsolved problem related to the formation of the FG is represented by its widened peristomial area, namely that lying between the AZM and the UM (~30° in width vs the 15° in normal organisms [66]): is it a feature of the temporarily determined cells? Is it acquired during activation? Is it the consequence of the many successive predatory events?

F The binary fission of the FG

When the FG recovers from the phase of drama- tic cellular changes undergone to begin its own existence, a binary fission occurs, which gives rise to the first generation of Steady State Giants (SSG): during this division the macro- and micro- nuclear DNA content increases heavily, reaching the quantitative strain-specific values characteriz- ing the SSG. The reason why this value for the macronucleus is ~3.7 times the normal value in + C9 [69], and of ~1.9-2.1 times the normal value in #89 [6], is still to be ascertained, as well as the origin of the extra-DNA itself: is it synthesized ex novo, or does it come from a recycling of that of

Giants as Adaptive Differentiations 401

the preys?

As to the micronucleus, both #C9 and S9 SSG show the same DNA content, about 1.9 times the normal quantity. The reasons why in different strains different increases of DNA content occur in the macronuclei and not in the micronuclei are not yet understood. The most relevant result, howev- er, is the finding itself; it is a well established topic among ciliatologists that the micronucleus is a sort of unchangeable, diploid genetic memory in each species. That of the O. bifaria giant is therefore quite a rare exception to the general rule, and it cannot but deserve to be investigated further, for instance from a cytological point of view: could any chromosomal alteration be detected in com- parison with what is described for normal cells [32]?

G_ The Steady State Giant (SSG)

SSG are those giants regularly undergoing a series of cell cycles, thus reproducing regularly through apparently regular binary fissions: their cell cycle is about twice as long as the normal one. The general kinetics of their growth is described by a logistic curve, where the log period is repre- sented by the ex novo dedifferentiation of the few First Giants, the log period is the expression of their intensive binary fissions, while the plateau is reached when the available preys become rarer and scattered through progressively wider spaces [66].

To describe such an interesting heteromorphic form thoroughly, the general morphology has been studied at the cytological [66] and ultrastructural levels [67]. The general shape of a normal O. bifaria (the upper half of a rotation ellipsoid) becomes more irregular in the SSG, due to the widening of the peristomial funnel at the anterior end of the giant (already mentioned in the para- graph about the First Giant) and to the preferen- tial accumulation of the food vacuoles in the posterior two thirds of the body, which assumes roughly the shape of half a pear. While all the particulars are described elsewhere [67], three major results deserve particular attention: (i) the normal 8 frontoventral cirri become 10-12 in SSG; (ii) the paraoral external (of the Undulating Mem- branelles) from a double ciliary array passes to a

triple or multiple ciliary array; (iii) the AZM strikingly increase the number of membranelles, exactly doubling the number of cilia per single membranelle: 3, 16, 22, 22 in normal cells, 3, 32, 44, 44 in a SSG.

These alterations well describe the change of the trophic niche of O. bifaria’s SSG, an organism which in fact needs wider peristomial and stronger “predating” organelles, being specialized in feed- ing on large, sturdy preys. The digestive system of a SSG, generally studied in [66], proved to be particularly interesting, being formed by many vacuoles, each containing either one single prey or a large amount of bacteria, the two foods being never mixed together within the same, single vacuole [67]; it seems likely, that in this way, different enzymatic batteries can be activated around different vacuoles, thus making the two digestive strategies more effective.

The last and perhaps the most important cyto- logical trait of an SSG is its already mentioned higher degree of micro- and macro-nuclear DNA content [60, 69], a biological trait actually charac- terizing the differentiating process.

Moreover the SSG proved to depend upon continuous active cellular contacts (=bumps) to maintain its differentiated state. The bumps must occur continuously between cortically normal cells: Con-A used separately either on SSG or on preys can interrupt this flux of information about the presence of preys, thus removing the con- tinuous trigger maintaining the gigantic state or, in other words, initiating the process of dedifferentia- tion. Experimentally to interrupt the normal re- production cycle of the SSG, it is sufficient also to isolate the giants in cell-free media: this clearly imitates the natural conditions at the end of a ciliate bloom in a certain volume. This shows that the maintenance of the gigantic state is continuous- ly controlled opportunistically by a sort of feed- back, capable of promptly revealing the progres- sive exhaustion of the specific pabulum (other ciliates): as soon as this situation occurs, the processes leading a SSG back to the normal state of the species are triggered.

To conclude about the nature of the SSG, we can well state that it represents a reversibly diffe- rentiated phase, specifically acquired by the spe-

402 N. Ricci AND R. BANCHETTI

cies, to exploit a new, predatory niche: specific traits facilitating such a role are developed by this primarily bacterivorous species: (a) wider peristo- mial funnel, (b) stronger ciliary organelles, and (c) non-oriented (=blindly casual) prey detecting system.

H The Dedifferentiation

In absence of a sufficient number of bumps, a SSG undergoes 4 binary fissions which specifically lead it back to its normal morpho-physiological state. Body size, nuclear size and DNA content have been measured during this process, and found to decrease progressively; in particular, the size and the DNA content of macronuclear appa- ratus pass from 220 um and 371 arbitrary units (AU), respectively, to 200 um and 380 AU after the 1st division, to 160 ~m and 257 AU after the 2nd, to 100 um and 177 AU after the 3rd division, to 102 wm and 108 AU after the 4th: being ~100, the control values of macronuclear size and DNA content, it becomes evident that, while the normal size is acquired after 3 cell divisions, the normal DNA content is established again only after the Ath division. Bacterized medium, autoclaved let- tuce medium and SMB (Synthetic Medium for Blepharisma, [44]) were used as dedifferentiating conditions: it was found that dedifferentiation occurs according to different kinetics, that in bacterized medium being capable of inducing the fastest dedifferentiation, the second being the autoclaved lettuce medium, the third the SMB [60]. Overall these results seems to indicate that this process clearly depends upon some energetical energy input, but more precise experiments are required before drawing conclusions from this story. We are at present carrying out a series of experimental sessions to draw the ethogram of these ex-giants, to test the general hypothesis [16] about the meaning of the newly “normal” indi- viduals as a sort of exploration shuttles (cf. the Tetrahymena’s swarmers), by which the species, after such a differentiation, quickly and efficiently spreads through the environment to find out new possible favourable conditions, where new popula- tions can settle and grow.

II The Protozoa and the “invention” of cell dif- ferentiation

First of all, we must recall the mental “Coperni- can revolution” invoked in the Introduction sec- tion: we must keep it clearly in mind that protozoa were the first eukaryotic living entities to colonize the primeval ocean for hundreds of millions of years before the pluricellular organization was reached: “All the pluricellular eukaryotes evolved from unicells, in which the fundamental traits of development appeared” [22]. Such a statement must guide our culture in approaching many biolo- gical problems correctly: with regard to “cell dif- ferentiation” we must recall that, though repre- senting a typical aspect of the biology of metazoa from a cultural point of view, it is actually one of the brightest “inventions” of protozoa, which used it to face several, dramatic survival problems. The definition itself of development has “developed” over recent years, passing from “the growth from one embrional stage to the next” (typically given for metazoa) to “the progress of an organism through its life cycle”, a general concept perfectly fitting the case of protozoa and of Oxytricha, as well, once we consider its life cycle as represented in Fig. 2.

If we proceed in this attempt to establish termi- nological parallelisms, which in turn underly sub- stantial similarities between protozoa and meta- zoa, we can consider now the main, three classical steps characterizing the development of a mul- ticellular organism: (i) cell differentiation, i.e. the process leading from one single stem cell to many different types; (ii) pattern formation (the process leading the different types of different cells to congregate in a certain, well defined organ); (iii) morphogenesis, i.e. the mechanical process under- lying organism shaping and tissue generation. According to the most commonly given definition of cell differentiation (“cells with one single genotype give rise to definitely different pheno- types”), there can be very little to debate, about the fact that the production of cysts and giants by O. bifaria actually are cell differentiations: stri- kingly different morphological and physiological states are generated, not through any casual abor- tive process (as has been suggested for the doub-

Giants as Adaptive Differentiations 403

lets of O. bifaria, [2]), but rather through geneti- cally encoded clearcut cytoplasmic reorganiza- tion(s) and specific protein synthesis.

This parallel holds further, if we consider that in the cell differentiation of the metazoan develop- mental processes two basic strategies have been described [28]: (a) long-range interactions medi- ated by soluble inductors (messages) such as those involved in the metamorphosis of Amphibians [37] and Insects [23, 26]; (b) short-range interactions

PROTOZOA

(cell-cell contacts), as those described by Jacobson [32] and Muthukkaruppan [45], for mouse lens development, Grobstein [27] for mouse metanephros development, Slavkin & Bringas [74] for odontogenesis, Cutler & Chaudhry [10] for rat submandibular gland development, Lehtonen et al. [38] for kidney tubules formation. Quite similar is the case of cell interactions described for proto- zoa, whose differentiative biology is controlled by soluble factors (Blepharisma’s preconjugation,

various messages (stimuli)

within tolerable values

Vv

METAZOA

various messages (stimuli)

+-aS a

ENVIRONMENT

PROTOZOON... .as an ORGANISM

it shows adaptive behaviour

(marrow range adjustments of physiological state)

aS a CELL it

undergoes adaptive cell differentiation reversible wide range adjustments of cell morphology and physiology to reshape the body )

ENVIRONMENT

METAZOON... .asan ORGANISM

it shows adaptive behaviour

BODY ,

a buffered environment XN for the cells, it controls

\ e ° ry q cell differentiation,

(imreversible morpho-physiological changes of the cells to build the body)

proper answers to match the environmental

challenges

LL = = =

proper answer to match the environmental challenges

,» 1.e. as a

Fic. 4. The parallelisms and the differences occurring between cell differentiation in Protozoa and Metazoa.

404 N. Ricci AND R. BANCHETTI

Miyake [44], Volvox carteri’s, Kochert [36] and Dictyostelium discoideum’s development, O’Day & Lewis [49]) or by direct cell-cell contacts (Chlamydomonas, Goodenough [24], Parame- cium’s, Hiwatashi [31]).

Among the other models, O. bifaria’s precon- jugative story is particularly interesting, inasmuch as the potential partners rely on a peculiar two step interaction process [52]: O. bifaria indeed releases the mating type specific soluble factors (gamones) to mediate the first-step cellular interactions be- tween potential partners occurring from a distance [15], while it uses a cell-bound-gamone strategy to guide the partners along the last, critical steps of the interactions leading to membrane fusion, meiosis, pronuclear exchanges etc. [61]: the spe- cies, in other words, is capable of using both these different strategies! Why does it use only cell contacts to trigger and to control the differentia- tion of giants? The opportunistic, highly conve- nient species-non-specificity of preys discussed above seems to us the evolutionary clue, possibly accounting for such an interactive strategy.

Coming back to the basic problem (can we speak of cell differentiation in the protozoan world?), the classic examples of Acetabularia [29, 35, 84] and of Naegleria [18] should cancel any doubt. On the other hand most of the efforts spent to convince zoologists working in the field of development that our attempt to make a terminological extrapola- tion actually represented quite a convenient (and perfectly correct) cultural jump, often came up against two major obstacles.

The first one was that the developmentalists have always said that “Cell differentiation is an irreversible phenomenon among metazoa”: in our opinion on the contrary the primigenial cell dif- ferentiation on the contrary was a reversible phe- nomenon (like that found in protozoa), while irreversibility was added only later when mul- ticellularity was reached. There would be no advantage for a protozoon in being capable of skipping environmental stresses by encysting, if it could not resume normal morphology and physiol- ogy on the return of favourable conditions. It is obviously true, moreover, that a concept like “irreversibility of the cell differentiation” could not be imagined except for a metazoon, which is a

multicellular organism capable of spending even millions of cells for just one function (electrical conducibility, contractility, distribution, efc.), not needing at all any dedifferentiation! This point of view seems well supported by the finding that whenever such a cellular “expendibility” is obtained by protozoa too, the same irreversibility is also realized: the case of Volvox seems para- digmatic!

The second objection to our efforts to establish strong parallelisms between protozoan and meta- zoan cell differentiation was somehow less critical, on the one hand, and yet more complicated to answer, on the other: the trigger and the control system of cell differentiation is “internal” in Meta- zoa, but “external” in Protozoa! If we consider the scheme in Fig. 4, the meaning of this observation is perfectly evident: in the case of protozoa, the environment modulates the behaviour of an organ- ism to induce the proper adaptive answers to the environmental changes under normal conditions, while it stimulates cell differentiation under ex- treme conditions, as a sort of deep morpho- physiological adjustment of the entire body to the environmental challenges. In the case of metazoa, on the contrary, the same environment, although acting clearly on the organism at the level of its adaptive behaviour, does not exert any direct effect on the cell differentiation itself, which relies upon specific messages released by another very peculiar and buffered “environment”, namely the whole body, which ends up by playing the role of a sort of fairly complex interface between the cells and the external environment.

Before concluding this part of the discussion, we should like to recall also “pattern formation” and “morphogenesis”, the other two stages of develop- ment after cell differentiation: if one considers the complex life cycle of V. carteri [36] and the soph- isticated one of D. discoideum [49], it seems to us that also as far as these two stages of development are concerned, protozoa show quite complex be- haviour and unexpected capabilities. Therefore it seems to us perfectly appropriate and justified to use the biological concepts (not only the termino- logy) of cell differentiation etc. also when speaking of protozoa.

Giants as Adaptive Differentiations 405

Inducing conditions

Small percentage of ex-novo differentiated organisms

Steady

State

hetero-

morphic

organism n® preys/cell 3

total digestion

return to the normal state through regulatory divisions

Fic. 5. A comparative, synoptic scheme of what reported in bibliography about the reversible cannibalism- gigantism-carnivourism in ciliates; =information ignored by the authors; (+ )=reported by the authors as a minor observation; + =yes!; — =no!

406 N. Ricci AND R. BANCHETTI

III The Protozoa and the “invention” of the ‘“‘con- sumers”: new ways of exploiting environmental resources.

There can be no doubt that modern protozoa represent the descendants of the first eukaryotes: they appeared in a primeval ocean completely colonized by prokaryotes, which represented a potentially limitless pabulum, for any organism capable of feeding on them [25]. This concept introduces to us the idea that the first protozoa found themselves literally “embedded” in a “full- food” substrate: the primary consumers, seen as new exploiters of the environmental resources, appeared therefore quite early in the evolutionary hystory of protozoa, which, due to both their “superior”, eukaryotic design and the endless food source (prokaryotes), underwent a tremendous adaptive radiation, concerning them so deeply and dramatically, that the widest morpho-physiological variations on the theme “eukaryotic cell” were realized. Of all the new adaptive solutions one in particular is to be considered in the context of our present paper, namely the conquest of the trophic niche of the secondary consumer, as a consequ- ence of that evergrowing, as yet ungrazed pabulum represented by the protozoa themselves. This second step made by the primitive eukaryotic unicells in the recent world of the consumers represented quite an important achievement, be- cause it completed the first food chain ever: deep and extensive studies of the classic examples of protozoan predation [1, 13, 43, 77] might guide us to a more correct comprehension of the phe- nomenon “carnivorism” in its general lines, in its essential, basic traits. The study of Litonotus lamella which predates specifically Euplotes cras- sus [68] led us to show that, although very simple and primitive, this organism had already realized an almost perfect carnivore: Litonotus, indeed, has specific systems to recognize and locate its prey (Ricci & Verni, in prep.), to kill it [73], to ingest and to digest it [74]; with respect to the more renowned wolves or lions, Litonotus lacks only the social dimension of the hunting pack.

In our opinion, the correct approach to the phenomenon of the reversible cannibalism- giantism-carnivourism (CGC) in protozoa is to con-

sider it as a very peculiar phenomenon, somehow representing an intermediate trophic niche, for the ciliate capable of behaving as herbivour or as carnivour organism. A hypothesis might even be put forward, concerning the possibility that the phenomenon CGC could represent, and somehow tetstify, that intermediate stage of evolution, when ciliates were neither fully primary consumers nor fully secondary consumers: could the advantage of being an active predator in the proper conditions lead a bacterivorous species to acquire the reversi- ble capability of becoming carnivorous, before becoming a fully irreversible carnivorous species?

We tried to reconsider as many bibliographic items as possible dealing with the CGC study: although a large number of reports is available in the field, it must be stressed, however, that they are only relatively useful in drawing a complex picture of this problem for several reasons: (a) the total number of the ciliate species studied for their CGC is very small, no more than fifteen out of thousands, characterized by extreme diversity and heterogeneity; (b) they belong to widely different systematic groups, so that any possible similarity among them might actually be due to different phylogenetic evolution and/or to different adap- tive strategies, thus making it impossible to unify interpretations, in terms of homologies/analogies, possibly accounting for the observed CGC process; (c) the reports range in time from 1853 [30] to 1993 [37]: this extremely wide span makes most of the papers only partially comparable, due to the diffe- rent minds, cultures, attitudes the researchers had not only toward the problem, but also toward the way of reporting the data: simple, direct, descrip- tive, naturalistic the earlier (roughly up to the sixties), eminently technical, elaborate and mainly focussed on cell biology aspects, the most recent.

Bearing all these considerations in mind, howev- er, we tried to put the things together, in an attempt to identify, where possible, common fea- tures and elements, indicating to us, at least tenta- tively, the basic fundaments of the adaptive logic of CGC phenomenon, in order to be able to propose a unifying working hypothesis to guide future research in this field toward more ordered, more comparable and more interpretable results.

The papers we studied deal with Frontonia [81],

Giants as Adaptive Differentiations 407

Tetrahymena [4-6, 14, 42, 73, 83], Blepharisma [20, 48, 50, 77], Platyophryides [51], Lembadion [37], Stentor [19, 39, 76], Gastrostyla [85], Oxyt- richa hymenostoma [11], Stylonychia [21], Onychodromus |86], O. bifaria [59, 60, 65, 66, 67, 69]. The results just described led us to put together a sort of patchwork (Fig. 5) describing the main characters (horizontal lines) known for the different species (vertical columns). The first consideration is that a very large number (58) of the 154 possibilities are black (=“unknown”), and that 7 more refer to information not reported directly by the Authors, but just found as small, secondary observations. In other words ~40% of the possible, basic information is not yet available even for the few species considered! In spite of all the handicaps so far mentioned, however, a sort of Common Denomenator, formed by several basic elements, seems to emerge from the scheme of Fig. 5.

A The induction conditions (internal and ex- ternal)

Independently from the different single, species- specific solutions, this general character well indi- cates a basic adaptive value of the CGC phe- nomena in ciliates: a physiological, internal, very likely epigenetic (sensu Nanney, [46]) predisposi- tion is controlled and determined by environmen- tal modulations (proper, weak preys-starvation- high cell densities).

B The number of the ex-novo produced heter- omorphic CGC organisms

With the only exception of Tetrahymena (which evolved the quite singular stomatin-induction of giants) the other species seem to share the same adaptive strategy: very few heteromorphic organ- isms differentiated ex-novo undergo intense cell binary fissions, for producing massively CGC. This finding seems to represent one of the clearest traits shared by most of the species studied. Why, on the contrary, should they prefere this solution (few changed FG reproduce many heteromorphic SSG instead of the Tetrahymena’s strategy (many changed SSG) is quite a complex question, far from being resolved, on the basis of the data so far available.

C_ The body size

Only Frontonia and Stentor do not enlarge the body to fulfill the new adaptive tasks, while the other 9 genera produce overdimensional indi- viduals. The sub-characters, namely wider peristo- mial structure, richer ciliary organelles and larger nuclei seem to follow unfailingly (automatically) the shift in body size, as possibly confirmed also by the case of Frontonia, where normal sized organ- isms have also normal peristome, ciliature and nuclei. The adaptive meaning of this deep, mor- phological reshaping seems to be quite clear, once the CGC nature of the heteromorphic phase is considered. It can be easily observed that (a) normal size cannibals (Frontonia), (b) gigantic cannibals (Blepharisma, Lembadion, Onychodro- mus) and (c) gigantic carnivores (Tetrahymena, Oxytricha bifaria) are described, while (d) no small carnivore has been found so far. If this element should be confirmed by future research, it might represent an important clue to the penetration of the intimate nature of a predator, which it is to be expected should be considerably larger and stron- ger than its preys, to work as an efficient carnivore. The case of Frontonia, in this context, seems to us the clearest example of a purely cannibalistic spe- cies, according to three major characters: (i) short lasting tendency to produce cannibals; (11) small size cannibals; (ii) three preys per cannibal, at maximum.

A final remark must be made with regard to the change in the body shape, when a giant is pro- duced: while all the ciliates seem just to “inflate” their body, regulating them isometrically (=“the giant is just a larger-normal individual), Hypot- richs, on the contrary, seem to have a highly conservative, untouchable portion of their body (namely their ventral surface, with all the locomo- tory organelles) and an extremely plastic, dorsal one, which, on the contrary, can undergo exten- sive variations (cf. the O. bifaria’s case, as the most paradigmatic in this sense.

D_ The adaptive meaning As to the adaptive strategy actually applied by a

CGC producing species we must be extremely cautious: many authors, indeed, speak of cannibals

408 N. Ricci AND R. BANCHETTI

merely because they only experienced that situa- tion in their experiments, without trying to check whether those forms could also be capable of predating non conspecific preys (carnivorous vs cannibal behaviour). According to our data, only the case of Lembadion’s giant was specifically tested in this sense and found to be a true, cannibalistic form. In general, the two strategies are followed by the different species, roughly in a 1:1 ratio: their basically different meanings have been already discussed in section I and II. To conclude this paragraph which deals with the possi- ble adaptive meaning of these heteromorphic stages in the biology of ciliates, a special word must be said about the Tetrahymena example: T. vorax specifically undergoes cell differentiation whenever it perceives the substances released by T. pyriformis. In our opinion, this seems to indicate that the species has chosen an extremely specialized form of carnivourism, feeding on only one species. This, in turn, cannot but remind to us the case of all those species of ciliates specifically predating only one prey (cf. the cases above men- tioned). The finding that also among the tempora- rily, reversibly differentiated predators, a similar strategy has been chosen not only extends the number of those cases, but seems to us moreover to strengthen Fenchel’s [16] hypothesis about the significant general advantages of such feeding be- haviour, which enables many different species of predators to survive also in one single habitat, without such a wide species diversification.

E The return to the normal state

The way by which the different species return to their normal morphology and physiology once the proper CGC conditions are over, namely the series of regulating fissions, represents another apparent- ly univocal solution found by quite different spe- cies to solve the same problem. The explanation of the reasons why this solution actually represents such a convenient dedifferentiation path for the species studied could represent a strong contribu- tion to a deeper understanding of the cell dif- ferentiation in general.

On the basis of the data already published and of the considerations so far made, we would like to recall how difficult and delicate the correct study of

these phenomena can be even for the most brilliant scientists [9, 87]: therefore it seems to us an absolute necessity, if we extend our knowledge of this theme, to carry on widespread new research, not only to complete what reported in Fig. 5, but also to extend it by adding new species and new elements: only after such a basic investigation phase, will any seriously indicative conclusion be- come possible.

CONCLUSIVE REMARKS

The data and the arguments above reported demonstrate that, as stated in the first part of this paper, protozoa actually help us in penetrating not only their biology and life philosophy, but also at least some of the most hidden aspects of the general adaptive strategies of Life: their peculiar, unique, double-sided nature, indeed, enables us to put together different fields of Biology, interpre- tating the same phenomenon according to the ideas and concepts typical of diverse areas. As a simple example, we cite Fig. 6, where the succes- sive steps of the process “O. bifaria giant forma- tion” are read as successive phases of a “cell differentiation” and as serial states of an “adaptive strategy”, as well.

In other words, Protozoa lend themselves as precious material in biological investigation, repre- senting a solid, polyhedric bridge between cell biology and whole organism adaptive biology: in protozoology, in fact, two quite different and basic aspects of biology (cell vs organism) become directly the two faces of the same coin and, in these conditions, both contribute to define the nature of the coin itself.

In our opinion, this is the most important thesis of the present review.

REFERENCES

1 Balbiani EG (1873) Arch Zool Exper 2: 363-394

2 Banchetti R, Ricci N (1986) Protistologica 22: 161- 168

3. Bradbury P, Hausmann K (1993) Ciliates: cells as organisms. Ed. by P Bradbury and K Hausmann, Springer Verlag Buhse HE3Jr (1966) J Protozool 13: 429-435

5 Buhse HESr (1967) J Protozool 14: 608-613

Giants as Adaptive Differentiations

induction

ingestion

differen- tiation’s

normal over-

threshold activation

culture

conditions

not differen- tiated cells

as cell

cell-cell | protein

contacts

specific

differen- trigger

tiation...

+a / stimulus , primary physiological adaptive from

consumer preparation

strategy. environment

synthesis

accom- plishment

exploitation

of a new

pabulum

409

giant binary Steady

first fission State

giants

cell-cell morpho- contacts

physiol.

nuclear regulation

dediffer-

maintain entiation

differen- tiation

changes

terminal the preys Bachito

secundary condition

the SSG

existence

determi- primary consumer fi WOE consumer

Fic. 6. The three levels of the story of the giants of O. bifaria: the story itself, the cell differentiation, the adaptive

16

ily

18 19 20 ~Al

22

Strategy.

Butzel HM, Fisher J (1983) J Protozool 30: 247-250 Cairns JJr (1991) Environm Auditor 2: 187-195 Corliss JO (1979) The Ciliated Protozoa. Pergamon Press, Oxford

Curds CR (1966) J Protozool 13: 155-164

Cutler LS, Chaudhry AP (1973) Dev Biol 33: 229- 240

Dawson JA (1919) J Exp Zool 29: 473-513

Dini F, Bracchi P, Luporini P (1975) Acta Protozool 14: 59-66

Dragesco J (1962) Bull Biol Fr Belg 96: 123-167 Dupy-Blanc J, Metenier G (1975) Protistologica 11: 159-167 ;

Esposito F, Ricci N, Nobili R (1976) J Exp Zool 197: 275-282

Fenchel T (1987) Ecology of Protozoa. The Biology of Free-living Phagotrophic Protists. Science Tech. Publishers. Madison, WI

Foissner W (1987) In “Progress in Protistology Vol 2” Ed. by JO Corliss and J Patterson, Biopress Ltd, pp 69-212

Fulton C, Walsh C (1980) J Cell Biol 85: 346-360 Gelei J (1925) Arch Protistenkd 52: 404-417 Giese AC (1973) Blepharisma. The Biology of a light-sensitive Protozoan. Stanford Univ. Press, Stanford, California, pp 123-134

Giese AC, Alden RH (1938) J Exp Zool 78: 117- 134

Gilbert LI (1988) Developmental Biology. Sinaver Associates, Inc, Sunderland, Massachusetts, U.S.A., pp 1-635

23

24

25

26

27 28 29 30 31

32

33

34

35

36 37

Gilbert LI, Goodman W (1981) In “Metamorph- Osis: a problem in developmental biology” Ed. by LI Gilbert and E Frieden, Plenum, New York, pp 139- 176

Goodenough UW (1977) In “Receptors and recog- nition Vol 3” Ed. by JL Reissig, Ser. B, Chapman & Hall, London, pp 323-351

Gould SJ, Raup DM, Sepkoski JJJr, Schopf TJM, Simberloff DS (1977) Paleobiology 3: 173-181 Granger NA, Bollenbacher WE (1981) In “Meta- morphosis: a problem in developmental biology” Ed. by LI Gilbert and E Frieden, Plenum, New York, pp 105-138

Grobstein C (1955) J Exp Zool 78: 539-547 Grobstein C (1956) Exp Cell Res 10: 424-440 Haemmerling J (1934) Wilhelm Roux Arch En- twicklungsmech Org 131: 1-81

Haime J (1853) Ann Sci Nat Zool 19: 109-134 Hiwatashi K (1969) In “Paramecium” Ed. by CB Metz and A Monroy, Academic Press, New York, pp 255-293

Jacobson AG (1966) Science 152: 25-34

Jahn TL, Bovee EC (1967) In “Research in Pro- tozoology” Ed. by TT Chen, Pergamon Press, Ox- ford, pp 41-200

Jennings HS (1906) Behaviour of the Lower Organ- isms. Indiana Univ. Press, Bloomington, IN Kloppstech K, Schweiger HG (1975) Differentia- tion 4: 115-123

Kochert G (1975) Symp Soc Dev Biol 33: 55-90 Kuhlmann HW (1993) Arch Protistenkd in press

38 39 40 41 42 43 44 45

46 47

48

410

Lehtonen E, Wartiovaara J, Nordling S, Saxen L (1975) J Embryol Exp Morphol 33: 187-203

Lennartz DC, Bovee EC (1980) Trans Am Microsc Soc 99: 310-317

Lueken W, Krueppel T, Gaertner M (1987) J Exp Biol 130: 193-202

Machemer H (1988) In “Paramecium” Ed. by HD Goertz, Springer Verlag, Berlin, pp 185-215 Metenier G (1981) Eur J Cell Biol 24: 252-258 Miller S (1968) J Protozool 15: 313-319

Miyake A (1968) Proc Jpn Acad 44: 837-841 Muthukkaruppan VR (1965) J Exp Zool 159: 269- 288

Nanney DL (1958) Proc Natl Acad Sci 44: 712-717 Nanney DL (1980) Experimental Ciliatology. Wiley and Sons, New York, pp 67-75

Nilsson JR (1967) Compt Rend Trav Lab Carlsberg 36: 1-24

49 O’Day DH, Lewis KE (1975) Nature 254: 431-432

50 51 52 53 54 55 56

57 58

59

60

61

62

Padmavathi PB (1961) Arch Protistenkd 105: 341- 344

de Puytorac P, Kattar MR, Groliere CA, da Silva Neto I (1992) J Protozool 39: 154-159

Ricci N (1981) In “Sexual interactions in Eukaryotic Microbes” Ed. by PA Horgen and DH O’Day, Academic Press, New York, pp 319=350

Ricci N (1989) Limnol Oceanogr 34: 1089-1097 Ricci N (1990) Anim Behav 40: 1048-1069

Ricci N (1991) Mar Pollut Bull 22: 265-268

Ricci N (1992) Proc 3rd Int Congr Neuroethology, August 9-14, 1992, Montreal, Canada, pp 79-80 Ricci N (1992) Acta Protozool 31: 19-32

Ricci N, Banchetti R (1981) Acta Protozool 20: 153-164

Ricci N, Banchetti R, Pelamatti R (1989) Hydro- biologia 182: 115-120

Ricci N, Bravi A, Grandini G, Cifarelli D, Gualtieri P, Coltelli P, Banchetti R (1991) Eur J Protistol 27: 264-268

Ricci N, Cetera R, Banchetti R (1980) J Exp Zool 211: 171-183

Ricci N, Erra F, Russo A, Banchetti R (1989) J

63

N. Ricci ANP R. BANCHETTI

Protozool 36: 567-571

Ricci N, Erra F, Russo A, Banchetti R (1991) Limnol Oceanogr 36: 1178-1188

Ricci N, Erra F, Russo A, Banchetti R (1992) J Protozool 39: 521-525

Ricci N, Erra F, Russo A, Banchetti R (1991) Eur J Protistol 27: 127-133

Ricci N, Riggio D (1984) J Exp Zool 229: 328-337

Ricci N, Rosati G, Verni F (1985) Trans Amer Microsc Soc 104: 70-78

Ricci N, Verni F (1988) Can J Zool 66: 1973-1981

Riggio D, Banchetti R, Seyfert HM, Ricci N (1987) Can J Zool 65: 847-851

Robinson H, Chaffee S, Galton VA (1977) Gen Comp Endocrinol 32: 179-186

Rosati G, Giari A, Ricci N (1988) Eur J Protistol 23: 343-349

Rosati G, Verni F, Ricci N (1984) Protistologica 20: 197-204 :

Ryals PE, Smith-Sommerville HE (1985) J Pro- tozool 33: 382-387

Slavkin HC, Bringas PJr (1976) Dev Biol 50: 428- 442

Smith HE (1982) J Protozool 29: 616-627

Tartar V (1961) The Biology of Stentor, Academic Press, New York, pp 1-413

Tulchin N, Hirshfield, HI (1962) J Protozool 9: 200- 203

Verni F (1985) Zoomorphol 105: 333-335

Verni F, Ricci N (1986) Symp Biol Hungar 33: 213- 216

Verni F, Rosati G, Ricci N (1984) Protistologica 20: 87-95

Vimala Devi, R (1964) J Protozool 11: 304-307 Visscher JP (1924) Biol Bull 45: 19-27

Williams NE (1961) J Protozool 8: 403-410 Wilson EB (1896) The Cell in Developmental Inheritance, Mac Millan, New York

Weyer G (1930) Arch Protistenkd 71: 139-228 Wicklow BJ (1988) J Protozool 35: 137-141

Yanbin P, Zuoren Z (1981) Acta Zool Sinica 27: 7— 11

ZOOLOGICAL SCIENCE 10: 411-415 (1993)

© 1993 Zoological Society of Japan

The Defective Color Vision in Juvenile Goldfish Does Not Depend on Used Training Task as the Measure of Discrimination: A Two-choice Response Measure

KEN ONISHI

Department of Physiology, Nara Medical University, Kashihara, Nara 634, Japan

ABSTRACT— Color discrimination ability of juvenile goldfish (Carassius auratus) was measured using a Y-maze training technique to test whether the defective color vision in juveniles found in a previous study [10] would be similarly observed using another training paradigm. The discrimination ability of juveniles was compared with that of adults that had been previously obtained with the same training paradigm [9]. Although the juveniles trained with green and red discriminative stimuli (colored papers) showed good discrimination ability comparable to that of the adults, the juveniles had great difficulty in discriminating blue from both green and red while the adults did not. These results are in agreement with findings in the previous study [10] using a “go/no-go” task. The defective color vision of juveniles is clearly not task-dependent but is rather a general property originating from the developmental

process of blue vision.

INTRODUCTION

A previous study [10] on the development of color vision in goldfish has clearly shown that juveniles have defective color vision, specifically for color blue, when it is examined with a “go/ no-go” training technique. It may be possible, however, that they can process blue information normally and discriminate blue from other colors as do adults when they are confronted with another training paradigm. The neural pathway (especially the blue information processing path- way) in juveniles that produces avoidance re- sponses might not yet have developed fully com- pared with other neural pathways subserving other types of behavior. It is still unclear whether or not the defective color vision observed in the previous study [10] is task-specific. Many behavioral studies on spectral sensitivity indicate that the property of any spectral sensitivity depends on the training task. For example, the reflex-like startle response tends to be triggered predominantly by long- wavelengths [2, 13]. A more complex behavior,

Accepted March 1, 1993 Received January 25, 1993

such as a two-choice discriminative response, is produced by short- and mid-wavelengths as well as long-wavelengths [6, 15, 16]. Yager [15] reported that short-wavelengths are most effective for the two-choice type of response. It seems that a definite spectral region is dominantly processed in a neural pathway when a certain type of behavioral pattern is used as the measure of a discriminative response.

In addition to such task-dependency, the color vision of goldfish also depends on the intensity levels of background illumination and discrimina- tive lights. The vision for red is defective under low levels of background illumination when the fish are trained on a “dark” test field [7, 8]. The color vision disappears in wavelength discrimina- tion performed at a low intensity of discriminative lights when trained on an “illuminated” test field [8]. In this condition, the fish use a “brightness” cue but not a “color” one. Considering the light- ing-condition dependence of color vision as well, in this study, the discriminative colored papers were presented under a relatively high intensity level of background illumination (about 1000 Ix) compared with that in the previous study (15 lx;

[10}).

412 K. OH8NISHI

MATERIALS AND METHODS

The same size juvenile goldfish (3-5 cm, age of <1 year) as those used in the previous study [10] were trained to discriminate between three types of paired colored papers, blue vs. green, red vs. blue and red vs. green, using the same Y-maze instrumental conditioning technique employed and described in detail in another former study [9]. In brief, juvenile fish were rewarded with food when they chose a correct stimulus in such a way that they swam into one of the choice chambers. For brightness control, the colored papers which had been adjusted to be of equal subjective brightness were adopted in this study as well. The maxima of reflectance (A,,,.x) of the colored papers was 480 nm in blue, 520 nm in green and 660-700 nm in red. In the two cases of blue/green and blue/red discriminations in juveniles, the blue papers of higher brightness (about 5% higher in relative brightness than for adults) were used, taking into account the juveniles’ lower blue sensitivity [10].

RESULTS

Juveniles, unlike adults, showed very poor abili- ties in color discrimination between blue and green and red and blue as shown in Figure 1 and 2a. In particular, in the discrimination between blue and green, the percentages of correct responses for the juveniles (n=11) were very low and not in- cremental (they did not increase over 60% in later sessions), whereas those for the adults (n=6) reached about 90% by later sessions: all data on the adults’ color discrimination used in this study were taken from a previous study [9]. Significant differences between the correct responses were observed in the two groups (for statistical analysis, a three-way analysis of variance was used: Groups, F(1, 15)=52.18, P<0.01; Days, F(7, 105)=8.33, P<0.01; Groups x Days, F(7, 105)=3.90, P< 0.01). Similarly, in the discrimination of blue/red, those for juveniles (n=23) were not as high as for adults (n=6; Groups, F (1, 27)=8.80, P<0.01; Days, F (7, 189)=24.42, P<0.01; Groups x Days, F(7, 189)=0.93, P>0.05). Some juveniles (n= 10), however, showed adult-like color discrimina- tion ability (Fig. 2b). No significant differences

o> Blue vs. Green = 100 — oO o 7) c fo} Q 0 50 - ® = 4 @ Adult (6/6) = O Juvenile (0/11) ro) (0) _————— oO 1 45 10 Sessions Fic. 1. Learning curves in juveniles (0, n=11) and

adults (@, n=6) trained with the blue vs. green stimuli. The fish were rewarded only when they responded to a correct stimulus (a stimulus less frequently responded to during the pretraining trials) during the training trials. The filled triangle indicates the start of the training trials. The num- bers in the parentheses indicate learners/trained fish. Vartical bars, +SD.

Red vs. Blue

100 ~~ ES iD [5}0))| P= =a o 7) Cc e Adult (6/6) ° oJuvenile (10/23) a o O i ? 100 Oo ® _ o 50 (6)

eLearner (10) oNon-learner (13)

{oS ao Sessions

Fic. 2. Learning curves in juveniles (0, n=23) and adults (@, n=6) trained with the red vs. blue stimuli (a) and those in learners (@, n=10) and non- learners (O, n=13) of the juveniles (b). Significant differences were observed in (a) but the juvenile learners showed a very similar learning curve to that of adults as shown in (b). See Figure 1 for further explanations.

Development of Color Vision in Goldfish

were observed between the correct responses of the juvenile learners (the fish which showed over 75% correct responses for at least 3 days succes- sively: all trained adults fulfilled this criterion) and adults (Groups, F(1, 14)=1.65, P>0.05; Days, F (7, 98)=24.54, P<0.01; Groups x Days, F(7, 98) =0.30, P>0.05). The percentages of correct responses of non-learners (n=13) in this task of discrimination were very low (Fig. 2b); they did not increase over 50% in later sessions like those of juveniles in the discrimination between blue and green. Contrarily, in the discrimination between red and green (Fig.3), the juveniles (n=12) showed a good, adult-like color discrimination ability. Their correct responses were not signi- ficantly different from those of adults (n=7; Groups, F(1, 17)=1.61, P>0.05; Days, F(7, 119) =9.37, P<0.01; Groups x Days, F(7, 119) =0.50, P>0.05). When the juveniles had acquired the learned responses, the tasks of discriminating be- tween the reinforced colored papers and gray papers with various brightnesses were performed to make sure that the fish did not discriminate among the colored papers on the basis of bright- ness. The fish correctly discriminated the rein- forced colored papers from the gray ones, showing similar percentages of correct responses to those for color discrimination. Thus, the juveniles clear- ly used a color cue, but not a brightness cue, in the color discriminations.

Red vs. Green

100

@ Adult (7/7) oJuvenile (9/12)

Correct responses (%)

———— es 9 1 LS 10

Sessions

Fic. 3. Learning curves in juveniles (0 n=12) and adults (@, n=7) trained with the red vs. green stimuli. No significant differences were observed between the curves. See Figure 1 for further ex- planations.

413

To test whether another type of visual discri- mination ability is normal in juveniles, brightness discrimination was performed using white and black papers. Figure 4 shows that juveniles (n= 12) could normally discriminate differences in brightness. Their correct responses clearly in- creased with the increment of training sessions as did those of adults (about 80% correct responses in later sessions) and were not significantly different from those of adults (n=6; Groups, F(1, 16)= 0.33, P>0.05; Days, F(7, 112)=20.29, P<0.01; Groups X Days, F(7, 112)=1.16, P>0.05). This indicates that brightness discrimination ability is normal in juveniles.

Brightness

100

50

@ Adult (6/6) © Juvenile (12/12)

Correct responses (%)

1 A 5 10 15

Sessions Fic. 4. Brightness discrimination in juveniles (O, n= 12) and in adults (@, n=6). A\ll trained juveniles showed adult-like clear learned responses. See Figure 1 for further explanations.

TABLE 1. Percentages of learners in the 3 types of color discrimination measured with 2 training techniques

blue vs. green red vs. blue red vs. green

Two-choice Juveniles 0 (0/11) 43 (10/23) 75 ( 9/12) Adults 100 ( 6/ 6) 100 ( 6/ 6) 100 ( 7/ 7) Go/No-go* Juveniles 0 ( 0/10) 0 (0/ 9) 70 ( 7/10) Adults 33 ( 4/12) 50 ( 6/12) 70 ( 7/10)

The numbers in the parentheses indicate learners/ trained fish. *The data (10 training periods) are from [10].

414 K. OHNISHI

The percentages of learners in this study, together with those measured in the previous study [10], are shown in Table 1. No juveniles trained to discriminate between blue and green could fulfill the criterion defining “learner” (0%), but some juveniles (43%) showed clear learned responses in the discrimination between red and blue. In the discrimination between red and green, many more juveniles (75%) could acquire the learned re- sponses. The brightness discrimination was easy for all trained juveniles to learn (100%). All trained adults fulfilled the criterion defining “lear- ner” in the three types of color discrimination as well as the brightness discrimination.

DISCUSSION

There are several important differences in the training methods used in this study and those of the previous one [10]. The correct responses of the fish were reinforced with a reward (food) and motivated by appetite in this study while they were reinforced with punishment (electro-shock) and motivated by fear in the previous study. The discriminative stimuli were presented simul- taneously using colored papers under a 1000 Ix background illumination in this study, while they were presented successively using monochromatic lights under a 15 lx background illumination in the previous study. It has been reported that the intensity level of background illumination con- siderably afftects the color vision of goldfish. The tetrachromatic color vision becomes trichromatic and, furthermore, the spectral sensitivity function becomes the luminosity function with the decrease in its intensity [7, 8]. Moreover, the fish seem to use a brightness cue at a low intensity (detectable level) of discriminative lights but they use a color cue for high intensity (about 1.0 log unit higher than the dectable level) of discriminative lights for wavelength discrimination when they are trained on an “illuminated” test field [8]. The evidence led to the assumption that the previously observed defective color vision in juvenile goldfish [10] may be due to the low intensity of background illumina- tion. Such a possibility, however, can be rejected because this study, which was performed under a high intensity (about 1000 Ix) of background illu-

mination, also demonstrated defective color vi- sion. Thus, the defect in color vision of juveniles is thought to originate from the developmental pro- cess of the neural system of blue vision.

Another aspect of interest in this study is that some juveniles showed an adult-like discrimination ability in the discrimination between red and blue but none for the discrimination between bule and green (Table 1: The training task of the present study was probably more easy for the fish to acquire discriminative responses than that of the previous study [10]. The numbers of adult learners in the present study were relatively large compared with those in the previous study [10].). This difference in the discrimination ability between blue vs. green and red vs. blue may be due to the different quantal absorption ratio of the mid- wavelengths to the long-wavelengths sensitive con- es among “blue”, “green” and “red” spectral re- gions. The quantal absorption ratio of the mid- wavelengths to the long-wavelengths sensitive con- es in “blue” spectral region is more similar to that in “green” spectral region than that in “red” spectral region [1]. In “blue” and “green” spectral regions, the quantal absorption of the long- wavelengths sensitive cones is not ignorable. Con- trary to this, in “red” spectral region, the quantal absorption of the mid-wavelengths sensitive cones seems to be insignificant. Thus, the discrimination between blue and green is probably more difficult than that between red and blue for juveniles who may have not the functionally matured short- wavelengths sensitive cones (The quantal absorp- tion of the ultraviolet sensitive cones is insigni- ficant in visible spectra [1].). Although this ex- planation is based on the immatureness of the receptor level, the possible immatureness of the post-receptor level should keep in mind. If the short-wavelengths sensitive cones have matured, some juveniles may have already developed a matured blue/red opponent processing pathway but not yet a blue/green one at the same stage of growth.

It is unknown, at present, whether the poor blue vision of juveniles originates from receptoral or post-receptoral elements. However, as already discussed before [10], some possible explanations of the poor blue vision of juveniles can be propo-

Development of Color Vision in Goldfish 415

sed, considering the morphological studies on the development of the retinal neurons [3-S, 11, 12, 14]. Those neurons of goldfish are very unique in such a respect that they continue to grow and further are added newly beyond larval stages into adult life, while in most vertebrates this neuro- genesis is completed during early postembryonic stages. Such unique neurogenesis may have a relation to the prolonged development of blue vision.

REFERENCES

1 Bowmaker JK, Thorpe A, Douglas RH (1991) Ultraviolet-sensitive cones in the goldfish. Vision Res 31: 349-352

2 Cronly-Dillon JR, Muntz WRA (1965) The spectral sensitivity of the goldfish and the clawed toad tad- pole under photopic conditions. J Exp Biol 42: 481- 493

3 Johns PR (1981) Growth of fish retinas. Amer Zool 21: 447-458

4 Johns PR, Easter SS (1977) Growth of the adult goldfish eyes. J Comp Neurol 176: 331-342

5 Meyer RL (1978) Evidence from thymidine labeling for continuing growth of retina and tectum in juve- nile goldfish. Exp Neurol 59: 99-111

6 Neumeyer C (1984) On spectral sensitivity in the goldfish: evidence for neural interactions between different “cone mechanisms”. Vision Res 24: 1223- 1231

7

10

11

12

13

14

15

16

Neumeyer C, Arnold K (1989) Tetrachromatic color vision in the goldfish becomes trichromatic under white adaptation light of moderate intensity. Vision Res 29: 1719-1727

Neumeyer C, Wietsma JJ, Spekreijse H (1991) Separate processing of “color” and “brightness” in goldfish. Vision Res 31: 537-549

Ohnishi K (1991) Goldfish’s visual information processing patterns in food-reinforced discrimina- tion learning between compound visual stimuli. J Comp Physiol 168: 581-589

Ohnishi K (1993) Development of color vision in goldfish: selective delayed maturation of blue vision. Vision Res 33: 1665-1672

Peng YW, Lam DMK (1991) Organization and development of horizontal cells in the goldfish re- tina, I: The use of monoclonal antibody AT101. Visual Neurosci 6: 357-370

Peng YW, Lam DMK (1992) Organization and development of horizontal cells in the goldfish re- tina, II: Use of monoclonal antibody MH1. Visual Neurosci 8: 231-241

Powers MK (1978) Light-adapted spectral sensitiv- ity of the goldfish: a reflex measure. Vision Res 18: 1131-1136

Raymond PA (1990) Horizontal cell axon terminals in growing goldfish. Exp Eye Res 51: 675-683 Yager D (1967) Behavioral measures and theoretic- al analysis of spectral sensitivity and spectral satura- tion in the goldfish Carassius auratus. Vision Res 7: 707-727

Yager D (1969) Behavioral measures of spectral sensitivity in goldfish following chromatic adapta- tion. Vision Res 9: 179-186

these > eal, while” mealeatiballe , € af yy y eed Ae ddbord: LD ctherwemnemiy plese wig pede staal ; ts m23¢t, Aealtahag sl ae vox i od Bho? 0k emerges eis igh ott myabes sAlw tS ay , wit} on re ape: o ah Oe at Whe pit , arene ie) «toon * rev ’ is pe : Si 1? We gj aay LAT ae Ai eagle, heetal duping ih Per ee ee | ea ah deuteron +6 eehedeiees So 1 PKG DM winked? ener ne eee TD ee ee ee FIRMA laaraempiiert iyo eoep tig Mt igh Th a Te. wet Mies > f Lipiel FRO a eer nr a A ya Pe dees ee ed 1S, incr tpiparrre Mal +4 aba y 2° 8 rar { ; wile tad saateh | sang side anane Ve : ’ nownette ful pe i) i Sf ae nya. eth rat ; Ti tle / tami hiy om is ett ; ; ai terres et a #7 i MALE pad . Asa AH Y Prey i t Fy fi ied Gads Bwaals ee 2 ES ap herb tieitc tl 4 Lc rh 2, teain of Pr = ue ‘Cen t (he im Dat FAT nibtale da dog ie : ; ' i fi rH a phy a: yr oma has uP * ’ eft . a * rs. wt oe ve ¢ yy Loire Dh mae aires » Aorhin woe Dn) an eee . ae E fil Pea A we in. Fe sty Val ieee ky Vint Ho po ; ; SAV OTE ty (iN, i ? x. ” £f0l ons ea! jos Fed ] ' — uit li Pinta ow ° i SOE yt roa

ZOOLOGICAL SCIENCE 10: 417-424 (1993)

Cardiotoxic Effects of Adrenochrome and Epinephrine on the Mouse Cultured Myocardium

Keiko Mikamt!, SHIGERU SATO*, MINoRU OTOKAWA?, ANNA PALUMBO“ and Krvoakt KuwASAWA>

‘Department of Legal Medicine, *Central Institute for Electron Microscopic Researches, Nippon Medical School, Tokyo 113, *Laboratory of Biology, Hosei University, Tokyo 194-02, Japan, *Stazione Zoologica, Napoli, Italy, and ° Department of Biology, Tokyo Metropolitan University, Tokyo 192-03, Japan

ABSTRACT—Effects were studied on cultured mouse ventricular myocardial cells of 10~°-10 7M epinephrine and adrenochrome, a) metabolite of epinephrine. Single ventricular myocardial cells in culture had negative chronotropic responses to adrenochrome (10 *M) and positive responses to epinephrine (10° M). However, cultured ventricular myocardial cells in confluent sheets had positive chronotropic responses to adrenochrome (2x 10~*M) and epinephrine (10°-°M). Cumulative LDH leakage from ventricular myocardial cells in confluent sheets increased remarkably after application of 10-*M adrenochrome but only somewhat after application of 10~*M epinephrine. Both adre- nochrome and epinephrine (10~° M) induced mitochondrial fusion and an increase of atrial specific granules, in ventricular myocardial cells in confluent sheets. Epinephrine induced an increase in size of myofibrils. In cultured ventricular myocardial cells, adrenochrome at 2X10~*M induced mitochon- drial damages, as shown by a swollen appearance, but epinephrine never induced such necrotic changes even at concentrations up to 10 -*M. The role of adrenochrome and epinephrine in catecholamine

© 1993 Zoological Society of Japan

cardiotoxicity is discussed.

INTRODUCTION

Myocardial necrosis which occurs during re- perfusion following myocardial ischemia might be, at least in part, due to excessive release of catecho- lamines in the heart [17, 21, 26]. Administration of excessive concentrations of epinephrine, nor- epinephrine or isoproterenol has been shown to induce myocardial necrosis in experimental ani- mals [4, 5, 20, 22]. Furthermore, adrenochrome is known as one of the non-physiological products of epinephrine [8, 19], but the in vivo formation of adrenochrome has been demonstrated in patholo- gical conditions [1, 14, 15]. Yates and his co- workers [29] assumed that in isolated perfused rat hearts the drug produced cell membrane injury and contractile impairment and might be partly responsible for necrogenesis in catecholamine-

Accepted March 8, 1993 Received February 1, 1993

induced cardiomyopathy. On the contrary, Wheatley and his colleagues [28] insisted that the catecholamine-induced myocardial cell damage in isolated rat hearts and papillary muscles was the result of epinephrine and norepinephrine stimula- tion, but not due to adrenochrome, since 10 +M of adrenochrome was required in order to produce a cardiotoxic effect, while the physiological peak concentration of epinephrine was 10-° M.

The cultured myocardium has not been em- ployed to evaluate the effects of adrenochrome and epinephrine on myocardial cells in catechol- amine-induced cardiotoxicity. In the present study, we used mouse cultured myocardial single cells and confluent sheets to examine the effects of the drugs on beat rates and ultrastructure in the preparations. We also examined the cumulative leakage of lactate dehydrogenase (LDH) from the cells in the confluent sheets to the culture medium, as an index of cellular damage [18].

418

MATERIALS AND METHODS

Mouse ventricular myocardial cells were pre- pared as reported previously [16]. Single myocar- dial cells were prepared by seeding cells in plastic dishes (Falcon, 35 mm) at concentrations of 2.5— 5 10* cells/dish. Myocardial cells in a confluent monolayer were prepared by seeding cells in a 4 Well Multidish (Nunclon, Nunc) at concentrations of 110° cells/well or, for an ultrastructural study, in MicroWell Modules (Nunc) at concentrations of 1.6 < 10* cells/well.

Adrenochrome was synthesized at the Universi- ty of Naples, as described by Sobotka and Austin [25]. Adrenochrome and epinephrine bitartrate (Sigma) were applied to the myocardial cells after cultivation for 5 days. The culture medium was replaced with Eagle’s minimum essential medium (MEM) containing 10% fetal bovine serum (FBS) buffered with 10 or 20mM BES (pH 7.3) 1-2 hr prior to the application of the drugs and the myocardial cells were then placed at 37°C in a CO,-free environment. Adrenochrome (10~°, 10~°, 10-*, 2x10~* and 107? M) or epinephrine (10-°, 10-°, 10-4 and 10 M) dissolved freshly in

50

Abeats/min °

-50

-100

(0) 20 40 60

Time after application of adrenochrome (min)

Fic. 1.

K. Mikami, S. Sato et al.

the MEM was applied to the myocardial cells by replacing the normal MEM.

Beating rates of cultured single and confluent myocardial cells were obtained by counting beats in the screen image of individual myocardial cells recorded by a television microscope (a phase con- trast inverted microscope: Nikon, TMD; a color camera: Hitachi, DK-7001N) coupled to a video system (Panasonic, AG-6720A). Statistical analy- ses were performed using the unpaired Student’s t-test. Activity of LDH in the culture medium was measured spectrophotometrically according to Vassault [27]. An ultrastructural study was per- formed by a conventional electron microscopic method previously reported [16].

RESULTS

Chronotropic effects on single ventricular myo- cardial cells

Most of the single ventricular myocardial cells beat spontaneously at a variety of rates ranging from about 50 to 350 per min. Adrenochrome (10-° and 107° M) did not induce chronotropic

100

50

Abeats/min

-50

0 20 40 60

Time after application of epinephrine (min)

Chronotropic effects of adrenochrome (a) and epinephrine (b) on single ventricular myocardial cells in

culture. Graphs express changes in beat rates per min of the cells from the values before applying the drugs

(means+S.D.). ©, control (a normal medium); A, 10~°M; v, 10>° M; ©, 10-7 M.

*** P< (0.001.

*, P<0.01; **P<0.02;

Cardiotoxicity of Catecholamines 419

effects on single ventricular cells in a 60 min period after the onset of drug applications (Fig. 1a), though about half of the cells ceased beating in the normal medium after a 60min exposure to a dosage of 10-°M. A dosage of 10~*M adreno- chrome clearly exerted negative chronotropic effects at 40 min and 60 min after the onset of the application, but not at 20min. At 40 min some cells ceased beating or died, and the beat rates of the beating cells decreased, statistically significant by Student’s ¢-test (Fig. la, P<0.001 vs. control). At 60 min about half of the cells ceased beating or died, and the remaining cells beated weakly and slowly (P<0.02 vs. control). All of the cells died even when they were returned in the normal medium after a 60min exposure to 10°*M adrenochrome.

Epinephrine (10~°, 10~° and 10-*M) induced positive chronotropic effects (Fig. 1b). There were statistical significances between the control value and the beat rates at 40 min (P<0.02) and 60 min (P<0.02) after the onset of application at a dose of 10-°M, and at 20min (P<0.01), 40min (P< 0.01) and 60 min (P<0.001) after the onset of application at a dose of 10°-*M. We never observed inhibitory effects of the dosages on the

Abeats/min

Time after application of adrenochrome

(hr)

cells such as bradycardia or beat arrest.

Chronotropic effects on myocardial cells in con- fluent sheets

Ventricular cells in the confluent sheets appeared to beat synchronously and regularly. Adrenochrome (10° and 10° M) induced tran- sient slightly positive chronotropic effects on the ventricular cells in confluent sheets at 1 hr after application (Fig. 2a). Adrenochrome at 2x 10-4 M maintained the positive chronotropic effects over a period of 3 hr (Fig. 2a), the reverse of the case for single ventricular cells (Fig. la). The difference from the control value at 2hr after application was statistically significant (P<0.05). At a higher dose (10°-*M) beats ceased im- mediately and completely.

Epinephrine (10~°-10~ M) effects on ventricu- lar cells in confluent sheets had a tendency to shift the whole configuration of the graph toward the area of chronotropic positivity (Fig. 2b). Epi- nephrine induced no negative chronotropic re- sponses up to 3 hr after the application.

LDH leakage

Cumulative LDH leakage from beating ven-

100

50 Ff

Abeats/min fo)

-50

(0) 1 2 3

Time after application of epinephrine (hr)

Fic. 2. Chronotropic effects of adrenochrome (a) and epinephrine (b) on cultured ventricular myocardial cells in confluent sheets. Graphs express changes in beat rates per min of the cells from the values before applying the

drugs (means+S.D.). O, control; A, 10~° M; v7, 10° M;

0.05.

, 10°-*M; ©,210~*M (a) and 10-7 M (b). *: P<

420 K. Mikami, S. Sato et al.

400 300 200

20

N

\ 10 N x | IN \ | |\ N | IN N | |N 0” 108. Ons Ome

Cumulative LDH leakage (mU/dish)

2x10 10

Concentration of adrenochrome (M) Fic. 3:

tricular cells in confluent sheets was almost the same in a control experiment, 1 hr after applica- tion of 10° °-2x10~*M adrenochrome. How- ever, LDH leakage was somewhat increased 2 hr after application of 10~° and 2x10~* M adreno- chrome (Fig. 3a). Application of 10°°M_ in- duced a significant increase of LDH leakage at | hr and 2 hr after the onset of the dosage (Fig. 3a), but not at 30 min (data not shown). These responses corresponded to the observation of the cell death accompanied with destruction of the cell mem- brane, as observed under a phase-contrast micro- scope.

Epinephrine (10 °-10~*M) induced no signi- ficant increase of LDH leakage from confluent ventricular cell sheets, though a dose of 10-*M appeared to induce increase of the leakage some- what (Fig. 3b). Indeed, we never observed cell death in epinephrine (10-°-10-?M) under a phase-contrast microscope.

Ultrastructure

At 30min after application of adrenochrome (10-°-10-*M) to ventricular cells in confluent sheets, mitochondria were already fusing (Fig. 4a). Atrial specific granules (atrial natriuretic peptide, [3, 12]) had increased, even among myofibrils, as

» (eo)

(mU/dish)

0 10° 10), 10m om

Concentration of epinephrine

(M)

Cumulative LDH leakage

Cumulative LDH leakage from cultured ventricular myocardial cells in confluent sheets after application of adrenochrome (a) and epinephrine (b) (means+S.D.).

: Lhr after application, NY : 2 hr after application.

well as in perinuclear regions (Fig. 4b). After a dosage of 2X10~*M, mitochondria having more electron-lucent matrix and parallel cristae appear- ed abundantly (Fig. 4c; arrows). Furthermore, some mitochondria were markedly swollen and had electron-lucent matrix, vesicular cristae and amorphous deposits (Fig. 4c; arrow head). But, the specific granules were scarcely seen at this point (Fig. 4c). Glycogen granules and rough endoplasmic reticulum had increased in number around mitochondria (10-°-2x10~*M; Fig. 4a- ©).

Application of epinephrine (10~°-10~* M) for 30 min induced mitochondrial fusion (Fig. 4d, arrows), giant mitochondria (Fig. 4d, arrow head), an increase in number of the atrial specific granules in perinuclear regions and among myofibrils (Fig. 4e, f, arrows) and an increase in size of myofibrils (Fig. 4e) in a dose-dependent manner, but no mitochondrial changes such as seen after the ap- plication of 2x 10~*M adrenochrome (Fig. 4c).

DISCUSSION

The response of single ventricular myocardial cells to adrenochrome was different from that of the ventricular cells in confluent sheets; first, the

Cardiotoxicity of Catecholamines 421

422 K. Mikami, S. Sato et al.

Fic. 4. Electron micrographs of cultured mouse ventricular myocardial cells in confluent sheets at 30 min after the onset of application of adrenochrome or epinephrine. (a) Mitochondrial fusion (arrows). 10~* M adrenochrome.

x 23,500. (b) Atrial specific granules (arrows) appeared.

10-*M adrenochrome. 4,700. (c) Almost all

mitochondria became slightly swollen (arrows), and occasionally some revealed ischemic appearances (arrow head). 2X10~*M adrenochrome. X18,900. (d) Mitochondrial fusion (arrows) and giant mitochondrion (arrow head). 10~°M epinephrine. X18,900. (e) Atrial specific granules increased in number and appeared even among myofibrils (arrows). Thickness of myofibrils was increased, compared with figure 4b (where myofibrils were of normal size). 10~*M epinephrine. 3,900. (f) Many atrial specific granules (arrows) appeared in the perinuclear region and there was abundant Golgi apparatus (arrow heads). 10~*M epinephrine. X23,500.

former response was negative chronotropic one (Fig. 1a) but the latter positive one (Fig. 2a). Secondly, the dose inducing a necrotic response was also different between the two culture sys- tems; single cells often revealed beat arrest in adrenochrome at a relatively low concentration, 10° M, and cell death at 10-4 M. But the cells in the sheets did not have such responses even at 2 X 10-4M up to 2 hr after application (Fig. 3a) and had at a higher dose, namely 10-7 M, although mitochondrial swelling had already appeared 30 min after the onset of application of 2x10-*M adrenochrome (Fig. 4c). In general, grouped cells may be more resistant to various attacks than

single cells, probably due to intercellular coopera- tion by transporting various substances through gap junctions [6, 7], and it is possible that this mechanism can function in setting up the differen- tial chronotropic responses of these culture sys- tems.

Mitochondrial swelling was observed after ap- plication of 2x10~*M adrenochrome (Fig. 4c), which was in agreement with the finding of Dhalla and his co-workers using perfused isolated hearts and intact hearts [23, 24, 29]. However, the dose of adrenochrome required to induce necrotic re- sponses was different between the isolated or intact hearts and the present cultured myocardial

Cardiotoxicity of Catecholamines 423

cells in sheets; in the former preparations in- tracellular edema, swelling of sarcoplasmic reticu- lum, contracted sarcomeres and dissolution of myo- fibrils were observed with application of adreno- chrome at a 10 *M order of concentration. However, these changes were not observed in the cultured myocardial cells at a dose of 2 OTS ML, and necrogenic responses were detected only at a higher dose, 10? M (Fig. 3a). We observed that LDH leakage from non-beating fibroblast-like cell clusters did not increase till 3 hr after application of 10-* M adrenochrome (data not shown). This observation may indicate that cell membrane in- jury responsible for an increase of LDH leakage might be aggrevated by contractile movements. Since the contractile movement of the cultured myocardium in sheets should be less active than that of isolated or intact hearts, it is likely that the necrogenesis of adrenochrome might be less po- tent in cultured ventricular cells in sheets.

The in vivo production of a high concentration of adrenochrome, such as an order of 10-7M, might not be predicted from the physiological serum peak concentration of 10° °M_ of epinephrine [28] except in limited myocardial loci, such as subneuronal terminals which could be exposed to high concentrations of epinephrine. It is unlikely that adrenochrome plays an important role in vivo in inducing catecholamine cardiotoxic- ity, although there remains a possibility that even a low concentration of adrenochrome exhibits a synergic cardiotoxic action in the presence of an additional cell injury factor such as other drugs, oxygen lack, and so on.

Epinephrine induced positive chronotropic effects on both single ventricular cells (Fig. 1b) and cells in sheets (Fig. 2b). The drug failed to induce necrogenic responses even at 10-°?M, as confirmed by enzyme leakage (Fig. 3b) and ultra- structural studies (Figs. 4d-f). This differs from the results with myocardial enzyme leakage of Wheatley et al. [28] using perfused isolated hearts, when the perfusion pressure was high, and also differs from the ultrastructural results of Ferrans et al. [4], using intact hearts. In our experimental model, a hypoxic effect could not be supposed to take place, since myocardial cells were soaked in oxygen-rich media. This is different from isolated

perfused hearts or intact hearts, which might be subjected to oxygen lack due to insufficient coron- ary circulation or to metabolic acceleration [9, 22].

Mitochondrial fusion was observed after ap- plication of either adrenochrome or epinephrine (Figs. 4a, d). This mitochondrial change seems to be a common characteristic of cultured myocardial cells subjected to ethanol attack, as previously reported [16]. The number of atrial specific gra- nules increased in the vicinity of the nucleus and also slightly among myofibrils, after application of either adrenochrome or epinephrine (Figs. 4b, e, f). Some occasions are known in which atrial specific granules are produced in cultured ventricu- lar myocardial cells [2, 10, 16]. The granules often appeared in hypertrophied [28] and volume- overloaded [13] rat ventricles in vivo. Both mitochondrial fusion and an increase in the gra- nules may indicate that the drugs have had a common pathway for induction of the effects, as in the case of some metabolic alterations.

There is a difference in the ultrastructural effects of epinephrine and adrenochrome; epinephrine induced an increase in size of myofibrils, but adrenochrome did not. This may suggest that metabolic acceleratory action of epinephrine causes myocardial hypertrophy. Consequently, it is likely that epinephrine per se up to 10° M is not able to induce cardiotoxicity in cultured myo- cardial cells, since fatal damages to cell membrane and mitochondria have not been perceived in cells to which epinephrine has been applied.

ACKNOWLEDGMENTS

The authors wish to thank Dr. R. B. Hill for revising the manuscript.

REFERENCES

1 Bindoli A, Rigobello MP, Galzigna L (1989) Toxic- ity of aminochromes. Toxicol Let 48: 3-20

2 Bloch KD, Seidman JG, Naftilan JD, Fallon JT, Seidman CE (1986) Neonatal atria and ventricles secrete atrial natriuretic factor via tissue-specific secretory pathways. Cell 47: 695-702

3 Cantin M, Gutkowska J, Thibault G, Milne RW, Ledoux S, MinLi S, Chapeau C, Garcia R, Hamet P, Genest J (1984) Immunocytochemical localiza- tion of atrial natriuretic factor in the heart and

11

12

13

424

salivary glands. Histochem 80: 113-127

Ferrans VJ, Hibbs RG, Cipriano PR, Buja LM (1972) Histochemical and electron microscopic stu- dies on norepinephrine-induced myocardial necrosis in rats. In “Recent Advances in Studies on Cardiac Structure and Metabolism” Ed by E Bajusz, G Rona, University Park Press, Baltimore, pp 495-525 Ferrans VJ, Hibbs RG, Weily HS, Weilbaecher DG, Walsh JJ, Burch GE (1970) A histochemical and electron microscopic study of epinephrine- induced myocardial necrosis. J Mol Cell Cardiol 1: 11-22

Goshima K (1977) Ouabain-induced arrhythmias of single isolated myocardial cells and cell clusters cultured in vitro and their improvement by quini- dine. J Mol Cell Cardiol 9: 7-23

Goshima K, Wakabayashi S (1981) Inhibition of ouabain-induced arrhythmias of ouabain-sensitive myocardial cells (quail) by contact with ouabain- resistant cells (mouse) and its mechanism. J Mol Cell Cardiol 13: 75-92

Green DE, Richter D (1937) Adrenaline and adre- nochrome. Biochem J 31: 596-616

Gremels H (1939) Uber Potentialstoffe. Ergeb Physiol 42: 53-106

Hassall CJS, Wharton J, Gulbenkian S, Anderson JV, Frater J, Bailey DJ, Merighi A, Bloom SR, Polak JM, Burnstock G (1988) Ventricular and artial myocytes of newborn rats synthesise and sec- rete atrial natriuretic peptide in culture: Light- and electron-microscopical localisation and chromato- graphic examination of stored and secreted molecu- lar forms. Cell Tissue Res 251: 161-169

Izumo S, Lompre A-M, Nadal-Ginard B, Mahdavi V (1987) Expression of the genes encoding the fetal isoforms of contractile proteins and atrial natriuretic factor (ANF) in the hypertrophied adult ventricles. J Am Coll Cardiol 9: 1A

Jamieson JD, Palade GE (1964) Specific granules in atrial muscle cells. J Cell Biol 23: 151-172

Lattion A-L, Michel J-B, Arnauld E, Corvol P, Soubrier F (1986) Myocardial recruitment during ANF mRNA increase with volume overload in the rat. Am J Physiol 251: H890-H896

Matthews SB, Hallett MB, Henderson AH, Camp- bell AK (1985) The adrenochrome pathway: A potential catabolic route for adrenaline metabolism in inflammatory disease. Ady Myocardiol 6: 367- 381

Matthews SB, Henderson AH, Campbell AK (1985) The adrenochrome pathway: The major route for adrenalin catabolism by polymorphonuclear leuco- cytes. J Mol Cell Cardiol 17: 339-348

16

17

18

19

20

21

22

23

24

25

26

27

28

29

K. Mikami, S. Sato et al.

Mikami K, Sato S, Watanabe T (1990) Acute effects of ethanol on cultured myocardial cells: An ultra- structural study. Alcohol Alcohol 25: 651-660 Muntz KH, Hagler HK, Boulas J, Willerson JT, Buja LM (1984) Redistribution of catecholamines in the ischemic zone of the dog heart. Am J Pathol 114: 64-78

Murphy MP, Hohl C, Brierley GP, Altschuld RA (1982) Release of enzymes from adult rat heart myocytes. Cire Res 51: 560-568

Palumbo A, d’Ischia M, Misuraca G, Prota G (1989) A new look at the rearrangement of adrenochrome under biomimetic conditions. Biochim Biophys Acta 990: 297-302

Pearce RM (1906) Experimental myocarditis: A study of the histological changes following in- travenous injections of adrenalin. J Exptl Med 8: 400-409

Rona G (1985) Catecholamine cardiotoxicity. J Mol Cell Cardiol 17: 291-306

Rona G, Chappel CI, Balazs T, Gaudry R (1959) An infarct-like myocardial lesion and other toxic manifestations produced by isoproterenol in the rat. Arch Pathol 67: 443-455

Singal PK, Yates JC, Beamish RE, Dhalla NS (1981) Influence of reducing agents on adre- nochrome-induced changes in the heart. Arch Pathol Lab Med 105: 664-669

Singal PK, Dhillon KS, Beamish RE, Kapur N, Dhalla NS (1982) Myocardial cell damage and cardiovascular changes due to I.V. infusion of adre- nochrome in rats. Br J Exptl Pathol 63: 167-176 Sobotka H, Austin J (1951) Betaine hydrazone of aminochromes. J Am Chem Soc 73: 3077-3079 Symes J, Poirier N, Michelborough L, Sniderman A (1977) Catecholamine release in ischemic cells: a mechanism of reperfusion necrosis. Clin Res 25: 673A

Vassault A (1984) Lactate dehydrogenase: UV- method with pyruvate and NADH. In “Methods of Enzymatic Analysis Vol 3” Ed by HU Bergmeyer, Verlag Chemie, Weinheim, 3rd ed, pp 118-126 Wheatley AM, Thandroyen FT, Opie LH (1985) Catecholamine-induced myocardial cell damage: Catecholamines or adrenochrome. J Mol Cell Car- diol 17: 349-359

Yates JC, Beamish RE, Dhalla NS (1981) Ventricu- lar dysfunction and necrosis produced by adre- nochrome metabolite of epinephrine: Relation to pathogenesis of catecholamine cardiomyopathy. Am Heart J 102: 210-221

ZOOLOGICAL SCIENCE 10: 425-430 (1993)

© 1993 Zoological Society of Japan

Immunochemical Detection of GTP-binding Protein in Cephalopod Photoreceptors by Anti-peptide Antibodies

Tatsuo Suzuki’, KinyA Narita”, KAzUO YOSHIHARA’,

Kazuo Nacat! and Yum Kito~

"Department of Pharmacology, Hyogo College of Medicine, Nishinomiya, Hyogo 663, *Department of Biology, Faculty of Science, Osaka University, Toyonaka, Osaka 560, 3Suntory Institute for Bioorganic Research, Shimamoto, Osaka 618, Japan

ABSTRACT— We prepared polyclonal antibodies (Pab) against the following peptides: partial sequ- ences of bovine transducin a subunit including the ADP-ribosylation sites sensitive to cholera toxin (CTX) and pertussis toxin (PTX) and the N-terminus of Drosophila GTP-binding protein qa (DGqN) ; Pab CTX, Pab PTX and Pab DGaN, respectively. These antibodies were specific to the peptides used as antigen and no crossreactivity was observed. Pab CTX and Pab PTX reacted with bovine transducin a subunit and the reactivity was lost by preincubation with the specific antigen peptide. Proteins of 41— 42 kDa in octopus and squid photoreceptor membranes were recognized by Pab DGqN but not by Pab CTX or Pab PTX. Anti-a antibody (GA/1) reacted with the same bands as Pab DGqN recognized. These results suggest that the major GTP-binding protein in cephalopod photoreceptors is a Gq-type,

similar to Drosophila Gq.

INTRODUCTION

Signal transduction of vertebrate photoreceptors has been well studied and the main sequence of events at the molecular level is well established [16, 20]. Light triggers rhodopsin to activate GTP-binding protein (G-protein), transducin, which in turn activates an effector enzyme, phos- phodiesterase, resulting in a decrease of cyclic guanosine monophosphate (cGMP) and the clo- sure of cGMP-dependent cation channels.

In invertebrate photoreceptors, electrophysiolo- gical studies have suggested that, besides cGMP, inositol trisphosphate (IP3) and/or calcium are candidates for intracellular second messengers [2, 6, 16]. Biochemical studies have shown light- dependent binding of non-hydrolysable GTP ana- logues to the photoreceptor membrane [17, 23] and light-dependent GTPase activity [3, 4, 13, 23], which indicate involvement of G-protein in photo- transduction in invertebrate photoreceptors.

Accepted March 18, 1993 Received January 4, 1993

Transducin, the key molecule in vertebrate phototransduction, is susceptible to modification (ADP-ribosylation) both by cholera and pertussis toxins [1, 24]. ADP-ribosylation by bacterial tox- ins had been applied to identify G-proteins of invertebrate photoreceptors (mainly in cephalo- pods) but the reported results were contradictory [7, 17, 22, 23], so the type of G-protein, as identified by the ADP-ribosylation, has remained uncertain.

Recently, G-protein was partially purified from squid photoreceptor membrane and its partial sequence was suggesting a Gq-type G-protein [15]. Genes encoding the a subunit of G-protein cou- pling to phosphoinositide-specific phospholipase C have been cloned and amino acid sequences were deduced from cDNAs in both vertebrates and invertebrates (Gqa, G;,a, DGqa) [12, 18]. These three Gq-type proteins have some unique sequ- ences which are not found in other G-proteins so far reported. In the present study, we prepared anti-peptide antibodies which react specifically with the N-terminus of Drosophila Gqa subunit (DGqa) and with peptides including ADP-

426 T. Suzuki, K. Narita et al.

ribosylation sites of transducin @ subunit. The G-protein of cephalopod photoreceptors was in- vestigated using these anti-peptide antibodies.

MATERIALS AND METHODS

Preparation of anti-peptide antibodies

We synthesized three peptides: EQDVLRSR- VKTTGI (residues 167-180) and DIIIKENLKD- CGLF (337-350), corresponding to amino acid sequences around the ADP-ribosylation sites of bovine transducin [8]; and CLSEEAKEQKR- INQE (4-19) of the N-terminal region of Dro- sophila Gqa [18]. The synthetic peptides were purified by reversed phase HPLC (C4 column), and amino acid composition was confirmed by amino acid analysis of each purified peptide. Each peptide of 5 mg was conjugated to 10 mg of bovine serum albumin (BSA, Sigma) using 1 mg of m- maleimidobenzoic acid N-hydroxysuccinimide es- ter (MBS, Sigma) as a cross-linker.

Japanese white rabbits were immunized with 1 ml emulsion of 1mg BSA-peptide conjugates mixed with Freund’s complete adjuvant. Three weeks later, the rabbits were boosted with the same amount of conjugates and Freund’s incom- plete adjuvant, and after two weeks antisera were obtained. The immunoglobulin fraction was purified by ammonium sulfate fractionation and DEAE-column chromatography. BSA-Sepharose complex was made by coupling BSA to CNBr- activated Sepharose 4B (Pharmacia) according to the manufacturer’s instruction, and anti-BSA anti- bodies were removed by the BSA-Sepharose col- umn. The antibodies against peptides around the ADP-ribosylation sites of transducin, sensitive to cholera and pertussis toxins, and against the N- terminus of Drosophila Gqa were named Pab CTX, Pab PTX and Pab DGqN, respectively.

Activities and specificities of the anti-peptide antibodies were determined by the method of enzyme-linked immunosorbent assay (ELISA), us- ing the above three peptides as adsorbed antigens (each 0.5 ug/well). Antibodies diluted serially were reacted with adsorbed peptides. The bound antibodies were determined by anti-rabbit IgG (goat)-peroxidase (Wako Chemicals) by the stan-

dard method with o-phenylenediamine/hydrogen peroxide as substrates; absorbance at 492 nm was determined with a microplate reader.

Preparation of photoreceptor membranes and transducin

Outer segment membranes of octopus (Octopus vulgaris), squid (Todarodes pacificus) and bovine photoreceptors were prepared by the method of sucrose-density-gradient centrifugation _[10]. Membranes floated on 32% sucrose were diluted with 50 mM Tris-HCl buffer (pH 7.4) and precipi- tated by centrifugation. The membrane prepara- tions were used for sodium dodecyl sulfate- polyacrylamide gel electrophoresis (SDS-PAGE) without further treatment.

Bovine transducin was prepared by the method of Kuhn [11]. Outer segment membranes were illuminated and washed three times with hypotonic buffer (5 mM Tris-HCl, pH7.4), and once with 100mM_ Tris-HCl buffer (pH7.4). The mem- branes were then washed three times with the hypotonic buffer containing 200 ~M GTP, and soluble fractions, containing transducin, were combined and concentrated.

SDS-PAGE and immunoblotting

The photoreceptor membranes were solubilized with sample buffer, subjected to SDS-PAGE and transferred to nitrocellulose sheet. The sheet was blocked with 1% gelatin/casein solution. Anti- peptide antibodies were diluted with 1% gelatin/ casein in phosphate buffered saline (PBS) contain- ing 1% BSA and preincubated at room tempera- ture for | hr (1/500 dilution for Pab CTX and Pab PTX, 1/1000 for Pab DGqN), and then incubated with the nitrocellulose sheet at 4°C overnight. The sheet was then treated with anti-rabbit IgG (goat)- peroxidase, and immunoreactive proteins were visualized with 3,3’-diaminobenzidine tetrahy- drochloride (DAB)/hydrogen peroxide solution.

Anti-a@ antibody (GA/1; specific to the GTP/ GDP-binding site, GAGESGKSTIVK; common to Gt, Gs and Go) and anti-@ antibody (SW/1, specific to the C-terminus of common f) were purchased from Daiichi Chemicals Co., and im- munoblot analyses were performed as above (1/ 300 dilution for GA/1 and 1/500 for SW/1).

G-protein in Cephalopod Photoreceptors 427

RESULTS

The specificity and reactivity of the anti-peptide antibodies are shown in Fig.1. Each antibody preparation reacted specifically with the peptide used as antigen, with no crossreactivity. The reactivities to BSA were 1/500-1/1000 of those to the specific peptides.

The transducin fraction extracted from bovine outer segment membranes was subjected to SDS- PAGE, transferred to nitrocellulose sheet and treated with the anti-peptide antibodies. The main components of this fraction were transducin a and B subunits (Fig. 2a). Pab CTX and Pab PTX recognized the transducin @ subunit: apparent molecular mass 39 kDa (Fig. 2b). The reactions were completely inhibited by preincubation of Pab CTX and Pab PTX with respective peptide antigen.

Immunoblot analyses of octopus, squid and bovine photoreceptor membranes were performed using the three anti-peptide antibodies, as shown in Fig. 3b. Pab DGqN recognized a 41 kDa pro- tein of octopus and a 42kDa protein of squid

Pab CTX

0.8

0.6

0.4

0.2

Relative absorbance at 492nm

Pab PTX

membranes but no bovine protein. The reactivity of Pab DGqN against the 41-42 kDa proteins was significantly reduced by preincubation with the N-terminal peptide of DGga (data not shown). Pab CTX and Pab PTX did not recognize the 41- 42 kDa proteins of octopus and squid membranes, though unknown 35kDa bands were weakly stained by Pab CTX. The results of analyses with anti-a antibody (GA/1) and anti-@ antibody (SW/ 1) are shown in Fig. 3c. GA/1 reacted with the same 41-42 kDa proteins in octopus and squid membranes as Pab DGqN recognized, although the reactivity was lower than with transducin, a. The bands with apparent molecular mass of about 35 kDa were equally recognized by anti-8 antibody (SW/1) in the membranes of three species. These results show that cephalopod photoreceptor mem- branes contain Gq-type G-protein and no trans- ducin-like G-protein.

DISCUSSION

The antibodies, Pab CTX and Pab PTX, were proved to be good tools for detection of G-proteins

Pab DGqN

G6 8 IO 12

Dilution (50x 2")

Fic. 1.

Reactivities and specificities of anti-peptide antibodies determined by ELISA. Three peptides corre-

sponding to: bovine transducin sequences around ADP-ribosylation sites sensitive to cholera toxin

(EQDVLRSRVKTIGI,

), and pertussis toxin (DIIIKENLKDCGLF, @); N-terminus of Drosophila Gqa

(CLSEEAKEQKRINOE, M), were used as adsorbed antigens.

428 T. Suzuki, K. Narita et al.

a) b)

Mr CBB Pab CTX Pab PTX (kDa)

94 -

7 -

43 - Fs

ae (Oise “eosilililal

30 = ie a

20) =

14 = :

peptide

Fic. 2. Specificities of immunoreactions of Pab CTX and Pab PTX against bovine transducin a subunit. a)

Coomassie brilliant blue stain (CBB). 14% polyacrylamide gel. b) Immunostains by Pab CTX and Pab PTX (1/ 500 dilution). Pab CTX preincubated with (+) or without (—) 20 ug/ml peptide (EQDVLRSRVKTTGI), and Pab PTX with (+) or without (—) 20 ug/ml peptide (DIIIKENLKDCGLF) at 25°C for 1 hr.

a) b) c) 94 - — ; Gi = -Rh -Rh 43 a a — — a cased, a - To. ie sate ‘ 4 - -o -o i... ome ai) = O S B OS 13 Oy S18 Oo S B ©. SB O Ses Fic. 3. Immunoblotting of octopus, squid and bovine photoreceptor membranes.a) Coomassie brilliant blue stain

(CBB). 12.5% polyacrylamide gel. O, octopus; S, squid; B, bovine. The bands stained by Pab DGqN (@) and by anti-# antibody (©) are indicated. The broad bands around 45 kDa in octopus and squid and 35 kDa in bovine samples are the rhodopsins (Rh). b) Immunostains by Pab CTX, Pab PTX (1/500 dilution) and Pab DGqN (1/ 1000 dilution). c) Immunostains by anti-a antibody (GA/1, 1/300 dilution) and anti-8 antibody (SW/1, 1/500 dilution).

G-protein in Cephalopod Photoreceptors 429

having ADP-ribosylation sites similar to bovine transducin. The peptide sequence corresponding to the N-terminus of Drosophila Gqa is highly conserved among Gqa, G;,a and DGqa, and has low homology with other G-protein @ subunits. The antibody Pab DGqN, therefore, is considered to be specific to these three members of Gq-class G-proteins. When these three antibodies were applied to immunoblot analyses, only Pab DGqN strongly reacted with the 41-42 kDa proteins in cephalopod rhabdomeric membranes.

Reports on ADP-ribosylation of cephalopod rhabdomeric photoreceptor membrane proteins by bacterial toxins have shown that proteins with apparent molecular weight around 40—42 kDa are ADP-ribosylated by pertussis toxin [7, 17, 22], 44- 45 kDa proteins by cholera toxin [7, 23] and some others by both bacterial toxins [17, 22]. The results of our present work show that cephalopod photo- receptor membranes contain no transducin-like G-protein which is detectable by both Pab CTX and Pab PTX (Fig. 3b). The sequence around the ADP-ribosylation site by pertussis toxin in Gi-type G-proteins (Gi, and Gi.) is very similar to that in transducin (12 of 14 residues are identical) [8]. Pab PTX did not react any proteins of squid and octopus membranes. This suggests that a Gi-type G-protein, if any, is below the detectable level. Gq-type G-protein lacks the cysteine residue of the ADP-ribosylation site [18] and so is not a substrate for ADP-ribosylation by pertussis toxin [14, 20]. The protein ADP-ribosylated by pertussis toxin, reported previously [7, 17, 22], is not Gq-type G-protein but may be minor component of G- proteins; Gi or Go. The proteins reported to be modified by cholera toxin [7, 23] are probably Gs-like G-proteins. The results in Fig. 3b, show- ing no Pab CTX-positive 44—45 kDa proteins, do not necessarily exclude the possibility that Gs-type G-protein is involved in rhabdomeric photorecep- tors, because the amino acid sequence around ADP-ribosylation site by cholera toxin in trans- ducin is considerably different from that in Gs [8].

The proteins recognized both by Pab DGqN had apparent molecular weight of 41-42 kDa, which is consistent with the reported molecular weight of a subunit of Gq, 42 kDa [14, 18, 21]. Anti-a anti- body (GA/1) reacted with the same proteins as

Pab DGgqN recognized. The reactivity of GA/1 with 41-42 kDa proteins was lower than with transducin a (Fig.3c), suggesting that the sequence of GITP/GDP binding site of cephalopod Ga is slightly different from the common sequence rec- ognized by GA/1. The Coomassie blue stain of the SDS-PAGE gel in Fig. 3a shows that the 41-42 kDa protein is a substantial component of cephalo- pod photoreceptor membranes, and that its ratio to rhodopsin is nearly the same as that of trans- ducin @ subunit to rhodopsin in bovine photorecep- tor membrane. These results suggest that the major G-protein in cephalopod photoreceptors is a Ggq-type, similar to Drosophila Gq. Our results are consistent with those of Pottinger et al. [15] suggesting that G-protein of squid photoreceptor is Gq-type.

The biochemical pathway of signal transduction is still obscure in invertebrate rhabdomeric photo- receptors. Both IP; and cGMP induce a depolariz- ing response mimicking light stimulation in Limu- lus ventral photoreceptor [2, 6, 9, 16]. Recent biochemical studies provide evidence suggesting that cGMP is not a second messenger in rhab- domeric photoreceptors [5, 19]. Our results sug- gesting that Gq-type G-protein is predominant in thabdomeric photoreceptors strengthen the IP3 hypothesis for invertebrate phototransduction.

ACKNOWLEDGMENTS

We thank Drs Komatsu and Okamura (Hyogo Coll Med) for technical suggestions in preparing anti-peptide antibodies and Dr Gleadall (Tohoku Univ) for reading the manuscript.

REFERENCES

1 Abood ME, Hurley JB, Pappone MC, Bourne HR, Stryer L (1982) Functional homology between sig- nal-coupling proteins: Cholera toxin inactivates the GTPase activity of transducin. J Biol Chem 257: 10540-10543

2 Bacigalupo J, Johnson E, Robinson P, Lisman JE (1990) Second messengers in invertebrate photo- transduction. In “Transduction in biological sys- tems” Ed by C Hidalgo, J Bacigalupo, E Jaimovich and J Vergara, Plenum, New York, pp 27-45

3 Blumenfeld A, Erusalimsky J, Heichal O, Selinger Z, Minke B (1985) Light-activated guano- sinetriphosphatase in Musca eye membranes resem-

10

11

12

13

14

430

bles the prolonged depolarizing afterpotential in photoreceptor cells. Proc Natl Acad Sci USA 82: 7116-7120

Brown JE, Combs A, Ackermann K, Malbon CC (1991) Light-induced GTPase activity and GTP[7S] binding in squid retinal photoreceptors. Vis Neuros- ci 7: 589-595

Brown JE, Faddis M, Combs A (1992) Light does not induce an increase in cyclic-GMP content of squid or Limulus photoreceptors. Exp Eye Res 54: 403-410

Frank TM, Fein A (1991) The role of the inositol phosphate cascade in visual excitation of inverte- brate microvillar photoreceptors. J Gen Physiol 97: 697-723

Fyles JM, Baverstock J, Baer K, Saibil HR (1991) Effects of calcium on light-activated GTP-binding proteins in squid photoreceptor membranes. Comp Biochem Physiol 98B: 215-221

Gilman AG (1987) G proteins: Transducers of receptor-generated signals. Ann Rev Biochem 56: 615-649

Johnson EC, Robinson PR, Lisman JE (1986) Cyc- lic GMP is involved in the excitation of invertebrate photoreceptors. Nature 324:468—470

Kito Y, Seki T, Hagins FM (1982) Isolation and purification of squid rhabdoms. In “Methods in Enzymology” Ed by L Packer, Academic Press, New York, Vol 81, pp 43-48

Kuhn H (1980) Light- and GTP-regulated interac- tion of GTPase and other proteins with bovine photoreceptor membranes. Nature 283: 587-589 Lee Y-J, Dobbs MB, Verardi ML, Hyde DR (1990) dgq: A Drosophila gene encoding a visual system- specific Ga molecule. Neuron 5: 889-898

Nobes C, Baverstock J, Saibil H (1992) Activation of the GTP-binding protein Gq by rhodopsin in squid photoreceptors. Biochem J 287: 545-548 Pang I-H, Sternweis PC (1990) Purification of unique a subunits of GTP-binding regulatory pro- teins (G proteins) by affinity chromatography with immobilized By subunits. J Biol Chem 265: 18707- 18712

15

16

il7/

18

19

20

21

22

23

24

T. Suzuki, K. Narita et al.

Pottinger JDD, Ryba NJP, Keen JN, Findlay JBC (1991) The identification and purification of the heterotrimeric GTP-binding protein from squid (Loligo forbesi) photoreceptors. Biochem J 279: 323-326

Rayer B, Naynert M, Stieve H (1990) Phototrans- duction: Different mechanisms in vertebrates and invertebrates. J Photochem Photobiol 7: 107-148 Robinson PR, Wood SF, Szuts EZ, Fein A, Hamm HE, Lisman JE (1990) Light-dependent GTP- binding proteins in squid photoreceptors. Biochem J 272: 79-85

Strathmann M, Simon MI (1990) G protein diversi- ty: A distinct class of @ subunits is present in vertebrates and invertebrates. Proc Natl Acad Sci USA 87: 9113-9117

Seidou M, Ohtsu K, Yamashita Z, Narita K, Kito Y (1993) The nucleotide content of the octopus photo- receptor cells: no changes in the octopus retina immediately following an intense light flash. Zool Sci 10: 275-279

Stryer L (1986) Cyclic GMP cascade of vision. Ann Rev Neurosci 9: 87-119

Taylor SJ, Smith JA, Exton JH (1990) Purification from bovine liver membrane of a guanine nuc- leotide-dependent activator of phosphoinositide- specific phospholipase C. J Biol Chem 265: 17150- 17156

Tsuda M, Tsuda T (1990) Two distinct light reg- ulated G-proteins in octopus photoreceptors. Biochim Biophys Acta 1052: 204-210

Vandenberg CA, Montal M (1984) Light-regulated biochemical events in invertebrate photoreceptors. 1. Light-activated guanosinetriphosphatase, guanine nucleotide binding, and cholera toxin catalyzed labeling of squid photoreceptor membranes. Biochemistry 23: 2339-2347

Van Dop C, Yamanaka G, Steinberg F, Sekura RD, Manclark CR, Stryer L, Bourne HR (1984) ADP- ribosylation of transducin by pertussis toxin blocks the light-stimulated hydrolysis of GTP and cGMP in retinal photoreceptor membranes. J Biol Chem 259: 23-26

ZOOLOGICAL SCIENCE 10: 431-438 (1993)

Serum Induced Cell Death

TAKESHI Kurita and HipEo Namixk1!

Department of Biology, School of Education, Waseda University, Shinjuku-ku, Tokyo 169-50, Japan

ABSTRACT—In the culture system of human fetal lung fibroblasts (TIG-1) using Eagle’s MEM containing various proportion of fetal bovine serum (FBS), FBS stimulated cell growth within the range below 40%. FBS at 60% and above inhibited cell growth in a dose-dependent manner and also induced cell death. As well as TIG-1, several types of cells presently tested also exhibited cell death in high concentration FBS. Bovine plasma also showed the same type of cytotoxicity as FBS. As a first step toward identifying the molecular basis of serum toxicity, FBS was divided into high and low molecular weight fractions by ultrafiltration and these were tested at various concentrations on variety of cell lines including TIG-1 human-fetal lung fibroblast, HeLa human epithelioid carcinoma cells, CPAE bovine aorta endothelial cell, FR Rat skin fibroblast, B16 mouse melanoma, and rat embryo primary cultures. When the macromolecular fraction was supplemented with inorganic salts and nutrients and its osmolarity was adjusted with Eagle’s MEM, it showed no toxic effects over a blood range of concentrations. By contrast, a high concentration of low molecular- weight fraction induced cell death. These data suggest FBS contains low molecular-weight (<1,000) factor(s) which cause cell growth

© 1993 Zoological Society of Japan

inhibition and cell death.

INTRODUCTION

When cells are cultured in vitro, serum is com- monly added to defined basal media as a source of nutrients and macromolecules essential for growth. As serum is complicatedly constituted of numerous types of biologically active substances, it’s roles in culture is also complex and difficult to clearly defined. A proposal that hormonally de- fined media could be substituted for serum- containing media [5] gave the concept of serum- free media. During the progress of development of the serum-free media, a number of factors such as growth factors, hormones, or cell attachment factors have been identified as essential factors for the cell culture. Despite the efforts, serum is generally still the best additive and many types of cells especially normal cells can survive for long period only with serum, implying the existence of other unknown factors in serum. Some of possible function of serum were postulated [4, 5]. Ever since Carrel and his co-workers reported in 1920’s

Accepted April 2, 1993 Received February 12, 1993 " To whom all correspondence should be addressed.

that sera from adult animals inhibit cell growth [1, 2], serum from young animals especially fetal bovine serum (FBS) has been most commonly used for cultures. Adult sera are known to contain much more immunoglobulins, complement, hor- mones, and lipids at higher concentrations than those from fetal and early animals, and these are thought to inhibit growth. In addition, some reports suggest that neoplastic transformation and chromosomal abnormalities occur less frequently in medium containing FBS as opposed to horse serum [3]. Nevertheless it is well known that FBS occasionally shows cytotoxicity [6], the cause of which is unclear. Since there are relatively few studies of this phenomenon, we began an inves- tigation of its molecular basis. Initially, we found that cells cultured in high concentration of FBS appended to undergo cell death. When the macro- molecular fraction (>M.W. 1,000) was separated from FBS by ultrafiltration and the low molecular weight fraction was substituted by Eagle’s MEM, this macromolecular FBS stimulated growth of TIG-1 human fetal lung fibroblast and some other cell types of cultures in a dose-depended manner without causing cell death. By contrast the low molecular weight fraction (<M.W. 1,000) caused

432 T. Kurita AND H. Namiki

cell death of TIG-1 cell. These results suggested that the FBS low molecular weight fraction con- tained cytotoxic factor(s). Serum, a product of blood coagulation, is formed at sites of tissue injury in vivo and may contain many factors in addition to those in plasma. Thus we also con- ducted a comparative experiment between serum and plasma in cytotoxicity, showing no practical difference.

MATERIALS AND METHODS

Cell cultures

Human fetal lung fibroblasts TIG-1, human epithelioid carcinoma HeLa, mouse melanoma B16 were obtained from JCRB. Mouse fetal skin FR was from ATCC. All cells were maintained in Eagle’s MEM containing 10% FBS. Cell cultures were observed and photographed using Nicon Di- aphot phase-contrast inverted microscope. Rat whole-embryo primary culture (RWEC) was established as follows. Wistar-lamichi rat (Japan Crea co.) embryos at 16 days of age were collected and minced in 60 mm petri-dishes. After digestion in 0.1% trypsin, 0.02% EDTA CM-PBS at 37°C for 30 min, the tissue was minced again and dis- persed in MEM+10% FBS. Undigested tissue clumps were removed with a nylon-mesh, and cells were collected by centrifugation. Cells were seeded in a 35 mm petri-dish/one embryo (Corn- ing) in MEM+10% FBS, and cultures were washed to remove cell debris and red blood cell. Medium was changed everyday, and when the cultures reached confluence, the cells were har- vested and used for experiments.

Serum

We preliminarily tested 14 lots of obtained from several commercial sources (Hyclone, Salmond Smith Biolab Ltd., GIBCO BRL, Boheringer Mannheim, Bio cell, IRVINE) with TIG-1. All showed almost identical cytotoxicity. Heat treat- ment (57°C, 30 min) had no effect on serum toxic- ity. Accordingly, we used FBS from Boheringer mannheim (Lot 562044) for all experiments with- out heat treatment.

Medium

Eagle’s MEM “Nissui” (Nissui Pharmaceutical Co., Ltd.) was used as the basal medium for all the experiments. Hanks solution “Nissui” (Nissui Pharmaceutical Co., Ltd.) was used for washing the cell.

Ultrafiltration of FBS

Ultrafiltration of FBS was performed using ultrafiltration membrane YM2 (M.W. 1,000) (Amicon Co.). FBS was concentrated 10-fold and diafiltrated with a 10-fold volume of deionized water to remove the low-molecular-weight frac- tion. The resultant macromolecular fraction of FBS (MM-FBS) was again concentrated 10-fold then MM-FBS was diluted to the original FBS volume with 10/9 concentrated Eagle’s MEM, and the pH and osmolality were adjusted to 7.2+0.2 and 290+ 10 mosmol/Kg-H,0O, as the ultrafiltrated FBS (UF-FBS). The FBS filtrate (<M.W. 1,000) was freeze-dried and dissolved in 90% volume of deionized water, and the insoluble fraction was removed by centrifugation. The pH of supernatant was adjusted to be pH 7.4 with 150 mM HCl, and MM-FBS was diluted 10-fold with this supernatant to prepare recombined FBS (RC-FBS). The osmolarity of RC-FBS was slightly higher than that of FBS (298-306 mosmol/Kg-H2O). The ultrafil- trate, low molecular fraction of FBS, was adjusted to pH 7.4 with 0.15 M HCl, almost equivalent to FBS in osmotic-pressure. This prepared ultrafil- trate was used as medium for culturing cells.

Preparation of plasma

Fresh bovine carotidal blood obtained from a local slaughterhouse was immediately added with sodium citrate (final conc. 0.5%) and kept at 15°C. The blood was centrifuged at 6,800 g from 30 min in a continuous centrifuge to remove the hemocyte fraction. The supernatant was stepwise microfil- tered and finally passed through a 0.22 um filter. After the keeping at 4°C for about 1 week, the supernatant was divided into three fractions and processed just before experiments as follows: 1. fibrin was simply spinned out (crude plasma), 2. treated with 1% of 100 unit/ml thrombin (Sigma) at 37°C for 16hr and contrifuged (thrombin-

Serum-Induced Cell Death 433

treated plasma), and 3. incubated at 37°C for 16 hr without thrombin treatment and centrifuged (incu- bated plasma).

Measurement of Serum effect

Cells were harvested from confluent 25 cm’- culture flasks (Corning) after treatment with 0.25% trypsin and 0.02% EDTA in CME-PBS. An appropriate number of cells in 100 pl/well MEM-+ 10% FBS were seeded into 96 well culture plates (Corning). The cell number used resulted in

1.0 10* cells/well in a 24-well plate, and after 24 hr incubation at 37°C the medium was changed to MEM containing the indicated concentrations of FBS. Six days later photographs were taken. Bar=100 pm.

1/20-1/5 confluence in each well, and they were as follows; TIG-1 4x 10°/well, CPAE 3x 10°/well, FR 4% 10°/well, HeLa 3x 10°/well, B16 6x 107/ well, RWEC 6 10?/well, and 6x 10*/well.

After overnight incubation at 37°C in 5% CO, the culture medium was removed. The wells were washed with Hanks balanced solution and 100 pl of test medium was added. After 6 days of culture, the cell number in each well was determined using a Coulter Counter ZM (Coulter Electronics).

e seeded at

434 T. Kurita AND H. Namiki

RESULTS

TIG-1, human fetal lung fibroblasts, were cul- tured in Eagle’s MEM containing various propor- tion of FBS. FBS stimulated cell growth within the range below 40%. FBS at 60% and above inhi- bited cell growth in a dose-dependent manner and also induced cell death (Fig.1, 2). The time required for the cell death was 2-3 days at 60% FBS during which slight cell growth was observed. On the other hand FBS at 80% and above induced complete cell death within 24 hr accompanied with no cell growth. FBS from several different com- mercial sources had essentially the equal effects on TIG-1 cells. Figure 1 shows the morphologices of TIG-1 cells cultured for 6 days in media containing several different concentrations of FBS. FBS cytotoxicity was dependent upon cell density affected for the degree of the cytotoxicity of FBS. The higher density, the slower the death of the TIG-1 cells (data not shown).

In order to estimate the molecular-weight of the cytotoxic factor(s), we separated FBS into a mac- romolecular fraction and a low molecular weight

60000 50000 40000

30000

CELL NUMBER / WELL

20000

10000

20 40 60 80 100 SERUM(%)

Fic. 2. Effect of serum concentration of TIG-1 cell growth. 0): FBS, @: UF-FBS, &: RC-FBS. Human fetal lung fibroblast TIG-1 cells were seeded at 4.0 x 10° cells/well in a 96-well plate, and after 24 hr incubation at 37°C medium was changed to MEM containing the indicated concentrations of FBS. Six days later, cell number was determined with a Coulter Counter.

fraction by ultrafiltration, and pH and osmolarity were then adjusted. Cytotoxicity was not retained above ultrafilter (Amicon Co.) of the following sizes: YM10 (M.W. 10,000), YM5 (M.W. 5,000), YM2 (M.W. 1,000) (data not shown). Accordingly we used ultrafilter YM2 for the experiments. UF- FBS showed no cytotoxicity and stimulated cell growth dose-dependently (Fig. 2). Recombined FBS (RC-FBS) promoted cell growth more effec- tively, but the cytotoxicity was comparable to intact FBS. These results indicated that the low molecular-weight fraction of FBS is necessary for the cytotoxicity.

To determine whether the low-molecular-weight fraction is sufficient for cytotoxicity, TIG-1 cells were cultured in various concentration of FBS- filtrate (from 10% to 100%). Cell death was maximal in 80% filtrate (Fig. 3). The FBS filtrate had cytotoxic activity, but it was slightly less than intact FBS. Thus most of the activity was able to pass through the filter.

4000 J _l Ww

= 3000 P= Ww a

= 2000 4 _ _ Ww )

1000

0

0 20 40 60 80 100 FILTRATE (%) Fic. 3. Cytotoxicity assay of the FBS ultrafiltrate. Hu-

man fetal lung fibroblasts (TIG-1) were seeded at 4.0 10° cells/well in a 96-well plate, and after 24 hr incubation at 37°C the medium was changed to MEM containing the indicated concentrations of FBS ultrafiltrate. Six days later, cell number was determined with a Coulter Counter.

The cell death induced by FBS is not a specific phenomenon for TIG-1 human fetal lung fibrlblasts. We observed that several other types of cells also ceased to grow and died in FBS. Figure 4 shows cultures of TIG-1 and HeLa cells in media

Serum-Induced Cell Death 435

containing FBS or UF-FBS. Figure 5 depicts the growth patterns of four types of cells cultured in media containing 10-100% of FBS. Though the cells were variously sensitive, all of the type showed cell death in FBS. Interestingly, growth of B16 mouse melanoma was inhibited by a high concentration of UF-FBS. This suggests that growth inhibitor remains in UF-FBS, although it is possible that a small amount of low molecular weight cytotoxic material did not pass completely through the filter. B16 melanoma was very sensi- tive to the cytotoxicity of FBS, and the cytotoxicity of the ultrafiltrate was less than FBS. In addition to TIG-1, HeLa, B16, CPAE, and FR cells, we tested human adult dermal fibroblasts (HDF) and Bovine carotid artery endothelial cells (HH). These culture also showed cell death in high con- centration FBS.

It is possible all of the cell lines examined lost their resistance to cytotoxic substances during long

a.TIG-1 FCS

term culture in media containing 10% FBS. We, therefore, tested rat whole-embryo primary cul- tures (RWEC) (Fig. 6), but the results were simi- lar. All of the cells died in FBS. When the cells from embryos were cultured in 100% FBS, we couldn’t observe any growing cells (data not shown). In the case of RWEC, cell-density was related to cell sensitivity to FBS cytotoxicity. Low density cultures were more sensitive than high density cultures. It is possible that cell-cell interac- tions or cell adhesion conditions affected cell toler- ance to FBS cytotoxicity.

Cytotoxicity of plasma as compared with serum was studied upon cultures using medium sup- plemented either with the crude plasma, thrombin treated plasma or incubated plasma (Fig. 7). As the result, paractically non of three kinds of plas- ma was different in cytotoxicity from FBS, and TIG-1 cells also exhibited cell death when cultured with high concentration of those plasma.

436 T. Kurita AND H. Namiki

b.HeLa FCS

Fic. 4. Effect of FBS on TIG-1 cells and HeLa cells. a. TIG-1 cells were seeded at 1.0 10* cells/well in a 24-well plate, and after 24hr incubation at 37°C the medium was changed to MEM containing the indicated concentrations of FBS (left) or UF-FBS (right). Six days later photographs were taken. Bar=100 um. b. Human epithelial carcinoma cells (HeLa) were seeded at 1.0 10° cells/well in a 24-well plate, and after 24 hr incubation at 37°C the medium was changed to MEM ontaining the indicated concentrations of FBS (left) or UF-FBS (right).

Six days later photographs were taken. Bar=50 pm.

DISCUSSION

Based upon these results, it is likely that endoge- nous cytotoxic factors exist among the low molecu- lar weight components of serum. FBS ultrafiltrate caused cell death, and the greater part of the cytotoxic factor(s) could pass through the YM2 filter (M.W. <1,000). The cytotoxicity, however, was slightly weaker than that of FBS. When the FBS fractions separated by ultrafiltration were recombined, the resultant mixture was more effec- tive for cell growth but was no less cytotoxic than intact FBS. It may be hypothesized that there may exist a macro-molecular co-factor(s) that increases toxicity or amplifies the signal for cell death. It also possible that some of the toxic factor(s) was

retained above ultrafilter by binding to macro- molecules. Furthermore, we found that in some cases small amounts of high molecular-weight of proteins (>M.W. 3000) leaked through the mem- brane (<10 ug/ml) (data not shown). It is also possible that low concentration of proteins play an important role in inducing cell death. At this point it is not possible to conclude with certainly that cell death was exclusively caused by low molecular weight factor(s). It is, however, clear that these factors are necessary for the cytotoxicity.

In answer to the question that the serum cytoto- xicity might extraordinarily be produced only at blood coagulation and, if so, plasma might not exhibit such the toxicity, the present data of Fig. 7 that non of preparations of plasma differed from

CELL NUMBER / WELL

25000

20000

15000

10000

5000

125000

CELL NUMBER / WELL

100000

75000

50000

25000

CPAE

B16

20 40 60 SERUM(%)

20 40 60

SERUM(%)

Serum-Induced Cell Death 437

100

80 100

CELL NUMBER / WELL

120000

100000

80000

CELL NUMBER / WELL 3 r=) ° r=)

0 20 40 60 80 100 SERUM(%) HeLa 0 20 40 60 80 100 SERUM(%)

Fic. 5. Effect of serum concentration of growth of CPAE, FR, HeLa, and B16 cell lines. 0: FBS, @: UF-FBS. Cells were inoculated at the following densities in 96-well plates; CPAE 3 x 10°*/well, FR 4 10°/well, HeLa 3x 10°/ well, B16 6x 10°/well. After 24 hr incubation at 37°C the medium was changed to MEM containing the indicated concentrations of FBS. Six days later, cell number was determined with a Coulter Counter.

CELL NUMBER / WELL

150000

125000

100000

75000

50000

25000

20

40 60 SERUM(%)

80 100

Fic. 6. Serum effect on the primary culture of a rat

whole-embryo. A primary culture of a rat whole embryo cells was seeded 6.0 10° (O: FBS, O: UF-FBS) and 6.0 10* (ml: FBS, @: UF-FBS) cells/ well in a 24-well plate. After 24 hr incubation at 37°C the medium was changed to MEM containing the indicated concentrations of FBS. Six days later, cell number was determined with a Coulter Counter.

438 T. Kurita AND H. Namiki

40000 - _1 30000 _ Wu = i ur 20000 = > Zz _ mt ul 10000

0 Seem [ T ast 20 40 60 80 100 PLASMA(%) Fic. 7. Effect of plasma concentration on growth of

TIG-1 fibroblast. @: crude plasma, 0: thrombin- treated plasma, A: incubated plasma. Human fetal lung fibroblast TIG-1 cells were seeded at 5.0 10° cells/well in a 96-well plate, and after 24 hr incuba- tion at 37'C medium was changed to MEM contain- ing the indicated concentrations of bovine plasma. Six days later, cell number was determined with a Coulter Counte.

FBS suggested the cytotoxicity existed normally in blood.

As to the question of why the toxic factor(s) do not cause cell death in vivo, one explanation may be that the cells are not exposed to the same high concentrations of these factors that are present in vitro. Most tissues are protected by an endothelial barrier and are not exposed to blood directly. If so, endothelial cells in vivo may be less sensitive to FBS cytotoxicity. CPAE bovine aorta endothelial cells showed cell death in FBS, however (Fig. 4B), it may be that the CPAE cell line lost this property during long-term culture. In a preliminary experi- ment we have that bovine aorta endothelial cells in the primary culture survived and even grew in 100% FBS (data not shown). By contrast, rat whole-embryo primary cultures did not grow and almost all the cells died in FBS (Fig. 6). It seems that most cell types except vascular endothelial cells can not survive in FBS in vitro.

The molecular identity of the cytotoxic factor is still unclear, but cells cultured in either Hank’s balanced salt solution and Dulbecco’s PBS did not showed cell death (data not shown). This suggests that cell death is not caused by simple starvation, nutritional imbalance, or a defect of physicoche- mical conditions of culture.

In vivo, all normal cell growth is well controlled, but once cells are transplanted in vitro, they start to rapidly proliferate in serum containing media. Vigorous cell growth may reflect the response at a site of injury. Therefore it is possible that cytotox- ic factor(s) is a negative growth regulator or an inducer of cell death in tissue injured in vivo.

ACKNOWLEDGMENTS

TIG-1 was provided by Japanese Cancer Research Resources Bank (JCRB). This work was supported in part by a donation from KIRIN brewery Co. LTD. Authors are also grateful to Dr. Robert, A. Shiurba, Mitsubishi-Kasei Institute of Life Sciences, for critical reading of the manuscript.

REFERENCES

1 Baker LE, Carrel A (1925) Lipids as the growth- inhibiting factor in serum. J Exp Med 42: 143-154

2 Carrel A, Ebeling AH (1921) Age and multiplication of fibroblasts. J Exp Med 34: 143-154

3 Evans JV, Jackson LJ, Andersen WF, Mitchell JT (1967) Chromosomal characteristic and neoplastic transformation of C3H mouse embryo cells in vitro in horse and fetal calf serum. J Nat Cancer Ins 38: 761- 769

4 Ham RG, Mckeehan WL (1978) Nutritional require- ment for clonal growth of nontransformed cell. Japan Sci Soc Press Tokyo, pp 63-115

5 Hayashi I, Sato G (1976) Replacement of serum by hormones permits growth of cells in defined medium. Nature 259: 132-134

6 Sato G (1981) A working hypothsis for cell culture Hormones and cell regulation. INSERM Europ Symp, Elsevier/North-Holland Biomedical Press, Amsterdam, 5: 1-13

ZOOLOGICAL SCIENCE 10: 439-447 (1993)

Effect of the Oesophageal Mesenchyme on the Differentiation of the Digestive-tract Endoderm of the Chick Embryo

SUSUMU MATSUSHITA

Department of Biology, Tokyo Women’s Medical College, Tokyo 162, Japan

ABSTRACT—The endoderm of various regions of the digestive tract of 6-day chick embryos was cultured in vitro in recombination with the oesophageal or other digestive-tract mesenchymes of 6-day chick embryos, and the differentiation of the epithelium was examined with attention given to the appearance of the stratified squamous epithelium characteristic of the mature oesophageal epithelium.

The oesophageal endoderm developed stratified squamous epithelium in high frequency in the presence of the oesophageal mesenchyme, but in low frequency when recombined with the other mesenchymes, which suggested the supporting action of the oesophageal mesenchyme on the proper differentiation of the oesophageal endoderm. Among the other endoderms of digestive tract, only the proventricular endoderm developed stratified squamous epithelium, when cultured in recombination with the oesophageal mesenchyme. The squamous cells of this epithelium contained numerous tonofilaments as those of the stratified squamous epithelium of the intact oesophagus. The ability to elicit stratified squamous epithelium in the proventricular endoderm was shown to be confined to the oesophageal mesenchyme. Thus, the oesophageal mesenchyme was likely to induce oesophagus-type differentiation at least in the proventricular endoderm.

© 1993 Zoological Society of Japan

INTRODUCTION

It is well known that the epithelial-mesenchymal interaction is prerequisite for the development of the digestive tract of the Aves [1, 14, 24]. Numer- ous studies have shown that the mesenchyme of the digestive tract can induce some endoderms to differentiate in a mesenchyme-dependent fashion. The duodenal mesenchyme of the young chick embryo was reported to elicit, in the gizzard endoderm associated to it, the intestine-like simple columnar epithelium forming villus- or previllous ridge-like structures, which developed brush- border structure and its enzymes [2, 5-7]. The proventricular mesenchyme induced the associated gizzard or oesophageal endoderm to form com- pound gland and to produce embryonic pepsi- nogen or pepsinogen mRNA, which was specific to the differentiated proventricular epithelium (3, 18-20]. Recently, it was demonstrated that the gizzard mesenchyme could induce the duodenal

Accepted March 30, 1993 Received December 28, 1992

endoderm to become an epithelium forming tubu- lar glands with prominent mucus production, which resembled the intact gizzard epithelium [12], though it is not yet proved whether the chemical nature of the mucus produced in the recombinates was the same as that of the gizzard mucus. Thus, it may be possible that the mesenchymes of most regions of the digestive tract of the young chick embryo possess the ability to induce the region- specific differentiation at least in some endoderms.

The effect of the oesophageal mesenchyme on the differentiation of the endodermal epithelium was also studied so far. The undifferentiated allantoic endoderm was reported to become pluri- stratified epithelium showing some resemblance to the embryonic oesophageal epithelium when re- combined with the oesophageal mesenchyme of the young avian embryo and cultured in vitro or in vivo on the chorioallantoic membrane of the chick embryo [21, 22], but it became intesinal epithelium with sucrase as well as cloacal epithelium when cultured long enough to achieve full differentiation in the coelomic cavity of the chick embryo [11]. The oesophageal mesenchyme was also reported

440 S. MATSUSHITA

to elicit the yolk sac endoderm to become pluri- stratified epithelium [9], but the intestinal epithe- lium with brush border and its enzyme activities also developed in the same type of recombination [9, 10]. The recombination study using the epithe- lia and mesenchymes of digestive tract demon- strated the heterotypic differentiation of some endoderms cultured in recombination with the oesophageal mesenchyme [23]. Thus, it is still obscure whether the oesophageal mesenchyme has the inductive ability of mesenchyme-dependent differentiation. The present study intended to examine whether the oesophageal mesenchyme has the region-specific inductive influence or not, by the recombination experiment carried out under in vitro culture condition. As for the marker of oesophageal differentiation, appearance of the stratified squamous epithelium was examined, since this epithelium was the characteristics of the fully differentiated oesophageal endoderm and was far from being mistaken for other types of epithelium. Though histological is the marker of oesophageal-type differentiation adopted in this study, the stratified squamous epithelium could be regarded as closely related to the cytodifferentia- tion of the oesophageal epithelial cells, since it appears only in the terminally-differentiated oesophagus of the chick near and after hatching [4, i, 1S, 16)).

MATERIALS AND METHODS

Animal

Embryos of the White Leghorn chick were used throughout the experiments.

Preparation of tissue fragments

The middle portion of the oesophagus and of the small intestinal fragment between the bile duct entrance and the yolk stalk, and the apex of gizzard body were removed from 6-day-old chick embryos (Fig. 1), and the endodermal epithelium and mesenchyme were separated with the aid of collagenase (Worthington, Code CLS, 0.03% in Tyrode’s solution at 38°C for 1h). Whole portion of the proventriculus was taken from 6-day chick embryos for the proventricular mesenchyme, while

digestive tract of 6-day chick embryo

collagenase treatment —-@Be

| epi. mes. intact recombinate fragment

millipore filter

grid explant

agar medium 1/2 day

Fic. 1. Diagram showing the mode of combination of epithelia and mesenchymes of the digestive tract. OE, oesophagus; PV, proventriculus; GZ, gizzard; SI, small intestine; epi., epithelium; mes., mesen- chyme.

liquid medium

only the posterior part was used for the proven- tricular endoderm or the intact proventricular frag- ment, for the purpose of eliminating the possible contamination of the apparent oesophageal tissue (Fig. 1). For the oesophageal mesenchyme of older stages, the fragment just anterior to the crop was taken from 8-day, 10-day and 12-day embryos. The isolated endoderm and mesenchyme were recombined after washing in serum-supplemented (50%) Tyrode’s solution and then in Tyrode’s solution.

In Vitro culture technique

After cultivation on an agar medium for about half a day at 38°C to ascertain the coherence of endoderm and mesenchyme [23], the recombinates were transfered onto a millipore filter (Nihon Millipore Kogyo K. K., pore size 0.8 wm). The filter with the recombinate was placed on a stain- less-steel grid in a small culture dish containing a liquid medium up to the level of the membrane filter, and the recombinate was cultured at 38°C in 95% air and 5% CO>. The intact fragments of digestive tract were also cultured in the same way.

Inductive Effect of Oesophageal Mesoderm 441

The culture medium consisted of 75% Medium 199 with Earle’s salt (Nissui Seiyaku), 20% 12-day chick digestive organ- and eye-free embryo extract (50% in Tyrode’s solution), 5% fetal bovine serum (GIBCO Lab.), and antibiotics (penicillin 100 units/ml, streptomycin 100 “g/ml). The medium was changed every third day. Since the stratified squamous epithelium was shown in a preliminary experiment to appear in the intact oesophageal explant after 12 days’ cultivation in a liquid medium, explants were cultivated for 12 days or longer up to 18 days.

Histology

After cultivation, the explants were fixed with ice-cold 95% ethanol for 4hr. Six wm paraffin sections were stained with alcian blue (AB)- hematoxylin. Some explants were fixed in 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer at 4°C for 2 h and then in 1% O,O, in the same buffer at 4°C for 1h. They were dehydrated with ethanols and embedded in Embedding Resin (TAAB). Ultra-thin sections were stained with methanolic solution of uranyl acetate and with lead citrate.

RESULTS

Appearance of stratified squamous epithelium in the digestive-tract endoderm in the intact digestive-tract fragments and in the recombinates with the oesophageal mesenchyme

In the explants of intact digestive-tract frag- ment, the stratified squamous epithelium appeared only in the explants of oesophageal fragment cul-

TABLE 1.

tured for 12 or 18 days (Table 1, Fig. 2). Columnar epithelium producing AB-stained mucus that might probably correspond to the oesophageal mucous gland epithelium was also found in the oesophageal explants (Fig. 2). In the proventricu- lar explants, columnar epithelium with apical AB- staining differentiated, part of which occasionally developed glandular invagination (Fig. 3). Giz- zard explants often formed lots of tubulargland- like structures or intra-epithelial invaginations, which were lined with AB-positive mucus- secretory epithelium. In the small-intestinal ex- plants, simple columnar epithelium with goblet cells appeared. Thus, the appearance of the stra- tified squamous epithelium was confirmed to be specific to the oesophageal-type differentiation. When cultured in the presence of the oeso- phageal mesenchyme, the proventricular en- doderm as well as the oesophageal endoderm developed the stratified squamous epithelium after 12 days’ cultivation (Table 1). It appeared in the recombinates with the oesophageal endoderm in wide area, while in those with the proventricular endoderm it appeared in various degree, which varied from only a small number of a tiny focus consisting of several squamous cells to a rather wide area (Figs. 4, 5). Transmission electron mi- croscopy showed that numerous tonofilaments were contained in the squamous cells of this stra- tified squamous epithelium, and ciliated cells were often found (Figs. 6a, b). These were characteris- tics found in the intact oesophagus during normal development [4, 15] and in the explants of intact oesophagus in the present study (Figs. 7a, b). The possibility that the stratified squamous epithelium found in these recombinates originated from the

Appearance of stratified squamous epithelium in the digestive-tract

endoderm in the intact digestive-tract fragments and in the recombinates with the oesophageal mesenchyme cultured in vitro

Grafts developing stratified squamous epithelium

Origin of endoderm

Intact fragments

Recombinates

Oesophagus 16/17 (94%) 7/ 9 (78%) Proventriculus 0/17 ( 0%) 20/21 (95%) Gizzard 0/11 ( 0%) 0/11 ( 0%) Small intestine 0/15 ( 0%) 0/12 ( 0%)

442 S. MATSUSHITA

2 gpe se TW S

we

a

Pe ra s

Inductive Effect of Oesophageal Mesoderm

oesophageal epithelial cells contaminated in the oesophageal mesenchyme was denied by the fol- lowing findings. Serial sections of the isolated oesophageal mesenchyme revealed the absence of oesophageal epithelial cells, and the cultivation of the mesenchyme alone for 12 days or longer never developed stratified squamous epithelium. In rare cases (1 out of 12), the isolated mesenchyme contained a small epithelial vesicle of unknown origin in the presumptive adventitial tissue outside of the muscular layer, and this epithelial cells did not become stratified squamous epithelium after cultivation. In the other region of the recombi- nates with the proventricular endoderm and in the recombinates with the gizzard endoderm, mucus- secreting epithelium with abundant undulations or short invaginations appeared (Figs. 4, 5, 8). These epithelial cells may be regarded as the differenti- ated proventricular or gizzard mucous cells, but discrimination from the oesophageal mucous cells was impossible in the present study. The small- intestinal endoderm became simple columnar epithelium with goblet cells. In a few explants, non-goblet mucous cells were also found.

Effect of various digestive-tract mesenchymes on the differentiation of the proventricular and oesophageal endoderm

The proventricular and oesophageal endoderms were found to develop stratified squamous epithe- lium in the presence of the oesophageal mesen- chyme. Then, the effect of other mesenchymes on these endoderms was analyzed.

443

As shown in Table 2, the proventricular en- doderm developed stratified squamous epithelium only in the recombinates with the oesophageal mesenchyme, in which AB-positive mucous epithelium also appeared. The AB-stained epithe- lium with occasional glandular invagination differ- entiated in the presence of the proventricular mesenchyme, and the mucus-secretory epithelium forming intra-epithelial invaginations or tubular gland-like structures developed in the presence of the gizzard mesenchyme. In the recombinates with the small-intestinal mesenchyme, AB-positive epithelium was less frequently found.

The oesophageal endoderm developed stratified Squamous epithelium in high frequency in the presence of the oesophageal mesenchyme. In the presence of other digestive-tract mesenchymes, stratified squamous epithelium was found in low frequency (Table 2). In the recombinates with the proventricular mesenchyme, AB-positive epithe- lium with occasional glandular invagination differ- entiated but even a small focus of stratified squamous epithelium rarely developed (Table 2, Fig.9). In the recombinates with the gizzard mesenchyme, AB-positive stratified or simple epithelium and stratified squamous epithelium appeared (Fig. 10). In the recombinates with the small-intestinal mesenchyme, AB-positive colum- nar epithelium appeared. Stratified squamous epithelium was found only in small areas of a few recombinates.

Fic. 2. An explant of intact oesophageal fragment cultured for 18 days. Stratified squamous epithelium (s) and

mucous epithelium (arrows) differentiated. 230. Fic. 3.

An explant of intact proventricular fragment cultured for 18 days. A glandular invagination (g) is formed,

and the epithelium is often stained with alcian blue in the apical portion (arrowheads). 230. Fic. 4. A recombinate of proventricular endoderm and oesophageal mesenchyme cultured for 18 days. Typical stratified squamous epithelium (s) and mucous epithelium (arrow) are seen. X230.

Fic. 5.

Small foci (arrows) of stratified squamous epithelium that appeared in the mucous epithelium found in a

recombinate of proventricular endoderm and oesophageal mesenchyme cultured for 18 days. 230.

Fic.

Fic.

6. Transmission electron microscopy of an epithelium of a recombinate of proventricular endoderm and oesophageal mesenchyme cultured for 18 days. a. Cells in the middle layer of the stratified squamous epithelium, which are interconnected with each other by desmosomes (arrowheads) and contained abundant tonofilaments (t). n, nucleus. 24000. b. A ciliated cell. 12000.

7. Transmission electron microscopy of an epithelium of an explant of intact oesophageal fragment cultured for 18 days. a. Cells in the middle layer of the stratified squampous epithelium. Desmosomes (arrowheads) and numerous tonofilaments (t) are seen. 24000. b. A ciliated cell. 12000.

444

S. MATSUSHITA

Effect of the oesophageal mesenchyme of aged embryos

As shown in Table 3, the stratified squamous epithelium appeared only in the presence of the oesophageal mesenchyme of 6-day embryos and not in the presence of the mesenchyme of older embryos, suggesting that the inductive ability of the oesophageal mesenchyme declined rather quickly after 6 days of incubation. The epithelium in the recombinates with the mesenchyme of older embryos became mucus-secreting epithelium with abundant undulations or short invaginations.

TABLE3. Appearance of stratified squamous epithelium in the 6-day proventricular endoderm cultured in the presence of oesophageal mesen- chyme of various developmental stages

Grafts developing stratified

Siagze squamous epithelium 6-day 20/21 (95%) 8-day 0/10 ( 0%) 10-day 0/ 8 ( 0%) 12-day 0/ 6 ( 0%)

Fic.8. A recombinate of gizzard endoderm and oesophageal mesenchyme cultured for 18 days. Mucus-secreting epithelium with abundant undula- tions or short invaginations appeared. X230.

Fic. 9. A recombinate of oesophageal endoderm and proventricular mesenchyme cultured for 18 days. Columnar epithelium with apical alcian blue- staining forming glandular invaginations appeared. x 230.

Fic. 10. A recombinate of oesophageal endoderm and gizzard mesenchyme cultured for 18 days. A small area of stratified squamous epithelium (s) isseen. X 230.

TABLE 2. Appearance of stratified squamous epithelium in the proventricular and oesophageal endoderms cultured in the presence of various digestive-tract mesenchymes

Grafts developing stratified squamous epithelium

Origin of mesenchyme Proventricular endoderm Oecsophageal endoderm

Oesophagus 20/21 (95%) 23/26 (88%)* Proventriculus 0/21 ( 0%)* 1/18 ( 6%) Gizzard 0/11 ( 0%) 6/11 (55%) Small intestine 0/12 ( 0%) 3/11 (27%)

*“, Explants of the intact fragments as well as the recombinates were included.

Inductive Effect of Oesophageal Mesoderm 445

DISCUSSION

The present study clearly showed that the stra- tified squamous epithelium appeared in high fre- quency in the proventricular endoderm cultured in recombination with the oesophageal mesenchyme, demonstrating the ability of the oesophageal mesenchyme to induce the differentiation of stra- tified squamous epithelium. The previous study [23] also reported an occasional appearance of the stratified squamous epithelium in the same type of recombination, but in the present study the stra- tified squamous epithelium appeared in much higher frequency, which might be due to a better culture condition or a longer cultivation period. The ultrastructural features of the stratified squamous epithelium found in the recombinates were shown to resemble those of the stratified squamous epithelium of intact oesophageal ex- plants, suggesting that both epithelia might repre- sent the same cytodifferentiated state. As shown in the present study, the appearance of stratified squamous epithelium in the proventricular en- doderm was confined to the recombinates with the oesophageal mesenchyme. Thus, it was likely that the oesophageal mesenchyme had the region- specific mesenchyme-dependent inducing ability to elicit the oesophagus-like differentiation at least in the proventricular endoderm. The small focus of stratified suamous epithelium would have de- veloped from a single induced cell or a group of a few induced cells in the recombined proventricular endoderm. The inductive ability of the oeso- phageal mesenchyme to elicit stratified squamous epithelium in the proventricular endoderm was shown to decline rather quickly after 6 days of incubation, while the ability to elicit pluristratified epithelium in the allantoic endoderm was reported to be retained until the 11th days of incubation [21]. This discrepancy might be due to the differ- ence in the responding tissue and/or to the differ- ence in the examined histological marker in the two experiments.

Regional difference in the responsiveness of the digestive-tract epithelium to the inductive stimuli of the mesenchyme was repetitively reported [3, 5, 20, 23]. The responsiveness to the oesophageal mesenchyme also showed regional difference in

the digestive-tract endoderms, and was found to be confined to the endoderm of the proventriculus, which situated next to the oesophagus, at least in the 6-day-old chick embryo. That the competence for the inductive action of the mesenchyme was high in the endoderm of its neighboring region(s) was reported also in the case of intestinal dif- ferentiation induced by the duodenal mesenchyme [5] or of the induction of pepsinogen or pepsinogen mRNA by the proventricular mesenchyme [3, 20]. As for the high responsiveness of the proventricu- lar endoderm, the following three possibilities may be considered; 1) some of the proventricular en- dodermal cells might be induced to change their fate and to become oesophageal-type epithelium; 2) undifferentiated cells capable to adopt oeso- phageal or proventricular fate according to the influence of the mesenchyme might be contained in the proventricular endoderm; 3) a small number of oesophageal cells which are to be lost through some selective mechanism(s) during normal de- velopment in the proventriculus might be distri- buted even in the posterior part of the proventricu- lus. It needs further extensive analysis to know which is the case.

The present study also showed that the dif- ferentiation of the oesophageal endoderm was influenced by the mesenchyme. The appearance of stratified squamous epithelium was low in the oesophageal endoderm cultured in recombination with the mesenchyme other than that of the oesophagus. Thus, the other mesenchyme might provide only an inappropriate condition for the differentiation of the stratified squamous epithe- lium or exert an inhibitory effect on its differentia- tion. The proventricular mesenchyme was known to possess an ability to induce proventricular dif- ferentiation [3, 20], and this effect may have contributed to rare appearance of the stratified squamous epithelium in the recombinates with the proventricular mesenchyme. On the contrary, the oesophageal mesenchyme provided a suitable con- dition for the normal differentiation of the oesophageal endoderm. Whether this effect of the oesophageal mesenchyme on the oesophageal en- doderm may be a nutritional one or be related to the inductive influence as exerted over the proven- tricular endoderm is not known. The appearance

446

of the stratified squamous epithelium even in the recombinates with the mesenchyme other than the oesophageal one, though in a rather low frequen- cy, suggests that the oesophageal endoderm of the 6-day chick embryo has already been determined to some extent, as suggested by the presence of self-differentiation potency in the oesophageal en- doderm of this stage [17].

At least at 6 days of incubation and probably at later stages during normal development, the oesophageal mesenchyme would interact mainly with the oesophageal endoderm likely to be en- dowed with the oesophageal fate. Since the oesophageal endoderm of 6-day embryos has also a potency of heterotypic differentiation into proventricular-type [3, 20], the inductive effect of the oesophageal mesenchyme might play, in con- cert with its probable supporting action that pro- vides the approprtiate condition(s), an important role in assurring the correct differentiation of the oesophageal endoderm, as suggested by Mizuno (1975) [13].

ACKNOWLEDGMENTS

The author wishes to express his deep gratitude to Professor Kazuo Utsugi of Tokyo Women’s Medical College.

REFERENCES

1 Haffen K, Kedinger M, Simon-Assmann P (1987) Mesenchyme-dependent differentiation of epithelial progenitor cells in the gut. J Ped Gastroenterol Nutr 6: 14-23

2 Haffen K, Kedinger M, Simon-Assmann PM, Lac- roix B (1982) Mesenchyme-dependent differentia- tion of intestinal brush-border enzymes in the giz- zard endoderm of the chick embryo. In “Embryonic Development, Part B: Cellular Aspects” Ed by MM Burger and R Weber, Alan R Liss, New York, pp 261-270

3. Hayashi K, Yasugi S, Mizuno T (1988) Pepsinogen gene transcription induced in heterologous epithe- lial-mesenchymal recombinations of chicken en- doderms and glandular stomach mesenchyme. De- velopment 103: 725-731

4 Hinsch GW (1967) Ultrastructural differentiation of the epithelium and mucous glands of the esophagus in the chick embryo. J Morph 123: 121-132

5 Ishizuya-Oka A, Mizuno T (1984) Intestinal cytodif- ferentiation in vitro of chick stomach endoderm

10

11

12

13

14

15

16

19

S. MATSUSHITA

induced by the duodenal mesenchyme. J Embryol Exp Morphol 82: 163-176

Ishizuya-Oka A, Mizuno T (1985) Chronological analysis of the intestinalization of chick stomach endoderm induced in vitro by duodenal mesen- chyme. Roux’s Arch Dev Biol 194: 301-305 Ishizuya-Oka A, Mizuno T (1992) Demonstration of sucrase immunoreactivity of the brush border induced by duodenal mesenchyme in chick stomach endoderm. Roux’s Arch Dev Biol 201: 389-392 Ivey WD, Edgar SA (1952) The histogenesis of the esophagus and crop of the chicken, turkey, guinea fowl and pigeon, with special reference to ciliated epithelium. Anat Rec 114: 189-211

Masui T (1981) Differentiation of the yolk-sac endoderm under the influence of the digestive-tract mesenchyme. J Embryol Exp Morphol 62: 277-289 Masui T (1982) Intestinalization of the area-vitellina endoderm cultured in association with digestive- tract mesenchymes. J Embryol Exp Morphol 72: 117-124

Matsushita S (1984) Appearance of brush-border antigens and sucrase in the allantoic endoderm cultured in recombination with digestive-tract mesenchymes. Roux’s Arch Dev Biol 193: 211-218 Matsushita S (1988) Effect of the mesenchyme on the differentiation of the duodenal and gizzard epithelia of the chick embryo. Zool Sci 5: 1264 Mizuno T (1975) Une hypothése sur l’organogenése du tractus digestif. C R Soc Biol 169: 1096-1098 Mizuno T, Yasugi S (1990) Susceptibility of epithe- lia to directive influences of mesenchymes during organogenesis: Uncoupling of morphogenesis and cytodifferentiation. Cell Differ Develop 31: 151-159 Mottet NK (1970) Mucin biosynthesis by chick and human oesophagus during ontogenetic metaplasia. J Anat 107: 49-66

Romanoff AL (1960) The digestive system. In “The Avian Embryo”, Macmillan, New York, pp 429-531 Sumiya M (1976) Differentiation of the digestive tract epithelium of the chick embryo cultured in vitro enveloped in a fragment of the vitelline mem- brane, in the absence of mesenchyme. Roux’s Arch Dev Biol 179: 1-17

Takiguchi K, Yasugi S, Mizuno T (1986) Gizzard epithelium of chick embryos can express embryonic pepsinogen antigen, a marker protein of proventri- culus. Roux’s Arch Dev Biol 195: 475-483 Takiguchi K, Yasugi S, Mizuno T (1988) Develop- mental changes in the ability to express embryonic pepsinogen in the stomach epithelia of chick embryos. Roux’s Arch Dev Biol 197: 56-62 Takiguchi K, Yasugi S, Mizuno T (1988) Pepsi- nogen induction in chick stomach epithelia by reag- gregated proventricular mesenchymal cells in vitro. Develop Growth & Differ 30: 241-250

21

22

Inductive Effect of Oesophageal Mesoderm 447

Yasugi S (1979) Chronological changes in the induc- tive ability of the mesenchyme of the digestive organs in avian embryos. Develop Growth & Differ 21: 343-348

Yasugi S, Mizuno T (1974) Heterotypic differentia-

‘tion of chick allantoic endoderm under the influence

of various mesenchymes of the digestive tract. Wilhelm Roux’s Arch 174: 107-116

23

24

Yasugi S, Mizuno T (1978) Differentiation of the digestive tract epithelium under the influence of the heterologous mesenchyme of the digestive tract in the bird embryos. Develop Growth & Differ 20: 261-267

Yasugi S, Mizuno T (1990) Mesenchymal-epithelial interactions in the organogenesis of digestive tract. Zool Sci 7: 159-170

> a

- i. a ae

nr steel | tay Cory moma ol weaken! =:

mene ed ee neal aan iy, Lait any, salon a ome Naa Penal ftw ane itn ea eo Oh acre ie anlsustl aeions tat SAT Witenet Wie (ptustbeeiaee ie bith =” uve we; Rohenedd we scl al lel pala Hagar: mow eee w ny ety Wl antmapee Hoy, eee wilt 7% Naat Yad Puctie Fania ‘ ; : neni ly clamp mel Waele “a ailing. Pinan » Anh Dew) 4 » WL. ue Sa 1 E eon? iyo Uke See per ony wth ape Ninh -ae- Te = { " rei} i ies nl ; ie j ioe 304 iovheg bo. Ay mio ati) a") Yahi tefnlgy

‘od

ZOOLOGICAL SCIENCE 10: 449-454 (1993)

Immunohistochemical Expression of Inhibin-a Subunit in the Developing Rat Gonads

SatosHI Koike! and Tetsuo NOUMURA

Department of Regulation Biology, Faculty of Science, Saitama University, Urawa, Saitama 338, Japan

ABSTRACT— In order to investigate the roles of inhibin for rat gonadal differentiation, the expression of inhibin-a subunit in the developing rat gonads was determined. Sprague-Dawley rat gonads from the gestational day (GD) 13 through the postnatal day (PD) 21 were fixed in Methacarn solution and immunohistochemically stained with a polyclonal antibody against [Tyr*°] porcine inhibin a-chain (1- 30)NHbz subunit raised in the goat. During fetal period, inhibin-a subunit was detected only in male gonads: slight or moderate staining was observed in Sertoli/supporting cells on GDs 14 and 15, and moderate or marked staining in Leydig cells from GD 17 to 20. Reactivity was negative in female gonads and in the Millerian and Wolffian ducts in both sexes during fetal period. After birth, inhibin-a subunit was moderately expressed in Sertoli cells and slightly in Leydig cells on PD 21 for males, and markedly in the granulosa cells for females. These results indicate that the expression of inhibin-a subunit is stage- and cell-specific during the gonadal development and the inhibin may participate in rat

© 1993 Zoological Society of Japan

testicular differentiation.

INTRODUCTION

Inhibin is a gonadal glycoprotein which reg- ulates pituitary FSH secretion [3, 16, 24] and synthesis [27, 29]. The purification, cloning and sequencing of inhibin cDNA showed that inhibin is a heterodimer comprised of an a-subunit (18 KD) and one of two related £-subunits (14 KD, 8A and AB) joined by a disulfide bond [2, 5, 8, 10, 19, 23]. After the purification of inhibin protein, it has been obvious that inhibin plays a variety of types of roles as hormone, paracrine and autocrine reg- ulators of cellular proliferation and functions in some mammals [for reviews, 4, 7, 14, 28, 30].

Expressions of inhibin a- and 6-mRNAs were seen by using in situ hybridization in the gonads of embryonic rats from gestational day (GD) 14 to birth [22]. Inhibin-subunits mRNAs were local- ized by the stage-specific and tissue-specific man- ners in the developing rat gonads: a-mRNA was seen in the seminiferous tubules and interstitial tissue from GD 14, BA-mRNA only in the intersti-

Accepted December 10, 1992 Received August 31, 1992

' Present address: Upjohn Pharmaceuticals Limited, Wadai, Tsukuba, Ibaraki 300-42, Japan

tial tissue just before birth and BB-mRNA over the tubules from GD 14. Immunohistochemical loca- lizations of a- and @-subunit proteins were exhi- bited in the germ, Sertoli and Leydig cells in fetal rat testes and the germ cells in the fetal ovaries [20]. These results indicate that inhibin may play roles for gonadal cell proliferation and differentia- tion, like the cases of immature and adult gonads.

In the presént study, immunohistochemical ex- pression of inhibin-a subunit was chronologically clarified in the fetal and prepubertal rat gonads from gestational day 13 to postnatal day 21.

MATERIALS AND METHODS

Experimental animals

Crj: CD (Sprague-Dawley) rats in 13 to 20 weeks of age were housed in constant temperature (22+2°C), relative homidity (55+10%) and light- dark cycle (lights on 7:00-19:00). The animals fed purina chow and took the tap water ad libitum. Cohabitation was done in the evening 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

450 S. KoikE AND T. Noumura

smear was found was designated as GD 0, and the day when litter was found was designated as post- natal day (PD) 0.

Preparation of tissues for immunostaining

Dams were sacrificed from GD 13 to 21 and neonates on PD 5, 11 and 21 by carbon dioxide. The gonads and genital ducts dissected from the fetuses and pups were fixed in Methacarn solution consisting of methanol, chloroform, and acetic acid, 6:3:1 in volume for a few hours to over- night. 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 um thickness.

Immunohistochemistry

Sections were deparaffinized with xylene and hydrated in decreasing concentrations of ethanol, and incubated in 6 M urea (ICN Biomedicals Inc.) at room temperature for 30min and to block endogenous peroxidases with 0.5% periodic acid (Sigma Chemical Co.) for 15 min. Sections were subsequently rinsed with 10mM phosphate buf- fered saline (PBS, pH 7.4, Sigma Co.) for 20 min, blocked non-specific staining with 1.5% normal rabbit serum in 10 mM PBS including 0.5% casein (Wako Pure Chemical Industries Ltd.) for 20 min and then incubated with avidin and biotin blocking solution (Vector Laboratory Inc.) for 15 min, each at room temperature. After that, sections incu- bated overnight at 4°C with the polyclonal anti- body against [Tyr*’] porcine inhibin a-chain (1- 30)NH); raised in the goat (gifted from Prof. Sasa- moto, Tokyo University of Agriculture and Tech- nology) at a dilution of 1:40,000 in 10mM PBS including 0.5% casein. Dose-response study indi- cated that this dilution of the antibody gave optim- al labelling results. Following this incubation the sections were rinsed with PBS and then treated with 0.5% biotinylated rabbit anti-goat secondary antibody (Vector Lab. Inc. ABC-peroxidase stain- ing kit Elite) diluted in 10mM PBS containing 0.5% casein for 30min at room temperature. Sections were again washed in PBS and subse- quently incubated with 2% avidin-biotin complex (Vector Lab. Inc. ABC kit Elite) in 10 mM PBS

for 60 min at room temperature. Avidin and biotin were prepared at least 30 min before applied to the sections to allow the complex to form. The sec- tions were again washed in PBS, and the bound antibody was visualized with 0.05% 3,3- diaminobenzidine tetrachloride (Sigma Chemical Co.) in 10mM Tris-buffered saline (Sigma Che- mical Co.) and 0.01% H>O> for 4 min.

Controls included (a) replacing the primary anti- body with normal goat serum, (b) using the prim- ary antibody that had been pre-incubated over- night at 4°C with 1 ~g/ml porcine inhibin-a subunit (1-32) (Peninsula Lab. Inc.) before this mixture was applied to the section in order to check the 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 anti- body to inhibin, with the immunoneutralized anti- body, and with normal goat serum were shown in Fig. 1. Inhibin antibody stained the immature rat Sertoli cells on PD 21, but the neutralized antibody or normal goat serum did not stain any cells. Therefore, these results showed that this polyclon- al antibody specifically stained inhibin-containing cells, because Sertoli cells might be regarded as the major source of inhibin in the immature male rat [see reviews].

Immunohistochemical localization

The immunohistochemical localizations of in- hibin-@ subunit in developing gonads were summa- rized in Table 1. The first positive staining was gained in most of the Sertoli cells on GD 14 (Fig. 2. A) and this sign was also seen in a few Sertoli cells on GD 15. At the second, the Leydig cells were positively stained from GD 17 to PD 5. The intensity of staining was marked on GDs 17 (Fig. 2. B) and 18 and gradually decreased with develop- ment. The number of the Leydig cells with posi- tive reaction was also decreased after GD 20. However, female gonads and _ mesonephric tubules, Miillerian and Wolffian ducts in both

Inhibin-a in Developing Rat Gonads 451

Fic. 1. Demonstration of staining specificity. Male gonads on PD 21 incubated with the inhibin-a antibody (A), with the neutralized antibody (B) or with normal goat serum (C). Staining is seen in the Sertoli cells (arrow) and the Leydig cells (arrow head) of A, but not those of B or C. Bars: 50 um.

TasLe 1. Immunohistochemical localization of inhibin-a in the developing rat gonads

GD: Gestational day, PD: Postnatal day, G: Germ cell, S: Sertoli/supporting cell, P: Peritubular cell, L: Leyding cell, W: Wolffian duct, M: Mullerian duct, MT: Mesonephric tubule, Gr: Granulosa/supporting cell, T: Theca cell, I: Interstitial/stromal cell, Grade, —: No detectable, +: Slight but above background levels, ++: Moderate, + + +: Marked staining, Shade box was shown that the cells or tissues were not found in that day.

452 S. KoIkE AND T. NoumMuRA

Fic. 2. Immunohistochemical localizations of inhibin-a in the perinatal gonads. (A) Male gonad on GD 14 shows the moderate staining in the Sertoli cells (arrow). (B) Male gonad on GD 17 expresses the marked staining in the Leydig cells (arrow head). (C) Male gonad on PD 21 is stained moderately in the Sertoli cells (arrow) and weakly in the Leydig cells (arrow head). (D) and (E) Female gonads on GDs 14 and 17, respectively, do not express any positive sign. (F) Female gonad on PD 21 expresses the marked-reactivity in the granulosa cells (arrow). Bars: 50

ym.

sexes were negative during prenatal period (Fig. 2. on PD 21 (Fig. 2. C and F), and the intensity of D and E). staining was moderate or marked. The slight and/ During postnatal period, positive staining was or faint staining was found in the Leydig cells on observed in most of the Sertoli cells for male | PDs 5 and 21. However, there were no detectable gonads and the granulosa cells for female gonads __ staining to the antibody in any other cells.

a

Inhibin-a in Developing Rat Gonads 453

DISCUSSION

Pattern of immunohistochemical expression of inhibin-@ subunit was partially consistent with the in situ identification of its mRNA [22]. Inhibin-a mRNA was localized in the tubular and interstitial tissues of male gonads during the third trimester gestational period from GD 14 to GD 21. Ogawa et al. [20] reported that inhibin-immunoreactivity was seen in the tubular and interstitial cells for males and the germ cells for females on GD 18.5. But in this experiment, inhibin-a subunit protein was only expressed in Sertoli cells on GDs 14 and 15, and in Leydig cells from GD 17. This incon- sistency might be depend on the applicaiton of different types of antibodies, because antigens used were [Tyr] inhibin-a subunit (1-20) by Ogawa et al. [20] but [Tyr*’] porcine inhibin a- chain (1-30)NHb in our experiment.

Positive reactions in Sertoli cells on GDs 14 and 15 suggest that inhibin might be related to the testicular differentiation of the gonad, especially to the formation of seminiferous tubules and aggrega- tion of the Sertoli cells around the germ cells like activin-A and -B, the homodimers of inhibin-BA and -£B subunits, in immature rat testis [13] be- cause the sexual differentiation of gonad begins from the late GD 13 [11, 12, 17]. However an elucidation of the exact role of inhibin and/or inhibin-a subunit for the seminiferous tubule formation or other developmental events will re- quire further investigation.

This is the first report of the positive staining in the fetal Leydig cells from GD 17. Immunoreac- tivity in purified Leydig cells which prepared from adult rats had been shown by Sharpe ef al. [25] and Risbridger et al. [21]. Sharpe et al. [25] reported that Leydig cells in adult rats made little contribu- tion to the intratesticular and blood levels of inhibin. However, inhibin modulates LH-induced androgen biosynthesis by testicular cells [9]. Leydig cells have synthetic ability of steroid hor- mones during the late gestation period. Therefore, inhibin expression in the fetal Leydig cells after GD 17 indicates that inhibin might modulate to the steroidogenesis in the fetal Leydig cells. Further study is needed to elucidate the role of inhibin on steroidogenesis and/or other physiological roles in

the fetal Leydig cells.

Postnatal expression of inhibin in the Sertoli and granulosa cells was consistent with many previous works [see reviews and 18]. Inhibin contributes to negative feedback loop of FSH secretion from the pituitary in the immature rats. And also inhibin is concerned with the paracrine roles for testicular and ovarian functions: steroidogenesis in the Leydig and thecal cells, increase in the number of follicles, and germ cell-Sertoli cell interaction in immature and mature rats [see reviews and 6, 26].

Recently, Matzuk et al. [15] reported that in- hibin-a was a tumor-suppressor gene of the gonad- al stromal tumors by using inhibin-a-deficient mice. Mice homozygous for the deleted a-subunit gene were susceptible to the development of diffe- rentiated gonadal stromal tumors as early as 4 weeks of age, but were normal for the develop- ment of the gonads and external genitalia before tumor development by histological examination. Therefore, it is suggested that other factors may be participated in the sexual differentiation of the gonads.

ACKNOWLEDGMENTS

We are very grateful to Prof. S. Sasamoto, Tokyo University of Agriculture and Technology, for supplying us with a polyclonal antibody against porcine inhibin a-chain.

REFERENCES

1 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

2 Bardin CW, Morris PL, Chen C-L, Shaha C, Vogl- mayr J, Rivier J, Vale, WW (1987) Testicular inhibin: structure and regulation by FSH, androgens and EGF. In “Inhibin-Non-Steroidal Regulation of Follicle Stimulating Hormone Secretion” Ed by HG Burger, DM de Kretser, JK Findlay, M Igarashi, Raven Press, New York, Serono symposia Publica- tions, 42: 179-190

3 Campen CA, Vale W (1988) Interaction between purified ovine inhibin and steroids on the release of gonadotropins from culture rat pituitary cells. En- docrinology 123: 1320-1328

4 de Jong FH, Grootenhuis AJ, Klaij IA, van Beur- den WMO (1990) Inhibin and related proteins: Localization, regulation, and effects. In “Circulating Regulatory Factors and Neuroendocrine Function”

10

11

12

13

14

15

454

Ed by JC Porter, D Jezova, Plenum Press, New York, pp 271-293

Esch FS, Shimasaki S, Cooksey K, Mercado M, Mason AJ, Ying S-Y, Ueno N, Ling N (1987) Complementary deoxyribonucleic acid (cDNA) cloning and DNA sequence analysis of rat ovarian inhibins. Mol Endocrinol 1: 388-396

Fauser BCJM, Hsueh AJW (1988) Growth factors: Intragonadal regulation of differentiated functions. In “Progress in Endocrinology 1988 Vol 1” Ed by H Imura, K Shizume, S Yoshida, Elsevier Science Publishers, Amsterdam, pp 451-456

Findlay JK, Clarke IJ, Luck MR, Rodgers RJ, Shukovski L, Robertson DM, Klein R, Murray JF, Scaramuzzi RJ, Bindon BM, O’Sea T, Tsonis CG, Forage RG (1991) Peripheral and intragonadal actions of inhibin-related peptides. J Reprod Fert, Suppl 43: 139-150

Fukuda M, Miyamoto K, Hasegawa Y, Nomura M, Igarashi M, Kanagawa K, Matsuo H (1986) Isola- tion of bovine follicular fluid inhibin of about 32 kDa. Mol Cell Endocrinol, 44: 55-60

Hsueh AJW, Dahl KD, Vaughan J, Tucker E, Rivier J, Bardin CW, Vale W (1987) Heterodimers and homodimers of inhibin subunits have different paracrine action in the modulation of luteinizing hormone-stimulated androgen biosynthesis. Proc Natl Acad Sci USA 84: 5082-5086

Knight PG (1991) Identification and purification of inhibin and inhibin-related proteins. J Reprod Fert, Suppl 43: 111-123

Magre S, Jost A (1980) The initial phases of testicular organogenesis in the rat. Arch Anat Mic- rosc Morphol Exp 69: 297-318

Magre S, Jost A (1991) Sertoli cells and testicular differentiation in the rat fetus. J Elect Microsc Tech 19: 172-188

Mather JP, Attie KM, Woodruff TK, Rice GC, Phillips DM (1990) Activin stimulates spermatogo- nial proliferation in germ-Sertoli cell cocultures from immature rat testis. Endocrinology 127: 3206- 3214

Mather JP, Woodruff TK, Krummen LA (1992) Paracrine regulation of reproductive function by inhibin and activin. Proc Soc Exp Biol Med 201: 1- 15

Matzuk MM, Finegold MJ, Su J-GJ, Hsueh AJW, Bradley A (1992) a-inhibin is a tumour-suppressor gene with gonadal specificity in mice. Nature 360: 313-319

McCullagh GR (1932) Dual endocrine activity of the testes. Science 76: 19-20

Merchant-Larios H, Villalpando I, Taketo-Hosotani T (1991) Gonadal morphogenesis in mammals. In “Reproduction, Growth and Development” Ed by

18

19

20

21

22

23

24

25

26

27

28

29

30

S. KoIkE AND T. NOUMURA

A Negro-Vilar, G Perez-Palacios, Raven Press, New York, Serono Symposia Publications 71: 1-11 Merchenthaler, I, Culler MD, Petrusz P, Negro- Vilar A (1987) Immunocytochemical localization of inhibin in rat and human reproductive tissues. Mol Cell Endocrinol 54: 239-243

Miyamoto K, Hasegawa Y, Fukuda M, Nomura M, Igarashi M, Kanagawa K, Matsuo H (1985) Isola- tion of porcine follicular fluid inhibin of 32 k deltons. Biochem Biophys Res Commun 129: 396-403 Ogawa K, Kurohmaru M, Shiota K, Takahashi M, Nishida T, Hayashi Y (1991) Histochemical loca- lization of inhibin and activin a, 8A and $B subunits in rat gonads. J Vet Med Sci 53: 207-212 Risbridger GP, Clements J, Robertoson DM, Drummond AE, Muir J, Burger HG, de Kretser DM (1989) Immuno- and bioactive inhibin and inhibin a-subunit expression in rat Leydig cell cul- ture. Mol Cell Endocrinol 66: 243-247

Roberts V, Vale W (1991) Expression of inhibin/ activin subunit messenger ribonucleic acids during rat embryogenesis. Endocrinology 128: 3122-3129 Robertson DM, Foulds LM, Leversha L, Morgan FJ, Hearn MT, Burger HG, Wettenhall RE, de Kretser DM (1985) Isolation of inhibin from bovine follicular fluid. Biochem Biophys Res Commun 126: 220-226

Schwartz NB, Channing CP (1977) Evidence of ovarian “inhibin”: suppression of the secondary use in serum follicle stimulating hormone levels in proestrous rats by injection of porcine follicular fluid. Proc Natl Acad Sci USA 71: 5721-5724 Sharpe RM, Kerr JB, Maddocks S (1988) Evidence for a role of Leydig cells in the control of the intratesticular secretion of inhibin. Mol Cell Endoc- rinol 60: 243-247

Skinner MK (1991) Cell-cell interactions in the testis. Endocr Rev 12: 45-77

Vale W, Rivier C, Hsueh A, Campen C, Meunier H, Bicsak T, Vaughan J, Corrigan A, Bardin W, Sawchenko P, Petraglia F, Yu J, Plotsky P, Spiess J, River J (1988) Chemical and biological characteriza- tion of the inhibin family of protein hormones. Rec Prog Horm Res 44: 1-38

Vale W, Hsueh A, Rivier C, Yu J (1991) The inhibin/activin family of hormones and growth fac- tors. In “Peptide Growth Factors and Their Recep- tors II” Ed by MB Sporn, AB Roberts, Springer- Verlag, New York, pp 211-248

Ying S-Y (1987) Inhibins and activins: Chemical properties and biological activity. Proc Soc Exp Biol Med 186: 253-264

Ying S-Y (1988) Inhibins, activins, and follistatins: Gonadal proteins modulating the secretion of folli- cle-stimulating hormone. Endocr Rev 9: 267-293

ZOOLOGICAL SCIENCE 10: 455-466 (1993)

The Development of the Hermaphroditic Gonad in Four Species of Dicyemid Mesozoans

HIDETAKA FURUYA, KAZUHIKO TSUNEKI

and YUTAKA KOSHIDA

Department of Biology, College of General Education, Osaka University, Toyonaka 560, Japan

ABSTRACT—The development of the functionally hermaphroditic gonad, the infusorigen, in four dicyemid species, Dicyema orientale, D. acuticephalum, D. japonicum, and D. misakiense, was studied in fixed materials with the aid of a light microscope. After an agamete (axoblast) undergoes the first division and excludes a paranucleus, the resulting cell undergoes the second division. Afterwards, three different types of cell lineage can be identified. (1) In D. orientale, the first oogonium is produced by the second division, and the axial cell of an infusorigen and the first spermatogonium are produced by the third division. (2) In D. acuticephalum, the first oogonium is produced by the second division, the axial cell is produced by the third division, and the spermatogonium is produced by the fourth division. The fourth division also produces the first oogonium of another egg line. (3) In D. japonicum and D. misakiense, the first spermatogonium is produced by the second division, and the axial cell and the first oogonium are produced by the third division. In all species examined, oogonia occupy the outer surface of the axial cell and spermatogonia are incorporated into the axial cell. In this way, the spermatogenesis proceeds within the cytoplasm of the axial cell. Mature infusorigens of these four species consist of about twenty cells. The respective numbers of oocytes and spermatozoa produced in each infusorigen

© 1993 Zoological Society of Japan

are roughly equal in these four species.

INTRODUCTION

Dicyemid mesozoans are found in the renal sac of benthic cephalopod molluscs. The bodies of dicyemids consist of only 20 to 40 cells and they are organized very simply [12, 14]. It has long been debated whether dicyemids are truly primitive multicellular animals [2, 7, 8, 11, 17], or whether they are actually organisms that have degenerated as a result of parasitism [5, 12, 14, 19].

As well known, two kinds of adult forms, nema- togens and rhombogens, are found in dicyemids. Asexual reproduction occurs within the axial cell of a nematogen, while sexual reproduction takes place within the axial cell of a rhombogen. The features of the sexual reproduction are unique [12, 15]. A hermaphroditic gonad, which is called an infusorigen, is formed within the axial cell of a rhombogen, and fertilization occurs around the

Accepted April 13, 1993 Received March 13, 1993

infusorigen. The zygote undergoes cleavages and develops into an infusoriform larva within the axial cell. The process of the fertilization and embryogenesis of infusoriforms have been de- scribed in detail [1, 3, 18], but the development of infusorigens has been studied only sporadically [10, 12, 14]. No research on the patterns of development of infusorigens has been performed from a systematic perspective. We examined the development of the infusorigens of Dicyema orien- tale, D. acuticephalum, D. japonicum, and D. misakiense, and found three different types of cell lineage. In this report, these three different cell lineages that can be followed during the develop- ment of infusorigens are described. In addition, we provide an estimate of the numbers of gametes that are produced and of the numbers of embryos that are generated in one infusorigen, and we discuss the reproductive capacity of infusorigens.

456 H. Furuya, K. TsuNEKI AND Y. KosHIDA

MATERIALS AND METHODS

Seventeen individual octopuses, Octopus vulga- ris, and three cuttlefish, Sepioteuthis lessoniana, were purchased or collected by the authors in the waters off the western coast of Japan. Dicyema orientale from Sepioteuthis lessoniana and three species of dicyemids, namely, D. acuticephalum, D. japonicum, and D. misakiense, from Octopus vulgaris were examined in the present study.

After the host cephalopods had been sacrificed, their renal sacs were taken out and smeared direct- ly on glass slides. Smeared dicyemids were im- mediately fixed with Carnoy’s fixative or with alcoholic Bouin’s solution (a mixture of absolute ethanol saturated with picric acid, formalin and acetic acid, 15:5:1, v/v). Specimens fixed with Carnoy’s fixative were stained with Feulgen’s stain

3 se

Fic. 1.

Photographs were taken at a magnification of 2000 diameters under an oil-immersion objective.

or by the PAS method and were poststained with Ehrlich’s hematoxylin and light green. Specimens fixed with alcoholic Bouin’s solution were stained with Ehrlich’s hematoxylin and light green only. The development of infusorigens in the axial cells of rhombogens was observed with the aid of a light microscope under an oil-immersion objective at a final magnification of 2000 diameters.

RESULTS

Dicyema orientale (Figs. 1, 2, 7, and Table 1)

At the stage of the transition from the nema- togen to the rhombogen, a number of agametes (axoblasts) degenerate. The remaining agametes, that number about 10 to 20, grow larger and

so IN (©)

ome +h ‘

Light micrograghs of developing infusorigens within the axial cells of rhombogens of D. orientale.

Scale bar

represents 10 um. (a): An agamete undergoing an unequal first division. (b): A progenitor cell of an infusorigen (PI) and a paranucleus (P) being produced after the first division. (c)-(i): Infusorigens. (c): The arrow indicates the zygotene stage of a primary spermatocyte. (d): The arrow indicates the zygotene stage of a primary oocyte. (e): The arrow indicates the metaphase of a primary spermatocyte. (f): The short arrow indicates the metaphase of a secondary spermatocyte viewed from the pole and the long arrow indicates the pachytene stage of a primary oocyte. The nucleolus is conspicuosly large in the primary oocyte at the pachytene stage. (g): The arrow indicates the metaphase of a secondary spermatocyte viewed from the side. (h): The arrow indicates the diplotene stage of

a primary oocyte. indicates the anaphase of a primary oocyte.

(i): The short arrows indicate the bouquet stage of a primary oocyte and the long arrow

NI, axial cell nucleus of infusorigen; O, oogonium; PO, primary oocyte; PS, primary spermatocyte; S,

spermatogonium; SP, spermatozoon.

Gonadal Development in Dicyemids 457

undergo an unequal division (Fig. la). Each small- er daughter cell loses its cytoplasm and becomes just a nucleus, known as a paranucleus, and it lies near the cell from which it arose (Fig. 1b). The

paranucleus stays within the axial cell of the rhom- bogen and grows to the same size as the axial cell nucleus of the rhombogen. The larger daughter cell, namely, the progenitor of the infusorigen,

Fic. 2. Sketches of the development of the infusorigen of D. orientale. Bar represents 5 ym. (a): A progenitor cell of infusorigen. (b) and (c): Two-cell stage. In (c), the metaphase of the third division is seen. (d)-(f): Three-cell stage. In (d) and (e), both the axial cell (AI) and the first spermatogonium (S) produced after the third division are shown. In (f), the first spermatogonium (S)is embedded in the axial cell (AI). (g): Four-cell stage. (h): Five-cell stage. The metaphase of the first spermatogonium (S)is seen. (i): Six-cell stage in optical section. (j): Eight-cell stage in optical section. Primary oocyte (PO) at the upper right corner is in the zygotene stage. (k): Nine-cell stage in optical section. The anaphase of the primary spermatocyte (PS) is seen. (1): The infusorigen. A primary oocyte (PO) at the beginning of anaphase is seen in the lower right corner. This oocyte contains a spermatozoon (SP) in the cytoplasm. Spermatozoa (SP) in the center are emerging from the axial cell of the

infusorigen.

AI, axial cell of infusorigen; O, oogonium; PO, primary oocyte; PS, primary spermatocyte; S, spermatogonium;

SP, spermatozoon.

458 H. Furuya, K. TSUNEKI AND Y. KosHIDA

undergoes a nearly equal division to the two-cell stage (Figs. 2b and c). One of these cells becomes the first oogonium. The other cell undergoes an equal division and produces the first spermatogo- nium and an axial cell of the infusorigen (Figs. 2c-e). The axial cell of the infusorigen does not divide further, but it increases in size and incorpo- rates the spermatogonium into its cytoplasm (Fig. 2f). The first oogonium remaining on the periphery of the infusorigen divides equally to generate a second oogonium and a primary oocyte (Fig. 2g). In the same way, the second oogonium produces a third oogonium and a primary oocyte. Very early primary oocytes can be distinguished from oogonia since the nucleolus in the former is larger than that in the latter (Figs. 2g and h). In primary oocytes, at the prophase of the first meio- tic division, the chromatin becomes aggregated on one side of the nucleus and the nuclear membrane becomes indistinct on the side opposite the aggregation of chromatin (Figs. 1i and 21). These features are characteristic of the so-called “bou- quet stage” of the prophase of the meiotic division. The primary oocytes gradually become larger and chromosomes become visible as thick threads in the nuclei, features that characterize the zygotene stage. At this stage, both ends of the chromo- somes are attached to the nuclear envelope (Figs. 1d and 2j). During the pachytene stage, the chromosomes become indistinct and the nucleus becomes similar in appearance to the interphase nucleus. At this stage, the nucleus includes a very large nucleolus (Figs. 1f and 2k-l1). When primary oocytes have grown to about 7 ~m in diameter, bead-like chromosomes appear in the nucleus. These oocytes are at the diplotene stage of meiotic prophase (Fig. 1h). The primary oocytes finally reach about 12 ~m in diameter (Fig. 11).

The first spermatogonium within the axial cell of the infusorigen undergoes an equal division and produces a spermatogonium and a primary sper- matocyte (Figs. 2h-j). The prophase of the first meiotic division of the primary spermatocyte pro- ceeds similarly to that of the primary oocyte, but the size of the primary spermatocyte does not change throughout the prophase (Figs. 1c, 1d, 2), and 2k). After the first meiotic division, a pair of secondary spermatocytes enters interkinesis. At

this stage, no chromosome structures can be seen. Within the axial cell of maturing infusorigen, usually two secondary spermatocytes are observed in addition to a spermatogonium and an axial cell nucleus. Soon after the second meiotic division, transformation of spermatids into spermatozoa occurs. Mature spermatozoa are composed of a small amount of deeply stained chromatin and a surrounding small clear area, interpreted as cyto- plasm. The cell membrane is hardly visible. The entire spermatozoon is about 2 ~m in diameter. The chromatin is usually horseshoe-shaped, but sometimes it is irregularly ring- or dot-shaped (Figs. 1h and 21). After emerging from the axial cell of an infusorigen, the spermatozoon enters the primary oocyte (Fig. 21). Fertilized oocytes remain adhering to the axial cell up to the time at which the first polar bodies are produced. Spermatozoa often adhere to the outer surface of the axial cell, to the oogonia, or to the primary oocytes, or they may appear between the oocyte and the axial cell (Fig. 21). The spermatozoon within the oocyte lies at the periphery of the metaphase plate of the first meiotic division of the oocyte (Fig. 21). The first polar bodies are composed of a mass of chromatin and a clear cytoplasmic area that is surrounded by a delicate membrane. They become detatched from the oocytes and often remain intact, but finally they degenerate.

The numbers of spermatogonia and primary spermatocytes, the number of spermatozoa within and on the surface of the infusorigen, and the numbers of oogonia and primary oocytes per in- fusorigen are shown in Table 1. D. orientale, being relatively long, has a large number of infusorigens and infusoriform embryos in the axial cell of a rhombogen (Table 1).

Dicyema acuticephalum (Figs. 3, 4, 7, and Table 1)

One or rarely two agametes become larger and undergo an unequal division at the beginning of the rhombogen stage (Figs. 3a and 4a). The larger daughter cell is the progenitor of an infusorigen, while the smaller cell becomes a paranucleus (Fig. 3b). The progenitor cell of the infusorigen under- goes an equal division, which results in the two-cell

Gonadal Development in Dicyemids 459

TaBLE 1. The numbers of infusorigens, gametes, and embryos within the axial cells of rhombogens

Body length of No. of No. of spermatogonia and Species rhombogens infusorigens primary spommatocytes (mm) per rhombogen per infusorigen’ Dicyema orientale” ~3.5 7-25 3.27+0.71 D. acuticephalum ~0.8 1- 2 3.38 £0.86 D. japonicum ~1.0 1- 2 4.06+1.28 D. misakiense ~1.0 1-2 4.04+1.24 No. of spermatozoa No. of oogonia and No. of infusorform Species per infusorigen” primary oocytes embryos per per infusorigen”? rhombogen D. orientale” 11.78+5.21 10.26+2.30 ~250 D. acuticephalum 12.67+4.70 9.25+1.43 ~ 20 D. japonicum 10.00+3.91 13.43 +3.47 ~ 35 D. misakiense 11.59+4.70 12.04+3.76 ~ 35

) Values represent means+S.D. and are based on results from 50 mature infusorigens. 2) D. orientale was described only with nematogens [16] and no rhombogens have been reported.

Fic. 3. Light micrographs of developing infusorigens within the axial cells of rhombogens of D. acuticephalum. Bar represents 10 um. (a): Agametes (A). A telophase figure of the first unequal division is seen in the center. The arrow indicates a smaller daughter cell that becomes a paranucleus. (b): A progenitor cell of an infusorigen (PI) and a paranucleus (P) within an axial cell (AR) of a rhombogen. (c): Five-cell stage. (d)-(g): Infusorigens. (d): The arrow indicates the metaphase of a primary spermatocyte viewed from the side. (e): The arrows indicate the interkinesis stage of the secondary spermatocytes. (f): The metaphase of a primary oocyte viewed from the side. The arrow indicates a sperm within the oocyte. (g): The large arrow indicates the anaphase of a secondary oocyte. The short arrow indicates a sperm within it. The primary oocyte (PO) is at the pachytene phase.

Al, axial cell of infusorigen; NI, axial cell nucleus of infusorigen; O, oogonium; PO, primary oocyte; PS, primary spermatocyte; S, spermatogonium; SP, spermatozoon.

460 H. Furuya, K. TSUNEKI AND Y. KOSHIDA

PO M

Fic.

oS ae) M @i = On: a b Cc d és

4. Sketches of the development of the infusorigen of D. acuticephalum. Bar represents 5 um. (a): The telophase of the first division of the agamete. The smaller cell becomes a paranucleus. (b): The progenitor cell of an infusorigen. (c): Two-cell stage. (d)-(e): Three-cell stage. In (e), the telophase of an oogonium (QO) is seen. The cell marked M is a mother cell of an oogonium and a spermatogonium. (f): Four-cell stage. (g): Five-cell stage. Two cells on the right side of the axial cell, produced by the division of the mother cell (M in e and f), become the first spermatogonium and the oogonium of one egg line, respectively. (h)-(i): Six-cell stage. In (i), the first spermatogonium (S) is embedded in the axial cell (AI) and the primary oocytes (PO) are at the bouquet stage. (j): Eight-cell stage in optical section. (k): An infusorigen in optical section. A secondary spermatocyte (SS) is seen in the axial cell (AI). (1): An infusorigen (surface view). The anaphase of a primary oocyte (PO)is seen in the upper right corner.

Al, axial cell of infusorigen; M, mother cell of the first spermatogonium and the oogonium; O, oogonium, PO, primary oocyte; PS, primary spermatocyte; S, spermatogonium; SP, spermatozoon; SS, secondary spermatocyte.

Gonadal Development in Dicyemids 461

stage (Fig. 4c). One of these cells becomes the first oogonium, and the other cell increases in size and divides unequally. The larger cell, the axial cell of the infusorigen, undergoes no further divisions (Fig. 4d), while the smaller cell divides equally and produces both the first spermatogonium and an oogonium (Fig. 4g). Both cells are so similar in size and appearance that they cannot be distin- guished until one of them enters the mitotic phase on the periphery of the infusorigen or is embedded in the axial cell (Fig. 41). D. acuticephalum has two egg lines (Fig. 7B). The oogonium of each egg line further divides and generates an oogonium and a primary oocyte (Figs. 4e-i). At around the six-cell stage, the first spermatogonium Is incorpo- rated into the axial cell of the infusorigen (Fig. 41). The process of spermatogenesis is similar to that observed in D. orientale. The chromatin of sper- matozoa usually forms a horseshoe or an irregular ring (Fig. 41). The numbers of spermatogonia and primary spermatocytes, and other numerical data, are shown in Table 1.

Fic. 5.

Dicyema japonicum and Dicyema misakiense (Figs. 5=7 and Table 1)

The developmental pattern of the infusorigen is the same in D. japonicum and D. misakiense. One or rarely two of the agametes become larger and undergo an unequal division at the beginning of the rhombogen stage. The smaller cell is trans- formed into a paranucleus, while the larger cell becomes the progenitor of an infusorigen (Fig. 5a). This latter cell divides unequally and pro- duces two cells (Figs. 5b and 6a). The smaller cell often undergoes an equal division (Fig. 6b), but one of the daughter cells soon degenerates. The remaining cell is the first spermatogonium (Fig. 6c). The larger cell divides equally and, thus, produces the first oogonium and an axial cell of the infusorigen (Figs. Sc and 6d). The axial cell under- goes no further divisions and increases in size. The first spermatogonium is incorporated into the cyto- plasm of the axial cell (Fig. 6e) and its nucleus increases in size (Fig. 6f). The oogonium gives rise

| PO.

Light micrographs of developing infusorigens within the axial cells of rhombogens of D. japonicum. Bar

represents 1() ~m. (a): A progenitor cell of an infusorigen (PI), the axial cell of a rhombogen (AR), a peripheral cell nucleus (NP), an axial cell nucleus of a rhombogen (NR), a paranucleus (P), and peripheral cell cytoplasm (PC) can be seen. (b): Two-cell stage. (c): Three-cell stage. (d)-(f): Infusorigens. (d): The arrow indicates the

metaphase of the primary spermatocyte.

(e): The arrows indicate the interkinesis stage of a secondary

spermatocyte. (f): The arrow indicates the metaphase of a secondary spermatocyte viewed from the pole. (g): The telophase of a primary oocyte. The long arrow indicates the first polar body. The short arrow indicates the

sperm within the primary oocyte.

Al, axial cell of infusorigen; NI, axial cell nucleus of infusorigen; NR, axial cell nucleus of rhombogen; O, oogonium; P, paranucleus; PO, promary oocyte; S, spermatogonium; SP, spermatozoon.

462 H. Furuya, K. TSuUNEKI AND Y. KOSHIDA

Fic. 6. Sketches of the development of infusorigens of D. misakiense. Bar represents 5 um. (a)-(c): Two-cell stage. As seen in (b), the first spermatogonium (S) often divides quite early. In (c), the mother cell of both the first oogonium and the axial cell is in metaphase. (d)-(e): Three-cell stage. In (e), a spermatogonium (S) is embedded in the axial cell (AI). (f): Four-cell stage. (g): Five-cell stage. The metaphase of a spermatogonium (S) is seen. (h): Eight-cell stage in optical section. The primary oocytes (PO) at the zygotene stage are seen. (i): The infusorigen in optical section. The primary oocytes (PO at the left and right) at the pachytene stage and another oocyte (PO at the upper left corner) at the zygotene stage are seen. The metaphases of the secondary spermatocytes (SS) are also seen from the side (center) and from the pole (right). (j): The infusorigen (surface view). Some spermatozoa (SP) are emerging from the axial cell.

Al, axial cell of infusorigen; O,oogonium; PO, primary oocyte; PS, primary spermatocyte; S, spermatogonium; SP, spermatozoon; SS, secondary spermatocyte.

Gonadal Development in Dicyemids 463

Paranucleus

A © -@ polar bodies

© agamate oogonia @ wo primary eee ei ake secondary oocyte progenitor cell @ = axial cell spermatids

of infusorigen O} Hite: (Ola © *% Om Sie She @) be spermatozoa ©- Ome ©—o

secondary spermatogonia spermatocytes

Paranucleus

B @-e

agamate oo

o) nS. oa Ea vouene®| (O—( @ ) mca

of infusorigen @) © __—— oogonia ole-

Ca — ®—e © mA © —= (0) © a ee

© —o

spermatogonia

Paranucleus Cc © e oogonia

eens i 5 le ex

@| ©

. @—(e) wie

progenitor cell © —= ©) of infusorigen © x @ © aan © @) oe @ spermatozoa

spermatogonia Fic. 7. The cell lineage of an infusorigen. (A): D. orientale. (B): D. acuticephalum. (C): D. japonicum and D. misakiense. The circle represents the cell and the central black area represents the nucleus. They are nearly scaled. The dotted circle means that the cytoplasm degenerates. At the site of the asterisk (+), the first spermatogonium is embedded in the axial cell. Short horizontal lines on right side of cells indicate that the cells proliferate further. In (c), the cross (x) means that the cell does not proliferate but degenerates.

464 H. Furuya, K. TsuNEKI AND Y. KosHIDA

to a new oogonium and a primary oocyte. Subse- quent oogenesis and spermatogenesis proceed in the same manner as in D. orientale and D. acu- ticephalum. he nucleus of the spermatozoa is horseshoe- or dot-shaped, and the spermatozoa often aggregate (Figs. 5e, Sf, and 6j). The number of spermatozoa and other numerical data are shown in Table 1.

DISCUSSION

In the four species of dicyemid mesozoans stu- died herein, an agamete enlarged at the beginning of the rhombogen stage. In other species of dicyemids, Nouvel [14] and McConnaughey [12] also reported that an agamete enlarges at the very early stage of the rhombogen. This enlargement is interpreted as a sign of the beginning of the development of infusorigens. Then the agamete undergoes an unequal division and the smaller daughter cell becomes a paranucleus without con- tributing to the formation of an infusorigen. Although we can offer no explanation for the formation of a paranucleus, it is a constant feature and may, thus, be essential to the development of infusorigens.

Several differences were apparent in the cell lineage of the infusorigens of the four species (Fig. 7). Nevertheless, two common features were apparent; one is that the first spermatogonium is incorporated into an axial cell, in which sperma- togenesis proceeds; and the other is that the oogo- nium remains at the periphery of the axial cell, where oogenesis occurs. The distinctive differ- ences occur early in the development in the various species. Two patterns of cell lineage are distin- guishable. One pattern is characterized by a second division that produces the first oogonium, and the other is characterized by a second division that produces the first spermatogonium. The for- mer is seen in D. orientale and D. acuticephalum, and the latter in D. japonicum and D. misakiense (Fig. 7). misakiense, all types of cell differentiate up to the

In D. orientale, D. japonicum, and D.

third division. In D. acuticephalum, by contrast, spermatogonia are generated only after the fourth division (Fig. 7B). In spermatogenesis, all four species examined have only one sperm line. In

oogenesis, D. acuticephalum has two egg lines, while the other three species have only one (Fig. 7). However, D. acuticephalum has a rather small number of oogonium and oocytes (Table 1). The body size may be a factor that limits the number of oocytes.

The development of infusorigens was reported by Lameere [10], Nouvel [14], and McConnaughey [12]. Lameere studied D. typus, while McCon- naughey did not specify the species that he studied. However, both these authors observed that the first oogonium is produced by the second division and the first spermatogonium and axial cell are generated after the third division. This previously reported cell lineage is, thus, the same as that observed in D. orientale. Nouvel [14] traced the development of the infusorigen of D. schulzianum. His findings are identical to ours in D. japonicum and D. misakiense in that the first spermatogonium is produced after the second division and the oogonium and axial cell are produced after the third division. However, the first spermatogonium of D. schulzianum is relatively large compared to that of D. japonicum and D. misakiense. In D. orientale, D. japonicum, and D. misakiense, the first spermatogonia, formed by unequal divisions, are relatively small before they are incorporated into an axial cell. They increase in size until they are as large as oogonia after having been incorpo- rated into an axial cell. In D. acuticephalum, the first spermatogonium is also a small-sized cell. The third division in this species proceeds unequally and generates a larger axial cell and a smaller cell. This smaller cell is the mother cell of both the first spermatogonium and the oogonium of one egg line. Thus, the differentiation of the first sperma- togonium occurs later in D. acuticephalum than in the other species (Fig. 7). The pattern of cell lineage observed in D. acuticephalum has not been reported in the earlier literature. Although pre- vious reports dealing with a few species [10, 12, 14] did not pay any attention to species-specific differ- ences in the cell lineages of infusorigens, a distinct difference does exist and could be a criterion for classification of dicyemid species.

In D. japonicum and D. misakiense, the pre- sumptive first spermatogonium often undergoes equal division before being incorporated into the

Gonadal Development in Dicyemids

axial cell. However, one of the daughter cells degenerates soon without differentiating into an oogonium, in contrast to the case in D. acuticepha- lum. This paticular division might be an “extra” cell division because it does not occur consistently in all infusorigens examined.

In D. aegira, which is about 1.5 mm in total length and has one or two infusorigens, two to three times as many spermatozoa are produced as oocytes [1, 13]. Among the species studied herein, D. orientale is a relatively large dicyemid, reaching 3.5 mm in length, and it has 7 to 25 infusorigens. D. acuticephalum, which is small, and D. japoni- cum and D. misakiense, which are medium-sized, have usually one, or sometimes two infusorigens [4, 16]. In the four species that we examined, the numbers of spermatozoa and the numbers of the oogonia and primary oocytes were roughly equal (Table 1). The numbers of oogonia and primary oocytes were 2.7 to 3.3 times those of spermatogo- nia and primary spermatocytes. If spermatogene- sis and oogenesis proceed at the same rate during the germ cell division and maturation, the number of spermatozoa can be estimated to be 13 to 16 (3.27 to 4.064). These values are 1.2 to 1.5 times the values for oocytes. This discrepancy may be attributed to a possibly lower rate of sperma- togenesis than of oogenesis [1]. The apparently small number of spermatozoa might be due to the limited space within the axial cell of the infusor- igen. A large number of spermatozoa may not be necessary in dicyemids, which perform self- fertilization within the axial cells of rhombogens. However, it is still unclear why the present four species have much smaller numbers of spermato- zoa than D. aegira. It is apparent that the size of rhombogens and the number of infusorigens per rhombogen do not affect the number of gametes produced.

In the nematode Caenorhabditis elegans, the cell lineage of the gonad has been studied in detail [6, 9]. Dicyemids have very simple gonads that are composed of a very small number of cells, and thus, they may also prove to useful as model systems for studies of the differentiation of ga- metes.

12

13

14

15

465

REFERENCES

Austin CR (1964) Gametogenesis and fertilization in the Mesozoan Dicyema aegira. Parasitology 54: 597-600

Eernisse DJ, Albert JS, Anderson FE (1992) Anne- lida and Arthropoda are not sister taxa: a phy- logenetic analysis of spiralian metazoan morpholo- gy. Syst Biol 41: 305-330

Furuya H, Tsuneki K, Koshida Y (1992) Develop- ment of the infusoriform embryo of Dicyema japoni- cum (Mesozoa: Dicyemidae). Biol Bull 183: 248- Di

Furuya H, Tsuneki K, Koshida Y (1992) Two new species of the genus D. (Mesozoa) from octopuses of Japan with notes on D. misakiense and Dicyema acuticephalum. Zool Sci 9: 423-437

Ginetsinskaya TA (1988) Trematodes, Their Life Cycles, Biology and Evolution. Amerind Publishing Co. Pvt. Ltd., New Delhi. (Translation of the origin- al Russian edition, 1968)

Hirsh D, Oppenheim D, Klass M (1976) Develop- ment of the reproductive system of Caenorhabditis elegans. Dev Biol 49: 200-219

Hyman L (1940) The Invertebrates, Vol I, McGraw Hill, New York

Hyman L (1959) The Invertebrates, McGraw Hill, New York

Kimble J, Hirsh D (1979) The postembryonic cell lineages of the hermaphrodite and male gonads in Caenorhabditis elegans. Dev Biol 70: 396-417 Lameere A (1918) Contributions a la connaissance des dicyémides (2e partie). Bull Biol Fr Belg 51: 347-390

Lapan EA, Morowitz HJ (1974) Characterization of mesozoan dispersal larva. Exp Cell Res 94: 277-282 McConnaughey BH (1951) The life cycle of the dicyemid Mesozoa. Univ Calif Publ Zool 55: 295- 336

McConnaughey BH, Kritzler H (1952) Mesozoan parasites of Octopus vulgaris, Lam. from Florida. J Parasitol 38: 59-64

Nouvel H (1948) Les Dicyémides. 2e partie: in- fusoriforme, tératologie, spécificité du parasitisme, affinités. Arch Biol 59: 147-223

Nouvel H (1947) Les Dicyémides. le partie: systé- matique, générations vermiformes, infusorigene et sexualité. Arch Biol 58: 59-220

Nouvel H, Nakao Y (1938) Dicyémides du Japon. Bull Soc Zool Fr 63: 72-80

Ohama T, Kumazaki T, Hori H, Osawa S (1984) Evolution of multicellular animals as deduced from 5S ribosomal RNA sequences: a possible early em- ergence of the Mesozoa. Nucleic Acids Res 12: 5101-5108

Vol V,

466 H. Furuya, K. TUNEKI AND Y. KOSHIDA

18 Short RB, Damian RT (1967) Oogenesis, fertiliza- 19 Stunkard HW (1954) The life history and systematic tion, first cleavage of Dicyema aegira (Mesozoa: relations of the Mesozoa. Quart Rev Biol 29: 230- Dicyemidae). J Parasitol 53: 185-195 244

ZOOLOGICAL SCIENCE 10: 467-471 (1993)

© 1993 Zoological Society of Japan

Hormonal Deficiency Causing Albinism in Locusta migratoria

SEWI TANAKA

Department of Insect Physiology and Behavior, National Institute of Sericultural and Entomological Science, Tsukuba, 305, Japan

ABSTRACT—Hoppers of an albino strain in the migratory locust, Locusta migratoria, are uniformly creamy white when reared under crowded conditions. Implantation of corpora cardiaca taken from a normal hopper caused some albino hoppers to turn grey, reddish, brown, or dark brown like the colors of normal isolated solitary individuals, and others to develop the black and orange coloration like that of normal gregarious hoppers. Other organs such as brain and thoracic ganglia also induced various colors. Therefore, the albinism of this locust is caused by deficiency of some hormonal factor(s) present in the corpora cardiaca, brain and other ganglia. Implantation of some organs obtained from 7 other species also caused albino locusts to develop dark body colors, suggesting that some hormonal factor(s) commonly present in different insects can promote dark pigmentation in albino locusts.

INTRODUCTION

Genetic albinism appears to be caused by lack of the enzyme tyrosinase, which is required to pro- duce the black pigment, melanin [6]. In the desert locust, Schistocerca gregaria, both enzyme and substrate for melanin formation are present, but melanization does not occur in albino individuals [10]. In the migratory locust, Locusta migratoria, the albino mutation is not uncommon and labora- tory lines have been established to study the genetic and behavioral aspects of the albinism [5, 14, 21]. However, no information is available on the underlying mechanism causing this phe- nomenon. I present here evidence suggesting that albinism in L. migratoria is caused by deficiency of some hormonal factor(s) normally present in the brain, corpora cardiaca and thoracic ganglia of this species, and that a similar factor(s) inducing the dark coloration in albino locusts also exists in other insects including a katydid, crickets and moths.

MATERIALS AND METHODS

A colony of albino locusts of L. migratoria was maintained under crowded conditions at a 14h (14L : 10D) photoperiod and 30°C since its estab-

Accepted March 10, 1993 Received December 29, 1992

lishment from albino individuals which appeared in a stock culture originating from Okinawa, Japan. The details of the rearing and handling methods have already been described [4, 20]. Various endocrine and nervous organs were re- moved from 1-day-old 5th instar hoppers, and transplanted with a small quantity of saline solu- tion (3 wl; 0.9% NaCl) into 1-day-old 4th albino instars. A small incision was made between the 2nd and 3rd abdominal sternites of the recipients, and implants were injected into the body cavity using a micropipette. All hoppers were chilled on ice for 10-20 min before the operation and almost no mortality was observed in the recipients which were continusouly kept as groups. Four or 5 days after operation, they ecdysed to the Sth instar, and the body color of each individual was recorded 10 days after the operation. The same procedure was followed for transplantation of organs from other insects into albino locusts, and whether the reci- pients turned darker (+) or not (—) was recorded 10 days after operation. The donor species used included Gryllus bimaculatus, Teleogryllus occipi- talis, Modicogryllus confirmatus, Metriptera hime, Leucania separata, Bombyx mori, and Cephanodes hylas.

RESULTS AND DISCUSSION

When various endocrine organs and ganglia

468 S. TANAKA

were taken from normal crowded hoppers, and transplanted into 4th-instar albino hoppers reared under crowded conditions, all implants except for suboesophageal ganglia (SG) promoted pigmenta- tion in the albino hoppers (Table 1). However, the body color obtained was dependent upon the kind of organ implanted. Albinos implanted with a pair of corpora allata (CA) turned green (Fig. 1E). They looked like normal solitary hoppers except that the ventral side of their body remained whit- ish. This is consistent with the previous findings that implantation of extra CA or administration of juvenile hormone (JH) or JH analogues, to

crowded hoppers induce the green solitary color [1, 2, 7, 8, 9, 15, 18, 19]. Implanted brains and thoracic ganglia induced various shades of grey, dark brown and straw yellow, the coloration being associated with normal non-green (=“homochro- me”) solitary hoppers [16]. A similar color change was observed when albinos were implanted with CC, but in this case some developed the black patterns as well as the orange background color and became indistinguishable from the normal gregarious hoppers (Fig. 1 B, H). The black component of this gregarious coloration is attri- buted mostly to cuticular melanins [8, 16]. After

Fic. 1. Photographs of Sth instar hoppers of L. migratoria; normal solitary (A), normal crowded (gregarious)(B), albino isolated (C), albino crowded hopper without implantation (D) and albino crowded hoppers implanted with normal corpora allata (E), brains (F), or corpora cardiaca (G and H).

Albinism in Locusta

ecdysis, the melanins remain in the exuviae, while the pigments associated with the solitary form are mainly ommochromes, bile pigments and caro- tenoids [16] and do not remain in the exuviae. When a similar experiment was carried out with organs derived from field-collected green solitary

TABLE 1. hoppers in L. migratoria

Donors and

469

hoppers, almost the same results were obtained (77=1.37; df=2; P>0.05; Table 1). This indi- cates that the factor promoting the gregarious black/orange coloration exists even in the CC of green solitary hoppers, although it either is not released or, if released, is inactive in the solitary

Effects of implantation of various organs from Sth instar hoppers on the body color of albino

Color obtained in albino recipients Solitary coloration

Greqarious coloration

Organs implanted

White Green

Light Dark

homochromy homochromy Black/orange

Normal (Crowded in lab)

corpora allata (2) 10 brain (1)

corpora cardiaca (2)

thoracic ganglia (3)

suboesophageal ganglia (3) 5

Normal (Field-collected solitary)

corpora cardiaca (2)

Albino (Crowded in lab)

corpora allata (2) 1 9 brains (1-2) 8

corpora cardiaca (4-12) 15

Albino (Crowded in lab) Control 20

10 5

Donors were field-collected as 4th instars and reared individually in an outdoor insectary until used on the day after molting to the 5th instar. Numbers in the table indicate the numbers of individuals with the color indicated. Numbers in parentheses indicate the number of organs implanted. Controls included 10 intact crowded albino hoppers and 10 sham operated ones which were treated in the same way as the other operated hoppers except that no organ was implanted. Light homochromy includes straw yellow and brown, and dark homochromy reddish

brown and dark brown.

TaBL_eE2. Effects of implantation of various organs from different insects on pigmentation in albino hoppers of L. migratoria

Donor species (Stage)

Organs impanted

Brain Brain-SG CC SG Gryllus bimaculatus (adult) + + Teleogryllus occipitalis (adult) + ste Modicogryllus confirmatus (adult) + ar Metriptera hime (larva) + + Leucania separata (larva) + + = Bombyx mori (pupa) + Cephanodes hylas (larva) + ar =

The recipients either turned darker (+) or remained whitish (—) (N=3—10).

470 S. TANAKA

hoppers. By contrast, when CC or brains were taken from albino hoppers and implanted into other albino individuals, no color change occurred in the latter, although CA taken from albinos induced the green color (Table 1). From these results, it is concluded that genetic albinism in L. migratoria is due to a deficiency of some hormonal factors normally present in the CC, brain and other nervous ganglia.

The brown or black coloration could also be induced in albino hoppers by implantation of CC, brain or other nervous organs from other insects (Table 2). Melanization in the common army- worm, L. separata, is known to be induced by a hormone (the melanization and reddish coloration hormone) secreted from the brain-CC-CA com- plex or the SG [11, 17]. Albino hoppers developed brown pigments when implanted with a brain-CC complex from the armyworm, but no pigmentation was elicited by implantation of even 6 SG. In the cricket, G. bimaculatus, melanization is suppres- sed by the juvenile hormone produced by the CA, but the factor causing melanization has not been identified [2]. Implantation of CA from three species of crickets into albino locusts turned the recipients green (data not shown) while that of CC induced strong melanization in the albino locusts (Table 2). These results suggest that some hor- monal factor(s) commonly present in different insects can promote dark pigmentation in albino locusts.

In this study, the gregarious coloration was obtained only in those albinos which had been implanted with CC from normal locusts, but whether the factor responsible for this coloration occurs only in CC or not remains uncertain. The involvement of CC and brain in the induction of dark pigments has been suggested using normal individuals of L. migratoria [3, 12, 13]. But, the substance(s) leading to the expression of the grega- rious black color and/or the solitary black homochromy has not been identified chemically. The major obstacles to such a study include (i) difficulty in obtaining a large number of solitary hoppers for a bioassay and (ii) inability to remove the underlying background color or homochromy from such solitary individuals. The albino strain may provide an excellent bioassay system for en-

docrinological studies on phase polychromism in locusts.

ACKNOWLEDGMENTS

I thank Noriko Kenmochi, Matsuji Kaneko, and Eishi Hasegawa for assistance in rearing insects, and Hiroshi Shinbo and Masatoshi Nakamura for giving me L. separ- ata and B. mori. Thanks are also due to M. Paul Pener (Hebrew University) and Lynn Riddiford (University of Washington) for helpful discussions and review of the manuscript.

REFERENCES

1 Fuzeau-Braesch S (1971) Contribution a l’etude de role de hormone juvénile dans la pigmentation. Arch Zool Exp Gen 112: 625-633

2 Fuzeau-Braesch S (1985) Colour changes. In “Com- prehensive Insect Physiology, Biochemistry and Pharmacology Vol 9 Behaviour” Ed by GA Kerkut, LI Gilbert, Pergamon Press, Oxford, pp 549-589

3 Girardie A, Cazal M (1965) Role de la pars inter- cerebralis et des corpora cardiaca sur la melanisation ches Locusta migratoria (L.). CR hebd Seance Acad Sci 261: 4525-4527

4 Hakomori T, Tanaka S (1992) Genetic control of diapause and other developmental traits in Japanese strains of the migratory locust, Locusta migratoria L.: Univoltine vs. bivoltine. Jpn J Ent 60: 319-328

5 Hunter-Jones P (1957) An albino strain of desert locust. Nature 180: 236-237

6 Hsia D Y-Y (1964) Inborn errors of metabolism. In “Diseases of Metabolism”. Ed by GG Duncan, Saunders, Philadelphia, pp 301-448

7 Joly L (1954) Résultats d’implantations systémati- ques de corpora allata a de jeunes larves de Locusta migratoria L. CR Seanc Soc Biol 148: 579-583

8 Joly P, Joly L (1954) Résultats de greffes de corpora allata chez Locusta migratoria. Ann Sci Natur Zool Ser 11, 15: 331-345

9 Joly P, Meyer AS (1970) Action de hormone juvenile sur Locusta migratoria (Orthoptere) en phase gregaire. Arch Zool Exp Gen 111: 51-63

10 Malek SRA (1957) Sclerotization and melanization: two independent processes in the cuticle of the desert locust. Nature 180: 237.

11 Matsumoto S, Isogai A, Suzuki A, Sonobe H (1981) Purification and properties of the melanization and reddish colouration hormone (MRCH) in the army- worm, Leucania separata (Lepidoptera). Insect Biochem 11: 725-733.

12 Nickerson B (1954) A possible endocrine mechan- ism controlling locust pigmentation. Nature 174: 357-358

13

15

16

17

Albinism in Locusta

Nickerson B (1956) Pigmentation of hoppers of the desert locust (Schistocerca gregaria Forskai) in rela- tion to phase coloration. Anti-Locust Bull 24: 1-34 Nolte DJ (1971) Two pleiotropic albino mutations. Proc 4th Cong SA Genet Soc (Pretoria 1970), pp 36-38

Fridman-Cohen S, Pener MP (1980) Precocenes induce effect of juvenile hormone excess in Locusta migratoria. Nature 289: 711-713

Pener MP (1991) Locust phase polymorphism and its endocrine relations. Adv Insect Physiol 23: 1-79 Ogura N (1975) Hormonal control of larval coloura- tion in the armyworm, Leucania separata. J Insect Physiol 21: 559-576

18

19

20

21

471

Roussel JP (1975) Action juvénilisante, chromatot- rope, gonadotrope, et cardiotrope de-JHIII sur Locusta migratoria. J Insect Physiol 21: 1007-1015 Roussel JP (1976) Activité comparee des hormones juveniles en C-18 (JH I) et en C-16 (JH III) chex Locusta migratoria. J Insect Physiol 22: 83-88 Tanaka S (1992) The significance of embryonic diapause in a Japanese strain of the migratory locust, Locusta migratoria (Orthoptera: Acrididae). Jpn J Ent 60: 503-520

Verdier M (1965) Mutation albinos de Locusta migratoria I. Origine et description (C.S.). Bull Soc Zool France 90: 493-501

7 72 . sted) ‘A enari _ 7 es +

Sf ery et Tey

at oft ep co 4, sl alia Walt > ye L Foal, 4a) iy meade yt. (tel) Md By 1) 1 Jal BI m4 2h wi ui ¢ rabies Vid’ 67 i } oe VabnAgT Ot nn png! t +> my? Sn 2 ‘ > ’ a jetzt " y ATO - : i ia) ) iw’

_

7 0.

wh a pe OL Lap

Vredarweas (iokreg a napyry ML WO ILE hey INA "pil dv diwwere OnidiN akyenicuala o 4g (OTUS ei scanaic hei bea

1 (RAs 2 wy ) ee rary bade me)

r i eal

. aA al vi

is : stipe on a heal?

erinds bi viel ta tapi | fi men! | wut alana) inna

ZOOLOGICAL SCIENCE 10: 473-480 (1993)

Use of Primary Cultures of Salmon Hepatocytes for the Study of Hormonal Regulation of Insulin-like Growth Factor I Expression In Vitro

CuNMING Duan!, Naoto HANZAWA, YASUHIRO TAKEUCHI, EISUKE HAMADA, SHIGETOH MIYACHI, and Tetsuya HirANO~

Kamaishi Laboratories, Marine Biotechnology Institute, Heita, Kamaishi, Iwate 026 and 7Ocean Research Institute, University of Tokyo, Minamidai, Nakano, Tokyo 164, Japan

ABSTRACT—Hepatocytes isolated from sockeye salmon (Oncorhynchus nerka) were seeded on an uncoated plastic dish with a positively charged surface, and cultured in hormone-free defined medium. Initially single cells began to aggregate and became flat after being attached to the dish. After approximately 6 days in culture, hepatocytes formed a monolayer sheet with cells in side-by-side contact, and remained in this form for at least 2 weeks. Hepatocytes in multicellular spheroids were also observed. The viability of the cells remained high (85-90%) during culture as judged by trypan blue exclusion. The cells actively incorporated [*H]leucine into cellular and secreted proteins. Salmon and bovine insulins stimulated the leucine incorporation in a dose-dependent manner, suggesting that the cells remained responsive to hormones during culture. The basal IGF-I mRNA level in cultured hepatocytes was much lower than that of in vivo status. Salmon GH increased the level of IGF-I mRNA, whereas salmon prolactin had no effect. These findings suggest that the primary culture system of salmon hepatocytes can be used for the in vitro study of hormonal regulation of IGF-I gene

© 1993 Zoological Society of Japan

expression in salmon.

INTRODUCTION

It has been well-documented that insulin-like growth factor I (IGF-I), which is a polypeptide with structure similarity to proinsulin, mediates many of the growth-promoting actions of growth hormone (GH) in mammals [7]. In teleosts, stud- les concerning the physiological functions of IGF- I has also been reviewed recently [3]. The nucle- otide sequences of IGF-I cDNAs have been deter- mined from coho salmon, Atlantic salmon, masu salmon, rainbow trout, and chinook salmon [2, 5S, 14, 31, 32]. Mammalian IGF-I is highly active in stimulating cartilage sulfation or body growth in teleost species [10, 11, 15, 23, 24]. IGF-I-like

Accepted March 16, 1993 Received January 18, 1993

' Present address and address correspondence to: School of Fisheries HF-15, University of Washington, Seattle, WA 98195 USA.

bioactivity in serum and liver extract, and tissue responsiveness to IGF-I have also been found to be GH-dependent in some teleosts [10, 11, 13, 15, 24]. All of these findings strongly suggest that the GH-IGF-I axis is well conserved throughout evolu- tion from teleost fishes to mammals.

In mammals, the major synthesis site of circulat- ing IGF-I is liver, and GH directly stimulates the hepatic IGF-I mRNA expression and IGF-I pep- tide production [7, 27]. In teleosts, Cao et al. [5] reported that injections of bovine GH increased hepatic IGF-I mRNA levels. Duan and Hirano [11] showed that hypophysectomy reduced, and injection of eel GH restored, IGF-I mRNA con- tent in the liver of the eel. They also reported that fasting significantly reduced hepatic IGF-I levels in the eel. Sakamoto et al. [30] reported that transfer of rainbow trout from fresh water to seawater increased IGF-I mRNA levels in liver, kidney and gill. In spite of some in vivo studies, detailed

474 C. Duan, N. HAnzawa et al.

knowledge concerning the regulation of IGF-I production in teleosts is still scarce. To further examine the regulatory mechanism of IGF-I in teleosts, an in vitro model has been called for.

Short-term culture of isolated hepatocytes from salmonid species has been extensively explored for their utility in toxicologic and metabolic studies [4, 25]. Long-term culture of salmon or trout hepato- cytes, however, has not been studied as extensive- ly. The purpose of the present study was to establish a long-term primary culture system of salmon hepatocytes and to examine the effect of GH on IGF-I mRNA expression in vitro.

MATERIALS AND METHODS

Animals

Sockey salmon (350-600 g body weight) were obtained from Sunlock, Ltd., Kamaishi, Iwate. The fish were held in seawater in a flow-through 1,000 / aquarium and maintained at 15-17°C under natural photoperiod. They were fed once per day with commercial pellets.

Hormones

Bovine insulin (24 U/mg) was purchased from Sigma. Salmon insulin was provided by Dr. Erika Plisetskaya, University of Washington, Seattle, WA. Recombinant chum salmon GH was pro- vided by Kyowa Hakko Kogyo, Co, Ltd.,Tokyo, and chum salmon prolactin was a gift from Dr. Hiroshi Kawauchi, Kitasato University, Sanriku, Iwate, Japan.

Isolation of hepatocytes

Fish were anesthetized by immersion in 0.02% MS222 prior to handling. A polyethylene cannula, connected to the perfusion reservoir, was inserted into the intestinal vein. The liver was perfused with an initial flow rate of 2 ml/min with 30 ml Ca’*-free Krebs-Ringer. Then, the liver was perfused with 50 ml of the perfusion medium con- taining collagenase (Wako Pure Chemical Indus- tries, 0.3 mg/l) for 25 to 30min. The liver was removed from the fish and minced with scissors. The minced tissue was transferred to a beaker with 50 ml Ca**-free Krebs-Ringer containing collage-

nase, and mixed by stirring with a magnetic bar for 30 min. The suspension was filtered through a filter unit (150-mesh screen) and centrifuged at 50 xg for 2 min. The supernatant was discarded and the cells were resuspended in Krebs-Ringer and centrifuged for 2 min. The washings were repeat- ed three more times. The final pellet was sus- pended with culture medium to a volume of 20 ml. Cells were counted in a Thoma haemacytometer.

Cell culture

Cells were suspended in William’s E medium (Sigma) with 5% of fetal bovine serum (Sigma) and plated at 3 x 10° cells/60-mm Falcon Primaria dish. After about 12 hr, non-adherent cells were discarded, and the medium was replaced with serum-free William’s E medium after washing 3 times by Dulbecco’s phophate buffered saline (PBS). The cells were incubated in 5%CO,/95% air at 15°C. Cells cultured for 5 or 6 days were used for the experiments.

Cell morphology

Morphology of the cells was examined by phase- contrast microscope and scanning electron micro- scope. For scanning electron microscope, the cells were fixed with 1% glutaraldehyde, postfixed with 1% OsO,, dehydrated through a graded series of ethanol, and finally soaked in isoamyl acetate. The cells on the plate were dried with a critical point dryer (Hitachi, HPC-1), coated with Pt-Pd using an Eiko Ion Coater, and then observed with a JSM-U3 scanning electron microscope.

Incorporation of [?H]leucine

After washing with PBS without Ca** and Mg** 3 times, the cells were incubated in 2 ml fresh William’s E medium with or without hormone for 16hr. Then, 2 Ci [*3H] leucine (NEN) was added and incubated for 4 hr. Four to five dishes/ group were used in each treatment. The measure- ment of incorporation of [*H]leucine into hepato- cyte as well as into the secreted proteins was performed as described by Plisetskaya et al. [28] with some modification. Briefly, the cells were separated from the medium by centrifugation for 5 min at 300g. The cell pellets were resuspended in ice-cold water and sonicated. The proteins from

IGF-I Expression in Salmon Hepatocytes 475

both the sonicated cells and the medium were precipitated by ice-cold 5% trichloroacetic acid (TCA). The TCA precipitates were washed twice with 5% TCA. Then, the pellets were dissolved in 1 ml 0.1M NaOH. The radioactivity was meas- ured in a scintillation counter (Packard TRI- CARB 300) after adding 6 ml scintillation fluid (Aquasol-2, NEN). The radioactivity in the cell fraction was expressed as intracellular protein, and that in medium fraction was expressed as ex- tracellular or secreted protein. The results were presented as DPM/10° cells viable at the beginning of the tracer experiment.

RNA isolation and Northern blot analysis

The hepatocytes cultured for 6 days were incu- bated in 2 ml fresh William’s E medium with or without hormone for 24 hr. After removal of the culture medium, cells were washed twice with Ca**/Mg?*-free PBS. Total RNA was isolated with a single-step extraction method using acid guanidium thiocyanate [6]. Poly [A*] RNA was purified by oligo(dT)-latex using a commercial kit (Oligotex-dT30, Takara). A 1.1 kb cDNA probe encoding the entire coding sequence for coho salmon IGF-I (provided by Dr. Stephen J. Duguay, University of Washington, Seattle, WA) and a chicken ~-actin cDNA probe (Sigma) were labelled with [°?P] by a multiprime DNA labelling kit (Amersham).

Poly [A*] RNA was denatured by 2.2 M formal- dehyde, electrophoresed on 1% agarose gels, and bound to a nylon filter (Hybond™) by capillary transfer in 20XSSC. The hybridization was per- formed as previously described [11]. Auto- radiography was performed by exposing the filters to Kodak X-Omat AR film at —70°C.

Statistical analysis

The results were expressed as means+SE, and differences among groups were evaluated statisti- cally by ANOVA followed by the Fisher Protected Least Significant Difference Test (PLSD) using the Statview 512+ [8].

RESULTS

Changes in hepatocytes morphology during pri- mary culture

The yield of liver cells obtained by the present method was about 1X10*cells/g liver. Paren- chymal cells (hepatocytes) were the major cell type (more than 98%). The viability was usually greater than 90% as judged by trypan blue exclu- sion test. Figure la shows the typical spherical shape of freshly isolated salmon hepatocytes. Iso- lated salmon hepatocytes were single, round and large cells.

The attachment of salmon hepatocytes was in- complete in ordinary non-coated culture dishes. The plating efficiency in these ordinally non- coated culture dishes was usually lower than 20%. Preliminary experiments showed that coating the dish with bovine fibronectin improved the attach- ment to some extent, but bovine collagen or gelatin was ineffective (data not shown). Subse- quent results showed that salmon hepatocytes attached firmly to Falcon Primaria culture dishes (Fig. 1b), which have positively charged surface, after overnight incubation (plate efficiency higher than 85%). Thus, cells were cultured in Primaria culture dishes in subsequent experiments.

Salmon hepatocytes showed significant morpho- logical changes after attachment to the dish sur- face. Initially, single, round hepatocytes began to aggregate to each other, and became flat in 3-5 days (Fig. 1b). After being cultured for 6 days, cells spread and formed typical monolayer (Fig. lc). The boundaries between cells became clear, and nuclei were clearly observed. They remained in this form for up to 3 weeks. Thereafter, they began to die, and more fibroblast-like cells appeared in the culture dishes (Fig. 1d). The multicellular spheroids with highly packed cells were observed sometimes (Fig. 2 a, b).

Protein synthesis and effect of insulin

Cultured hepatocytes actively incorporated [*H] leucine into intracellular and extracellular pro- teins. Actinomycin D, which is an inhibitor of RNA synthesis and can affect protein synthesis via inhibition of RNA synthesis, decreased the [*H]

476 C. Duan, N. HAnzawa et al.

Fic. 1. Phase-contrast microscopy of salmon hepatocytes in primary culture. Cells at day 0 (a); day 3 (b); day 10 (c); day 30 (d). All magnification are x 200.

Z a a end : £ ay

re <i = ; Wee ee) “i AG = ¥ a

Fic. 2. Multicellular spheroids of salmon hepatocytes. (a) Phase-contrast microscopy (magnification, x 200); (b) scanning electron microscopy (magnification, < 1,000).

IGF-I Expression in Salmon Hepatocytes 477

* Intracellular

DPM X 10° /10° cells N lo)

Extracellular

DPM X 10 3/10° cells

Cont 0.18 1.75 17.5 175 10

Actinomycin D (ug/ml) Fic. 3. Effect of insulin on the incorporation of [°H] leucine into intracellular and extracellular proteins i primary cultures of salmon hepatocytes. Data are means+SE (n=5). *Significantly different from the control (p<0.05).

Bovine insulin (nM)

leucine incorporation into proteins in a dose- dependent manner. At a dose of 10 g/ml, it significantly reduced the incorporation of labelled leucine into intracellular and extracellular proteins (26 and 36% of the control levels, respectively) (Fig. 3), indicating that most of the [*H] leucine measured was indeed incoporated into newly synthesized proteins in this system.

As shown in Figure 3, bovine insulin stimulated the [°H] leucine incorporation into both intracellu- lar proteins and extracellular proteins in a dose- dependent manner. The minimum effective dose was 1.75nM for intracellular protein synthesis, and 0.18 nM for protein secretion. Similar dose- dependent stimulation was also obtained with homologous salmon insulin (data not shown).

IGF-I mRNA expression and effect of GH

Figure 4 shows the results of a Northern blot with 15 yg of poly [A*] RNA isolated from intact salmon liver or hepatocytes cultured without any hormone for 6 days. Both liver and hepatocyte RNA showed a major class of IGF-I mRNA (approximately at 3.9kb). The signal of hepato- cyte RNA was much weaker than liver RNA so

IGF-I

12

B-actin

Fic. 4. Northern blot analysis of RNA. Fifteen micro- grams of poly [A*] RNA isolated from salmon liver (land 1) and six-day-old cultured salmon hepato- cytes (lane 2) were used. (a) IGF-I mRNA. The exposure time was 2 days for liver RNA (lane 1) and 7 days for hepatocyte RNA (lane 2). (b) f-Actin mRNA. The same blot was rehybridized with chicken £-actin CDNA. The exposure time was 2 days.

IGF-|

12345678910

B-actin

Fic. 5. Effects of salmon GH and prolactin on IGF-I mRNA levels in primary cultures of salmon hepato- cytes. Lanes 1-2: prolactin 1,000 ng/ml; lanes 3-4: prolactin 100 ng/ml; lanes 5-6: control; lanes 7-8: GH 100 ng/ml; lanes 9-10: GH 10 ng/ml. Approx- imately two microgram of poly [A*] RNA was used for each lane. The exposure time was 7 days for both IGF-I and @-actin mRNA.

478 C. Duan, N. HAnzawa et al.

that the blot had to be exposed for longer time (7 days as compared to 2 days of liver RNA), suggest- ing low expression level of IGF-I mRNA in hor- mone-free culture condition in vitro.

The effect of chum salmon GH and prolactin on IGF-I mRNA expression was then examined in primary cultures of salmon hepatocytes. Approx- imately 2 ug poly [A*] RNA were obtained from each dish, and used for Northern blot analysis. Clear hybridization signals of IGF-I mRNA were obtained in GH-treated group (Fig. 5a, lanes 7-8, GH 100 ng/ml), whereas no apparent hybridiza- tion signal of IGF-I mRNA was seen in the control group or prolactin-treated groups after 7 days of exposure (Fig. 5a, lanes 1-6). When rehybridized the same blot with a #-actin cDNA probe, similar levels of clear 8-actin mRNA signal were obtained with all groups (Fig. 5b, lanes 1-10).

DISCUSSION

In this study, salmon hepatocytes were cultured in a serum-free medium for up to several weeks. The cultured cells were responsive to insulins in term of stimulation of protein synthesis. GH stimulated IGF-I mRNA expression in primary cultures of hepatocytes, whereas prolactin had no effect.

There are many reports on culture of fish hepatocytes [25, 4]. To date, most of the studies used freshly isolated liver cells in suspension or primary culture cells over very short periods (with- in 24 h), whereas few studies were carried out with long-term primary culture system in a defined medium. This is in sharp contrast to the extensive studies on long-term cultures of mammalian hepatocytes [26]. This was partly because the attachment of fish hepatocytes to the surface of culture plate was slow and remained relatively weak, making it difficult to manipulate the cultures without disrupting of the cells [17, 22]. To solve this problem, efforts have been made to coat the culture plate various extracellular matrix compo- nents such as mammalian fibronectin, collagen, gelatin and fish serum or fish skin extract [4, 16, 18, 20, 22]. While some of the matrix components such as mammalian and fish fibronectins improved the efficacy of attachment [18], it is time-

consumimg to prepare the coated plates. In this study, we have shown that salmon hepatocytes attached to Primaria dishes firmly without any treatment, and survived well up to several weeks.

The freshly isolated salmon hepatocytes seemed undamaged when viewed under phase-contrast microscope and scanning electron microscope (data not shown). The viability of the cells was approximately 90%. During the culture, cells reaggregated, became flat, and spread to form a monolayer. While most cells were in the mono- layer form, some spheroids were also observed. In mammals, two distinct features of primary culture of adult hepatocytes are observed when the cells are cultured in Primaria dishes: monolayer culture with proliferative activity and multicellular sphe- roids with poor proliferative activity and high albumin-producing ability [21]. It is not clear whether the biochemical and physiological prop- erties of these two forms of salmon hepatocytes in culture are similar to those of mammalian hepato- cytes.

It has been well documented in teleosts that insulin stimulates protein synthesis both in vivo and in vitro [1, 19, 28]. In the present study, a dose-dependent stimulation was obtained with mammalian and salmon insulins. The minimum effective dose was 1.75nM, which is within the physiological range of circulating insulin salmonids [29]. The dose dependency and minimum effective dose of insulins are consistent with previous findings obtained in intact animals, cultured liver slices, and suspension of isolated liver cells [1, 19, 28], suggesting our primary culture system is suit- able for study of hormone action.

Although some in vivo studies have indicated the GH-regulated hepatic IGF-I mRNA express- ion in teleosts [5, 11, 9, 30], there was no published in vitro information to date. In the present study, Northern blot analysis revealed a major band of 4 kb with cultured hepatocyte RNA, which was identical to that obtained with intact liver RNA. A longer exposure time, however, was necessary to visualize the hybridization signals with hepatocyte RNA (7 days) than with liver RNA (2 days), indicating a lower level of IGF-I mRNA express- ion in vitro as compared to that of in vivo. The lower expression level is likely due to the lack of

IGF-I Expression in Salmon Hepatocytes 479

hormone stimulation, because the cells were cul- tured in hormone-free medium. Addition of sal- mon GH to the medium at a concentration of 100 ng/ml markedly increased the IGF-I mRNA levels. There was no apparent signal in the control and prolactin-treated groups. Since /-actin mRNA levels were all similar, the absence of IGF-I mRNA in the control and prolactin-treated groups is most likely to be due to low expression levels of IGF-I mRNA. These results suggest that GH has a direct effect on hepatic IGF-I mRNA expression in salmonid species, and that our pri- mary culture system can be used as an in vitro model to study the hormonal regulation of IGF-I expression.

While our study demonstrated for the first time that GH stimulates hepatic IGF-I mRNA express- ion in vitro in a teleost species, the accurate quantitation of the results was difficult due to the low sensitivity of Northern blot and the low ex- pression level of IGF-I. As mentioned above, using Northern blot, we were unable to detect IGF-I mRNA in poly [A*] RNA (2 yg) prepared from hepatocytes of the control groups even after a long exposure unless using larger amount of RNA sample. Thus, a more sensitive and reliable assay is necessary for further study of hormonal regula- tion of IGF-I mRNA expression in vitro in salmo- nids. RNA dot blot or slot blot is a simpler version of Northern blot technique, and is slightly more sensitive and semi-quantitative. However, results of dot-blot or slot blot can be misinterpreted if the RNA samples are contaminated by DNA. In comparison, another technique, solution hybridi- zation/RNase protection assay, is more sensitive, specific and fully quantitative. Recently, we have developed a solution hybridization/RNase protec- tion assay for salmon IGF-I mRNA [9]. Quantita- tive data obtained by this assay confirmed the finding of the present study: GH but not prolactin stimulates IGF-I mRNA expression in primary cultures of salmon hepatocytes. Studies concern- ing the detailed regulation mechanism of IGF-I mRNA expression in salmon are in progress using the primary culture system of hepatocytes estab- lished in the present study.

ACKNOWLEDGMENTS

We are grateful to Mr. Takeshi Yamanome, Sunlock, Ltd., for providing the experimental fish. We thank Drs. Erika M. Plisetskaya and Stephen J. Duguay, School of Fisheries, University of Washington, for their critical reading of the manuscript. Thanks are also due to Mr. Yoji Kikuchi, Kamaishi Laboratories, Marine Biotech- nology Institute, for his technical assistance. This work was supported in part by grants from the Ministry of Education, Science and Culture and also from Fisheries Agency, Japan to T.H., and a grant from the Association for Overseas Technical Scholarship (AOTS) of Japan to C.D.

REFERENCES

1 Ablett RF, Sinnhuber RO, Selivonchick DP (1981) The effect of bovine insulin on “C-glucose and H-leucine incorporation in fed and fasted rainbow trout (Salmo gairdneri). Gen Comp Endocrinol 43: 211-217

2 Aoyagi K, Kojima T, Tanabe Y, Soma G, Saneyoshi M (1993) Total nucleotide sequence and expression in E. coli of cDNA encoded insulin-like growth factor I (IGF-I) from cherry salmon, Oncorhynchus masou. Mol Marine Biol Biotech (in press)

3 Bern HA, McCormick SD, Kelley 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 poikilothermic verte- brates. In “Modern Concept of Insulin-like Growth Factors” Ed by EM Spencer, Amsterdam, Elsevier, pp 85-96

4 Blair JB, Miller MR, Pack D, Barnes R, Teh SJ, Hinton D (1990) Isolated trout liver cells: Estab- lishing short-term primary cultures exhibing cell-to- cell interactions. In vitro Cell Dev Biol 26: 237-249

5 Cao QP, Duguay SJ, Plisetskaya EM, Steiner DF, Chan SJ (1989) Nucleotide sequence and growth hormone-regulated expression of salmon insulin-like growth factor I mRNA. Mol Endocrinol 3: 2005- 2010

6 Chomcezynski P, Sacchi N (1987) Single-step method of RNA isolation by acid guanidium thiocyanate- pheno-chloroform extraction. Anal Biochem 162: 156-159

7 Daughaday WH, Rotwein P (1989) Insulin-like growth factor I and II. Peptide, messenger ribonuc- leic acid and gene structures, serum, and tissue concentrations. Endocrine Rev 10: 68-91

8 Dowdy S, Wearden S (1989) “Statistics for Re- search”. John Wiley & Sons, New York, pp 303-305

9 Duan C, Duguay SJ, Plisetskaya EM (1993) Insulin- like growth factor I (IGF-I) mRNA expression in coho salmon, Oncorhynchus kisutch: Tissue dis-

10

11

12

13

14

15

16

17

18

19

20

21

480

tribution and effects of growth hormone/ prolactin family proteins. Fish Physiol Biochem (in press) 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 cartilages of the Japanese eel, Anguilla japonica. Gen Comp Endocrinol 79: 3220- 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 Endo- crinol 79: 326-331

Duguay SJ, Park LK, Samadpour M, Dickhoff WW (1992) Nucleotide sequence and tissue distribution of three insulin-like growth factor I prohormones in salmon. Mol Endocrinol 6: 1202-1210

Gray ES, 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 131: 57-66

Haschemeyer A, Mathews R (1983) Temperature dependency of protein synthesis in isolated hepato- cytes of Antarctic fish. Physiol Zool 56: 78-87 Hayashi S, Ooshiro Z (1985) Primary culture of the freshly isolated liver cells of the eel. Nippon Suisan Gakkaishi 51: 765-771

Hayashi S, Ooshiro Z. (1986) Primary culture of the eel hepatocytes in the serum-free medium. Nippon Suisan Gakkaishi 52: 1641-1651

Inui Y, Ishioka H (1983) Effects of insulin and glucagon on amino acid transport into the liver and opercular muscle of the eel in vitro. Gen Comp Endocrinol 51: 213-218

Kocal T, Quinn BA, Smith IR (1988) Use of trout serum to prepare primary attached monolayer cul- tures of hepatocytes from rainbow trout (Salmo gairdneri). In Vitro Cell Dev Biol 24: 304-308

Koide N, Sakaguchi K, Koide Y, Asano K, Kawa-

22

23

24

25

26

27

28

29

30

31

32

C. Duan, N. Hanzawa et al.

guchi M, Matsushima H, Takenami T, Shinji T, Mori M, Tsuji T (1990) Formation of multicellular spheroids composed of adult rat hepatocytes in dishes with positively charged surface and under other nonadherent environments. Exp Cell Res 186: 227-235

Maitre JL, Valotaire Y, Guguen-Guillouzo Z (1986) Estradiol-17f stimulates of vitellogenin synthesis in primary culture of male rainbow trout hepatocytes. In vitro Cell Dev Biol 22: 337-343

McCormick SD, Kelley KM, Young G, Nishioka RS, Bern HA (1992) Stimulation of coho salmon growth by insulin-like growth factor I. Gen Comp Endocrinol 86: 398-406

McCormick SD, Tsai PI, Kelley KM, Nishioka RS, Bern HA (1992) Hormonal control of sulfate incor- poration in branchial cartilage of coho salmon: Role of IGF-I. J Exp Zool 262: 166-171

Moon TW, Walsh PJ, Mommsen TP (1985) Fish hepatocytes: A model metabolic system. Can J Fish Aquat Sci 42: 1772-1782

Nakamura T (1987) “Methodology of Primary Cul- ture of Hepatocytes”, Jap Sci Soc Press, Tokyo, pp 5-53 (in Japanese)

O’Neill IE, Houston B, Goddard C (1990) Stimula- tion of insulin-like growth factor-I production in primary cultures of chicken hepatocytes by chicken growth hormone. Mol Cell Endocrinol 70: 41-47 Plisetskaya E, Bhattacharya S, Dickhoff WW, Gorbman A (1984) The effect of insulin on amino acid metabolism and glycogen content in isolated liver cells of juvenile of coho salmon, Oncorhynchus kisutch. Comp Biochem Physiol 78A: 773-778 Plisetskaya EM, Dickhoff WW, Paquette TC, Gorb- man A (1986) The assay of salmon insulin by homologous radioimmunoassay. Fish Physiol Biochem 1: 35-41

Sakamoto T, McCormick SD, Hirano T (1992) Mode of action of growth hormone in salmonid osmoregulation: Mediation by insulin-like growth factor I. Fish Physio Biochem (in press) Shamblott, MJ, Chen, TT. Identification of a second insulin-like growth factor in a fish species. Proc Natl Acad Sci USA 89: 8913-8917

Wallis AE, Devlin RH (1993) Duplication insulin- like growth factor-I genes in salmon display alterna- tive splicing pathways. Mol Endocrinol 7: 409-422

ZOOLOGICAL SCIENCE 10: 481-488 (1993)

Metabolism of Androgens by the Mouse Submandibular Gland and Effects of Their Metabolites

KAZUHIKO SAWADA AND TETSUO NOUMURA

Department of Regulation Biology, Faculty of Science, Saitama University, Urawa, Japan

ABSTRACT— In the mouse submandibular gland, sexual difference becomes evident on day 30, when the granular convoluted tubules (GCT) of the gland rapidly grow in response to drastically increased levels of circulating testosterone and 5a-dihydrotestosterone (DHT) in the male. We studied the in vitro metabolism of both androgens by the mouse gland and the effects of their metabolites on the gland.

By the glands on days 20 and 30, *H-testosterone was not further converted to any derivatives. In contrast, *H-DHT used as a substrate was metabolized partly to 7=H-Sa-androstane-3a,17-diol (3a-diol) and *H-5a-androstane-3,17/-diol (3-diol) in both sexes (approximately 9 and 1%, respectively). On the basis of the metabolites identified, the gland contained both 3a- and 38-hydroxysteroid dehy- drogenase activities, which were not sexually different.

Neonatally-castrated mice were given daily injections of 3a-diol, 3@-diol, testosterone and DHT for 1-10 days from day 20. The relative occupied area (ROA) of GCT in the gland increased in both sexes by treatment with 3a-diol or DHT for 4 days, or by 3f-diol or testosterone for 10 days. The order of ROA gain after 4-10 days was DHT >3a-diol >testosterone >3f-diol. The mitotic activity of GCT increased in both sexes by treatment with 3a-diol or DHT for 4 days, but in only males by 3f-diol or testosterone for 10 days.

The results suggest that in the mouse gland testosterone exerts its effects without any intraglandular conversion, and that DHT primarily acts in its form, whereas its intraglandular metabolite, 3a-diol may

© 1993 Zoological Society of Japan

be supplementary as a stimulant.

INTRODUCTION

In rodents, the male submandibular gland is larger than the female one and has a more complex morphology. The glandular contents of biological- ly active polypeptides, including nerve growth factor, epidermal growth factor, renin and pro- teases, are higher in the male than in the female, being responsive to androgens [3, 6, 12, 14, 16, 19, 28]. By histological, ultrastructural and mor- phometrical studies, both sexes experience a simi- lar morphogenesis of the gland during perinatal development, and then the sexual difference arises at 3-4 weeks of age, when the granular convoluted tubules (GCT) grow more rapidly in the male than in the female [10, 11, 13, 21]. In a completely androgen-independent state (e.g., neonatal castra- tion or androgen-insensitive Tfm mutation), the

Accepted April 17, 1993 Received March 2, 1993

glands carry out the feminine development [21, 22]. The masculine development of the gland is caused by circulating androgens, testosterone and 5a-dihydrotestosterone (DHT), the latter being more effective [23]. In male mice, their circulating levels are drastically increased between days 20 and 30 [23], when sexual difference of the gland occurred morphologically.

Therefore, in order to determine how testoster- one and DHT exert their effects on the mouse submandibular glands, we examined in vitro meta- bolism of these androgens by the glands, and the effects of the resulting metabolites on the glands by morphometry.

MATERIALS AND METHODS

Animals

CD-1 mice were obtained from Charles River Japan Co. and maintained by randomly mating in

482 K. SAWADA AND T. NOUMURA

our laboratory. The animals were given a com- mercial diet (CRF-1: Charles River Japan Co.) and tap water ad libitum and were kept at 23+1°C under 12-hr artificial illumination (from 8:00 to 20:00).

In vitro androgen metabolism by the mouse sub- mandibular gland

Both male and female mice were killed on days 20 and 30. Approximately 50 mg tissue of the submandibular glands were minced into about 2 mm-fragments. Minced tissues were incubated with 10 nM [1, 2, 6, 7--H]testosterone and [1, 2, 4, 5, 6, 7--H]5a-dihydrotestosterone (DHT) (specific activity, 70 and 117 Ci/mmol, respectively; Amer- sham) in 2ml Hanks’ solution (Nissui Phar- maceutical Co., Ltd., Tokyo, Japan) containing 4 mM glucose-6-phosphate-2NA (Boehringer Man- nheim GmbH, Germany) for 1 hr at 37°C in an atmosphere of 95% air and 5% CO). After 1 hr, the incubation was stopped with 3 ml ethyl acetate, containing each 100g of the following non- radioactive reference steroids: 4-androstene-3,17- dione (androstenedione), Sa-androstane-3,17- dione (Sa-androstanedione), Sa-androstan-3a-ol- 17-one (androsterone), 4-androsten-17/-ol-3-one (testosterone), Sa-androstan-17-ol-3-one (5Sa-di- hydrotestosterone, DHT), 5a-androstane-3a,17/- diol (3a-diol) and 5a-androstane-3,17f-diol (3£- diol) (Sigma Chemical Co., St. Louis, Mo, U.S.A.). After shaking, the ethyl acetate fractions were transferred to a new set of glass tubes, and the water phase was reshaken twice each time with 3 ml ethyl acetate. The resulting 9 ml ethyl acetate fraction was washed with 1 ml saturated NaHCO; and twice with 1 ml distilled water, and was evapo- rated. The extracted steroids were redissolved in 20 ul methanol-dichloromethane (1:1, vol/vol), and applied to TLC Silica gel plate 60 HFo54+366 (Merck, Darmstadt, Germany). The plates were developed in the solvent system, dichloromethane- diethyl ether (4:1, vol/vol), and were scanned to detect the areas of radioactivity by Chromatoscan- er (Aloka, Tokyo, Japan). The radioactive areas were scraped off and steroids were extracted from the Silica gel three times with 2 ml methanol- dichloromethane (1:1, vol/vol) each time. The dried residues of radioactive steroids, except for

5a-androstanediols, were transferred to scintilla- tion vials. After adding 4 ml scintillation fluid (6 g DPO, 0.3 g POPOP in 1,000 ml toluene), radioac- tivity was measured in a Packard Liquid Scintilla- tion Counter, model 3255 with an efficiency of 50% for 7H. The fraction of 5a-androstanediols was acetylated by dissolving in equal volumes of pyridine and acetic anhydride (0.2 ml, total volume) and allowing the mixture to stand over- night at room temperature. The mixture was evaporated and redissolved in 2041 methanol- dichloromethane (1:1, vol/vol), applied to the TLC plates and developed twice in dichlor- omethane. The results were expressed as the percentages of total radioactivity. The protein content of each sample was determined by BCA protein assay reagent (Pierce, Rockford, IL, USA) and enzyme activities of 3a- and 38-hydroxysteroid dehydrogenases were expressed as f mol of 3a- and 36-diols produced from DHT per hour per mg protein, respectively. Each datum was the mean of 2-4 independent experiments. The procedure losses were less than 10% after first chromatogra- phy and less than 20% after acetylation and second chromatography (for separation of 3a- and 3/- diols), and the results were corrected for these losses.

Morphological studies

Mice of both sexes were castrated on day 0 and were subcutaneously given daily treatments with each 100 ug of 3a-diol, 3@-diol, testosterone and DHT, and the vehicle alone (0.1 ml sesame oil) for 1, 4 and 10 days starting on day 20, and they were killed on days 21, 24 and 30, respectively. All animals were subcutaneously given a single injec- tion of colchicine (5 “g/g body weight) dissolved in 0.9% NaCl, 5hr before they were killed to arrest cell division at metaphase. Submandibular glands were weighed and fixed in Bouin’s solution, embedded in paraffin and sectioned at 8 um. Sec- tions were stained with Delafield’s hematoxylin and eosin.

The sectional figures of submandibular glands in the microscopic enlargements (400) were traced by the camera lucida, and the area of GCT was measured by the picture analyzer (Logitec K-510, Kantou Denshi Co., Japan) connected with a

Androgens and Mouse Submandibular Glands 483

microcomputer (NEC PC-9801DA). The area was expressed as percentages of the total area of the gland (the relative occupied area, ROA). Cell height of GCT was measured in randomly chosen five. sections.

Cells in division per 500-1,000 cells were counted in the GCT of the gland. Mitotic rate was estimated by counting the cells at metaphase per 5 hr in 100 cells.

Statistical analysis

Data were statistically analyzed by Student’s t-test or by Cochran-Cox test.

RESULTS

In vitro androgen metabolism by the mouse sub- mandibular gland

The results in Table 1 are expressed as the percentage conversion of each substrate to metabolites/hr/mg of wet tissue, and a proposed metabolic pathway of androgens in the mouse submandibular gland is shown in Figure 1. By the glands on days 20 and 30, *H-testosterone was not further converted in vitro to any radioactive metabolites, including DHT, and its recovery was

TABLE 1. Jn vitro androgen metabolism by the mouse submandibular gland

% recovery

substrate animals age in days testosterone DHT 3a-diol 3-diol testosterone male 20 98.8 ND ND ND 30 96.1 ND ND ND female 20 96.7 ND ND ND 30 95.6 ND ND ND DHT male 20 — 79.4 9.6 1.1 30 — 83.5 9.0 1.4 female 20 — 78.9 8.2 il 2 30 — 79.3 8.7 1.4

The results are expressed as the percentage conversion of each substrate to metabolites/hr/mg wet tissue. Values are the mean of 2-4 independent experiments. DHT=5a-dihydrotestosterone, 3a-diol=5Sa- androstane-3a,17-diol, 3-diol=5Sa-androstane-38,17f-diol, ND=not detected.

5a-reductase OH '

snes sre Za oA O 2 H

OH

3a-HSD H Or

H OH : 5a-androstane-3 a,17£ -diol

(3a-diol)

OH testosterone 5 a-dihydrotestosterone oo

(DH T)

38-HSD HO H

5a-androstane-3,17 8 -diol (38-diol)

Fic. 1. A proposed metabolic pathway of androgens in the mouse submandibular gland. HSD=hydroxysteroid

dehydrogenase.

484 K. SAWADA AND T. NoumMuURA

TABLE 2. 3a- and 3f-Hydroxysteroid dehydrogenase activities in the mouse submandibular gland

enzyme activity (fmol-mg ‘ protein-hr~‘)

animals age in days 3a-HSD 38-HSD male 20 369.8 43.4 30 645.4 98.1 female 20 314.3 47.6 30 574.6 93.2

Values are the mean of 2-4 independent experiments. HSD=hydroxysteroid dehydrogenase.

more than 95% (Table 1). On the other hand, 3H-DHT was metabolized in limited amounts to tritiated 3a- and 3f-diols (approximately 9 and 1%, respectively). The amounts of metabolic TO products from DHT did not differ between sexes at each age (Table 1). On the basis of the metabo- lites identified, the gland contained both 3a- and 38-hydroxysteroid dehydrogenase (HSD) activi- ties. The activities of these enzymes increased

oT y

50

e oll

between 20 and 30 days, but were not sexually © 3a-diol different at both ages (Table 2). oe ae ee

v DHT

Effects of 5a-androstanediols on the mouse sub- mandibular gland

female By the mouse glands, some amounts of 3a- and 100 38-diols could be converted in vitro from DHT. Therefore, the effects of 3a- and 3f-diols on the glands were examined morphometrically at 30 days of ages, and were compared with those of testos- terone and DHT. In 30-day-old mice, the male glands were superior to the female glands in all morphometric parameters used: the weight, the ROA of GCT, the cell height of GCT, and the mitotic activity of GCT. By neonatal castration, these sex differences were abolished, and the 0 intraglandular structures in the males resembled to treatment periods (days) those in the normal females, showing lower values in all four parameters [21]. In females, castration

weight of the glands (mg)

50

5 10

Fic. 2. Change in the weight of the submandibular glands of neonatally-castrated male and female

could not induce any changes in their glands [21]. mice, treated with each 100 ug of Sa-androstane- The net weight of the glands in castrated males 3a,17B-diol (3a-diol), 5a-androstane-3,17/-diol gradually increased and became significantly (38-diol), testosterone and Sa-dihydrotestosterone

heavier by 10 daily treatments with all four (DHT), and the vehicle alone for 1, 4 and 10 days from day 20. Values are mean+S.E.M., n=4-S per

EUhOWonS, ciNOmyA ne cuieoss Gl DIR wee group. a: P<0.001 vs oil control (Student’s test),

greater than those of other three androgens (Fig. b: P<0.05, c: P<0.02 vs oil control (Cochran-Cox 2). In castrated females the gland weight gain was test).

Androgens and Mouse Submandibular Glands 485

male 50

e oll © 3a-diol 4 38-diol 0 testosterone vy DHT

25

female

50

ROA of GCT (%) °

25

Co) 5 10 treatment periods (days)

Fic. 3. Change in the relative occupied area (ROA) of the granular convoluted tubules (GCT) of the sub- mandibular glands of neonatally-castrated male and female mice, treated with each 100 yg of 5a- androstane-3a,17£-diol (3a-diol), 5a-androstane- 38,17f-diol (36-diol), testosterone and 5a-dihydro- testosterone (DHT), and the vehicle alone for 1, 4 and 10 days from day 20. Values are mean+ S.E.M., n=8-10 per group.

caused only by treatment with 3a-diol or DHT after 10 days (Fig. 2).

Both the ROA and the cell height of GCT in castrated animals were increased by 4 daily treat- ments with 3a-diol or DHT and by 10 daily treat- ments with 3f-diol or testosterone (Figs. 3 and 4). After 10 daily treatments, 3a-diol and DHT were nearly equal to increase both the ROA and the cell height of GCT, but the former was slightly inferior to the latter after 4 days of treatments (Figs. 3 and 4). The effects of 3f-diol on these two parameters were less effective than those of testosterone, and were sexually different, the male GCT being more responsive than the female GCT (Figs. 3 and 4). Both 3a- and 3-diols could not cause any in-

male 25 ® oll ° 3a-diol A 3f-diol 207? Oo testosterone a) 20 v DHT a ~ iS S15 @) = ® 2 eee eo - (0) iS) female GS 25 ~ ° = a > c ® a c 20 a b a 15 eee eee (0) 5 10

treatment periods (days)

Fic. 4. Change in the cell height of the granular convo- luted tubules (GCT) of the submandibular glands of neonatally-castrated male and female mice, treated with each 100 ug of Sa-androstane-3a,17-diol (3a- diol), 5a-androstane-3,17-diol (3-diol), testoster- one and 5a-dihydrotestosterone (DHT), and the vehicle alone for 1, 4 and 10 days from day 20. Values are mean+S.E.M., n=8-10 per group. a: P <0.001 vs oil control (Student’s t-test), b: P<0.005, c: P<0.001 vs oil control (Cochran-Cox test).

creases in the ROA of other three regions, the acini, the intercalated ducts and the excretory striated ducts, in both the male and female glands (data not shown).

3a-Diol induced to increase the mitotic activity of GCT in both sexes in a manner similar to DHT, that is, the mitotic activity was significantly in- creased after 4 days, and then gradually decreased to the control levels by 10 days (Fig.5). By treatment with 3f-diol or testosterone for 10 days, the mitotic activity of GCT increased somewhat in castrated males (Fig.5). In castrated females, these two androgens failed to increase the mitotic activity of GCT (Fig. 5). Neither 3a- nor 3-diol had effects on the mitotic activities of other three

486 K. SAWADA AND T. NoumuRA

male

e oil

° 3a-diol

4 3£-diol

© testosterone v DHT

female

mitotic rate of GCT cells (%)

(0) 5 10 treatment periods (days)

Fic. 5. Change in the mitotic rate of the granular convoluted tubules (GCT) of the submandibular glands of neonatally-castrated male and female mice, treated with each 100 ug of 5a-androstane- 3a,17B-diol (3a-diol), Sa-androstane-38,17-diol (3f-diol), testosterone and 5a-dihydrotestosterone (DHT), and the vehicle alone for 1, 4 and 10 days from day 20. Values are mean+S.E.M., n=8-10 per group. a: P<0.001, b: P<0.002 vs oil control (Student’s ttest), c: P<0.001 vs oil control (Cochran-Cox test).

regions in the gland of either sex (data not shown).

DISCUSSION

The mouse submandibular glands contain both cytosolic and nuclear androgen receptors [15, 17, 20, 26, 27]. The cytosolic receptor in the male glands increases during postnatal development and attains adult levels by day 20 [17, 26], while circulating levels of androgens (testosterone and DHT) begin to rise on day 20 [17, 23]. In our male mice on day 30 [23], circulating DHT levels (1.0+ (0.21 ng/ml), which is produced by testes [24] and metabolized from testosterone in the peripheral tissues [1, 9] including liver, prostate, epididymis,

seminal vesicle and phallus, are approximately 4-fold over and the serum DHT/testosterone ratio (1:3 in the ratio) is 2-fold higher, compared with those of pubescent rats reported [8]. We have reported that administration of both testosterone and DHT to neonatally-castrated mice for 20 days from day 20 causes to increase all the gland weight, the ROA of GCT, the cell height of GCT, and the mitotic activity of GCT, and that DHT is more effective in comparison with testosterone [23]. In the present study, testosterone was not further converted in vitro to its derivatives, including DHT, by the submandibular glands of mice on days 20 and 30 (Table 1), suggesting that testoster- one itself acts without intraglandular conversion to DHT and other metabolites, whereas DHT used as a substrate could be converted partly to 3a- and 38-diols in amounts of less than 10% (Table 1). Daily injections of these two metabolites to neona- tally-castrated mice for 10 days from day 20 caused to increase all four parameters in the gland (Figs. 2-5). The morphometrical changes in the gland induced by 3a-diol were nearly as much as those by DHT, but both the ROA and the cell height of GCT were slightly lower by treatment with 3a-diol than by DHT after 4 days (Figs. 3 and 4). On the other hand, 3(-diol was less effective than testos- terone (Figs. 3 and 4). Thus these DHT metabo- lites produced by the gland have also androgenic effects on the gland growth, and particularly 3a- diol has a potency similar to DHT. These results indicate that, in vivo, DHT itself acts primarily, and that its metabolites, 3a-diol may act secondari- ly on the gland.

As described before [23], by treatment with DHT, the ROA and cell height of GCT in the glands of both sexes are increased for 7-20 days of treatments, whereas the mitotic activity is in- creased for 2-7 days only. Similar results were obtained by treatment with 3a-diol (Figs. 3-5). These androgens primarily affect the cell nucleus to initiate the mitosis, followed by the sequence of cellular events in GCT. In contrast, the mitotic activity was increased slightly and only in males by treatment with 36-diol or testosterone for 10 days (Fig. 5). Although ample time may be allowed for peripheral conversion of the latter androgens to active mitogenic androgens and for accumulation

Androgens and Mouse Submandibular Glands

of their effects, different effects among these four hormones can be explained solely as a consequ- ence of different binding affinities to the androgen receptor.

Metabolism of testosterone by the submandibu- lar salivary glands has been reported in some mammalian species. By the domestic boar glands, testosterone is converted in vitro to its 5a-reduced products, a large amount of DHT and 3a-diol, and lesser amounts of 3(-diol [4]. Androstenedione is produced as the main metabolite by the rat glands in vitro [2] and by the canine glands in vivo [29]. In the present study, testosterone was not further converted in vitro to its derivatives by the mouse glands, corresponding to the results from the homogenate of the adult murine gland [7]. There- fore, it is suggested that androgen pathway in the submandibular glands may be different among mammalian species.

We demonstrated that the murine glands con- tained 3a- and 3@-HSD activities (Table 2). By histochemical study, the steroid metabolizing en- zymes, including 3a- and 38-HSDs, are mainly found in the excretory striated ducts (SD) in human [25] and pig [5]. In these two species, the GCT are absent in their glands [18]. The SD of the rodent gland are transformed to the GCT during puberty [10, 11, 13, 18, 21]. Therefore, the SD in human and pig have been considered to be identi- cal with both the GCT and SD in rodents. In our mice, the ROA of GCT plus SD are superior in the males to those in the females on days 30-90 [21], but the activities of 3a- and 38-HSDs were not sexually different on day 30 (Table 2). In order to solve these inconsistent results, further research is required to examine the histochemical localization of these enzymes in the glands of our mice.

REFERENCES

1 Albert J, Geller J, Stoeltzing W, Loza D (1978) An improved method for extraction and determination of prostate concentrations of endogenous androgens. J Steroid Biochem 9: 717-720

2 Baldi A, Charreau EH (1972) 17@-Hydroxysteroid dehydrogenase activity in rat submaxillary glands. Its relation to sex and age. Endocrinology 90: 1643- 1646

3 Barka T (1980) Biologically active polypeptides in

10

11

12

13

15

16

17

487

submandibular glands. J Histochem Cytochem 28: 836-859

Booth WD (1977) Metabolism of androgens in vitro by the submaxillary salivary gland of the mature domestic boar. J Endocr 75: 145-154

Booth WD, Hay MF, Dott HM (1973) Sexual dimorphism in the submaxillary gland of the pig. J Reprod Fert 33: 163-166

Byyny RL, Orth DN, Cohen S (1972) Radioimmu- noassay of epidermal growth factor. Endocrinology 90: 1261-1266

Coffey JC (1973) Steroid metabolism by mouse submaxillary glands. I. Jn vitro metabolism of testos- terone and 4-androstene-3,17-dione. Steroids 22: 247-257

George FW, Johnson L, Wilson JD (1989) The effect of a Sa-reductase inhibitor on androgen phy- siology in the immature male rat. Endocrinology 125: 2434-2438

Goldstein JL, Wilson JD (1972) Studies on the pathogenesis of the pseudohermaphroditism in the mouse with testicular feminization. J Clin Invest 51: 1647-1658

Gresik EW (1980) Postnatal developmental changes in submandibular glands of rats and mice. J His- tochem Cytochem 28: 860-870

Gresik EW, MacRae EK (1975) The postnatal development of the sexually dimorphic duct system and of amylase activity in the submandibular glands of mice. Cell Tissue Res 157: 411-422

Gresik EW, Schenkein I, van der Noen H, Barka T (1981) Hormonal regulation of epidermal growth factor and protease in the submandibular gland of the adult mouse. Endocrinology 109: 924-929 Jayasinghe NR, Cope GH, Jacob S (1990) Mor- phometric studies on the development and sexual dimorphism of the submandibular gland of the mouse. J Anat 172: 115-127

Kasayama S, Yoshimura M, Oka T (1989) The regulation by thyroid hormones and androgen of epidermal growth factor synthesis in the subman- dibular gland and its plasma concentrations in mice. J Endocr 121: 269=275

Kyakumoto S, Kurokawa R, Ohara-Nemoto Y, Ota M. (1986) Sex difference in the cytosolic and nuclear distribution of androgen receptor in mouse submandibular gland. J Endocr 108: 267-273 Michelakis AM, Yoshida H, Menzie J, Murakami K, Inagami T (1974) A radioimmunoassay for the direct measurement of renin in mice and its applica- tion to submaxillary gland and kidney studies. En- docrinology 94: 1101-1105

Minetti CASA, Valle LBS, Fava-De-Moraes F, Romaldini JH, Oliveira-Filho RM _ (1986) Ontogenesis of androgen receptors in the mouse submandibular gland: correlation with the develop-

18

19

20

21

22

23

24

488

mental profiles of circulating thyroid and testicular hormones. Acta Endocr 112: 290-295

Pinkstaff CA (1980) The cytology of salivary glands. Int Rev Cytol 63: 141-261

Roberts ML (1974) Testosterone-induced accu- mulation of epidermal growth factor in the subman- dibular salivary glands of mice, assessed by radioim- munoassay. Biochem Pharmacol 23: 3305-3308 Sato N, Nemoto T, Baba R, Ota M (1985) Dialysis- induced transformation of mouse submandibular androgen-receptor. Biochem Int 10: 771-776 Sawada K, Noumura T (1991) Effects of castration and sex steroids on sexually dimorphic development of the mouse submandibular gland. Acta Anat 140: 97-103

Sawada K, Noumura T (1992) Sexually dimorphic duct system of the submandibular gland in mouse with testicular feminization mutation (Tfm/Y). Acta Anat 143: 241-245

Sawada K, Noumura T (1992) Differential effects of testosterone and 5a-dihydrotestosterone on growth in mouse submandibular gland. Zool Sci 9: 803-809 Sheffield JW, O’Shaughnessy PJ (1988) Testicular steroid metabolism during development in the nor-

K. SAWADA AND T.

25

26

27

28

29

NOUMURA

mal and hypogonadal mouse. J Endocr 119: 257-264 Sirigu P, Cossu M, Perra, MT, Puxeddu P. (1982) Histochemistry of the 3@-hydroxysteroid, 17/- hydroxysteroid and 3a-hydroxysteroid dehyd- rogenases in human salivary glands. Arch Oral Biol 27: 547-551

Takuma T, Nakamura T, Hosoi K, Kumegawa M. (1977) Binding protein for Sa-dihydrotestosterone in mouse submandibular gland. Biochim Biophys Acta 496: 175-181

Verhoeven G (1979) Androgen binding proteins in mouse submandibular gland. J Steroid Biochem 10: 129-138

Walker P, Weichsel ME Jr, Hoath SB, Poland RE, Fisher DA (1981) Effect of thyroxine, testosterone, and corticosterone on nerve growth factor (NGF) and epidermal growth factor (EGF) concentrations in adult female mouse submaxillary gland: dissocia- tion of NGF and EGF responses. Endocrinology 109: 582-587

Weiner AL, Ofner P, Sweeney EA (1970) Metabol- ism of testosterone-4-'*C by the canine submaxillary gland in vivo. Endocrinology 87: 406-409

ZOOLOGICAL SCIENCE 10: 489-496 (1993)

Binding of Thyroid Hormone to Freshwater Perch Leydig Cell Nuclei-rich Preparation and its Functional Relevance

Niwar R. JANA and SAMIR BHATTACHARYA!

Department of Zoology, Visva-Bharati University, Santiniketan-731 235, W. Bengal, India

ABSTRACT—Leydig cells were isolated from the testis of a freshwater perch, Anabas testudineus, belonging to the prespawning stage of reproductive cycle. Cells were sonicated and a pure nuclei preparation was obtained for binding assay. Under optimum assay conditions of pH 7.4 at 30°C and 90 min incubation binding of ['*°I] 3,5,3’-triiodothyronine (T3) to Leydig cell nuclei was saturable at 1.7 nmol/1 concentration. A Scatchard analysis of T3-binding exhibited a Kd of 26.810? mol/] and a maximum binding capacity (Bmax) aS 1.66 pmol/mg DNA. Competitive inhibition studies demon- strated that binding of T; to Leydig cell nuclei is analogue specific. Biological relevance of T3 binding to perch Leydig cell was evaluated by adding varied concentrations of T3 to Leydig cell incubation (1 x 10° cells/incubation). Twenty five ng to 100 ng of T3 resulted in a dose dependent increase in androgen release while 200 ng of T3 had no additional effect over 100 ng under present incubation system. Stimulation of androgen release by T; was significantly inhibited (p<0.01) by cycloheximide. T3 (100 ng/ml) increased protein synthesis in Leydig cell (70% as compared to control) which was significantly inhibited (p<0.01) by cycloheximide. Results indicate that binding of T3 to Leydig cell of perch testis

© 1993 Zoological Society of Japan

triggers the synthesis of a protein (or proteins) which in turn stimulates androgen release.

INTRODUCTION

Thyroid has long been implicated in the repro- duction of fish [11, 13-16, 20, 26, 29]. Annual cycles of thyroid activity in teleostean fish have been found to be correlated with the gonadal maturation [6, 8, 17]. Experimental hypothyroid- ism in teleosts effected retardation of ovarian development [23] and administration of thyroid hormone resulted in the maturation of oocytes in stellate sturgeon [10], goldfish [16] and freshwater perch [30]. All these reports clearly indicate an influence of thyroid hormone on teleostean gonad- al function, but how it does so is still unclear.

Recent reports from this laboratory demon- strated high affinity and low capacity triiodothyr- onine (T3) binding sites in the ovarian nuclei of perch [7, 22]. These information indicate a direct involvement of thyroid hormone in piscine ovarian function. However, report regarding the influence

Accepted April 17, 1993 Received August 19, 1992 ' To whom correspondence should be addressed.

of thyroid hormone on male reproductive function in fish is lacking. The aim of this paper is to show T; binding to Leydig cell nuclei isolated from the testis of a freshwater perch, Anabas testudineus and also to demonstrate that the addition of T3 to Leydig cell incubation in vitro causes significant increase in androgen release.

MATERIALS AND METHODS

Aniaml

Anabas testudineus is a freshwater perch pre- dominantly found in Eastern India. It breeds during the monsoon and its reproductive cycle can be divided into four phases [6]. Perch belonging to prespawning stage were used for the present ex- periment as thyroid hormone level in plasma re- mains high during this phase [6]. Adult male perch (25-30 g) length (10-12 cm) were acclimatized in laboratory aquaria at 28°-30°C for at least 7 days prior to experiment. They were fed ad libitum with commarcial fish food (Shalimar Fish Food, Bird and Fish food Manufacturers, Bombay, India).

490 N. R. JANA AND S. BHATTACHARYA

Testis weight varied from 200-300 mg.

Preparation of Leydig cells from perch testis

The testis of the perch is encapsulated by tunica albuginea, a thin delicate membrane which en- velopes the fibrous connective tissue. The connec- tive tissue fiber from the peripheral stroma divides the lumen of testis into a large number of semi- niferous tubules of varied shape and size. Each tubule contains germ cell at different developmen- tal stages and Sertoli cell. Germ cell can be categorised as primary and secondary spermatogo- nia, spermatocyte, spermatid, and spermatozoa. Perch testis has discrete interstitial cell or Leydig cell between the seminiferous tubules. In histolo- gical section of perch testis stained with haema- toxylin and eosin, interstitial or Leydig cells appear as a large cell with oval nuclei. Plenty of Leydig cells from perch testis could be observed under microscope. This is in contrast to other freshwater teleosts like the Indian catfish, M. vittatus and murrel, C. gachua where testis are devoid of true interstitium and characterised by possession of tubule boundary cells. In these fishes it is difficult to identify the Leydig cell. Since this perch, Anabas testudineus contains clearly defined Leydig cell in considerable number, the testes were collected by killing the perch, sliced with a surgical blade and then subjected to the prepara- tion of Leydig cell according to the method de- scribed by Yu et al. [36] and Yu and Wang [37] which is a modification of the procedure reported by Dufau ef al. [12] and Van Damme et al. [32]. Briefly small pieces of testis from perch were transferred to Erlenmyer flask containing prein- cubation medium (MEM with 25mM HEPES, 0.1% BSA, 0.1% sodium bicarbonate, pH 7.4; penicillin 1000 U/ml; streptomycin 50 ug/ml). The testis pieces were gently dispersed for 10 min with a magnetic stirrer placed on ice bath and a completely homogeneous sunpension was obtained. The cell suspension was then filtered through a fine nylon mesh and preincubated for one hr at 30°C with shaking at 50 cycles/min. This was then cooled under ice and centrifuged at 400 g for 10min at 4°C. Pelleted Leydig cells were suspended in the incubation medium which was a preincubation medium with 0.125mM _ xanthine

plus sodium heparin, 0.5% (v/v). It appears that testicular Leydig cells of perch differs from that of mammalian Leydig cells as the latter could be harvested by centrifugation at 250 g [37].

Purity of Leydig cells

To examine the purity of the Leydig cell prepa- ration, histochemical staining for 3/-hydroxy- steroid dehydrogenase (38-HSD) enzyme was per- formed by a modification of the method of Wiebe [35]. The reaction mixture contained 0.1 M Na-K phosphate buffer, pH 7.2, 0.14 ~mol 5a-androstan 38-ol-17-one, 0.6 ~mol nitroblue tetrazolium, 7 pmol NAD. 500 yl of reaction mixture was added to 1x10° cells and the suspension was incubated for one hr at 30°C. We have described earlier that 30°C was the optimum temperature for incubating the tissues from this fish [7, 8, 29]. Sa-A-3-ol-17- one was omitted from control incubation. A drop of the cell suspension was placed on a glass slide and allowed to dry. After embedding in gelatin, positively stained cells was estimated by counting a minimum of six field under the light microscope. Approximately more than 85% cells were found to be 3-hydroxysteroid dehydrogenase positive in- dicating them to be Leydig cells.

Binding incubation

T3 binding to nuclear preparations from Leydig cells was performed by following the description reported earlier from this laboratory [7]. Leydig cells were lysed by ultrasonication (Labsonic 2000, B. Braun, West Germany) at 150 KHz and nuclei were isolated according to our earlier procedure [7, 22] adopted from the description of Lawson et al. [19]. To determine the optimum binding condi- tions of [!*°I]-T3 to Leydig cell nuclear prepara- tions, incubations were performed at different temperatures pHs and time intervals. It was found that 90 min of incubation at pH 7.4 and tempera- ture 30°C were optimum conditions of radiolabelled T; binding (data not shown). 30 ug of DNA was incubated in a final volume of 500 «al with varying concentrations of ['*°I]-T3 (0.26-2.2 nmol/1) in the absence (total binding) or presence of 500 fold excess of unlabelled T3 (nonspecific binding) at pH 7.4 and 30°C for 90 min in a shaking water bath. After termination of incubations, free

T; Receptor in Leydig Cell 49]

and bound radioactivity were separated by the addition of ice cold 40% polyethylene glycol (PEG-mol. wt. 6000) and by centrifugation at 3000 gin a refrigerated centrifuge. The supernatant was aspirated and radioactivity of the pellet was meas- ured in a Gamma Counter (1282 Compugamma CS, LKB, Sweden).

Incubation of Leydig cells

For in vitro incubation, Leydig cells were sus- pended in MEM (300 yl, containing 1X 10° cells) and was added to each well of multiwell module (NUNC, Denmark). Incubations were performed at 30°C with gentle shaking under an atmosphere of oxygen. Cells were allowed to incubate for 2 hr and at 2hr hormone and other chemicals were added, incubation was then continued for another 3 hr. Hence total incubation period was 5 hr. A 2 hr preincubation time was necessary for the recov- ery of cells. Viability of cells were determined at the beginning and at the end of incubation by Trypan blue (0.1%) dye exclusion method which showed 90-80% viability of the cells. For each experiment, functional ability of isolated Leydig cells in in vitro incubation was checked by the addition of purified carp gonadotropin [2], which released androgen in the medium. T3 was added in increasing concentrations (25-200 ng/ml) to the Leydig cell incubation. T3 was dissolved by the addition of 10 ul of NaOH (1N) to 1ml of T; solution (distilled water). 10 “1 volume was fixed for each concentration of T3 which was added to 300 pl of Leydig cell incubation. Same concentra- tion and volume of NaOH was added to control incubation (without T3) as vehicle. Since MEM contained 25 mM HEPES, there was no changes of pH due to this addition. After termination of incubation, each incubate was centrifuged at 500 g, medium was collected and stored at —20°C until androgen RIA.

Protein synthesis in Leydig cells

To monitor the protein syunthesis in Leydig cells in response to T3, cells (1 10°/well) were incu- bated with ['C]-leucine (specific activity —300 mCi/mmole) in MEM supplemented with 1 mM of eighteen different amino acids except leucine. The pH was adjusted to 7.4. Protein synthesis of

Leydig cells were determined according to the procedure previously described from this labora- tory [7]. Briefly, cells were incubated in the presence of T3 (100 ng/ml) alone or T3 (100 ng/ ml) plus cycloheximide (50 ~g/ml) or in the ab- sence of either chemicals (control). The cells were then separated from the medium by centrifuga- tion, resuspended in distilled water and subjected to ultrasonication at 150 KHz. Amount of protein in the lysed preparation was determined by follow- ing the method of Lowry et al. [21] taking bovine serum albumin as the standard. The sonicated material was precipitated with a final concentra- tion of 10% TCA. The pellet obtained after centrifugation at 3000 g in a refrigerated centrifuge was first washed with 10% TCA followed by another wash with 5% TCA containing cold leucine (1 mM). 7% TCA was then added to the pellet and the sample was heated for 30 min at 95°C to denature nucleic acids. After centrifuga- tion at 3000 g and two subsequent washes with ethanol: ether (1:1), the final precipitate was dis- solved in 500 yl of 1(N) NaOH and counted in a Liquid Scintillation Counter (LSS 20, ECIL) in 10 ml of toluene based cocktail containing PPO, POPOP and methyl cellosolve and TCA precipit- able radioactivity is expressed as dpm/mg protein.

Radioimmunoassay of androgen

RIA of androgen was carried out according to the procedure described by Yu and Wang [37] which was a modification of a method reported earlier [1]. The medium collected after Leydig cell incubation was subjected to androgen RIA. It was found that androgen values were similar with or without extraction by diethylether. The antiserum used (provided by Dr. John Y. L. Yu, Academia Sinica, Taiwan, batch no. ASIZ-T-3-02) had 100% crossraction with testosterone, 74% with 5a-dihy- drotestosterone, 1.23% androstenedione, 0.59% androstenediol and very negligible crossreactions with other steroids. Sample, [°H]-testosterone and antitestosterone serum were added and the sam- ples were incubated for 18hr at 4°C. Dextran coated charcoal was used to separate bound from free steroid. After centrifugation at 3000 g, super- natant containing the bound hormone was counted in a Liquid Scientillation Counter. The sensitivity

492 N. R. JANA AND S. BHATTACHARYA

of the androgen RIA was 5.0 pg and the linear range of standard curve was 5-300 pg. The in- terassay co-efficient of variation was less than 10%.

Statistics

Data were analysed by one way analyses of variance (ANOVA). Where F value indicated significance means were compared by post hoc multiple range test. All values are expressed as mean +standard error of the mean (SEM).

RESULTS

T; binding to perch Leydig cell nuclei

Nuclear preparation from Leydig cells were in- cubated at 30°C for 90min at pH7.4 and T3 binding was one half maximal by 45 min and maximal at 90 min under these conditions. To obtain the specific binding of T3, 30 ug of DNA were incubated with increasing concentrations of ['*°I]-T3 (from 0.26 nmol/I to 2.2 nmol/l). Figure

0.05

0.04

BOUND ( pmol)

Osteo Ws 4 oes FREE (nmol/l)

Fic. 1. Saturation curve of T3 binding to Leydig cell nuclei. 30 ug of DNA from the Leydig cell nuclei were incubated with 0.26-2.2 nmol/I of ['*°I]-T3 for 60 min at 30°C. Non-specific binding (NSB) is in the presence of 500 fold excess cold T3. Specific binding (SB) is the difference between total binding (TB) and non-specific binding. Data represents the mean +SEM of six determinations.

aR Kd = 26.8 x10 mol/|,“mg DNA Bmax=1-66 pmol/mg DNA

012 010

LJ

WwW

te 0-08

3

2 0:06

(aa)

0.04

0-02

20 .40, 60 80. 100mmza BOUND (pmol/l)

Fic. 2. Scatchard analysis of the data obtained from Fig. 1. The X axis represents p mol/I concentration of bound T; and Y axis represents bound to free ratio. Kd: Dissociation constant. B,,ax: Maximum binding capacity.

1 showes that specific binding of T3 increased linearly from 0.26 nmol/! to 1.7 nmol/! and then reached a plateau indicating saturation of binding sites. Specific binding data were analysed by Scatchard plot which showed a Kd of 26.8X 107° mol/l and maximum binding capacity (Bmax) was 1.66 pmol/mg DNA (Fig. 2). Competitive binding experiments were conducted at optimal assay con- ditions with variation in T3, TRIAC (3,3’,5- triiodothyroacetic acid), T, and carp GtH (cGtH- Banerjee et al. [2]) concentrations. Results clearly indicate that T; binding sites in Leydig cell nuclei are analogue specific (Fig. 3). 3 inhibited the binding of [!*°I]-T3 very efficiently but less com- petitively than TRIAC while Ty was a poor com- petitor. cGtH did not inhibit [!*I]-T; binding. Tissue specificity of T3; binding was observed by incubating nuclear preparations of testis, liver, tail kidney and small intestine. Maximum specific binding occurred with liver nuclei followed by testis. Binding was negligible with small intestine and tail kidney nuclei (Fig. 4).

Effect of T; on androgen release from Leydig cells

Addition of increasing concentrations of T3 to

T; Receptor in Leydig Cell 493

8

75

50

25

1251-13 BOUND(e of total)

50 100 200 400 800 1600 UNLABELED HORMONE (ng)

Fic. 3. Competitive inhibition of ['*°I]-T; binding to perch Leydig cell nuclei by TRIAC, T3, T4, and cGTH. All values are mean+SEM of six deter- minations.

T3 BOUND

*fo

Sik

BS

L

Fic. 4. [!*°I]-T3 binding to nuclei obtained from differ- ent tissues. Liver (L), Testis (T), Small intestine (SI), Tail kidney (TK). All values are mean+SEM of six determinations.

Leydig cell incubation resulted in dose dependent increase of androgen release. Increase of androgen release was observed from 25 ng to 100 ng, further increase in dose had no additional effect in the present incubation system (Fig. 5).

Fic. 5.

0-4

ANDROGEN RELEASED(ng/10® cells)

50 100 150 200 CONCENTRATION OF T3 (ng/ml)

Effect of T; on androgen release by Leydig cell of perch. After a preincubation of 2 hr the cells were incubated for another 3 hr in the absence of T; (control) or in the presence (increasing concentra- tions) of T3. Data represents the mean + SEM of six determinations.

TABLE 1. Inhibition of T3-stimulated androgen re- lease by cycloheximide

Androgen released System (pe/ 1s 10° cells) Control 49.9 +6 T3 (100 ng/ml) 345* +9.8 T; (100 ng/ml)+ AQ sei

Cycloheximide (50 g/ml)

P*<0.01 as compared to control. P**<0.01 as compared to T3 added experiment. Values are mean+SEM of six determinations.

This stimulation of androgen release by T3 was inhibited by cycloheximide (Table 1).

Stimulation of Leydig cell protein synthesis by T3

Incubation of Leydig cell with [/*C]-leucine and T3 resulted in a significant (p<0.01) increase in TCA precipitable radioactivity which was cyc- loheximide sensitive (Fig. 6). T3 caused about 70% increase in protein synthesis as compared to control while cycloheximide inhibited this stimula- tion to about 53%.

494 N. R. JANA AND S. BHATTACHARYA

DPMx103/mg PROTEIN

T3+CHX

C18)

Fic. 6. Protein synthesis in the Leydig cell in reponse to T3. Incubation of ['C]-leucine into TCA precipit- able protein was determined in presence (100 ng/ ml) or in the absence of (control) of T3; or T3 (100 ng/ml) with 50 “g/ml cycloheximide (CHX). Data represents the mean+SEM of six determinations.

DISCUSSION

The present study demonstrates that nuclear T3 binding sites are present in Leydig cell of perch testis. Binding of T3 to teleostean Leydig cell nuclei is a new report and suggests an influence of thyroid hormone on testicular function. This freshwater perch breeds once a year /.e. during monsoon or rainy season and its annual reproduc- tive cycle consists of four stages [6]. We have selected prespawning stage male fish because gonadal metabolic activity is in peak during this stage. Standardisation of T3 binding assay showed maximum binding at 30°C with pH 7.4. These two characteristics clearly coincided with the natural environment of prespawning stage (April-May) when water temperature varied between 28 -32°C and pH of the water is weakly alkaline (7.3-7.5). This indicates that optimal binding conditions may have physiological relevance.

Liver is the conventional binding site for thyroid hormone and commonly used for thyroid hormone

receptor assay in teleost. The Scatchard plot analysis of our data shows that T3 binding affinity to perch Leydig cell nuclei is comparable to hepa- tic nuclei of the same fish. Ka=0.041 x10? M~ ‘in perch Leydig cell nuclei and Ka=0.06<10? M! in the hepatic nuclei of the same fish [7]. These values are closer to the affinity of hepatic nuclei of rainbow trout (Ka=0.22x10’M~—') [33] while hepatic nuclei of coho salmon has much higher affinity (Ka=1.03 10° M~') [9]. Surprisingly T3 binding affinity of perch ovarian nuclei is much higher (Ka=0.11<10°M~') as compared to Leydig cell nuclei [7]. Occupation of receptor site by T3 shows considerable variation. T3 binding capacity was considerably higher in perch ovarian nuclei (Bmax=4.312 pmol/mg DNA) [7] as com- pared to Leydig cell nuclei (Byax=1.66 pmol/mg DNA). Hepatic nuclei of perch showed five fold greater T3 binding capacity (Bmax =8.882 pmol/mg DNA) [7] in comparison to Leydig cell nuclei. However T3 binding to Leydig cell nuclei of perch is analogue and tissue specific.

Specific binding of T3 is believed to result in the initiation of at least some of the effect of thyroid hormone [24]. Oppenheimer et al. [25] detected low concentration of nuclear T3 binding sites with- in rat brain, spleen and testis, these tissues are supposed to be unresponsive to thyroid hormone. Unresponsiveness has been defined as a failure of thyroid hormone in increasing Oj consumption. But in testis with relatively low thyroid hormone binding sites a small change in O, consumption induced by thyroid hormone may not be detectable [28]. Recently T; binding sites have been detected in rat testis Sertoli cell nuclei and a role of thyroid hormone in the regulation of growth and matura- tion of Sertoli cell has been suggested [18]. A novel C-erb-A gene encodes a protein which is a receptor for thyroid hormone and this gene has also been isolated from human testis library [3, 27, 34]. The binding of T3 to perch Leydig cell nuclei is therefore not entirely unexpected but it obvious- ly raises a question whether perch testis is func- tionally responsive to thyroid hormone or not.

To answer this question perch Leydig cells were incubated in vitro. Addition of T3; to Leydig cell incubation resulted in androgen release. Increas- ing concentrations of T3; causes a dose dependent

T; Receptor in Leydig Cell

increases in androgen release indicating a biologi- cal response. Although it appears to be a very unusual effect of thyroid hormone, earlier report shows that thyroid hormone treatment may lead to a 1.5—2.5 fold increase in in vitro testicular testos- terone synthesis in rat [28]. However what is more surprising in the present investigation is the inhibi- tion of T3 stimulated androgen release from the Leydig cell by cycloheximide. This suggests in- volvement of protein synthesis in the increased release of androgen. Role of thyroid hormone in protein synthesis is well known. It has been shown that T3 specifically interacts with nuclear binding sites, triggering metabolic events which lead to the modulation of nuclear activity and results in sti- mulation of protein synthesis [4, 5, 31]. The above mentioned finding prompted us to observe T3 effect on Leydig cell protein synthesis. Results clearly showed 70% increase in protein synthesis in Leydig cell by T3 which is effectively inhibited by cycloheximide.

Our findings therefore show that a freshwater perch testicular Leydig cell nuclei contains a single species of high affinity and low capacity T3 recep- tors which are saturable and analogue specific. These T3 binding sites are possibly physiologically active as T3 stimulates androgen release from Leydig cells. This stimulation appears to be not a direct effect of T3 but mediated by a protein(s) induced by thyroid hormone. It is indeed a very interesting aspect and requires further validation in other teleosts.

ACKNOWLEDGMENTS

This work was supported by a grant from the Depart- ment of Sceince and Technology, Ministry of Science and Technology, New Delhi, India (SP/SO/CO3/88). The authors wish to express their sincere gratitude to Dr. John Y. L. Yu, Academia Sinica, Taipei, Taiwan for the generous gift of antiserum for androgen RIA.

REFERENCES

1 Anderson PH, Fukushima K, Schiller HS (1975) Radioimmunoassay of plasma testosterone with use of pollyethylene glycol to separate antibody bound and free hormones. Clin Chem 21: 708-714

2 Banerjee PP, Bhattacharya S, Nath P (1989) Purification and properties of pituitary gonadotropic

10

11

12

13

14

15

16

17

495

hormone from Indian teleosts: Freshwater murrel (Channa punctatus) and carp (Catla catla). Gen Comp Endocrinol 73: 118-128

Benbrook D, Pfatl M (1987) A novel thyroid hor- mone receptor encoded by a C-DNA clone from a human testicular library. Science 238: 788-791 Bernal J, Coleoni AH, Degroot LJ (1978) Triiodothyronine stimulation of nuclear protein synthesis. Endocrinology 102: 452-459

Bernal J, Degroot LJ (1980) Mode of action of thyroid hormones. In “The thyroid gland” Ed by M De Vesseher, Raven Press, New York, pp 123-143 Chakraborti P, Bhattacharya S (1984) Plasma thyr- oxine levels in freshwater perch: Influence of season gonadotropins and gonadal hormones. Gen Comp Endocrinol 53: 179-186

Chakraborti P, Moitra G, Bhattacharya S (1986) Binding of thyroid hormone to isolated ovarian nuclei from a freshwater perch, Anabas testudineus. Gen Comp Endocrinol 62: 239-246

Chakraborti P, Rakshit DK, Bhattacharya S (1983) Influence of season gonadotropins and gonadal hor- mones on the thyroid activity of freshwater perch Anabas testudineus (Bloch). Can J Zool 61: 359-364 Darling DS, Dickhoff WW, Gorbman A (1982) Comparison of thyroid hormone binding to hepatic nuclei of .the rat and a teleost (Oncorhynchus kisutch). Endocrinology 111: 1936-1942

Dettlaff TA, Davydova SI (1979) Differential sensi- tivity of cells of follicular epithelium and oocytes in the stellate sturgeon to unfavourable conditions and correlating influence of triiodothyronine. Gen Comp Endocrinol 39: 229-232

Dodd JM (1975) The hormones of sex and repro- duction and their effects in fish and lower chordates twenty years on. Am Zool 15 (suppl 1): 127-171 Dufau ML, Mendelson CR, Catt KJ (1974) A highly sensitive in vitro bioassay for luteinizing hormone and chorionic gonadotropin: testosterone produc- tion by dispersed Leydig cells. J Clin Endocrinol Metab 39: 610-613

Eales JG (1979) Thyroid hormones in cyclostomes and fishes. In “Hormones and Evolution Vol 1” Ed by EJW Barington, pp 341-346

Fontaine YA (1975) Hormones in fishes. In “Biochemical and Biophysical perspective in marine biology Vol 2” Ed by DC Malins and JR Sargents, Academic Press, Orlando, Fla, pp 139-212 Gorbman A (1969) Thyroid function and its control in fishes. In “Fish Physiology Vol 2” Ed by WS Hoar and DJ Randal, Academic Press, New York, Vol 2: pp 241-274

Hurlburt ME (1977) Role of thyroid gland in ova- rian maturation of the goldfish, Carassius auratus. L Can J Zool 55: 1906-1913

Ichikawa M, Mori T, Kawashima S, Ueda K, Shira-

18

19

20

21

22

23

24

25

26

27

28

496

hata S (1974) Histological changes in the thyroid and interrenal tissue of kokanee (Oncorhynchus nerka) during sexual maturation and spawning. J Fac Sci Univ Tokyo, 4: 175-182

Jannini AE, Olivieri M, Francavilla S, Gulino A, Ziparo E, D’armienoto M (1990) Ontogenesis of the nuclear 3,5,3 -triiodothyronine receptor in the rat testis. Endocrinology 126: 2551-2526

Lawson GM, Tsai MJ, Tsai SY, Minghetti PP, McClure ME, O’Malley BW (1982) In “Laboratory Methods Manual for Hormone action and Molecular Endocrinology”. Ed by H Ballin and BW O’Malley, 6th Ed Houston Biological Assn Inc, pp 5-7 Leatham J (1972) Role of Thyroid. In “Reproduc- tive Biology” Ed by H Ballin and S Glasser, Experta Medica, Amsterdam, pp 857-876

Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with folin phenol rea- gent. J Biol Chem 193: 265-275

Maitra G, Bhattacharya S (1989) Seasonal changes of Triiodothyronine binding to piscine ovarian nuc- lei. Zool Sci 6: 771-775

Matty AJ (1960) Thyroid cycles in fish. Symp Zool Soc London, 2: 1-15

Oppenheimer JH (1979) Thyroid hormone action at the cellular level. Science 203: 971-979 Oppenheimer JH, Schwartz HL, Surks MI (1974) Tissue differences in the concentration of triiodothyronine nuclear binding sites in the rat: liver, kidney, pituitary, heart, brain, spleen and testis. Endocrinology 95: 897-903.

Sage M (1973) The evolution of thyroidal function in fishes. Am Zool 13: 899-905

Sap J, Munoz A, Damm K, Goldberg Y, Ghysdael J, Leutz A, Beug H, Vennstrom B (1986) The C-erb-A protein is a high-affinity receptor for thyr- oid hormone. Nature 324: 635-640

Schneider G, Kopach K, Ohanian H, Bonnefond V, Mittler JC, Ertel NH (1979) The hypothalamic- pituitary-gonadal axis during hyperthyroidism in the

29

30

31

32

33

34

35

36

3)7/

N. R. JANA AND S. BHATTACHARYA

rat. Endocrinology 105: 674-679

Sen S, Bhattacharya S (1981) Role of thyroid and gonadotropin on the mobilization of cholesterol in a teleost, Anabas testudineus (Bloch). Indian J Exp Biol 19: 408-412

Sen S, Bhattacharya S (1982) Hormonal influence of perch ovarian 17f-hydroxysteroid dehydrogenase activity in in vitro system. Indian J Exp Biol 20: 664— 667

Tata JR, Widnell CC (1966) Ribonucleic acid syn- thesis during early action of thyroid hormones. Biochem J 98: 604-620

Van Damme MP, Robertson DM, Diczfalusy E (1974) An improved in vitro bioassay method for measuring luteinizing hormone (LH) activity using mouse Leydig cell preparations. Acta Endrocrinol 77: 655-671

Vander Kraak JG, Eales JG (1980) Saturable 3,5,3’- triiodothyronine binding sites in liver nuclei of rain- bow trout (Salmo gairdneri). Gen Comp Endocrinol 42: 437-448

Weinberger C, Thompson CC, Ong SE, Lebo R, Gruol JD, Evans MR (1986) The C-erb-A gene encodes a thyroid hormone receptor. Nature 324: 641-646

Wiebe PJ (1978) Steroidogenesis in rat Leydig cells: Effect of gonadotropin on the activity of 5-Ane and 5-Ene 3a- and 3-hydroxysteroid dehydrogenases during sextual maturation. Endocrinology 102: 765— 784

Yu JYL, Chang TY, Hsu HK, Liao CF, Wan WCM (1981) Androgen/testosterone synthesis by the dis- sociated testicular cells from mice of different ages in response to rat LH stimulation in vitro. Bull Inst Zool Acad Sin 20: 57-65

Yu JYL, Wang LM (1987) Comparative effect of diverse vertebrate gonadotropins on androgen formation in vitro from testes of roosters and mice. Biol Reprod 36: 816-824

ZOOLOGICAL SCIENCE 10: 497-504 (1993)

Developmental Changes of Testicular Gonadotropin Receptors and Serum Gonadotropin Levels in Two Strains of Mice

KoIcuHI SHIRAISHI, TIPPAWAN SINGTRIPOP, MIN KYUN PARK

and SElCcHIRO KAwASHIMA!

Zoological Institute, School of Science, University of Tokyo Tokyo 113, Japan

ABSTRACT—The relationship between testicular gonadotropin receptors and serum gonadotropin concentrations during sexual maturation was studied in two inbred mouse strains ( BALB/cAJcl and ICR/Jcl strains). The specific binding of ['*°I]iodo-FSH per unit testicular weight was the highest at 21 days of age in the BALB/cAJcl strain and 7 days in the ICR/Jcl strain, followed by a rapid decrease thereafter. The total FSH binding in the ICR/Jcl strain was significantly decreased at 35 days from the level at 28 days and remained low thereafter. However, such decrease was not observed in the BALB/ cAJcl strain. Scatchard plot analyses showed that the changes in FSH binding were due to the changes in the number of binding sites and not in the affinity of binding. The specific binding of ['**I]iodo-LH per unit testicular weight reached a peak around puberty , and the total LH binding tended to remain constant after puberty in both strains. During early testicular developmental stages, the increase in the serum FSH levels was well correlated with the number of FSH receptors. Prepubertal increase in LH binding sites occurred after the rise in the serum FSH levels. To conclude, we found a strain difference in the regulation of the number of FSH receptors, indicating that down-regulation is not unique to the

© 1993 Zoological Society of Japan

CS7BL/6NCrj strain, but that it is not universal among strains of mice.

INTRODUCTION

The action of gonadotropins on the gonad is mediated by specific gonadotropin receptors local- ized in the plasma membrane of target cells. For example, follicle-stimulating hormone (FSH) first binds to the plasma membrane receptors of Sertoli cells [2, 8] and elicits various biochemical re- sponses after stimulation of adenylyl cyclase [5, 12]. Therefore, the number of FSH receptors is one of the most important factors determining the sensitivity of Sertoli cells. Tsutsui [19] recently suggested that FSH and testosterone acted syner- gistically to induce FSH receptors during testicular development in the rat.

In immature and adult male rats, exogenous administration of large amount of FSH was effec- tive in decreasing the number of FSH receptors in Sertoli cells [4, 10]. In male CS57BL/6NCrj mice,

Accepted April 26, 1993 Received April 6, 1993 ' To whom requests of reprints should be addressed.

endogeneous FSH was effective in decreasing its own receptors after puberty, and hypophysectomy when adult induced a significant increase in the number of FSH receptors in Sertoli cells [17]. This is the discovery showing that physiological level of FSH is effective in inducing down-regulation of FSH receptors. As the mechanism for this down- regulation of FSH receptors, internalization of FSH-FSH receptor complexes was claimed by autoradiographic and kinetic studies [13-15]. These findings in the mouse point out that the mouse testis serves as a suitable model for the study of down-regulation and the relationship be- tween gonadotropin receptors and gonadotropin levels.

The present study was designed first to examine whether the down-regulation of FSH receptors found in male CS7BL/6NCrj mice is unique to this strain or general phenomenon among various strains of mice, and secondly to afford fun- damental knowledge on the developmental changes of gonadotropin receptors and serum gonadotropin concentrations in male mice of

498 K. SHiRAIsHI, T. SincTRipop et al.

strains other than CS7BL/6NCrj.

MATERIALS AND METHODS

Animals

Male mice of the BALB/cAJcl and ICR/Jcl strains maintained in this laboratory were used. They were housed in a temperature-controlled room(25+0.5'C) under daily photoperiods of 12- hr light and 12-hr dark cycles (lights on at 0600 h), and were given laboratory chow and tap water ad libitum.

Preparation of serum and receptor samples

Intact mice of various ages during testicular development were sacrificed by decapitation be- tween 1000-1200h. Trunk blood was collected and allowed to clot at room temperature for one hour. After centrifugation at 1,800 g for 20 min, serum was stored at —20°C until radioimmunoas- say (RIA) for FSH and LH. In order to secure sufficient volume of serum for assay in younger mice, serum from one to six animals was pooled. Immediately after the blood collection, the testes were taken out and weighed. The testes from one to six mice were pooled (the number of pooled testes differed according to the weight). The testes were homogenized in 0.04 M Tris-HCI! buffer (pH 7.4) containing 0.005 M MgSO, and 0.1% BSA. The concentration of the homogenates was ad- justed to contain 40 mg fresh testes per 100 wl in each age group. The homogenates were centri- fuged at 11,000 xg for 20 min at 4°C. The resulting pellets were resuspended in cold Tris-HCl buffer and adjusted to contain 20 mg original tissue/100 pil. The suspension was instantaneously frozen on dry ice-ethanol and stored at —70°C until the binding assay was performed. For the binding assay, the frozen samples were quickly thawed and diluted in cold Tris-HCI buffer. The final receptor preparations were adjusted to contain 5 mg equivalent original tissue/100 yl.

Hormone preparations

NIDDK-rat FSH(rFSH)-I-7 and NIDDK-rat LH(rLH)-I-7 were radioiodinated for the assay of FSH and LH receptors, respectively. Unlabeled

ovine FSH (oFSH) and ovine LH (oLH) (Bio- Active Chem. Lab.,Tokyo) were used to correct for non-specific binding throughout the assays of FSH and LH receptors, respectively. NIDDK-rat FSH-RP-2 and NIDDK-rat LH-RP-3 were used as reference preparations in RIA of FSH and LH, respectively. These hormone preparations were the gifts from Dr. S. Raiti the National Institute of Arthritis, Diabetes and Digestive and Kidney Dis- eases, NIH, and Dr. A. F. Parlow, Pituitary Hormone and Antisera Center, University of Maryland School of Medicine.

Binding assay

For the assays of FSH and LH receptors highly purified FSH (NIDDK-rFSH-I-7) and LH (NIDDK-rLH-I-7) were radioiodinated with !°I in the presence of lactoperoxidase and hydrogen peroxide using the method described previously [16]. The specific activities of ['*°I]iodo-FSH and [’*I]iodo-LH were 50 and 40 uCi/yug, respec- tively.

In radioreceptor assay (RRA) of FSH, receptor preparation and ['*I]iodo-FSH (1.0 ng) were incu- bated at 35°C for 2.5 hr with or without unlabeled oFSH. For RRA of LH, receptor preparation and ['*I]iodo-LH (1.0 ng) were incubated at 35°C for 2.5 hr with or without unlabeled oLH. At the end of incubation, the reaction tubes were centrifuged at 11,000 g for 20 min at 4°C and washed three times by adding 1 ml Tris-HCl buffer containing 0.1% BSA to each tube and centrifuged. Radioac- tivity of resulting pellets was counted in an auto- well gamma counter. Specific binding was calcu- lated by total binding minus non-specific binding. In Scatchard plot analysis, saturation binding ex- periments were performed, where the receptor preparation (100 ul) and labeled ligand were incu- bated with or without different amount (0.2-256 ng) of unlabeled ligand. Final volume of reaction mixture for saturation-binding experiments was adjusted to 400 ul. The equilibrium constant of dissociation (Kd) was determined with the Scatch- ard plots, constructed from the data obtained from the saturation-binding experiments.

RIA of FSH and LH Serum FSH and LH were measured by RIA

FSH and LH Receptors in Mouse Testis 499

using a double antibody method [18]. Concentra- tions of FSH and LH (both being assayed in 100 pl serum) were expressed in terms of ng NIDDK-rat FSH-RP-2 and NIDDK-rat LH-RP-3 per ml serum, respectively.

Statistical analysis Differences between testicular weights, FSH

and LH bindings, and serum FSH and LH concen- trations were analyzed by Student’s /-test.

RESULTS

Changes in testicular weight

Male mice of the BALB/cAJcl strain were sac- tificed at 14, 21, 28, 35, 42, 56 and 77 days of age. Fig. 1 shows the changes in the testicular weight. The testicular weight steadily increased from 14 to 42 days (P<0.001). After 42 days the testicular weight continued to increase much slowly (Fig. 1, upper panel).

BALB/cAuJcl

__ 200

)

E

iS

12)

2 100

(2)

g&

w

®

ld

8 14 21 28 35 42 49 56 63 70 77 Age in days 200 ICR/Jcl

fo)

E

=

a

= 100

?p)

®

D

®

i

c 7 14 21 28 35 42 49 56 63 70 77 Age in days

Fic. 1. Changes with age in the testicular weight in

BALB/cAJcl (upper panel) and ICR/Jcl mice (low- er panel). Each point represents the mean + SEM (n =4).

In ICR/Jcl strain, mice were killed at 7, 14, 21, 28, 35, 42, 56 and 77 days. The testicular weight increased progressively from 7 to 42 days (P< 0.001), and after that the weight of testes showed no significant increases (Fig. 1, lower panel).

Changes in FSH binding during testicular develop- ment

In BALB/cAJcl mice, the specific binding of ['*I]iodo-FSH /5 mg tissue equivalent was the greatest at 14 or 21 days (Fig. 2, upper panel). It rapidly decreased from 21 to 35 days, followed by a gradual decrease until 77 days of age. The specific binding of ['*°I]iodo-FSH per two testes increased rapidly from 14 to 21 days (P<0.001), and then continued to increase slowly until 42 days of age. The levels at 56 and 77 days were almost the same as the level at 42 days (Fig. 2, lower panel).

In ICR/Jcl mice, the level of specific binding of ['*I]iodo-FSH/5 mg tissue was the highest at 7 days of age. A rapid decrease was observed between 7 and 21 days of age (P<0.001). The

a FSH-R z 9 BALB/cAJcl D4 Eo £2 ex Boe a 2 2 5] Fa o 0 14 21 28 35 42 49 56 63 70 77 Age in days gs FSH-R io) BALB/cAuJcl 8 See) Ea 4 we i= = a o@2 8 0 q 14 21 28 35 42 49 56 63 70 77 Age in days Fic. 2. Changes with age in specific binding of

['?*I]iodo-rat FSH per 5 mg testicular tissue (upper panel) and specific binding of ['*I]iodo-rat FSH per two testes (lower panel) in BALB/cAJcl mice. Each point represents the mean+SEM (n=4). In- cubation of 2.5 hr at 35°C.

500 K. SHIRAISHI, T. SINGTRIPOP et al.

% FSH-R

Biyive ICR/Jcl

D

Ex

Qi’o 4

2 x

OE

5 82

2

is]

® 7 14 21 28 35 42 49 56 63 70 77

Age in days

” FSH-R

23

B ICR/Jel

ions

2% 2

2

26

6S oY

eS)

5

3 (0) SS EE EE ES EE eee 7 14 21 28 35 42 49 56 63 70 77

Age in days Fic.3. Changes with age in specific binding of

['**I]iodo-rat FSH per 5 mg testicular tissue (upper panel) and specific binding of ['?°I]iodo-rat FSH per two testes (lower panel) in ICR/Jcl mice. Each point represents the mean+SEM (n=4). Incuba- tion for 2.5 hr at 35°C.

level at 77 days was not significantly different from the level at 35 days (Fig. 3, upper panel). The specific binding of ['*°I]iodo-FSH per two testes markedly increased from 7 to 14 days (P<0.001) and from 21 to 28 days (P<0.001), followed by a rapid decrease at 35 days (P<0.05). Thereafter, it tended to show a slow decrease between 35 and 77 days of age (Fig. 3, lower panel). In contrast to the marked changes in the specific binding, the non- specific binding on unit weight basis was low and almost constant regardless of age and strain (data not shown).

In order to calculate the number of binding sites, Scatchard plot analysis of the specific binding of FSH was carried out in testicular preparations from BALB/cAJcl mice at 21 days of age and ICR/JcL mice at 7 and 56 days of age. Scatchard plots showed straight lines in all groups, indicating the presence of one kind of binding sites. Kd calculated from the fitted lines of the plots in the testis from BALB/cAJcl mice at 21 days of age was 0.15 nM and those from ICR/Jcl mice at 7 and 56 days of age were 0.15 nM and 0.12 nM, respec-

BALB/cAucl 0.3

0.2

0.1

100 200

500

0.25 0.2

0.15

B/F

0.1 >. Kd=0.12nM

0.05

=~ . > ~

100

200 B (pg)

300

Fic. 4. Scatchard plots of the binding of [!*"Iiodo-rat FSH to the receptor preparations of mice at 21 days of age in BALB/cAJcl mice (upper panel) and at 7 and 56 days of age in ICR/Jcl mice (lower panel).

tively. The number of binding sites in the testis of 21-day BALB/cAJcl mice was 2.09 fmol/mg tissue and 7- and 56-day-old ICR/Jcl mice were 2.48 fmol and 0.91 fmol/mg tissue respectively (Fig. 4).

Changes in LH binding during testicular develop- ment

In BALB/cAJcl mice, the specific binding of [!**I]iodo-LH /5mg tissue equivalent slightly in- creased until 28 days of age, followed by a rapid increase at 35 days, and thereafter the high level was kept constant (Fig.5, upper panel). Total binding per two testes continuously increased from 14 to 56 days (P<0.01) (Fig. 5, lower panel).

In ICR/Jcl mice, the specific binding of ['*°I]Jiodo-LH /5 mg tissue slightly increased from 7 to 14 days, followed by a rapid increase between 14 to 28 days of age (P<0.05, 14 days vs. 28 days) (Fig. 6, upper panel). Then, the level tended to

FSH and LH Receptors in Mouse Testis 501

@ LH-R fa BALB/cAuJcl = 3 | Sas ws 1e)) < 2 Se eae: Smal ‘SG ye) “A 144 21 28 35 42 49 56 63 70 77 Age in days LH-R 2 BALB/cAJcl 3 9 $s = SL oo 6 c x SE 683 Bo) a 0 —— —t__1__ ee ee 2 14 21 28 35 42 49 56 63 70 77

Age in days

Fic.5. Changes with age in specific binding of ['?°*I]iodo-rat LH per 5 mg testicular tissue (upper panel) and specific binding of ['"I]iodo-rat LH per two testes (lower panel) in BALB/cAJcl mice. Each point represents the mean+SEM (n=4). In- cubation for 2.5 hr at 35°C.

o LH-R B 3 ICR/Jcl oD

5% 2

=2

£ x

OE

5 81

=

(3)

@o

BO Af oanc a Sa oa SE ae Be ea 2 7 14 21 28 35 42 49 56 63 70 77

Age in days

e LH-R 3 ICR/Jel g 3 :

§ =

£2

D = 2

& x

OE

£eq

ign

a

(i)

ae)

2 7 14 21 28 35 42 49 56 63 70 77

Age in days

Fic. 6. Changes with age in specific binding of

['*°I]iodo-rat LH per mg testicular tissue (upper panel) and specific binding of ['*°I]iodo-rat LH per two testes (lower panel) in ICR/Jcl mice. Each point represents the mean+SEM (n=4). Incuba- tion for 2.5 hr at 35°C.

decrease until 77 days. The specific binding per two testes gradually increased from 7 to 21 days (P <0.01), followed by a marked increase during 21 and 28 days (P<0.05) (Fig. 6, lower panel). Dur- ing 28 and 77 days the high level was maintained.

Changes in serum FSH level in male mice

In BALB/cAJcl mice, the serum FSH level increased rapidly until 28 days of age (P<0.05, 7 days vs. 28 days). Thereafter, the serum FSH concentrations were almost constant (Fig. 7, upper panel). The period of rapid increase in the serum FSH level coincided well with the period of in- crease in the total binding shown as a broken line adopted from Fig. 2 (lower panel).

FSH-R VS. FSH

| BALB/cAJcl B = 2) ee one 2 30 26 x 42% iP 20 OE E romsy 2 40 2 2 ® ©) cS a

0 EE oe (0)

14 21 28 35 42 49 56 63 70 77 Age in days FSH-R VS. FSH

] ICRJcl 43 8 = 80 Bs £ g 2 60 4 SSS / 2 Eom x= ne) - 5 ! > S ie 40 g iI = 220; 4 2 2 a

OL 0

7 14 21 28 35 42 49 56 63 70 77 Age in days Fic. 7. Age-related changes in serum FSH levels (solid

line with redrawn profile of specific binding of ['?*I]iodo-rat FSH per two testes (broken lines) in BALB/cAJcl (upper panel) and in ICR/Jcl (lower panel). Concentrations of serum FSH are expressed as nanograms of NIDDK-rat FSH-RP-2 per ml. Each point represents the mean+SEM (n=4).

In ICR/Jcl mice, the serum FSH concentrations rapidly increased from 7 to 21 days of age (P< 0.001) (Fig. 7, lower panel). Gradual increase was observed during 21 to 42 days. After that the serum FSH level was almost constant until 77 days

502 K. SHrrRaAIsHi, T.

of age (Fig. 7, lower panel). The period of rapid increase in the serum FSH level fitted well with the period of increase in the total FSH binding (Fig. 7 lower panel). Increase in the serum FSH level was preceded by the increase in the total FSH binding at prepubertal period in ICR/Jcl strain, and the curve for the serum FSH level and that for FSH binding were roughly mirror-imaged between 21 and 35 days (Fig. 7, lower panel). Increase in the total LH binding occurred after the increase in the serum FSH at prepubertal period in both strains (Fig. 8).

LH-R VS. FSH l 10 s = 40 8° £ 3 5 2 30 26 xe 6 2x it 20 Ae 6. E ao 5 2 5 10 2 : 0 (ee ES ee em 14 21 28 35 42 49 56 63 70 77 Age in days LH-R VS. FSH

ICR/Jel

oO oo So ©

Serum FSH (ng/ml)-— aS fo} Spcific binding /two testes (cpm x 10°)> °

20 of & 0 7 14 21 28 35 42 49 56 63 70 77 Age in days Fic. 8. Age-related changes in serum FSH levels (solid

line) with redrawn profile of specific binding of ['°*I]iodo-rat LH per two testes (broken lines) in BALB/cAJcl (upper panel) and in ICR/Jcl (lower panel). Concentrations of serum FSH are expressed as nanograms of NIDDK-rat FSH-RP-2 per ml. Each point represents the mean+SEM (n=4).

Changes in serum LH level in male mice

The serum LH levels were generally variable in both strains. In BALB/cAJcl mice, the serum LH concentration was high at 14 days of age and suddenly declined at 21 days. Thereafter, it tended to increase until 56 days of age (Fig. 9, upper panel). In ICR/Jcl mice, the serum LH level steadily increased during 7 to 28 days of age, and

SINGTRIPOP et al.

LH-R VS. LH oe BALB/cAJcl 2 = ue jon E Ei gs ‘D) (0).8) 8> oes 0.6 6 aeex 8 03 ee 3 (0) —- tht —————— te wt 0 Q 14 21 28 35 42 49 56 63 70 77 2 Age in days LH-R VS. LH I ICR/Jel 3 E6 Dey 2 ae 5° Zs ge 5 2 158 ® ” L o ar 8 (0) a Se 02 7 14 21 28 35 42 49 56 63 70 77 @ Age in days Fic. 9. Age-related changes in serum LH level (solid

line) with redrawn profile of specific binding of ['°*IJiodo-rat LH per two testes (broken line) in BALB/cAJcl (upper panel) and in ICR/Jel (lower panel). Concentrations of serum LH are expressed as nanograms of NIDDK- rat LH-RP-3 per ml. Each point represents the mean+SEM (n=4).

gradually decreased from 28 to 77 days of age (P< 0.05) (Fig. 9, lower panel). In accordance with the increase in the total LH binding the serum LH tended to increase.

DISCUSSION

Tsutsui ef al. [17] have reported that down- regulation of testicular FSH receptors was a phy- siological phenomenon during sexual maturation in C57BL/6NCrj mice. In other species, such as Japanese quail [16] and photostimulated hamster [18], it is suggested that the increase in FSH levels stimulates the induction of its own receptors (up- regulation). In latter species the total number of FSH receptors is closely related to the plasma FSH levels. First aim of the present study was to examine whether the down-regulation of FSH re- ceptors is unique to male CS7BL/6NCrj mice or general phenomenon among various strains of mice. We found that the specific binding of ['°°I]iodo-FSH per unit tissue weight decreased at

FSH and LH Receptors in Mouse Testis 503

28 days of age from the level at 21 days in BALB/ cAJcl strain. However, the highest specific binding per unit tissue was observed as early as at 7 days of age in the ICR/Jcl strain. Total FSH binding per two testes was closely correlated with the increase in circulating FSH levels in both strains of mice. Scatchard plot analyses showed that the increase in FSH binding was due to the increase in the number of binding sites rather than the increase in the affinity of binding in ICR/Jcl mice. The present results were in conformity with the findings of Ketelslegers et al. [7] who reported that the early development of testicular FSH receptors was accompanied by a prominent rise in the plasma FSH. However, after the attainment of a peak in the total specific binding of ['I]iodo-FSH, it decreased significantly in ICR/Jcl mice. Whereas, significant decrease in the total specific binding was not observed in BALB/cAJcl mice. These results clearly show a difference in the two strains of mice. For the decrease from the pubertal level in the ICR/Jcl strain, down-regulatory effects of endoge- nous FSH might be a possible cause, because the serum level of FSH was much higher in the ICR/ Jcl strain than BALB/cAJcl strain. We may conclude that down-regulation seems to be not unique to the C57BL/6NCrj strain, but it is not universal among strains of mice.

The present study showed that the specific bind- ing of ['*I]iodo-LH per two testes increased until 56 days of age in BALB/cAJcl mice and 42 days in ICR/Jcl mice. After these days the high levels were maintained. These results are in accord with the findings of Pahnke et al. [11] who reported that the specific binding of [!*°I]iodo-hCG increased with age until 60 days in rats and that during this period the number of Leydig cells increased. Simi- lar results have been reported by Mori et al. [9] in mice. Mori et al. [9] showed that total LH binding markedly increased during 19 to 40 days, when the number of Leydig cells rapidly increased.

Ketelslegers et al. [7] found that the develop- ment of testicular LH receptors coincided with the phase of rise in the plasma FSH. In the present study, the specific binding of [!**I]iodo-LH per two testes increased rapidly in the prepubertal period in male mice in two strains, accompanied by an increase in the serum FSH levels. This finding

indicates that the serum FSH is an important factor for the induction of LH receptors. In contrast to the serum FSH, the level of serum LH failed to show a good correlation with the number of LH receptors. In this conjunction, stimulatory effects of Sertoli cells to Leydig cell function have been well documented [1, 3, 6]. However, factors secreted from Sertoli cells by the stimulation of FSH and inducing LH receptors in Leydig cells have not yet been identified.

ACKNOWLEDGMENTS

This study was supported by a Grant-in-Aid for Scien- tific Research from the Ministry of Education, Science and Culture, Japan and a Research Grant from Zenyaku Kogyo Ltd. to S.K. The authors are grateful to Dr. S. Raiti and Pituitary Hormone Distribution Program, National Institute of Arthritis, Diabetes, and Digestive and Kidney Diseases (NIDDK) for the supply of rat gonadotropins, and to Dr. A. F. Parlow, Pituitary Hormone/ Antisera Center, University of Maryland School of Medicine for the supply of RIA kits.

REFERENCES

1 Benahmed M, Reventos J, Tabaone E, Saez J M (1985) Cultured Sertoli cell-mediated FSH stimula- tory effect on Leydig cell steroidogenesis. Am J Physiol 248: E176-E181

2 Castro AE, Alonso A, Mancini RE (1972) Localiza- tion of follicle-stimulating and luteinizing hormones in the rats testis using immunohistological tests. J Endocrinol 52: 129-136

3 Chen YD, Payne AM, Kelch RP (1976) FSH stimulation of Leydig cell function in the hypophysectomized immature rat. Proc Soc Exp Biol Med 153: 473-475

4 Francis GL, Brown TJ, Bercu BB (1981) Control of Sertoli cell response to FSH: Regulation by homolo- gous hormone exposure. Biol Reprod 24: 955-961

5 Fritz IB (1978) Sites of action of androgens and follicle stimulating hormone on cells of the semi- niferous tubule. In “Biochemical Actions of Hor- mones Vol 5” Ed by G Litwack, Academic Press, New York, pp 249-281

6 Janecki A, Jakubowiak A, Lukaszyk, A (1985) Stimulatory effect of Sertoli cell secretory products on testosterone secretion by purified Leydig cells in primary culture. Mol Cell Endocrinol 42: 235-243

7 Ketelslegers JM, Hetzel WD, Sherins RJ, Catt KJ (1977) Developmental changes in testicular gona- dotropin receptor: plasma gonadotropins and plas- ma testosterone in the rat. Endocrinology 103: 212-

10

11

504

222

Mancini RE, Castro AE, Seiguer AC (1976) Histo- logic localization of follicle-stimulating hormone in the rat testis. J Histochem Cytochem 15: 516-525 Mori H, Tsutsui K, Kawashima S (1985) Develop- mental changes in testicular luteinizing hormone receptors and Leydig cell number in mice with reference to change in plasma gonadotropins and testosterone. J Sci Hiroshima Ser B, Div 1, 32: 143- 156

O’Shaughnessy P J, Brown P S (1978) Reduction in FSH receptors in the rat testis by injection of homologous hormones. Mol Cell Endocrinol 12: 9- 15

Pahnke VG, Leinberger FA, Kunzig HJ (1975) Correlation between hCG (LH)-binding capacity, Leydig cell number and secretory activity of rat testis throughout pubescence. Acta Endocrinol 79: 610-618

Risbridger GP, Hodgson YM, de Kretser DM (1981) Mechanism of action of gonadotropins on the testis. In “Testis” Ed by H Burger and D. de Kretser, Raven Press, New York, pp 195-212 Shimizu A, Tsutsui K, Kawashima S (1987) Auto- radiographic study of binding and internalization of follicle-stimulating hormone in the mouse testis minces in vitro. Endocrinol Japon 34: 431-442

Shimizu A, Kawashima S (1989a) Kinetic study of

15

16

17

18

19

K. SHIRAISHI, T. SincTRIPOP et al.

inter-nalization and degradation of '*'I-labeled folli- cle-stimulating hormone in mouse Sertoli cells and its relevance to other systems. J Biol Chem 264: 13632-13638

Shimizu A, Kawashima S (1989b) Derivation and application of mathematic model for kinetics of '5!]-follicle-stimulating hormone bound to recep- tors. J Biol Chem 264: 13639-13641.

Tsutsui K, Ishii S (1978) Effects of follicle- stimulating hormone and testosterone on receptors of follicle-stimulating hormone in the testis of the immature Japanese quail. Gen Comp Endocrinol 36: 297-305

Tsutsui K, Shimizu A, Kawamoto K, Kawashima S (1985) Developmental changes in the binding of FSH to the testicular preparations of mice and the effects of hypophysectomy and administration of FSH on the binding. Endocrinology 117: 2534-2543 Tsutsui K, Kawashima S, Masuda A, Oishi T (1988) Effects of photoperiod and temperature on the binding of follicle-stimulating hormone (FSH) to testicular preparations and plasma FSH concentra- tion in the Djungarian hamster, Phodopus sungorus. Endocrinology 122: 1094-1102.

Tsutsui K (1991) Pituitary and gonadal hormone- dependent and -independent induction of follicle- stimulating hormone receptors in the developing testis. Endocrinology 128: 477-487

ZOOLOGICAL SCIENCE 10: 505-509 (1993)

© 1993 Zoological Society of Japan

The Behavioral Reactions of a Snake and a Turtle to Abrupt Decreases in Gravity

RicHarD WasseErsuG! and AKeEmI IzuMI-KUROTANI“

‘Department of Anatomy and Neurobiology, Faculty of Medicine, Sir Charles Tupper Medical Building, Dalhousie University, Halifax, Nova Scotia, B3H 4H7, Canada, and "Space Utilization Research Center, Institute

of Space and Astronautical Science, 3-1-1, Yoshino-dai Sagamihara, Kanagawa 229, Japan

ABSTRACT—We report here on the behavioral reaction of two reptiles to abrupt decreases in gravity. One striped rat snake, Elaphe quadrivirgata, and three striped-neck pond turtles, Mauremys japonica, were exposed to microgravity (u-G) on parabolic flight, during the filming of a documentary for the NHK television station in Japan. The video films revealed that the snake reflexively responded to the shift from hyper- to hypogravity by taking up a defensive posture—on the first parabola, the snake struck at itself. The turtles actively extended their limbs and hyper-extended their neck in p-G, a posture which is identical to that displayed during their contact “righting reflex”, when placed upside-down in normal gravity. The aggressive display of the snake was unexpected, although the righting response of the turtles was consistent with that shown by other vertebrates, including fish and mammals, exposed to u-G. An implication of these observations is that the afferent signal for the righting reflex of vertebrates in normal gravity must be the unloading of ventral receptors in the sensory system, rather than the loading of dorsal receptors. These are the first behavioral records for any

reptiles exposed to hypogravity.

INTRODUCTION

In the nearly half century that man has been exploring space, a plethora of organisms have been exposed to microgravity (4-G) either briefly, on parabolic flights, or for longer periods, on orbital missions. Several birds, reptiles, amphibians and fishes, as well as a large variety of mammals have all been exposed to z-G [2]. Most of this work has focused on the physiological consequences of re- duced gravity. Surprisingly little attention has been given to the reflexive behaviors of these organisms upon their initial exposure to micro- gravity. It is not known, for example, whether there are consistent patterns in the reflexive re- sponses of animals to abrupt changes from hyper- to hypo-G. Do the responses of various verte- brates correlate with their phylogenetic rela- tionships or way of life? Can we predict how any

Accepted March 16, 1993 Received October 28, 1992

animal will react, for example, in freefall?

Early work with teleost fishes [e.g. 5, 6] suggests that aquatic vertebrates reflexively pitch down- ward in «-G and that this leads to looping (forward somersaults). More recent studies, however, with a larger variety of aquatic species (Wassersug, unpublished data) suggest that such forward loop- ing is not characteristic of all lower vertebrates exposed to 4-G. Some amphibian larvae, for example, make forward somersaults (e.g. Xeno- pus), whereas others make backward somersaults (e.g. Bufo). Others (e.g. Rana) float freely and still others twist and roll along their long axis while swimming forward in short dashes [3, 7, 9, 10]. Clearly, more animals will have to be examined before systematic patterns in the behavioral re- sponses of vertebrates to decreased gravity can be revealed. Ultimately, if behavioral research in gravitation biology is to become a predictive sci- ence, such patterns should exist and be identi- fiable.

As a step towards producing a broader data base on the behavioral responses of vertebrates to de-

506 R. WASSERSUG AND A. IZUMI-KUROTANI

creased gravity, we report here on the behavior of a snake and three turtles exposed to changing gravity in parabolic flight. The behavior was videotaped in the early spring of 1992 by the Japanese television station NHK, in preparation for their April 6th, 1992 television program “Kura- bete Mireba” (Comparative Ethology for the General Public). Although the film sequences discussed here were not used in the final produc- tion of the documentary, they were graciously made available to us for study.

To the best of our knowledge, there are no other records of the behavior of reptiles in parabolic flight. The great morphological difference be- tween a snake and a turtle and their dramatically different responses, as noted below, emphasize how different the responses of organisms to de- creased gravity can be.

MATERIALS AND METHODS

The snake used in this experiment was a single specimen of the common striped rat snake of Japan, Elaphe quadrivirgata. This an active, semi-arboreal snake. The turtles used were speci- mens of the Japanese striped-neck pond turtle Mauremys [Clemmys] japonica. All specimens were on loan to NHK from a local pet store in Nagoya, Japan, and were returned to that store at the end of the mission. Species identifications were confirmed by the supplier, but regrettably no other information on the specimens was recorded. Judging from the size of the containers holding the animals during the flight, we estimated that the snake was 75cm long. The carapace lengths for the three turtles were similarly estimated at 6-8 cm, 10-12cm and 13-15 cm respectively. All animals were active and in clearly good health.

The behavior of an additional turtle of the same species and size of the largest turtles exposed to p-G was examined in the laboratory at the Biology Department of Shimane University, Matsue, Japan.

The snake was flown alone; the three turtles were flown together. For flight the animals were housed in a simple, smooth-sided aquarium with- out water. The maximum and minimum dimen- sions of their containers were on the order of half a

meter to a third of a meter respectively. The aquarium was braced in the pressurized, tempera- ture-controlled cabin of a Mitsubishi MU 300 aircraft. A fix-mounted Sony Hi 8 video camera continuously filmed the animals throughout their parabolic flights. The snake and turtles were each exposed to eight parabolas. The parabolic ma- neuvers were performed between 6.6 and 8.5 km elevation and provided in excess of 15-18 sec of uw-G (i.e. G<10~%), with intervening hyper-G episodes of approximately 2-G.

RESULTS

Snake

Before the first parabola, the snake was actively exploring its cage. The animal had climbed the vertical edge with the front half of its body and was probing an upper corner with its head, as if searching for a crack or crevice to enter. As the aircraft entered »-G on the first parabola, the snake immediately retracted its head into three tight curves. The animal thrashed about violently, twisting and rolling. Three seconds into u-G, a mid-portion of the snake’s body came within approximately 10 cm of its head. At that moment the snake struck at itself (Fig. 1). As soon as the head hit the body, it was withdrawn. The action was so fast that we could not confirm from the video images (only alternate ones of which are shown in Fig. 1) whether the mouth was open during this snout-body contact or whether teeth were planted in the skin.

For the remainder of the parabola the snake’s head stayed cocked, as if prepared to strike again. At the initiation of reduced gravity on each subse- quent parabola, the snake immediately cocked its head into the pre-strike posture just described. Additional strikes, however, were not observed. On the 3rd, 4th and 6th parabolas, the animal successfully braced a portion of its body between the top and bottom of its aquarium and stayed in contact with the container throughout «-G. This bracing eliminated most of the chaotic twisting and rolling seen in the earlier parabolas.

The snake’s response to the hyper-G phase of the parabolic trajectory before its first and subse-

Reptiles in Microgravity 507

quent exposure to 4-G was, in contrast, subdued. Whenever G increased from <1 to >1 the snake was, understandably, forced toward the bottom of its container. When the animal had a portion of its

aah

body elevated during hyper-G, it could be seen straining to maintain its posture. However, at no time in hyper-G did it respond either hyperactively or aggressively, as it did during the shift from hyper-G to hypo-G on the first parabola.

Turtles

The reaction of the turtles to reduced G, in contrast to that of the snake’s, was slower and more subdued. At the onset of u-G, the turtles either stayed in contact with a substrate or were thrown between surfaces due to fluctuation in their acceleration relative to that of the aircraft. The turtles had no control over their trajectory in 4-G and tumbled until they hit a wall or one another. They did not withdraw into their shells. On the contrary, they extended their necks and raised their heads while at the same time fully extending and elevating their limbs (Fig. 2). The digits, most notably on the hind limbs, were spread while the foot was rotated along the long axis of the limb,

Fic. 2. A video image of a pond turtle, Mauremys japonica, in microgravity. Note the hyper-extension of the neck and the asymmetric dorsal deviation of the extended limbs. The same hyper-extension of head and limbs is used by M. japonica in normal gravity to generate torque through its long axis and thus to right itself when upside-down by rolling over. The video image has been computer enhanced as in Fig. 1.

Fic. 1. Four sequential video frames of the rat snake, Elaphe quadrivirgata, in microgravity. In the first frame (a) the snake’s head is in the middle of the field. In the subsequent two frames (b, c), the snake’s snout strikes its body then descends to the bottom left of the field (d). The frames are sepa- rated by 20msec. The video images have been computer enhanced to remove extraneous reflec- tions from the container’s walls.

508 R. WASSERSUG AND A. IZUMI-KUROTANI

bringing the anterior, leading edge upward. The impression created was one of the animal trying to reach or contact, with its head and limbs, objects dorsal to it. The limbs on both sides were not extended or elevated equally; that is, the maneu- ver was not performed symmetrically.

The dorsal extension just described was exhi- bited by a turtle during normal gravity in the aircraft when it “landed” upside-down. The pos- ture was used by the turtle specifically to right itself. If was completely duplicated by M. japonica in the laboratory at Shimane University by simply putting the turtle upside-down on a styrofoam sheet. In that position, the turtle was able to contact the surface with its extended snout and claws and hold itself in place, such that it could quickly flip over, i.e. “roll” in aviation parlance. On the smooth surface of the flight container, the maneuver was futile except in one case involving the smallest turtle.

The turtles showed little response to either the initial or subsequent episodes of hyper-G. If they were free floating when G shifted from <1 to >1, they were propelled toward the bottom of their aquarium. They did not, however, retreat into their shells or exhibit other signs of undue stress in hyper-G. No progressive changes in the behavior of the turtles were seen as they went through the multiple parabolas.

DISCUSSION

The behavior of reptiles in decreasing G can be readily understood as a manifestation of behaviors that they show in 1-G under special circumstances. There is nothing in their behavioral repertoire during parabolic flight that can be said to be either G or “space specific.” The behavior that the snake exhibited was, however, quite different from that of the turtle.

The cocked neck and strike shown by E. quadri- virgata as G decreased was clearly of a defensive sort. The writhing and twisting, particularly in early parabolas, is indistinguishable from that of the normal frightened and threatened E. quadri- virgata. The fact that the snake actually struck at itself suggests that the animal suffered loss of proprioceptive clues about body position in “-G.

On the other hand, Elaphe snakes are naturally preyed upon by hawks in Japan and they occa- sionally experience aerial suspension when cap- tured. In this situation, it is in the snake’s best interest to strike, even to strike blindly! Such a response may have been demonstrated by E. quadrivirgata during the shift from hyper- to hypo- G.

Snakes vary greatly in their response to tilting in 1-G, depending on whether they are terrestrial or arboreal [4]. In this regard, it would be particular- ly interesting to see whether other snakes that are more or less arboreal than Elaphe show the same defensive response in #-G and parabolic flight.

The turtles responded to »-G as if they were upside-down [1]. This has implications to under- standing the stimulus for their righting reflex in 1-G. Specifically, the afferent signal for the hyper- extension of the neck and limbs used in the right- ing maneuver cannot be the loading of dorsal sensory receptors. Rather it must be unloading of ventral receptors.

Hyper-extension of the axial skeleton with asymmetrical limb movements has been observed in other vertebrates in ~-G. The combined limb and axial movements lead to long-axis rotation, or rolling. In 1-G this appears to be an important part of a righting reflex best known in arboreal and semi-arboreal vertebrates, such as the cat. The same maneuver is seen, however, in a variety of less arboreal animals in «-G. These range from aquatic animals, such as the adult clawed frog (Xenopus) and salamander larvae [3, 8], to strictly terrestrial mammals, such as the rabbit (on the NHK video film, unpublished data). What distin- guishes these behaviors in s«#-G from the 1-G situation is that they are more repetitive and protracted in freefall, simply because they are not effective. The ability to duplicate these displays in turtles and many other animals by simply inverting them supports the conclusion that they are man- ifestations of a natural righting reflex.

ACKNOWLEDGMENTS

We thank Dr. M. Yamashita and NHK television for facilitating access to videofilms of animals in microgray- ity. Laboratory space at Shimane University was gra- ciously provided by Dr. T. Naitoh. Drs. R. Marlow and

Reptiles in Microgravity 509

B. Young helped us access relevant literature. Mr. S. Pronych and Ms. M. Fejtek prepared the illustrations. Dr. H. Lillywhite, Mr. S. Pronych and Ms. M. Fejtek reviewed drafts of the manuscript. This work was supported by Canada’s Space Agency and its Natural Science and Engineering Research Council, as well as a grant from the Japan Science and Technology Fund.

REFERENCES

1 Ashe V (1970) The righting reflex in turtles: a description and comparison. Psychon Sci 20: 150- 152

2 Ballard RW, Mains RC (1990) Animal experiments in space: a brief overview. In “Fundamentals of Space Biology” Ed by M Asashima and G Malacin- ski, Japan Scientific Societies Press, pp 21-41

3 Izumi-Kurotani A, Wassersug RJ, Yamashita M, Naitoh T, Nagaoka S (1992) Frog behavior under microgravity. 9th ISAS Space Utilization Symp, pp 112-114

4 Lillywhite H (1987) Circulatory adaptations of snakes to gravity. Amer Zool 27: 81-95

5 Scheld HW, Baky A, Boyd JF, Eichler VB, Fuller PM, Hoffman RB, Keefe JR, Kuchnow KP,

10

Oppenheimer JM, Salinas GA, von Baumgarten RJ (1976) Killifish hatching and orientation (Experi- ment MA-161), Apollo-Soyuz Test Project Pre- liminary Science Report. In “NASA Technical Memorandum X-58173” Washington DC, pp 281- 305

Von Baumgarten RJ, Baldrighi G, Shillinger Jr GL (1972) Vestibular behavior of fish during diminished G-force and weightlessness. Aerospace Med 43: 626-631

Wassersug RJ (1992) The basic mechanics of ascent and descent by anuran larvae (Xenopus laevis). Copeia 1992: 890-894

Wassersug RJ, Izumi-Kurotani A, Yamashita M, Naitoh T (1993) Motion sickness in amphibians. Behav Neural Biol: in press

Wassersug RJ, Pronych SP, Schofield SM, Schmidt GK (1991) On the way of life of aquatic vertebrates and their behavioral responses to reduced gravity. ASGSB Bull 5: 42

Wassersug RJ, Pronych SP, Souza KA (1991) Studying the effects of microgravity on lower verte- brate development and behavior: a progress report. Proceedings Spacebound 91, CSA Publication No. 5891, pp 146-150

= - . ¥ = ia yh | a-. —- V. ed i L - fal to) 7 ae Pe Vinegonme val al aE A eT AN SS ihe ery) , i ; =. I Ten Ed aly CRED MNS ford! Th cheeeangys > Arie At) : , Ha z P mu lim gee he TERRA) (i (eve ET si pe T

LAR andl rneseny

T yiytptcgl nor Ay Bed tg ie ie | oe

. anette ® >* % \ 7 ut ek Ae rail ari w : oe CUO erga mo ji» ihe ; Ari wdinliw » i thé IMPLI Gew wh wel > ‘ La 9 + { j i isn 7 12 TT i a : ui div ort Vex ’ | oer ih 4s fist ii wA tré * t Ay

ibe lid

om

herr Yay ,

dvsjoit: Ming

CH, HN, A igen a

lewin¥ i han ep ey i te Thaw be ikea

: ‘baie’ (PN CE ball See PSGAPT Pw ATA an the bt fr

ee - penah -<ors en A

& Spoictey Bi ease nly Ai (at weal rll H “ny 00 4a sarong tne} :) tis ths) LT Brit? ‘Beiclald TP bai eta (FE 4 easiest a Ve ictrerr RH, : v" ralea wl werkerk reer: rovy Ny oT apenas) seu

1 See L.) PA wal a rn edu id ar | aly os mien wv were N a Cat wae aa ste"

or (nh ‘ ' % i Ca] ‘ { TP « 4 £8 i \

ZOOLOGICAL SCIENCE 10: 511-519 (1993)

Differences in the Responses of Two Mudskippers to Terrestrial Exposure

YuEN K. Ip, Cuee Y. Lee, Suit F. Cuew', Wai P. Low and KAH W. PENG

Department of Zoology, National University of Singapore, 10 Kent Ridge Crescent, Singapore 0511, Republic of Singapore

ABSTRACT—The concentration of NH,* in the plasma and muscles of Periophthalmodon schlosseri and Boleophthalmus boddaerti increased significantly after exposure to terrestrial conditions for 24 hr. Such increases in NH,‘ concentrations in B. boddaerti were not accompanied by any significant increase in urea concentrations. However, significant increases in urea concentrations occurred in the plasma, liver and muscle of P. schlosseri kept away from water. Results obtained indicate that the liver may be a major organ involved in urea production in this mudskipper. In addition, there were significant increases in the concentrations of total free amino acids in the plasma, liver and muscle of P. schlosseri, but not in those of B. boddaerti, after 24 hr of terrestrial exposure. Terrestrial exposure also affected the aminating and deaminating activities of glutamate dehydrogenase from the liver of P. schlosseri, \eading to a significant increase in the aminating/deaminating ratio. It is concluded that P. schlosseri, having developed a greater affinity to land than B. boddaerti, has also acquired a greater

© 1993 Zoological Society of Japan

capacity to detoxify ammonia.

INTRODUCTION

Mudskippers are gobioid teleosts usually found in mangrove swamps in the estuaries of rivers. They are amphibious and spend a substantial part of their lives out of water. In Singapore, Boleoph- thalmus boddaerti and Periophthalmodon schlos- seri inhabit mud flats which are periodically inun- dated by the tide. The former makes burrows on the lower regions of the intertidal zone while the latter burrows on higher ground. At low tide, both are found on the mudflats. At high tide, B. boddaerti stays in its water-filled burrows and resurfaces only when the tide ebbs while P. schlos- seri often swims with its snout and eyes above water along the water’s edge.

Recent studies by low et al. [17, 18] reveal that the gills of P. schlosseri are better adapted to a terrestrial than an aquatic environment. It has

Accepted March 29, 1993 Received December 7, 1992

1 Present address: Biology Division, School of Science, National Institute of Education, Nanyang Technologic- al University, 469 Bukit Timah Road, Singapore 1025, Republic of Singapore

relatively fewer and shorter gill filaments, and its gills exhibit intrafilamentary secondary lamellar fusions which reduce coalescence of the respira- tory surfaces upon terrestrial exposure. However, such fusions would render branchial gaseous ex- change in water via the counter-current mechan- ism inefficient. In contrast, the gills of B. boddaer- ti are better adapted for aquatic respiration as they exhibit relatively longer gill filaments, and most of their secondary lamellae are aligned to the respira- tory water current.

In general, the nature of the major nitrogenous end-product of a species is correlated with that species’s environment: aquatic species are ammo- notelic, whereas terrestrial species are either ureo- telic or purinotelic [3, 21]. Gordon et al. [9] first demonstrated that the Madagascan mudskipper, Periophthalmus sobrinus, might have a potential for the transition to ureotelism while it was out of water. Similar results were obtained for the Chinese mudskipper, Periophthalmus cantonensis [10]. However, Morii et al. [22, 23] and Morii [21] showed that NH,* was mainly accumulated, and its conversion to urea was hardly performed, in the bodies of P. cantonensis and Boleophthalmus pecti-

$12 Y. K. Ip, C. Y. LEE et al.

nirostris during the period out of water. Iwata et al. [15] obtained results, similar to those of Morii [21] and Morii et al. [22, 23], on P. cantonensis in a separate study.

To date, whether mudskippers have a potential for the transition towards ureotelism during a terrestrial excursion is still disputable. This is mainly due to the lack of knowledge on their differences in terrestrial affinities. It would appear that P. schlosseri is a more important candidate to be examined in this regard because its gills show the highest degree of adaptation, amongst the mudskippers, to respire terrestrially. Since no such information on this mudskipper is available, the present study was undertaken to elucidate the mechanisms by which ammonia was detoxified in P. schlosseri when exposed to a terrestrial environ- ment. Experiments were also performed on the relatively more aquatic species, B. boddaerti, for comparison. During the course of the study, it was observed that, different from B. boddaerti, P. schlosseri accumulated in its tissues not only urea and NH,"*, but also significantly higher concentra- tions of total free amino acids (TFAA) after 24 hr of terrestrial exposure. Therefore, the possible involvement of glutamate dehydrogenase (GDH) in ammonia detoxification in the latter mudskipper was also examined.

MATERIALS AND METHODS

Collection and maintenance of mudskippers

P. schlosseri and B. boddaerti were collected along the estuarine canal at Pasir Ris, Singapore. They were maintained in 50% (15%o salinity) sea- water (SW) at 25°C in the laboratory and the SW was changed daily. The aquaria were tilted slight- ly, so that the fish were free to be in or out of water. No attempt was made to separate the sexes. P. schlosseri and B. boddaerti were fed small guppies and a manufactured product (Goldfish and Staple Flake, Everyday Co., Singapore), respec- tively.

Exposure of mudskippers to experimental condi- tions

Fish were exposed to terrestrial conditions at

25°C for 24 hr in aquaria lined with one layer of cotton wool and two layers of Whatman no. 1 filter paper steeped with 50% SW at the bottom. After 24 hr, fish were anaesthetized for 10 min in an atmosphere saturated with diethylether.

Fish submerged in aerated 50% SW at 25°C for 24 hr were used as controls for comparison. After 24 hr in the SW, fish were anaesthetized by the introduction of 3-aminobenzoic acid ethyl ester (MS222) at a final concentration of 0.005%.

Sample preparation for the analyses of NH4*, urea and free amino acids (FAA)

Anaesthetized fish were killed immediately by pithing. The lateral muscle and the liver were quickly excised. No attempt was made to separate red and white muscles. The excised tissues and organs were immediately freeze-clamped in liquid N> with precooled tongs [6]. The whole procedure was completed within 30s. Frozen samples were kept at —80°C until analysis.

The frozen samples was weighed, ground to a powder under liquid N> and placed in either 5 vol (w/v) (for urea and NH,~ analyses) or 15 vol (w/ v) (for FAA analyses) of ice-cold 6% trichloro- acetic acid (TCA). The sample was homogenized thrice using an Ultra-Turrax homogenizer (Janke and Kunkel Gmbh & Co., Germany) at maximum speed for 20 sec each with 10 sec off intervals. The sample was then centrifuged at 10,000xg at 4°C for 10min using a Kokusan H251 refrigerated centrifuge (Kokusan Enshinki Co., Japan). The supernatant fluid obtained was analyzed for NH,*, urea and FAA.

A separate group of fish exposed to similar conditions was used for the collection of blood samples. The caudal peduncle of the anaesthetized fish was severed and blood exuding from the caudal artery was collected in heparinized capillary tubes. The tubes were centrifuged at 4,000X¢g at 4°C for 10 min to obtain the plasma. The collected plasma was deproteinized in 2 vol (v/v) of ice-cold 6% TCA and centrifuged at 10,000Xg at 4°C for 15 min. The resulting supernatant fluid was kept at —80°C until analysis.

Determinations of NH4*, urea and FAA

For the analysis of NH4*, the pH of the depro-

Terrestrial Adaptations in Mudskippers 513

teinized sample was adjusted to 5.5—6.0 with 5 M KHCO;. The NH,*t content was determined according to the method of Kun and Kearney [16]. The reaction medium, in a total of 2.7 ml, con- tained 100mM Tris-HCl (pH8.0), 10mM a- ketoglutarate (ccKG), 0.19 mM NADH, 24.3 1U GDH (Sigma Chemical Co., MO) and 0.2 ml sample. Freshly prepared NH,Cl solution was used as the standard for comparison.

For the determination of urea, the pH of the sample was neutralized with 5M K;CO3. To 0.2 ml of this neutralized sample, analysis was per- formed using a Sigma Urea Assay Kit Procedure 535 (Sigma Chemical Co., MO). To another 0.2 ml of the same sample, similar analysis was per- formed after incubating for 15 min at 30°C with 0.2 ml of 20 mM imidazole buffer (pH 7.0) containing 2 IU urease (Sigma Chemical Co, MO). The difference in absorbance obtained from samples with and without urease treatment was used for the estimation of the urea concentration in the sample. Urea obtained from Sigma Chemical Co. (USA) was used as a standard for comparison.

For the analysis of FAA, the sample was ad- justed to pH2.2 with 4M LiOH and diluted appropriately with 0.2 M lithium citrate buffer (pH 2.2). FAA were analyzed using a Shimadzu LC- 6A Amino Acid Analysis System with a Shim-pack ISC-07/S1504 Li type column. Since taurine was found to be present in much greater concentrations than other FAA in the various tissues and organs of these two mudskippers (except in the muscle of B. boddaerti) and since terrestrial exposure had no significant effect on its concentrations, the authors decided not to take it into account in the calcula- tion for the concentration of TFAA to better reflect the overall changes in the concentrations of other FAA.

Results were expressed as «M/g for muscles and livers, and ~M/ml for plasma.

Preparation of samples for enzymatic analyses

Samples for GDH assays were prepared accord- ing to Chew and Ip [4] with some modifications. The excised muscle was homogenized in 5 vol (w/ v) of an ice-cold buffer, which contained 300 mM sucrose, 0.1 mM EDTA and 3 mM Tris-HCl (pH 7.4), using an Ultra-Turrax homogenizer at mini-

mum speed for 10sec. The liver sample was homogenized three strokes in 10 vol (w/v) of the same buffer using a teflon-glass homogenizer. The | homogenized sample was centrifuged at 600 x g for 15 min at 4°C in a Beckman J2-21 M/E refriger- ated centrifuge (Beckman Instruments, USA) to remove any unbroken cells and nuclei. The super- natant fluid obtained was further centrifuged for 15 min at 10,000 x g to obtain the mitochondria. The mitochondria were washed twice by resuspension and resedimentation in the same buffer and soni- cated before assaying of GDH activity.

Enzyme assays

Enzyme assays were performed by monitoring changes of absorbance at 25°C using a Shimadzu UV-160A spectrophotometer.

Activity of GDH in the aminating direction was determined according to the method of Iwata et al. [15] with some modifications. The reaction mix- ture in a total volume of 2.8 ml contained 200 mM triethanolamine-HCl (pH7.8), 0.1mM NADH, 0.12 mM ADP, 10 mM @KG and 250 mM ammo- nium acetate. A control assay was conducted without aKG. The oxidation of NADH was moni- tored at 340 nm. Specific activities were expressed as uymol NADH oxidized/mg protein per min.

Activity of GDH in the deaminating direction was assayed using a modified colorimetric method of Beutler and Michal [2]. The absorbance change was monitored at 492 nm. The reaction medium, in a total volume of 1.3 ml, comprised 200 mM glycine-NaOH (pH 8.8), 0.4mM NAD, 0.12 mM ADP, 0.8 mM iodonitrotetrazolium chloride, 0.17 IU/ml diaphorase (Sigma Chemical Co., USA) and 100 mM glutamate. The specific activity was expressed as “M formazan formed/mg protein per min.

Protein content of the sample was determined according to the method of Bradford [1]. Bovine gamma globulin (Sigma Chemical Co., USA) dis- solved in 25% glycerol was used as the standard for comparison.

Determination of LTs9 for mudskippers exposed to various concentrations of NH4Cl

Because of the great difference in ammonia tolerance between the two mudskippers, different

514 Y. K. Ip, C. Y. Lee et al.

concentrations of NH4Cl were used to obtain the respective LTs9 values. Ten B. boddaerti (20-30 g) were exposed individually at 25°C to 3.7 liter of aerated SW containing 20 or 50mM of NH,Cl (Merck Chemical Co., Germany). Similarly, 10 P. schosseri (90-120 g) were exposed to 150mM NH,Cl. The time of mortality for each group was recorded. The LTs59 values were determined graphically.

Statistical analyses

Results were presented as mean+SE. Student’s t-test was used to compare differences between means. Differences with P< 0.05 were regarded as statistically significant.

RESULTS

The concentrations of NH,* in the plasma and muscles of P. schlosseri and B. boddaerti increased significantly after exposure to terrestrial conditions for 24 hr (Table 1). Under both the submerged and terrestrial conditions, the NH4* concentra- tions in the liver of P. schlosseri were approximate- ly 4 times higher than those of B. boddaerti. The increases in NH,4* concentrations in the plasma

TABLE 1.

and muscle of B. boddaerti were not accompanied by any significant increase in urea concentrations. However, significant increases in urea concentra- tions occurred in the plasma, liver and muscle of P. schlosseri kept away from water (Table 1). The increase in the concentration of urea, without an accompanying increase in that of NH,4”, in the liver of P. schlosseri exposed terrestrially led to a significant increase in the urea/NH,* ratio in this organ.

The liver of P. schlosseri contained a significant- ly greater concentration of TFAA than that of B. boddaerti (Table 2). However, the TFAA concen- tration in the muscle of the latter was significantly higher than that of the former (Table 3). This was due to the presence of approximately 20 times more glycine in the muscle of B. boddaerti com- pared to that of P. schlosseri. The taurine concen- trations in the liver (Table 2) and muscle (Table 3) of P. schlosseri were significantly higher than those of B. boddaerti.

After exposure to terrestrial conditions for 24 hr, there were significant increases in the concen- trations of TFAA in the liver (Table 2), muscle (Table 3) and plasma (Table 4) of P. schlosseri, but not in those of B. boddaerti. In the latter

Concentrations (.M/ml] plasma and «M/g liver or muscle) of urea and ammonia and their ratios

(urea/ammonia) in the plasma, liver and muscle of B. boddaerti and P. schlosseri fully submerged in 50% SW or exposed to terrestrial condition for 24 hr’

Urea Fish Tissues Conditions Urea Ammonia —= Ammonia B. boddaerti Plasma Submerged 1.34+0.09 0.44+0.03 2.90+0.32 Terrestrial 1.61+0.13 0.72 +0.05* 2.09 +0.27 Liver Submerged 0.55 +0.07 0.64+0.07 0.91+0.17 Terrestrial 0.59+0.15 0.88 +0.09 0.66+0.13 Muscle Submerged 0.81+0.05 - 0.88 +0.07 0.93+0.12 Terrestrial 0.79 +0.12 1.46+0.20* 0.59+0.11 P. schlosseri Plasma Submerged 1.14+0.16 0.54+0.06 2.17+0.29 Terrestrial 2.34+0.32* 0.92 +0.05* 2.19+0.33 Liver Submerged 0.61+0.13 2.54+0.21 0.23+0.05 Terrestrial 1.66+0.21* 3.46+0.54 0.50 +0.08* Muscle Submerged 0.89+0.14 0.96+0.13 1.05+0.31 Terrestrial 1.47+0.13* 1.64+0.09* 0.90 +0.08

' Results represent means+SE of three of five determinations on separate preparation from different animals. * Significantly different from the corresponding value of the submerged fish.

Terrestrial Adaptations in Mudskippers S15

TaBLE 2. Concentrations (~M/g) of various free amino acids (FAA) and total FAA (TFAA) in the livers of B. boddaerti and P. schlosseri fully submerged in 50% SW or exposed to terrestrial condition for 24 hr’

B. boddaerti

P. schlosseri

a Submerged Terrestrial Submerged Terrestrial Ala 0.22 +0.02 0.15 +0.03 1.17 +0.12 4.13 +0.74* Arg 0.020 + 0.002 0.020 + 0.002 0.18 +0.03 0.19+0.04 Asp 0.26 +0.04 0.24 +0.03 0.27 +0.03 0.43 +0.04* Glu 1.96 +0.32 3.10 +0.47 4.20 +0.29 5.39 +0.31* Gly 1.09 +0.08 0.61 +0.15* 1.04 +0.12 0.68 +0.09* His 0.17 +0.02 0.16 +0.03 0.26 +0.02 0.33+0.04 Ile 0.040 + 0.006 0.039 + 0.005 0.080 + 0.007 0.16+0.01* Leu 0.130 +0.020 0.094 + 0.014 0.21 +0.02 0.37 +0.02* Lys 0.15 +0.01 0.14 +0.03 0.39 +0.02 0.98 +0.30 Phe 0.038 + 0.007 0.052 + 0.003 0.21 +0.02 0.22 +0.04 Pro 0.24 +0.05 0.28 +0.05 0.36 +0.03 0.59 +0.08* Ser 0.17 +0.01 0.13 +0.02 0.10 +0.01 0.18 +0.02* Thr 0.040 + 0.004 0.042 + 0.007 0.94 +0.10 0.39 +0.05* Tyr 0.17 +0.05 0.13 +0.009 0.09 +0.02 0.16+0.02 Val 0.18 +0.04 0.22 +0.04 0.24 +0.02 0.60+0.11* TFAA 5.13 +1.06 5.59 +0.80 10.3 +0.8 15.5+1.4* Tau 7.54 +0.34 WAR) se AY) Mbp, ee} 20.2+2.3

' Results represent means + SE of six to seven determinations on separate preparations from different animals. * Significantly different from that of the corresponding value of the submerged fish.

TABLE 3. Concentrations («M/g) of various free amino acids (FAA) and total FAA (TFAA) in the muscles of B. boddaerti and P. schlosseri fully submerged in 50% SW or exposed to terrestrial condition for 24 hr’

B. boddaerti P. schlosseri

Submerged Terrestrial Submerged Terrestrial Ala 1.95+0.09 2.94 +0.26* 1.30+0.12 2.99 +0.32* Arg 0.49+0.05 0.28+0.01* 0.18+0.02 0.33 +0.03* Asp 0.33+0.05 0.12 +0.02* 0.13+0.01 0.11+0.02 Glu 0.42+0.06 0.31+0.02 0.23+0.05 0.22 +0.03 Gly 22.6+1.0 23.0+1.3 1.07+0.13 1.30+0.16 His 0.71+0.13 1.03 +0.03* 0.25+0.02 0.32 +0.03 Ile 0.22+0.02 0.20+0.01 0.12+0.01 0.35 +0.03* Leu 0.41 +0.04 0.36+0.01 0.27 +0.02 0.66 +0.05* Lys 2.17+0.07 2.08+0.11 1.16+0.14 1.63+0.14 Phe 0.07+0.01 0.11+0.03 0.17+0.03 0.24+0.02 Pro 0.19+0.01 0.27 +0.02* 0.10+0.02 0.24 +0.01* Ser 0.90 +0.07 0.92+0.02 0.23+0.01 0.44 +0.05* Thr 0.32 +0.06 0.54+0.04* 0.31+0.02 0.42 +0.04* Tyr 0.25+0.01 0.17+0.02* 0.10+0.03 0.18+0.02 Val 0.28+0.03 0.26+0.02 0.21 +0.02 0.51+0.04* TFAA 31.6+0.8 Sf a Mog) 5.99 +0.33 9.93 +0.78* Tau 7.40+0.49 9.05+ 1.93 19.4+1.9 19.1+2.9

" Results represent means+SE of five to six determinations on separate preparations from different animals. * Significantly different from that of the corresponding value of the submerged fish.

516

Y. K. Ip, C. Y. Lee et al.

TABLE 4. Concentrations («M/ml) of various free amino acids (FAA) and total FAA (TFAA) in the plasma of B. boddaerti and P. schlosseri fully submerged in 50% SW or exposed to terrestrial condition for 24 hr!

B. boddaerti

P. schlosseri

ee Submerged Terrestrial Submerged Terrestrial

Ala 0.043 +0.002 0.071 +0.004* 0.140 +0.018 0.242 +0.019* Arg 0.023 +0.002 0.025 +0.007 0.044 +0.019 0.056 +0.003 Asp 0.0070 + 0.0012 0.0067 + 0.0006 0.0041 +0.0001 0.0034 + 0.0002 Glu 0.013 +0.002 0.014 +0.002 0.014 +0.001 0.012 +0.003 Gly 0.248 +0.013 0.260 +0.020 0.076 +0.008 0.078 +0.009 His 0.019 +0.001 0.021 +0.005 0.022 +0.004 0.027 +0.003 Ile 0.051 +0.005 0.047 +0.001 0.063 +0.004 0.156 +0.006* Leu 0.115 +0.012 0.103 +0.005 0.109 +0.009 0.256 +0.015* Lys 0.020 +0.003 0.056 +0.007 0.057 +0.019 0.139 +0.031* Phe 0.023 +0.002 0.022 +0.004 0.034 +0.004 0.049 +0.007* Pro 0.0090 + 0.0006 0.0110+0.0010 0.018 +0.001 0.036 +0.008* Ser 0.016 +0.001 0.017 +0.001 0.024 +0.002 0.025 +0.003 Thr 0.020 +0.001 0.025 +0.004 0.072 +0.008 0.069 +0.009 Tyr 0.022 +0.003 0.024 +0.001 0.021 +0.002 0.036 +0.005* Val 0.073 +0.007 0.066 +0.003 0.121 +0.007 0.195 +0.026* TFAA 0.721 +0.030 0.762 +0.041 0.886 +0.075 1.359 +0.088* Tau 8.26 +1.21 9.14 +0.98 10.0 +2.5 12:25 aE Ae4

' Results represent means+SE of five to six determination on separate preparations from different

animals

* Significantly different from that of the corresponding value of the submerged fish.

TABLE 5.

Specific enzyme activities of glutamate dehydrogenase in the amination (uM NADH oxidized/mg protein per min) and deamination (uM formazan formed/mg protein per min) directions, and their ratios (amination/deamination) from the livers and muscles of B. boddaerti and P. schlosseri fully submerged in 50% SW or exposed to terrestrial condition for 24 hr’

Amination Fish Tissues Conditions Amination Deamination

Deamination

B. boddaerti Liver Submerged 0.18 +0.03 0.0072 + 0.0008 PM Kehae 7472 Terrestrial 0.25 +0.03 0.0068 + 0.0006 28.2+2.7

Muscle Submerged 0.028 + 0.001 0.0022 + 0.0001 12.8+0.8

Terrestrial 0.025 +0.001 0.0022 + 0.0001 IBIR/SE Uf P. schlosseri Liver Submerged 1.62 +0.21 0.0372 + 0.0060 58.1+12.0 Terrestrial 0.85 +0.06* 0.0088 + 0.0014* 116 +21*

Muscle Submerged 0.13 +0.03 0.0075 + 0.0015 17.7+0.6

Terrestrial 0.20 +0.03 0.0098 + 0.0018 19.3+2.0

' Results represent means+SE of five to six determinations on separate preparations from different animals. * Significantly different from the corresponding value of the submerged fish.

Terrestrial Adaptations in Mudskippers 517

mudskipper, only the concentration of alanine in the plasma was significantly increased after terrest- rial exposure. However, similar exposure in- creased the concentrations of alanine, isoleucine, leucine, lysine, phenylalanine, proline, tyrosine and valine in the plasma of P. schlosseri (Table 4). There were significant decreases in glycine concen- trations in the livers of both B. boddaerti and P. schlosseri exposed to terrestrial conditions (Table 2). On the other hand, the concentrations of alanine, aspartate, glutamate, isoleucine, leucine, proline, serine and valine increased significantly in the liver of only P. schlosseri after 24 hr of terrest- rial exposure (Table 2). Terrestrial exposure signi- ficantly decreased (arginine, aspartate and tyro- sine) and increased (alanine, histidine, proline and threonine) the concentrations of several FAA in the muscle of B. boddaerti (Table 3). In compari- son, only significant increases in the concentrations of 8 FAA (alanine, arginine, isoleucine, leucine, proline, serine, threonine and valine) were observed in the muscle of P. schlosseri exposed to terrestrial conditions (Table 3). Terrestrial expo- sure had no significant effect on the concentrations of taurine in the liver (Table 2), muscle (Table 3) and plasma (Table 4) of both mudskippers.

The specific activities of GDH, in the aminating and deaminating directions, in the liver and muscle of P. schlosseri were significantly higher than those of B. boddaerti (Table5). Terrestrial exposure had no significant effect on the aminating and deaminating activities of the GDH from the liver and muscle of B. boddaerti. In contrast, similar exposure decreased the aminating and deaminat- ing activities of the GDH from the liver of P. schlosseri, but the amination/deamination ratio was significantly increased (Table 5).

The LTs 9 of P. schlosseri for 150 mM NH,Cl was 13 hr. For B. boddaerti, the LTs9 for 20 and 50 mM NH,Cl were 75 and 36 min, respectively.

DISCUSSION

Since branchial NH,* excretion would be in- efficient when no external water current is avail- able to irrigate the gills during terrestrial exposure, the plasma and muscles of P. schlosseri and B. boddaerti exposed terrestrially for 24 hr accumu-

lated significantly greater quantities of NH4* com- pared to those of the respective fish submerged in water. Similar to P. cantonensis and B. pectiniros- tris [21-23], terrestrial exposure had no significant effect on the urea concentrations in the plasma, liver and muscle of B. boddaerti. Hence, the accumulated NH,‘ does not seem to be channel- led into urea production in B. boddaerti. In contrast, P. schlosseri accumulated significantly greater concentrations of urea in the plasma, liver and muscle after being kept away from water for 24 hr. Thus, a conversion of NH,4* into urea might have occurred in this mudskipper during terrestrial exposure. The authors speculate that the liver may be the major organ involved in urea production in P. schlosseri, because this organ exhibited the greatest increase (2.5 fold) in urea content but without any significant accumulation of NH,4* after 24hr of terrestrial exposure, leading to a significantly greater urea/NH,* ratio compared to that of the submerged fish.

The pathway involved in the increased urea synthesis in P. schlosseri during terrestrial expo- sure 1s not elucidated in this study. Some teleosts can synthesize urea through purine catabolism in their liver [5, 8, 20]. Gregory [11] detected urate oxidase, allantoicase and allantoinase, which are involved in converting uric acid through allantoin and allantoic acid to glyoxylate and urea, in the mudskippers, Periophthalmus expeditionium and Periophthalmus gracilis. Glyoxylate could lead to the formation of glycine through transamination reactions. Since the glycine level in the liver of P. schlosseri exposed to 24 hr of terrestrial conditions was significantly lower than that of the submerged ones, it can be deduced that the increased urea production in this fish during terrestrial exposure may not be due to the degradation of uric acid. The ornithine-urea cycle is an alternative pathway for urea synthesis in some other teleosts [13, 20, 24]. However, Gregory [11] was able to detect only two (arginase and ornithine carbamoyltrans- ferase) out of five enzymes of the ornithine-urea cycle in the liver extracts from the mudskippers, P. expeditionium, P. gracilis and Scartelaos histophor- us. Since no information concerning this aspect is available for P. schlosseri, we are currently investi- gating if all the enzymes of this cycle are present in

518 Y. K. Ip, C. Y. LEE et al.

this mudskipper.

Iwata et al. [15] reported that the muscle of P. cantonensis accumulated FAA, especially the non- essential ones, after terrestrial exposure. How- ever, in P. schlosseri, accumulations of some essential amino acids, e.g. the branched-chain FAA, also occurred. Hence, there might be an increase in the mobilization of protein as carbon and energy sources in P. schlosseri exposed to terrestrial conditions. Such a shift in metabolic pathways may be related to the high terrestrial affinity of this mudskipper. It may be metabolical- ly more active under the terrestrial condition as its gills are better adapted to respire on land than in water [17, 18]. Indeed, it is the only fish known to exhibit bradycardia when submerged in water ([7]; Ip, unpublished results).

A closer examination of the FAA profiles re- veals that the major FAA accumulated in the tissues of P. schlosseri exposed to terrestrial condi- tions was alanine. The percentages of increases in TFAA which could be accounted for by the respec- tive increases in alanine contents in the liver, muscle and plasma of P. schlosseri exposed terres- trially were approximately 57%, 22% and 56%, respectively. Such a phenomenon was not observed in P. cantonensis [15]. It has been demonstrated that most of the amino acids re- leased by proteolysis in the white muscle of the spawning salmon during migration are collected and transported to other tissues as alanine [19]. Therefore, it is possible that a similar strategy is adopted by P. schlosseri during terrestrial expo- sure. According to our current understanding of pathways of amino acid metabolism, most of the free amino acid pool can be converted to alanine [12]. The overall quantitative energetics of these processes in white muscles appear to be quite favorable; the net conversion of glutamate to alanine would yield 10 mol ATP per mole alanine formed, and this value can be even higher for proline and for arginine conversion to alanine [12]. Since this favourable ATP yield from partial amino acid catabolism is not accompanied by a net re- lease of NH,4*, it would be advantageous to P. schlosseri which is confronted with the problem of NH,* excretion when it moves away from water. For such a system to work, the NH,* released

from the partial catabolism of various FAA must be incorporated into glutamate, which can then undergo transamination to transfer the amino group to pyruvate, leading to the formation of alanine. It would appear that a portion of the NH,* produced by P. schlosseri during the 24 hr of terrestrial exposure could be shuttled to the liver, and be converted into glutamate through amination of aKG. Although the specific activities (approaching V,yax) of GDH obtained do not necessarily reflect the enzyme activities in vivo, the amination/deamination ratio in the liver of this mudskipper significantly increased after 24 hr of terrestrial exposure, which correlates well with the significant increase in glutamate content in this organ.

Contrary to P. chrysospilos [15] and P. schlos- seri, B. boddaerti did not exhibit any increase in the TFAA and branched-chain FAA contents in its tissues when exposed terrestrially. Since there was no change in the glutamate level and the specific activities of GDH, in the aminating and deaminat- ing directions, in the liver of B. boddaerti after 24 hr of terrestrial exposure, the formation of alanine in the muscle of this fish might not involve the amination of aKG through GDH in the liver as in P. schlosseri. Instead, the formation of alanine from certain FAA in the muscle of B. boddaerti during terrestrial exposure might involve trans- amination reactions within this tissue.

That the increases in concentrations of urea and FAA in the tissues of P. schlosseri exposed to terrestrial conditions indeed signal a greater capac- ity to detoxify ammonia, and not merely an in- crease in the mobilization of protein and amino acids, is reflected by the high tolerance of this mudskipper to NH,* added to the ambient SW. Although it is generally accepted that fish are especially intolerant of dissolved ammonia [3], P. schlosseri could survive for at least 10 days in 100 mM NH,Cl (Ip, unpublished results), and its LTso for 150 mM NH,Cl was 13 hr. To the best of the authors’ knowledge, P. schlosseri exhibits one of the highest tolerance to NH,* amongst teleosts, and even amongst mudskippers [14].

Hence, it can be concluded that, even though both are mudskippers, P. schlosseri and B. bod- daerti have developed different degrees of adapta-

tion to survive in a terrestrial environment.

Terrestrial Adaptations in Mudskippers

The

former mudskipper has acquired a greater capacity to detoxify ammonia than the latter. Thus, NH,* excretion would be less of a problem to P. schlos- seri than B. bodaerti during a terrestrial excursion.

10

11

12

REFERENCES

Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye bind- ing. Anal Biochem 72: 248-259

Beutler HO, Michal G (1974) L-Glutamate: Deter- mination with glutamate dehydrogenase, diaphorase and tetrazolium salts. In “Methods of Enzymatic Analysis Vol. 4” Ed by HU Bergmeyer, K Gawehn, Academic Press, New York, 2nd ed, pp 1708-1713 Campbell JW (1973) Nitrogen excretion. In “Com- parative Animal Physiology “Ed by L Prosser, Saun- ders College Publishing, Philadelphia, 3rd ed pp 279-316

Chew SF, Ip YK (1990) Differences in the response of two mudskippers, Boleophthalmus boddaerti and Periophthalmus chrysospilos to changes in salinity. J Exp Zool 256: 227-231

Cvancara VA (1969) Distribution of liver allan- toinase and allanatoicase activity in freshwater tele- osts. Comp Biochem Physiol 29: 631-638

Faupel RP, Seitz HJ, Tarnowski W, Thiemann V, Weiss CH (1972) The problem of tissue sampling from experimental animals with respect to freezing technique, anoxia, stress and narcosis. Arch Biochem Biophys 148: 509-522

Gary WF (1962) Cardiac responses of fishes in asphyxial environments. Biol Bull 122: 362-368 Goldstein L, Forster RP (1970) Nitrogen metabol- ism in fishes. In “Comparative Biochemistry of Nitrogen Metabolism, Vol. 2” Ed by JW Campbell, Academic Press, New York, pp 495-518

Gordon MS, Boetius I, Evans DH, McCarthy R, Oglesby LC (1969) Aspects of the physiology of terrestrial life in amphibious fishes. I. The mudskip- per Periophthalmus sobrinus. J Exp Biol 50: 141- 149

Gordon MS, Ng WWS, Yip AYW (1978) Aspects of the physiology of terrestrial life in amphibious fishes. III. The Chinese mudskipper Periophthalmus cantonensis. J Exp Biol 72: 57-75

Gregory RB (1977) Synthesis and total excretions of waste nitrogen by fish of the Periophthalmus (muds- kipper) and Scartelaos families. Comp Biochem Physiol 57A: 33-36

Hochachka PW, Guppy M (1987) Metabolic Arrest and the Control of Biological Time. Harvard Uni- versity Press, London

13

14

15

16

17

18

19

20

21

22

23

24

519

Huggins AK, Skutsch G, Baldwin E (1969) Ornithine-urea cycle enzymes in Teleostean fish. Comp Biochem Physiol 28: 587-602

Iwata K (1988) Nitrogen metabolism in the muds- kipper, Periophthalmus cantonensis: Changes in free amino acids and related compounds in various tissues under conditions of ammonia loading, with special reference to its high ammonia tolerance. Comp Biochem Physiol 91A: 499-508

Iwata K, Kakuta M, Ikeda G, Kimoto S, Wada N (1981) Nitrogen metabolism in the mudskipper, Periophthalmus cantonensis: A role of free amino acids in detoxification of ammonia produced during its terrestrial life. Comp Biochem Physiol 68A: 589- 596

Kun E, Kearney EB (1974) Ammonia. In “Method of Enzymatic Analysis, Vol. 4” Ed by HU Berg- meyer, K. Gawehn, Academic Press, New York, 2nd ed, pp 1802-1806

Low WP, Ip YK, Lane DJW (1990) A comparative study of the gill morphometry in three mudskippers- Periophthalmus chrysospilos, Boleophthalmus bod- daerti and Periophthalmodon schlosseri. Zool Sci 7: 29-38

Low WP, Lane DJW, Ip YK (1988) A comparative study of terrestrial adaptations of the gills in three mudskippers-Periophthalmus chrysospilos, Boleo- phthalumus boddaerti and Periophthalmodon schlos- seri. Biol Bull 175: 434-438

Mommsen TP, French CJ, Hochachka PW (1980) Sites and patterns of protein and amino acid utiliza- tion during the spawning migration of salmon. Can J Zool 58: 1785-1799

Mommsen TP, Walsh PJ (1991) Urea synthesis in fishes: evolutionary and biochemical perspectives. In “Biochemistry and Molecular Biology of Fishes, 1. Phylogenetic and Biochemical Perspectives” Ed by PW Hochachka, TP Mommsen, Elsevier, New York, pp 139-163

Mort H (1979) Changes with time of ammonia and urea concentrations in the blood and tissue of muds- kipper fish, Periophthalmus cantonensis and Boleophthalmus pectinirostris kept in water and on land. Comp Biochem Physiol 64A: 235-243

Mort H, Nishikata K, Tamura O (1978) Nitrogen excreton of mudskipper fish, Periophthalmus can- tonensis and Boleophthalmus pectinirostriis in water and on land. Comp Biochem Physiol 69A: 189-193 Mori H, Nishikata K, Tamura O (1979) Ammonia and urea excretion from mudskipper fishes Perioph- thalmus cantonensis and Boleophthalmus pectiniros- tris transferred from land to water. Comp Biochem Physiol 63A: 23-28

Read LJ (1971) The presence of high ornithine-urea cycle enzyme activity in the teleost Opsanus tau. Comp Biochem Physiol 39B: 406-413

_— eh Wi Sulkin Thehioanet

av

vw Toe

twee ree wha

vat ancy |

lo au t ‘ape het og pO aT) MMO pepe bs. ios é ¥ ery piven i ij} biberki < ; a q*) ? i «Ae - a } Lis | a4

Ln |

,

axslinng oeteaicalilies Uacyubrep

oS

ant ‘Preaigns winasty heap URAL VHT OPE elite

GHA 1) "aed del “Str a7 Se oop aed wn

i Avi a7

(JTISAM al Pennie &

ie +m wha Lar gp p> TU gape wie | had ah vot RT| ly mn é

PAy bicebsiant). ame, an apaiien f SAAT oN eee are it : miele ornate ed My ‘ L200) 4. be-tet ayer

vei, "aul Peat aes “i uN,

oe

ona

iz <i ile Lei bene Paw ef vingonebetel viele TwwG 4 al \ aN

iy -a ig r.

ti * F olyasttrai) * vy }

wientt aw oe PA ;

‘GANS he Oo yaa E

; waert'«

LMT Oe APD Se a

yf

nivie surg t Oskimiyg ;

i, See litte tft hy g! 3) is ay he fF, as uae

’ C's Te

‘ neal “ira r. « SOR UDA VE jiu) ty Gi incon nae

¥

rts al PR eyeten's ty al " =m *. waked, Ipeeree

Me

DA Mepagt 9 ey ri sbaiidl Hy Ri: y wih ae

ZOOLOGICAL SCIENCE 10: 521-527 (1993)

Photoperiodic Responses of the Linden Bug, Pyrrhocoris apterus, under Conditions of Constant Temperature and under Thermoperiodic Conditions

Hipewaru Numata!, Ara H. SAuLicy?

and TaTyANA A. VOLKOVICH~

‘Department of Biology, Faculty of Science, Osaka City University, Sumiyoshi, Osaka 558, Japan, and *Laboratory of Entomology, Biological Research Institute, St. Petersburg University, St. Petersburg, 198904 Russia

ABSTRACT—Photoperiodic responses were examined in the linden bug, Pyrrhocoris apterus (L.) (Heteroptera, Pyrrhocoridae), from Belgorod, Russia, under conditions of constant temperature and under thermoperiodic conditions with a 12-hr cryophase and a 12-hr thermophase. The critical daylength for the induction of adult diapause was longer at lower mean temperatures. The induction of diapause under thermoperiodic conditions was determined principally by the mean temperature of the thermoperiod, although thermoperiodic conditions enhanced the induction of diapause somewhat in the threshold zone. Development under thermoperiodic conditions was more rapid than at constant temperatures that were equivalent to the mean temperature under thermoperiodic conditions. Developmental retardation was observed under photophases a little shorter than the critical value for

© 1993 Zoological Society of Japan

the induction of diapause, both at constant temperatures and under thermoperiodic conditions.

INTRODUCTION

Insects in the wild develop under conditions of daily and seasonal fluctuations in temperature and these fluctuations, together with changes in the photoperiod, regulate the seasonal life cycles of insects [7, 8, 21]. In some insects, the thermo- period has been shown to influence the parameters of photoperiodic responses by amplifying or reduc- ing the tendency to enter diapause, or by changing the critical daylength [2]. We must consider this influence in discussions of seasonal adaptation in Insects, in particular in species that live in a continental climate with large daily changes in temperature. However, photoperiodic responses of insects have usually been studied at constant temperature in the laboratory.

The linden bug, Pyrrhocoris apterus, has an adult diapause that is controlled by a long-day photoperiodic response [12, 18, 22]. Saunders [18] showed that photoperiod also affects the develop-

Accepted March 29, 1993 Received March 4, 1993

mental rate. Hodek [13] described the induction of diapause by brief exposure (2-4 days) to low temperatures in P. apterus. Furthermore, Honék [14] reported that the fecundity of this species is higher under natural alternating temperatures than at a constant temperature in the laboratory. However, no experiments under thermoperiodic conditions have been performed to examine the photoperiodic response in P. apterus. In this study, we examined the photoperiodic response in P. apterus, from an inland region of Russia, at constant temperatures and under thermoperiodic conditions, and we discuss the effects of tempera- ture on this response.

MATERIALS AND METHODS

Adults of P. apterus, after overwintering, were collected from the field in the reservation “Forest on the River Vorskla” (50°N, 36E) which is located in the forest-steppe zone in the Belgorod Region, Russia. Their eggs were used for the experiments. The insects were reared in the laboratory by the method by Schlagbauer [19].

522 H. Numata, A. H. SAULICH AND T. A. VOLKOVICH

Nymphs were reared in 500-ml glass bottles. The density of the nymphs was kept at 100-150 for early instars and was reduced to 50-75 for the final (fifth) instar. Seeds of the linden tree Tilia corda- ta, strewn on the ground were collected and served as food. Puparia of the blue meat fly, Calliphora vicina were also supplied to late-instar nymphs and adults. Water was provided in test tubes with cotton plugs. Food and water were replaced every two or three days. Zigzag-folded strips of filter paper were used to increase the inner surface area. After adult emergence, the insects were kept in male and female pairs in Petri dishes. Thirty days after adult emergence, the females that had not laid eggs were dissected. The stage of their ovaries was examined, and the females with immature ovaries were judged as being in diapause [12].

Experiments were performed in boxes in which both temperature and photoperiod were control- led [5]. The light intensity in the boxes ranged between 180 and 250 lx and was supplied by 20 W fluorescent lamps. The deviation from the set temperatures was not more than 1.0°C. The humidity fluctuated between 50 and 70%. Photo- periodic conditions from a photoperiod with a 12-hr photophase and a 12-hr scotophase (12L- 12D) to 18L-6D were used; the longest photophase corresponds to the maximum daylength in the region in which insects were collected, including civil twilight.

A square-wave type of thermoperiod was made by transferring the rearing bottles between temperatures at 9:00 and 21:00 every day. Thus, the thermoperiod was composed of a 12-hr cryo-

12L-12D Je 15L-9D 16L-8D 17L-7D 18L-6D

Time of day

Fic. 1. The combinations of thermoperiod and photo- period used in the experiments. Shaded bars, scotophase; open bars, photophase, in 24-hr cycles.

phase and a 12-hr thermophase (12C-12T). The thermophase coincided with the photophase, and the cryophase with the scotophase. Under 12L- 12D the thermoperiod and photoperiod were com- pletely synchronized, whereas longer photophases extended into the cryophase (Fig. 1).

RESULTS

Induction of diapause at constant temperatures

First, we compared the photoperiodic response

100 30°C

50

100 27C

100f 945°C

23°C

Incidence of diapause (%)

100 20°C

12 14 16 18 Photophase (h/day)

Fic. 2. The photoperiodic response curves for the in- duction of adult diapause at various constant temperatures in Pyrrhocoris apterus. n=20-32 (20°C), n=7-61 (23°C), n=154-190 (24.5°C), n= 17-36 (27°C), n=14-39 (30°C).

Photoperiodic Response of Linden Bug 523

with respect to induction of diapause for five constant temperatures: 20, 23, 24.5, 27 and 30°C. At each temperature, the insects showed a long- day photoperiodic response. As the temperature was raised from 20°C to 27°C, the critical day- length was shortened from about 18 hr to about 15.5 hr. However, the critical daylength was about 15.5 hr both at 27°C and at 30°C, even though at 30°C some females laid eggs even under 14L-10D

(Fig. 2).

Induction of diapause under thermoperiodic condi- tions

Next, we examined the photoperiodic response under three sets of thermoperiodic conditions: a thermoperiod with a 13°C cryophase and a 27°C thermophase (abbreviated as 13-27°C; mean,

100 20-32°C

50

13-33°C

Incidence of diapause (%)

12 14 16 18

Photophase (h/day)

Fic. 3. The photoperiodic response curves for the in- duction of adult diapause under thermoperiodic conditions (12C-12T) in Pyrrhocoris apterus. n= 31-56 (13-27°C), n=37-53 (13-33°C), n=13-24 (20-32°C). The photoperiodic response curves at constant temperatures equivalent to the means of the thermoperiodic conditions are shown by thin lines.

20°C); 13-33°C (mean, 23°C); and 20-32°C (mean, 26°C). The former two conditions are similar to the natural temperatures in the summer days in the native habitat of P. apterus. Under 13- 27°C conditions, all adults entered diapause with any photoperiod other than 18L-6D, as they do at a constant temperature of 20°C, which is equiva- lent to the mean temperature under the thermo- periodic conditions (Fig. 3). However, under 18L- 6D, the proportion of adults in diapause was significantly higher under 13—-27°C conditions than at a constant 20°C (P<0.05 by Fisher’s exact probability test). Furthermore, the preoviposition period was significantly shorter in nondiapause females reared under 13-27°C conditions (n=8, median=18, range=14—-24 days) than in those reared at a constant 20°C (n=12, median=23, range= 17-34 days; P<0.01 by Mann-Whitney U test, U=13.0). The photoperiodic response curve under 13-33°C conditions coincided with that at a constant 23°C, which is equivalent to the mean temperature under the thermoperiodic conditions. The proportion of diapause adults in 17L-7D was a little higher under 13-33°C conditions than at a constant 23°C (Fig. 3), but the difference was not significant (P>0.3 by Fisher’s exact probability test). Although we did not examine the photo- periodic response at a constant 26°C, which is equivalent to the mean temperature of 20-32°C conditions, the insects showed an intermediate response between constant temperatures of 24.5°C and 27°C (Fig. 3).

Thus, photoperiodic response curves for induc- tion of diapause under thermoperiodic conditions were similar to those obtained at constant temper- atures equal to the mean temperatures of the thermoperiods, although there was a little differ- ence under 18L-6D. Next we examined the induc- tion of diapause in 18L-6D under six other thermo- periodic conditions: 13-24.5°C, 15-27°C, 15- 33°C, 20-27°C, 20-33°C and 27-33°C. At a cryophase temperature of 13°C, the number of diapause adults varied from 5.7 to 100% of the total depending on the temperature of the thermo- phase. At a cryophase temperature of 15°C the percentage of diapause adults varied from 0 to 56.5%, depending again on the temperature of the thermophase. At a thermophase temperature of

524 H. Numata, A. H. SAULICH AND T. A. VOLKOVICH

TABLE 1.

Effects of thermophase and cryophase temperatures on the induction of

adult diapause under 18L-6D and 12C-12T* in Pyrrhocoris apterus

Temperature (C) Incidence of diapause n Cryophase Thermophase Mean ()

13 24.5 18.75 100 Dil 13 27 20 78.9 38 13 33 23 S)5 1 53 15 27 21 56.5 23 15 33 24 0 15 20 27 DBRS 0 22 20 32 26 0 13 20 33 26.5 0 15 Dy 33 30 0 19

* See text for details.

27°C, the percentage of diapause adults varied from 0 to 78.9% depending on the temperature of the cryophase (Table 1). Therefore, neither the temperature of the cryophase nor that of the thermophase independently defined the photo- periodic response. By contrast, the insects entered diapause at lower mean temperatures and failed to do so at higher mean temperatures both at con- stant temperatures and under thermoperiodic con- ditions. However, the threshold temperature was about 21°C under thermoperiodic conditions, and was a little higher than at constant temperatures (Fig. 4).

From the results at constant temperatures and under thermoperiodic conditions, we can conclude that the mean of the day and night temperatures

= fo) (S)

Incidence of diapause (%) a oO

15 20 25 30 Mean temperature (°C)

Fic. 4. Induction of adult diapause under 18L-6D in Pyrrhocoris apterus at various constant tempera- tures (solid circles) and under thermoperiodic condi- tions (12C-12T) (open circles).

defines for the most part the parameters of the photoperiodic response in P. apterus, although thermoperiodic conditions enhanced the induction of diapause to a slight extent near the critical daylength.

Developmental rate

Then we compared the duration of egg and nymphal stages under various photoperiodic con- ditions at three constant temperatures (20, 24.5, 27°C) and under two sets of thermoperiodic condi- tions (13-27°C, 13-33°C). Under all conditions, the females developed a little more rapidly than the males, although the results are combined in Figure 5. Both at constant temperatures and under thermoperiodic conditions, retardation of development took place near the critical daylength for the induction of diapause. The insects de- veloped most slowly under a photophase that was a little shorter than the critical value under all temperature conditions (Fig. 5). Apart from the developmental retardation in the threshold zone, the insects developed much more rapidly under 13-27°C conditions than at a constant 20°C, and more rapidly under 13-33°C conditions than at a constant 24.5°C (Fig.5). Thus, development under thermoperiodic conditions was much more rapid than that estimated from the mean tempera- ture of the thermoperiod. Figure 6 shows the developmental rate, designated as the reciprocal of the duration of egg and nymphal stages, under

Photoperiodic Response of Linden Bug 525

Developmental period (days)

Photophase (h/day)

Fic. 5. Effects of photoperiod on the developmental period in Pyrrhocoris apterus at various constant temperatures and under thermoperiodic conditions (12C-12T). Solid circles with vertical lines, duration of the period from oviposition to adult emergence (mean+S.D., n=31-112). Arrows, photoperiods that induced adult diapause in 50% of insects.

15L-9D, which was not very much affected by the developmental retardation in the threshold zone. In Figure 6, we show the simple linear regression lines for the developmental rate at constant temperatures and under thermoperiodic condi- tions. The lower threshold temperature for de- velopment, shown as the intercept on the abscissa, was 12.5°C and 9.0°C at constant temperatures. and under thermoperiodic conditions, respectively. However, there was only a slight difference in the accumulation of heat required for complete de- velopment, which is indicated by the reciprocal of the slope. It was 416 and 429 degree-days at

04

y=-0.0211+0.0023x

02

y=-0.0300+0.0024x

Developmental rate (1/day)

10 15 20 25 30 Mean temperature (°C)

Fic. 6. Developmental rate per day under 15L-9D in Pyrrhocoris apterus at various constant tempera- tures (solid circles) and under thermoperiodic condi- tions (12C-12T) (open circles). Lines, simple linear regressions.

constant temperatures and under thermoperiodic conditions, respectively (Fig.6). Thus, under thermoperiodic conditions the insects developed as rapidly as at a constant temperature that was higher by about 3.5°C than the mean temperature under the thermoperiodic conditions.

DISCUSSION

In P. apterus from Belgorod, the photoperiodic response for the induction of adult diapause de- pended on temperature (Fig. 2). At lower temper- atures, the critical daylength was longer, as it is in many other insects with a long-day photoperiodic response [7, 8, 21]. However, increasing the temperature above 27°C had no effect on the critical daylength (Fig. 2). The critical daylength was found to be about 15 hr both at 20°C and at 25°C in P. apterus from Erevan, Armenia (40°N) [22]. Thus the dependence on temperature of the critical daylength varies between geographic populations within a species.

In many insects with a long-day photoperiodic response, thermoperiodic conditions enhance the photoperiodic induction of diapause under the natural phase relationship, i.e., when the cryo- phase is concurrent with the scotophase and the thermophase coincides with the photophase [1, 4, 6, 9, 10]. These results have been explained by the hypothesis that scotophase temperatures strongly influence the induction of diapause [2, 8, 10, 21]. However, experimental results obtained in recent years indicate that scotophase temperatures do not always determine the induction of diapause. For example, in the predatory bug Podisus macu- liventris (Heteroptera, Pentatomidae, Asopinae), thermoperiodic conditions with a cool scotophase slightly reduce the incidence of diapause under short-day conditions but do not affect the critical daylength [11]. In another species in the same subfamily, Perillus bioculatus, thermoperiodic conditions with a cool scotophase reduce the inci- dence of diapause under near-critical photoperiods and, consequently, the critical daylength becomes shorter and equivalent to that at the temperature of the photophase [23]. Here we provide another example to contradict the view that the scotophase temperature has stronger effects on the induction

526 H. Numata, A. H. SAULICH AND T. A. VOLKOVICH

of diapause than the photophase temperature [2, SelON2 1]

In P. apterus, diapause under thermoperiodic conditions was determined principally by the mean temperature of the thermoperiodic conditions, although such conditions increased the incidence of diapause only slightly in the threshold zone (Figs. 3, 4). This weak enhancement of diapause induction in the threshold zone may be due to the low temperature in the scotophase as shown in other species [1, 4, 6, 9, 10]. However, we cannot deny the possibility that the 12C-12T theroperiod itself showed a short-day effect because in the present experiments the duration of thermophase did not coincide with that of photophase except 12L-12D (Fig. 1).

P. apterus developed under thermoperiodic con- ditions much more rapidly than at constant temperatures equivalent to the mean temperatures of the respective thermoperiods (Fig. 4). This response does not agree with the classic law of temperature summation but has been demon- strated in some other insects [2]. With respect to the rate of the development, thermoperiodic con- ditions corresponded to a higher temperature than the mean value, in contrast to the induction of diapause. In the European corn borer Ostrinia nubilalis, in which the effect of thermoperiodic conditions has been studied most intensively, ther- moperiodic conditions produce essentially the same rate of larval development as do the appropriate mean temperatures [3, 17], although a low temperature in the scotophase increases the incidence of diapause [1, 4]. Thus, in both species, day and night temperatures are integrated in dif- ferent ways for the development and for the mod- ification of the photoperiodic response.

Saunders [18] reported that P. apterus from Prague, Czechoslovakia (50°N), developed more slowly around the critical photoperiod for the induction of diapause than under short-day or long-day conditions at a constant temperature of 25°C. Our present experiments confirmed these findings not only at various constant temperatures but also under thermoperiodic conditions. Fur- thermore, the peak of retarded development moved along the photoperiodic axis in a tempera- ture-dependent manner, together with the critical

daylength for the induction of diapause. Conse- quently, insects developed most slowly under a photophase a little shorter than the critical value for the induction of diapause under all tempera- ture conditions (Fig. 5). Saunders [18] suggested that this developmental retardation promotes the seasonal synchronism in the univoltine population of P. apterus in Prague. Kidokoro and Masaki [16] showed a similar retardation of nymphal develop- ment near the critical photoperiod for the induc- tion of egg diapause in both univoltine and bivol- tine populations of the band-legged ground crick- et, Dianemobius nigrofasciatus, and they attri- buted the main function of this response to a fine tuning of the growth rate so that the diapause stage would be reached at the appropriate time. Our field observations indicate that the Belgorod population of P. apterus, which lives at the same latitude as Prague, produces two generations, at least in warmer years (unpublished). Even in Czechoslovakia, P. apterus is not always univoltine [15, 20]. The developmental retardation in P. apterus must have an ecological significance even in bivoltine life cycles, as in D. nigrofasciatus. Development under thermoperiodic conditions was more rapid than that estimated from the mean temperature of the thermoperiod (Figs. 5, 6). Furthermore, nymphs of P. apterus select warmer habitats and, therefore, their body temperature is sometimes much higher than the air temperature [15]. Therefore, under natural conditions, P. apterus may develop much more rapidly than in experiments at constant temperatures that were equivalent to the mean air temperatures. It is not appropriate to use the developmental parameters obtained from experiments at constant tempera- tures for predicting the seasonal development under natural conditions in P. apterus. Furthermore, in P. apterus photoperiodic re- sponse for the induction of diapause depended much on temperature (Fig. 2), and the tempera- ture is variable from year to year in Belgorod region. Therefore it is difficult to discuss the seasonal adaptation of this local population based only on the present results. We now plan to make further observations in the field and to perform experiments under natural conditions in an effort to determine the ecological significance of the

Photoperiodic Response of Linden Bug

effect of photoperiod and temperature on diapause and development in P. apterus.

10

11

REFERENCES

Beck SD (1962) Photoperiodic induction of di- apause in an insect. Biol Bull 122: 1-12

Beck SD (1983) Insect thermoperiodism. Annu Rev Entomol 28: 91-108

Beck SD (1983) Thermal and thermoperiodic effects on larval development and diapause in the European corn borer, Ostrinia nubilalis. J Insect Physiol 29: 107-112

Beck SD (1985) Effects of thermoperiod on photo- periodic determination of larval diapause in Ostrinia nubilalis. J Insect Physiol 31: 41-46

Braun VA, Goryshin NI (1978) Climate chambers with programming of photoperiodic and tempera- ture rhythms for ecological research. Vestn Lenin- grad State Univ 3: 26-34 (in Russian)

Chippendale GM, Reddy AS, Catt CL (1976) Photoperiodic and thermoperiodic interactions in the regulation of the larval diapause of Diatraea grandiosella. J Insect Physiol 22: 823-828

Danks HV (1987) Insect Dormancy: An Ecological Perspective. Biological Survey of Canada, Ottawa, p 439

Danilevsky AS (1961) Photoperiodism and Seasonal Development of Insects, Leningrad State Univ, Leningrad, p 243 (in Russian)

Danilevsky AS, Kuznetsova IA (1968) The intra- specific adaptations of insects to the climatic zona- tion. In “Photoperiodic Adaptations in Insects and Acari” Ed by AS Danilevsky, Leningrad State Univ, Leningrad, pp 5-51 (in Russian)

Goryshin NI (1964) The influence of diurnal light and temperature rhythms on diapause in Lepido- ptera. Entomol Obozr 43: 86-93 (in Russian) Goryshin NI, Volkovich TA, Saulich AH, Vagner M, Borisenko IA (1988) The role of temperature and photoperiod in the control over development and diapause of the carnivorous bug Podisus macu- liventris (Hemiptera, Pentatomidae). Zool Zh 67: 1149-1161 (in Russian)

12

13

14

15

16

17

18

19

20

21

22

73)

527

Hodek I (1968) Diapause in females of Pyrrhocoris apterus L. (Heteroptera). Acta Entomol Bohemo- slov 65: 422-435

Hodek I (1971) Induction of adult diapause in Pyrrhocoris apterus L. by short cold exposure. Oecologia 6: 109-117

Honék A (1986) Enhancement of fecundity in Pyrrhocoris apterus under alternating natural condi- tions (Heteroptera, Pyrrhocoridae). Acta Entomol Bohemoslov 83: 411-417

Honék A, Sramkova K (1976) Behavioral regula- tion of developmental cycle in Pyrrhocoris apterus L. (Heteroptera: Pyrrhocoridae). Oecologia 24: 277-281

Kidokoro T, Masaki, S (1978) Photopeiriodic re- sponse in relation to variable voltinism in the grownd cricket, Prteronemobius fascipes Walker (Orthoptera: Gryllidae). Jpn J Ecol 28: 291-298 Matteson JW, Decker GC (1965) Development of the European corn borer at controlled and variable temperatures. J Econ Entomol 58: 344-349 Saunders DS (1983) A diapause induction- termination asymmetry in the photoperiodic re- sponses of the linden bug, Pyrrhocoris apterus and an effect of near-critical photoperiods on develop- ment. J Insect Physiol 29: 399-405

Schlagbauer N (1966) Eine Methode zur Massen- und Dauerzucht der Feuerwanze Pyrrhocoris apter- us. Beitr Entomol 16: 199-202

Socha R, Sula J (1992) Voltinism and seasonal changes in haemolymph protein pattern of Pyrrho- coris apterus (Heteroptera: Pyrrhocoridae) in rela- tion to diapause. Physiol Entomol 17: 370-376 Tauber MJ, Tauber CA, Masaki S (1986) Seasonal Adaptations of Insects. Oxford Univ. Press, New York, p 411

Volkovich TA, Goryshin NI (1978) Evaluation and accumulation of photoperiodic information in Pyr- rhocoris apterus L. (Hemiptera, Pyrrhocoridae) dur- ing the induction of oviposition. Zool Zh 57: 46-55 (in Russian)

Volkovich TA, Kolesnichenko LI, Saulich AH (1990) The role of thermal rhythms in the develop- ment of Perillus bioculatus (Hemiptera, Pentatomi- dae). Zool Zh 69: 70-81 (in Russian)

vo fh done week remmnnetinae 7 $f) eames rat a

SU a Sai

i

ie) Bee

‘ - " * ’ ' ' vy “A +ri- j 1 i +h ; i ors Shi i “ 4) 4% ‘ i ie ee feo | at le ae } il

- Py may)

=e ' wrt ee ¥ é ide deter one

7 — ewer ats. +) ye She {Uae er BM), otaikatabat Kins od

: ie ae 4: Ai; Ft | id 4 i Ota aett AcE alley

43 9) Ke hi ver

[Pipa ty aie! i Pe ba ¥ i). 71) “abkwerteteiy leery BU i ratte

i

matt} ea Un age

ridatcn ite: Fe ia.

2M de Detoennindy Tone APR | aid nd ‘ovo i belbtalald rea) steal VHT OR Te ee =p ty aster elas

| ime ate : Jjeaaliay

( trsaggrae pt itt) Sf

) 7 ’ b 4 ie > a od pol rade " en AGGMSOONT

pany ipikulibalete te lei ad

' ‘457 At ietre@y a tl wal onenid * F _ hy hepa d tate \ ro Oeley. nage ' \ Ree don't ‘ft a See } i ‘ “ SR Se 6) HG bi ! Wit i: i SAF) AT a sMallias yt aaa rH I e ity i fe rs ‘ ; mitt ral vig ; i opel ot A i * ie ‘ yi tiey I” {ee 2,

8 AP Neuniqaen M RHE ao) gyi grirlS set

reliant ju. lhe

i - J ia ed " RES ae

. ar 3 %

be OF i

ZOOLOGICAL SCIENCE 10: 529-536 (1993)

© 1993 Zoological Society of Japan

Phylogenetic Position of Acoel Turbellarians Inferred from Partial 18S rDNA Sequences

TomogE KatayAMa!, MASsAMIcHI YAMAMOTO’, Hiroshi Wada”

and Noriyuki SATOH”

'Ushimado Marine Laboratory, Faculty of Science, Okayama University, Ushimado, Okayama 701-43, and *Department of Zoology, Faculty of Science, Kyoto University, Kyoto 606-01, Japan

ABSTRACT—Primitive platyhelminths, especially Acoel turbellarians, are thought to be key to understanding the origin and evolution of metazoa. In order to infer their phylogenetic position within the phylum Platyhelminthes, we determined and compared the complete nucleotide sequence of a region of about 750 base pairs in the central part of an 18S rDNA for ten turbellarians, including two species of the group Acoela, six species of the group Polycladida, and two species of the group Tricladida. The deduced phylogenetic tree suggests that the three groups examined form discrete and separate entities. In addition, the tree suggests an earlier emergence of the Acoel turbellarians than the other platyhelminths. This animal may not be derived by means of secondary reduction from advanced acoelomates but may be nearest to its metazoan ancestors.

INTRODUCTION

The origin and evolution of multicellular ani- mals (metazoans) have received considerable attention over the years [4, 9-11, 14, 26]. It is generally accepted that the metazoa arose from protozoa, perhaps 700-1000 million years ago. Models of the metazoic origin suggest that either colonial flagellates [10] or syncytial ciliates [9, 11] became acoel flatworm-like creatures (phylum Platyhelminthes), from which various modern metazoans have diverged. Current debate con- cerns whether the metazoa are monophyletic or polyphyletic and what the ancestral metazoa were like. Recent studies of the molecular phylogeny of invertebrates have suggested that the metazoa are polyphyletic (Cnidarians arose from a protist ancestry different from the Bilateria) [5, 8] and within the Bilateria, an early split gave rise to Platyhelminthes and the coelomate lineage [8, 13], although a monophyletic origin of the metazoa has also been proposed [16]. Regardless of route, primitive platyhelminths, especially Acoel tur-

Acctpted March 5, 1993 Received November 9, 1992

bellarians, are key animals in understanding the origin and evolution of modern metazoans.

In the present investigation, we compared molecular sequence data derived from 18S rDNAs to estimate the phylogenetic position of Acoel turbellarians. The 18S rRNA or its gene (18S rDNA) is ideally suited for phylogenetic studies of distantly related organisms because it is rich in information and the sequencing methodology per- mits the rapid accumulation of very large data- bases. According to Pace et al. [18], these mole- cules are conservative in overall structure and constitute a single gene family, so that problems of establishing homology among paralogous genes are avoided. The rRNA gene seems to be free of artifacts of lateral gene transfer between phylo- genetically distant organisms. This and the ab- sence of paralogous genes mean that 18S rRNA sequences accurately reflect phylogenetic rela- tionships among the organisms from which the tRNA were prepared. The rRNA are present in large amounts in all organisms and are easily isolated. The conservative nature of the rRNA structure extends to the nucleotide sequence. Some regions of the molecule are highly conserved among distantly related species. Using these uni-

530 T. Katayama, M. YAMAmoto et al.

versally conserved regions, direct rapid sequencing of 18S rRNA (DNA) was achieved. Enough data were accumulated from 18S rRNA sequences for statistically significant comparisons. Since each of the 1000 or so sequenced nucleotides constitutes a characteristic state, the evolutionary information content of rRNA sequences is very high.

Here we determined the complete nucleotide sequence of a region of about 750 base pairs (bp) in the central part of the 18S rDNA for ten turbellarian species consisting two species of Acoela, six of Polycladida and two of Tricladida. The phylogenetic relationship among these tur- bellarians was analyzed by sequence comparison. Our primary interest was in determining whether the acoel turbellarian is a primitive platyhelminth or a secondary reduction from advanced acoelo- mates.

MATERIALS AND METHODS

Animals

The nine turbellarian species examined included two species of the order Acoela, Convoluta naikaiensis (Yamasu) and Amphiscolops species (Yamasu), six species of the order Polycladida, Notoplana koreana (Kato), Planocera multitenta- (Kato), Stylochus orientalis (Bock), Pseudostylochus obscurus (Stimpson), Stylocho-

culata

plana pusilla (Bock), and Thysanozoon brocchii (Risso), and two species of the order Tricladida, Dendrocoelopsis lactea (Ichikawa et Okugawa) and Dugesia japonica (Yeri et Kaburaki). Whole animals were frozen in liquid nitrogen and kept at —80°C until use.

DNA isolation

Frozen and powdered samples were lysed in TE buffer (10 mM Tris-HCl, 0.1M EDTA, pH 8.0) that contained 0.5% sodium dodecyl sulfate. After digesting samples with proteinase K (100 g/ml) at 50°C for 3 hr, DNA was extracted with phenol and precipitated in ethanol and an equal volume of 5.0 M ammonium acetate. Samples resuspended in TE buffer were further purified by RNase A (20 pg/ml) digestion at 37°C for 1 hr followed by ethanol precipitation.

Adults of Convoluta naikaiensis have symbiotic alga. To avoid contamination of algal DNA, that of the turbellaria was extracted from embryos.

Amplification of the central region of 18S rDNA and sequencing

A region of about 1000 bp from the central part of 18S rDNA was amplified by the polymerase chain reaction (PCR) [20] in a Perkin Elmer Cetus thermal cycler. Amplifications were performed in 100 ul of 50mM KCl, 1.5mM MgCh, Tris-HCl (10mM, pH9.0), 0.1% Triton X-100, with 0.2 mM each dNTP, 100 pM primer, template DNA (10-100 ng) and 2U Taq DNA polymerase (Promega). Primer-1 [5 -CAG(CA)CCCGCGG- TAAT(TA)C-3’] and primer-2 [5’--ACGGGCG- GTGTGT(AG)C-3’], the latter being identical to Primer C of Field et al. [8], were used for am- plification. One of the primers was kinased prior to PCR at the 5’ terminal phosphate. The temper- ature regimen for 30 cycles was 1 min at 92°C, 2 min at 55°C, and 3 min at 72°C.

According to the method described by Higuchi and Ochman [12], single-stranded DNA was obtained by digesting the amplified product with lambda exonuclease. The nucleotide sequence of the single-stranded PCR products was directly determined by dideoxy chain-termination [22] us- ing Sequenase ver 2.0 (USB) and [*S]-dATP (Amersham). In addition to primer-1 and -2, primer-3 (5’-TTGGCAAATGCTTTCGC-3)), primer-4 (antisense of primer-3), primer-5 [S5’- ATTCCTTT(AG)AGTTTC-3)] and __ primer-6 (antisense of primer-5) were used in sequence determination.

Comparison of sequences and inferences about phylogeny

Sequences were aligned on the basis of max- imum nucleotide similarity. Using the aligned sequences, evolutionary distance values were calculated pairwise as described by Jukes and Cantor [15]. The phylogenetic tree was inferred from an analysis of results by the neighbor-joining method of Saitou and Nei [21]. The degree of support for internal branches of the tree was further assessed by bootstrapping [7].

The corresponding sequences of Saccharomyces

Molecular Phylogeny of Turbellarians 531

cerevisiae [17] and Neurospora crassa [17] provided data of outgroup organisms.

RESULTS

Partial nucleotide sequences of the 18S rDNAs from ten turbellarians

The complete nucleotide sequences of a region of about 750 bp in the central part of the 18S rDNA of ten species of turbellarians are summa- rized in Figure 1. The sequences correspond to positions 910-1646 in the sequence of human 18S tRNA [17]. Alignment of the nucleotide sequ- ences from the ten planarian species revealed some key features. In some regions, the nucleotide sequences were highly conserved. For example, among the ten species, few changes were evident in the sequences from positions 277 to 380. In contrast, the nucleotide sequences of other re- gions, such as positions 214-267 and 543-614, were highly variable. Sequences of yet another region varied only moderately. These regional differences may reflect differences in the function- al constraints upon the regions in the 18S rRNA.

The sequence alignment showed that the six species of the order Polycladida share common nucleotide sequences, as do the two species of the order Tricladida. However, the nucleotide sequ- ences of these two groups are very different, suggesting their discrete separation. In addition, the nucleotide sequence of the two acoela species Convoluta and Amphiscolops differed from those of Polycladida and Tricladida so much that almost the entire region of the 18S rDNA we examined was changed. This suggests a large evolutionary divergence of the acoel turbellarians from the two other groups of planarians.

Phylogenetic relationships among the ten species of turbellarians

Structural similarity and evolutionary distance values were calculated pairwise as described [15] between the sequences aligned in Figure 1. The results are summarized in Table 1. A phylogenetic tree was constructed by the neighbor-joining method [21], by reference to the distance values of Table 1. The phylogenetic tree shown in Figure 2

indicated that the ten turbellarians can be subdi- vided into three groups, which correspond to the orders Acoela, Polycladida and Tricladida, and that the acoel group emerged first among the three. The internal branches of the tree were supported by the high bootstrapping values. Among the six species of Polycladida, the bran- ching of Thysanozoon belonging to suborder Cotylea from the five species of Acotylea was evident (Fig. 2).

The phylum Platyhelminthes is subdivided into four major groups, Turbellaria (planarians), Monogenea (monogenetic flukes), Trematoda (digenetic flukes), and Cestoda (tapeworms) [4, 14]. Recent studies of the phylogenetic rela- tionships among several parasitic platyhelminths by comparison of partial 18S rRNA sequences [2, 17] support the hypothesis that platyhelminths are monophyletic; that planarians emerged first, then flukes and finally tapeworms evolved. Together with the results of the present study, we con- structed a phylogenetic tree to examine the posi- tion of acoel turbellarians within the phylum Platy- helminthes (data not shown). The tree suggests that Convoluta and Amphiscolops are more primi- tive members within Platyhelminthes.

Phylogenetic position of acoel turbellarians in the animal kingdom

The complete or partial nucleotide sequences of 18S rRNA (or DNA) of various phyla, including Artemia salina [17], Caenorhabditis elegans [17], Ciona savignyi [19], Xenopus laevis [17], and Homo sapiens {17| have been reported. The phylogenetic tree (Fig. 3) shows our positioning of the Acoel turbellarians, Polycladida and Tricladi- da in the animal kingdom. It is evident that these three groups of turbellarians are not interrupted by any other animal groups. Also, the branching point of the acoel turbellarians is earlier than those of the other turbellarians compared in this study.

DISCUSSION

In this investigation, we determined and com- pared the complete nucleotide sequence of a re- gion of about 750 bp in the central portion of the 18S rDNA from ten species of turbellarians. The

532

i

1

1

Re

T. Katayama, M. YAMaAmoro eft al.

Seat, corey aioe er eis coi ocan aa coca ic onions tao ncmodarescdorrra.ca 7 1. Stylochus TIGTTGGTTTTCGGAACA--————' TGAAGTAATGATTAAGAG———GGACAGAC—GGGGGCATTCGTATTGC 2. Planocera - 3. Pseudostylochus 4. Notoplana 5. Stylochoplana 6. Thysanozoon (<) 7. Dendrocoelopsis - A AT PA LEC AT GcT 8. Dugesia A - A A c- AT GcT 9. Convoluta AAC CGG G CTGTT GcTCG G GA A G T 0. Amphiscolops Ac C TACCG AGTTTCCCTCAG TGC A CTTA GA TC Lup teiioute tesseiloseiry ve seyteyatnteo leh wowspvelis save cteavtoneifong enayalisl sts wa ver’eripiey sisage'ctlaweue, © fp auene to cb asteaelloy peavex cise takers pag Ree ae eee eae 1. -GGTGGG-AGAGGTGAAATTCTTGGATCATCGCAAGAC—GCCCTAC---AGCGAAAGCATTTGCCAAGAATGTTTCCATTAATCAAGAAC 2. - 3. 4. 5. 6. A GA ls T AA AGC AAA T T 8. cT GA AGC GAAA 1 T 9. TAC TT GG AAA T ae 0. TAG TTC CA AAA TACT T 3 otto ; vA oe ayayphacavea Hialece-e.poa avacacke auauerpieccratectisvete avgroveloie He tere pratte ere cere Ghee aCe Eee eee 1. GAAAGTCAGAGGTTCGAAGACGATCAGATACCGTCCTAGTTCTGACCATAAACGATGCCAACTGGCAA-TCCGTTGCGATTGCAAGTICG 2. G 3. 4. 5. 6. CTG Cc CG ics A T A G TA CGAA G AATT CAATA 8. T A T eT A G G TA CGAA GA ATT-TAATC Go TA TA G T C GA A TTA TA G TT cC TG Cc =—-7TC 0. TA T G T clhUG OA a8 T A TA G CTT CCAAAG CAT- CCAT nig. wileh oat isstel apie fab gk eS oud ip oe NGE RR etallpaalabnce isis. o. 8 pos) Seek at ebay «ee sic tater, pe wdencahersite (cise eee 1. ATCCAACGGGCAGCCT-—CCGGGAAACCAAAGTCTTTIGGGTTCCGGGGGAAGT-ATGGTTGCAAAGC TGAAAC TTAAAGGAATTGAC—GGA 2. T 3. 4. Do 6. AT 7. TC TTIG TAA TA A 8. TTG TAA TA A G 9. CTGGG AA A TTTAA TAT ATC 0. cCTTGG AAA TTAA TAT AT becovcnncruech eer Biss ava) poave suse weve aviplane etcsties erehese cacnee chet ates pleeene beter +....,--430

CU MN HY Pwd Pe a aaa qgqaQaaa

epeeee

. AAGGCACCACAAGGAGTGGAGCC TGCG-CTTAATTTGAC TCAACACGGGAAAAC TC TACCCGGCCCGGACACTGTGAGGATTGACAGATT

G

epeeee

HHHAHA

Molecular Phylogeny of Turbellarians 533

Rap T RSME rt. APT eis hovel BAND ARE RATS RIE Ueay NAL We tee ELSE. oan Ae SIO0 1. Stylochus GAAAGCTCTITCTTGATTCGGTGGGTGGTGGTGCATGGCCGTT-CTTAGTTIGGTGGAGCGATTTGTCTGG 2. Planocera 3. Pseudostylochus 4. Notoplana = 5. Stylochoplana 6. Thysanozoon TCAT G 7. Dendrocoelopsis ay 8. Dugesia ay A 9. Convoluta TTA- C AA CAA Cc A T Cc 10. Amphiscolops T- A ATA Cc A Tm uc PRPs = Uta ta epee eta ait o. aualarjpeouelsucisteuexessapiaucnevottereie| @lece Sasisot eat isi plaitanete Petal pe eaters peoples cheaie 590 1. TTAATTCCGATAACGAACGAGACTCT-——AGCCTACTAAATAGTACAC-TCATCCAT—-TTGTGTGAGTGCC ———GACTTCTTAGGGGGAC 740 Sr TG 4. TG 5 cr = Be 6. CT CA AG = Tlic A G T TG GTT AA- CCA - T IN JX Ge 8. ATG GT T T MTTTAAC C CAA —- AAT C AC AA AT 9. TCT G AT GAGCAG ACG AG A TGCAAA- AG x (e 10. TcTCc cT ATT T A TG CA GCATAA T A PT ca eM Repo atie ctonsiin cshereutiore akeresevcvelnitienercuenphcviuslie hayes § ofp avoinpectiessiaystp/avegs«tajeus,onpialovaisit says dynes 600 1. AA-GTG—GCATTCAAGCCATA-CGAAA--TTGAG-—CAATAACAGGTCTGTGAT—GCCC TTAGA-TGTCCGGGGCCGCACGCGCGCTACAC Xe Se G 4. cc CG Br = G 6. A CG T A Gas ie -A G T- TRA LT G A 8. AA ( CaS Cleny ZA ae AA A G A 9. TC TCCA T ATACGGAAAGCGAAG GT Lee Aa 10. T AGCA A - ATAC GCTAGTGAAG AT he TA — A PPT AS CN Mot ate testa elstcee shasdeterottctersstaeate i clasye tahoe cdatsh erat vipcateale Slaiecrplana scones OTT 1. TGAATAGCGTAACGAGTTATTCTCCTGATCCGA-AAGGGTCGGGCAACCTGTTGAAACCTATTCGTG 2. 3. = G fe 4. @ 8)5 Ty Cc 6. TC G Cc & A Ta GCAGTAAC Ae AGCTA aw © T ot ACTG A 8. GCAGTTCC A A CTTAACCTA TA T ue Ake ACTG A oF T CTCTAC GA T G AAAA A TAGAG AT A T ACAGA CGGAG AATT T 10. TG GCTTC TAT A AATC T TGA G TcT A T ACAAA G A TCCTAAA Fic. 1. Alignment of the central regions of the 18S rDNAs from ten Turbellarians. All the bases are shown for

Stylochus orientalis and only bases different from these are shown for other species. Bars indicate deletions.

534 T. Katayama, M. YAMAmoTo et al.

TABLE 1.

Structural similarity and evolutionary distance data for turbellarian 18S rDNA sequences

Genera il. De 3. 4.

De 6. T: 8. 3. 10. 11. 12"

Stylochus Planocera Pseudostylochus Notoplana

1

2

3

4

5. Stylochoplana

6. Thysanozoon 7. Dendrocoelopsis 8. Dugesia

9. Convoluta

10. Amphiscolops 11. Saccharomyces 12. Neurospora

0.000 0.004 0.002 0.005 0.004 0.002 0.005 0.002 0.002 0.048 0.100 0.131

0.004 0.049 0.102 0.134 0.221

0.048 0.100 0.131 0.221 0.214 0.184 0.188 0.048 0.100 0.131 0.221 0.214 0.184 0.188 0.223 0.219 0.188 0.191 0.216 0.186 0.188 0.049 0.098 0.134 0.223 0.216 0.188 0.188 0.134 0.170 0.270 0.248 0.228 0.231 0.049 0.231 0.209 0.177 0.173

0.250 0.228 0.205 0.200

0.117 0.236 0.245

0.233 0.238

0.067

The values are average numbers of nucleotide substitutions per sequence position determined by the Jukes and

Cantor formula [15].

ae 12.Saccharomyces 11.Neurospora

10.Amphiscolops

9. Convoluta 7.Dendrocoelopsis 8.Dugesia 6.Thysanozoon 5.Stylochoplana 4.Notoplana 3.Pseudostylochus 2.Planocera

1.Stylochus

0.1 unit

Fic. 2. The phylogenetic tree of ten turbellarians, as deduced by the neighbour-joining method using the fraction of observed substitution difference over all sites. The scale bar indicates an evolutionary dis- tance of 0.1 nucleotide substitution per sequence position. Numbers at each branch indicate the percentage of times a node was supported in 500 bootstrap pseudoreplications by the neighbour- joining method.

phylogenetic tree (Fig. 2) indicated that the ten species can be subdivided into three groups, which correspond to the orders Acoela, Polycladida and Tricladida. In addition, together with the phy- logenetic tree including other groups of the animal kingdom (Fig. 3), early emergence of the acoel

turbellarians in metazoic evolution is evident.

11.Saccharomyces

12.Neurospora 10.Amphiscolops 9.Convoluta 7.Dendrocoelopsis

8.Dugesia

2.Planocera 5.Stylochoplana Caenorhabditis

Artemia Ciona

Xenopus

Homo sapiens

0.1 unit

Fic. 3. Phylogenetic relationships of the turbellarians with other organisms (Neurospora crassa, Sacchar- omyces cerevisiae, Caenorhabditis elegans, Artemia salina, Ciona savignyi, Xenopus laevis and Homo sapiens). The tree was constructed by the neigh- bour-joining method. The scale bar indicates an evolutionary distance of 0.1 nucleotide substitution per sequence position. Numbers at each branch indicate the percentage of times a node was sup- ported in 500 bootstrap pseudoreplications by the neighbour-joining method.

Therefore, they might be some of the closest multicellular animals to their metazoan ancestors. The early emergence of acoel turbellarians will be substantiated further by investigating with other

Molecular Phylogeny of Turbellarians 535

molecular probes such as 28S rDNA.

A phylogenetic tree constructed from the 5S rRNA sequences suggested that the branching points of Dugesia japonica (Tricladida) and nema- todes (eg, Caenorhabditis elegans) are slightly ear- lier than those of other metazoans, including Pla- nocera reticulata (Polycladida) [9]. That means turbellarians including Tricladida and Polycladida are polyphyletic. However, that view was not supported in the present study (see Figs. 2 and 3). Even when nematodes are included in the 18S rRNA (DNA) tree (Fig. 3), Tricladida and Polyc- ladida are categorized into one larger group, or Turbellarians.

It seems a consensus that all platyhelminths except for the lower turbellarians, including Acoela, Catenulida and Nemertodermatida, are of monophyletic origin [1-3, 6, 23, 24]. However, the relationships among the Catenulida, the Nemertodermatida-Acoela and the other orders of Platyhelminthes, are problematic [1, 3, 6, 23, 24]. Most views of the relationships within the Platyhel- minthes support the hypothesis that a primary acoelomate condition is derived from an acoel- type worm, and as the phylum evolved, the gut became more elaborate and the division into dis- tinct germ layers more pronounced. The Acoela, which are virtually solid with a ventral mouth .and central ‘syncytial’ area acting as the gut, have many primitive features, and have been considered nearest to the ancestral group within the phylum [11]. Catenulida and Nemertodermatida, with fully a developed gut and more distinct mesodermal matrices, are assumed to represent the next level of sophistication. Recently, how- ever, several investigators have argued that Cate- nulida and Nemertodermatida, the latter, in par- ticular, may be more primitive [24, 26]. Acoela are unusual in having modified cilia, in lacking any intercellular matrix [24], and in showing a peculiar form of ‘duet’ cleavage similar to that of gastro- trichs and nematodes (cf. [26]). They might there- fore be a derived group, an example of secondary reduction. The present 18S rDNA sequence analy- sis, however, did not support this view, namely the branching point of Convoluta was much earlier than those of other turbellarians. Further studies of the molecular phylogeny of Catenulida and

Nemertodermatida will offer more conclusive evi- dence of turbellarian evolution.

ACKNOWLEDGMENTS

We are very grateful to Dr. Takashi Miyata and Mr. Naruo Niko, Department of Biophysics, Kyoto Universi- ty for their help in constructing the phylogenetic trees. We also thank Dr. Sachiko Ishida, Department of Biolo- gy, Hirosaki University for providing us Dendrocoelopsis lactea species, Dr. Terufumi Yamasu, Department of Biology, Ryukyu University for his kind help in collect- ing acoel species, and Dr. Ken-ichi Tajika, Department of Biology, Nihon University School of Medicine for identifying the three species of Polycladida (Notoplana, Planocera and Stylochus). This research was supported in part by grants from the Research Institute of Marine Invertebrates (Tokyo) to N.S. and by the Narishige Zoological Science Award to M.Y.

REFERENCES

1 Ax P (1985) The position of the Gnathostomulida and Platyhelminthes in the phylogenetic system of the Bilateria. In “The Origins and Relationships of Lower Invertebrates”. Eds. by SC Morris, JD George, R Gibson, HM Platt, Clarendon Press, Oxford, pp 168-180

2 Baverstock PR, Fielke R, Johnson AM, Bray RA, Beveridge I (1991) Conflicting phylogenetic hypoth- eses for the parasitic platyhelminths tested by partial sequencing of 18S ribosomal RNA. Intern J Para- sitol 21: 329-339

3 Brooks DR (1989) A summary of the database pertaining to the phylogeny of the major groups of parasitic platyhelminths, with a revised classifica- tion. Can J Zool 67: 714-720

4 Brusca RC, Brusca CJ (1990) Sinauer, Sunderland

5 Cristen R, Ratto R, Baroin A, Perasso R, Grell KG, Adoutte A (1991) An analysis of the origin of metazoa, using comparisons of partial sequences of the 28S RNA, reveals an early emergence of triplob- lasts. EMBO J 10: 499-503

6 Ehlers U (1986) Comments on a phylogenetic system of the Platyhelminths. Hydrobiologia 132: 1- 12

7 Felsenstein J (1985) Confidence limits on phy- logenesis: an approach using the bootstrap. Evolu- tion 39: 783-791

8 Field KG, Olsen GJ, Lane DJ, Giovannoni ST, Gheselin MT, Raff EC, Pace NR, Raff RA (1988) Molecular Phylogeny of the animal kingdom. Scien- ce 239: 748-753

9 Hadzi J (1963) The Evolution of Metazoa. Perga-

Invertebrates.

10

11

12

13

14

15

16

17

18

19

536

mon Press, Oxford

Haeckel E (1874) The gastraea-theory, the phy- logenetic classification of the animal kingdom and the homology of the germ-lamellae. Quart J microsc Sci 14: 142-165

Hanson ED (1977) The Origin and Early Evolution of Animals. Pitman, London

Higuchi RG, Ochman H (1989) Producing of single- stranded DNA templates by exonuclease digestion following the polymerase chain reaction. Nucleic Acids Res 17: 58-65

Hori H, Osawa, S (1987) Origin and evolution of organisms as deduced from 5S ribosomal RNA sequences. Mol Biol Evol 4: 445-472

Hyman LH (1951) The Invertebrate Vol 2. McGraw-Hill, New York

Jukes TH, Cantor RC (1969) Evolution of protein molecules. In “Mammalian Protein Metabolism”. Ed. by HN Munro, Academic Press, New York, pp 21-132

Lake JA (1990) Origin of the metazoa. Proc Natl Acad Sci USA 87: 763-766

Neefs J-M, Van de Peer Y, Hendriks L, De Wachter R (1990) Compilation of small ribosomal subunit RNA sequences. Nucleic Acids Res 18: 2237-2317 Pace NR, Stahl DA, Lane DJ, Olsen GJ (1985) Analyzing natural microbial populations by rRNA sequences. Amer Soc Microbiol News 51: 4-12 Rieger RM (1985) The phylogenetic status of the acoelomate organization within the Bilateria. In

20

21

22

23

24

25

26

T. KatayAmMa, M. YAMAmoTo et al.

“The Origins and Relationships of Lower Inverte- brates”. Eds. by S C Morris, JD George, R Gibson, HM Platt, Clarendon Press, Oxford, pp 101-122 Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi R, Horn GT, Mullis KB, Erlich HA (1988) Primer- directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239: 487— 491

Saitou N, Nei M (1987) The neighbor-joining method: a new method for reconstructing phy- logenetic trees. Mol Biol Evol 4: 406-425

Sanger F, Nicklen S, Coulson AR (1977) DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74: 5463-5467

Smith JPS, Teyler S (1985) The acoel turbellarians: kingpins of metazoan evolution or a specialized offshoot? In “The Origins and Relationships of Lower Invertebrates”. Eds. by SC Morris, JD George, R Gibson, HM Platt, Clarendon Press, Oxford, pp 123-142

Smith JPS, Teyler S, Rieger RM (1986) Is the turbellaria polyphyletic? Hydrobiologia 132: 13-21 Wada H, Makabe KW, Nakauchi M, Satoh N (1992) Phylogenetic relationships between solitary and col- onial ascidians, as inferred from the sequence of the central region of their respective 18S rDNAs. Biol Bull 183: 448-455

Willmer P (1990) Invertebrate Relationships. Cam- bridge Univ. Press, Cambridge

ZOOLOGICAL SCIENCE 10: 537-542 (1993)

[RAPID COMMUNICATION]

© 1993 Zoological Society of Japan

The Pattern of Volatile Compounds in Incubated and Fresh Preputial Fluid of Male Mice

Kyoxo Ninomtya!, Isso NOHARA?, TAKAAKI ToYoDA”

and Takes! Kimura!

'Department of Biology, College of Arts and Sciences, University of Tokyo, Komaba, Tokyo 153, and *Central Research Center, Takasago International Corporation, Kamata, Tokyo 144, Japan

ABSTRACT—The volatile compounds from incubated and fresh preputial fluid of male mice were analyzed by capillary gas chromatography-mass spectrometry. We found quantitative and qualititative changes in the volatile compounds of preputial fluid after incubation. These chemical changes in the incubated preputial fluid seem to be necessary for female attraction.

INTRODUCTION

Both androgen-dependent compounds in the urine and androgen-independent compounds from preputial glands which are found in excreted urine of male mice have been shown to be necessary for female attraction [4, 5]. We have also demon- strated that, when mixed with preputialectomized male urine, incubated preputial gland fluid, which may have been processed by oxidation, bacterial action or other chemical changes after excretion, is significantly more attractive to females than fresh preputial fluid [4], but the chemical nature of the male’s preputial attractant factor remains un- known. Lipid extracts of the preputial gland secretions of male mice contain a number of long-chain alkyl acetates which are testosterone dependent [7]. Recently Novotny’s group re- ported that the two major compounds in the fresh preputial fluid of male mice are E,E-a-farnesene and E-#-farnesene. These compounds are found at high levels in the urine of dominant male mice and

Accepted April 21, 1993 Received March 19, 1993

are aversive to subordinate males [6].

In the present study, the major volatile com- pounds of both fresh and incubated preputial fluid from male mice were analyzed by capillary gas chromatography-mass spectrometry(GC-MS) and the distribution of these compounds was com- pared.

MATERIALS AND METHODS

Fluid was taken directly from the preputial glands of 11 male ICR/JCL mice at 8 weeks of age. These mice were purchased from Nihon Clea Co. (Tokyo) at 4 weeks of age and housed in groups of 5 or 6 in 24X17X12 cm plastic cages in a room which was maintained at about 23°C and kept on a 14hr L: 10hr D photoperiod. They received OA-1 mice chow (Nihon Clea Co.) and water ad libitum. Mice were anesthetized with sodium pentobarbital (Abbott) and their preputial glands were surgically removed. The glands were then cut open and the fluid was stored in glass vials. Half of the preputial fluid was incubated at 37°C for two days, and the other half was kept at minus 80°C until being analyzed.

Volatile compounds of incubated and frozen preputial fluid were analyzed using the head space technique, which employed a porous polymer (Tenax TA) as a collection medium. The volatiles were sparged from 0.1-ml preputial samples at room temperature with purified nitrogen gas, and absorbed onto a precolumn packed with Tenax TA

538 K. NINOMIYA

for 2hr. The volatiles were desorbed by heating Tenax TA at 250°C for 6 min with herium gas, and then retrapped in the portion of the capillary column, which was cooled by dry ice acetone, to concentrate the vapor before GC-MS analysis. A chemical bonded PEG-20M column (0.25 mm ID x50 m) was used as a analytical column. Column temperature was programmed from 70° C to 220°C at a rate of 4°C per min. Volatile constituents were identified with a Hitachi M-80B mass spectrometer connected to a gas chromatograph (Hewlett Pack- ard HP-5790GC), using electron impact ionization at 20 eV.

RESULTS AND DISCUSSION

The major volatile compounds found in incu- bated and fresh preputial fluid are compared in Figure 1. The initial part of each chromatogram is enlarged in Figures 2 and 3 in order that each peak can be clearly labelled. Identified compounds, listed in Table 1, were categorized into three groups: those common to both incubated and fresh preputial fluid (Group C), those found only in incubated preputial fluid (Group I) and those found only in fresh preputial fluid (Group P). The C-14 peak of fresh preputial fluid in Figure 1 was contaminated with an unidentified compounds. The C-20 peak of fresh preputial fluid in Figure 1 was contaminated with benzyl acetate. The I-39 peak appears to be an isomer of tetradecenyl acetate. It seems possible that diethyl phthalate (1-41), benzoic acid (C-37) and dibutyl phthalate (C-38) originated from the porous polymer used as the collection medium.

The first of the two major peaks in the incubated preputial fluid (C-21) was identified as E-f- farnesene, and the second (C-24) as E,E-a- farnesene. Although these sesquiterpenic com- pounds are also known to be elevated in the urine of dominant male mice [1], there are several reasons why it does not seem likely that they are necessary for female attraction. First, although these two compounds were found in incubated and fresh preputial fluid at almost the same concentra- tion, fresh preputial fluid is not effective for female attraction [4]. Second, Jemiolo et al. [2] demon- strated that these compounds must be presented at

, I. NoHARA et al.

50 to 100 times higher in concentration than their natural level to attract virgin females. Such high concentrations should not be necessary for female attraction because sex attractant factors work at a natural low concentration. Third, we found these compounds also in female preputial (clitoral) glands (unpublished data). Although the ex- perimental analysis is not yet completed, it seems that these sesquiterpenes may not be important for female attraction because they are found in both male and female preputial glands. These facts suggest that compounds other than E-@-farnesene and E,E-a-farnesene must play an important role in producing the female attractant factors of male preputial glands.

In some of the compounds in Group C, we found quantitative changes between incubated and fresh preputial fluids. For example, the concentra- tion of hexadecyl acetate (C-36), which is reported as a major constituent in lipid extracts of preputail glands [7], was greatly increased after incubation. Other compounds including some aldehydes (hep- tanal, octanal, £-2-heptenal, etc), terpene (cymene, limonene, etc), alkyl acetates (amyl acetate, hexyl acetate, etc) are found only in incubated preputial fluid. The highest peak (I-7) in these compounds was identified as limonene. Limonene (1-7, I-8) has been identified as an alarm pheromone of termites [2], but the behavioral effects of limonene in mice have not been investi- gated.

The present study provides several possible ways that preputial factor may influence female attrac- tion. First, the new compounds which are pro- duced after incubation (Group I) might be effec- tive. Second, the quantitative changes in volatile compounds during the process of incubation might be important. Moreover, the combination of the quantitative changes as those observed in C-28, C-35, C-36 and the aforementioned qualitative changes might be necessary for female attraction. We are presently comparing the volatile com- pounds of the urine of preputialectomized males with those of intact males to explore these possibil- ities.

Preputial Fluid of Male Mice

ge-9 ——— Z o> (oe) — + Ww 2 vp-l ra Le-9 wae by-l 96-9 —— se-9 —— np) Olas 8e-I =a) = ° ee-9——_— S ce-9 ——— Ss)

pt-9 — SG vt-I—— Lz-9 — Lz-9 — 61-9 —— ° =) - Zta-4-1 LL-O-4-9 l-d-}-d 8-9-+4-9 <q a

CO SO ee

z o2 a LU = = ° fo) (50) = (e) N = fe) =

Gas-chromatographic profiles of (A) preputial fluid from male mice following 2 days of incubation, and (B) fresh preputial fluid from male mice. The section between the arrows are enlarged in Fig. 2 and 3. Peak numbers

Fic. 1.

539

For details see text.

refer to Table 1.

K. Ninomiya, I. Nowara et al.

540

"Ty ons

ul ping yenndoid ysory (q) Jo uoNn1od podiejuq “¢ “‘DI4 ‘T oinsiy ur pny yeyndoid poyeqnout (vy) Jo uoniod posiejuq -°Z “514

ANAT I | ae I wal AE she Me HA RES Bah Ker [x8 ll W/E ol oks “Ltd ° ab Ba In Ir “ | | lei ~ : {}}° oi 1 v I : te : S| V9 iS ° 5 Pr = | | = It | 2 4 i We > i = o T S| N

Preputial Fluid of Male Mice

541

TaBLeE 1. Compounds found in both incubated and fresh preputial fluid (Group C), only in incubated preputial fluid (Group I) or only in fresh preputial fluid (Group P)

Group C Group I Group P

Peak Compounds Peak Compounds Peak Compounds

C-1 octane I-1 chroloform P-1 methanol

C-2 nonane I-2 decane P-2 ethanol

C-3 is-pentanal I-3 butanol P-3 1-buten-3-one+ benzene C-4 toluene I-4 myrcene P-4 propanol -+2-methyl-3-buten-2-ol C-5 is-butylacetate I-5 amyl acetate P-5 3-xylene

C-6 hexanal I-6 heptanal P-6 2-xylene

C-7 4-xylene I-7 limonene P-7 3-methyl-2-buten-2-ol C-8 ethylbenzene I-8 limonene P-8 4-cymene

C-9 octan-2-one I-9 amyl alcohol P-9 Z-3-hexenyl acetate C-10 6-methyl-5-hepten-2-one I-10 styrene P-10 1,4-dichlorobenzene C-11 hexanol I-11 hexyl acetate P-11 valeric acid

C-12 acetic acid I-12 octanal P-12 geraniol

C-13 heptanol-+ furfural I-13 tridecane P-13 ethyl laurate

C-14 2-ethylhexanol I-14 E-2-heptanal P-14 -ionone

C-15 benzaldehyde C-16 linalool+?

C-17 octanol

C-18 butan-1,4-olide C-19 acetophenone C-20 cyclohexyl is-thiocyanate C-21 E-f-farnesene C-22 dodecanal

C-23 Z,E-a-farnesene C-24 E,E-a-farnesene C-25 tridecanal

C-26 caproic acid C-27 dodecyl acetate C-28 tetradecanal C-29 dodecanol

C-30 phenol

C-31 capryic acid C-32 tetradecyl acetate C-33 tetradecenyl acetate C-34 nonanoic acid C-35 capric acid

C-36 hexadecyl acetate C-37 benzoic acid C-38 dibutyl phthalate

I-15 I-16 I-17 I-18 I-19 I-20 I-21 I-22 I-23 I-24 I-25 I-26 1-27 I-28 21-9 I-30 I-31 I-32 I-33 I-34 1-35 1-36 I-37 I-38 I-39 1-40 I-41 I-42 I-43 1-44

heptyl acetate nonan-2-one nonanal tetradecane 4-is-propenyltoluene octyl acetate 2-decanone propionic acid nonyl acetate 2-undecanone nonanol carvone pentan-1,5-olide E,E-2,4-decadienal 2-tridecanone octan-1,4-olide benzothiazole enathic acid 10-undecanol nonan-1 ,4-olide cinnamaldehyde pentadecanal 4-cresol tetradecanol

tetradecenyl acetate?

decanoic acid diethyl phthalate undecanoic acid hexadecanol lauric acid

P-15 diethyleneglycol P-16 nerolidyl formate

542 K. Ninomiya, I. Nowara et al.

ACKNOWLEDGMENTS

We wish to thank Dr. Richard E. Brown and Ms. Heather Schellinck, Dalhousie University, for their helpful comments on the manuscript. KN is the recipient of a JSPS Research Fellowship. This study was partly supported by Grants-in Aid (Nos. 63540597 and 02640579) from the Japan Ministry of Education, Science and Culture.

REFERENCES

1 Harvey S, Jemiolo B, Novotny M (1989) J Chem Ecol 15: 2061-2072

2

3

Jemiolo B, Xie TM, Novotny M (1991) Physiol Behav 50: 1119-1122

Lindstrém M, Norin T, Valterova I, Vrkoc J (1990) Naturwissenshaften 77: 134-135

Ninomiya K, Kimura T (1988) Physiol Behav 44: 791-795

Ninomiya K, Kimura T (1990) Naturwissenschaften 77: 586-588

Novotny M, Harvey S, Jemiolo B (1990) Experientia 46: 109-113

Spener F, Mangold HK, Sansone G, Hamilton JG (1969) Biochim Biophys Acta 192: 516-521

ZOOLOGICAL SCIENCE 10: 543-546 (1993)

[RAPID COMMUNICATION]

© 1993 Zoological Society of Japan

Neural Control of Muscle Differentiation in the Leg of the Fleshfly Sarcophaga bullata

PAKKIRISAMY SIVASUBRAMANIAN

Department of Biology, University of New Brunswick, Fredericton, Canada E3B 6EI

ABSTRACT—The tubular leg muscles of an adult fly differentiate from embryonic adepithelial cells present within the leg imaginal discs. Transection of the disc stalk from the ventral ganglion in a prepupa results in the formation of an externally normal looking leg that lacks locomotor muscles. The absence of motor axons in the nerve of in the experimental leg is suggestive of the importance of motor innervation in the development of leg muscles.

INTRODUCTION

Several studies using both in vitro and in vivo procedures have demonstrated the importance of innervation in the differentiation of vertebrate muscles [3, 6, 9, 14, 15, 19]. Among invertebrates insects have received special attention in this regard. The influences of nerves on insect muscle differentiation was first demonstrated by Kopec [8]. When he removed thoracic ganglia from gypsy moth caterpillars the resulting adults were devoid of thoracic muscles. Similar results were also obtained for silkmoths [11, 21]. By a more refined procedure Sivasubramanian and Nassel [18] re- cently demonstrated the role of mesothoracic larval nerve in the differentiation of thoracic flight muscles in the fly, Sarcophaga bullata. Although these studies did indicate the importance of in- nervation for the differentiation of flight muscles, additional information can be obtained from ex- amining leg muscle differentiation in holometabo- lous insects such as flies which arise de novo from

Accepted April 12, 1993 Received February 4, 1993

bags of embryonic cells called imaginal discs [4, 13, 20]. This is in contrast to the thoracic flight muscles that differentiate using the histolysing larval muscles as templates [5]. Further, the imaginal discs are attached to the larval ventral ganglion by means of stalks which also contain nerves that apparently act as guides for the growing adult axons between the legs and the ventral ganglion [1]. In larger species of flies these disc stalks along with their guiding pioneer axons can be surgically severed. The present study examines the legs developed from stalk transected imaginal discs of the fleshfly, Sarcophaga bullata. The results strongly suggest the importance of motor innervation for leg muscle differentiation.

MATERIALS AND METHODS

The fleshfly Sarcophaga bullata was reared in the laboratory under constant conditions of tempera- ture (25°C) and photoperiod (16L:8D). Adults were fed with ample supply of sugar and water. Larvae were raised in fresh beef liver. Post- feeding, mature third instar larvae were collected from culture containers and used for experiments. All operations were performed on freshly formed prepupae (1-2 hr post pupariation). Mesothoracic legs lacking motor innervation were produced according to the method of Sivasubramanian and Nassel [17]. Since the stalk of the leg disc contains axons that presumably act as pioneers to guide the adult nerves disconnecting the stalk deprives the growing motor neurons from finding their targets. A “V” shaped incision was made on the anterior

544 P. SIVASUBRAMANIAN

Fic. 1. Schematic diagram showing the lateral view of the larval brain (Br) and ventral ganglion (Vg). M and P are, respectively, mesothoracic and protho- racic leg imaginal discs. The arrow depicts the location of stalk transection.

ventral side (on segments 5 and 6, i.e., 2nd and 3rd segments after anterior spiracles) of freshly formed prepupa. The triangular flap was lifted up and the exposed stalk of left mesothoracic leg disc was transected. The site of transection is shown in Figure 1. The flap was left back on the slit and the dried hemolymph closed twe wound. The oper- ated specimens kept in humid chambers at 25°C. metamorphosed into normal looking flies in 12 days.

Twenty experimental (left mesothoracic) legs were examined histologically and compared with normal, right mesothoracic legs. The femurs of the

legs were embedded in epon and 2 um thick plastic sections were stained with toluidine blue.

RESULTS AND DISCUSSION

Morphologically, the legs developed from stalk transected imaginal discs appear quite normal like the control legs (Fig. 2A and C) with full comple- ment of segments, bristles etc. However, these experimental legs exhibit very little movement. The flies simply drag these legs while walking. Examination of histological sections reveal the reason for the absence of motor function in these legs. A cross section of the femur of a normal (right) mesothoracic leg is shown in Figure 2B. It is filled with tibial levator and depressor muscles. Compared to this the experimental leg is devoid of both sets of these muscles (Fig. 2D). There is a

nerve bundle in the operated leg (Fig. 2D, arrow). Our previous study of horseradish peroxidase fillings of such stalk transected legs [17] as well as the absence of any movement in these legs suggest that this nerve is probably composed mostly sensory axons. However, these histological and behavioral observations cannot rule out the pre- sence of few, albeit, nonfunctional motor axons in the nerve bundle.

Within the imaginal discs there are irregular shaped adepithelial cells which are the progenitors of adult leg muscles [13]. During metamorphosis, these cells organize themselves into columns, become multinucleate and eventually differentiate into the tubular muscles of the leg [16]. Are these adepithelial cells dependent on motor innervation for their differentiation?

The results of earlier studies were ambiguous about the role of innervation on leg muscle differentiation. When Drosophila mesothoracic leg discs were cultured in vivo muscle specific enzymes were expressed in the transplanted leg discs [20]. However, when the discs were cultured in vitro muscles failed to differentiate in the otherwise morphologically normal looking legs [9]. The present study used a different approach to examine the same problem. Here an in situ leg disc was deprived of motor innervation by transecting the disc stalk prior to metamorphosis [17]. When the stalk was selectively severed leaving the disc’s attachment with the epidermis an externally nor- mal looking leg developed. It is quite conceivable that the absence of muscle in the stalk transected legs is due to the lack of motor axons in the nerves of these legs.

Myogenesis is dependent on motor innervation in chick embryos [3]. Destruction of motor neurons by bungarotoxin inhibits myotube forma- tion in rat skeletal muscles [6]. Presumably, the motor neuron terminals have some trophic or inductive influence on myogenic cells [15]. In grasshopper embryos motor neurons contact mus- cle pioneers very early in development [2] and perhaps this contact influences them from the very beginning. Precisely what stage of leg muscle differentiation in flies is controlled by motor innervation is not clear at the moment. However, preliminary screening of histological preparations

Leg Muscle Differentiation in a Fly 545

Fic. 2. Control (A) and stalk transected (C) legs indicating the approximate site of cross section (arrow heads). B. Cross section of a normal leg with levator (le) and depressor (de) muscles. D. Cross section of a stalk transected leg. Note the leg is completely devoid of muscles. Both legs have a leg nerve (arrow).

of developing stalk transected legs indicates that the proliferation of adepithelial cells (myoblast ACKNOWLEDGMENTS precursors) is unaffected implying the post mitotic This sesoaa wes smppoded by a qeut Rom dhe

events of myogenesis to be highly dependent on _ Natural Sciences and Engineering Research Council of motor innervation. Canada.

546

REFERENCES

Auerbach C (1936) Trans Roy Soc Edin 58: 787-815 Ball EE, Ho RK, Goodman CS (1985) J Neurosci 5: 1808-1819

Bonner PH, Adams TR (1982) Develop Biol 90: 174-184

Crossley AC (1978) In “The Genetics and Biology of Drosophila Vol2b” Ed by M Ashburner, TRF Wright, Academic Press, pp 499-560

Fernandes J, Bate M, Vijayraghavan K (1991) Development 113: 67-77

Harris AJ (1981) Phil Trans Roy Soc Lond Ser B. 293: 257-277

Jan YN, Ghysen A, Christoph I, Barbel S, Jan LY (1985) J Neurosci 5: 2453-2464

Kopec C (1923) J Exp Zool 37: 15-25

Mandaron P (1970) Develop Biol 22: 298-320 McLennan IS (1983) Develop Biol 98: 287-294

Nassel DR (1983) In “Functional Neuroanatomy” Ed by NJ Strausefeld, Springer Verlag, pp 44-91

12

13 14 15 16 17 18 19

20

21

P. SIVASUBRAMANIAN

Nutesch H (1985) In “Comprehensive Insect Physiol- ogy, Biochemistry and Pharmacology Vol 2” Ed by G Kerkut, LI Gibert, Pergamon Press, pp 425-452 Poodry CA, Schneiderman HA (1970) Roux’s Arch Dev Biol 166: 1—44

Popiela H (1978) Exp Neurol 62: 404-416

Popiela H, Ellis S (1981) Develop Biol 83: 266-277 Reed CT, Murphy C, Fristrom D (1975) Roux’s Arch Dev Biol 178: 285-302

Sivasubramanian P, Nassel DR (1989) Develop Growth Diff 31: 341-349

Sivasubramanian P, Nassel DR (1990) Zool Sci 7: 223-228

Sohal GS, Holt RK (1980) Cell Tiss Res 210: 383- 393

Ursprung HM, Conscience-Egli M, Fox DJ, Walli- mann T (1972) Proc Nat Acad Sci USA 69: 2812- 2813

Williams CM, Schneiderman HA (1952) Anat Rec 113: 560-561

D eve | O m e n t Published Bimonthly by the Japanese Society of Pp Developmental Biologists

e ° ge Distributed by Business Center for Academic Growth & Differentiation Societies Japan, Academic Press, Inc.

Papers in Vol. 35, No. 3. (June 1993)

26. REVIEW: T. Kuwana: Migration of Avian Primordial Germ Cells toward the Gonadal Anlage

27. M. Takenaka, A. Suzuki, T. Yamamoto, M. Yamamoto and M. Yoshida: Remodeling of the Nauplius Eye into the Adult Ocelli during Metamorphosis of the Barnacle, Balanus amphitrite hawaiiensis

28. L. Bosco, C. Valle and D. Willems: Jn Vivo and in Vitro Experimental Analysis of Lens Regeneration in Larval Xenopus laevis

29. A. Yoshiki, K. Moriwaki, T. Sakakura and M. Kusakabe: Histological Studies on Male Sterility of Hybrids between Laboratory and Wild Mouse Strains

30. Y. Kamata, S. Furuya and I. Yasumasu: Proteins ADP-ribosylated in Nuclei and Plasma Membrane Vesicles Isolated from Sea Urchin Embryos at Various Stages of Early Development

31. K. Shimizu and K. Yoshizato: Involvement of Collagen Synthesis in Tissue Reconstitution by Dissociated Sponge Cells

32. K. Hayashi, Y. Hagiwara and E. Ozawa: Vimentin Expression Pattern is Different between the Flank Region and Limb Regions of Somatopleural Mesoderm in the Chicken Embryo

33. S. Takiya and Y.i Suzuki: Role of the Core Promoter for the Preferential Transcription of Fibroin Gene in the Posterior Silk Gland Extract

34. Y. Shiroya and Y. T. Sakai: Organization of Actin Filaments in the Axial Rod of Abalone Sperm Revealed by Quick Freeze Technique

35. Y. Nagasawa, S. Kurata, H. Komano and S. Natori: Purification and Heterogeneous Localization of Sarcophaga Lectin Receptor on the Surface of Imaginal Discs of Sarcophaga peregrina (Flesh Fly)

36. D.C. Ludolph, J.-A. Cameron, A. W. Neff and D. L. Stocum: Cloning and Tissue Specific Expression of the Axolotl Cellular Retinoic Acid Binding Protein

37. S. Yonezawa, Y. Maejima, N. Hagiwara, T. Aratani, R. Shoji, T. Kageyama, S. Tsukada, K. Miki and M. Ichinose: Changes with Development in the Expression of Cathepsin E in the Fetal Rat Stomach

38. R. Créton, G. Zwaan and R. Dohmen: Specific Developmental Defects in Molluscs after Treatment with Retinoic Acid during Gastrulation

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.

ergonomic design.

aes OLYMPUS

Y-‘E:S is the answer.

The BX Series was developed with your needs and our ideals in mind.

Y-shaped ergonomic design ensures maximum comfort. Easy operation. Easy recognition. And fatigue-free use. To the question of what SYSTEM MICROSCOPE constitutes the ideal microscope from the user’s point of view, our answer is a unique Y-shaped design. We invite you to discover unprecedented operational ease and comfort.

Excellent optics provide superior image quality.

Developed through the application of more than 70 years of experience and what the latest technology has to offer, Olympus’ own UIS (Universal Infinity System) infinity-corrected optical system assures exceptional image quality. You benefit BX40/BX50 |

from both higher contrast and outstanding flatness throughout the field of view.

System versatility answers diverse research requirements. NEW Frame rigidity and stability have been increased dramatically through use of

computer simulation (FEM analysis). You'll find that reliability has been enhanced,

too, when the time comes for system expansion, via devices for image processing,

microscopic light metering, photography and so on.

OLYMPUS OPTICAL CO.,LTD. OLYMPUS (JAPAN) CO.,LTD. For free catalog, please contact: OLYMPUS (JAPAN) CO.,LTD. Ryumeikan Bldg., 3-4, Kanda-surugadai, Chiyoda-ku, Tokyo Tel. (03) 3251-8971

The Ultimate Name in Micromanipulation Ease of operation, and the most advanced

Can be combined with New WR & MX series E

Model MX-2 3-D Micromanipulator

Model MxX-1 3-D Micromanipulator

Model WR-87

One Dimensional Aqua Purificate Micromanipulator

Model SR-6 Model PP -83 Stereotaxic Instrument for Rat Glass Microelectrode Puller

* * %& Send enquiries for request for MODIFICATIONS or IMPROVEMENTS * * * Physiological, Pharmacological, Zoological & Neurosciences Research Equipments

THARISHIGE) NARISHIGE SCIENTIFIC INSTRUMENT LAB.

\ JBEAS i) 9-28 KASUYA 4-CHOME SETAGAYA-KU, TOKYO 157, JAPAN PHONE (INT-L)81-3-3308-8233, FAX (INT-L) 81-3-3308-2005

CABLE:NARISHIGE LABO, TELEX, NARISHIGE J27781

ee ae Se

Se = oe " 7 : ae tt Ting S ~~ GPs 62a) ® ¢ 2H aE ni emBlt eben oi ge isVvbOs Jezom ert pres 08999 °. 48k

ey

ya mn 7

a

; tow wecele ‘ ae |e ‘ a ie —— Th, ar a

Ped a ES Rina \ ViA@A tu

wo

(Contents continued from back cover)

of male mice (RAPID COMMUNICA- TION)

Enviromental Biology Ip, Y. K., C. Y. Lee, S. F. Chew, W. P. Low, K. W. Peng: Differences in the responses of two mudskippers to terrestrial exposure

Numata, H., A. H. Saulich, T. A. Volkovich: Photoperiodic responses of the linden bug,

Pyrrhocoris apterus, under conditions of con- stant temperature and under thermoperiodic CONGILIONS <yee fete he sferrsaiseere te eae arcreprsseiencters 521

Systematics and Taxonomy Katayama, T., M. Yamamoto, H. Wada, N. Satoh: Phylogenetic position of acoel tur- bellarians inferred from partial 18S rDNA SEQUENCES ise e navstsrsciste suyrsran es sstasrate sree are 529

ZOOLOGICAL SCIENCE

VOLUME 10 NUMBER 3

JUNE 1993

CONTENTS

REVIEWS Ooka, H.: Proliferation of anterior pituitary

cells in relation to aging and longevity in the Tat sotanedeccm ern deners arene Meee saci 385

Ricci, N., R. Banchetti: The peculiar case of the giants of Oxytricha bifaria (Ciliata, Hypotrichida): A paradigmatic example of cell differentiation and adaptive strategy

ORIGINAL PAPERS

Physiology Ohnishi, K.: The defective color vision in juvenile goldfish does not depend on used training task as the measure of discrimina- tion : A two- choise response measure ...411 Mikami, K., S. Sato, M. Otokawa, A. Palum- bo, K. Kuwasawa: Cardiotoxic effects of adrenochrome and epinephrine on _ the mouse cultured myocardium Suzuki, T., K. Narita, K. Yoshihara, K. Nagai, Y. Kito: Immunochemical detection of GTP-binding protein in cephalopod photore- ceptors by anti-peptide antibodies

Cell Biology

Kurita, T., H. Namiki: Serum-induced cell death

Developmental Biology Matsushita, S.: Effect of the oesophageal mesenchyme on the differentiation of the digestive-tract chick embryo

endoderm of the

Sivasubramanian, P.: Neural control of mus- cle differentiation in the leg of the fleshfly Sarcophaga bullata (RAPID COMMUNI-

CATION)

Reproductive Biology

Koike, S., T. Noumura: Immunohisto- chemical expression of inhibin-a subunit in the developing rat gonads Furuya, H. K. Tsuneki, Y. Koshida: The development of the hermaphroditic gonad in four species of dicyemid mesozoans

Endocrinology Tanaka, S.: Hormonal deficiency causing albinism in Locusta migratoria Duan, C., N. Hanzawa, Y. Takeuchi, E. Hamada, S. Miyachi, T. Hirano: Use of primary cultures of salmon hepatocytes for the study of hormonal regulation of insulin- like growth factor I expression in vitro ...473 Sawada, K., T. Noumura: Metabolism of androgens by the mouse submandibular gland and effects of their metabolites ....481 Jana, N. R., S. Bhattacharya: Binding of thyroid hormone to freshwater perch leydig cell nuclei-rich preparation and its functional

relevance: <.4..4. 9a .dcs eae 489 Shiraishi, K., T. Singtripop, M. K. Park, S. Kawashima: Developmental changes of

testicular gonadotropin receptors and serum gonadotropin levels in two strains of mice

Bhavior Biology The be- havioral reactions of a snake and a turtle to

Wassersug, R., A. Izumi-Kurotani:

abrupt decreases in gravity

Ninomiya, K., I. Nohara, T. Toyoda, T. Kimura: The pattern of volatile com- pounds in incubated and fresh preputial fluid

(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 June 15 Printed by Daigaku Letterpress Co., Ltd., Hiroshima, Japan

ih

¢

ie fone ytd

EH ie

ve aw be

Malas ea

has

a cried

tat

Hehe die dey

¥ we

“wyare at

Il

SMITHSONIAN INSTITUTION LIBRARIES

ii

3 9088 01261 2784