iy son » Cire dime eecadgeh kl het i i ens BHM Pagegsitecs Aes ae 4 } bares & Vata resee Vath ts st mk eee: 23 i | ste | seat | eres 13 AA 4 hetero 8 tot aa Nears Mele! ye nt ; : one tLe ara eat NZ Got oyigiae Gut oA ths Ae eee’ he hat RELSCYS,, Bye: ssa hing’ the Se eNeost ks ; | : 2 Newnan : ‘ Leyva : | : | ener 7 fhe esl & ea NN Ae RES voaeken' i com Nase Seether : { Hina : Vesta Sv ine ‘ Sees ste (EEO Y Aba STS A ye Se se Be CREATAS OE adem ahs rarer en! Fae ake yepue qu arena Tepes : Gra Meet e CARS Line eae pn u tes : : Dein as eae aenee i ta : thy sue vite Apes RR EEOE Vr Size pup sen ; ‘ sf iSeiafe Ra tie tren Rae oe Hee eT Pty aay net Lae Lad ¢ cy Being aay oe ice) ‘ Fey aie : j ot : : cya : ; : 1 < (Sie tain ty Mh peas oh ei rete EE reget! peg Conc yn eis Sr erkon Do Ced ragh Tete yeas COUN UNIO 4 Bitte ath Py bor dows ¥ Seka a Say 197 “Ty ha ynecin! f, ne peter ee . | ieee a nln Pete é Tt ae LE Hearse ate ise tp tN “weit Beals ene ay, " t { een is sera ; ea tty Bt Nrive se ; i z ) Tas at i : a yey eevee wR he ap ior rye sido Seine sere sypcoaye oe Oe NEw Renee Cpa ees chyumecn yan WS nS ae open et ete sya esta” i Ue spta pee sh i - ZOOLOGICAL SCIENCE An International Journal VOLUME 9 Og, published by The Zoological Society of Japan CONTENTS OF VOLUME 9 NUMBER 1, FEBRUARY 1992 REVIEWS Morita, H.: Transduction process and im- pulse initiation in insect contact chemorecep- CO EE fe cfd Rio coh esa esol ait covnln ie gouyiviee cia os 1 Ugrumov, M. V.: Development of the hypothalamic monoaminergic system in on- togenesis. Morpho-functional aspects ...17 ORIGINAL PAPERS Physiology Okumura, T., C. Han, Y. Suzuki, K. Aida and I. Hanyu: Changes in hemolymph vitel- logenin and ecdysteroid levels during the reproductive and non-reproductive molt cy- cles in the freshwater prawn Macrobrachium UD DOMENS CIMA AA whe, CRED, Vo eh) 31) Azuma, K. and N. Iwasaki: Effects of Ba?* on the photoresponses of isolated single rods POMMUOC MEUM A: - rnin canara sd oo etes ol ees 47 Hirata, J. and H. Michibata: Electron spin resonance spectrometry of vanadium ions in the blood cells of the ascidian, Ascidia gem- mata (COMMUNICATION) Hara, T., R. Hara, A. Kishigami, Y. Koshida, S. Horiuchi, M. Yoshida, M. Yamamoto, T. Goto and U. Raj: Localization of re- tinochrome in the retina of a tetrabranchiate cephalopod, Nautilus pompilius (COM- MUNICATION) Yasuyama, K., T. Kimura and T. Yamaguchi: Proctolin-like immunoreactivity in the dorsal unpaired median neurons innervating the accessory gland of the male cricket, Gryllus bimaculatus Cell Biology Nagaishi, H. and N. Oshima: Ultrastructure of the motile iridophores of the neon tetra Sato, I.: Origin of multinucleality of muscle cells during myogenesis of the newt Yano, J.andM.Suhama: Effect of actinomy- cin D on nuclear events during conjugation in the ciliate Stylonychia pustulata Genetics Kosaka, T.: Autogamy and autogamy inheri- tance in Euplotes woodruffi syngen 1 (Ciliophora) Immunology Zhang, H., Z. Huang, K. Yamaguchi and S. Tomonaga: Granulocytes and macropha- ges in amphioxus Biochemistry Iino, T., K. Dohke and M. Tsusue: The Purification and characterization of sepiap- terin reductase from fat body of the silkworm Bombyx mori Nakauchi, U. and K. Maruyama: Immunob- lot detection of vertebrate-type of connectin (titin) in ascidian bodywall muscle and tad- pole (COMMUNICATION) Developmental Biology Satomi, D.: Developmental changes of gluta- mate decarboxylase and 2’,3-cyclic nuc- leotide 3’-phosphodiesterase in the organ- otypic culture of newborn mouse cerebellum Watanabe, M.: Egg maturation in labora- tory-reared females of the swallowtail but- terfly, Papilio xuthus L. (Lepidoptera: Papi- lionidae), feeding on different concentration solutions of sugar Endocrinology Yada, T. and T. Hirano: Influence of seawa- ter adaptation on prolactin and growth hor- mone release from organ-cultured pituitary of rainbow trout Fukuda, M., R. Hashimoto, K. Yamanouchi, Y. Arai, F. Kimura and M. Takada: ll Effects of unilateral hypothalamic lesion on serum gonadotropin in hemiovariectomized rats (COMMUNICATION) Kiriishi, S., H. Nagasawa, H. Kataoka, A. Suzuki and S. Sakurai: Comparison of the in vivo and in vitro effects of bombyxin and prothoracicotropic hormone on prothoracic glands of the silkworm, Bombyx mori... 149 Suzuki, M.,S. Hyodo and A. Urano: Cloning and sequence analyses of vasotocin and iso- tocin precursor cDNAs in the masu salmon, Oncorhynchus masou: evolution of neurohy- pophysical hormone precursors Takahashi, S., K. Okamoto, H. Sonobe, M. Kamba and E. Ohnishi: Jn vitro synthesis of ecdysteroid conjugates by tissue extracts of the silkworm, Bombyx mori Iwasawa, A., S. Tanaka, Y. Hanaoka and K. Wakabayashi: Development and applica- tion of time-resolved fiuoroimmunoassay for gonadotropin of a wide range of amphibian species Behavior Biology Chiba, Y. and K. Tomioka: Entrainability of diphasic circadian activity of the mosquito, Culex pipiens molestus to 24-hour light-dark cycles: a physiological significance of critical light-to-dark ratio Enviromental Biology Chew, S. F. and Y. K. Ip: Tolerance of the mudskipper, Boleophthalmus boddaerti, to a lack of oxygen (COMMUNICATION) Taxonomy Kubota, S.: Eucheilota intermedia Kubota is a distinct taxon and the third form of Eutima Japonica Uchida (Hydrozoa; Leptomedusae) (COMMUNICATION) Matsui, M., T. Sato, S. Tanabe and T. Hayashi: Local population differentiation in Hynobius retardatus from Hokkaido: an electrophoretic analysis (Caudata: Hyno- biidae) Nomura, K. and K. Hayashi: Rhynchocinetes striatus, a new species (Decapoda, Caridea, — Rhynchocinetidae) from southern Japan sls elite udeclaes ie. wean 199 ERRATUM >...) 002.5... Se eee 237 Instructions to, Authors” 752.85 eee eee 238 NUMBER 2, APRIL 1992 REVIEWS Kanzaki, R. and T. Shibuya: Olfactory pro- cessing pathways of the insect brain Ogawa, K.: Primary structure and function of audyneimpmotormmoleculeiay sen sass ee 265 ORIGINAL PAPERS Physiology Harmon, J. S. and M. A. Sheridan: Previous nutritional state and glucose modulate gluca- gon-mediated hepatic lipolysis in rainbow trout, Oncorhynchus mykiss Cell and Molecular Biology Kuroda, Y., Y. Shimada, B. Sakaguchi and K. Oishi: Effects of sex-ratio (SR)-spiroplas- ma infection on Drosophila primary embry- onic cultured cells and on embryogenesis Takagi, K. and S. Kawashima: Attempts to improve survival of neurons derived from neonatal rat hypothalamus-preoptic area in serum-free media Ichikawa, T. and K. Ajiki: Development of an in situ hybridization histochemistry for choline acetyltransferase mRNA with RNA probes Biochemistry Harbige, L. S., K. Ghebremeskel, G. Williams and M. A. Crawford: Hepatic fatty acids in wild rockhopper (Eudyptes crestatus) and magellanic (Spheniscus magellanicus) penguins before and after moulting Lawrence, J. M. and P. Moran: Proximate composition and allocation of energy to body components in Acanthaster planci (Linnaeus) (Echinodermata: Asteroidea) Developmental Biology Tanimura, A. and H. Iwasawa: Origin of Somatic cells in Bidder’s organ and the gonad proper in the toad, Bufo japonicus formosus (COMMUNICATION) Fujino, Y., A. Fujiwara, I. Yasumatsu and T. Fujii: Chromogranin A-like proteins in the heat-stable fraction of sea urchin eggs, embryos and the substances secreted with sperm Funakoshi, K., Y. Fukue and S. Tabata: Tooth development and replacement in the Japanese greater horseshoe bat, Rhino- lophus ferrumequinum nippon (COM- MUNICATION) Kearn, G. C., K. Ogawa and Y. Maeno: Hatching patterns of the monogenean para- sites Benedenia seriolae and Heteraxine heter- ocerca from the skin and gills, respectively of the same host fish, Seriola quinqueradiata (COMMUNICATION) Miyata, S., Y. Nishibe, M. Sendai, I. katayama and H. K. Kihara: Changes in timing and site of appearance of a protease in Xenopus embryos Reproductive Biology Takahashi, T., Y. Tsuchiya, Y. Tamanoue, T. Mori, S. Kawashima and K. Takahashi: Occurrence of a novel 350-kDa serine pro- teinase in the fluid of porcine ovarian folli- cles and its increase during their maturation Endocrinology Saad, A. H. and W. Ali: Seasonal changes in humoral immunity and blood thyroxine levels in the toad, Bufo regularis Nakazawa, T. S., T. Machida and S. Kawashi- ma: Effects of unilateral and _ bilateral orchidectomy on laterality of neurons of the preoptic area and plasma levels of gonado- tropins and testosterone in male mice ....357 Matteo, L. D., S. Minucci, M. D’Antonio, S. Fasano and R. Pierantoni: Effects of a gonadotropin-releasing hormone analog ili (HOE 766) on germinal and interstitial com- partments during the annual cycle in the green frog: Rana esculenta Amano, M., K. Aida, N. Okumoto and Y. Hasegawa: Changes in salmon GnRH and chicken GnRH-II contents in the brain and pituitary and GTH contents in the pituitary in female masu salmon, Oncorhynchus masou, from hatching through ovulation Nozaki, M. and A. Gorbman: The question of functional homology of Hatschek’s pit of amphioxus (Branchiostoma belcheri) and the vertebrate adenohypophysis Jacob, M.: In spermatogenesis in Oryctes rhinoceros (Coleoptera, Scara- baeidae): the role of ecdysone and juvenile hormone (COMMUNICATION) Kikuta, T. and H. Namiki: Identification of intracellular localization of laminin in the rat anterior pituitary (COMMUNICATION) vitro Behavior Biology Kasuya, E., T. Kumaki and T. Saito: Vocal repertoire of the Japanese treefrog, Rha- cophorus arboreus (Anura: Rhacophoridae) (COMMUNICATION) Asada, N., K. Fujiwara, H. Ikeda and F. Hihara: Mating behavior in three species of the Drosophila hypocausta subgroup Systematics and Taxonomy Hirose, E., T. Nishikawa, Y. Saito and H. Watanabe: Minute protrusions of ascidian tunic cuticle: some implications for ascidian phylogeny Kubota, S. and T. Horita: A new hydro- medusa of the genus Eirene (Leptomedusae; Eirenidae) from Toba, Japan Furuya, H., K. Tsuneki and Y. Koshida: Two new species of the genus Dicyema (Mesozoa) from octopuses of Japan with notes on D. misakiense and D. acuticephalum NUMBER 3, JUNE 1992 REVIEWS Nagai, Y.: mammals Nunomura, W.: C-reactive protein (CRP) in animals: its chemical properties and biologi- cal functions Primary sex determination in ORIGINAL PAPERS Physiology Kanzaki, R., N. Sugi and T. Shibuya: Self- generated zigzag turning of Bombyx mori males during pheromone-mediated upwind walking Matsuoka, T. and K. Taneda: Step-up and step-down photoresponses in Blepharisma Griffond, B., J. Van Minnen and C. Colard: Distribution of APGWa-immunoreactive substances in the central nervous system and reproductive apparatus of Helix aspersa Cell Biology Suzuki, T. and S. Funakoshi: Isolation of a fibronectin-like molecule from a marine bivalve, Pinctada fucata, and its secretion by amebocytes Khoo, G., V. P. E. Phang and T. M. Lim: The confocal laser scanning microscope as a tool for studying xanthophores of the sword- tail (Xiphophorus helleri) (COMMUNICA- TION) Immunology Zhang, H., T. Sawada, E. L. Cooper and S. Tomonaga: Electron microscopic analysis of tunicate (Halocynthia roretzi) hemocytes Biochemistry Roliti( Es i Suveude;santosyandpls Vande Linares: Phospholipids and fatty acids in intact and regenerating Dugesia anceps, a fresh water planaria (COMMUNICATION) Mita, M.: Diacyl choline phosphoglyceride: the endogenous substrate for energy me- tabolism in sea urchin spermatozoa Developmental Biology Makabe, K. W., S. Fujiwara, H. Nishida and N. Satoh: Failure of muscle myosin heavy- chain gene expression in quarter ascidian embryos developed from the secondary mus- cle lineage cells Kurabuchi, S. and Y. Kishida: Effect of delay in anterior or posterior amputation on regen- eration of short fragments of planaria ....575 Roudebush, W. E. and J. G. Kim: Mouse embryo biopsy: abnormal development with trophoblastic vesicle formation Iwamatsu, T.: Morphology of filaments on the chorion of oocytes and eggs in the meda- kanya: aids isk eee 589 Reproductive Biology Ishijima, S. A., M. Okuno, Y. Nakagome, H. — Odagiri, T. Mohri and H. Mohri: Separa- _ tion of x- and y-chromosome-bearing murine sperm by free-flow electrophoresis: evalu- ation of separation using PCR Endocrinology Zairin, M. Jr., K. Asahina, K. Furukawa and K. Aida: Plasma steroid hormone profiles during HCG induced ovulation in female walking catfish Clarias batrachus Tezuka, Y., H. Kobayashi and H. Uemura: Dipsogenic action of brain natriuretic pep- tide and endothelin-1 in the Japanese quail, Coturnix coturnix Japonica Shirai, M. and Y. G. Watanabe: Effect of brain on proliferative activity of adeno- hypophysial primordial cells in vitro De Jesus, E. G., T. Hirano and Y. Inui: Gonadal steroids delay spontaneous flounder metamorphosis and inhibit T3-induced fin ray shortening in vitro Gobbetti, A., M. Zerani and V. Botte: A possible involvement of prostaglandin E> in the reproduction of female crested newt, Triturus carnifex Tsutsui, K., S. Kawashima, V. L. Saxena and A. K. Saxena: Binding properties and photoperiodic influence of follicle-stimu- lating hormone receptors in the subtropical wild quail George, J. C. and T. M. John: Flight be- Vv haviour and thyroid hormone regulation in homing pigeons (COMMUNICATION) Ecology Smith, J. I. and H. Yu: The association be- tween vocal characteristics and habitat type in Taiwanese passerines NUMBER 4, AUGUST 1992 REVIEWS Matsumoto, A.: Hormonally induced synap- tic plasticity in the adult neuroendocrine AES ORE s, Py. GS ATRAE LA 2s eth ks 679 Bock, W. J.: The species concept in theory MGR RACICS heats. L3G. Wernvoug. Jered. (ae 697 ORIGINAL PAPERS Physiology Naitoh, T. and R. J. Wassersug: The emetic response of urodele amphibians Miro, J. L., S. Araneda and B. Canguilhem: Origin of serotonergic innervation of olfac- tory bulbs in the Europian hamster, Cricetus cricetus: An autoradiographic study using (7H]5-HT retrograde labelling Yamashita, S.: Effect of monochromatic illu- mination of the brain on the phototactic behavior of orb weaving spiders, Argiope amoena and Nephila clavata (COMMUNI- CATION) Developmental Biology Endo, K., S. Ueno, M. Matsufuji and Y. Kakuo: Photoperiodic control of the deter- mination of two different seasonal diphen- isms of the Asian comma butterfly, Polygo- nia c-aureum L. Ukeshima, A.: Scanning electron microscopy of differentiating chick ovaries during embryonic period Murakami, M., I. Iuchi and K. Yamagami: Isolation of intact yolk spheres of fish embryos, which contain the majority of lyso- somal acid phosphatase responsible for yolk phosphoprotein metabolism (COM- MUNICATION) Nakamura, S., R. Kagotani, H. Fujisaki and M. K. Kojima: The acid-insoluble organic matrix of spicules in the sea_ urchin LCHAUCENITOLUS SPU CICTTUIUS re. eee 741 Ohya, Y., K. Watanabe, N. Shimamoto and M. Amano: Scleral fibroblasts of the chick embryo can proliferate without transferrin in protein-free culture Inoue, C. and Y. Kakinuma: Symbiosis be- tween Cytaeis sp. (Hydrozoa) and Niotha livescens (Gastropoda) starts during their larval stage Amemiya, S. and Y. Nakajima: First electron microscopical study on the sperm morpholo- gy of the sea lily (Crinoidea, Echinodermata) (COMMUNICATION) Endocrinology Kobayashi, M., M. Amano, Y. Hasegawa, K. Okuzawa and K. Aida: Effects of olfactory tract section on brain GnRH distribution, plasma gonadotropin levels, and gonadal stage in goldfish Madsen, S. S. and H. A. Bern: Antagonism of prolactin and growth hormone: Impact on seawater adaptation in two salmonids, Salmo trutta and Oncorhynchus mykiss Kai-ya, H., J. Okuyama, T. Ishijima, Y. Sasayama, H. Yoshizawa and C. Oguro: Effects of Ca concentrations in culture medium on the release of calcitonin from incubated ultimobranchial glands of the bull- frog, Rana catesbeiana Kobayashi, Y., S. Kawashima, S. Takahashi and K. Wakabayashi: Effects of chronic Vi treatment with chlorpromazine on the aging of hypothalamo-pituitary-ovarian axis in the WA et cennta cam Pao PES HE ee a cheer 72) Sawada, K. and T. Noumura: Differential effects of testosterone and 5a-dihydrotes- tosterone on growth in mouse submandibu- lar gland Morphology Shirai, S. and K. Nakaya: Functional mor- phology of feeding apparatus of the cookie- cutter shark, Jsistius brasiliensis (Elasmo- branchii, Dalatiinae) Ando, K. and S. Arai: Neuropeptide Y in- nervation of cerebral arteries in microchiro- pteran bats Ecology Matsumoto, T.: Familial association, nym- phal development and population density in the Australian giant burrowing cockroach, Macropanesthia (Blattaria: Blaberidae) rhinoceros Taxonomy Nagatomi, A.: Notes on the phylogeny of various taxa of the orthorrhaphous Brachycera (Insecta: Diptera) Ohtsuka, S., R. Huys, G. A. Boxshall and T. It6: Misophriopsis okinawensis sp. nov. (Crustacea: Copepoda) from hyperbenthic waters off Okinawa, South Japan, with definitions of related genera Misophria boeck, 1864 and Stygomisophria gen. nov. Huys, R., S. Ohtsuka, G. A. Boxshall and T. It6: Itoitantulus misophricola gen. et sp. nov.: First record of Tantulocarida (Crus- tacea: Maxillopoda) in the North Pacific re- gion NUMBER 5, OCTOBER 1992 REVIEWS Takahashi, S.: Heterogeneity and develop- ment of somatotrophs and mammotrophs in thepratee hase aaain te Ae eee 901 Gerencser, G. A. and B. Zelezna: Chloride pumps in biological membranes .......... O25 ORIGINAL PAPERS Physiology Azuma, K., N. Iwasaki, M. Azuma, T. Shino- zawa and T. Suzuki: HPLC Analysis of retinoids extracted from the _ planarian, Dugesia japonica Fukuta, S., T. Ikata and I. Miura: Effect of disuse on muscle energy metabolism studied by in vivo 31-phosphorus magnetic reso- nance spectroscopy Ootsubo, T. and M. Sakai: Initiation of sper- matophore protrusion behavior in the male cricket Gryllus bimaculatus DeGeer Cell Biology Lee, Y. H. and C. E. Lee: | Ultrastructure of spermatozoa and spermatogenesis in Nepo- morpha (Insecta: Heteroptera) with special reference to phylogeny Immunology Rinkevich, B. and Y. Saito: Self-nonself rec- ognition in the colonial protochordate Botryllus schlosseri from Mutsu bay, Japan Rinkevich, B., M. Shapira, I. L. Weissman and Y. Saito: Allogeneic responses between three remote populations of the cosmopoli- tan ascidian Botryllus schlosseri Saad, A. H.: Sexual development of im- munocompetence in the toad, Bufo regularis (COMMUNICATION) _...... eis oleae 1081 Fabry, H. and J. L. Hedrick: Antibody pro- duction in the goat: Immunokinetics and epitope specificity using a glycoprotein im- munogen Biochemistry Asami, K.: Appearance of a nuclear histone H1 kinase at the start of DNA synthesis of regenerating rat liver ..-..-..-....-.--. 1001 Developmental Biology Kobayashi, H. and T. Hishida: Electron- microscopic observation of an ectopic PGC- like cell in the teleost Oryzias latipes (COM- MUNICATION) 1087 eeceee eee eee ee ee eee ese ee Reproductive Biology Suzuki, T., H. Yamanaka, K. Suzuki, K. Nakajima, K. Kanatani, M. Kimura and N. Otaki: Immunohistochemical demonstra- tion of metallothionein in the rat epididymis and spermatic cord 1009 Kubokawa, K., S. Ishii, K. Tanabe, K. Saitou and H. Tajima: Analysis of sex steroids in feces of giant pandas 1017 eee eee eee ee eee ee ee Endocrinology Paolucci, M., M. M. Di Fiore and G. Ciarcia: Oviduct 17-estradiol receptor in the female lizard, Podarcis s. sicula, during the sexual cycle: Relation to plasma 17(-estradiol con- centration and its binding proteins 1025 Vii trols humoral immunity in the lizard, Chal- CUACSHOCEILALUS Marne hn, eee nr 1037 Hasumi, M. and H. Iwasawa: Dependence of prolactin-stimulated tail fin growth and molt- ing on water in male salamanders (Hynobius nigrescens) (COMMUNICATION) . 1093 Kezuka, H., M. ligo, K. Furukawa, K. Aida and I. Hanyu: Effects of photoperiod, pinealectomy and opthalmectomy on circu- lating melatonin rhythms in the goldfish, COW ASSIUSTQUN AUIS) is ae oe ok ae 1047 Peter, V.S. and O. V. Oommen: _Intermedi- ary metabolism in_ castrated/thyroid- ectomized Calotes versicolor: Regulation by thyroxine and testosterone .......... 1055 Arakawa, E., T. Kaneko, K. Tsukamoto and T. Hirano: Immunocytochemical detection of prolactin and growth hormone cells in the pituitary during early development of the Japanese eel, Anguilla japonica ........ 1061 Morphology Ishikawa, Y.: Innervation of the caudal-fin muscles in the teleost fish, medaka (Oryzias Saad, A. H., M. H. Mansour, M. E. Yazji and TQUBCS)) <P mesh nes ett esd Peeps 5525 1067 N. Badir: Endogenous testosterone con- NUMBER 6, DECEMBER 1992 REVIEWS tics: Potential confusion before order 1113 Scharrer, B.: Recent progress in comparative Proceedings of the 63rd Annual Meeting of the MEULOUMIMUNOIOSY. 2.2... .. 2602-2 22-2 oes > 1097 ZoologicalsSociety.of Japan 92-2 a2. = 42. D7 Gorodilov, Y. N.: Rhythmic processes in ZANNOUNCEMENM(S. chetys ace ost nae se eee 1307 lower vertebrate embryogenesis and their INEKMOWIEGCEMENS Seana cat sccies cass: eta 1308 role for developmental control ......... eT OME UTI O TaN ONG Ee err ace ea slain 1309 Suzuki, K. and D. G. Furth: What is a clas- sification? A case study in insect systema- Contents of ZOOLOGICAL SCIENCE, Vol. 9, Nos. 1-6 ish aA, ha ‘aon * anei fb einen nary ee ES * - ae p a < ‘ ae : hee oer 7 - : ar Brier = Re’ : te.” q dome, = : ‘ : | Z - . res ae ; 7 : 7 ; d i ; * : . } t j fe ¥ = j b ) - \ \ i ‘ a \ . bs ) i Development Published Bimonthly by the Japanese Society of Developmental Biologists e ° ge Distributed by Business Center for Academic Growth & Differentiation Societies Japan, Academic Press, Inc. Papers in Vol. 34, No. 6. (December 1992) 66. REVIEW: K. Yoshizato: Death and Transformation of Larval Cells during Metamorphosis of Anura 67. A. Yanagi: Nuclear Differentiatin in Paramecium caudatum: Analysis by the Monoclonal Antibody against a Macronuclear Antigen 68. L. Bosco and S. Filoni: Relationship between Presence of the Eye Cup and Maintenance of Lens-Forming Capacity in Larval Xenopus laevis 69. A. Hikosaka, T. Kusakabe, N. Satoh and K. W. Makabe: Introduction and Expression of Recombinant Genes in Ascidian Embryos 70. R. Collura and K. S. Katula: Spatial Pattern of Expression of Cyl Actin-(-Galactosidase Fusion Genes Injected into Sea Urchin Eggs 71. Z.-S. Ji, K. Kubokawa, S. Ishii and S-I. Abé: Differentiaiton of Secondary Spermatogonia to Primary Spermatocytes by Mammalian Follicle-Stimulating Hormone in Organ Culture of Testes Fragments from the Newt, Cynopus pyrrhogaster 72. K. Kinoshita, Y. Fujii, Y. Fujita, K. Yamasu, T. Suyemitsu and K. Ishihara: Maternal Exogastrula-Inducing Peptides (EGIPs) and Their Changes during Development in the Sea Urchin Anthocidaris crassispina 73. T. Sawai: Effect of Microtubular Poisons on Cleavage Furrow Formation and Induction of Furrow-like Dent in Amphibian Eggs 74. K. Mitsunaga-Nakatsubo, M. Kanda, K. Yamazaki, H. Kawashita, A. Fujiwara, K. Yamada, K. Akasaka, H. Shimada and I. Yasumasu: Expression of Na*, K*-ATPase a-Subunit in Animalized and Vegetalized Embryos of the Sea Urchin, Hemicentrotus pulcherrimus 75. T.-C. J. Wu, L. Wang and Y.-J. Y. Wan: Differential Expression of Retinoic Acid Receptor mRNA during Mouse Embryogenesis 76. A.A. Oohata: Induction of Cell Differentiation of Isolated Cells in Dictyostelium discoideum by Low Extracellular pH 77. §S. Komazaki: Ultrastructural Localization of Calcium in the Presumptive Ectodermal Cells in Gastrulae of the Newt, Cynops pyrrhogaster, by Cytochemistry and X-Ray Microanalysis 78. T. Iwamatsu, M. Kikuyama and Y. Hiramoto: Fertilization Reaction without Changes in Intracellular Ca7* in Medaka Eggs—An Experiment with Acetone-Treated Eggs 79. K. Yamada, S. Eguchi, T. Yamamoto, K. Akasaka and H. Shimada: Cis-Acting Elements for Proper Ontogenic Expression of Arylsulfatase Gene of Sea Urchin Embryo Author Index to Volume 34 Development, Growth and Differentiation (ISSN 0012-1592) is published bimonthly by The Japanese Society of Developmental Biologists, Department of Developmental Biology, 1990: Volume 32. Annual subscription for Vol. 33, 1991: U.S.$ 162,00, U.S. and Canada: U.S. $ 178,00, 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, 1 East First Street, Duluth, MN 55802, 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, 1 East First Street, Duluth, MN 55802, U.S. A. In Japan: Send nonmember subscription orders and notices of change of address to Business Center for Academic Societies Japan, 16-3, Hongo 6-chome, Bunkyo-ku, Tokyo 113, Japan. Send inquiries about membership to Business Center for Academic Societies Japan,4—16, Yayoi 2-chome, Bunkyo-ku, Tokyo 113, Japan. Air freight and mailing in the U.S. A. by Publications Expediting, Inc., 200 Meacham Avenue, Elmont, NY 11003, U.S.A. The Ultimate Name in Micromanipulation 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 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 TRARISHIGE) NARISHIGE SCIENTIFIC INSTRUMENT LAB. aay 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 XL 7 (04 NH ZOOLOGICAL SCIENCE An International Journal ZOOLOGICAL SCIENCE The Official Journal of the Zoological Society of Japan Editors-in-Chiet: The Zoological Society of Japan: Semen@ | aves (CTO) Toshin-building, Hongo 2-27-2, Bunkyo-ku, Goda | epeyesal (TP eis7o) Tokyo 113, Japan. Tel. (03) 3814-5675 Managing Editor: Ofncon. Bounce Vague n Clana) President: Hideo Mohri (Chiba) pssistaneE (Mots: Secretary: Hideo Namiki (Tokyo) wigslai’se, [Samson CO xeyeonine) Treasurer: Makoto Okuno (Tokyo) SUES oS vans) Librarian: Masatsune Takeda (Tokyo) Akiyoshi Niida (Okayama) —— 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) Hiroshi Watanabe (Tokyo) Mayumi Yamada (Sapporo) Ryuzo Yanagimachi (Honolulu) ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and communications in the broad field of Zoology. The journal was founded in 1984 as a result of unification of Zoological Magazine (1888-1983) and Annotationes Zoologicae Japonenses (1897-1983), the former official journals of the Zoological Society of Japan. ZOOLOGICAL SCIENCE appears bimonthly. An annual volume consists of six numbers of more than 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, 6—16—3 Hongo, 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. No part of this publication may be reproduced or stored in a retrieval system in any form or by any means, without permission in writing from the copyright holder. © Copyright 1992, The Zoological Society of Japan 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 9: 1-16 (1992) © 1992 Zoological Society of Japan REVIEW Transduction Process and Impulse Initiation in Insect Contact Chemoreceptor HIROMICHI Morita Emeritus Professor of Kyushu University, 6-21-3 Kasumigaoka, Higashi-ku, Fukuoka 813, Japan INTRODUCTION It was generally accepted that the local potential evoked in sense organs by adequate stimuli gener- ates afferent impulses in an adjacent region of sensory neurons (generator potential [1]), but it was after 1950 that the generator potential was actually recorded in many sensory neurons [2-9]. Since then, the word “receptor potential” gradual- ly took the place of “generator potential”. Now, the word “generator potential” is rarely found in current papers. This can be understood as we are now most interested in sensory receptor mecha- nisms. It has been suggested for these several years that cyclic AMP (cAMP) or cyclic GMP (cGMP) plays a part in sensory cells as in the target cells for some hormones [10-13]. We recently wrote a review article “chemoreception physiolo- gy” in insects [14], where we could not describe the above aspects of studies on insect chemoreception. Main purpose of the present review is therefore to describe and discuss the development of the field since then. Thus, cell biological aspects are more important here. However, generation of impulses in chemosensilla of insects is so unique in relation to their structures that this problem is also dis- cussed. There are many excellent papers describing in- sect chemoreception with behavioral responses [15-24], but studies of individual chemosensory cells have been made possible only after elec- trophysiology was introduced. Electrophysiologi- Received October 7, 1991 cal study of the insect chemosensory system had been thought difficult because of its tiny size compared with the vertebrate system. As early as 1941, Pfaffmann recorded afferent impulses from taste nerve fibers of the cat [25], while Nodai published a paper in 1953 to report electric poten- tial changes of the afferent nerve by chemical stimulation of tarsal chemosensilla of an insect, orthopteran [26]. Thereafter, many efforts were made in vain to record chemosensory impulses from the tarsal nerve of insects [27]. Once the technique was established for recording of im- pulses [28] and receptor potential [9, 29], however, it soon became clear that insects were suited for study of individual sense cells because of their small number in a single sense organ, sensillum. Now, for example, the labellar sugar receptor cell of the blowly, Phormia regina, and the fleshfly, Boettcherisca peregrina, is one of the most thor- oughly investigated chemosensory cells in the animal kingdom. SENSILLUM STRUCTURE AND IMPULSE INITIATION When we first recorded impulses from the tip of the tarsal chemosensillum of the butterfly, Vanessa inidica, we were surprised to find that the impulse represents an increase in positive potential at the tip with reference to the base of the sensillum [27]. This was true also for the labellar chemosensillum of the green bottle-fly, Lucilia caeser [30, cf. 31]. We expected that impulses would be initiated in a region near the sensillum tip and would therefore 2. H. Morita represent an increase in negative potential at the tip with reference to the base. This polarity was understood as the result of initiation of impulses somewhere near the sensillum base [30, 32]. The next question is why impulses are initiated in the basal region instead of the tip region where the first step of reception occurs. To answer the question, we have to learn morphology of insect sensilla. All the investigated hairs, bristles, pegs and cones on the body surface of insects contain primary sense cells (sensory neurons) that send their axons to the central nervous system (CNS). The number of the sense cells in a sensillum varies with types of sensilla and species of insects, but is generally one to several. The sensilla containing only one neuron are mostly mechanosensitive. The most intensively studied distance chem- osensory (olfactory) sensillum of Bombyx contains two neurons. The labellar contact “chemosensory (gustatory or taste) hair” of flies contains one mechanosensory and four chemosensory neurons (i.e., it is a mechano- and chemo-sensitive sensil- lum). All the neurons are bipolar, whose cell body (perikaryon) is located rather deeply beneath the base of the sensillum. The dendrite extends from the cell body toward the base of sensillum, on the way to which it terminates into a cilium connecting itself with the distal segment. This segment is differentiated into a sensor of the adequate stimu- li. Such is the case in the rod and cone of vertebrate retina: the cilium connects the cell body with the outer segment; this segment is differenti- ated into a photo sensor. As shown by Figure 1 [33], the distal segments of dendrites (dds) are covered together with the dendritic sheath (dsh). This sheath appears the same as cuticle as far as transmission electron microscopy shows, and sepa- rates the inside of the sensillum into two lumina: inner lumen containing the distal segments of dendrites and outer lumen containing nothing but fluid. The inner lumen proximally opens to the ciliary sinus (cs), and the outer lumen to the sensillar sinus (ss). The perikarya of sensory neurons (7) are triply wrapped by inner (i), in- termediate (m) and outer (0) sheath cells. The inner sheath cell forms the ciliary sinus (cs) around the connecting cilia (clm) and is also thought to secrete the dendritic sheath (i.e., thecogen cell). Fic. 1. Structure of a typical sensillum. bf, body fluid; clm, connecting cilium; cs, ciliary sinus; cu, cuticle, dds, dendritic distal segment; dsh, dendritic sheath; e, epidermis; g, glia cell; 7, inner sheath cell; m, intermediate sheath cell; n, nerve cell; 0, outer sheath cell; ss, sensillar sinus. Adapted with permis- sion from [33], Pergamon Press PLC. The outer and intermediate sheath cells (trichogen and tormogen cells, respectively) together form the sensillar sinus (ss). The sensillar tip is 50 to 100 mV positive with reference to the body fluid (bf) [34, 35, 36]. Thurm [35] called this potential “transepitherial potential” (TEP), and showed that TEP requires an energy dependent process, i.e., active transport of some ions to or from the sensillar sinus. In contact chemosensilla of many genera of flies, the concentric arrangement of two cylinders (den- dritic sheath and outer cuticular wall), as shown in Figure 1, changes into a very eccentric one, 1.e., Insect Contact Chemoreception 3 Fic. 2. Cross section of labellar chemosensillum at one quarter of its length from the tip (Phormia regina), < 28,000. By courtesy of Prof. Y. Tominaga, Fukuoka University. the dendritic sheath is fused with the outer wall around nearly a half of its circumference at any level of cross section (Fig. 2). Thus, the inside of the shaft of sensillum is divided into two halves with a partition wall of the remaining non-fused part of dendritic sheath. Such a sensillum appears as two-toned under an optical microscope with transmitted light: dark through the inner lumen and light through the outer lumen. The largest type of labellar chemosensilla of fleshfly or other big flies are so large that a microelectrode tip of glass pipette can easily be inserted into the outer lumen. It is somewhat difficult to insert a mi- croelectrode into the inner lumen, but not impossi- ble. Figure3 shows the records of the same impulse taken by two microelectrodes inserted into the inner (7) and outer (0) lumina, respectively, at the same level above the sensillar base. The common reference electrode was introduced into the proboscis through crushed head, and was grounded via a calibration pulse generator of low resistance. In this case, a rectangular pulse, +1 mV of 2ms duration, was added to show the fidelity of the amplifiers used. The spike height of the impulse recorded from the inner lumen is more than three times higher than that from the outer lumen: the inner and outer lumina are not equipotential. This indicates that the partition wall i a Fic. 3. Records of the same impulse obtained from the inner (z) and outer (0) lumina of the labellar chem- osensillum (Boettcherisca peregrina) [40]. 1 msecx5 Q Cie vr is highly resistive against electric currents. When an adequate stimulus such as a sugar solution is applied to the tip of the sensillum, a slow sustained potential accompanied by trains (or a train) of impulses is recorded between two microelectrodes inserted into the outer lumen, one in the distal region and the other in the proximal of the sensillum (Fig. 4). The polarity of the slow potential is negative at the distal electrode with reference to the proximal: an electric current be- gins to flow toward the tip within the outer lumen prior to initiation of impulses, and is sustained during the application of stimulus. To illustrate such a situation [42], the structure of the labellar chemosensillum is shown in Figure5, and the corresponding equivalent electric circuit in Figure 6A, the equivalent or effective resistances to the generator current for impulses being emphasized. Terminals in, out, ilt and is in Figure 6A represent the same lettered points in Figure 5, respectively. It is assumed in Figure 6A that the inner lumen is completely insulated from the outer lumen by the partition wall except at its two ends. A canal communicating the inner and outer lumina at the sensillar tip is assumed to exist because the current flows toward the tip within the outer lumen. Pre- cise location and structure of the canal need fur- ther discussion (see below). Thus, TEP (its magni- 4 H. Morita |\/64 M A 1/32 M B 1/16 M C 1/8 M D 1/4 M E 1/2 M F | i : 7 2 mV i] 0.1 Ss Fic. 4. DC records of responses of the single labellar chemosensillum (Boettcherisca). Stimulated by glucose solution of the indicated concentration [40]. out Fic. 5. Illustration for the equivalent circuit diagram in Figure 6. i/t, tip of inner lumen; in and out, inside and outside at the receptor membrane, respectively; is, impulse initiation site; Eypp, electromotive force of trans epithelial potential (TEP). Others are the same as in Figure 1 [42]. Insect Contact Chemoreception 5 B in Uy 1/ng I Ey out out (E) (0) Fic. 6. Equivalent circuit diagrams for impulse generator current (or receptor current). The terminlals represent the same lettered points in Figure 5, respectively [42]. tude, E7,p) effectively polarizes the dendrite distal segment (dds) between both ends of the partition wall (arrows in Figure 5). An adequate stimulus increases the conductance of receptor membrane (ng) and an inward current (/) flows across the receptor membrane (from out to in). Current J is divided into J; and JL, the latter flowing out from the distal segment most intensively at the proximal end of the partition wall (impulse initiation site, is) under the strong influence of TEP (the intensity of I, is the highest at is and decays exponentially along the distance from it). The canal at the tip is electron microscopically difficult to show, but was demonstrated by diffu- sion of Ag* into the inner and outer lumina when AgNOs solution was applied to the tip [14]. The process of Ag* diffusion is clearly visible under a light microscope as a growth of precipitation, most probably, of AgCl. Sometimes, dense precipita- tion makes a plug in the inner or outer lumen, beyond which no diffusion occurs. When the plug is made in the outer lumen, we can see the growth of precipitation only in the inner lumen. Such an observation does not necessarily prove the exist- ence of communication canal at the tip, but is also possible if each lumen opens outward at the tip. In fact, there is an electron microscopical study on outlets of the outer lumen at the tip [39]. If the inner lumen were completely insulated from the outer lumen by the partition wall even at both ends, i.e., if resistances R, and R> in Figure 6A were infinitely high, none of the current of inflow at the tip would flow out from the distal segment except at loci proximal to the proximal end of the partition wall, i.e., /=J,. Even though insulation at the tip and root of the distal segment is not perfect, the partition wall assists the current to flow within the distal segment from the tip to root effectively for generation of impulses. In a previous review [14], we presented the results of calculation to show how the partition wall favors transmission of signals (membrane potential changes) from the tip to root without contribution of TEP. Many investigators of insect sensilla, especially mechanosensilla, believe that TEP is the only force driving the generator current to flow. It is because K~ concentration was quite high (100 to 200 mM) in the receptor-lymph cavity (sensillar sinus, ss) [37 cited in 38], and the receptor mem- brane immersed in this lymph is considered to be already depolarized before any adequate stimulus is applied. Recently, Kijima et al. (personal com- munication) studied effects of TEP on the initia- tion of impulses in the labellar chemorecptor of the blowfly, and found a close correlation between impulse frequency and TEP which was lowered by metabolic inhibitors. However, the receptor still responded to sugar solutions with low frequency of impulses even when TEP was completely dimi- nished. Further, he found that the concentration of Na” is higher than that of K~ in the sensillar shaft. Such results suggest that the depolarization of receptor membrane on stimulation contributes to the generator current in the labellar chem- oreceptor of the blowfly. Figure 6B shows the equivalent circuit, which I have used for a model describing the relation between simulus strength and response magnitude 6 H. Morita since I published it 1969 [40]. This circuit is the same as that adopted in non-linear summation of membrane potential changes at the muscle end- plate [41]. Now, the potential at out with reference to in is defined as E when the receptor membrane current J flows for n, the number of the channels opened by the stimulus, g being the conductance per opened channel. For simplicity, n is assumed as zero at rest (without stimulus). Therefore, the potential change V by stimulus becomes as V= E— Em, Where E,, is the resting membrane potential. The theoretical maximum V,,, is written as V,,=E, —E,,, where £, is a) reversal potential of the receptor membrane. Then, we obtain: V=V,,/ (1+G/ng), and I=—GV. If we plot V against log ng, a sigmoid curve is obtained. Considering a contribution of TEP to impulse initiation, we obtain the same equation by rede- fining V,,=E,—(Em+Erep) and V=E—(E,,+ Eryep) for an ideal case where no current flows through conductance G;, [14]. Actually, however, it is most probable that a current J, flows across the receptor membrane from a source just adjacent to it. For such a case, the circuit diagram of Figure 6C is available. This is reduced from Figure 6A, where for J, to flow, R3 is negligible compared with 1/Gz, conductance 1/R; is negligible compared with 1/R3, and R> can then be included in 1/G,. Defining V and V,,, as above, we obtain an equa- tion for J, as 1h=—Vm G2/(1+ G/ng) + Eqep/(ng/G,G2+ WG Gs) where G=G,+G). The equation reminds us that a current, 5 ,—0, flows through G> even at rest (n =0). Only a current increase by stimulation is of interest for us, and it becomes Bf, p=0—= > GoxV n+ Ever Gi/G)/(1+ Ging). Defining Vi=(VintEree Gi/G)/(1+G/ng), (1) we can write the actual maximum as Vs=(Vint Erep Gi/G)/(1+G/sg), (2) where s is the total number of the channels. Asuuming that the channel is opened by sucrose making 1:1 complex with its receptor site, 1.e., introducing n=s/(1+K/C), (3) C being the sucross concentration and K the dissociation constant between sucrose and the re- ceptor site, we obtain the same equation as that published before [40]: Ve— Vel (ape (4) where K,=K/(1+sg/G). (5) The response in impulse frequency (r) in an initial steady state was ascertained to be proportional to the current response [,—I, ,— 0 [40], so that we can write equation (4) as r=r,/(1+K,/C), where r, is the actual maximum of response in impulse frequency. From this equation, we can derive so-called Beidler’s tasts equation as Clr= Clr.“ Rel te Beidler [43] obtained this relation assuming that response is proportional to the number of stimulus molecules binding to the receptor site, and, there- fore, regarded K, as the dissociation constant. TRANSDUCTION AND ADAPTATION Based on the assumption that the channel is opened directly by sucrose binding to the receptor site in 1: 1 manner, the intensity-response relation is described above. This assumption is valid for all sugars except for monosaccharide, where the bind- ing is in 2:1 manner (2 sugar molecules to 1 receptor site of the channel) [44]. Such an assump- tion should be rewritten by a model which numer- ically accounts for the maximum response and Hill coefficient [45-48], but the result may be much the same as far as we assume that the channels is directly operated by sugars. Recently, however, Amakawa et al. [49] obtained the results suggesting that cyclic GMP (cGMP) is involved in the sensory transduction as Insect Contact Chemoreception 7 the second messenger within the sugar receptor cellin Phormia. They applied a membrane perme- able cGMP analogue, dibutyryl cyclic GMP (dbcGMP), to the tip of the sensillum, and re- corded responses from the sugar receptor cell (but not from the other receptor cells). The adaptation (decline of impulse frequency) to dbcGMP sti- mulation was quite slow compared with that to sucrose stimulation in a period more than 3 s after the beginning of stimulus; so slow that the cell continued to discharge impulses of almost a con- stant frequency over 30 s. Furthermore, the cell discharged impulses for a while even after the end of stimulus, when stimulated by dbcGMP mixed with a membrane permeable inhibitor of cyclic nucleotide phosphodiesterase, theophylline or iso- butyl-methyl-xanthine (IBMX). The channel has been assumed to be directly opened by sugar binding, partly because the gener- ator current of impulses (receptor current) begins to flow without any detectable delay after the onset of stimulation. By noise analysis of the sugar receptor current in Boettcherisca, Kijima et al. [50] obtained the results: (a) The autocorrelation func- tion, or covariance function [51], C(t), could be approximately described by an exponential term. (b) The time constant of this exponential term differed with different sugars even when the recep- tor currents were the same in amplitude. General- ly, C(t) is defined as CA=((T)— <b (T+)-—D)p, where /(T) and J(T+12) are receptor currents at times T and T+1, respectively, <J is the average of the receptor currents, and the angle brackets indicate the average taken over 7. Therefore, C(t) precisely represents the time course of relaxation toward equilibrium for the open channel deviated from equilibrium by an amount of the variance of I, C(0)=o0°(I). The relaxation time course should be the same for the same equilibrium state (the same value of <I>, 1.e., the same amplitude of receptor current), if the channel would be brought to this equilibrium state by the same concentration of intracellular cGMP. This is quite contrary to the above results (a) and (b). Thus, it is strongly suggested that the channel is directly operated by sugars. The transduction mechanism in Phormia might be different from that in Boettcherisca. There is indeed a report in Boettcherisca that dbcGMP and its derivatives scarcely cause the sugar receptor cell to discharge impulses, but do the salt receptor cell [52] (cf. the case in Phormia: dbcGMP causes only the sugar receptor cell to discharge impulses [49]). Nevertheless, an involvement of cGMP in the sensory transduction [49] does not necessarily con- tradict a direct operation of the receptor channel by sugars [50]. It is possible that the receptor channel bound to intracellular cGMP is opened only after extracellular binding of sugars: both cGMP- and sugar-bindings are necessary for the opening, and the sugar-binding is the rate limiting step. In such a case, the time course of relaxation differs with sugars and their concentrations as shown by Kijima et al. [50]. Amakawa et al. [49] pointed out the possibility of extracellular receptor site for cGMP and dbcGMP, and there is no experiment of purely intracellular application of cGMP. The extracellular binding of stimulants may not only open the channel, but may also decrease the concentration of intracellular cGMP in some way or other (cf. rods and cones of vertebrates, [53]), resulting in a decrease of open channels. Thus, it is conceivable that the mem- brane permeable dbcGMP, whether be intracellu- larly converted to cGMP or take the part of cGMP, sustains the receptor current as long as dbcGMP stimulation continues. _ The receptor current rapidly decreases to zero after a stimulus solution is broken contact with the sensillum tip. This rapid decrease has been thought to be due to active processes within the receptor cell, because the current tends to last after the end of stimulus in deteriorated prepara- tions. Amakawa et al. [49] obtained the result suggesting that one of the above active processes is hydrolysis of cGMP by phosphodiesterase (cf. [75]). Amakawa and Ozaki [54] tried intracellular ap- plication of activators or inhibitors of protein kinase C by exposing (1 or 2 min) the sensillar tip to a solution containing one of the reagents. The response to sucrose was recorded 5 min after the treatment, and was compared with that after the treatment with the same solution lacking the rea- 8 H. Morita gent. The activator (phorbol ester; DPBA or TPA) depressed the response to sucrose, while the inhibitor (H-7) raised it. Both effects were de- tected with a delay of 0.4-0.5 s after the onset of stimulation, and were thought to be related to an adaptation to long continued stimuli. If we define adaptation simply as a decline of impulse frequen- cy in response to a continued stimulus, the adapta- tion started 0.4—0.5 s after the beginning of stimu- lus in the normal sugar receptor cell. The mem- brane impermeable agent, DPBA or H-7, was made to penetrate into the receptor cell, being mixed with the treatment solution of 0.03% sodium deoxycholate (DOC). Treatment with DOC itself accelerated the adaptation: the adapta- tion started immediately after the onset of stimula- tion. The effect of DOC treatment was antago- nized by a Ca’*-chelator EGTA mixed with DOC. Protein kinase C is activated by Ca** and di- glycerides (or instead, phorbol esters), and there- fore the kinase C was suggested to promote the adaptation to sucrose stimulation. As discussed before, the response to a long lasting stimulus seemed to depend on the in- tracellular concentration of cGMP [49]. As one of the other ways for regulation of response, the kinase C inactivates the receptor molecule in the membrane, resulting in a decrease of the mole- cules available for response. Such might be the case as a kind of desensitization, since neither activators nor inhibitors of kinase C changed the response at the beginning of stimulus. In contrast with kinase C, inhibitors of kinase A or G reduced the response at the beginning [54]. Thus, cGMP and protein kinases (C and G or A) are most probably involved in regulation of re- sponse in the sugar receptor cell of the blowfly. At present, however, we cannot describe a whole story of sensory transduction and adaptation in the sugar receptor cell of the fly. Amakawa et al. [49] ascertained in Phormia that the sugar receptor cell responds to dbcGMP, by examining the proboscis extension response to dbcGMP as well as the spike height of impulses. As noted before in Boettcherisca, the salt receptor cell responds to dbcGMP or its derivatives, but the sugar receptor cell does scarcely [52]. Thus, the same receptor site is possible to occur in a different category of receptor cell (see below). RECEPTOR MOLECULES Multiple receptor sites in a sugar receptor cell Taking into account conformations of sugars and related carbohydrates, Evans [55] proposed that the sugars which stimulate taste receptors of the blowfly combine with two or more distinct receptor sites, each with unique structural requirement, and that these sites are associated with the same recep- tor cell. This is the first definite statement of multiple receptor sites in a sugar receptor cell, referring to the glucose and fructose sites. There- after, we have been able to effectively discuss the structure-activity relationship in the blow- and flesh-flies. Shimada et al. [56] showed in the fleshfly that a 3 min treatment of a single sugar receptor cell with 0.5 mM p-chloromercuribenzoate (PCMB) almost completely depressed its response to D-glucose, but did not affect the response to D-fructose. This was the first experiment discriminating the two receptor sites in a sugar receptor, and it became possible to decide with which receptor site a sugar combines. Figure 7 shows an example: D-fructose, D-fucose and D-galactose react mainly with the fructose site, while D-arabinose, L-fucose, L- arabinose, L-sorbose, D-xylose, L-glucose, as well as D-glucose react wich the glucose site. Sucrose and maltose are judged to react with the glucose site as an a-D-glucopyranoside. Monosaccharide such as fructose has many con- formations as well as a, 8 anomers in an aqueous solution. Therefore, we had first of all to make clear what conformation and which anomer is effective. Chasing the time course of changes in stimulating effectiveness accompanied by muta- rotation, Hanamori et al. [57] concluded that D- fructose is effective as the §-D-furanose form. This and the forgoing results clearly indicated that the glucose site combines with polyols in the form of a-pyranose, while the frustose site combines with those in the form of 8-furanose. Therefore, the glucose and frustose sites are now called pyra- nose (P) and furanose (F) sites, respectively. The structural requirement of stimulants for activating Insect Contact Chemoreception 9 VILILLILLLALAA LALLA LL aR O'5 VLLLL F D-Fucose G F G Relative response after PCMB treatment D-Fructose 22222277277772777 777 PPA D- Galactose ¥2¢2/2/ 478 F D-Arabinose G F I L- Arabinose #— th RH ® ® @ @ w © e W o w S ” vo) ro) L © OuyoOulouso S,/ueoOuUZO > — = 0) x< G) G) Ww) = I ! ! J ai) (a) zy Q Fic. 7. Comparison of responses to some mono- and disaccharides after PCMB treatment. The response (with standard deviation) is normalized so that the response before the treatment is unity [56]. P site is the existence of three juxtaposed equato- rial hydroxy groups in a chair form [56, 57]. That of F site is so. strict that mehyl-@-D- fructofuranoside is ineffective [57, 58]. Shiraishi and Kuwabara [59] stimulated the sugar receptor of the fleshfly with 19 L-amino acids, out of which 6 amino acids (valine, leucine, isoleucine, methionine, phenylalanine, and tryp- tophan) were effective. Shimada and Isono [60] divided F site further into F and T sites by treat- ment with pronase, which did not affect the re- sponse to D-fructose but markedly depressed the response to L-valine. Thus, T site was responsible for reception of aliphatic amino acids. They thought that aromatic amino acids reacted with F site, because the treatment with pronase did not affect the response to phenylalanine and tryp- tophan. There was a definite stereospecificity of F site for aromatic amino acids [61]. On the other hand, there was a different rigid stereospecificity of F site for furanose and its derivatives. This was ascer- tained with 2,5-anhydro-D-hexitols of definite con- figurations [62]. The above two specificities were so different that the sites for aromatic amino acids and for furanoses must not be the same. Accord- ingly, Shimada et al. [62] proposed a new site Ar (Ar=aromatic or aryl) for aromatic amino acids, and also renamed the site for aliphatic acids R site instead of T site (R=alkyl). So far four receptor sites in total have been proposed for the receptor sites in the sugar recep- tor cell of the fleshfly. They are P site for pyranose and its derivatives, F site for furanose and its derivatives, Ar site for aromatic amino acids and its derivatives, and R site for aliphatic amino acids, small peptides, and fatty acids. In addition, the neucleotide site was proposed for the sugar recep- tor cell in Phormia as mentioned before [49], but it was for the salt receptor cell in Boettcherisca [52]. Recently, Shimada et al. [63] found in Boettcher- isca that certain 1,6-anhydro-$-D-hexopyranoses (1,6-anhydro-~-D-galactose, -altrose, -talose and -gulose) were effective in stimulating the salt re- ceptor cell: there is a sugar receptor site in the salt receptor cell. 10 H. Morita It is well known that ingestion of food in flies is released by afferent impulses of sugar receptor and water receptor cells and is blocked by those of salt receptor cells. Water is essential for the mainte- nance of life in flies, but is reyjected when con- taminated with salts of high concentrations. However, water with such salts would still be valuable if it contains nutritious sugars. The response of water receptor cell to water in itself is completely blocked by dissolved 0.05 M sodium citrate, but is recovered by addition of some sugars in it. Wieczorek and Koppl [64] discovered this phenomenon in Protophormia and called it reac- tivation by sugars. The receptor site was a kind of F site; indeed, a sugar receptor site in the water receptor cell. Search for receptor molecules—a-glucosidase Eluting the legs of flies (Phormia) with water and testing for enzyme activity in the eluate, Dethier [65] found an a-glucosidase activity. This brought about Hansen’s biochemical investigations on the receptor mechanism in Phormia [66]. He demonstrated a positive correlation between the regional a-glucosidase activity and the regional density of chemosednsilla and between the sub- strate specificity of the enzyme and the stimulating effectiveness of a-glucosides (for the above- mentioned P site). Kijima et al. [67] picked up this proposal in Phormia, separated isozymes [68], characterized them [69], and found a membrane- bound a-glucosidase in the chemosensillum [70, 71]. Over a wide spectrum of substances, they disclosed good parallelisms in substrate- and in- hibitor-specificities between the enzyme and the Sugar receptor, but only one result of a great discrepancy was enought to discard the hypothesis: the inhibition constant of nojirimycin for the a- glucosidase was in a range of 10° to 10°-° M, while that for the sugar receptor was above 4x 10-* M [72, 73]. Dethier wrote in his recent review [74]: “They (including Hansen et al.) proved that the enzymes were intimately and exclusively associated with the receptors. Exactly what part the glucosidases play in the process of transduction is still a mystery.” Whatever a mystery may be, it is certain that sucrose is split into glucose and fructose on the outside of the intact sugar receptor membrane. It amounts to 0.5 pmol per sensillum per hour, i.e., 1.4x10~'° mol s~', when the sugar receptor is stimulated by 0.1M sucrose [72]. This is just comparable with the number of monovalent cat- ions crossing the receptor membrane for 1s. The split glucose and fructose molecules might be picked up into the distal segment of the sugar receptor cell, as suggested by Hanamori [75]. He exposed the sensillar tip to '*C-D-glucose or 3-O- mehtyl-'*C-D-glucose, effective or ineffective stim- ulant for the sugar receptor cell (Phormia), respec- tively, and studied the uptake and movement of isotopic activity. When the tip was pretreated with 75 mM colchicine for 30 min, the taken up and remaining activity within the sensillum shaft in- creased for glucose, but not for 3-O-mehyl- glucose. This was explained as a result of dis- assembly of microtubules, which otherwise could transport glucose, but not 3-O-methyl-glucose, proximally to the cell body. That is, the glucosi- dase might play a part in the rapid stop of response after the end of stimlus. Affinity electrophoresis Ozaki [76] at last found a most probable candi- date for P site in Phormia, with affinity elec- trophoresis. Her strategy was based on her own finding that starch does not evoke but inhibits the response of the sugar receptor cell [77]. Starch has no electric charge and is polydisperse with huge molecular weights in an aqueous solution. There- fore, a protein molecule under electrophoresis moves only in a state free from formation of the starch-protein complex. As the mobility of protein is thus proportional to the period of this free state, we obtaine (the fraction of free state is derived from equation (3) as (s—n)/s=1—1/(1+K/C), where n represents the number of molecules mak- ing complex in 1:1 manner, s the total number of molecules): m,/m=1+[I]/Ki, (6) where m and m, is the mobility of protein with and without starch, respectively, [I] the concentration of starch, and K; the dissociation constant of the protein-starch complex [78]. Further, because a protein molecule can be assumed to have the same Insect Contact Chemoreception 11 mobility when it forms complex with a small uncharged molecule of sugar, we obtain an equa- tion as in competitive inhibition, m /(m,—m’)= Ki +[s]/Ka)/[I], (7) where m’ is the mobility of the protein in the presence of starch and the sugar, [s] the sugar concentration, and K, the dissociation constant of the protein-sugar complex [79]. In two-dimensional polyacylamide gel elec- trophoresis, starch was added to the gel of the first run. The second run was carried out through a slab gel that were the same in composition except for lacking starch. Proteins having no interaction with starch should align on a diagonal line starting at the origin of electrophoresis. In fact, a protein was detected at a spot aside from the diagonal line, at shorter distances on the first run with higher concentrations of starch (Fig. 8). The dissociation constant of the protein-starch complex obtained from equation (6) was 0.7%, which agrees well with the value, 0.6%, obtained before in her physiological study [77]. The dissociation con- -05 0 O5 10 15 20 CONCENTRATION OF STARCH, II] ‘/o B =—»> stants of the protein-sugar complex obtained from equation (7) also agreed with the values predicted by equation (5): K, is larger than K, by a factor of (1+sg/G), which was previously estimated to be about 5 in Phormia [80]. In her study, a-glucosidase did not interact with starch, suggesting that this novel protein is differ- ent from a-glucosidase mentioned above. MAINTENANCE OF RECEPTOR MEMBRANE If we examine the chemoreceptor activity in flies collected in the field, we will find that many chemoreceptor cells have lost their activity. The contact chemoreceptor cells are exposed to many dangerous substances while the fly searches for food. Therefore, it is important for the fly to recover the activity of injured chemoreceptors. Shimada [81] is the first to report the recovery from an injury resulting from contact with deter- gents in Boettcherisca. Being exposed to 0.1% sodium deoxycholate (DOC) for 5 min at the sensillar tip, the chemoreceptor cells discharged injury spikes for a while, ceased to discharge within the exposure time, were silent for 10-20 min, then began responding again, and finally recovered about 80% response at about 50 min after the detergent exposure. The destruction and reorganization of the recep- tor membrane were reinvestigated in detail in Phormia. Recording the receptor membrane cur- rent during the DOC treatment, Kashihara et al. (unpublished) follwed changes from a partial to full lysis of the membrane at the sensillar tip. The full lysis was characterized by a large continued injury current accompanied by an intense dis- charge of impulses. However, as in Boettcherisca, the response to 0.1 M sucrose (0.1 M sucrose re- sponse) recovered about 90% at 90 min after the Fic. 8. (A) Determination of dissociation constant Kg after equation (6). (B) Determination of m and m,. In the presence of starch (1st run), starch- interacted protein migrated to position p._ If starch had not existed in the first run, the protein should have migrated to position q on the diagonal line in an extension of horizontal line op. The diagonal line is defined by the protein stain (dotted area). Then, m,/m=oq/op. Adapted from [76]. 12 H. Morita DOC treatment. Ninomiya ef al. [80] found that colchicine has a unique effect on the recovery. The sugar receptor cell was treated with colchicine (5- 25mM), whose effects were not detected at all unless we examied the recovery from destruction by the succeeding DOC treatment; the recovery was deeply depressed when the receptor mem- brane was pretreated with colchicine of concentra- tions above 5 mM for 2 min. The recovery process clearly consisted of two processes, colchicine de- pendent and colchicine independent ones. The colchicine independent process was fast, approxi- mated by an exponential term with time constant of 15 min; this process is thought to be physical, i.e., desorbtion of DOC monomer from the recep- tor membrane. The colchicine dependent process was rather slow, and its time constant was 50 min wher approximated by an exponential term; this process is thought to be biological, i.e., regenera- tion of the distal segment of dendrite as well as repairs of the receptor membrane. Observing the 0.1 M sucrose response as long as 15 to 25 hr, Ozaki et al. [82] revealed a long lasting effect of the treatment with 25 mM colchicine for 2 min. As shown by Figure 9, after the colchicine- DOC treatment (25 mM colchicine treatment for 2 min succeeded by 0.3% DOC treatment for 2 min), the 0.1 M sucrose response was recovered to about 50% level through the colchicine independ- ent process. This level was kept for 12 hr thereaf- ter. The second step of recovry in response then began, whereas recovery in latency (from the onset of stimulus to the first impulse) occurred 5 hr WW Ww uJ 2 Ww) : 3 2 a. mn Mp) = WW = Ss leq lJ we Tp) (eo) 5 n > Ww (Ss) fod Zé WT colchicine tal > 0OC < = J} < J jeg TIME AFTER OOC TREATMENT Fic. 9. Changes in the response (©) and the latency (@) (from the onset of stimulus to the first impulse) after the colchicine-DOC treatment [82]. earlier. This implies that the effect of the colchi- cine terminated 7 hr after the treatment. The effect of colchicine was also detected with- out DOC treatment. The 0.1 M sucrose response decreased 12 hr after the single colchicine treat- ment, but recovered to the normal at 4 hr thereaf- ter. When the colchicine treatment was repeated every 3 hr, the response was reduced to 50% at 16 h after the first treatment. Calculation on the diffusion equation showed that the concentration of colchicine in the receptor cell was at most 5.6 10~° M at 3 hr after the treatment. The remaining effect of colchicine at such an extremely low concentration could not be explained unless we assumed involvement of microtubule system. When microtubules were disassembled into tubu- lin dimers by binding to colchicine, the resultant colchicine-tubulin dimers may cap the growing end of microtubule to prevent its reassembly even at extremely low concentrations of free colchicine, as postulated by Margolis and Wilson [83]. It was also ascertained that vinblastine had the effect but lumicolchicine did not [80]. Thus, it is most likely that microtubules is involved in the maintenance of © the receptor membrane as well as the distal seg- ment of receptor cell. As above, electrophysiology tells us living states of the receptor cell in real time. The long latency in response and lack of the initial phasic response (Fig. 10a) pointed out existence of a distance be- tween the tip of sensillum and that of distal seg- ment of the cell which had recovered the response after the colchicine-DOC treatment. This was ascertained with electron microscopy [82], and led us to calculate the actual concentration of sucrose (C’) at the tip of the distal segment at the begging of recovery. Further, assuming the time constants of the colchicine-independent and -dependent re- covery processes as 15 (7) and 50 min (7), respec- tively, and using equations (1)-(5), we can calcu- late the time course of recovery, i.e., the ratio V’,(t)/V,, where V’,(t) is the response recovered with time after the colchicine-DOC treatment. Thus, Ninomiya et al. [80] obtained the theoretical curves of recovery process as shown in Figure 11. The agreement with the experiments is excellent: after the same treatment (DOC treatment without colchicine), the recovery is seen slower in 0.01 M Insect Contact Chemoreception 13 betore 1 mV after a MTA TOYA UU LA) LLU LaLa nilbal bd san cabin 01s ot pape) x abt ee ded ae Di all Lh tt ical lh ali Fic. 10. Typical DC records of the 0.1 M sucrose responses, obtained before and after the colchicine-DOC treatment (continuous a to b). RESPONSE RELATIVE % 30 60 90 min TIME AFTER BEGINNING OF RECOVERY Fic. 11. Theoretical curves and results of experiments. (©) 0.1M sucrose response, pretreatment without colchicine; (@) 0.01 M sucrose response, without colchicine; (x) 0.1M sucrose response, pretreat- ment with 0.5mM colchicine; (A) 0.1 M sucrose response, with 5 mM colchicine [80]. sucrose response than in 0.1 M sucrose response; these recovery processes are described by the theoretical curves with the same set of parameter values (7}=15 min, »m=50 min, C’/C=0.6, and The horizontal bar represents the period of stimulus [80]. sg/G=4). Figure 12 shows the concentration-response curves before and after the DOC-colchicine treat- ment. The concentration of colchicine was so high that the colchicine-dependent recovery was com- pletely suppressed. The averaged values of K, (+ SEM) were 14+4 and 75+4 mM, respectively, before and after the treatment. The above set of the parameter values determines in turn the values of several constants: K=70+20 mM, (sg/G)’= 0.56 (the value of sg/G after the treatment), V’,/ V,=0.45 (V’, is the maximum response after the treatment). The value of 0.45 is somewhat lower than that of experiments, but is located within the standard error. In fact, if we take the value K=90 mM, the value V’,/V, becomes 0.625, which is the same as the result of experiments in Figure 12. As mentioned above, the colchicine-DOC treat- ment produces a preparation suited for reconstruc- tion of the receptor membrane in vivo. Pretreat- ment with colchicine (25 mM, 2 min) completely depresses the colchicine dependent recovery pro- cess after destruction by DOC (7.2 mM, 2 min) treatment, and retains the 0.1 M sucrose response 14 H. Morita RESPONSE RELATIVE 1073 1072 10° mM 1@! CONCENTRATION OF SUCROSE Fic. 12. Concentration-response curves for sucrose stimulation before (@) and after (©) the colchicine (25 mM)-DOC treatment. theoretical curves after equations (2), (4) and (5). the treatment [80]. to a level below a half of the intact response. Ninomiya et al. (unpublished) incubated the sensil- lar tip for 20 min, after the colchicine-DOC treat- ment, with microsomal fraction (P-fraction) which was prepared from the homogenate of labella of the blow fly. This incubation improved the recov- ery especially in the latency in response. However, when the sensillar tip was incubated with tubulin- tyrosine ligase, the recovery was significantly im- proved in the magnitude of response. In other words, P-fraction seemed to elongate the des- tructed distal segment, while tubulin-tyrosine ligase seemed to increase the number of receptor molecules. In this review, I described and discussed only a few selected subjects. For a comprehensive review before 1976, see reference [84]. Each points is the mean (+ SEM) of six preparations. The smooth curves represent the Arrows indicate K, values before (14 mM) and after (75 mM) ACKNOWLEDGMENTS I would like thank Drs. H. Kijima, I. Shimada, T. Amakawa and M. Ozaki for reading the manuscript and valuable comments. REFERENCES 1 Granit, R. (1947) Sensory Mechanisms of the Re- tina. Oxford Univ. Press, London. 2 Katz, B. (1950) J. Physiol., 111: 261-282. 3. Davis, H., Tasaki, I. and Goldstein, R. (1952) Cold Spr. Harb. Symp. Quant. Biol., 17: 143-154. 4 Alvarez-Buylla, R. and Ramirez de Arellano, J. (1953) Am. J. Phsiol., 172: 237-244. 5 Gray, J. A. B. and Sato, M. (1953) J. Physiol., 122: 610-636. 6 Eyzaguirre, C, and Kuffler, S. W. (1955) J. Gen. Physiol., 39: 87-119. 7 Eyzaguirre, C, and Kuffler, S. W. (1955) J. Gen. Physiol., 39: 121-153. 8 Edwards, C. and Ottoson, D. (1958) J. Physiol., 33 34 35) 36 37] Insect Contact Chemoreception 15 143: 138-148. Morita, H. (1959) J. Cell. Comp. Physiol., 54: 189- 204. Fesenko, E. E., Kolesnikov, S. S. and Lyubarsky, A. L. (1985) Nature, 313: 310-313. Nakamura, T., and Gold, G. H. (1987) Nature, 325: 442-444. Avenet, P., Hofmann, F. and Lindemann, B. (1988) Nature, 331: 351-354. Tonosaki, K. and Funakoshi, M. (1988) Nature, 331: 354-355. Morita, H. and Shiraishi, A. (1985) In “Compre- hensive Insect Physiology, Biochemistry and Phar- macology” Ed. by G. A. Kerkut and L. I. Gilbert. Pergamon Press, Oxford, vol. 6: 133-170. McIndoo, N. (1914) J. Exp. Zool., 16: 265-346. von Frisch, K. (1919) Zool. Jahrb., 37: 1-238. von Frisch, K. (1921) Zool. Jahrb. Abt. f. allg. Zool. u. Physiol., 38: 449-516. von Frisch, K. (1935) Z. vergl. Physiol., 21: 1-156. Mirmnich Dre. (1921) Jr Exp: Zool-. 33> 173=203. Minnich, D. E. (1926) Biol. Bull., 51: 166-178. Minnich, D. E. (1926) Anat. Rec., 34: 126. Minnich, D. E. (1929) Z. vergl. Physiol., 11: 1-55. Minnich, D. E. (1931) J. Exp. Zool., 60: 121-139. Dethier, V. G. (1955) Quart. Rev. Biol., 30: 348- SWAN. Pfaffmann, C. (1941) J. Cell. Comp. Physiol., 17: 243-258. Nodai, H. (1953) Kyushu Shigaku Z., 6: 1-3. (In Japanese) Morita, H., Doira, S., Takeda, K., and Kuwabara, M. (1957) Mem. Fac. Sci. Kyushu Univ. Ser. E., 2: 119-139. Hodgson, E. S., Lettvin, J. Y. and Roeder, K. D. (1955) Science, 122: 417-418. Morita, H. and Yamashita, S. (1959) Science, 130: OW. Tateda, H. and Morita, H. (1959) J. Cell. Comp. Physiol., 54: 171-176. Hodgson, E. S. and Roeder, K. D. (1956) J. Cell. Comp. Physiol., 48: 51-75. Dethier, V. G. (1956) Amer. Inst. BIol. Sci., Publ. No. 1: 1-30. Zacharuk, R. Y. (1985) In “Comprehensive Insect Physiology, Biochemistry and Pharmacology” Ed. by G. A. Kerkut and L. I. Gilbert. Pergamon Press, Oxford, vol. 6: 1-69. Wolbarsht, M. L. (1958) J. Gen. Physiol., 42: 413- 428. Thurm, U. (1972) Olfaction and Taste, 4: 95-101. Wieczorek, H. (1982) J. Comp. Physiol. A, 148: 303-311. Kuppers, J. (1974) In “Mechanoreception” Ed. by 38 39 40 4] 42 43 44 45 46 47 48 49 50 Si 52 33 54 SS) 56 D7) 58 59 60 61 62 J. Schwartzkopff, pp. 387-394. Abh. Rhein. Westf. Akad. Wiss. Opladen. McIver, S. B. (1985) In “Comprehensive Insect Physiology, Biochemistry and Pharmacology” Ed. by G. A. Kerkut and L. I. Gilbert. Pergamon Press, Oxford, vol. 6: 71-132. Menco, B. Ph. M. and van der Wolk, F. M. (1982) Cell & Tissue Res., 223: 1-27. Morita, H. (1969) Olfaction and Taste, 3: 370-381. Martin, A. R. (1955) J. Physiol., 151: 1-23. Morita, H. (1991) Biomed. Res., 12, Supplement 2: 225-227. Beidler, L. M. (1954) J. Gen. Physiol., 38: 133-139. Morita, H. and Shiraishi, A. (1968) J. Gen. Phy- siol., 52: 559-583. Kijima, H. and Kijima, S. (1978) J. Theor. Biol., 71: 567-585. Kijima, H. and Kijima, S. (1980) J. Theor. Biol., 82: 425-463. Kijima, H., Kijima, S., Noguchi, M. and Morita, H. (1980) J. Theor. Biol., 83: 131-161. Kijima, H. and Kijima, S. (1989) In “Biological Transduction Mechanisms” Ed. by M. Kasai, et al. Japan Sci. Soc. Press, Tokyo/Springer-Verlag, Ber- lin. pp. 17-45. Amakawa, T., Ozaki, M. and Kawata, K. (1990) J. Insect Physiol., 36: 281-286. Kijima, H., Nagata, K., Nishiyama, A. and Morita, H. (1988) J. Gen. Physiol., 91: 29-47. Neher, E. and Stevens, C. F. (1977) Ann. Rev. Biophy. Bioeng., 6: 345-381. Shimada, I. (1991) In “Molecular Mechanism of Chemosensory Transduction” (Science Report for Grant-in-Aid from Min. of Educ. Sci. and Cult. Jap., 63304063) Ed. by H. Kijima, pp. 66-67. Pugh, E. and Altman, J. (1988) Nature, 334: 16-17. Amakawa, T. and Ozaki, M. (1989) Physiol., 35: 233-237. Evans, D. R. (1963) Olfaction and Taste, 1: 165- 176. Shimada, I., Shiraishi, A., Kijima, H. and Morita, H. (1974) J. Insect Physiol., 20: 605-621. Hanamori, T., Shiraishi, A., Kijima, H. and Morita, H. (1974) Chemical Senses and Flavor, 1: 147-166. Jakinovich, W. Jr., Goldstein, I. J., Baumgarten, R. J. von and Agranoff, B. W. (1971) Brain Research., 35: 369-378. Shiraishi, A. and Kuwabara, M. (1970) J. Gen. Physiol., 56: 768-782. Shimada, I. and Isono, K. (1978) J. Insect Physiol., 24: 807-811. Shimada, I. and Tanimura, T. (1981) J. Gen. Physiol., 77: 23-39. Shimada, I., Horiki, H., Ohrui, H. and Meguro, H. (1985) J. Comp. Physiol., 157: 477-482. J. Insect 63 64 65 66 67 68 69 70 mh M2 73 74 16 Shimada, I., Ohrui, JASTS, 23: 99-102. Wieczorek, H. and Koppl, R. (1978) J. Comp. Physiol., 126: 131-136. Dethier, V. G. (1955) J. Econ. Entmol., 48: 235- 239. Hansen, K. (1969) Olfaction and Taste, 3: 383-391. Kijima, H., Koizumi, O. and Morita, H. (1973) J. Insect Physiol., 19: 1351-1362. Kawabata, K., Kijima, H., Shiraishi, A. and Mori- ta, H. (1953) J. Insect Physiol., 19: 337-348. Koizumi, O., Kijima, H. and Morita, H. (1974) J. Insect Physiol., 20: 925-934. Amakawa, T., Kijima, H. and Morita, H. (1975) J. Insect Physiol., 21: 1419-1425. Kiima, H., Amakawa, T., Nakashima, M. and Morita, H. (1977) J. Insect Physiol., 23: 467-479. Morita, H., Enomoto, K., Nakashima, M., Shima- da, I. and Kijima, H. (1977) Olfaction and Taste, 6: 39-46. Nakashima, H, Enomoto, K., Kijima, H, and Mori- ta, H. (1982) Insect Biochem., 12: 579-585. Dethier, V. G. (1990) Ann. Rev. Neurosci., 13: 1- 13). H. and Meguro, H. (1989) H. Morita 73 76 77 78 79 80 81 82 83 84 Hanamori, T. (1976) Chemical Senses and Flavor, 224229239) Ozaki, M. (1988) Zool. Sci., 5: 281-290. Hara, M. (1983) J. Insect Physiol., 29: 113-118. Takeo, T. and Nakamura, S. (1972) Arch. Biochem. Biophy., 153: 1-7. Takeo, T. and Nakamura, S. (1978) In “Affinity Chromatography” Ed. by O. Hoffmann-Ostenhof, M. Breitenbach, F. Koller, D. Kraft and O. Schneider, Pergamon Press, Oxford and New York, pp. 67-70. Ninomiya, M., Ozaki, M., Kashihara, Y. and Mori- ta, H. (1986) J. Gen. Physiol., 87: 1003-1016. Shimada, I. (1975) J. Insect Physiol., 21: 1565- 1574. Ozaki, M., Ninomiya, M., Kashihara, Y. and Mori- ta, H. (1986) J. Gen. Physiol., 87: 533-549. Margolis, R. L. and Wilson, L. (1977) Proc. Nat. Acad. Sci., 74: 3466-3470. Dethier, V. G. (1976) The Hungry Fly. Harvard University Press, Cambridge, Massachusetts and London, England. ZOOLOGICAL SCIENCE 9: 17-36 (1992) © 1992 Zoological Society of Japan REVIEW Development of the Hypothalamic Monoaminergic System in Ontogenesis. Morpho-functional Aspects MICHAEL V. UGRUMOV Institute of Developmental Biology, U.S.S.R. Academy of Sciences, Moscow 117808, U.S.S.R. CONTETS |UCROGINCTOMN 35 eo aae dancer Anarene teese On aameee ore acl rma erin ler SOE aR Ye Sc on a 5 Se a ee 17 Dov ClOpmei tol MONOAMINE SICISVStCMIS epee cs-cee- de eak ona vedas Meade deans on dep dee ese eae a oe ne 18 Pe ALCCHOLATMINEL PLC SV SLOT gai dels, lesa niiae otto. clad gout ncrec oueten hats aa Ne ns aoe woes ncidetgd ove Sota 18 AME ANTIG ITIVE CLOMICS erence pn reece erste Bohne eas neee dee so sew zh oa hc MAAC Ped asicceee a hee ge 18 bb) ee Monpho-tunctional Chanactenisttcs) s:c.9. 4-6. esas nae ea sche onecePeosadee terete een aco 21 PPPS CTOLOMMERLIC) (Sartell SV STEM Aer et Ssh eon eh Sac ca ben taseledat quel ca cneeedeateeM eee ae de eraandecsasetan: 24 AEP NTCMICE CCOMICS cd. ISSUER ERE ee Uae cece ct, com airtime WIE L 4, I PTT Te ateinniieyt 24 b) ae MonphotunctiomaluchanactenmiStt Sey. sass. ck cee ge ee seesece sn ctcce cee eeernc teach emeetoe cnclsecacee oh 26 III. Neurohumoral control of development of monoaminergic systeMS ...................0ccecee eee ee sees 26 Pa @ ALC CHOLATINITICT CIC SV StEMM i nan tea usta Seater: snes cn canie conte Ee nick Goble state acinar iemichewiee Suaiacih 26 A) MES ONUAINSTCTOINS ee ere ee anette ee het ade seek ete MMHG ch aioe on sr Aw Mya eM oh cred ect Nee 26 ) MOMS TAMORIMOMES A Fee Set ss heer Mesa s es dace cincscmiin Lutidia sens dec Se MO Ne acilte ac cnet ea eteiacronre eae Diy PRES CTOLOMIMET CICS VSUC INP eee ese EERE fence Se alts Feo te vale g TUT es « Saa aielhah seu Wee Gok bane MRE a cecal 28 Zi) SKEET SUETTOTONS, sero o cts se TE SPF cs SERIE ISIE A aN 5 28 MeO Chemhonmonese. seat saee er eternity ten sah none seca oreds uate an neh nana ee car ae 29 IW FUNC HONAlSISMICAaNnce Ol MONOAMINES 7. 2s. eke faeces snes eect es onan sect ve cae coc tue Wan cesecsceeeSeone 29 Lo (CRUSE NOLETTNNT Ec iaos8 3p ee ai ok ee Sake Sean Ae ae aCe Un Es ee Rea HTT ne eee 29 ORES CRO COMMU A eG eRe RRR TNL MOAN Lo bea Sssia chase site walelddie aaivia MERE OAN GMA ong od ecles duane see 32 Whe (COWNGIIISTIOINS “gescbecigtacaededs dAeG Seces ATOR ARON GAS 6 Se MISES ie Morr at recive ri nana ane an er ea 38) PNERERE INCE SMR ey es tere Pete eB Pe cate eer acs atic Aescial corso stores Sesoine della leas elaratinate eeceeae a te 32) I. INTRODUCTION Over the last two decades a large body of evidence have been accumulated on the key role of hypothalamic monoamines (MAs): dopamine (DA), noradrenaline (NA), adrenaline (A) and serotonin (5-HT), in the neuroendocrine regula- tion [1, 2, 3]. The MAs play the role of: a) Neurotransmitters controling the functional activ- ity of the target neurons; b) Neuromodulators regulating the release of neurohormones from adjacent neurosecretory axons; and c) Neurohor- Received November 15, 1991 mones delivered by portal blood and cerebrospinal fluid (CSF) to the target adenohypophysial cells List of addreviations AD-adrenaline, adrenergic; CA-catecholamine, catech- olaminergic; CSF-cerebrospinal fluid; DA-dopamine, dopaminergic; E-embryonic day; EM-electron micros- copy, electron microscopic; 5-HT-serotonin, seroto- ninergic; IP-immunoposivive; LM-light microscopy, light microscopic; MA-monoamine, monoaminergic; ME-median eminence; MFB-medial forebrain bundle; (M)PA-(medial) preoptic area; n-nucleus, nuclei; NA- noradrenaline, noradrenergic; P-postnatal day; pCPA- p-chlorophenylalanine; SCN-suprachiasmatic nucleus; SS-sexual steroids. 18 M. V. UGrumov / IV Mo) oo Se | ee ae 7 ! Vi Fic. 1. Schematic representation of participation of hyporthalamic monoamines in the neuroendocrine regulation. I-serotoninergic neurons (SE) of raphe nucleus; II-noradrenergic neurons (NA) of locus coeruleus; III-hypothalamic neurosecretory nuclei with dopaminergic (DA) and peptidergic (PE) neurons; IV-median eminence with capillary loops of primary portal plexus; V-portal veins; VI- adnohypophysis with the secondary portal plexus; V-ventricle. Specialized contacts of monoaminergic axons; small arrowhead-axo-ventricular; large arrowhead-synapse; middle arrow-axo-axonic; large arrow-axo-vascular. small arrow-direction of blood stream in hypophysial portal circulation. and neurons (Fig. 1) [4]. Moreover, MAs provide their influence on the development first of the whole embryos and then of the central nervous system including hypo- thalamus long before the onset of the neurotrans- mitter function [5, 6, 7, 8]. In turn, the differentia- tion of MA neurons and expression of their feno- type are under the neurohumoral control [9]. This review attempts to summarize the results mainly on the development of the hypothalamic MA system and to a lesser extent on its role in the neuroendocrine mechanisms in ontogenesis. Il. DEVELOPMENT OF MONOAMINERGIC SYSTEMS 1. CATECHOLAMINERGIC SYSTEM a) Architectonics Cell bodies. The combination of long survival [°>H]thymidine autoradiography with immunocytochemistry of tyrosine hydroxylase (TH), the key-enzyme of the catecholamine (CA) synthesis, has shown that the genesis of the hypothalamic CA neurons in rats occurs from the 11th fetal day (E11) until E17. The neurons of the dorsal accumulations (zona incerta, etc.) are mainly originated at E12, or even — earlier, while those of the ventral accumulations (arcuate region) later, at E15 [10]. TH immunocytochemistry occured to be the most useful approach in the study of the dif- ferentiation of the CA neurons, as this enzyme appears early in ontogenesis and is widely distri- buted through neurons [11]. First rare TH im- munopositive (IP) neurons were observed in the lateral hypothalamus at E13 (Fig. 2) [12] that is in agreement with the biochemical observation of TH activity in this region [13]. Over subsequent 2-3 days the THIP neurons increased in number giving rise to bilateral dorsomedial accumulations (Fig. 3) [12, 14]. From E18 onwards, THIP neurons were widely distributed concentrating in some hypothalamic nuclei (n.) (Figs. 4, 5). In the ante- rior hypothalamus, occasional neurons or their small clusters were regularly seen in the medial preoptic area (MPA), suprachiasmatic and ante- rior hypothalamic n. [12]. A considerable number of THIP neurons are concentrated dorsomedially overlapping the paraventricular n., rostrally, and zona incerta, dorsomedial and posterior hypotha- Monoaminergic System in Ontogenesis 19 Fics. 2, 3. lamic n., caudally. Ventromedially these clusters are connected with a massive periventricular accu- mulation of THIP neurons first appeared on E20 2k In the mediobasal hypothalamus two small ven- tral bilateral accumulations of THIP neurons observed only in perinatal rats (E18-P3) are local- ized lateral to the median eminence (ME) (Fig. 4). By P3 only few neurons keep their initial position while a similar neuronal accumulation appears in the arcuate n. (Fig. 5) [12, 14]. It is still uncertain whether this is a result of neuronal migration or transient expression of TH synthesis. Thus, CA neurons are originated long before birth. Over perinatal period they occupy their definitive positions concentrating mainly in zona incerta, periventricular and arcuate n., as in adults Schematic representation of catecholaminergic (THIP) neurons (black dots) in the hypothalamus and septum in fetal rats at E13 (2) and E15 (3). Each dot represents 2-3 THIP neurons. fm-foramen of Monro; lv-lateral ventricle; of-optic fissure; or-optic rudiment; p-pituitary; pa-preoptic area; v-3rd ventricle. (Modified [12]). [15]. Nerve fibers. One of the most important characteristic of the development of the CA system is the growing of the THIP fibers to the target neurosecretory nuclei and circumventricular organs. THIP fibers belong- ing to neurons of mesencephalon and pons [11, 16] first enter the hypothalamus via the medial fore- brain bundle (MFG) at E13. Then, they project into the optic chiasma and tracts as well as into the primordium of the ME from E15, the anterior comissure, the diagonal band and septum from E18, and into zona incerta, dorsomedial and para- ventricular n., etc. perinatally [17]. It is obvious that the places of the high concentrations of THIP neurons become occupied by networks of THIP 20 M. V. UGRUMOV Fics. 4, 5. neurons in the hypothalamus and septum in fetuses at E18 (4) and in neonatal rats at P9 (5). a-arcuate nucleus (n.), ac-anterior comissure; ah-anterior hypothalamic n.: db-diagonal band; dm-dorsomedial n.: It-lamina terminalis; mb-medial forebrain bundle; me-median eminence; oc-optic chiasma; of-optic fissure; on-optic nerve; ot-optic tract; pe-periventricular n.; pv-paraventricular n,; s-septum; sc-suprachiasmatic n.; so-supraoptic n.; vm-ventromedial n.; zi-zona incerta. For other abbreviations see Figs. 2, 3. (Modified from [12, 47].) fibers, however, the latter are also observed in the hypothalamic regions almost lacking the THIP neurons [17]. In the septal region and diagonal band, the THIP fibers terminate on luteinizing hormone releasing-hormone neurons in neonatal rats [18]. The number of smooth neurons innervated by THIP fibers remains at the constant level from P2 until P90, while that of neurons possessing spine- like processes and innervated by THIP fibers tri- pled [18]. This is in agreement with physiological observations of the onset of NA stimulation of the luteinizing hormone releasing-hormone release in prepubertal (P29) female rats [19]. In the SCN, though THIP fibers first appear at E18, their concentration remains insignificant until the end of fetal life. Over the first nine days of postnatal life, the concentration of THIP fibers increases abruptly, thus, highly exceeding that in the SCN of adults [20]. Further EM analysis identified many THIP fibers as axons contained synaptic and dense core vesicles. By the end of intrauterine development the THIP axons make immature synapses (presynapses) with the im- munonegative dendrites. After birth, first the symmetric (at P2) and then asymmetric (at P9) synapses appear. CA fibers might be suggested as a regulator of the presumptive oscillators, pep- Monoaminergic System in Ontogenesis pM tidergic neurons, which become functionally active perinatlly [21]. As to the paraventricular n., few THIP fibers arrive there even in fetuses (E19, 22; E17, 23), though they surround neurosecretory neurons only one-two weeks after birth [17, 23]. In addition to DA and NA, AD fibers also contribute to this innervation from Pl [24]. A number of EM double-labeling studies showed that in adults the CA fibers innervate vasopressin- [25] thyrotropin releasing-[26] and corticotropin releasing-[27] hor- mone-producing neurons in this region. The growing of the CA fibers to the ME is considered as the essential characteristic of the maturation of the tubero-infundibular system. THIP fibers first terminate there at E18. After the final settling of THIP neurons in the arcuate n., the ME receives the increased number of THIP fibers mainly from this region [14, 17] reaching maximum by puberty [22, 28, 29, 30]. From the physiological studies in adults, DA modulates the peptide release from the ME [31, 32]. Therefore, the establishment of the close relations between CA and peptidergic axons is of great importance. Double labeling technique showed first at LM [29, 33] and then at EM level [34], that the axo-axonal contacts between luteiniz- ing hormone-releasing hormone and CA fibers in the ME are established from P8 until P14, while the DA “innervation” of somatostatin axons is delayed for a week [29]. Although THIP fibers regularly spread to the periventricular region after E15, they reach the ventricular lumen only at E18-E20. This suggests the release of CA into the CSF, followed by their wide distribution through the brain. After E20, the frequency of these contacts drops and becomes practically absent in adults [17]. The same tenden- cy has been marked in phylogenesis [35]. Thus, from E13 the hypothalamus of rats is innervated by the CA fibers either belonging to the hypothalamic neurons or arriving via the MFB from the outside. Since this time, CA fibers progressively innervate such target regions as the diagonal band, septum, SCN, paraventricular n., etc. Besides, CA fibers sprout to the ME and the 3rd ventricle. b) Morpho-functional characteristics According to LM immunocytochemical observa- tions, THIP neurons undergo striking morpho- logical modifications relating to their differentia- tion. The earliest neurons (Fig. 6a) possess one or two short unbranched processes terminating with the growth cone. Two days later, in addition to uni- and bipolar, multipolar neurons first appear. From E18 onwards, THIP cell bodies increase in size and their processes in length [12]. Although the majority of THIP neurons undergo similar morphological differentiation, from E20 at least two neuronal populations could be recognized [12]. The first one consists of small uni- and bipolar neurons with short and narrow unbranched processes. These neurons are localized in ventral, mainly, the mediobasal hypothalamus (arcuate n.) (Fig. 6b, c). The second population includes large multipolar neurons with long highly ramified pro- cesses concentrated dorsally presumably in the zona incerta (Fig. 6d)[12]. Up-to-date, only THIP neurons of the SCN and arcuate n. (first population) have been studied in ontogenesis at the EM level [36]. The authors have failed to find out any principal (qualitative) differences in the ultrastructure of THIP neurons from E22 until P21. They are characterized by the relatively large nucleus and scanty cytoplasm as well as by the well-diveloped Golgi complex (Fig. 6e). As the differentiation of THIP neurons proceeds, their size increases slightly, the nuclei become indented, the dense core vesicles appear, thus, showing the secretory activity. The THIP neurons, at least from E22, are rarely innervated by immunonegative axons containing dense core vesicles and synaptic vesicles. These axons give rise mainly to axo-dendritic immature synapses (presynapses) in fetuses, transformed into the sym- metric and rarely asymmetric synapses after birth (Fig. 6f). The synaptogenesis continues at least till P21 that is manifested in the increase of both the number and complexity of synapses. Biochemical and histofluorescent techniques have provided the information on the synthesis of the final product, CA, and on their transfer via axons towards the target regions. According to biochemical data, TH activity first detected at E13 Di. M. V. UGruMov Fic. 6. THIP (a, b, d-f) and [*H]dopamine-labeled (c) catecholaminergic neurons and fibers in the hypothalamus of fetal (a) and postnatal (b, c, e-P9; f-P21) rats at light (a-d) and electron microscopic (e, f) levels. V-3rd ventricle; small arrow-axo-dendritic synapse; large arrow-randioactively labeled neuron; arrowhead-radioactively labeled fiber. a-ventral hypothalamus; b, c-arcuate n.; d-zona incerta; e, f-suprachiasmatic n. For other abbreviations see ipa 4 25. Shows a 12-fold increase prenatally and, then, [13]. 4-fold rise in four postnatal weeks. The activity of With histofluorescent technique first CA con- this enzyme is highly stimulated in embryonic taining neurons were detected in the hypothalamus tissue culture by depolarizing agent, veratridine at E12 [37]. By E18 the frequency of the fluores- Monoaminergic System in Ontogenesis 2 cent neurons increases significantly. In the anter- ior hypothalamus they are mainly concentrated in the paraventricular and periventricular n., while in the middle hypothalamus they are localized dorsal- ly, in the zona incerta and dorsomedial n. as well as in the mediobasal hypothalamus [28, 37]. From E18 to the puberty the modifications of the CA system are rather quantitative than qualita- tive. Until E20 the intensity of fluorescence in- creases progressively, followed by its drop by P9 and then it abruptly increase by P21 [28]. Accord- ing to biochemical, HPLC data, the content of DA at E21 reaches its level at P11. On the contrary, the concentration of NA increased four-fold over the same perinatal period. Aderenaline remains at the lowest level comparing with other CA both in fetus and neonates [38]. From comparison of histofluorescent and biochemical data follows that in parallel to the gradually increasing synthesis of CA their release predominates at the end of fetal life or soon after the birth. The specific uptake and K*-stimulated release of CA is the next essential functional characteristic of CA neurons. First radioautographic signs of the specific uptake of intraventricularly injected Birth = (e) fo) (e) a (e) (e} (?H)dopamine uptake (com/mg tissiuve) 45th 15th 16th 18th 20th 9th Age (days) Fic. 7. (Modified [28]). (°H)dopamine release (cpm/mg/ftaction) [H]DA by hypothalamic CA neuronal elements have been detected in fetal rats on E18 [39]. Further biochemical “isotopic” study in vitro has shown the simultaneous onset of the uptake and K*-stimulated Ca*?-dependent release of CA at E16 (Fig. 7) [28]. The uptake increased in 2,6 times for subsequent two days and, then, doubled between E20 and P9 reachng the adult level. As to K*-stimulated Ca*?-dependent release of CA, it increases considerably on E17 and remains at the same level in older fetuses and neonates, while doubled between P9 and P45 (Fig. 7) [28]. Thus, four stages in the development of the hypothalamic CA system could be distinguished. At the first stage (E12-E13), rare CA neurons and fibers of the MFB appear. The second stage (until E16) is characterized by the most intense neuro- nogenesis, followed by the onset of CA synthesis, uptake and K *-stimulated release. Over the third stage E16-E18, the final settling of CA neurons occurs simultaneously with the stimulation of synthesis, uptake and ion regulating release of CA. The fourth stage (until puberty) is mainly character- ized by completing of the CA neuron differentia- tion and the establishment of the axonal pathways Kt KiCa=O Kt . i ’ { * 15th day x 16th day o Y9thday di e e 45th day 60 N 30 | : ee me Nene” NON oo Ne e Ay, 40 50 60 70 8C 90 Time (min) Specific [7H]dopamine uptake and K *-stimulated release in the hypothalamus of fetal and postnatal rats. 24 M. V. UGRUMOvV for CA transfer to the target neurons and circum- ventricular organs. 2. SEROTONINERGIC SYSTEM a) Architectonics Cell bodies. Dorsal and median raphe n. are known to be the main sources of the hypothalamic 5-HT innerva- tion in adult mammals [40]. A number of im- munocytochemical [41, 42, 43] studies have shown that 5-HT neurons in the raphe n. of rats first appear at E12-E13. This is followed by their differentiation and axonal growing to the target regions. The question on the hypothalamic localization of the 5-HT neurons still remains opened. First radioautography has shown that some hypo- thalamic neurons have an ability for specific up- take of [PH]5-HT following its intraventricular injection [41]. These neurons are localized in the SCN of fetuses (E18) and in the dorsomedial n. of neonatal (P9) (Fig. 8a) [41] and adult [44] rats. Further 5-HT immunocytochemistry attempted to detect either 5-HT accumulating or 5-HT synth- esizing neurons in the hypothalamus in ontogene- sis. As 5-HTIP cells have not been observed in the hypothalamus of intact animals [41, 45, 46, 47], the 5-HT synthesis and its reuptake was stimulated by the pretreatment of fetuses (E18), neonatal (P9), and adult rats with L-tryptophan, precursor of 5-HT synthesis, and pargyline, inhibitor of monoamine oxidase [45, 47, 48]. In these experi- ments, two populations of 5-HTIP neurons, in the anteriolateral hypothalamus and the dorsomedial *~ % n., were observed in fetuses and neonates (Fig. 8b,c) [47] while in adults they were localized only in the dorsomedial n. [45]. In fetuses both populations, but in neonates only the first one, consist of intensely immunostained bipolar elongated neurons with long processes (Fig. 8b). In neonates [47] and adults [45] the neurons of the dorsomedial n. characterized by a fairly weak immunostaining, are oval, small in size with one or two short processes (Fig. 8c). 5-HT immunostaining provoked by pharmacological pretreatment has been prevented by the prelimin- ary infusion of the uptake inhibitor, fluoxetine. It means that immunostaining is accounted for rather by the ability for specific uptake of 5-HT from the enviroment than by its synthesis from L- tryptophan catalyzed by tryptophan hydroxylase [47, 48]. On the contrary, decarboxylation of 5- hydroxytryptophan to 5-HT with decarboxylase of the aromatic amino acids is rather probable, as both neuronal populations become 5-HT immuno- positive following pretreatment of aminals with pargyline and 5-hydroxytryptophan instead of L- | tryptophan [47]. Such characteristics of 5-HTIP neurons as an ability for the specific uptake and decarboxylation of the 5-HT precursor made them similar to the cultivated in vitro hypothalamic 5-HT-like neurons of fetal mice [49], on the one hand, and to the APUD cells [50], on the other. Over the last decade, the attention has been focused on the origin, functional significance and fate of these cells, however, this problem is far from understanding [45]. According to our sugges- tion, the 5-HTIP neurons might serve as a tempor- Fic. 8. [°H]5-HT-labeled (a, arrowhead) and 5-HTIP (b, c) neurons in the anteriolateral region (b, large arrow) and dorsomedial n. (a, c; small arrow) of the hypothalamus in rats at P9. Monoaminergic System in Ontogenesis WS al store of 5-HT either released from the 5-HT axons of the MFB, or circulated in the CSF [47]. This is in agreement with phylogenic observations of 5-HTIP neurons in the periventricular hypotha- lamic area in intact lower vertebrates. These cells are supposed to provide an intermediate link in the neurohumoral regulation of the neuroendocrine centers [35]. As to the fate of the 5-HTIP cells in the anterolateral hypothalamus of fetal and neonatal rats (see above), probably, they repre- sent the transient neuronal population partly ex- pressed fenotype of 5-HT neurons [47]. In conclusion, the hypothalamus of fetuses and neonates contains two populations of neurons which become 5-HTIP after the pargyline and L-tryptophan pretreatment, apparently due to the 5-HT absorption from the environment. More- over, these cells are characterized by the specific uptake of 5-HT and the capacity to decarboxylase 5-hydroxytryptophan into 5-HT. Thus, the 5-HT neurons of the raphe nucleus are the main source of the hypothalamic innervation, while the presence of the hypothalamic 5-HT neurons is still under question. Nerve fibers. Although 5-HTIP neurons first appear in the raphe n. of the rat as early as E12, their axons first 6150 0 ~ oO a) = =) ~ i 2 D Birth c £ £ 1000 2409 4 _ 2 D o o xs = © v Ss E s i 500 + 2 50 2 Se Te SOD 16th 18th 20th 9th 45th Age (days) reach the hypothalamus at E14-E16 [41, 42, 43, 46]. The concentration of the 5-HT fibers in- creases abruptly from E16 until E18, that is parti- cularly evident for the perichiasmatic area. Since this time, 5-HTIP fibers invade preoptic area (PA), septum and the diagonal band [42], 1.e., those regions which already contain their targets, e.g., luteinizing hormone releasing-hormone neurons [51]. The special attention has been paid to the 5-HT innervation of the SCN because of its role in the control of the circadian rythmic activity of neuroendocrine and other functions [52]. Accord- ing to LM radioautographic and immunocyto- chemical data, 5-HT fibers first invade the SCN at E18 [41, 53]. By the end of fetal life, the concen- tration of 5-HT fibers particularly in its ventro- lateral region increases considerably reaching the adult level around P9 [41, 42, 46]. Further EM analysis showed that in the oldest fetuses 5-HTIP axons make axo-dendritic immature synaptic con- tacts (presynapses) which are transformed neona- tally into symmetric, and rarely asymmetric synapses [53]. In evaluating the functional significance of 5-HT input into the SCN, it is necessary to take into account that in adults the 5-HT axons innervate vasoactive intestinal peptide neurons [54] control- Ktca=o Kt # 16th day x 17th day © 9thday e 45th day od “Se Br 40 50 60 70 80 90 Time (min) Fic. 9. Specific [PH]5-HT uptake and K*-stimulated release in the hypothalamus of fetal and postnatal rats. (Modified [57]). 26 M. V. UGRuMov ing their functional activity [55]. These cells are settled in the SCN at E19 [52], and become functionally active as spontaneous oscillators by the end of intrauterine development [56]. In the middle (mediobasal) hypothalamus, 5-HT fibers are widely distributed from E18 [41] invad- ing the arcuate and dorsomedial n. as well as the ME. In the ME, 5-HT fibers terminate in the external zone or even penetrate to the area of the primary portal plexus abutting on the portal capil- laries [41]. . The essential characteristic of the 5-HT system both in pre- and postnatal life is the progressive penetration of 5-HT fibers to the cerebral ventri- cles. 5-HT released to the CSF is believed to reach the target regions of the brain, or is transferred through the ME to the hypophysial portal circula- tion and finally to adenohypophysis [41]. Thus, 5-HT axons first arrived at the hypothala- mus at E14-E16, spread to the target regions over the perinatal period, followed by the establishment of specialized contacts with target neurons (synapses) as well as with the hypophysial portal circulation and cerebral ventricles. b) Morpho-functional characteristics According to radioautographic data, the specific uptake of intraventricularly injected [*H]5-HT by the hypothalamic nerve fibers in vivo is already evident at E18 [41]. Using biochemical “isotopic” technique the specific uptake by the hypothalamic tissue in vitro has been first detected even earlier, at E16, while the [*H]5-HT release by the depolar- izing concentrations of K* in a Ca‘?-dependent manner is evident from E17 [57] (Fig. 9). It means that Ca’*-channels become functionally active and S-HT is intravesicularly stored from E17 onwards. The most rapid increase of the [?H]5-HT uptake occurs from E16 until E18 [57] that is apparently a result of the most intense growing of 5-HT fibers to the hypothalamus as well as of the significant activation of the 5-HT-binding protein synthesis in the brain [58]. From E18 to P45, the value of the specific uptake retains at the constant level. This shows that despite the progressive increase in the number of 5-HT fibers, their number per mg of tissue does not change significantly. In contrast to uptake, the Ca‘?-dependent and K*t-stimulated relase of [*H]5-HT increases considerably over the perinatal period reaching an adult level by P9 (Fig. 9) [57]. Thus, the hypothalamic 5-HT system could be functionally active from E16-E17, i.e., shortly after first arrival of 5-HT fibers, that is manifested by the onset of the specific uptake and K*- stimulated release of 5-HT. It. NEUROHUMORAL CONTROL OF DEVELOPMENT OF MONOAMINERGIC SYSTEMS 1. CATECHOLAMINERGIC SYSTEM a) Sexual steroids Biochemistry The sexual difference in the content of CA, mainly of NA, in the hypothalamus of rats from P10-P12 onwards, suggests the role of sexual ster- oids (SS) in the development of the CA system [59, 60]. This has been checked comparing the content of CA in neonatally castrated and sham-operated male rats as well as in neonatally androgenized females. The dynamic of the NA content in androgenized females resembles that in control males [59, 60], though the maximum has been achieved earlier, by P60. Similarly, the NA level in the castrated males gradually rises until P180 as that in the control females excepted the peak at P60 (Fig. 10a) [59]. In contrast to NA, no clear sexual difference in hypothalamic DA content has been observed in neonates [60, 61]. The castration of males does not modify the DA content until P120, while at P180 its level highly exceeds that in control males and androgenized females. Similarly to NA, the DA content shows a decline in its level with age both in control males and androgenized females (Fig. 10b) [59]. The authors suggested that the initial in- crease of CA content in the control males and androgenized females is accounted for the neona- tal androgen promotion of the growing of the CA fibers. Sexual differences in the DA content were also observed in the arcuate n. and ME: the DA level of males exceeds that in females. The neonatal Monoaminergic System in Ontogenesis 27 Noradrenaline nmol/g wet weight of tissue Age In days Dopamine nmol/g wet welgnt of tissue 120 Age in days Fic. 10a, b. Mean (+SEM) concentrations of noradrenaline (a) and dopamine (b) in the hypothalamus at various postnatal ages. ook <005585 Sh 00i as hae ().001e ns-not significant. Upper panel-male; 1 sham- castrated; M castrated. Lower panel-female; 0 control; M androgenized [59]. castration of males causes the reduced DA level in adults, while neonatally androgenized females show an increased level of this neurotransmitter [62, 63]. Though the mechanism of this action is still uncertain, the effect of SS on the TH activity in ontogenesis is rather probable [13]. Morphology Further immunocytochemical study using anti- bodies to TH and DA-/-hydroxylase attempted to verify the role of SS in the development of the hypothalamic CA system [64]. The perinatal (E16- P10) androgenization of females causes decrease in both the number of THIP neurons and fibers belonged not only to the hypothalamic but also to the mesencephalic neurons. Conversely, postnatal injection of testosterone “masculinizes” only the content of cell bodies. These observations suggest that the critical period of the differentiation of hypothalamic CA neurons corresponds to the neonatal life while that of mesencephalic neurons begins earlier, i.e., in fetal life [64]. According to recent data, sexual dimorphism of DA neurons of diencephalon is first manifested without any influence of SS [65]. Thus, the DA neurons taken into tissue culture from fetuses, at E14, 1.e., before the onset of SS secretion, show more intense outgrowth of processes and higher uptake of [*H]DA in females than in males. The neurons taken into the tissue culture three days later show the only difference in specific uptake. Probably this sexual dimorphism is accounted for the earlier genesis of DA neurons in females comparing with males [65]. The role of either neurotropic factors or steroids secreted by glia cells in tissue culture in the sexual dimorphism of DA neurons [65, 66] also cannot be excluded. b) Other hormones Thyroid hormones The convincing evidence of the role of thyroid hormones in the differentiation of CA neurons has been obtained in the hypothalamic tissue culture of fetal mice (E16). Triiodthyronine results in the increased size of DA neurons and stimulates the 28 M. V. UGRUMOV growth and arborization of DA processes associ- ated with a rising of specific uptake of [7H]DA for 35%. The authors suggest either the direct or indirect, through glial cells, trophic action of thyr- oid hormones on the DA neurons [67]. To evaluate the role of thyroid hormones in the development of the hypothalamic CA system in vivo, the rats at P8-P10 were pretreated with thyroxine. This results in the increased contents of NA and DA at P16-P21, followed by their levelling or even reversion [68]. The authors hold the opinion that the thyroid hormones provide a tem- poral stimulating effect on the brain development. Adenohypophysial hormones Very recently, the stimulating effect of hor- mones of the intermediate lobe on the differenti- ation of the hypothalamic DA neurons has been suggested. In fact, the co-cultivation of the embryonic hypothalamus with the intermediate lobe results in transient increase of the specific uptake of [PH]DA [69]. More clear data were obtained on the neonatal imprinting effect of a-melanotropin in the feed- back regulation of DA neurons of the arcuate n. [70]. This regulation is established in rats between P4 and P8, that coincides with the peak in plasma melanotropin [71]. The passive immunization with antiserum to melanotropin at P5-P6, but not at P11-P12, makes the DA neurons of the arcuate n. non-sensitive to this hormone both in adult males and females [70]. In addition to the hormones of the developing organism, maternal hormones also control the maturation of the hypothalamic CA system. For example, maternal prolactin penetrated to pups in the course of suckling provides a long-term effect on the functional activity of DA neurons of the arcuate n. [72]. The depletion of maternal prolac- tin in pups at P2-P5, but not at P9-P12, by daily injections of bromcriptine to lactating females results in depression of DA turnover and in the elevation of the plasma prolactin when tested at P30-P35 [72]. Thus, the hormones control the development of the hypothalamus modifying the neuronal dif- ferentiation and metabolism, networks of nerve fibers, synaptic organization, neuroendocrine reg- ulation and behaviour. Sexual steroids seem to be the most potent factor involved in so-called “sex- ual differentiation” of the hypothalamus and its MA system. 2. SEROTONINERGIC SYSTEM a) Sexual steroids Biochemistry Up-to-date, there are controversial data on the ontogenetic sexual difference in the hypothalamic content of 5-HT. Watts and Stanley [73] failed to find out any sexual difference in the 5-HT content in period from P1.5 until P80, while other authors demonstrated its transient divergence [74]. Still, clear difference in the hypothalamic 5-HT content in adult male and female (higher level) rats [75] rises a question on the role of SS in the develop- ment of the 5-HT system. After castration of male rats at P1 the hypotha- lamic level of 5-HT is not changed comparing with control by P12. On P60 and P75, 5-HT depletes significantly, and is no longer seen at P90. Follow- ing three months, the concentration of 5-HT be-. comes higher in castrated rats than in control [76], even reaching its level in adult females [75]. In neonatally castrated adult males as in adult females the 5-HT content in the anterior hypotha- lamus exceeds that in the control males for 67%, while in the posterior hypothalamus for 46% [75]. Morphlogy The sexually dimorphic medial preoptic n. has been chosen to study the influence of SS on the distribution of 5-HT fibers. In both sexes this nucleus includes the medial part with the low concentration of 5-HT fiber, particularly in its center, and the lateral part with the extensive network of 5-HT fibers [77]. In perinatally or neonatally androgenized females, as in the control males, the area of the low fiber density (medial and central medial preoptic n.) is considerably larger, while the area of the high fiber density (lateral MPA) is proportionally smaller comparing with the control females (Fig. 11). This is more evident in perinatally androgenized females than in post- natally androgenized ones. Adult gonadectomy does not affect the distribution of 5-HT fibers in Monoaminergic System in Ontogenesis 29 Male + oil Female + oil C7 a MePO SV 7 SBS aa eee psy , y Ee is PyPO fpr ea A AS rhe f eS Qa EEE i wf ie OY BN uae Biter = TRIS RAN ARS oe nd eB m i ie NY . oe F177 \) PERE: c ee Hp as << ¥- LPO . \ anes cus WR 2 MPO 54 OS e eis ee SSeS Female+TP Fic. 11. Line drawings of the medial preoptic area at the level of the medial preoptic n. (MPN) illustrate the distribution of 5-HT-stained fibers in this nucleus and adjacent regions in oil-treated female (A), oil-treated male (B), and perinatally testosterone propionate-treated female (C) rats. m, c, l-medial, central, and lateral parts of the MPN, respectively. MPO-medial preoptic area. any area. These data show that the sexually dimorphic distribution of 5-HT fibers in the MPA is determined by the perinatal SS environment [77]. Physiology In this study, such physiological integrative char- acteristics of the hypothalamic 5-HT system as specific uptake and K*-stimulated Ca*?- dependent release of [PH]5-HT in vitro have been also compared in intact and neonatally castrated adult males, as well as in adult females [78]. In intact adult males, both parameters were signif- icantly lower than in females, while reaching the female level after castration. These data are considered as further evidence of the supressive effect of androgens on the development of the hypothalamic 5-HT system. The membrane effect of SS modifying their permeability also cannot be exculded [78]. Neuroendocrine regulation It is well-known that the neonatal androgeniza- tion of the hypothalamus results in the establish- ment of the tonic gonadotropin secretion. Con- comitantly, the mechanism of the control of gona- dotropin secretion seems to become non-sensitive to the regulative effect of 5-HT. In fact, intact For other abbreviations see [77]. females and neonatally castrated males, in contrast to intact males, show the luteinizing hormone release in response to 5-hydroxytryptophan, a 5- HT precursor. On the contrary, neonatal androgenization abolished the lutenizing hormone response to 5-HT in prepubertal female rats [79]. Thus, the sexual differences in the hypothalamic 5-HT system are manifested in the wide range diapason, from the 5-HT content to the neuroen- docrine regulations, that is the consequence of the testosterone masculinizing effect over the critical perinatal period. b) Other hormones In adition to SS, thyroid hormones, corticoster- oids, etc. contribute to the control of the dif- ferentiation of 5-HT system but this problem has been evaluated somewhere else. IV. FUNCTIONAL SIGNIFICANCE OF MONOAMINES 1. CATECHOLAMINES The role of CA in “sexual differentiation” of the hypothalamus occurred to be the most attractive problem. Their action is conveyed via pre- and postsynaptic a- and f-receptors. Presynaptic a- 30 M. V. UGRUMOov Tyrosine ay | TOH Metabolites Fie. 12. Neuron terminal Presynapse Synapse Postsynapse A proposal for the metabolism of neuronal norepinephrine in the neonatal rat. The following numbered steps refer to the numbered pathways in the figure. 1) Norepinephrine (NE) is synthesized from tyrosine by - tyrosine hydroxylase (TOH), the rate limiting enzyme. 2) Newly synthesized NE is incorporated into the synaptic storage vesicles or becomes part of the cytoplasmic pool. 3) With a neural stimulus, the synaptic vesicle fuses with the synaptic membrane, and NE 1s released into the synapse. 4) Synaptic NE will bind to postsynaptic a-receptors (step 5) or B-receptors (step 6). Synaptic NE (step 4) can also bind to a presynaptic a-receptor (step 7), which will inhibit the synthesis of TOH and the subsequent formation of new NE. 8) Synaptic NE and its physiological stimulus are terminated by rapid uptake of NE back into the neutron. The uptake mechanism is a-receptor linked. NE taken back into the neuron can be reincorporated into synaptic storage vesicles (step 9), metabolized by monoamine oxidase (MAO; step 10), or leak back out into the synapse (step 11) [81]. receptors are related to the specific reuptake of CA and to the negative feedback regulation of TH (a). Postsynaptic a; and f-receptors (/8;, 2) are responsible for the adequate reaction of target- cells (Fig. 12). Although (; and (-receptors are similar in their structure, the first ones are mainly sensitive to NA while the second to AD. Moreov- er, f)-receptors are mainly related to the post- synaptic membrane, while /5-receptors are also localized extrasynaptically [80, 81]. Morphogenesis The enlargement of the sexually dimorphic n. of the PA in the adult male and female rats were observed after the postnatal (P1-P5) stimulation of f2-receptors by agonist, salbutamol [82]. It is noteworthy that the pre- and neonatal treatment of males even with high concentrations of androgens or estrogens fails to increase the volume of this region [83]. According to the authors’ suggestion the effect of salbutamol could be accounted for by either its promotion of the neuron migration to the centre of this region, or by preventing of the neurons from the cell death [82]. The regulative influence on neuronogenesis is probably low, be- cause, as in the hypothalamic region, it practically completes by the end of fetal life [84]. Neuroendocrine regulation A number of physiological and pharmacological Monoaminergic System in Ontogenesis Sil studies have shown that SS, CA ligands and the drugs interfering with the CA metabolism, when applied neonatally, contribute to the “sexual dif- ferentiation” of the hypothalamus (Table 1) [6]. The mechanisms of their action remain uncer- tain, as substances mentioned above are multifunc- tional. Thus, SS apparently control the number of a-and #-receptors [6], on the one hand, and in- fluence the metabolism of NA, on the other [85]. TABLE 1. metabolism of catecholamines In turn, NA influences the concentration of SS receptors in the hypothalamic region [86]. Moreover, NA stimulated {-receptors are sup- posed to prevent androgenization of the hypotha- lamus by inhibition of either aromatization of testosterone or of the estradiol uptake by the neuronal nucleus. The last mechanism is closely related to the SS modulation of genome [81, 87]. Moreover, the stimulation of /-receptors inten- Effects of neonatally applied sexual steroids and drugs related to either the physiological action or Experimental Physiological Mechanism eld ast ‘ Og of action Ecce Neonatal Persistent a) phenoxybenzamine aAN { androgenization estrus in b) phentolamine aAN | of females [81] adulthood c) propranolol BAN — dja&c — Sines C — f) methyl-p-tyrosine inhibitor of a CA synthesis g) tyramine Stimulator of NA release Neonatal Female a) clonidine aA — androgenization sexual b) prazosine a, AN — of female & behavior c) yohimbine a,AN -— gonadectomy in adulthood [6]; Intraventricular Nuclear accumulation a) phenoxybenzamine aAN { injection of [*H]estradiol b) isoproterenol BA { of [*H]testo- in hypothalamus c) isoxuprine BA | sterone to d) hydroxybenxylpindolol BAN — female at P4 e)akd t [87] f)b&d t g)c &d t Neonatal LH response a) clonidine aA — treatment of to estradiol b) prazosine a, AN t females with & progesterone c) yohimbine a,AN t drugs & gonad- in adulthood d) salbutamol oA t ectomy in e) alprenolol BAN | adulthood [6, 88] f) isoprenaline BA | Neonatal female sexual a) clonidine aA | treatment of behavior b) prazosine a, AN — females with c) yohimbine a,rAN — drugs & gonadec- d) salbutamol AoA t tomy in e) alprenolol BA t adulthood [6] f) isoprenaline Neonatal female sexual a) isoprenaline aA t treatment of behavior b) salbutamol 5A — males with c) alprenolol BAN = drugs. Castration & estradiol and progesterone in adulthood [6]. a: a-receptor; A: agonist; AN: antagonist; 8: 8-receptor; f: increased |: decreased; “—”: no effect. 32 M. V. UGRuMov sifies the female sexual behavior, but it inhibits the female type of gonadotropin secretion. On the contrary, the blockade of a-receptors results in the increase of the LH-surge release, while decreasing of female sexual behavior. Activation of ap- receptors supresses the female sexual behavior, though has no effect on the LH-release (Table 1) [88]. From the physiological point of view, the in- terfering of CA, mainly DA, and SS in their action on the gonadotropin secretion is of particular interest. Thus, the gonadotropin secretion is under the DA inhibitory control in female rats, but not in males, until P20. The physiological response of both females and males are completely reversed after their neonatal androgenization and castra- tion, respectively [89]. The hypothesis that the DA neurons in the anteroventral periventricular n. of the PA mediate the SS effects on gonadotropin secretion seems to be reasonable. The failure to show a phasic gonadotropin secretion in response to estrogen and progesterone in adult females after neonatal androgenization that caused the reduc- tion of the population of DA neurons, is in favor of this idea [64]. The direct inhibiting action of DA on prolactin secretion begins in rats only after birth, at P3. First at this time, the inhibitor of DA receptors in adenohypophysis (pimozide) becomes effective in the increase of prolactin plasma level [90]. Thus, CA are believed to mediate the organizing effects of SS on the hypothalamus in the critical period of ontogenesis inducing the permanent sex- ual differences in morphogenesis of the hypothala- mic target regions, e.g., sexually dimorphic n. of the PA, the gonadotropin secretion pattern, and sexual behaviour. 2. SEROTONIN Autoregulation The good example of the autoregulation is the population of the hypothalamic 5-HT-like neurons possenssing the specific uptake of 5S-HT and synth- esizing, = 5-H from its precursor, ~ 5- hydroxytryptophan (see above). The repeated treatment of the embryonic hypothalamic tissue culture with 5-HT agonist, 8-hydroxy-2-[di(n- propyl)amino]tetralin, results in an increased num- ber of 5-HT-like neurons with the elevated level of aromatic-L-amino acid decarboxylase. Both events are highly suppressed by a specific blocker of 5-HT receptors, metergoline [49]. Neuronogenesis One of the most important evidence on the role of 5-HT in the differentiation of the target neurons in vivo has been obtained evaluating the prolifera- tive activity of cell-precursors with long-survival [(°H]thymidine radioautography following pharma- cological 5-HT depletion by pCPA in fetuses [7]. In some hypothalamic areas, zona incerta and paraventricular n., the retardation of the neuronal genesis (prolongation of cell-precursors prolifera- tion) has been detected, while in the others, the SCN and ventromedial n., the effect of 5-HT depletion is not evident. According to the authors’ suggestion these neurons are originated before the first delivery of 5-HT either via MFB or with the CSF. Sexual differentiation Although androgens are known to be the most important inductor of hypothalamic “sexual dif- ferentiation”, 5-HT is believed to play an in- termediate or even a leading role in its action. Thus, the neonatal inhibition of the 5-HT synthesis in female rats results in the increased volume of the sexually dimorphic n. to that of intact males, while it does not alter the SS level in the blood (Cay). These data have been significantly revised by further study of Jarzab and Dohler [92] who used physiological models of female rats neonatally treated with: a) Testosterone; b) Testosterone and pCPA; and c) Testosterone and L-tryptophan. According to their results, the postnatal inhibition of 5-HT synthesis does not modify the effect of testosterone either on the gonadotropin secretion or on the expression of sexual behavior. Still, postnatal stimulation of 5-HT synthesis inhibits considerably the expression of sexual behavior while it does not modify the gonadotropin release [92]. The discrepancies in the above mentioned re- sults are apparently accounted for by the interfer- Monoaminergic System in Ontogenesis ing of many biologically active substances during the critical periods. On the other hand, the pharmacological models are usually far from ideal, e.g. pCPA inhibits the 5-HT synthesis only for 50- 70%, while modifying the metabolism of CA, proteins, etc. at the same time. Neuroendocrine regulation One of the most importnat characteristic of the 5-HT is functioning in the maturation of the recep- tors on the targets. Thus, the 5-HT binding ability of the hypothalamic tissue has been estimated in postnatal female rats. The specific 5-HT-binding sites were already observed at P5; their number doubled between P16 and P27 and continued to increase at least until P37 [93]. It is of particular interest that even the fetal 5-HT receptors are capable of responding to changes in the 5-HT level: the maternal administration of the 5-HT depleter, pCPA, results in the elevated number of receptors in newborn pups, whereas 5-HT agonist, 5-methoxytryptamine, provides the opposite effect. It means that the 5-HT receptors are functional in fetuses and show the same plastic changes as in adults [94]. In other words, 5-HT may serve as a potential developmental signal even during intrauterine development [95]. Further physiological experiments have proved the functional potency of 5-HT receptors in the developing hypothalamus. Thus, 5-HT is ineffec- tive as the modulator of the SS action on LH secretion in female rats at P5. This effect first appears between P8 and P12 while increasing abruptly and remaining at the same high level until P18. Then, the response to 5-HT decreases prog- ressively until P50 [93]. This dynamic is quite different from the 5-HT induced release of prolac- tin in ontogenesis. The 5-HT regulative effect on prolactin secretion being absent over first postnatal week appears at P12 followed by its strengthening until puberty [93]. Thus, several aspects of the 5-HT action on the developing hypothalamus could be distinguished: autoregulation of 5-HT neuron differentiation, control of neuronogenesis, the regulation of “sex- ual differentiation’, etc. Vv. CONCLUSIONS 1) Hypothalamic catecholamine- and _ seroto- ninergic systems undergo synchronous develop- ment that is manifested first in the genesis of catecholamine- and serotoninergic neurons in the hypothalamus and brain stem long before birth (E12-E13). Further neuron differentiation results in the expression of the key enzymes and monoamine synthesis, the onset of specific uptake and potassium-stimulated release of monoamines. One of the most important attribute of the dif- ferentiation is the growing of axons to the target neurons, hypophysial portal circulation and III ventricle, followed by the establishment of special- ized contacts: synapses, axo-vascular and axo- ventricular. 2) The development of the monoaminergic sys- tems is under the hormonal control. Some hor- mones (thyroid hormones) provide a temporal stimulating effect on the neuron differentiation, while others, e.g., sexual steroids, have a program- ing irreversible action that occurs in the critical perinatal period. In the last case, the hormones modify considerably the neuron differentiation and metabolism, networks of fibers, synaptic orga- nization, and, finally, neuroendocrine regulation and behaviour. 3) The hypothalamic monoamines appeared ear- ly in ontogenesis are characterized by a wide range spectrum of their physiological action. First, they control the genesis and further differentiation of the target neurons. Later, over critical perinatal period, they mediate the hormonal programing influence, thus, modifying the architectonics, metabolism, receptors, etc. of the hypothalamic target regions and, finally, the adenohypophysial hormone secretion and behaviour. REFERENCES 1 Barraclough, C. A. and Wise, P. M. (1982) Endo- crine Rev., 3: 91-119. 2 Collu, R. (1989) In “Pediatric Endocrinology”. Ed. by R. Collu, J. R. Ducharme and H. J. Guyda, Raven Pres, New York, pp. 1-36. 3 Fuller, R. W., Clemens, J. A. (1981) In “Serotonin: Current Aspects of Neurochemistry and Function”, Ed. by B. Haber, S. Gabay, M. R. Issidorides and S. 13 14 15 16 7) 18 19 20 21 Upp 34 G. A. Alivesatos, Plenum Press, New York, Adv. Exp. Med. and Biol., 133, pp. 431-444. | Jackson, I. and Mueller, G. P. (1982) In “Biological Regulation and Development. Hormone Action”, Ed. by R. F. Goldberg and K. R. Yamamoto, Plenum Press, New York, pp. 127-202. Buznikov G. A. (1991) Neurotransmitters in Embryogenesis. Harwood Acad. Publ., London, 256 (in press). Jarzab, B., Kokocinska, D., Kaminski, M., Gubala, E., Achtelik, W., Wagiel, J. and Dohler, K. D. (1990) In “Hormones Brain and Behavior in Verte- brates. I. Sexual Differentiation, Neurochemical Aspects, Neurotransmitters and Neuropeptides. Comp. Physiol.” Ed. by J. Balthazart. Karger, Bas- al, Vol. 8, pp. 41-50. Lauder J. M. (1983) Hormonal and humoral in- fluences on brain development. Psychoneuroendoc- rinology, 8: 121-155. Lauder, J. M., Wallace, J. A. and Krebs, H. (1981) Adv. Exp. Med. and Biol., 133: 477-506. Jonakait, G. M. and Black, I. B. (1989) In “Hand- book of Experimental Pharmacology. Catechola- mines II”, Ed. by U. Trendelenburg and N. Weiner, Vol. 90/II: pp. 137-160. Borisova, N. A., Balan, I. and Ugrumov, M. V., unpublished. Specht, L. A., Pickel, V. M., Joh, T. H. and Reis, D. J. (1981) J. Comp. Neurol., 199: 233-253. Ugrumov, M. V., Taxi, J., Tixier-Vidal, A., Thi- bault, J. and Mitskevich, M. S. (1989) Neurosci- ences, 29. 135156: Friedman, W. J., Dreyfus, C. F., McEwen, B. S. and Black, I. B. (1989) Dev. Brain Res., 48: 177- 185. Daikoku, S., Kawano, H., Okamura, Y., Tokuzen, M. and Nagatsu, I. (1986) Dev. Brain Res., 28: 85- 98. Chan-Palay, V., Zaborszky, L., Kohler, C., Gold- stein, M. and Palay, S. L. (1984) J. Comp. Neurol. 227: 467-496. Specht, L. A., Pickel, V. M., Joh, T. H. and Reis, D. J. (1981) J. Comp. Neurol., 199: 255-276. Ugrumov, M. V., tixier-Vidal, A., Taxi, J., Thi- bault, J. and Mitskevich, M. S. (1989) Neurosci- ence, 29: 157-166. Wray, S. and Hoffman, G. (1986) Brain Res., 399: 327-331. Clough, R. W., Hoffman, G. E. and Sladek, C. D. (1988) Brain Res., 446: 121-132. Van den Pol, A. N. and Tsujimoto, K. L. (1985) Neuroscience, 15: 1049-1086. Reppert, S. M. and Uhl, G. R. (1987) Endocrinolo- gy, 120: 2483-2487. Hyyppa, M. (1969) Z. Zellforsch., 98: 550-560. 23 24 25 26 ZA 28 29 30 Sil 32 33) 34 36 37 38 39 40 M. V. UGRuMov Khachaturian, H. and Sladek, J. R. (1980) Pep- tidkeSawle 7-95. Foster, G. A., Schultzberg, M., Goldstein, M., Hokfelt, T. (1985) J. Comp. Neurol. 236: 348-381. Nakada, H. and Nakai, Y. (1985) Brain Res., 361: DAV=297 Nakai, Y., Shioda, S., Ochiai, H. and Kozasa, K. (1986) Curr. Top. Neuroendocrinol., 7: 135-160. Liposits, Zs., Sherman, D., Phelix, C. and Paull, W. K. (1986) Brain Res., 85: 95-106. Borisova, N. A., Sapronova, A. Y., Proshlyakova, E. V. and Ugrumoy, M. V. (1991) Neuroscience, 43: 223-229. Ibata, Y., Fukui, K., Obata, H. L., Tanaka, M., Hisa, Y., Sano, Y. (1982) Brain Res. Bull., 9: 407- 415. Loizou, L. A. (1971) Z. Zellforsch., 114: 234-252. Kordon, C., Enyjalbert, A., Hery, M., Joseph- Bravo, P. I., Rodsztein, W. and Ruber, M. (1980) In “Physiology of the Hypothalamus. Handbook of the Hypothalamus”, Ed. by P. J. Morgane and J. Panksepp, Dekker Inc, New York, Vol. 2, pp. 253- 306. Kuljis, R. O. and Advis, J. P. (1989) Endocrinolo- gy, 124: 1579-1581. Ibata, Y., Tani, N., Obata, H. L., Tanaka, M., Kubo, S., Fukui, K., Fujimoto, M., Knoshita, H., Watanabe, K. and Sano, Y. (1981) Cell Tiss. Res., — 216: 31-38. Ugrumov, M. V. (1991) Int. Rev. Cyt., 129: 207- AST. Parent, A., Poitras, D. and Dube, L. (1984) In “Handbook of Chemical neuroanatomy. Vol. 2. Classical transmitters in the CNS. PArt I”, Ed. by . Bjorklund and T. Hokfelt, Elsevier, Amsterdam, pp. 409-439. Ugrumov, M. V., Vladimirov, S. V., Kasmambeto- va, S., Popov, A. P. and Thibault, Neuroscience, submitted. Lichtensteiger, W. and Schlumpf, M. (1981) In “Steroid Hormone regulation of the Brain”, Eds. K. Fuxe, J. A. Gustafsson and Wetterberg. Oxford, Pergamon Press, pp. 161-172. Trembleau A., Ugrumov M., Bernabe J. and A. Calas Development of the hypothalamic catechola- minergic system and and its pharmacological deple- tion during ontogenesis. Society for Nueroscience, 21st Annual Meeting, 1991, Abstracts Vol. 17, Part Jay Do. JIVE). Ugrumoyv, M. V., Mitskevich, M. S., Halasz, B., Kiss, J. and Borisova, N. A. (1986) Int. J. Dev. Neurosci., 4: 101-111. Steinbusch, H. W. H. and Nieuwenhuys, R. (1981) In “Serotonin: Current Aspects of Neurochemistry and Function”, Ed by B. Haber, S. Gabay, M. R. 4] 42 43 44 45 46 47 48 49 50 51 2 53 54 55 56 S// 58 a9 60 61 62 63 Monoaminergic System in Ontogenesis 35 Issidorides and S. G. A. Alivisatos, Plenum Press, New York, pp. 7-35. Ugrumov, M. V., Taxi, J., Mitskevich, M. S. and Tramu, G. (1986) Dev. Brain Res., 30: 75-84. Wallace, J. A. and Lauder, J. M. (1983) Brain Res. Bull., 10: 459-479. Aitken, A. R. and Tork, I. (1988) J. Comp. Neurol., 274: 32-47. Beaudet, A. and Descarries, L. (1979) Brain Res., 160: 231-243. Frankfurt, H., Lauder, J. and Azmitia, E. (1981). Neurosci. Lett., 24: 227-232. Lidov, H. G. W. and Molliver, M. E. (1982) Brain Res. Bull., 8: 398-430. Ugrumovy, M. V., Taxi, J., Steinbusch, H. W. M., Tramu, G. and Mitskevich, M. S. (1989) Dev. Brain Res., 46: 233-241. Ugrumovy, M. V., Taxi, J., Steinbusch, H. W. M.., Tramu, G. and Mitskevich, M. S. (1988) C. R. Acad. Sci. (Paris), 307: 505-509. De Vitry, F., Hamon, M., Catelon, J., Dubois, M., Thibault, J. (1986) Proc. Nati. Acad. Sci. (USA), 83: 8629-8633. Sakumoto, T., Sakai, K., Jouvet, M., Kimura, H. and Maeda, T. (1984) Brain Res. Bull., 12: 721- 133) Horvath, J., Sétalo, G. and Flerk6, B. (1982) En- docr. Exp., 16: 229-238. Meijer, J. H. and Rietveld, W. J. (1989) Physiol. Rev., 69: 671-707. Ugrumov, M. V., Bladimirov, S. V., Kasmambeto- va, S., Popov, A. P., Novodjilova, A. P. and Tramu, G., Neurosicence, submitted Bosler, O. and Beaudet, A. (1985) J. Neurocytolo- gy, 14: 749-763. Shimatsu, A., Kato, Y., Matsushita, N., Katakami, H., Ohta, H., Yanaihara, N. and Imura H. (1983) Brain Res., 264: 148-151. Maegawa, M. (1987) Tokushima J. exp. Med., 34: 29-43. Ugrumov, M. V., Proshlyakova, E. V. and Sapro- nova, A. Y. (1989) Neuroscience, 32: 127-131. Liu, K. P., Tamir, H., Hsiung, S., Adlerberg, M. and Gershon M. D. (1987) Dev. Brain Res., 32: 31- 41. Siddiqui, A. and Gilmore, D. P. (1988) Endocr. (Copenh.) 118: 483-494. Reznikov, A. and Nosenko, N. (1990) In “Neuroen- docrinology: New Frontiers”. Ed by G. Wollmann and R. Tubingen. Wilson, W. E. and Agrawal, A. K. (1979) Dev. Neurosci., 2: 195-200. Crowley, W. R., O’Donohue, T. L. and Jacobowitz, D. M. (1978) Acta Endocr. (Copenh.), 89: 20-28. Demarest, K. T., McKay, D. W., Riegle, G. D. and Acta 64 65 66 67 68 69 70 71 2 73 74 75 76 V1 78 79 80 81 82 83 84 Moore, K. E. (1981) Neurooendocrinology, 32: 108-113. Simerly, R. B., Swanson, L. W., Handa, R. J. and Gorski, R. A. (1985) Neuroendocrinology, 40: 501- 510. Reisert, I., Engele, J. and Pilgrim, Ch. (1989) Cell Tiss. Res., 255: 411-417. Costa, E., Baulieu, E. E., Krueger, K. E., Purdy, R. H. and Puia, G. (1991) Abstracts Third IBRO Congress, Montreal, P. 8, W12. Puymirat, J., Barret, A., Picart, R., Vigny, A., Loudes, C., Faivre-Bauman, A. and Tixier-Vidal, A. (1983) Neuroscience, 10: 801-810. Lengvari, I., Branch, B. J. and Taylor, A. N. (1980) Dev. Neuroci., 3: 59-65. Charli, J. L., Faivre-Bauman, A., Loudes, C. and Kordon, C. (1991) In XXeme Colloque de la Societe de Neuroendocrinologie Experimentale, Geneve, 18-20 Septembre 1991, (in press). Lichtensteiger, W. and Schlumpf, M. (1986) Brain Res., 368: 205-210. Davis, M. D., Lichtensteiger, W., Schlumpf, M and Bruinink, A. (1984) Neuroendocrinology, 39: 1-12. Shyr, S. W., Crowley, W. R. and Grosvenor, C. E. (1986) Endocrinology, 119: 1217-1221. Watts, A. G. and Stanley, H. F. (1984) Neuroen- docrinology, 38: 461-466. Reznikov, A. G., Nosenko, N. D. and Demkiv, L. P. (1979) Endokrinologie, 73: 11-19. Borisova, N. A., Manukhin, B. N. and Selivanova, G. P. (1989) Bull. Exp. Biol. Med., 57: 135-136 (Russ). Siddique, A. and Gilmore, D. P. (1987) Neurosci- ence, Suppl. Vol. 22, p. 154. Simerly, R. B., Swanson, L. W. and Gorski, R. A. (1985) Brain Res., 340: 91-98. Borisova, N. A., Proshlyakova, E. V., Sapronova, A. Ya. and Ugrumov, M. V. Bull. Exp. Biol. Med., (in press) (Russ). Moguilevsky, J. A., Faigon, M. A., Scacchi, P. and Szwarcfarb, B. (1987) Neuroendocrinology, 45: 274-277. Jarzab, B., Sickmoller, P. M., Geerlings, H. and Dohler, K. D. (1987) Hormones and Behavior, 21: 478-492. Raum, W. J. and Swerdloff, R. S. (1981) Endocri- nology, 109: 273-278. Jarzab, B., Kaminski, M., Gubala, E., Achtelik, W., Wagiel, J. and Dohler, K. D. (1990) Brain Res., 516: 257-262. Dohler, K.-D., Coquelin, A., Davis, F., Hines, M., Shryne, J. E. and Gorski, R. A. (1984) Brain Res., 302: 291-295. Altman. J. and Bayer, S. A. (1986) Adv. Anat., Embryol. and Cell Biol., 100: 1-173. 85 86 87 88 89 90 91 36 Reznikov, A. G. and Nosenko, N. D. (1983) Exp. Clin. Endocr., 81: 91-93. Blaustein, J. D. and Turcotte, J. (1987) Brain Res., 436: 253-264. Raum, W. J., Marcano, M. and Swerdloff, R. S. (1984) Biol. Reproduction, 30: 388-396. Jarzab, B., Gubala, E., Achtelik, W., Lindner, G., Pogorzelska, E. and Dohler, K.-D. (1989) Exp. Clin. Endocrinol., 94: 61-71. Lacau de Mengido J., Becu-Villalobos D. and Libertun C. (1987) Dev. Brain Res., 35: 91-95. Ojeda, S. R. and McCann. S. M. (1974) Endocri- nology, 95: 1499-1505. Handa, R. J., Hines, M. Scoonmaker J. N., Shryne, J. E. and Gorski, R. A. (1986) Dev. Brain Res., 30: 92 93 94 95 M. V. UGRUMOV 278-282. Jarzab, B. and Dohler, K. D. (1984) In “Sex differences in the Brain”. Ed. by G. J. De Vries. J. P. C. De Bruin, H. B. M. Uylings and M. A. Corner, Elsevier, Amsterdam, Prog. Brain Res., Vol. 61, pp. 119-126. Becu-Billalobos, D. and Libertun, C. (1986) Dev. Brain Re., 25: 111-116. Hamon, M., Bourgoin, S., El Mestikawy, S. and Goetz, C. (1984) In “Handbook of Neurochemis- try”, Ed. by A. Lajtha, Plenum Publ., New York, Vol. 6, pp. 107-143. Whitaker-Azmitia, P. M., Lauder, J. M., Shemmer A. and Azmitia, E. C. (1987) Dev. Brain Res., 33: 285-289. © 1992 Zoological Society of Japan ZOOLOGICAL SCIENCE 9: 37-45 (1992) Changes in Hemolymph Vitellogenin and Ecdysteroid Levels during the Reproductive and Non-Reproductive Molt Cycles in the Freshwater Prawn Macrobrachium nipponense Takuyt OKUMURA, CHANG-HEE Han!, Yuzuru SUZUKI, Katsumi ArpA~ and Isao HANyu Laboratory of Fish Physiology, Department of Fisheries, Faculty of Agriculture, The University of Tokyo, Bunkyo, Tokyo 113, Japan ABSTRACT—Hemolymph vitellogenin and ecdysteroid levels were traced during the reproductive and non-reproductive molt cycles in adult females of the freshwater prawn Macrobrachium nipponense, in which the reproductive development coincides with the molt cycle. During the reproductive molt cycle, vitellogenin levels began to increase at stage B. High levels were maintained during stages C,-D3 in concomitance with ovarian growth, and these levels decreased following the prespawning ecdysis. Vitellogenin was at undetectable levels during the non-reproductive molt cycle, and correspondingly, Ovarian maturation did not advance. The predominant ecdysteroids in the hemolymph of M. nipponense were determined by HPLC-RIA analysis as 20-hydroxyecdysone (20E) and immunoreactive high polarity products (HPP). 20E and HPP levels peaked at stage D3 during both molt cycles. 20E levels declined rapidly to low levels after ecdysis in the reproductive molt cycle. On the other hand, 20E levels decreased slowly during stages A-C, in the non-reproductive molt cycle. HPP levels immediately declined prior to spawning at stage Ag and showed a small peak after spawning at stage A, in the reproductive molt cycle, whereas in the non-reproductive molt cycle levels were still high at stage A and declined slowly during stages B-C,. These differences in ecdysteroid levels suggest a possible relationship between ecdysteroid levels and vitellogenesis. is completed during the subsequent molt cycle. In most crabs, ovarian development, spawning, and brooding occur within a single intermolt period. Endocrinological mechanisms possibly integrat- INTRODUCTION The integration of reproduction and molting appears to be a physiological necessity in female crustaceans. Most species continue to molt and grow even after reaching sexual maturity. The timely occurrence of ecdysis ensures that the egg mass brooded abdominally is not lost with the shedding of the exoskeleton. Various manners of coordination of the molting and reproductive cy- cles exist (see [1, 2]). In a number of prawn species and in several isopods, ovarian development occurs during the intermolt period, and spawning takes place after ecdysis. Brooding of the egg mass Accepted August 13, 1991 Received June 10, 1991 " Present address: Department of Biology, College of Natural Science, Dongeui University, Busan, Korea. * To whom correspondence should be addressed. ing the processes of molting and vitellogenesis are only partially understood. A correlation between the processes of molting and reproductive activity persists after ablation of the eyestalk in the shrimps Lysmata seticaudata [3] and Palaemon serratus [4]. The eyestalk includes X-organ sinus gland complex, a secretion site of neurohormones such as vitellogenesis-inhibiting hormone and molt-inhibiting hormone. These results suggest that while the neurohormones of the eyestalk may play indirect roles, they are not indispensable to the integration of molting and reproduction. It is more likely that hormones of other sources play key roles in this integration. Removal of the Y-organ, a site of ecdysteroid production via 38 T. Okumura, C.-H. HAN et al. cholesterol modification, induces the arrest of vitellogenesis in amphipods and isopods [5-7]. These results suggest the possibility that ecdyster- oids not only control molting but also regulate vitellogenesis. In order to gain more understand- ing of the functions of ecdysteroids in vitellogene- sis, the correlation between hemolymph ecdyster- oid and vitellogenesis should be further investi- gated. However, this has been examined in only a limited number of species (see [2]). To our knowl- edge, patterns of hemolymph ecdysteroid levels in adult females have not been priorly compared between the vitellogenic and non-vitellogenic molt cycles in any species. In order to examine the relationship between vitellogenesis and molting, the freshwater prawn Macrobrachium nipponense was employed in this study. This is considered a suitable species, as the reproductive cycle of female coincides with the molt cycle and the reproductive status can be manipulated [8]. Female prawns repeatedly undergo the reproductive molt cycle during the spawning season from June to August in Japan. During the reproductive molt cycle, ovaries de- velop cyclically in concurrence with the molt cycle, and mating and spawning occur following each ecdysis. However during the non-spawning sea- son, females molt successively without the occurr- ence of ovarian development (non-reproductive or common molt cycle). The environmental factors influencing the annual reproductive cycle have also been investigated [8]; the initiation of the spawn- ing season is induced by increased temperature regardless of photoperiod, and the termination of the spawning season is caused by shortened day- length. Therefore, the reproductive molt cycle and the non-reproductive molt cycle are easily repro- ducible in this species under artificial conditions. Prawns undergo continuous cycles of reproductive molt at 13.5-day intervals when reared at 28°C under 15L9D, but successively repeat the non- reproductive molt cycle under 28°C and 12L12D at 14.8-day intervals [8]. In this study, we examined changes in hemolymph levels of vitellogenin (Vg), the precur- sor of yolk protein, during the molt cycle of M. nipponense in order to gain more insight into the coincidence of vitellogenesis and molting. Further- more, we follwed the fluctuation of hemolymph ecdysteroid during the molt cycles. Differences in ecdysteroid levels between the reproductive molt cycle (occurrence of vitellogenesis) and the non- reproductive molt cycle (absence of vitellogenesis) and relationships between ecdysteroid and Vg levels are discussed. MATERIALS AND METHODS Animals Female freshwater prawns M. nipponense origi- nating from the Kasumigaura Lake in Ibaraki Prefecture were purchased from local fishermen, and kept in outdoor tanks under ambient water temperature and photoperiod until use in experi- ments conducted the following year. For experiments, female prawns (body weight ranging from 1.4—4.6 g) were transferred to plastic aquaria (604040 cm), and reared under two separate conditions: 15L9D at 28°C from August to October in 1987 and 1988, and 12L12D at 28°C from October to November in 1988. Prawns were fed commercial pellets (Taiyo Fishery Co. Ltd.) ~ and Tubifex worms daily. In order to avoid any effects due to male prawns, only female prawns were placed in aquaria. Therefore, prawns in the reproductive molt cycle were not able to mate and brood, but could spawn continuously. After acclimation periods of more than three weeks, prawns were sacrificed at differing molt stages. Hemolymph samples were taken from the posterior aorta with a capillary tube after cutting the sixth abdominal segment. The hemolymph samples were stored at —70°C until analyses of hemolymph Vg and ecdysteroid levels. Gonado- somatic indices (GSI) were calculated as gonad weight (g) < 100/body weight (g). Molt stage Molt stages were determined according to the methods of Han [8] by observing setal develop- ment in the pleopods under a differential phase contrast microscope. As stage E is of a very short period, prawns were not sacrificed at this stage. Molt stage criteria is shown in Table 1. Vitellogenin and Ecdysteroid in Prawns 3) TABLE 1. Criteria for the determination of the molt stages in M. nipponense S Approximate R c tape duration* aaa Postmolt 2 days A Exoskeleton is very soft. Setae are filled with translucent cellular matrix. During the reproductive molt cycle, stage A is subdivided into stage Ao (before oviposition) and A, (after oviposition). B Setal matrix is constricted. Internal cones are forming inside setae. Intermolt 5 days Co Exoskeleton is fully formed. Completion of formation of internal cones is observed. C, New setae begin to appear in internal cones, but the epidermis is not yet separated from the base of setae. Premolt 7 days Do The epidermis begins to retract from the old cuticle. D, The epidermis separates from the old cuticle completely. Space between the epidermis and old cuticle widens. D, New setae form in the cellular matrix under the retracted epidermis. D; New barbules form on the new setae. Ecdysis 10 sec J8) Old cuticle is shed. *: rearing condition 28°C Enzyme-immunoassay (EIA) for Vg Vitellin was purified from mature ovaries of M. nipponense using gel-filtration and DEAE-ionex- change chromatography [8], and was used as the standard in this EIA system. Antisera were raised against the purified vitellin in rabbits [8]. The only hemolymph protein to which the antisera exhibited cross-reactivity was Vg, validating its specificity for assay use [8]. The purified vitellin was dissolved in CB (0.1 M Na,CO; buffer, pH 9.6) containing 0.01% male hemolymph and a series of 2-fold dilutions were made, ranging from 8.2 ng/ml (0.82 ng/well) to 1,050 ng/ml (105 ng/well). The hemolymph sam- ples were diluted 10,000-fold in CB, and subse- quently diluted with CB containing 0.01% male hemolymph to the concentration adequate for assay range, and subjected to EIA. The antiserum was diluted 3,000-fold with PBS-Tween (0.073 M Na,HPO, and 0.036 M KH>PO, buffer, pH 7.2, containing 0.05% polyoxyethylenesorbitan-mono- laurate, 0.04 M NaCl and 0.02% NaNs3). EIA procedures were as follows; one hundred yl of either vitellin in standard solutions or Vg in sample solutions was adsorbed to wells of 96-well EIA plates (Nunc-Immuno Plate MaxiSorp F96 Certificate, A/S Nunc) by incubation at 4°C over- night. Subsequently, solutions in wells were dis- carded, and wells were washed with PBS-Tween three times. In order to block adsorption of antibodies for the purpose of the following steps, wells were coated with 1% bovine serum albumin solution (250 1) in PBS (0.073 M NazHPO, and 0.036 M KH>PO, buffer, pH 7.2, containing 0.04 M NaCl and 0.02% NaN;), by incubation at 27°C for 2 hr. After washing, diluted antiserum (100 pl) was added to each well followed by incubation at 27°C for 2hr. After washing, 100 1 of goat anti-rabbit IgG (whole molecule) alkaline phos- phatase conjugate (Sigma Chemical Co.) in PBS- Tween (1 :300) was added to wells with incubation at 27°C for 2hr. After washing, 100 ul of 0.07% p-nitrophenylphosphate, disodium salts in 9.7% diethanolamine buffer, pH 9.8, containing 0.047% MgCl, and 0.02% sodium azide was added to each well, and incubation at room temperature was carried out for 30 min. Continuous absorbance of substrate solutions in each well was measured at 405nm by a Tosoh Co. MPR-A4 microplate- reader. The concentration of samples was calcu- lated from the standard curve, and expressed in 40 T. Oxumura, C.-H. HAN et al. milligram eugivalents of vitellin per milliliter of hemolymph. Validation of EIA Parallelism between the vitellin standard curve and the absorbance curve for serially diluted hemolymph of vitellogenic female prawns was examined. To determine the recovery rate, differ- ing doses of vitellin were added to hemolymph of non-vitellogenic female prawns, and the amounts of vitellin recovered were measured by EJA. The relationship between the amounts of vitellin added and recovered was determined by linear regression analysis. The intra-assay variabilities were also examined. Radioimmunoassay (RIA) for ecdysteroids Hemolymph ecdysteroids were assayed by re- verse-phase high performance liquid chromatogra- phy (HPLC) and RIA, as described previously [9]. Hemolymph samples at each molt stage (2-7 prawns) were pooled for analysis, as hemolymph volume of this species was too small to determine individual levels. Ecdysteroids were extracted from sample hemolymph with methanol, and partitioned be- tween 70% methanol and n-hexane (1:1). The 70% methanol phases were subjected to HPLC separation using a reverse phase column (ODS- 80TM, 4.6250 mm, Tosoh Co.). The starting solvent, methanol-water (1:1) was run for 30 min, after which the solvent system was changed to a linear gradient reaching methanol-water (3:2) in 15 min. The flow rate was 0.8 ml/min. Fractions were collected at a rate of one fraction per minute. Ecdysteroid in each fraction was measured by a double antibody RIA method. The antibody used in this study (provided by Dr. M. Nagata, The University of Tokyo) showed similar cross- reactivities toward 20-hydroxyecdysone (20E) and ecdysone. The standard curve was obtained by serial dilutions of 20E (Sigma Chemical Co.); therefore, results are expressed in nanogram equivalents of 20E per milliliter of hemolymph. Statistics Duncan’s multiple range test was used for statis- tical analysis after log-transformation of data. RESULTS Validation of EIA The EIA system used in this study was validated according to standard criteria. The absorbance curve of serially diluted hemolymph was parallel to the standard curve (Fig. 1), indicating that the adsorption of Vg in hemolymph samples to wells showed the same dose-dependence as did the standard. The amounts of standard vitellin added and recovered were significantly correlated (r= 0.984, p<0.01). The mean recovery rate was 87.4 +3.6% (mean+SE). The intra-assay coefficients of variation were under 10% in concentrations ranging from 3.8 ng/well to 90.3 ng/well. 4 Dilution of female hemolymph (10 ) e—e Gb 82 AG 8 4 2 0.5 @ 7) = o = re) a 2 = ¢ 0.1 0.05 8.2 16.4 32.8 65.6 131 263 525 1050 Vitellin (ng/ml) o—o Fic. 1. EIA system for vitellogenin. Parallelism be- tween the standard curve for purified vitellin and the absorbance curve for serial dilutions of female hemolymph. GSIs and Vg levels during the molt cycles The variations of GSIs and hemolymph Vg levels were followed during the two types of molt cycles (Fig. 2). During the reproductive molt cycle, GSIs were low during stages A;-Co after spawning, and started to increase at stage C; (p< 0.05). Values peaked at stage D3 and then de- Vitellogenin and Ecdysteroid in Prawns 4] =» Reproductive Molt Cycle o Non-Reproductive Molt Cycle e Reproductive Molt Cycle © Non-Reproductive Molt cycle 10 ue 5) Cc = ga =] : : QO. =v z : . e Sy o Y) = al O = (= oT) Vv ss 5) ” 2 S o = > O-----So7------ O------ Oils O----... (“esnsa Qauisoae Foote O------. fe) 9) O-------------- } ----- O----- O----- t}----- f}----- i----- i}----- Oo @) Reproductive MolteGycie: (2). (3). (4)_..(5) (5) (7). (6), (6). @, (2) Non-Reproductive Molt Cycle : (6) (Seer) (6) hs (S) ee (A) eG) t(5)) tS) Molt stage Fic. 2. GSI and hemolymph vitellogenin levels during the reproductive and non-reproductive molt cycles in Macrobrachium nipponense. Data for stage A during the reproductive molt cycle (divided into stages Ag and A,), and during the non-reproductive molt cycle (undivided). The data for stages A and Ay are shown twice. Values represent the mean+SE. Number of animals assayed is given in parentheses. Vitellogenin levels were calculated as vitellin equivalents. clined at spawning during stages Ag-A, (p<0.01). Hemolymph Vg levels were low during stages Ao-A,, increased at stage B (p<0.01), remained at high levels during stages C,-D3, and then declined (p<0.01) sharply after the prespawning ecdysis. During the non-reproductive molt cycle, GSIs remained constantly low (0.6-0.7%). Hemolymph Vg levels were nondetectable (less than 0.01 mg/ ml) throughout the molt cycle. Identification of ecdysteroids in hemolymph Ecdysteroids in hemolymph were separated by HPLC before RIA. The HPLC profile of hemolymph ecdysteroid at stage D3 of the repro- 42 T. Oxumura, C.-H. HAN et al. 50 HPP % Of total ecdysteroid 0 10 20 Figs: D3 Stage Reproductive Molt Cycle 60 % MeOH 50 30 40 Retention time (min) Analysis of immunoreactive ecdysteroids in the hemolymph of female prawns at stage D3 during the reproductive molt cycle by HPLC-RIA system. Hemolymph ecdysteroids were separated into fractions by reverse-phase HPLC on a methanol-water solvent system using a linear gradient, and were detected by RIA. Results are expressed as the percent of the total ecdysteroid present in the hemolymph. Arrows indicate elution . times of references; 20-hydroxyecdysone (20E), makisterone A (M), ecdysone (E), 20-hydroxyecdysone-22- acetate (A), 2-deoxy-20-hydroxyecdysone (DO20E), ponasterone A (P), 2-deoxyecdysone (DOE), respectively. HPP, high polarity products; LPP, low polarity products. ductive molt cycle is shown representatively in Fig. 3. Two peaks were identified as 20E and ecdysone, respectively, by coelution with authentic stan- dards. Several other peaks of RIA activity were also detected. However, no peaks were observed to elute at the retention times of reference com- pounds makisterone A, 20-hydroxyecdysone-22- acetate, 2-deoxy-20-hydroxyecdysone, ponaster- one A or 2-deoxyecdysone. One major unknown peak was detected prior to 20E and termed high polarity products (HPP). Small peaks which eluted after ecdysone were collectively termed low polarity products (LPP). Total ecdysteroid was calculated as the sum of all immunoreactive sub- stances eluted by HPLC. Ecdysteroid levels during the molt cycles Changes in hemolymph ecdysteroid levels dur- ing the reproductive and non-reproductive molt cycles are shown in Fig. 4. Ecdysteroid profiles differ between the two molt cycles. During the reproductive molt cycle, total levels in hemolymph declined sharply after ecdysis, and showed a small peak at stage A, after spawning. On the contrary, during the non-reproductive molt cycle, levels decreased more gently during stages A-C;. During both molt cycles, levels were low. during stages C,-Do, showing an increase during stages D,-Dp, and peaking at stage D3. 20E and HPP were the predominant im- munoreactive ecdysteroids during both molt cylces. 20E levels rapidly declined to low levels after ecdysis in the reproductive molt cycle, where- as 20E levels slowly decreased during stages A-C, in the non-reproductive molt cycle. The levels peaked at stage D3 during both molt cycles. HPP levels rapidly declined just prior to spawn- ing at stage Ap and showed a small peak at stage A, in the reproductive molt cycle. On the other hand, in the non-reproducive molt cycle, the levels Vitellogenin and Ecdysteroid in Prawns 100 Reproductive Molt Cycle — ° TOTAL ° e g ) o HPP 8 Oo m 20E o v 2 (5) 50 oF: J a 3 une 2 WwW Cc : ES Spawning o a 7) yp ™ a fe} eo © 0 eee peat Re yes eer Ag Ai B Co Ci Do Di D2 D3 Ao Molt stage Non-Reproductive Molt Cycle ” a > ° TOTAL pees Lu o HPP Vv 400 = 20E i i 4 LPP ° Ecdysteroid (ng/ml) moO as 0 o4—pa— a so4— rE ina A B Co Ci Do Dy; Molt stage D2 D3 Fic. 4. Changes in total ecdysteroid levels, and in the levels of four HPLC-separated ecdysteroids from the pooled hemolymph during the reproductive and non-reproductive molt cycles. ecdysteroid levels are shown twice. The data for stages A and Ag Ecdysteroid levels were calculated as 20-hydroxyecdysone equivalents. 20E, 20-hydroxyecdysone; E, ecdysone; HPP, high polarity products; LPP, low polarity products. were still high in stage A, and declined slowly during stages B-C,;. HPP levels reached maxima at stage D3 in both molt cycles. Ecdysone levels were low, but also peaked at stage D3 in both molt cycles. LPP was remained at trace levels throughout both molt cycles. DISCUSSION Increases in Vg levels and the advancement of ovarian development were correlated with the molt stages during the reproductive molt cycle in this species; on the other hand, vitellogenesis and ovarian development did not occur during the non-reproductive molt cycle. The observance of high Vg levels was followed by rapid GSI increases during stages C,-D3. Histological observation re- vealed that yolk globules were accumulated in the oocytes, and that oocyte diameter increased [8]. Vg levels decreased concomitanly with the cessa- tion of GSI increase at stage Ap (before spawning), suggestive of the termnation of Vg synthesis and additionally, of Vg uptake. These results concern- ing Vg levels were as expected given that the cycle of spawning is synchronous with the molt cycle in this prawn. The profile for Vg levels during the reproductive molt cycle was similar to that of hemolymph Vg levels in the giant freshwater prawn Macrobrachium rosenbergii ({10], Okumura et al., unpublished data). This correlation in vitellogenesis and molt cycle suggests that they are linked endocrinologically. During stages Ap(A)-Co, levels of 20E and HPP in the reproductive molt cycle were lower than those in the non-reproductive molt cycle. At stage B of the reproductive molt cycle, hemolymph Vg levels began to increase. These observations may suggest that lower levels of 20E and HPP are involved in the initiation of vitellogenesis. Howev- er, in this study, it is unclear whether this differ- ence in 20E and HPP levels is indicative of their participation in vitellogenesis or rather, reflective of differences in ecdysteroid metabolism between the reproductive and non-reproductive molt cy- cles. Hemolymph 20E levels peaked at stage D3 and declined just after ecdysis during the reproductive molt cycle. This rapid change was concomitant 44 T. Oxumura, C.-H. HAn et al. with the decrease of Vg levels. A similar phe- nomenon was observed in M. rosenbergii; Meusy and Payen [2] examined total ecdysteroid levels, and Okumura et al. employed HPLC-RIA analysis to determine levels of 20E, as well as of other ecdysteroid species (unpublihsed data). Further- more, 20E levels and the synthesis levels of Vg in fat body were correlated similarly in the amphipod Orchestia gammarella [11]. Such correlations are considered to suggest that hemolymph 20E partici- pates in the modulation of the termination of vitellogenesis. Ecdysteroid levels were linked to Vg levels in this study, but the physiological significance of these phenomena is not yet clear. The function of 20E in controlling vitellogenesis has been investi- gated previously. However, results in these re- ports appear inconsistent. Y-organ removal de- creased Vg synthesis in O. gammarella [5]. The arrest of oocyte growth was. observed after Y- organ ablation in the terrestrial isopod Armadilli- dium vulgare [7]. In the isopod Porcellio dilatatus, a decrease in hemolymph V¢g levels resulted after Y-organ ectomy, and this effect could be compen- sated for by 20E injection [6]. On the other hand, administration of 20E just after ecdysis inhibited the onset of vitellogenesis in L. seticaudata |12| and in O. gammarella [13]. Variances among such reports might be attributed to differing molting stages of animals used and ranges of 20E dosage utilized. It has been reported that maturing ovaries con- tain ecdysteroid in the shore crab Carcinus maenas [14], in O. gammarella [15], in the spider crab Acanthonyx lunulatus [16] and in M. rosenbergii [17], suggestive of vitellogenic ovaries as a possible site of ecdysteroid production and/or metabolism. Ovarian development may have some influence upon ecdysteroid metabolism and hemolymph ecdysteroid levels. HPP is thought to consist of substances con- verted from ecdysone via metabolic pathways, and to have no biological activity. Several types of reactions leading to high polarity ecdysteroids have been considered in ecdysteroid metabolism: hydroxylation at the C-20 and/or C-26 position, ecdysonic acid formation following hydroxylation, and conjugation (see [18]). HPP was also detected in the hemolymph of P. serratus [19], and the mysid Siriella armata [20]. Enzymatic hydrolysis has revealed that the HPP in S. armata consists mainly of conjugated forms of ecdysteroid [20]. Thus, the difference in HPP levels seen between reproductive and non-reproductive individuals appears to reflect differences in ecdysteroid me- tabolism. Hemolymph ecdysteroid levels peaked at stage D; during both molt cycles in M. nipponense. The timing of this peak was within the range of results in other crustaceans which varied from D, to D3 (see [21]). In almost all crustaceans, ecdysteroid levels decline sharply just prior to ecdysis and are maintained at low basal levels during the postmolt period. However, in M. nipponense, ecdysteroid levels showed a small peak at A, stage in the reproductive molt cycle and were relatively high during the postmolt period of the non-reproducive molt cycle. A small peak in hemolymph ecdyster- oid, in addition to a large peak prior to ecdysis, was also observed at stage B in P. serratus [19], and at the end of the intermolt period in the fiddler crab Uca pugilator [22]. The physiological implica- | tions of such patterns are not understood. In M. nipponense, as vitellogenesis cycle was in synchronization with the molt cycle, we hypothe- size that 20E may be involved in controlling both vitellogenesis and molting. However, the exis- tence of several factors for the control of vitel- logenesis are additionally postulated (see [2, 23]). It is possible that such factors also take part in the integration of vitellogenesis and molting. Further investigations are required in order to clarify the function of ecdysteroids in vitellogenesis in M. nipponense. ACKNOWLEDGMENTS We express our appreciation to Dr. Masao Nagata, The Univeristy of Tokyo for kindly providing us with antiserum. We thank Marcy N. Wilder, The University of Tokyo for reading this manuscript. This study was supported in part by Grant-in-Aid from the Ministry of Education, Science and Culture of Japan. REFERENCES 1 Adiyodi, R. G. (1985) Reproduction and its con- 10 At 12 Vitellogenin and Ecdysteroid in Prawns trol. In “The Biology of Crustacea”. Ed. by D. E. Bliss and L. H. Mantel, Academic Press, Orland FL, VOI. 9, pp. 147-215. Meusy, J.-J. and Payen, G. G. (1988) Female reproduction in Malacostracan Crustacea. Zool. Silo, 52 ZARA OS) Charniaux-Cotton, H. and Touir, A. (1973) Con- trdle de la prévitellogenése et de la vitellogenése chez la Crevette hermaprodite Lysmata seticaudata Risso. C. R. Acad. Sci. Paris, Ser D, 276: 2717- 2720. Faure, Y., Bellon-Humbert, C. and Charniaux- Cotton, H. (1981) Folliculogenése et vitellogenése secondaires chez la Crevette Palaemon serratus (Pennant); contrdle par les pédoncules ocularies et Yorgane X de la medulla externa (MEX). C. R. Acad. Sci. Paris, Ser III, 293: 461-466. Meusy, J.-J., Blanchet, M.-F. and Junera, H. (1977) Mue et vitellogenése chez le Crustacé Amphipode Orchestia gammarella Pallas. Il. Etude de la synth- ése de la vitellogénine (“fraction protéique femelle” de ’hemolymphe) aprés destruction des organes Y. Gen. Comp. Endocrinol., 33: 35-40. Souty, C., Besse, G. and Picaud, J.-L. (1982) Sti- mulation par |’ecdysone du taux hémolymphatique de la vitellogénine chez le Crustacé Isopode terres- tre Porcellio dilatatus Brandt. C. R. Acad. Sci. Paris, Ser III, 294: 1057-1059. Suzuki, S. (1986) Effect of Y-organ ablation on oocyte growth in the terrestrial isopod, Armadilli- dium vulgare. Biol. Bull., 170: 350-355. Han, C.-H. (1988) “Physiological studies on the reproductive cycle of a freshwater prawn Macro- brachium nipponense (de Haan).” Ph. D. thesis, The University of Tokyo. Okumura, T., Nakamura, K., Aida, K. and Hanyu, I. (1989) Hemolymph ecdysteroid levels during the molt cycle in the Kuruma prawn Penaeus japonicus. Nippon Suisan Gakkaishi, 55: 2091-2098. Derelle, E., Grosclaude, J., Meusy, J.-J. Junéra H. and Martin, M. (1986) ELISA titration of vitel- logenin and vitellin in the freshwater prawn, Mac- robrachium rosenbergii, with monoclonal antibody. Comp. Biochem. Physiol, 85B: 1-4. Blanchet-Tournier, M.-F. (1982) Quelques aspects des interactions hormonales entre la mue et la vitellogenése chez le Crustacé Amphipode Orchestia gammarellus (Pallas). Reprod. Nutr. Dévelop. 22: 325-344. Touir, A. and Charniaux-Cotton, H. (1974) In- fluence de l’introduction d’ecdystérone sur |’exuvia- tion et la démarrage de la vitellogenése chez la Crevette Lysmata seticaudata Risso. C. R. Acad. 13 14 IS) 16 7) 18 19 20 21 LD, yep) Sci. Paris, Ser D, 278: 119-122. Blanchet, M.-F., Junéra, H. and Meusy, J.-J. (1975) Mue et vitellogenése chez Orchestia gammarella (Crustacé Amphipode): étude de la synthése de la fraction protéique femelle aprés introduction d’ec- dystérone. Experimentia, 31: 865-867. Lachaise, F., Goudeau, M., Hetru, C., Kappler, C. and Hoffmann, J. A. (1981) Ecdysteroids and ovarian development in the shore crab, Carcinus maenas. Hoppe-Seyler’s Z. Physiol. Chem., 362: 521-529. Blanchet, M.-F., Porcheron, P. and Dray, F. (1979) Variations du taux des ecdystéroides au cours des cycles de mue et de vitellogenése chez le Crustacé Amphipode Orchestia gammarellus. Int. J. Inver- tebr. Reprod., 1: 133-139. Chaix, J.-C. and De Reggi, M. (1982) Ecdysteroid levels during ovarian development and embryogene- sis in the spider crab Acanthonyx lunulatus. Gen. Comp. Endocrinol., 47: 7-14. Wilder, M. N., Okumura, T. and Aida, K. (1991) Accumulation of ovarian ecdysteroids in synchro- nization with gonadal development in the giant freshwater prawn, Macrobrachium _ rosenbergii. Zool. Sci., 8: 919-927. Lafont, R. and Koolman, J. (1984) Ecdysone metab- olism. In “Biosynthesis, Metabolism and Mode of Action of Invertebrate Hormones”. Ed. by J. A. Hoffmann and M. Porchet, Springer-Verlag, Berlin, pp. 196-226. Baldaia, L., Porcheron, P., Coimbra. J. and Cas- sier, P. (1984) Ecdysteroids in the shrimp Palaemon serratus: Relation with molt cycle. Gen. Comp. Endocrinol., 55: 437-443. Cuzin-Roudy, J., Strambi, C., Strambi, A. and Delbecque J.-P. (1989) Hemolymph ecdysteroids and molt cycle in males and females of Siriella armata M.-Edw. (Crustacea: Mysidacea): possible control by the MI-ME X-organ of the eyestalk. Gen. Comp. Endocrinol., 74: 96-109. Skinner, D. M. (1985) Molting and regeneration. In “The Biology of Crustacea”. Ed. by D. E. Bliss and L. H. Mantel, Academic Press, Orland FL, Vol. 9, pp. 43-146. Hopkins, P. M. (1983) Patterns of serum ecdyster- oids during induced and uninduced proecdysis in the fiddler crab, Uca pugilator. Gen. Comp. Endocri- nol., 52: 350-356. Charniaux-Cotton, H., and Payen, G. (1988) Crustacean reproduction. In “Endocrinology of Selected Invertebrate Types”. Ed. by H. Laufer and R. G. H. Downer, Alan R. Liss, New York, pp. 279-303. aa ha “apa Ap Sh ase hes naps a deat es Fototald : be pfs ty yng a) bas fee iors a | Hive gibi iis a. | te ere yale SEP sia Y. Je 2 oe Mtoe Ao, hiner Di afl apres) | meee s aap i i iv ea ang an Li icoeteil pees Py tah. é (tds 2 b . i ons " f ‘ Ae he * re bo | a : Sea oe ai se “ EF (ate ee: ; 7 xs i sah} TA ; gees oe ibe e ie peyote) vee : ; ag eG RMT) ot: Bata ny ZOOLOGICAL SCIENCE 9: 47-52 (1992) © 1992 Zoological Society of Japan Effects of Ba?* on the Photoresponses of Isolated Single Rods from Frog Retina Katsu AZUMA and NAOHIKO IWASAKI Department of Biology, Osaka Medical College, Sawaragicho 2-41, Takatsuki, Osaka 569, Japan ABSTRACT—Effects of extracellular barium (Ba**) on the photoresponses of mechanically isolated single rods from the retina of the frog (Rana catesbeiana) were investigated by sucking the rod inner segments into tightly fitting pipettes. The addition of Ba** to a control solution prolonged the duration of the rod response to a flash of light (flash response), and increased its sensitivity to the flash (flash sensitivity). The response of the rod to a step of light (step response) in the control solution ascended to an initial peak, and then descended to a lower steady level within a few seconds (relaxation), showing light adaptation. Ba** abolished the relaxation of the step response. The flash sensitivity of the rod in the Ba**-solution drastically decreased with increasing intensity of adaptation light in the manner different from that in the control solution. These effects were specific to Ba**, Sr°* or Co** being ineffective. The effects were not explicable by the role of Ba** as a K*-blocker, because other K* -blockers, tetraethylammonium chloride (TEA), 4-aminopyridine (4-AP) and Cs* were ineffective. INTRODUCTION In the previous paper, we reported that lowered extacellular calcium concentration shortened the recovery phase of the photoresponse of isolated single rods from frog retinas [1]. Ca*t has an important role in light adaptation in rods and cones [2]. On the other hand, Brown and Flaming [3] showed by the intracellular recording in retinal rods of toad that Ba** induced several effects on the rod photoresponses. They described as fol- lows. Ba’t increased the rod sensitivity and markedly delayed the decay of the rod response. Ba’? also induced the oscillations of membrane potentials of the rods. Therefore, we intended to investigate whether or not Ba** would affect on the rod under no interaction with other rods or second order cells. In this study we examined the effects of Ba** on mechanically isolated single rods from the frog retina using the suction pipette method. We shall first demonstrate that Ba** prolongs the decay of the flash response and also increases the flash sensitivity. Next we shall show that Ba*t Accepted September 25, 1991 Received July 16, 1991 abolishes the descending phase of the step re- sponse and causes a significant effect on the prop- erties of the light adaptation in isolated single rods. MATERIALS AND METHODS The experiments described in this paper were very similar to those in the already published one [1]. Adult bullfrogs were dark-adapted overnight. The animals were double-pithed and then the eyeballs were enucleated. Retinas were isolated from the enucleated eyeballs. The isolated retina was placed in normal physiological saline in a Petri dish and shaken to detach the rods. The rod suspension in the saline (about 0.5 ml) was put into the measuring chamber using a disposable syringe. The chamber was set on the stage of the inverted microscope (Nikon-MAD). A single rod which had outer and inner segments was sucked at the inner segment following the methods described by Hodgkin et al. [4]. All these procedures were carried out under infrared light using an infrared image coverter (NEC-NVR2015). The physio- logical saline was used as a control superfusate which contained (in mM) NaCl, 85; KCl, 2.5; NaHCOs, 25; MgCh, 1.2; glucose, 25; HEPES, 3, and CaCl, 1, and was adjusted to pH7.6 by 48 K. AZUMA AND N. IWASAKI adding 1N-HCl. Experiments on the effects of Ba’* were carried out by adding various concen- trations of BaCl, to the control superfusate (hereafter, called Ba**-solution). The flow rate of superfusion was 2 ml min '. The temperature of the superfusate was maintained at 20+1°C. The apparatus for optical stimulation was essen- tially the same as that described by Baylor et al. [5]. The unattenuated light intensity for flash (20 ms) or step of 500-nm light was 710° photons yum *s_'. The stimulus intensity was expressed by the negative logarithm of the neutral density filter. RESULTS Effects of Ba** on the flash response The upper recordings in Figure 1 are the flash responses of a single rod superfused with the control solution. The flash intensities were indi- cated by the numerals under horizontal lines. After recording these responses, the superfusate was switched to 0.2 mM Ba’*-solution. As shown by the middle recordings in Figure 1, Ba** pro- longed the durations of the flash responses at both intensities of —4.0 and —3.0 in log unit. Ba** also markedly increased the amplitude of the flash response as seen at intensity of —4.0 log unit. It should be noticed that there is no oscillation of membrane current at any intensity of flash stimu- lus. After 30 min of superfusion with the Ba7+ -solution, the same rod was again superfused with the control solution. After 40 min of the superfu- sion (washout-period), each response almost rec- overed to its control waveform, as shown by the lower recordings in Figure 1. It was found that the prolongation of responses was induced by Ba**+ above 0.1 mM (data not shown). Figure 2 is the response-intensity plots obtained from the experiments similar to Figure 1. The peak amplitudes of flash responses at the control, 0.2mM Ba’? and the recovery were plotted against the longarithm of the relative intensitites of the flahses. The peak amplitudes of all responses were normalized against those of corresponding maximal responses. The smooth curves, 1, 2 and 3, were obtained from calculations based on the following quation [6]: control \ 0.2 mM Ba2t < BE GOVERY, Flash na | ee ere, -4.0 Fic. 1. Effects of Ba** on the flash responses. The upper recordings are control flash responses at the. intensities indicated by the numerals under horizon- tal lines. After recording these responses, the su- perfusate was changed from the control solution to 0.2mM_ Ba’*t-solution. The middle responses were recorded after 20 min of the superfusion with the Ba**-solution. The lower responses were re- corded after 40 min of recovery in control solution. R/Ro ee) Here R represents the peak amplitude of each response, with maximum (R,ax), and I represents flash intensity with half-saturating intensity (Io). The constants, Ip(n) were determined by using the log-linear, least-squared error method to fit each set of data points. When the value of n was 1.3, the values of Ip at the control, 0.2 mM Ba’* and the recovery were —3.7, —4.2 and —3.8, respec- tively. Thus 0.2mM Ba*t shifted Ig into the direction of lower intensity by 0.5 log unit, which indicated an increase in the rod sensitivity. Effects of Ba** on the step response Figure 3 shows the step responses at various Ba**-effect on isolated single rods 140 5 Control 2: 0.2 mM Bact W S recovery = Ss =>} O ne) Sj is Ons oO = See (@) = 0 -4 =. log relative intensity Fic. 2. Effects of Ba** on the response-intensity relationship. These results were obtained from the experiments similar to those used in Figure 1. The measurements were carried out consecutively with the same rod. Curve 1: control, Curve 2: 0.2 mM Ba**, Curve 3: recovery. The peak amplitudes of the flash response are plotted versus the logarithms of relative intensity, normalized against the peak amplitude of the corresponding maximal responses. -5.6 10 | pA -5.0 -4.6 -4.0 =3.0 60 Ss Fic. 3. Step responses recorded in the control solution. The intensity of step light was varied from —5.6 to —3.0 in log unit, as indicated to the upper of each response. The duration of each step light is indi- cated by a horizontal line as 60s. intensities in control solution. The step responses at the lower intensities, —5.6, —5.0 and —4.6 in log unit, rose to the initial peaks then rapidly recovered to the lower levels (relaxation) and tended to gradually recover to the dark levels. The relaxation was usually and markedly observed at the intensity of —5.0 log unit. At the stronger intensities, —4.0 and —3.0 in log unit, both re- sponses hardly indicated the relaxations and were steady at their peak levels. The upper recording in Figure 4 is the step response at the intensity of —5.0 log unit in the control solution. After the recording, the rod superfusate was changed to 0.5mM_ Ba*?- solution. The middle recording was obtained after 20 min of superfusion with the Ba’ *-solution. The recording showed that Ba** completely abolished the relaxation of the step response. The same rod was again superfused with the control solution after the recording. The step response almost recovered to its control waveform after 40 min of the superfusion, as shown by the lower recording in Figure 4. 50 K. AZUMA AND N. IWASAKI COMEGO JL -5.0 0.5 mM Ba’* 10 pA recovery 60 Ss Fic. 4. Effects of Ba*t on the step responses. The upper recording is the step response in the control solution. After recording this response, the super- fusate was changed from the control solution to 0.5 mM Ba’*-solution. The middle recording was obtained after 20 min of continuous superfusion with the Ba**t-solution. The lower recording was obtained after more than 40 min of recovery in the control solution. The intensity and duration of the step light are —5.0 log unit and 60s (indicated by a horizontal line), respectively. Light adaptation of the rod The flash responses of a single rod were re- corded at various intensities in the control solution and in the dark. Following these recordings, the responses to the flashed superimposed on the background light at the intensity of —5.5 log unit were recorded. Finally the flash responses were again recorded in the dark after turning-off the background light. The response-intensity plots obtained from those experiments are indicated in Figure 5A. The peak amplitudes of all flash re- sponses were plotted against the logarithms of the relative intensities of flashes. The peak amplitude of each flash response in the initial dark period. Smooth curves, 1, 2 and 3, were drawn after the manner of the case in Figure 2. Io(n) of curves, 1, 2 contro] normalized current =) i= (eB) = = 3 2) fo) = 0.5 ‘S = fe} Cc 0 log relative intensity Fic. 5. Response-intensity relationships obtained from - two rods. A, control solution: B, 0.5mM Ba??. Curves 1 were obtained from the flash responses at various intensities in the dark. Curves 2 were obtained from the responses to the flashés superim- posed on the background light intensity of —5.5 log unit. Curves 3 were obtained from the flash re- sponses recorded during 10 to 20 min after turning- off the background light. The peak amplitudes of all flash responses are plotted against the logarithms of the relative intensities of flashes. The peak ampli- tude of each flash response is normalized to that of a maximum flash response in the initial dark period. and 3, were —3.7 (1.1), —3.3 (0.9) and —3.7 (1.1), respectively. The background light induced a little decrease in maximum response amplitude and shifted Ip into the direction of higher intensity by 0.4 log unit. The response-intensity plots in Figure 5B were obtained from the experiments similar to those used in Figure 5A except the presence of Ba’*. Ip (n) of curves, 1, 2 and 3, were —4.1 (1.8), —3.6 (1.0) and —4.1 (1.6), respective- ly. The background light induced a significant decrease in maximum response amplitude and Ba’ *-effect on isolated single rods shifted Ip into the direction of higher intensity by 0.5 log unit. Figure 6 was obtained from the experiments similar to those described in Figure 5 (A and B). In Figure 6 the flash sensitivity (log unit) during the background light was plotted against the log relative background intensity in control solution or 0.5 mM Ba’t-solution. Flash sensitivity during the background light was defined as the logarithm of the ratio of the flash intensity needed to elicit 4 pA-photocurrent in the dark to that during the background light. We restricted the experiments presented in Figure 6 to those in which a nearly full recovery in the flash response was obtained after turning-off the background light. In control solu- tion (circles) flash sensitivity against background intensity showed the slope of approximate 1 rang- ing from —5.5 to —3.5 in log unit, where Weber’s law held. On the other hand, the slope in the case of Ba**-solution (triangles) is 2.8 ranging from — 5.5 to —4.5 in log unit, where Weber’s law did not hold. Thus Ba*t markedly modified the flash sensitivity of a single rod under background light. control flash sensitivity (log units) -/ -6 -5 -4 -3 log relative background intensity Fic. 6. Flash sensitivities (log unit) plotted against the logarithms of the intensities of background light in the control (circles) and 0.5mM Ba** (triangles) solutions. These results were obtained from the experiments similar to those described in Figure 5 (A and B). Results in the control and Ba** solu- tions were obtained from 4 cells and 3 cells, respec- tively. The data points and vertical bars indicate averaged values and standard deviations, respec- tively. DISCUSSION As already described, Ba** of 0.2 mM increased the rod sensitivity by 0.5 log unit (see Figure 2). This result was close to that shown by Brown and Flaming using the intracellular recording in retinal rods of toad [3]. They described that Ba’* of 0.6 mM increased the rod sensitivity by 0.4 log unit. Ba** also prolonged the decay of the single rod response but induced no oscillation in membrane current of the single rod. Thus this study com- pletely supported the description of Brown and Flaming [3]; i.e., Ba**-effects except the oscilla- tion were directly exerted on the rod. New findings in this study were as follows. Ba** abolished the relaxation of a step response (see Figure 4). The flash sensitivity of a single rod in the Ba**-solution drastically decreased with in- creasing intensity of background light, where We- ber’s law did not hold (see Figure 6). Thus Ba** also caused significant effects on the proporties of light adaptation in a single rod. We tested whether or not other divalent cations, Ca**, Mg’*, Sr?* and Co*+t, would induce similar effects on the rods. These ions were less effective except lower Ca** which enhanced the relaxations of step re- sponses (data not shown). It has been well known that Ba’? blocks K* -conductances of rods [3, 7] and Miiller cells [8]. Fain & Quandt [7] found that extracellular TEA (6-12 mM) and 4-AP (10 mM) blocked the K* -outward current of toad rods. Bader et al. [9] reported that extracellular TEA, Cs* and Co? blocked the current carried predominantly by K* in solitary rod inner segments from the salamander retina. However, we found that 10 mM 4-AP, 12 mM TEA and 5 mM Cs* did not indicate effects similar to those of Ba’? , whether the rod outer segments or inner segments were superfused with the physiological solution containing those K~* -blockers (data not shown). Probably the effects of Ba’? is not due to the inactivation of the KT -conductance. However, we can not thoroughly exclude the possibility, because we have not yet tested the intracellular actions of TEA, 4-AP and Csi: According to the prevailing model for the light adaptation of rod photoreceptors, a light-induced 52 K. AZUMA AND N. IWASAKI decrease in intracellular Ca** acts on guanylate cyclase and causes an increase in cyclic GMP level, which opposes the light-activated decline in the cyclic GMP level (negative feedback) [10]. The relaxations of step responses may be related to the increase in the cyclic GMP level caused by the light-induced decrease in intracellular Ca** [2]. On the other hand, it has been found that Ba** evokes an increase in cyclic GMP in the cilia of Paramecium and the Ba**-effect is specific as Mg’*, Ca** or Sr** does not elicit an increase in cyclic GMP [11]. Although such effects have not been known in the rods, the findings in Para- mecium are suggestive taking it into consideration that the outer segment of rod is phylogenetically a rudimentary ciltum. It is not unreasonable that Ba’* acts on the enzymatic process related to cyclic GMP and inhibits the negative feedback. Although the mechanism of the Ba** effect is not yet elucidated, the Ba** manipulation may be useful to study the behaviour of rods in the ab- sence of the negative feedback underlying light adaptation. REFERENCES 1 Azuma, K. (1989) Effect of low extracellular cal- cium concentration on photosensitivity of isolated rods from frog retina. J. exp. Biol., 145: 255-262. 2 Nakatani, K. and Yau, K.-W. (1988) Calcium and 10 Lh light adaptation in retinal rods and cones. Nature 334: 69-71. Brown, K. T. and Flaming, D. G. (1978) Opposing effects of calcium and barium in vertebrate rod photoreceptors. Proc. Natl. Acad. Sci. USA 75: 1587-1590. Hodgkin, A. L., McNaughton, P. A., Nunn, B. J. and Yau, K.-W. (1984) Effect of ions on retinal rods from Bufo marinus. J. Physiol., Lond., 350: 649- 680. Baylor, D. A., Lamb, T. D. and Yau, K.-W. (1979) The membrane current of single rod outer segments. J. Physiol., Lond., 288: 589-611. Naka, K. I. and Rushton, W. A. H. (1966) S- potentials from colour units in the retina of fish (Cyprinidae). J. Physiol., Lond., 185: 536-555. Fain, G. L. and Quandt, F. N. (1980) The effects of tetraethylammonium and cobalt ions on responses to extrinsic current in toad rods. J. Physiol., Lond., 303: 515-533. Bolnick, D. A., Walter, A. E and Sillman A. J. (1979) Barium suppresses slow PIII in perfused bullfrog retina. Vision Res., 19: 1117-1119. Bader, C. R., Bertrand, D. and Schwartz, E. A. (1982) Voltage-activated and calcium-activated cur- rents studied in solitary rod inner segments from the salamander retina. J. Physiol., Lond., 331: 253-284. Yau, K.-W. and Nakatani, K. (1985) Light-induced, reduction of cytoplasmic free calcium in retinal rod outer segment. Nature, 313: 579-582. | Schultz, J. E., Thomas, P. and Klumpp, S. (1986) Voltage-gated Ca** entry into Paramecium linked to intraciliary increase in cyclic GMP. Nature, 322: 271-273. ZOOLOGICAL SCIENCE 9: 53-64 (1992) Proctolin-Like Immunoreactivity in the Dorsal Unpaired Median Neurons Innervating the Accessory Gland of the Male Cricket, Gryllus bimaculatus Koust YAsuyaMa!, TETSUYA KimurA* and TSUNEO YAMAGUCHI’’* "Department of Biology, Kawasaki Medical School, Kurashiki 701-01, and *Department of Biology, Faculty of Science, Okayama University, Okayama 700, Japan ABSTRACT—The dorsal unpaired median neurons (DUMR’7s) bilaterally extend bifurcating axons to the accessory gland via the left and right branches (Br3s) arising from the seventh nerve roots of the terminal abdominal ganglion in the male cricket, Gryllus bimaculatus. Proctolin-like immunoreactivity (PLI) was demonstrated in these DUMR7 neurons by the postembedding immunogold method combined with retrograde horseradish peroxidase (HRP) labeling and ligation of one of the Br3s. Ligation of one Br3 and HRP-labeling of the opposite Br3 produced labeling of both the somata of DUMR7 neurons and the ligated axons containing numerous granular vesicles (100 nm in diameter). PLI was found within the granular vesicles (SO—200 nm in diameter) in HRP-labeled DUMR7 somata as well as within similar granular vesicles accumulated in the HRP-labeled axons. Axons containing granular vesicles (100-115 nm in diameter) with PLI were noted to associate extensively with the visceral muscles winding around the accessory gland. It is therefore evident that the HRP-labeled axons with accumulation of granular vesicles originated from the DUMR7 neurons on the basis of the characteristic bilateral morphology of these neurons. These findings indicate the possibility that DUMR7 neurons produce a proctolin-like substance and transport it through their axons to the accessory gland musculature, where this substance may serve as a neuromuscular transmitter or © 1992 Zoological Society of Japan cotransmitter. INTRODUCTION Our interest lies in investigating the neural con- trol of visceral muscles associated with the male cricket accessory gland, which produces various spermatophore-forming materials and then re- leases them at a certain period in the mating cycle according to a temporal motor program [1]. Our previous morphological and electrophysiological studies have suggested a proctolinergic innervation for the accessory gland musculature [2, 3]. Proctolin was originally isolated from the cock- roach, Periplaneta americana [4, 5], and is a pen- tapeptide with the structure Arg-Tyr-Leu-Pro-Thr- OH [6]. This peptide was initially proposed to be a neurotransmitter associated with the hindgut visceral muscle [4, 5]. Subsequently, proctolin has been detected in many arthropods by high- Accepted October 5, 1991 Received September 10, 1991 * To whom reprint requests should be addressed. performance liquid chromatography (HPLC), im- munocytochemistry, and radioimmunoassay (re- viewed in [7]), and multiple physiological functions have been proposed for this peptide. These in- clude action as a peripheral neurotransmitter or cotransmitter influencing on visceral and skeletal muscles [e.g., 8-16], action as a central neurot- ransmitter or neuromodulator [e.g., 17-23], and a role as a neurohormone [e.g., 24, 25]. With respect to the visceral musculature of the oviduct, there are a number of pieces of evidence for the myotropic action of proctolin in cock- roaches [14, 15], locusts [13], diptera [9, 26], and hemiptera [16]. By the use of HPLC and bioassay, the presence of proctolin in this tissue and its release by high potassium saline and neural sti- mulation have been demonstrated [16, 27-30]. Furthermore, immunocytochemical studies using antisera to proctolin have revealed proctolin-like immunoreactive nerve fibers in the locust oviduct and neurons with proctolin-like immunoreactivity 54 K. YASUYAMA, T. KIMURA AND T. YAMAGUCHI (PLI), which may be related to these fibers, in the penultimate (seventh) abdominal ganglion [28, 31]. The musculature of the cricket accessory gland is innervated by bilaterally paired neurons (LC neurons) and dorsal unpaired median neurons (DUMR7 neurons), with the axons running through the branches (Br3s) of the bilateral seventh nerve roots arising from the terminal abdominal ganglion (TAG) [2]. Selective antidro- mic stimulation of these DUMR7 and LC neurons evokes contraction of the accessory gland which Squeezes out spermatophore-forming material from the glandular tubules, while orthodromic stimulation of both types of neurons evokes not only contraction but also a transient reduction of the basal tonus after relaxation [3]. Synthetic proctolin produces sustained contraction of the accessory gland at a concentration as low as 107” M [3]. Using HPLC and bioassay, it has been demonstrated that proctolin is present in the inner- vated accessory gland, but not in the gland without innervation, and that this peptide is released from the accessory gland in response to the application of high potassium saline [3]. These findings sug- gest that the DUMR7 neurons are proctolinergic excitatory motoneurons and the LC neurons are inhibitory motoneurons [3]. In the present paper, we will provide additional evidence for the proctolinergic innervation of the accessory gland musculature in the male cricket, as a result of studies involving electron microscopic immunocytochemistry combined with retrograde HRP labeling. These studies have demonstrated that PLI is present in the DUMR7 neurons inner- vating the the accessory gland musculature. MATERIALS AND METHODS Animals The TAGs and accessory glands of adult male crickets (Gryllus bimaculatus) were used in this study. The animals were obtained from a labora- tory culture reared under a 12 hr light-12 hr dark Giles a ZC Antiserum preparation and testing We obtained rabbit anti-proctolin antiserum from Seikagaku Kogyo Co. (Osaka, Japan). The rabbits were immunized with glutaraldehyde- linked conjugates of synthetic proctolin and bovine serum albumin (BSA) or bovine thyroglobulin [11, 32, 33]. Proctolin antiserum was purified by affini- ty chromatography according to the method of Eckert and Ude [33]. The purified anti-proctolin- BSA serum was mainly used in light microscopic immunochemistry, and the purified anti-proctolin- thyroglobulin serum was used for the electron microscopic study. Controls were performed as follows: purified proctolin antiserum was preincu- bated with synthetic proctolin at a concentration of 0.5 mg/ml diluted antiserum for 16 hr, and then used to process sections for light and electron microscopy. Tests were also performed in which the primary antiserum was left out of the staining procedure. All immunoreactivity was eliminated by such treatment. Light microscopic immunocytochemistry Each of the left and right nerve branches (Br3s) arising from the seventh nerve roots of the TAG was tied by a single knot using a hair. After 6 hr, the ganglion was isolated together with the ligated Br3s. Both the ligation and the dissection proce- dures were undertaken in buffered cricket saline [3]. The tissue was then fixed for 15 hr at 4°C ina glutaraldehyde-picric acid mixture [34] containing 2.5% glutaraldehyde and 15% saturated picric acid in 0.1 M phosphate buffer (pH 7.3). After washing for 4hr in the same buffer, the tissue was dehy- drated, embedded in paraffin, cut into 10 um sections, and mounted on gelatin-coated slides. Immunoreactivity was visualized using the Vectas- tain avidin-biotin-peroxidase complex (ABC) method (Vector Lab.). After deparaffinization and rehydration, the sections were rinsed in 0.01 M phosphate-buffered saline (PBS, pH 7.4) for 10 min, incubated in methanol (100%) with 0.3% H,O, for 30 min to block endogenous peroxidase activity, and rinsed again in PBS. The sections were then incubated for 30 min in normal goat serum diluted to 1:50 with PBS, and were subse- quently incubated with the primary antiserum for Proctolin-Like Immunoreactivity in DUM Neurons 55 48 hr at 4°C. Proctolin antiserum was used after dilution from 1:200-1:800 with PBS containing 0.5% BSA and 0.01% Triton X-100. The sections were washed three times in PBS for 5 min each, incubated at room temperature for 40 min in biotinylated goat anti-rabbit IgG diluted with PBS (1: 200), and washed further three times in PBS for 5 min each. Thereafter, the sections were incu- bated with the PBS-diluted ABC reagent (1: 100) for 1-2 hr, washed three times in PBS for 5 min each, and incubated for 5-30 min with 0.05% 3,3'-diaminobenzidine tetrahydrochloride (DAB) and 0.01% HO; in 0.05 M Tris buffer (pH 7.2). Finally, the sections were washed in distilled water for 15min, dehydrated, and mounted in Per- mount. Electron microscopic immunocytochemistry com- bined with retrograde HRP labeling One of a pair of Br3s was tied by a single knot using a hair, and the proximal cut end of the other Br3 was plunged into a thin capillary filled with 5% HRP for 24hr at 4°C, and then the TAG plus attached Br3s were fixed in 2.5% glutaraldehyde in 0.1M cacodylate buffer (pH. 7.2) for 3hr, fol- lowed by washing overnight in the buffer. Next, DAB enzyme reaction was performed according Nassel [35]. After the enzymatic reaction, the tissues were washed in the same buffer for 5 hr, dehydrated and embedded in epoxy resin (Luveak, Nakarai Tesque). No osmium postfixation was carried out. For electron microscopic immunocy- tochemistry without HRP tracing, the accessory glands or TAG (with Br3s ligated for 6 hr) were fixed for 3 hr after dissection, washed in buffer, dehydrated, and embedded in the resin. The postembedding protein A-colloidal gold method [36] was employed as follows: ultrathin sections were collected on formvar-carbon coated nickel grids and soaked for 30 min with drops of PBS with 1% BSA. The proctolin antiserum was then applied at 1:200-1:1,000 (in PBS with 1% BSA for 24hr at 4°C). Afterwards, the grids were attached magnetically to the inner wall of a beaker and washed in PBS by furious agitation with a magnetic stirrer for 30 min. The protein A-gold complexes (Janssen Pharmaceutical; particle size: 10 or 15 nm) were applied at a concentration of 1:40 (in PBS with 1% BSA) for 1 hr at room temperature. After washing (as above) for 40 min, the sections were stained with uranyl acetate and lead citrate. For the sections of HRP labeled tissues, lead citrate staining was eliminated. Final- ly, sections were examined with a Hitachi H 500 electron microscope. RESULTS Proctolin-like immunoreactive neurons in the TAG The proctolin antiserum revealed both of bilaterally symmetrical and medial immunoreac- tive neurons in the TAG. Symmetrically arranged neurons with proctolin-like immunoreactivity (PLI) were prominent in embryonic segments 7 and 9 of the TAG, which are derived from diffe- rent segmental ganglia during embryonic develop- ment [37]. Figure 1 shows an example of dorso- laterally situated bilaterally symmetrical neurons with PLI in the region corresponding to segment 8. Other neural elements with PLI are also visualized in both the dorsal commissure and the dorsolateral neuropil. These neurons and neural elements were constantly stained in all the preparations. Medial neurons with PLI were observed in the dorsal and ventral midline of the TAG, although the number of neurons and the degree of immunoreactivity varied from preparation to preparation. Figure 2 shows that the dorsomedial neurons with PLI are located in the caudal region corresponding to segment 10 of the TAG. In male crickets, the dorsal unpaired median (DUMR7 neurons) innervating the accessory gland form three clusters along the midline in positions corres- ponding to segments 8-10 or 11 of the TAG [2, 38]. Dissection or ligation of the nerve branches (Br3s) through which the DUMR7 neurons send their axons to the accessory gland, led to inten- sified immunoreactivity of the medial neurons in the caudal TAG. In the cockroach, it has been reported that dissection of the nerve running from the sixth abdominal ganglion to oviduct results in an increase of PLI in the ganglion [39]. The intensification of PLI in the caudally located me- dial neurons following the dissection or ligation of Br3s suggests that some of these medial neurons neurons 56 K. YASUYAMA, T. KIMURA AND T. YAMAGUCHI Fic. 1. A pair of dorsolateral, proctolin-like immunoreactive neurons (arrows) in the terminal abdominal ganglion. Immunoreactive fiber profiles are also seen in the dorsal commissure (arrowheads) and in the dorsolateral neuropil (double arrowheads). may be DUMR7 neurons. The section shown in Figure 2 was made from the cricket in which the Br3s were ligated for 6hr before fixation. The medial neurons were generally stained rather weakly, but the ventral commissure and dorsome- dial neuropil constantly revealed strong PLI (Fig. 2). This ventral commissure carries the bilaterally bifurcating axons of DUMR7 neurons in the caud- al part of the ganglion (Fig. 3). This observation supports the possibility that some DUMR7 neurons are proctolinergic. Electron microscopic observation of proctolin-like immunoreactive neurons The protein A-gold method revealed the wide distribution of the axons with PLI in the muscula- ture associated with the accessory gland, 1.e., in the single-fibrillar muscles winding around most of each glandular tubule, as well as in the multifibril- lar thick muscles surrounding the opening of each tubule in the anterior portion of the ejaculatory duct. PLI was restricted to large electron-dense granular vesicles with a diameter of 100-115 nm in the axons (Fig. 4). These immunoreactive vesicles were also detected in the ligated Br3s (Fig. 5). Our previous study [2] showed that the ligation of Br3s caused a heavy accumulation of large elec- Cross-section at the level of segment 8. x 140. tron-dense granular vesicles in the axoplasm pro- ximal to the ligature. Figure 5 shows an axon containing proctolin-like immunoreactive vesicles — (ca. 100 nm in diameter) which accumulated after ligation. These findings suggest that the proctolin- like immunoreactive vesicles are produced by neurons in the TAG. It should also be noted in Figure 5 that there is a non-immunoreactive axon with smaller granular vesicles (ca. 80 nm in dia- meter) than those in the immunoreactive axons. In order to localize PLI in the DUMR7 neurons, electron microscopic immunocytochemistry com- bined with HRP labeling and ligation was per- formed. When one of the Br3s was ligated and HRP was applied to the other Br3, labeling was noted of both the somata of DUMR7 neurons and the axons at the site of ligation, where numerous large granular vesicles accumulated at the ligature. Figure 3 shows HRP-labeled DUMR7 neurons and their secondary neurites running through the later- al commissure located in the caudal region of TAG. The protein A-gold method was applied to the region of the TAG shown in Figure 3. A few proctolin-like immunoreactive, electron-dense granular vesicles were found dispersed in the HRP labeled somata (Fig. 6). The size of the vesicles with PLI ranged from 50 to 200 nm and the mean Proctolin-Like Immunoreactivity in DUM Neurons Fic. 2. Dorsomedial proctrolin-like immunoreactive neurons (arrows) in the caudal region of the terminal abdominal ganglion. Cross-section. 230. Fic. 3. HRP-labeled DUMR7 neurons in the caudal region of the terminal abdominal ganglion. Immunoreactive fiber profiles are also seen in the ventral commissure (arrowheads). Retrograde HRP lebeling through the right Br3 revealed a number of DUMR7 somata and their axons (arrowheads) running through the ventral commissure in the caudal region. 2. Cross-section. 230. diameter was 108 nm (n=88). Vesicles with PLI were seen to associate with Golgi bodies, but were not seen in the cisternae of these structure. Proc- tolin-like immunoreactive vesicles were also seen close to lysosome-like structures (Fig. 7), as pre- viously described for PLI neurons in the cockroach hypocerebral ganglion [23]. Application of the The level of this section corresponds to that shown in Fig. protein A-gold method to the ligated Br3, revealed PLI in the large electrondense granular vesicles (ca. 100 nm in diameter) accumulated in the HRP labeled axons (Fig. 8). PLI could not be localized in the somata and axons strongly labeled with HRP, as the dense HRP reaction products prevented the identifica- 58 K. YASUYAMA, T. KIMURA AND T. YAMAGUCHI Fic. 4. An axon (ax) containing proctolin-like immunoreactive, electron-dense granular vesicles and running within the proximal part of the accessory gland tubules where multilayered muscle fibers (mf) are located. HRP labeling was not performed. The section was stained with uranyl x 50,000. Part of a transverse section of a ligated Br3. Numerous vesicles are accumulated in the axons (ax1, ax2) at the labeling was restricted to the vesicles. acetate for 5 min and with lead citrate for 1 min. Fic. 5. Immunogold site of ligation. An axon (ax1) containing large electron-dense granular vesicles with proctolin-like immunoreac- tivity, is seen, together with an axon (ax2) containing smaller non-immunoreactive vesicles. HRP labeling was not performed. The section was stained with uranyl acetate for 5 min and with lead citrate for 1 min. 48,000. tion of vesicles or the cellular organelles. Howev- er, aS shown in Figures 6 and 8, PLI could be detected in the electron-dense granular vesicles of moderately HRP labeled somata and axons. Although PLI was weaker in the vesicles of HRP- labled axons when compared with unlabled axons (Fig. 5), the gold label could still be clearly seen present over the accumulated vesicles in the for- mer group of axons (Fig. 8). It seems obvious that HRP applied to severed Br3s was transported to the contralateral axons of DUMR7 neurons through their somata in the Proctolin-Like Immunoreactivity in DUM Neurons 59 Fics. 6, 7. Parts of HRP-labeled somata of DUMR7 neurons in the caudal region of the terminal abdominal ganglion. (Fig. 6). 40,000. Proctolin-like immunoreactive, electron-dense granular vesicles are found dispersed in the cytoplasm Immunoreactive vesicles are also seen close to a lysosome-like structure (ly) (Fig. 7). The sections were stained with uranyl acetate alone for 1 min. x 40,000. Fic. 8. Proctolin-like immunoreactive vesicles in the HRP-labeled axon (ax) within the ligated Br3. This section was obtained from a preparation in which one Br3 was ligated and the other Br3 was retrogradely labeled with HRP. HRP labeled the axons with accumulated vesicles in the ligated Br3. granular vesicles is seen. TAG due to the characteristic bilateral morpholo- gy of these neurons. Therefore, these results suggest that some DUMR7 neurons are procto- linergic: a proctolin-like substance may be pro- duced by these DUMR7 neurons and transported through their axons to the muscles as a neuro- muscular transmitter or cotransmitter. The section was stained with uranyl acetate alone for 1 min. Gold labeling of electron-dense x 40,000. DISCUSSION A number of studies have provided evidence that proctolin plays a role as a neuroregulatory substance in the visceral muscles of the female reproductive organs of insects [e.g., 9, 13-16, 26]. Our present and previous studies [2, 3] have concerned the proctolinergic innervation of the visceral muscles associated with the male repro- ductive organs of crickets. 60 K. YASUYAMA, T. KIMURA AND T. YAMAGUCHI The previous study demonstrated that the crick- et accessory gland muscles are innervated by dor- sal unpaired median neurons, DUMR7_ neurons, with axons running bilaterally through Br3s of the TAG seventh nerve roots [2]. The present experi- ment showed that the ligation of one Br3 and retrograde HRP labeling of the opposite Br3 caused labeling of the somata of DUMR7 neurons and of axonal granular vesicles which accumulated in the axoplasm proximal to the ligature. Thus, the HRP-labeled axons containing these vesicles were identified as originating from DUMR7 neurons and proctolin-like immunoreactive vesicles were localized in HRP-labeled DUMR7 somata and axons by the postembedding protein A-gold method (Figs. 6-8). Our present results reinforce the supposition that DUMR7 neurons are procto- linergic excitatory motoneurons that induce neurogenic contraction of the accessory gland mus- culature by releasing proctolin [3]. The ultrastructure of the proctolin-like im- munoreactive vesicles has been characterized in the ganglia and peripheral nerves of a few insects by the preembedding PAP technique [17, 19, 25, 40] and the postembedding immunogold technique [23]. The immunoreactive terminals in the dorso- caudal neuropil of the cockroach TAG were de- monstrated to contain large dense vesicles with a diameter of 140-150 nm and numerous small clear vesicles [19]. In the blowfly, immunoreactive granular vesicles with a diameter of about 100 nm were found in the immunoreactive somata of the protocerebrum and in the terminal of the pars intercerebralis [25]. The immunoreactive termin- als in the lateral abdominal nerves were reported to contain clear vesicles of about 75 nm in dia- meter and granular vesicles with a diameters of ca. 115nm [40]. The postembedding immunogold technique has demonstrated three types of procto- lin-like immunoreactive vesicles in the cockroach: small electron-dense vesicles having a mean dia- meter of 80nm in the frontal and hypocerebral ganglia and in the musculature of the oviduct and hindgut, large electron-dense vesicles (150 nm) in the frontal ganglion, and large electron-lucent vesicles (150 nm) with flocculent contents in the hypocerebral ganglion [23]. In the DUMR7 neurons of the cricket, the proctolin-like im- munoreactive vesicles in HRP-labeled somata ranged from 50 nm to 200 nm in diameter (mean diameter: 108 nm) (Figs. 6, 7). In the HRP- labeled axons within the ligated Br3, and in the axons innervating the accessory gland muscula- ture, immunoreactive vesicles with a diameter of 100-110 nm were noted (Figs. 4, 8). Ude and Eckert [23] have suggested that the different sizes of vesicles with PLI might be due to the colocaliza- tion of proctolin with various neuroactive subst- ances in the vesicles. The neurons producing these vesicles were not demonstrated in the present study, though ligation of Br3s also revealed the accumulation of non-immunoreactive vesicles (ca. 80 nm in diameter) (Fig. 5). Light microscopic observation showed a weak PLI of the medial neurons in the cricket TAG (Fig. 2). This finding was in accordance with electron microscopic abservation, which showed that the vesicles with PLI were dispersed in the DUMR7 somata (Figs. 6, 7). It seems probable that the immunoreactive vesicles produced by these neurons are rapidly transported to the peripheral organs. This is supported by the observation that — the accumulation of vesicles occurs in a rather short period of time (3 hr or less) after Br3 ligation (Yasuyama, unpublished data). Dorsal unpaired median (DUM) neurons have been described in several insects [e.g., 41-48]. Hoyle [49] first suggested that the DUMETi neuron innervating the extensor tibiae mucsle of the locust is octopaminergic, and the presence of octopamine in DUMETi soma was definitely de- monstrated by Evans and O’Shea [50]. DUM neurons of the firefly TAG have also been shown to contain significant levels of octopamine [51]. It is now established that octopamine is the substance produced and released as a neuromodulator by DUMETi neuron and probably by other DUM neurons as well. Octopamine and other biogenic amine-containing cells in invertebrates stain speci- fically with the dye neutral red [e.g., 52-54] and DUMR7 neurons are stained selectively by this dye [38]. In the cricket accessory gland, octopa- mine was effective in increasing the frequency of myogenically evoked contractions at concentra- tions as low as 10°’ M [3]. Therefore, it is possible that octopamine coexists with proctolin in the Proctolin-Like Immunoreactivity in DUM Neurons 61 DUMR7 neurons. Furthermore, as suggested by Pfluger and Wat- son [48], it is possible that there are the neurons which are morphologically similar to DUM neurons but have a different role in the innervation of visceral muscle. In the female locust, Lange and Orchard [46] reported that retrograde labeling of the branches of the oviduct nerve revealed a total of eight neurons (three pairs of bilaterally symmet- rical neurons and two DUM neurons [DUMOV neurons]) within the seventh abdominal ganglion. Subsequently, it was shown that DUMOV neurons contain octopamine by using a radioenzyme assay [55], and that bilaterally paired neurons are prob- ably proctolinergic by immunocytochemistry [28]. In addition, the presence of another anterior clus- ter of DUM neurons within the seventh ganglion of the female locust was also revealed by retrog- rade labeling of the oviduct nerve, [31, 48, 56]. Kiss et al. [56] reported two types of axon termin- als containing granular vesicles in the locust ovi- duct musculature: one of them makes conventional neuromuscular junctions but the other type con- taining larger vesicles does not. The DUMETi neuron is thought to release transmitter adjacent to the muscle fibers without forming an anatomi- cally specific neuromuscular junction [57]. On the basis of these facts, Pflliger and Watson [48] sug- gested that the anterior cluster of DUM neurons within the seventh abdominal ganglion of the female locust may be motoneurons having a diffe- rent function from the posterior octopaminergic DUMOV neurons. This possibility is also sug- gested by the detection of a number of DUM neurons following retrograde labeling of the genit- al nerve running from the male locust TAG. In the case of cricket DUMR7 neurons, the somata con- tain large electron-dense vesicles which are iden- tical with those found in the axon terminals making neuromuscular junctions with the accessory gland musculature and with those in their neurites [2]. Our previous physiological study suggested that the DUMR7 neurons are probably motoneurons mediating the squeezing of secretion from the tubules of the accessory gland [3]. Slow coxal depressor of Ds motoneuron is well known to be proctolin-containing motoneuron [11, 58]. This neuron lies bilaterally in the third thoracic ganglion of the cockroach and innervates the muscles of a proximal segment of the leg. O’Shea and Bishop [11] used a combination of intracellular dye injection and immunocytochemis- try to demonstrate that the Ds motoneuron had PLI, while the DUM neurons in the third thoracic ganglion had no PLI. In female locusts, no PLI has been located in the dorsomedial region within the seventh abdominal ganglion [28]. On the other hand, a number of studies on the mapping of PLI in the insect nervous system have shown proctolin-like immunoreactive neurons along the midline of the ganglia [e.g., 17, 18, 20, 25, 40, 59-64]. Most of these neurons appear to be or are proposed to be paired neurons, and only a few unpaired median neurons with PLI have been reported. In the cockroach TAG, one of the four ventral unpaired median (VUM) neurons inner- vating the oviduct muscle and the sphincter muscle was characterized as having PLI [64]. In lepidop- tera, Davis et al. [62] have rerpoted the dorsome- dial unpaired neuron with PLI (K neuron). This neuron occurs in all of the abdominal ganglia and appears to project bilaterally to the perivisceral organs. The K neuron was proposed to have a nuerohumor function like the midline bilateral (MB) cells in the abdominal ganglia of Manduca sexta, which have bilateral projections to the peri- visceral organs and produce cardioacceleratory peptides [65]. In addition, neurons with PLI have been found in the caudal region of the TAG or the fused abdominal ganglion in several insects (e.g., the grasshopper [20]; the cockroach [17, 18]; the Col- orado potato beetle [63]; and diptera [40, 61]. Some of these neurons were proposed to innervate the hindgut [e.g., 17, 40, 61]. In grasshoppers, the anteromedial (AM) cells with PLI in the terminal (fifth) abdominal ganglion have been reported to project to the hindgut [66]. DUMR7 neurons in male crickets form three clusters in the posterior half of the TAG, and our present study showed that the most caudal cluster of DUMR7 neurons innervating the male reproductive organs had PLI. Further experiments are needed to explore the coexistence of proctolin and octopamine in DUMR7 neurons and to explore the functional role of individual DUMR7 neurons in mediating 62 the contraction of the accessory gland according to a temporal motor program. ACKNOWLEDGMENTS This work was suported in part by a Grant-in-Aid for scientific research from Ministry of Education, Science and Culture of Japan. 10 11 12 13 REFERENCES Loher, W. (1974) Circadian control of sperma- tophore formation in the cricket Teleogryllus com- modus Walker, J. Insect Physiol., 20: 1155-1172. Yasuyama, K., Kimura, T. and Yamaguchi, T (1988) Musculature and innervation of the internal reproductive organs in the male cricket, with special reference to the projection of unpaired median neurons of the terminal abdominal ganglion. Zool. Sci., 5: 767-780. Kimura, T., Yasuyama, K. and Yamaguchi, T. (1989) Proctolinergic innervation of the accessory gland in male cricket (Gryllus bimaculatus): detec- tion of proctolin and some pharmacological prop- erties of myogenically and neurogenically evoked contractions. J. Insect Physiol., 35: 251-264. Brown, B. E. (1975) Proctolin: a peptide transmit- ter candidate in insects. Life Sci., 17: 1241-1252. Brown, B. E. and Starratt, A. N. (1975) Isolation of proctolin, a myotropic peptide, from Periplaneta americana. J. Insect Physiol., 21: 1879-1881. Starratt, A. N. and Brown, B. E. (1975) Structure of pentapeptide proctolin, a proposed neurotrans- mitter in insects. Life Sci., 17: 1253-1256. Orchard, I., Belanger, J. H. and Lange, A. B. (1989) Proctolin: a review with emphasis of insects. J. Neurobiol., 20: 470-496. Piek, T. and Mantel, P. (1977) Myogenic contrac- tions in locust muscle induced by proctolin and by wasp, Philanthus triangulum, venom. J. Insect Phy- siol., 23: 321-325. Cook, B. J. and Meola, S. (1978) The oviduct musculature of the horesefly, Tabanus sulcifrons, and its response to 5-hydroxytryptamine and procto- lin. Physiol. Entomol., 3: 273-280. O’Shea, M. and Adams, M. E. (1981) Pentapeptide (proctolin) associated with an identified neuron. Science, 213: 567-569. O’Shea, M. and Bishop, C. A. (1982) Neuropeptide proctolin associated with an identified skeletal motoneuron. J. Neurosci., 2: 1242-1251. Adams, M. E. and O’Shea, M. (1983) Peptide contrasmitter at a neuromuscular junction. Science, 221: 286-288. Lange, A. B. and Orchard, I. (1984) Some pharma- 14 5) 16 7! 18 19 20 7 22 23 24 a) 26 K. YASUYAMA, T. KIMURA AND T. YAMAGUCHI cological properties of neuromuscular transmission in the oviduct of the locust, Locusta migratoria. Arch. Insect Biochem. Physiol., 1: 231-241. Cook, B. J. and Holman, G. M. (1985) The role of proctolin and glutamate in the _ excitation- contraction coupling of insect visceral muscle. Comp. Biochem. Physiol., 80C: 65-73. Stoya, G. and Penzlin, H. (1988) The effect of different putative transmitters on spontaneous pha- sic contractions of cockroach oviduct. Zool. Jb. Physiol., 92: 301-311. Lange, A. B. (1990) The presence of proctolin in the reproductive system of Rhodnius prolixus. J. Insect Physiol., 36: 345-351. Eckert, M., Agricola, H. and Penzlin, H. (1981) Immunocytochemical identification of proctolinlike immunoreactivity in the terminal ganglion and hind- gut of cockroach Periplaneta americana (L.). Cell Tissue Res., 217: 633-645. Bishop, C. A. and O’Shea, M. (1982) Neuropeptide proctolin (H-Arg-Tyr-Leu-Pro-Thr-OH): Im- munocytochemical mapping of neurons in the cen- tral nervous system of the cockroach. J. Comp. Neurol., 207: 223-238. Agricola, H., Eckert, M., Ude, J., Birkenbeil, H. and Penzlin, H. (1985) The distribution of a procto- lin-like immunoreactive material in the terminal ganglion of the cockroach, Periplaneta americana. Cell Tissue Res., 239: 203-209. : Keshishian, H. and O’Shea, M. (1985) The distribu- tion of a peptide neurotrasmitter in the post- embryonic grasshopper central nervous system. J. Neurosci., 5: 992-1004. Marder, E., Hooper, S. L. and Siwicki, K. K. (1986) Modulatory action and distribution of the neuropep- tide proctolin in the crustacean stomatogastric ner- ous system. J. Comp. Neurol., 243: 454-467. Siwicki, K. K. and Bishop, C. A. (1986) Mapping of proctolinlike immunoreactivity in the nervous sys- tems of lobster and crayfish. J. Comp. Neurol., 243: 435-453. Ude, J. and Eckert, M. (1988) Submicroscopic characterization of proctolin-like immunoreactivity in the nervous system of the cockroach, Periplaneta americana L. Cell Tissue Res., 254: 197-202. Schwarz, IT. L., Lee, G. M -E; Siwickin Ke K-. Standaert, D. G. and Kravitz, E. A. (1984) Procto- lin in the lobster: the distribution, release, and chemical characterization of a likely neurohormone. J. Neurosci., 4: 1300-1311. Nassel, D. R. and O’Shea, M. (1987) Proctolin-like immunoreactive neurons in the blowfly central ner- vous system. J. Comp. Neurol., 265: 437-454. Cook, B. J. (1981) The action of proctolin and 5-hydroxytryptamine on the oviduct of the horesefly, Tabanus proximis. Int. J. Invert. Re- Di 28 29 30 31 32 33 34 35 36 37 38 39 Proctolin-Like Immunoreactivity in DUM Neurons 63 - prod., 3: 209-212. Holman, G. M. and Cook, B. J. (1985) Proctolin, its presence in and action on the oviduct of an insect. Comp. Biochem. Physiol., 80C: 61-64. Lange, A. B., Orchard, I. and Adams, M. E. (1986) Peptidergic innervation of insect reproductive tis- sue: the association of proctolin with oviduct visceral musculature. J. Comp. Neurol., 254: 279-286. Orchard, I. and Lange, A. B. (1987) The release of octopamine and proctolin from an insect visceral muscle: effects of high-potassium saline and neural stimulation. Brain Res., 413: 251-258. Orchard, I. and Lange, A. B. (1987) Cockroach oviducts: the presence and release of octopamine and proctolin. J. Insect Physiol., 33: 265-268. Eckert, M., Varanka, I. and Benedeczky, I. (1989) Proctolinergic innervation of the locust oviduct. Zool. Jb. Physiol., 93: 471-479. Bishop, C. A., O’Shea, M. and Miller, R. (1981) Neuropeptide proctolin (H-Arg-Tyr-Leu-Pro-Thr- OH): immunological detection and neuronal loca- lization in insect central nervous system. Proc. Natl. Acad. Sci. USA, 78: 5899-5902. Eckert, M. and Ude, J. (1983) Immunocytochemic- al techniques for the identification of peptidergic neurons. In “Functional neuroanatomy. Springer Series in Experimental Entomology”. Ed. by N. J. I. Strausfeld, Springer, Heidelberg, Berlin, New York, pp. 267-301. Newman, G. R., Jasani, B. and Williams, E. D. (1983) A simple post-embedding system for rapid demonstration of tissue antigens under the electron microscope. Histochem. J., 15: 543-555. Nassel, D. R. (1983) Horseradish peroxidase and other heme protens as neuronal markers. In “Func- tional Neuroanatomy. Springer Series in Ex- perimental Entomology”. Ed. by N. J. I. Satraus- feld, Springer, Heidelberg, Berlin, New York, pp. 44-9]. Roth, J. (1983) The colloid gold marker system for light- and electron microscopic cytochemistry. In “Techniques in Immunocytochemistry”, vol. 2. Ed. by G. R. Bullock and P. Petrusz, Academic Press, London, New York, pp. 217-284. Panov, A. A. (1964) Correlations in the ontogenetic development of the central nervous system in the house cricket Gryllus domesticus L. and the mole cricket Gryllotalpa gryllotalpa L. (Orthoptera, Gryl- loidea). Entomol. Rev., 45: 179-185. Yamaguchi, T., Kushiro, N. and Waki, T. (1985) Sexual dimorphism of the terminal abdominal gang- lion of the cricket. Naturwiss., 72: 153-154. Sobek, L., Eckert, M., Penzlin, H. and Reissmann, S. (1986) Evidence for proctolinergic innervation of the cockroach oviduct. Zool. Jb. Physiol., 90: 461- 466. 40 4] 42 43 44 45 46 47 48 49 50 51 52 33 Nassel, D. R., Holmgqvist, B. I. and Movérus, B. J. A. (1989) Vasopresin- and proctolin-like im- munoreactive efferent neurons in blowfly abdominal ganglia: development and ultrastructure. J. Comp. Neurol., 283: 450-463. Plotnikova, S. I. (1969) Effector neurones with several axons in the ventral nerve cord of the Asian grasshopper Locusta migratoria. J. Evol. Biochem. Physiol., 5: 276-278. Hoyle, G., Dangan, D., Moberly, B. and Colqu- houn, W. (1974) Dorsal unpaired median insect neurons make neurosecretory endings on skeletal muscle. J. Exp. Zool., 187: 159-165. Casaday, G. B. and Camhi, J. M. (1976) Meta- morphosis of flight motor neurones in the moth Manduca sexta. J. Comp. Physiol., 112: 143-158. Davis, N. T. and Alanis, J. (1979) Morphological and electrophysiological characteristics of a dorsal unpaired median neuron of the cricket, Acheta domesticus. Comp. Biochem. Physiol., 62A: 777- 788. Christensen, T. A. and Carlson, A. D. (1981) Symmetrically organized dorsal unpaired median (DUM) neurones and flash control in the male firefly, Photuris versicolor. J. Exp. Biol., 93: 133- 147. Lange, A. B. and Orchard, I. (1984) Dorsal un- paired median neurons, and ventral bilaterally paired neurons, project to a visceral muscle in an insect. J. Neurobiol., 15: 441-453. Watson, A. H. D. (1984) The dorsal unpaired median neurons of the locust metathoracic ganglion: neural structure and diversity, and synapse distribu- tion. J. Neurocytol., 13: 303-327. Pfluger, H.-J. and Watson, A. H. D. (1988) Struc- ture and distribution of dorsal unpaired median (DUM) neurones in the abdominal nerve cord of male and female locusts. J. Comp. Neurol., 268: 329-345. Hoyle, G. (1975) Evidence that insect dorsal un- paired median (DUM) neurons are octopaminergic. J. Exp. Zool., 193: 425-431. Evans, P. D. and O’Shea, M. (1978) The identifica- tion of an octopaminergic neurone and the modula- tion of a myogenic rhythm in the locust. J. Exp. Biol., 73: 235-260. Christensen, T. A., Sherman, T. G., McCaman, R. E. and Carlson, A. D. (1983) Presence of octopa- mine in firefly photomotor neurons. Neuroscience, 9: 183-189. Evans, P. D. and O’Shea, M. (1977) An octopa- minergic neurone modulates neuromuscular trans- mission in the locust. Nature, 270: 257-259. Wallace, B. G., Talamo, B. R., Evans, P. D. and Kravitz, E. A. (1974) Octopamine: selective asso- ciation with specific neurons in the lobster nervous 54 a5 56 i 58 59 60 61 64 system. Brain Res., 74: 349-355. Stuart, A. E., Hudspeth, A. J. and Hall, Z. W. (1974) Vital staining of specific monoamine- containing cells in the leech nervous system. Cell Tissue Res., 153: 55-61. Orchard, I. and Lange, A. B. (1985) Evidence for octopaminergic modulation of an insect visceral muscle. J. Neurobiol., 16: 171-181. Kiss, T., Waranka, I. and Benedeczky, I. (1984) Neuromuscular transmission in the visceral muscle of locust oviduct. Neuroscience, 12: 309-322. Hoyle, G., Colquhoun, W. and Williams, M. (1980) Fine structure of an octopaminergic neuron and its terminals. J. Neurobiol., 11: 103-126. Witten, J. L. and O’Shea, M. (1985) Peptidergic innervation of insect skeletal muscle: immunoche- mical observations. J. Comp. Neurol., 242: 93-101. Lange, A. B., Orchard, I. and Barrett, F. M. (1988) The presence and distribution of proctolin in the blood-feeding bug, Rhodnius prolixus. J. Insect. Physiol., 34: 379-386. Breidbach, O. and Dircksen, H. (1989) Proctolin- immunoreactive neurons persist during metamorph- osis of an insect: a developmental study of the ventral nerve cord of Tenebrio molitor (Coleoptera). Cell Tissue Res., 257: 217-225. Anderson, M. S., Halpern, M. E. and Keshishian, 62 63 64 65 66 K. YASUYAMA, T. KIMURA AND T. YAMAGUCHI H. (1988) Identification of the neuropeptide trans- mitter proctolin in Drosophila larvae: characteriza- tion of muscle fiber-specific neuromuscular endings. J. Neurosci., 8: 242-255. Davis, N. T., Velleman, S. G., Kingan, T. G. and Keshishian, H. (1989) Identification and distribu- tion of a proctolin-like neuropeptide in the nervous system of the Gypsy moth, Lymantria dispar, and in other Lepidoptera. J. Comp. Neurol., 283: 71-85. Veenstra, J. A., Romberg-Privee, H. M. and Schooneveld, H. (1985) A proctolin-like peptide and its immunocytochemical localization in the Col- orado potato beetle, Leptinotarsa decemlineata. Cell Tissue Res., 240: 535-540. Stoya, G., Agricola, H., Eckert, M. and Penzlin, H. (1989) Investigations on the innervation of the oviduct muscle of the cockroach, Periplaneta amer- icana (L.). Zool. Jb. Physiol., 93: 75-86. Tublitz, N. J. and Truman, J. W. (1985) Identifica- tion of neurones containing cardioacceleratory pep- tides (CAPs) in the ventral nerve cord of the tabacco hawkmoth, Manduca sexta. J. Exp. Biol., 116: 395- 410. Keshishian, H. and O’Shea, M. (1985) The acquisi- tion and expression of a peptidergic phenotype in the grasshopper embryo. J. Neurosci., 5: 1005- LOIS: ZOOLOGICAL SCIENCE 9: 65-75 (1992) ULTRASTRUCTURE OF THE MOTILE IRIDOPHORES | OF THE NEON TETRA Hirosut NAGAISHI and Noriko Osuma! Department of Biomolecular Science, Faculty of Science, Toho University, Miyama, Funabashi, Chiba 274, Japan ABSTRACT—As a contribution to efforts aimed at the explanation for the motile mechanism of iridophores in the neon tetra (Paracheirodon innesi), the ultrastructure of the cells was studied by electron microscopy. One or two layers of iridophores are present under the subepidermal collagenous lamella, and melanophores lie beneath these layers. Within each iridophore, two stacks of light- reflecting platelets are arranged symmetrically, overlying the slender nucleus. Each platelet was found to be surrounded by two indenpendent membranes, namely, an inner and an outer membrane. Microtubules were observed in the cortical cytoplasm, and microfilaments were also found around the stacks of platelets. The edges of the platelets in each stack appeared to be joined together by microfilaments. The possible involvement of the organelles in changing the angle of inclination of the © 1992 Zoological Society of Japan reflecting platelets is discussed. INTRODUCTION Accompanying recent progress in studies on light-absorbing chromatophores, our understand- ing of light-reflecting chromatophores has also been enhanced. In particular, unique iridophores, which exhibit cellular motility and are referred to as “motile iridophores”, have lately attracted con- siderable attention. From their studies of the iridophores responsible for the generation of the blue-green color of the lateral stripes of the neon tetra, Lythgoe and Shand [1] and Clothier and Lythgoe [2] reported that the cells are directly affected by light and that the spectrum of light reflected from the cells shifts according to changes in the ambient illumination. Actually, the color of the stripe changes in vivo from green in the daytime to violet-blue at night. The motile activity of the iridophores was also found to be controlled by the sympathetic nervous system [3]. The fluorescence-like coloration produced by the iri- dophores appears to result from the phenomenon of non-ideal thin-film interference [4, 5]. A shift in the peak wavelength of light reflected Accepted October 1, 1991 Received September 2, 1991 ' To whom correspondence should be addressed from the iridophores (i.e., a change in color) is thought to be due to a change in the distance between adjacent platelets [1, 6, 7]. In our pre- vious paper, we proposed the “theory of venetian blinds”: a change in the distance between the reflecting platelets is caused by a change in the angle of inclination of the platelets [6]. However, the mechanism controlling the inclination of the platelets has not yet been elucidated. In case of the blue damselfish, Chrysiptera cyanea, a spherical nucleus is present in the upper part of the iridophore and many stacks of small thin platelets (guanine) are distributed radially from the nucleus. It was suggested that the change in the distance between the platelets occurs simul- taneously in each stack, resulting in a change in the lengths of the stacks themselves [7, 8]. The changes may resemble the expansion and contrac- tion of the bellows of an old-fashioned camera. The tubulin-dynein system may be involved in the motile activities of damselfish iridophores since the motility is inhibited by the treatment with colchi- cine, vinblastine or podophyllotoxin, antimitotic reagents, and by _ erythro-9[3-(2-hydroxyno- nyl)|adenine, an inhibitor of ATPase activity [9]. However, such a possibility has not yet been supported by observations at the electron micros- copic level. 66 H. NAGAISHI AND N. OSHIMA In the present study, the ultrastructure of the unique motile iridophores of the neon tetra was examined mainly under the electron microscope in order to facilitate the understanding of the mechanism responsible for changes in the angle of inclination of the reflecting platelets within the cells. MATERIALS AND METHODS Individuals of both sexes of the neon tetra, Paracheirodon innesi, 21-25 mm in body length, were purchased from a local commercial source and reared in an aquarium for several days before experiments. After decapitation, the body of the fish was sliced in half along the backbone and the skins, including lateral stripes, were cut into small pieces with a clean, sharp razor blade. These pieces were then fixed in a solution of 2.5% glutaraldehyde and 1% paraformaldehyde in 0.1 M phosphate buffer Te tensile CL rs: Fic. 1. Electron micrograph of a vertical section through iridophores (1) of the neon tetra. (pH 7.2) for 2 hr at room temperature. After a rinse with the buffer for 30 min, samples were postfixed with 1% OsQO, in 0.1 M phosphate buffer (pH 7.2) for 1 hr at 4°C. Specimens were dehy- drated through a graded ethanol series, treated with methyl glycidyl ether and embedded in Epoxy resin (Quetol 812; Nisshin EM, Tokyo). Thin sections were cut with a glass or a diamond knife on a Porter-Blum MT-1 ultramicrotome (Ivan Sor- vall, Newtown), mounted on formvar-coated grids, and then stained with 3% uranyl acetate and lead citrate. Some sections were coated with carbon after the staining. Specimens were observed under a Hitachi HU-12A electron micro- scope at 75 kV. One ym sections stained with toluidine blue and the entire skin preparation were observed through a microscope equipped both with epi-illumination and usual transmission optical systems (Optiphoto, XT-BD, with CF-BD plan objective lens; Nikon, Tokyo). ie ese “he Iridophores are situated just under the subepidermal collagenous lamella (CL) in two layers and a network of melanophores (M) is beneath them. A large slender nucleus (N) is present in the lower region of the cell. The reflecting platelets (RP) are stacked regularly at a fixed distance over the nucleus, while most of them were detached during the preparation of sections. MF: Muscle fiber. Mt: Mitochondria. Bar=2pm. Ultrastructure of Motile Iridophore RESULTS General morphology The iridophores of the neon tetra are shaped like a rectangular parallelepiped of about 60-70 ym in length, 20-30 «m in breadth, and 1-2 ~m in height. As shown in Figure 1, electron micro- graphs at low power revealed that layers of iri- dophores are situated just under the subepidermal collagenous lamella, being backed by a network of melanophores, in the stripe skin. Observation of the entire skin preparation under a light-field epi-illumination microscope showed that the iri- dophores were densely distributed like paving stones in the region of the lateral stripes in two layers, though they were in a single layer in some areas. When the iridophores in the upper layer were focused, ones in the lower stratum were also visible, though out of focus. In vertical sections of one um thickness, the two layers of iridophores |m asc po F = en B ML Ficss: —~ Y 00 fan nC ; hy) i} 1 | poe ' AREY Schematic drawing showing the distribution of the motile iridophores in the stripe skin. Fic. 2. Micrograph of one pm vertical section of the stripe skin. Two layers of the iridophores (1) are recognized, but in the right part of the photograph, cells are in a single layer. CL: collagenous lamella. M: Melanophore. Bar=20 um. were also recognized (Fig. 2). The longitudinal axis (long axis) of each iridophore was found to run parallel with the dorso-ventral axis of the fish. The distribution of iridophores in the fish skin is illustrated in Figure 3. The iridophores overlap each other with thinned portions of cytoplasm to form a thin reflective stratum (Fig. 4A). The thinned portions of cyto- plasm in adjoining cells are often tangled as shown p 1 ( A: Lateral view of the fish. Part of the lateral stripe consisting of the iridophores is magnified in the middle frame and the arrangement of the reflecting platelets within the cells is drawn in the right one. Dorso-ventral axis indicated as D—V corresponds to D—V in (A) of this figure. and the arrangement of the reflecting platelets are drawn in the frames. ML: Muscle layer. Epidermis. M: Melanophore. Scale. I: Iridophore. B: Frontal view of the fish. Distribution of the iridophores CL: Collagenous lamella. E: N: Nucleus. RP: Reflecting platelet. S: 68 H. NAGAISHI Fic. 4. Parts of vertical sections through iridophores showing that the thinned portions of cytoplams in adjoining iridophores overlap one another (A) or are tangled (B). Arrowheads indicate the border of each cell. Bar=1 pum. we 4 SE am OO. Fic.5. Electron micrograph showing a desmosome (arrow) by which neighboring iridophores are con- nected. Bar=1 ym. AND N. OSHIMA in Figure 4B. Occasionally, the junctional appa- ratus of membranes can be observed (Fig. 5). In vertical sections of the stripe skin that were cut parallel to the longitudinal axes of the iridophores, a large slender nucleus is found in the lower part of each iridophore (Fig. 1). In cross sections of such cells, the triangle-shaped nucleus is located in the central region of the cell (Fig. 6). Mitochondria are present mainly in the cortical region of the cells, and in the thinned portions of cytoplasm (the edges of the cells), where the reflecting platelets are absent. Few mitochondria are seen in the region between the nucleus and the cell membrane (Figs. 1 and 6). Inside the cell membrane, many pinocytotic vesicles are found here and there (Fig. 7). Golgi apparatus (Fig. 8) and smooth endoplas- mic reticulum (Fig. 9) are sometimes observed. In addition, small electron-dense particles, which may be glycogen granules, are distributed through- out the cytoplasm, although they are scarcely found in the spaces sandwiched between the reflecting platelets. Arrangement of the light-reflecting platelets In vertical sections of the iridophores, many large reflecting platelets are observed over the nucleus (Figs. 1 and 3B). They are stacked reg- ularly at an equal distance from one another, and Fic. 6. Electron micrograph of a cross section of an iridophore. In this photograph, the iridophore is in a single layer. A triangle-shaped nucleus (N) is present in the central region of the cell. A stack of reflecting platelets (RP) is present on each side of the nucleus. CL: Collagenous lamella. M: Melanophore. Mt: Mitochondria. Bar=2 pm. Ultrastructure of Motile Iridophore 69 Sa = : : por STE Lge ernes Cua Fic. 7. A part of a horizontal section through an iri- dophore showing the pinocytotic vesicles. The section was cut through the plane near the cell membrane. Empty spaces in which the reflecting platelet has been present are seen on the left side of the photograph. The electron-dense tubular net- work (arrowheads) is thought to represent the in- vaginations of the cell membrane. Bar=1 yum. Fic. 8. Electron micrograph showing the Golgi appar- atus (arrow). The open space shows the area in which a reflecting platelet has been located. N: Nucleus. ae Sen eee ear e eRe eS aS ate’ 35 Fic. 9. Electron micrograph showing the smooth endo- plasmic reticulum (arrow). N: Nucleus. Bar=0.5 pm. they are inclined at a constant angle. Our light- microscopic observations of fixed skin under light- field epi-illumination revealed that each platelet is hexagonal in shape (guanine crystal; unpublihsed data). The long axis of most platelets is about 15- 20 um in length, and the guanine crystals become smaller at both ends of the stack. The actual reflecting platelets themselves are usually not observed in sections. As shown in Figures 1, 6, 10 etc., the platelets are represented by empty spaces since they become detached in the process of cutting or staining of sections. Moreover, the spaces are bulged by the exposure of sections to the electron beam during the observations. In cross sections of the iridophores, a pair of stacks is positioned above the nucleus or is present on both sides of the nucleus (Fig. 6). Occasionally, three stacks are seen in a cell (photographs not shown). In horizontal sections of the iridophores, each stack lies parallel to the longitudinal axis of the cell. Each platelet is inclined against the axis and the incline of the platelets in neighboring stacks is symmetrical (Figs. 10 and 3A). On the basis of these observations, three-dimensional construc- tion of iridophores is shown in Figure 11. Vesicles that include a fine guanine crystal are often observed in the proximity of both ends of the stack of the platelets (Fig. 12). Inner and outer membranes surrounding the platelet For the preservation of the reflecting platelets themselves, some sections were coated with car- bon after staining. Each platelet was found to be surrounded by two independent membranes; namely, an inner and an outer membrane, as shown in Figure 13. When the platelet is preserved completely, the inner membrane adheres closely to the guanine platelet and the outer membrane, giving the appearance of the fused membrane (Fig. 14). Therefore, the inner membrane is often barely visible. When the platelet is partly broken or detached, the inner membrane can be recognized as a very thin line (Fig. 13). A relatively electron- lucent space is present between these two mem- branes. Observations at high magnification suggest that 70 H. NAGAISHI AND N. OsHIMA Fic. 10. Horizontal section through an iridophore showing a pair of stacks of reflecting platelets (RP) parallel to the long axis of the cell. The arrangement of the platelets in both stacks is symmetrical. Mt: Mitochondria. N: Nucleus. Bar=2 yum. Reflecting Platelet Nucleus ; ee Dorsal Side Fic. 11. Illustration showing three-dimensional construction of the neon tetra iridophore. A slender nucleus is present in the lower part of the cell along the cell axis. Many reflecting platelets in hexagonal shape arrange regularly inclining at a constant angle. Long axis of each platelet is also out of perpendicular to the long axis of the cell. A pair of stacks of the platelets is positioned above the nucleus. As for the arrangement of the platelets within the cell, refer to the schematic drawings in Figure 3. the guanine platelets are not more than 5 nm thick (Fig. 14). However, guanine crystals partially re- maining in empty sacs are apparently very thick As shown in Figures 15 and 16, microtubules, (cf. Figs. 10 and 13), probably as a result of the — which run parallel or perpendicular to the longitu- deformation of the crystals. dinal axis of the iridophore, are observed in the Microtubules and microfilaments Ultrastructure of Motile Iridophore Tal ee at 7 Y é Me ARS zs ee s Oil a Pak A 2 Ee % +r A 3 i 3 Ear = = en x S LE 4 cs fy Pe “s < eg : < ‘ a, = ° Z < < ‘ = aus ca see) B = Fic. 12. Electron micrograph showing the vesicles that include a fine guanine crystal (arrowheads). Com- pare the size of the crystal with that of the reflecting platelets seen in Figures1, 4 and 8. N: Nucleus. Bar=1 um. Fic. 13. Electron micrograph showing the inner (arrow- heads) and outer (arrows) membranes that surround each reflecting platelet. Although sections were coated with carbon after electron staining in this series, the platelets themselves were partly or com- pletely broken. At the both ends of sacs, thick electron-dense lines are visible, since pieces of broken platelets adhered to the inner membrane. Bar=0.2 um. cortical cytoplasm of the iridophore. Compara- tively large numbers of microtubules are distri- buted in the lower part of the cells, and also on the edges of the cells, whereas there are only a few in the region sandwiched between the nucleus and the stacks of reflecting platelets (Fig. 15). No microtubules are seen in the spaces between the platelets. | Microfilaments of about 8 nm in diameter are found around the stacks of platelets. The edges of the platelets appear to be linked together by the microfilaments (Fig. 17). In cross sections, very long microfilaments are frequently found, as shown in Figure 18, whereas they are seldom seen showing a perfectly preserved reflecitng platelet (arrow). The platelet appears to be about 5 nm thick. When the platelet is perfectly preserved, no inner membrane can be recognized since this mem- brane adheres closely to the platelet. Each platelet is also surrounded closely by the outer membrane (arrowheads), although the membrane has become partially detached from the platelet. At the edges of the platelets, vesicular structures are visible. This section was coated with carbon after electron staining. Bar=0.2 um. Fic. 15: the microtubules (arrows) that run parallel to the Part of a cross section of an iridophore showing longitudinal axis of the cell. While microtubules are present mainly at the bottom of the iridophore, they are not abundant in the cell. Note the bundle of the microtubules within the melanophore. M: Melanophore. Mt: Mitochondria. N: Nucleus. RP: Reflecting platelets. Bar=0.5 um. in vertical sections. On the other hand, thin filamentous structures, which might connect the outer membrane that surrounds one platelet with an adjacent one, are also observed (Fig. 18). These filaments are about 6nm in diameter. The long 8-nm filaments running in parallel with one another are sometimes linked by fine fibrous struc- tures, which are very similar to those seen between the platelets (Fig. 18). The microfilaments of about 8 nm in diameter are often connected to the cell membrane by the fine filamentous structures, Fic. 16. Part of a horizontal section through an iri- dophore showing the microtubules (arrows) that run perpendicular and parallel to the longitudinal axis of the cell. They are probably present near the cell membrane, since many pinocytotic vesicles are observed around the microtubules. The large arrow indicates the longitudinal axis of the cell. Bar=0.5 um. Zo NS * Sehr Fic. 17. Cross section of an iridophore showing the microfilaments (arrows) that connect the edges of stacked reflecting platelets (RP). Fine filament- like structures (arrowheads) are derived from the microfilaments and some of them are connected to the cell membrane. The platelets themselves are no longer present (see text). Bar=0.2 um. showing microfilaments and fine filament-like struc- tures. Several long microfilaments (arrows) run along the side of the reflecting platelet (RP), and many fine filament-like structures (arrowheads) are seen between the adjacent platelets. The microfila- ments and the outer membrane that surrounds the reflecting platelet seem to be connected by the filamentous structures. The outer membranes are also mutually connected by these structures. Bar= 0.2 um. . NAGAISHI AND N. OSHIMA as shown in Figure 17. The 8-nm filaments are also distributed randomly in the cytoplasm, and some- times they form a lattice (photographs not shown). However, these filaments are not inserted deeply into the spaces between the reflecting platelets. DISCUSSION In the present study, electron microscopy clearly revealed the way in which the iridophores are distributed in the integument of the lateral stripe region of the neon tetra, and the three- dimensional arrangement of the reflecting platelets within the iridophores was clarified. Generally, within the cell, there are two stacks of more than 100 platelets and each platelet makes an acute angle of depression with respect to the median plane of the fish’s body (cf. Fig. 3B). This observa- tion accords well with our previous observation that the highest reflectance is obtained when illu- mination is applied from the dorsal side of the fish. Since the long axis of the reflecting platelets in each stack is not perpendicular to the longitudinal axis of the iridophore (cf. Figs. 10 and 11) and, furthermore, since some stacks are more or less curved, irradiation from the direction of the head or tail should also result in a low level of reflection; a phenomenon that has been observed in practice [5]. Two layers of iridophores were generally found in the stripe skin. The presence of the two layers may have the effect of purifying the interference colors reflected from the lateral stripes and may augment the reflection because of the resultant increase in the number of platelets that participate in the multilayered thin-film interference [4, 5]. Microbutules were found within the iridophores of the neon tetra. They have also been observed in iridophores of the Atlantic salmon, Salmo salar L. [10], and of the freshwater goby, Odontobutis obscura [11]. Although the role of microtubules in the cells of the latter two species has not been discussed in detail, it is assumed that the microtu- bules function as the cytoskeleton. In case of the neon tetra iridophores, the microtubules are not likely to act as the cytoskeleton since only a relatively small number of microtubules is observed and they are mainly found in the lower Ultrastructure of Motile Iridophore 73 part of the cell. Each of a pair of stacks itself may maintain the shape of the neon tetra iridophores rather than common cytoskeletal elements such as microtubules. In the dermal iridophores of the lizard, Anolis carolinensis, in the iridophores in the scales of the goldfish and Holocentrus ascensionis, and in the tapetal iridophores of the cardinal tetra, all of which are physiologically inactive iridophores, both thin (6.5 nm in diameter) and thick (10 nm) filaments are present [12, 13]. Not only thin filaments but also 10-nm filaments are considered to be cytoskeletal elements. By contrast, in phy- siologically active iridophores, such as the dermal iridophores of the leopard frog, Rana pipiens |14] and the iridophores in the scales of the cardinal tetra [13] and the freshwater goby, Odontobutis obscura [11], thin filaments are abundant and thick filaments are rarely found. In the neon tetra iridophores, thick filaments of about 8 nm in dia- meter are frequently observed in addition to thin- ner filamentous structures, although the cells are of the motile type. It seems that the 8-nm filament and the thinner fibrous structures correspond to the thick and the thin filaments present in non- motile iridophores, respectively. However, a small difference in the diameter of the two respective types of filament should be noted. Rohrlich [13] suggested that the 6.5-nm fila- ments present in physiologically active iridophores might serve to hold the crystalline sheets in their parallel array and might also participate in the movement of the platelets. These filaments were supposed to be composed of actin. Thin filament- like structures in the neon tetra iridophores, which are present in the cytoplasm between the guanine platelets, also my act as the spacers required for the parallel arrangement of the platelets. In addi- tion, the presence of such structures might keep the surface of the platelets flat. In the neon tetra iridophores, closer attention should be paid to the 8-nm filaments, which seem to surround the stacks of the platelets, linking their edges together. Thus, via the action of these filaments, the guanine crystals may be piled up evenly. Furthermore, it is proposed that the 8-nm filaments may be involved in changes in the in- clination of the platelets. In order to unequivocal- ly identify the two types of filaments and confirm the spatial distribution of them, immunohistoche- mical analyses are in progress. For researchers interested in the motile activity of pigment cells, an essential problem is the elu- cidation of the way in which the motive force is generated. The role of microtubules [15, 16, 17, 18], microtrabecular systems [19] and actin fila- ments [20] has been suggested. Above all, the tubulin-dynein system seems to be a highly prob- able candidate for the driving force behind pig- ment migration [21, 22]. Pharmacological analyses have revealed that antimitotic reagents, colchicine and vinblastine [23], and erythro-9[3-(2-hydroxy- nonyl)|adenine (EHNA), an ATPase inhibitor, block the translocation of pigment [24]. In case of the motile iridophores of the blue damselfish, Oshima and Fujii [9] also reported that antimitotic reagents and EHNA inhibit the responses of the cells. In the motile iridophores of the neon tetra, we suggest the possible involvement of the 8-nm fila- ments in changing the inclination of the reflecting platelets. If slipping between the 8-nm filaments that link the edges of platelets and the microtu- bules, or the network of filaments with which the cell membrane is backed, occurs on the upper and lower sides (epidermal and dermal sides) in oppo- site directions, the inclination of all platelets can be changed synchronously. If the microtubules participate in this phenomenon, sliding might take place only on the dermal side since the microtu- bules are present mainly in the lower part of the iridophores. In such a case, the upper end of each platelet should be fixed. However, the structural relationship between the microtubules and the filaments has not yet been fully defined. Moreov- er, it is a matter of concern that the number of microtubules might be not sufficient to allow them to discharge their duties as a “foothold”. In the physiological experiments, we have found that colchicine, vinblastine and EHNA block re- versibly the change in stripe color caused by the application of norepinephrine, implying that the motive force responsible for changing the inclina- tion of the platelets is generated by dynein-tubulin interactions. Surprisingly, cytochalasin B also in- hibited the iridophore response (unpublised data, 74 H. NAGAISHI AND N. OSHIMA Oshima and Nagaishi). This fact suggests that the motile mechanism of the neon tetra iridophores may differ from that of Chrysiptera cells, whose response to norepinephrine is never inhibited by cytochalasin B [9]. The possible role of the actin filaments in the motile activity of the neon tetra iridophores also should be considered. Therefore, we must be in a hurry for the identification of 6-nm and 8-nm filaments seen in these cells. In the present study, we found that the guanine platelets are enveloped in double bound mem- branes, which resemble the fused membrane. This finding is in accord with the results from fish and amphibian iridophores reported by Kamishima [25, 26]. However, Matsuno and Iga [11] reported that each reflecting platelet contained in the iri- dophores of the freshwater goby (Odontobutis obscura) is surrounded by a limiting membrane, and that, at its lateral margins, the platelet is supported by two membranes.. A similar observa- tion was reported by Rohrlich and Porter [12] in a study of the iridophores of Anolis carolinensis. Further comparative observations at the ultra- structural level are clearly required. ACKNOWLEDGMENTS We thank Prof. R. Fujii of our Department for his interest and encouragement. The authors are also grate- ful to Dr. K. Miyaji of Toho University for his kind technical advice. This work was supported by a Grant-in- Aid from the Ministry of Education, Science and Culture of Japan to N.O. (No. 93540590), and by the Science Research Promotion Fund from the Japan Private School Promotion Foundation. REFERENCES 1 Lythgoe, J. N. and Shand, J. (1982) Changes in spectral reflexions from the iridophores of the neon tetra. J. Physiol., 325: 23-34. 2 Clothier, J. and Lythgoe, J. N. (1987) Light- induced colour changes by the iridophores of the neon tetra, Paracheirodon innesi. J. Cell Sci., 88: 663-668. 3 Nagaishi, H. and Oshima, N. (1989) Neural control of motile activity of light-sensitive iridophores in the neon tetra. Pigment Cell Res., 2: 485-492. 4 Huxley, A. F. (1968) A theoretical treatment of the reflexion of light by multi-layer structures. J. Exp. Biol., 48: 227-245. 5 10 11 2 13 14 15 16 17 18 19 Land, M. F. (1972) The physics and biology of animal reflectors. Prog. Biophys. Molec. Biol., 24: 75-106. Nagaishi, H., Oshima, N. and Fujii, R. (1990) Light-reflecting properties of the iridophores of the neon tetra, Paracheirodon innesi. Comp. Biochem. Physiol., 95A: 337-341. Kasukawa, H., Oshima, N. and Fujii, R. (1987) Mechanism of light reflection in blue damselfish motile iridophore. Zool. Sci., 4: 243-257. Oshima, N., Sato, M., Kumazawa, T., Okeda, N., Kasukawa, H. and Fujii, R. (1985) Motile iri- dophores play the leading role in damselfish colora- tion. In “Pigment Cell 1985: Biological, Molecular and Clinical Aspects of Pigmentation”. Ed. by J. T. Bagnara, S. N. Klaus, E. Paul and M. Schartl. Univ. Tokyo Press, Tokyo, pp. 241-246. Oshima, N. and Fujii, R. (1987) Motile mechanism of blue damselfish (Chrysiptera cyanea) iridophores. Cell. Motil. Cytoskel., 8: 85-90. Harris, J. E. and Hunt, S. (1973) The fine structure of iridophores in the skin of the Atlantic salmon (Salmo salar L.). Tissue and Cell, 5: 479-488. Matsuno, A. and Iga, T. (1989) Ultrastructural observations of motile iridophores from the freshwa- ter goby, Odontobutis obscura. Pigment Cell Res., 2: 431-438. Rohrlich, S. T- and Porter," K- "R2"@972) Fine - structural observations relating to the production of color by the iridophores of a lizard Anolis caro- linesis. J. Cell Biol., 53: 38-52. Rohrlich, S. T. (1974) Fine structural demonstra- tion of ordered arrays of cytoplasmic filaments in vertebrate iridophores: A comparative survey. J. Cell Biol., 62: 295-304. Taylor, J. D. and Bagnara, J. T. (1972) Dermal chromatophores. Am. Zool., 12: 43-62. Bikle, D., Tilney, L. G. and Porter, K. R. (1966) Microtubules and pigment migration in the mela- nophores of Fundulus heteroclitus L. Protoplasma, 61: 322-345. Murphy, D. B. and Tilney, L. G. (1974) The role of microtubules in the movement of pigment granules in teleost melanophores. J. Cell Biol., 61: 757-779. Schliwa, M. and Euteneuer, U. (1978) Quantitative analysis of the microtubule system in isolated fish melanophores. J. Supramol. Struct., 8: 177-190. Obika, M. and Negishi, S. (1985) Effect of hexylene glycol and nocodazole on microtubules and melano- some translocation in melanophores of the medaka Oryzias latipes. J. Exp. Zool., 235: 55-63. Byers, H. R. and Porter, K. R. (1977) Transforma- tions in the structure of the cytoplasmic ground substance in erythrophores during pigment aggrega- tion and dispersion. I. A study using whole-cell preparations in stereo high voltage electron micros- 20 21 22 28 Ultrastructure of Motile Iridophore TS copy. J. Cell Biol., 75: 541-558. Akiyama, T. and Matsumoto, J. (1983) The block- ade of pigment displacement in_ swordtail erythrophores by microinjection of antiactin anti- body. J. Exp. Zool., 227: 405-411. Nakamura, N., Ikeda, Y. and Obika, M. (1987) Video and electron microscopic studies on pigment transport in Gambusia melanophores. Jpn. J. Ichthyol., 34: 351-360. Oshima, N., Inagaki, H. and Manabe, T. (1990) Evidence for involvement of dynein-tubulin system in pigment aggregation within tilapia melanophores. Comp. Biochem. Physiol., 96A: 517-523. Obika, M., Turner, W. A., Jr., Negishi, S., Menter, 24 DS 26 D. G., Tchen, T. T. and Taylor, J. D. (1978) The effects of lumicolchicine, colchicine and vinblastine on pigment migration in fish chromatophores. J. Exp. Zool., 205: 95-109. Beckerle, M. C. and Porter, K. R. (1982) Inhibitors of dynein activity block intracellular transport in erythrophores. Nature, 295: 701-703. Kamishima, Y. (1979) Electron microscopic study on two types of reflecting cells in the ventral skin of the sand eel, Ammodytes personatus. Proc. Jpn. Acad., 55B: 141-146. Kamishima, Y. (1990) Organization of reflecting platelets in vertebrate iridophores. Zool. Sci., 7: 1020. el ae hari aie See a oF 7 t 2. 4 = F , i =i Ny = 3 1 = o / t ZOOLOGICAL SCIENCE 9: 77-88 (1992) © 1992 Zoological Society of Japan Origin of Multinucleality of Muscle Cells during Myogenesis of the Newt IWANE SATO* Department of Biology, College of general Education, Osaka University, Toyonaka 560, Japan ABSTRACT—The process of multinucleality formation in muscle cells was traced in a Japanese newt. Mitotic figures of muscle cell nuclei are fairly abundant during the embryonal and larval stages in general. Cytokinesis occurs during the myoblast stages, but after a cell has become slender, and is fixed between myosepta, cytokinesis does not occur. No evidence of cell fusion between muscle cells was observed. Macerated specimens of muscle tissue were useful for the counting of the total muscle cell numbers in one myotome. The results show that the muscle cell number in one myotome does not change remarkably during the embryonal and larval stages. The invation or fusion of other kinds of cells was not observed. Hence the multinucleality of muscle cells is the result of mitosis without cytokinesis of one myoblast, and not the product of cell fusion. INTRODUCTION There are two opposing opinions as to the origin of multinucleality in muscle cells of vertebrates. One is “by cell fusion” and the other is “by mitosis without cytokinesis”. In the first case the result would be a syncytium, and in the second, a plas- modium. There have been neither decisive mic- rophotographs nor illustrations for either theory and the controversy continues. The purpose of this paper is to critique this discrepancy by the most effective material from the standopoints of embryology, histology, cytology and karyology. MATERIALS AND METHODS Myotome muscle tissues of the developing embryos and larvae of a Japanese newt, Cynops pyrrhogaster pyrrhogaster Boie, were used. The larvae of each developmental stage were fixed as a whole in Bouin’s fluid. Serial sections in 10 ~m thickness were made. Horizontal section series were most useful. Longitudinal and transverse section series of each stage were also made respec- Accepted October 19, 1991 Received August 7, 1991 * Retired in March, 1973. Author’s present address: Kamibettocho 36, Kitashir- akawa, Sakyoku, Kyoto 606, Japan tively as the occasion demanded. Double staining in Delafield’s haematoxylin and eosin was applied universally. To count the total muscle cell number in one myotome, the maceration technique was used. As larval tissue is soft and fragile, trypsin medium was unnecessary. The unilateral single certain myotome was dissected from a living larva under a dissection microscope in Holtfreter’s solution. Myotome tissue was then immersed in Ca**-free Holtfreter’s solution for about 10 minutes, and the softened unilateral myotome was transferred as a whole onto a slide with one drop of the same medium. The weight of a cover slip was just sufficient to isolate a myotome muscle into free cell groups. After counting the total cell number, the slide was immersed in 10% -formalin solution, stained and mounted for preservation. The larvae were developmentally staged according to the table by Y. K. Okada (1947) [1]. OBSERVATIONS Myogenesis “in vivo” starts as somitogenesis in the trunk region initially. When somites are formed, the cells are randomly arranged, but soon after, they differentiate into three groups. The superficial one is dermatome. The deeper one is myotome whose cells become arranged rather 78 I. Sato Fic. 1. a: somite, b: myocoel c: notochord. radially around the myocoel (Fig. 1). The third group, sclerotome, occupies a small area and is not conspicuous in later stages in this species (Fig. 1). The myoblasts are full of yolk granules with a small amount of melanin granules. They continue to divide mitotically with each following cytokinesis. The chromosome number of this species is 24 in diploid (Fig. 2). During the late neurula stage, myoblasts be- come slender and are arranged longitudinally. Such myoblasts are fixed between anterior and posterior myosepta which are the products of screlotome cells. The fixed end of a cell forms a small flat disc. In the early tail bud stage (stage 30), all the cells are mononucleated and full of yolk granules (Fig.3). Some nuclei also show the condensed chromosomes preparing for the coming mitotic activity. There is a problem in ascertaining the initial total number of myoblasts precisely, as it is dif- ficult to identify the myoblasts from other kinds of Longitudinal section through the notochord including bilateral somites of the tail bud stage. Sta2655, x60), elements. In a typical example of cell count in an embryo in stage 29, the total number of myoblasts in the fourth unilateral myotome is about 66. Among eight inidividuals examined, the number of myoblasts was not more than 75, and not less than 58. After initial cell organization in a myotome is completed, myoblasts start to divide mitotically. During mitotic activity the cell body becomes short and round. In most cases the rounding occurs on the inner surface of each posterior myoseptum. In this case both ends of the mitotic cell body are still fixed on myosepta, one end becoming slender and string-like (Fig. 4B). Sometimes such a figure might be mistaken as a mitotically dividing cell about to invade muscle tissue. Close attention must be given to the long, slender portion of the cells during mitotic activity. After mitosis cells return to the initial slender shape between the two myosepta. Cytokinesis does not follow and the cells remain binucleated retaining the myotube Multinucleality of Muscle Cells 79 ie: % Fic. 2. Mitosis of a myoblast. St. 26. Fic. 3. Mid-sagittal section of head and trunk myotomes c: myotome, d: yolk. structure. In Figure 5A, all the muscle cells are binucleated after the first mitosis. The number of yolk granules are decreased. Formation of vacuoles in the cytoplasm seems to be related to the dissolving yolk granules. Tetanucleated cells x 1,000. in early tail-bud stage. St. 28. a: brain, b: notochord, can also be seen. In this figure the inner chromatin pattern of two adjacent nuclei is quite similar. This shows that they are sister nuclei after one mitotic division. These nuclei are slightly rectangular in this specimen, which might be artifactual due 80 ca Fic. 4. The first mitosis of muscle cells. A: prophase, B: metaphase. SE 3k x500! Fic. 5. A: Binucleated stage after first mitosis. B: Pairs of sister nuclei after first mitosis. St. 32.) Aix<250) By x 600. Multinucleality of Muscle Cells 81 eh Fic. 6. primarily to heavy shrinkage along the long axis of the muscle cell during chemical fixation. In Stage 32 (Fig. 6), isolated muscle cells from the fourth myotome evidently show the presence of en- velopes (sarcolemma) around each cell. At around this stage (Stage 33) some cells are binucleated and others are tetranucleated (Fig. 7), yolk granules are diminished and large vacuoles within cyto- plasm have become more conspicuous. The de- velopment of myofibrillae starts in the early tail bud stage within the peripheral cytoplasm of a myotube. This phenomenon starts initially in the anterior trunk region, and propagates backward gradually. Large spinal ganglia are also completed corresponding and contacting with each side of respective myotomes (Fig. 8A). As a result of motor unit completion, an embryo can react against external stimuli, shaking head and trunk laterally in it’s egg capsule at this stage. Mitotic division is still going on, within the myotube structure. Figure 9 shows a cell with eight nuclei, A is in the tissue and B is an isolated one. All Isolated cell group fixed and stained after counting. St. 32. 400. « these nuclei are situated in the central cytoplasm, still retaining the myotube organization. As the cell is greatly shrunk, the transverse bands become clear under a light microscope. After hatching the yolk mass decrease rapidly, the muscle cells become longer and thicker, and the myotube structure is lost gradually. The nuc- lei, situated in the central cytoplasm in a row, gradually move into the myofibrill bundle and also to the peripheral portion of the cell. Sometimes a nucleus is situated just beneath the cell membrane, surrounded by a small amount of cytoplasm. Mi- totic division still occurs, but simultaneous nuclear division may no longer be expected, as was seen in one pair of sister nuclei in the common central cytoplasm. So such cells with nuclei of odd num- bers in one muscle cell are a common result. Examples of such mitotic division in the peripheral cytopasm are shown in Figure 10A, B and C. After hatching, the tail extends rapidly, but at the tip myotome muscle cells are still mono- nucleated, though myofibrillae are already well 82 I. SATO ee Fic. 7. Tetranucleated stage. St. 33. 400. a: myoseptum, b: vacuole. as ss Fic. 8. A: Longitudinal section of lower trunk region including bilateral myotomes. x4 ganglion. B: prophase. c: metaphase. St. 33. 400. 0. a: spinal cord, b. spinal Multinucleality of Muscle Cells 83 aw Fic. 9. Octanucleated muscle cell. St. 34. contrast microscope. Xx 600. ae: Fic. 10. Mitotic nuclear division in cells with well developed myofibrillae. developed in their cytoplasm. Soon mitotic activ- ity follows, and cells become binucleated. But in the tail region of larvae, muscle cells in the somite are rather shorter in length, and the occurrences of polynucleality are fewer. In Figure 11 one nucleus A: Trunk myotome cell. 400. B: isolated cell under a phase & 5 va PES After hatching. St. 34. is in late prophase with condensed chromosome groups. On the upper left side, one mitotic metaphase plate with mitotic apparatus is extruded together with a small amount of cytoplasm due to the remarkable shrinkage of myofibrillae during 84 I. SATO Fic. 11. The most caudal myotome in tail tip. All cells still mononucleated and with well developed myofibrillae. St. 32. 800. Upper left: metaphase with extruded nuclear plate. Fic. 12. A long muscle fiber in a cranial myotome with 16 nuclei. St. 34. 400. Multinucleality of Muscle Cells 85 Fic. 13. Longitudinal section through right forelimb. S52: chemical fixation. In this larva, it is shown that sixteen or slightly less nuclei are contained in one cell in the cranial myotome (Fig. 12). A significant comment about the myogenesis in appendages should be added (Fig. 13). The arms and legs of this urodelan amphibian are small and slender, so the total cells which constitute each appendage muscle are not abundant. Consequent- ly it is easy to count the total cell number even in the largest forelimb muscle, M. triceps brachii. The total cell number in this muscle is about 72. Both ends of the skeletal muscle cell, attaching to bones, is slender in shape, altogether different from myotome muscle cells whose cell ends form flat discs. But the process of multinucleality formation is quite similar to that seen in somito- genesis. Results of these observations are schema- tized in Figure 14. x 60. a & b: two heads of M. triceps brachii, c: humerus. DISCUSSION There are several papers and textbooks which refer to mitotic division of muscle cell nuclei in embryos and larvae of vertebrates. For instance, Weichert [2] said that “myoblasts begin to differ- entiate and become spindle-shaped and are arranged in bundle”. He also said “Mitotic divi- sion then occur, resulting increase in mass (chick embryo)”. Almost similar descriptions are seen in book by Schneider [3], Schaffer [4], Bucher [5] and so on. But recently, these observations and de- scriptions are mosly forgotten, or are neglected, and the presence of mitotic nuclear division in muscle tissue is brushed aside in the study of myogenesis. The cells of tailed amphibians are extraodianrily large and are convenient materials for the study of myogenesis. The author’s observations on newts are in complete accordance with these previous papers. 86 I. Sato myotube myoblast Fic. 14. Recently there appeared an argument worthy of notice, that is, the mitotic figures in developing muscle tissue are not the division of muscle cell nuclei itself, but are the nuclei of the so-called “satellite cells”. Concerning the nature of these cells, there are some precise and instructive discus- sions by Fawcet and Bloom [6], and Schultz [7]. It is generally known that there might exist some other kinds of dividing mesodermal elements and invading cells in muscle tissue. But in the author’s preparations the mitotically dividing cells show the presence of myofibrillae in every case. Moreover, it is clear that the dividing nuclei belong neither to the invading cells nor to some other kind of mesodermal element. It should be noted that in early developmental stages the cytoplasmic mass containing a mitotic structure sometimes protrudes from the neighboring muscle cells, and that such mass looks as if a cell with mitotic structure is invading the muscle tissue. For instance, the extrusion of the mitotic apparatus together with a = = ee t Hees ED Schema of myogenesis. small amount of cytoplasm (Fig. 11) might be of artifact, presumably due to the remarkable shrink- age and powerful contraction of the muscle tissue during chemical fixation. According to the au- thor’s experience, cells in mitotic activity, in most cases, become stationary and not motile. Even mesodermal wandering cells show a similar tendency. Consequently, invading figures of cells with mitotic structures seem somewhat ambiguous as to their origin “in vivo”. In “in vitro” specimens, on the contrary, the situation is quite different. When myoblasts are cultured “in vitro”, they show almost no mitotic activity after they start myofibrill formation in their cytoplasm. This fact was strongly emphasized by Holtzer [8]. His statement may be right as far as “in vitro” cultured cells are concerned. But in “in vivo” condition, that is to say, in normally de- veloping embryos and larvae, the mitotic division of muscle cell nuclei is common even after myo- fibriogenesis has begun. Multinucleality of Muscle Cells 87 Another serious problem is “cell fusion”. Ele- ments such as mesenchymal cells, fibroblasts and also myoblasts, tend to be highly thigmotactic, and they easily fuse to each other in “in vitro” condi- tion. In “in vivo” condition, such as within living organisms, the cells of mesodermal origin, fibro- blasts and histiocytes for example, envelop foreign substances like surgical strings or plastic tubes for medical use. In such cases, the cells fuse to each other and become multinucleated [9, 10]. The same phenomenon occurs in “in vitro” conditions between cultured cells and the inner surface of a culture bottle or a cover slip. The appearance of “Fremdk6rperriesenzellen” of Lambert [11] is a problem directly related with the fundamental nature of biological membrane. In normal physiological conditions “in vivo”, cell fusion seems to rarely occur. Between the differentiating muscle cells, fusion was never observed. Moreover, the author has never found plausible and reliable figures of cell fusion in published papers on the normal developmental course of myogenesis. The extrusion, sometimes observed, never means cell fusion, but merely means shrinkage in processing. The features of the cell body in question should be totally and careful- ly examined. To ascertain the formation of a multinucleated cell, the total cell number in the initial cell group as well as the multinucleated cell group should be analyzed. If cell fusion occurs among 130 mono- nucleated cells, the resultant binucleated cells might total 65, and the succeeding fusion might produce 33 tetranucleated cells. If the 3rd fusion occurs, 16 octanucleated cells might appear, and so on. This does not represent an actual case. According to MacCallum [12], the total number of muscle cells in the sartorius muscle in human embryos of various ages does not differ greatly from each other. Subsequently, according to Mac- Callum’s data, the increase in size of a muscle simply depends upon the growth of the individual cells. The author’s observations on myogenesis in a myotome are in agreement with this conclusion. In modern biology “in vitro” culture is conve- nient and an excellent method for cytological studies. Many findings on myofibriogenesis were made chiefly by cultured materials and under an electron microscope. On the other hand, the discrepancy between the “in vivo” and “in vitro” states on the behavior of cells as a whole should also be clarified. The author’s experience shows that no remarkable differences among those two can be seen as long as cell organellae are con- cerned, if cells are cultured properly “in vitro” and if tissue cells are physiologically normal “in vivo”. Because cell organellae, such as mitochondria, Golgi apparatus, myofibrillae and so on, are situ- ated in the cytoplasm of their own mother cell in both cases, organellae can exhibit their normal physiological behavior and show their original morphological features in both cases. But as to the behavior of cells as a whole, the condition might not always be the same in both cases. The differ- ence between culture media and normal internal milieu, and some differences in the cell environ- ment may be inevitable. These discrepancies might surely be diminished or removed in the near future by improvement of culture techniques. The simulation of physiological environmental condi- tions of muscle cells to each other during myogenesis will hopefully be achieved before long. ACKNOWLEDGMENTS The author is deeply grateful to those who have been friendly and sympathetic to this study. He also wishes to express his gratitude to his wife, Misako for her en- couragment and helpful preparation of this manuscript. REFERENCES 1 Okada, Y. K. (1947) Developmental Stage of Japanese Newt (Triturus pyrrhogaster). In “Ex- perimental Morphology, Vol. 3”. Kodansha, Tokyo. 2 Weichert, C. K. (1959) Element of Chordate Ana- tomy. McGraw Hill, New York, p. 500. 3 Schneider, K. C. (1902) Lehrbuch der Vergleichen- den Histologie. Gustav Fischer, Jena, s. 813. 4 Schaffer, J. (1920) Vorlesungen tiber Histologie und Histogenese. Wilhelm Engelmann, Leipzig, s. 209. 5 Bucher, O. (1973) Cytologie und mikroskopische Anatomie des Menschem. Stuttgart, Wien, s. 202. 6 Fawcett, D. W. and Bloom, W. (1975) A Textbook of Histology. Saunders, Philadelphia, p. 297. 7 Schultz, E. (1978) Changes in the Satellite Cells of Growing Muscle following Denervation. Anat. Rec., 190: 299-311. 8 Holtzer, H. (1970) Myogenesis. In “Cell Dif- 10 88 I. Sato ferentiation”. Ed. by O. A. Schjeide and J. de Vellis, Van Nostrand Co., New York, pp. 476-503. Kissane, J. M. (1990) Andersons’s Pathology, 9th ed. Vol. 1. The C. V. Mosby Co., St.Louis, p. 979. Martin, M. B. and Epstein, W. L. (1974) Formation of Multinucleate Giant Cells in Organized Epithe- lioid Cell Granulomas. Am. J. Pathol., 74: 263-274. 11 12 Lambert, R. A. (1912) The production of foreign body giant cell in vitro. J. exp. Med., 15: 510-515. MacCallum, J. M. (1895) On the histogenesis of striated muscle fibers and the growth of the human sartorius muscle. Johns Hopkins Hosp. Bull., 1808, cited from G. Clark (1971) The Tissue of the Body, Clarendon Press, Oxford, p. 147. ZOOLOGICAL SCIENCE 9: 89-99 (1992) Effect of Actinomycin D on Nuclear Events during Conjugation in the Ciliate Stylonychia pustulata JUNII YANO and MIkIo SUHAMA Zoolgical Institute, Faculty of Science, Hiroshima University, Kagamiyama, Higashi-hiroshima 724, Japan ABSTRACT—In conjugation of the ciliate Stylonychia pustulata, the relationship between RNA synthesis and nuclear behavior was examined by the use of actinomycin D (AmD) and *H-uridine. At different stages of conjugation, pairs were immersed in 50 ug/ml AmD. When AmD treatment was begun at the early and late stages of meiosis I, conjugation was prevented at 3 defined stages; meiosis I, postmeiotic division and formation of synkaryon. When AmD treatment was begun at meiosis II or at the early period of second postzygotic division, two nuclei produced by first postzygotic division were prevented from dividing at ana-telophase of second division (PZII), but separation of pairs into exconjugant-type cells and degradation of old macronuclei occurred. In many exconjugant-type cells, one of the 2 nuclei which did not divide in PZII swelled like a macronuclear anlage. When AmD treatment was begun at the stage of pronuclear differentiation, nuclear event of conjugation was blocked at the stage of macronuclear development. However, polytene chromosomes were visible in the macronuclear anlage. Autoradiographs of *H-uridine showed that labeled precursors were extensively incorporated into conjugants during meiotic division, but incorporation declined to zero or near the zero level until the stage of pronuclear differentiation. These results suggested that the nuclear events including meiosis, postmeiotic division, pronuclear differentiation and macronuclear develop- ment were largely controlled by RNA synthesized during meiosis. The sequential relationship between © 1992 Zoological Society of Japan RNA and protein syntheses during conjugation is discussed. INTRODUCTION In conjugation of ciliated protozoa, two mature cells of complementary mating type temporarily unite to form a pair. The conjugation includes the following nuclear events: meiosis of micronuclei (germinal nuclei), cross-fertilization, development of new micro- and macronuclei (somatic nuclei) and degradation of old macronuclei [1, 2]. The occurrence of RNA synthesis during the early period of conjugation has been known in Stylony- chia mytilus [3, 4] and Tetrahymena thermophila [5-8]. In the former the long-lived messenger RNA controls the further developmental process of conjugation. In T. thermophila, however, RNA synthesis again occurred in the late stage of con- jugation and in exconjugants [5-7]. In Para- mecium aurelia a net decrease of the cytoplasmic RNA content followed by a drastic increase occur- Accetped October 28, 1991 Received February 7, 1991 red during conjugation [9]. Thus, the initiation and duration of RNA synthesis in conjugation process are different among ciliates. In S. mytilus and 7. thermophila, however, the. relationship between RNA synthesized at the early period of conjugation and its role for nuclear changes in conjugation is unclear. How far does RNA affect nuclear changes which occur sequentially during conjugation? The aim of the present study is to analyze the pattern of RNA synthesis of S. pustulata in the whole process of conjugation and to compare it with the pattern of protein synthesis. Our previous study has shown that pair formation and initiation of meiosis in S. pustulata require RNA and protein syntheses during pair formation [10] and that the progress of furhter nuclear events of conjugation is also controlled by protein synthesized during con- jugation [11]. A detailed comparison of the patterns of RNA and protein synthesis in conjugation revealed two unique findings: synthesis of RNA corresponding 90 J. YANO AND M. SUHAMA to protein synthesis at different stages of meiosis and synthesis of RNA regulating almost all the cytological events after meiosis II. Such a com- parison has not yet been conducted in other cili- ates. The experiment was conducted by treating pairs with an inhibitor of RNA synthesis, acti- nomycin D, or labeling them with a RNA precur- sor, “H-uridine, at different stages of conjugation. MATERIALS AND METHODS Two stocks of complementary mating types, HH2 (mating type II) and TM6 (IV), of Stylony- chia pustulata were used. The stocks were cultured in a wheat infusion medium containing green algae Chlorogonium elongatum as food [12]. Cells from each mating type were separately washed twice with exhausted medium, and them mixed together. Pair formation started about 2 hr after mixing [13]. When pair formation reached near the maximum in the mixtures, A-shaped pairs were transferred within 15 min into wells of depression slides con- taining exhausted medium. This time was adopted as zero time of the formation of conjugation. At intervals of 30 min, about 100 pairs were selected at random from wells of the depression slides and fixed for 10min with acetic-alcohol (1 :3) on cover glass coated with Mayer’s albumin. The specimens were hydrolyzed for 40 min in 4N HCl at room temperature and stained for 30 min in Schiff’s reagent. The length of duration of each nuclear event in conjugation was estimated by observing the stained cells as follows. The dura- tion was defined as the time from the point that 50% of the pairs in a given time possessed the same nuclear stage to the point that 50% of the pairs in a later time had the next nuclear stage. Actinomycin D (AmD) dissolved in culture medium was added to the pairs in depression slides at known times after the onset of pairing. The last addition of AmD was made at the time when the pairs separated into exconjugants. The concentra- tion of AmD was determined as follow. Vegeta- tive cells immediately after cell division were treated at 10, 25, 50 and 100 ug/ml AmD. The replication bands which duplicated DNA in the macronuclei were not formed at 50 and 100 ug/ml AmD. At lower concentrations, the replication bands passed through the macronuclei. Thus, 50 pg/ml AmD was employed in this experiment. About 200 pairs were treated at the final concen- tration of 50 ug/ml. Every hour after the treat- ments, 20 pairs were randomly selected and stained with acetic carmine or Feulgen. The nuclear events were observed till 24 hr after the treatments. Each experiment was repeated three times at 23°C. The pairs treated with 50 ng/ml AmD showed normal morphology for 24 hr. For autoradiography, about 100 pairs were pulse-labeled with 100 ~Ci/ml °H-uridine (New England Nuclear; 26.2 Ci/ml) for 1 hr at intervals of one hour from the onset of pairing. On the other hand, the pairs were pre-treated with 50 yug/ ml AmD for 30 min, and then labeled with 100 uCi/ml *H-uridine for 1 hr in the presence of AmD. At the end of labeling procedure, the pairs were washed twice with the exhausted medium. The old macronuclei were isolated to distinguish the labeled RNA in them from the labeled RNA in the cytoplasm. Some labeled pairs were destroyed by an isolation medium containing 0.05% Triton X-100 and 0.01% spermidine phosphate (Sigma) with distilled water [15]. Pairs and isolated macro- nuclei were fixed in MFG solution (methanol: formalin: acetic acid=17:2:1) for several sec, again fixed in acetic-alcohol (1:3), spread on the subbed slide and then air-dried. Unincorporated label was eliminated by washing the slides with 5% trichloroacetic acid for 10 min at 4°C. In testing the specificity of isotope incorporation, the dupli- cated slides were treated with 0.05% ribonuclease A (RNase; Boehringer Mannheim) in 10mM phosphate buffer (pH 7.0) for 7 hr.at 37°C. Auto- radiographs were prepared with Sakura NR-M2 liquid emulsion. Exposure time was 14 days at 4°C. The labeled nuclei were stained with 2% methyl green through the emulsion after develop- ment. : Morphological changes in each nuclear event during conjugation have been described previously [13]. Vegetative cells of S. pustulata possess two micronuclei and two macronuclei. There are eight major micronuclear events: first and second maturation divisions (MI, MII), postmeiotic divi- sion (PM), pronuclear disivion (PD), formation of synkaryon (SK), first and second postzygotic di- RNA Synthesis during Conjugation 91 vison (PZI, PZII) and macronuclear development (MD). Stage MI will be described in detail later. Stage MII: two spherical nuclei rapidly undergo MII and two disk-like nuclei degenerate later. Stage PM: all four nuclei produced by MII begin to swell after a short time of interphase and then only two nuclei survive and take part in nuclear divi- sion. Four nuclei separate into anterior two and posterior two, since the mitotic spindles are oriented along the longitudinal axis of the con- jugant. The other two nuclei shrink and then degenerate. Resorption of old macronuclei into its cytoplasm begins. Stage PD: though four nuclei produced by PM swell after a short time of interph- ase, one nucleus in the anterior group differenti- ates into a migratory pronucleus and one nucleus in the posterior group into a stationary pronucleus (Fig. 3G). The remianing nuclei shrink and de- generate later. Stage SK: exchange of migratory pronuclei and subsequent fusion of migratory and stationary pronuclei occur. Stages PZI and PZII: two divisions of the synkaryon follow to yield four nuclei. Stage MD: of the four nuclei arranged in a file, the second and fourth sister nuclei from the anterior differentiate into new micronuclei, the third nucleus develops into a macronuclear anlage, and the first nucleus degenerates later (Fig. 3H). Each pair separates into two exconjugants. They possess oral apparatus (OA) with a small buccal cavity and small cytostome without paroral mem- branes (exconjugant-type OA). Polytene chromo- somes are developed in anlage. The macronuclear anlage shrinks and then amitotically divides into parts and the products become new macronuclei. The exconjugant-type OA is reorganized into a functionally normal OA at the time. RESULTS Duration of the nuclear events during conjugation Morphological changes in seven of the eight nuclear events except for MI during conjugation have already been described above. However, the exact time schedule of all the nuclear events has not yet been presented. By the method mentioned above, the micronuclear events among pairs pro- gressed synchronously to a considerable extent as shown in Figure 1. Moreover, the duration of each micronuclear events was estimated from the data l il ll IV VII Mill PM PZII MD 100 e-e-© “0-0 BEER he Ho OAKS 718) x in +-+ v » [" O V + Ww oy | vill © O © S 5 VI 2 A = % © 0) =F v Vv 5 50 & © O Ge : fo) ° oe ® io) e \ ‘ vf VW a ! O—- @-—g¢—_L_ 9 +—- 5 Ne -A- x -O-2© wey ee [KX hint it O 2 4 6 8 10 7 : 14 a 18 20 Time (h) Fic. 1. Progression of micronuclear changes during conjugation of Stylonychia pustulata. MI and MII, first and second meiotic division; I, compact nuclei; II, swollen nuclei; III, nuclei at early stage of parachute; IV, nuclei at mid stage of parachute; V, nuclei at late stage of parachute; VI, prometaphase I; VII, metaphase I; VIII, ana-telophase I; PM, postmeiotic division; PD, pronuclear differentiation; SK, formation of a synkaryon; PZI, first postzygotic division; PZII, second postzygotic division; MD, development of macronuclear anlage. 92 J. YANO AND M. SUHAMA 0 5 10 15 20 ee ee) Ae AmD / + tH _——— eee WHT —_—_—_—______$____—__+> — ———<$_————————————— _ Re a Ae eS Chx ~ es --_ =_— '— . /~ -——e. Fic. 2. Duration of each micronuclear event in conjugation (top) and effects of actinomycin D (AmD) and cycloheximide (Chx) [11] on micronuclear changes (bottom). The left end of arrow shows the onset of drug treatment, and the right end shows the end point of normal nuclear progress in the presence of drug. The length between the left and right ends shows the duration of normal nuclear progress. (Fig. 2). Especially, the micronuclear behavior from the beginning of pairing to the end of MI is newly subdivided into eight stages (stage I-VIII) on the basis of morphological features of micro- nuclei. At stage I (0-1.3 hr), two micronuclei in each conjugant remain compact after the onset of pairing. At stage II (1.3-3.8 hr), micronuclei swell from the usual size of 3 ~m in diameter to 10 um. Thin chromatin fibers first appear (Fig. 3A), fol- lowing chromatin granules in the periphery of nuclei (Fig. 3B). In each conjugant, the anterior macronucleus elongates. At stage III (3.8-6.7 hr), chromatin threads unilaterally emerge from the chromatin mass in the swollen micronuclei (Fig. 3C). The anterior macronucleus breaks into two pieces. At stage IV (6.7-8.8 hr), more chromatin threads are concentrated in a hemispherical shape, and the micronuclei assume a typical parachute- shape (Fig. 3D). At stage V (8.89.8 hr), conden- sation of chromatin threads takes place, but the small chromatin mass is still on the opposite side (Fig. 3E). At stage VI (9.8-10.2 hr), many small chromatin rods or chromosomes are dispersed in the nuclei (Fig. 3F). At stage VII (metaphase I, 10.2-10.8 hr), the chromosomes are arranged on the eugatorial plate. At stage VIII (ana-telophase I, 10.8-11.4hr), one daughter nucleus shows a spherical shape, and the other is a disk-like in shape. The duration of the other stages is measured as follows: stage MII, 11.4-12.4 hr; stage PM, 12.4- 15.3 hr; stage PD, 15.3-16.5 hr; stage SK, 16.5- 17.2 hr; stages PZI and PZII, 17.2-18.6 hr; and stage MD, 18.6-19.1 hr. Resorption of old macro- nuclei into the cytoplasm and separation of pairs occur about 14hr and 19.5 hr after the onset of pairing, respectively. Full appearance of polytene chromosomes and shrinkage of macronuclear anlange are observed about 20 and 60 hr after the separation of pairs, respectively. Effects of AmD on conjugation and nuclear events in conjugation Effect of AmD on micronuclear behavior is summarized in Figure 1. When A-shaped pairs were treated by AmD, they developed into com- plete pairs. The micronuclei proceeded from stage I to stage II, but the swollen nuclei showed unusual elongation in the presence of AmD (Fig. 4A). When pairs 2 and 3 hr after the onset of pairing RNA Synthesis during Conjugation 93 ¥ : <. : > Fic. 3. Nuclear state in different stages of conjugation. micronucleus of stage II 3 hr after pairing. Chromatin granules are located in one side of the nucleus. \H Bar=5 wm. A-F. Meiotic prophase of meiosis I. micronucleus (arrow) of stage II 2hr after pairing. The chromatin fibers appear in the nucleus. A.A B. A C.A micronucleus of stage III (early parachute) 5 hr after pairing. D.A micronucleus of stage IV (mid parachute) 7 hr after pairing. E. A micronucleus of stage V (late parachute) 9 hr after pairing. Many chromatin strands occupy in one side of the nucleus. F. A micronucleus of stage VI (prometaphase I) 10 hr after pairing. G. pronuclear differentiation 16hr after pairing. MP and SP show a migratory and a stationary nucleus, respectively. Arrows indicate degenerating nuclei. H. Differentiation of macronuclear anlage (MA) and new micronuclei (MI) 19 hr after the onset of pairing. Arrow indicates a degenerating nucleus. were treated with AmD, micronuclei proceeded from stage II to stage III, and their chromatin threads aggregated into one irregular shaped mass (Fig. 4B). Macronucleus fragmentation occurred, but resorption of the fragments scarcely took place. When pairs 4 and Shr after the onset of pairing were treated with AmD, micronuclei pro- ceeded from stage III to stage IV. However, the chromatin threads which appeared on one side of the nuclei became arranged in parallel rows about Shr after the treatment (Fig. 4C). Then, these nuclei were elongated and became spindle shaped, but the chromosomes were not arranged on the equatorial plate unlike those of the control pairs 94 J. YANO AND M. SUHAMA % aM _ 97. Fic. 4. Abnormalities of micronuclear events caused by actinomycin D (AmD). Bar=5 ym. A. Elongated micronuclei of a conjugant 5 hr after a pair was treated 15 min from the onset of pairing. B. Chromatin threads aggregating in micronuclei of a conjugant 5 hr after a 2-hr-old pair was treated. C. Chromatin threads arranged in parallel rows in two micronuclei of a conjugant 5 hr after a 4-hr-old pair was treated. D. Two micronuciei showing irregular spindle shapes. The nuclear stage is later than C. E and F. Two nuclei (arrows) blocked at anaphase (E) and telophase (F) of PM in conjugants 5 hr after a 9-hr-old pair was treated. G-H. Nuclei in conjugants treated 12 hr after the onset of pairing. G. Two nuclei (arrows) blocked at PZII remain unchanged in an exconjugant-type cell 10 hr after the onset of treatment. H. One of the 2 nuclei blocked at ana-telophase of PZII swells like macronuclear anlage (arrow) in an exconjugant-type cell 10 hr after the onset of treatment. (Fig. 4D). When AmD treatment was begun at 6 became arranged in parallel rows, but they were and 7 hr after pairing (late stage III or early step of | longer than chromosomes of metaphase I (stage stage IV), micronuclei entered stage V. However, VII) and often adhered each other. Then nuclei 3 hr after the treatment, the chromatin strands assumed a spindle shape like Figure 4D. When RNA Synthesis during Conjugation 95 AmD treatment was begun at 8 hr (stage IV), nuclear division was blocked at ana-telophase I (stage VIII). Two daughter nuclei were connected with the chromatin materials. When AmD treatment was begun at 9 and 10 hr (stage V or VI), MI and MII occurred, and then two nuclei entered into PM. However, the nuclear division was disrupted at ana- or telophase in the presence of AmD. In the former cases, two daughter neuclei stopped at anaphase, and then two daughter nuclei remained adhered (Fig. 4E). In the latter cases, two nuclei were blocked at telophase and remained connected with some chromatin materials (Fig. 4F). When AmD treat- ment was begun at 11 hr (stage VII or VIII), PM and PD occurred normally, but the formation of BIG.35;. For distribution of grains, see text. Arrows indicate macronuclei. FV, food vacuole. macronuclei at stage PZII. Autoradiographs of pairs and isolated macronuclei pulse-labeled for 1 hr with *H-uridi ee a ne at different times. Bars in A and E are 40 um, and in the others 10 um. A. A pair at stage III. B. Isolated old macronuclei at stage MII. isolated from a pair treated with AmD at stage III and labeled for 1 hr. D. Macronuclei E. A pair at stage PM. F. Isolated old 96 J. YANO AND M. SUHAMA synkaryon failed. When AmD treatment was begun at 12, 13 and 14 hr (stage MII or PM), nuclear event of conjuga- tion was blocked at ana-telophase of PZII. Each pair was separated into two cells 19 hr after pairing or slightly later. The cells possessed an excon- jugant-type OA. About 50% of these cells had neither macronuclear anlage nor new micronuc- leus (Fig. 4G), and others had one swollen nucleus like macronuclear anlage derived from one of the two nuclei which failed to divide in PZII (Fig. 4H). In such nuclei, polytene chromosomes were not found. The old macronuclei degraded. When AmD treatment was begun at 15, 16, 17, 18 and 19 hr (late PM stage or before the separation of pair), pairs separated at almost the same time as the non-treated control. Exconjugants possessed a macronuclear anlage and new _ micronuclei. Polytene chromosomes were observed in the mac- ronuclear anlage. However, unlike the non- treated control, the macronuclear anlage did not shrink. Moreover, the reorganization of excon- jugant-type OA into a functionally normal OA was inhibited by AmD treatment. RNA synthesis pattern during conjugation When pairs were pulse-labeled with *H-uridine at stages I-VIII and autoradiographed, silver grains were observed on the macronuclei and cytoplasm (Fig. 5A, B, C). Many of the grains disappeared in conjugating cells and isolated mac- ronuclei when the cells were treated with AmD (Fig. 5D) or when the preparations were treated with RNase. These suggest that the grains indicate specific incorporation of *H-uridine into RNA which is synthesized in macronuclei. RNA synth- esis actively occurred in macronuclei at stages MI and MII (Fig. 5B, C). The number of grains on micronuclei of meiotic prophases was almost simi- lar as those in the background of the cytoplasm. Macronuclear and cytoplasmic labels remark- ably decreased to zero or near the zero at stages PD, PZI, PZII and MD (Fig. 5E, F). When exconjugants with polytene chromosomes were labeled for 4 hr or more, a few grains were observed on the macronuclear anlage but not on the old macronuclei (Fig. 6A). Thus, the anlage may synthesize RNA weakly in this stage. The Fic. 6. Autoradiographs of an exconjugant and a young vegetative cell. Bar=20 um. A. An exconjugant cell 24 hr after the separation of pair and labeled for 4hr. The silver grains of macronuclear anlage (dotted line) are slightly greater in number than those of cytoplasm. No grains are located in the old macronuclei (arrows). B. A young vegetative cell having the normal oral apparatus after the reorga- nization and labeled for 1 hr. Two macronuclei (arrows) have much more silver grains than the macronuclear anlage in Fig. 6A. exconjugants had small cytostome without paroral membranes (exconjugant-type OA) and the incor- poration of precursors into the food vacuoles of exconjugants was at an extremely low level. This may influence RNA synthesis of macronuclear analge. Silver grains were observed on the new macronuclei and cytoplasm in cell with normal OA about 70 hr after separation of a pair (Fig. 6B). Therefore, most of RNA precursors may be in- corporated into exconjugant through its small cytome. DISCUSSION RNA synthesis in the macronucleus Macronuclei in the conjugating cells of Stylony- chia pustulata synthesize RNA from the onset of pairing to stage MII. When pairs were split at 10 min after pairing, the split cells underwent auto- gamy [14]. Ten min after the start of pairing, two partners of a pair loosely unite with their peris- tomes [14] and micronuclei remain compact [13]. Therefore, full conjugation is probably not neces- sary for the maintenance of this RNA synthesis. Duration of RNA synthesis in conjugation of S. RNA Synthesis during Conjugation 97 pustulata is different from the case of S. myztilus. In the latter, synthesis of stable mRNA regulating cytological events during conjugation occurs from 5 hr to 6 hr after the onset of pairing [3, 4]. Two cells of a pair loosely unite during the first 5 hours of their union. If these pairs are split, the split cells return to the vegetative stage. In the following 1 hr, the pairs become firmly attached. If the tight pairs are separated forcibly, the split cells undergo autogamy. In the Shr to 6hr following pair formation, micronuclei are slightly elongated [3]. In both species of Stylonychia, RNA synthesis of old macronuclei was inactive in conjugants after PM and in exconjugants. On the contrary, old macronuclei actively undergo RNA synthesis in Blepharisma musculus [16], Euplotes woodruffi [17], P. aurelia [18] and T. thermophila [19]. New macronuclear anlage with polytene chromosomes in S. pustulata may slightly synthesize RNA as the case of S. mytilus [3, 20]. In the latter, RNA synthesis is supported by the fact that ribonuc- leoprotein particles are present in the chromo- somes and the karyolymph of macronuclear anlagen in exconjugant [21]. However, it has been reported in Oxytricha by Spear and Lauth [23] and in S. mytilus by Ammermann [24] that polytene chromosomes scarcely incorporate RNA precur- sors. These results may be caused by the low incorporation of labeled RNA precursors from exconjugant-type cytostome into the exconjugant. In B. musculus [16], P. aurelia [25] and T. thermo- phila |6, 19], macronuclear anlagen which have no polytene chromosomes actively synthesize RNA from the early stage of development. Relationship between RNA synthesis and protein synthesis at the early period of conjugation In S. pustulata compact micronuclei swelled and the swollen micronuclei entered the middle or late parachute stage in the presence of AmD (Fig. 2). The parachute stage corresponds to the zygotene stage in meiosis of other organisms [2]. As shown in Figure 2, on the other hand, Cycloheximide (Chx) also blocked the progress of meiosis if it was added at the early stage of MI [11], suggesting that proteins which were synthesized during this period were needed for meiosis. Our present data suggest that RNA synthesized from the onset of pairing to the early parachute stage may be responsible for protein synthesis. Similar synthesis of RNA and proteins in the early period of meiosis has also been observed in T. thermophila [5]. In T. thermo- phila the synthesis of conjugation-specific mRNA reaches the maximum level just prior to crescent stage (pachytene), while conjugation-specific pro- teins are actively synthesized during meiotic prophase. In the male meiocytes of Lilium, RNA is synthesized at the meiotic prophase [26, 27]. Moreover, three meiotic-specific proteins partici- pating in crossing-over are activated at zygotene or pachytene [28]. Thus, syntheses of RNA and proteins from zygotene to pachytene may be com- mon phenomena in various organisms including ciliates. Differences of the effects of AmD and Chx at stages III to IV In S. pustulata, there is a difference between the effect of AmD and that of Chx. Micronuclei of the early parachute stage entered the late parachute stage in the presence of AmD, but the micronuclei entered prometaphase I or metaphase I in the presence of Chx. Thus, the progress of meiosis from the early parachute stage (stage III) to prom- etaphase I (stage VI) may required RNA synthesis rather than protein synthesis during this period. Since chromosomes were longer than those of metaphase I and often adhered each other, RNA may be involved in chromosomal behavior such as crossing-over or segregation of chromosomes. However, there is no evidence showing a sufficient deposition of silver grains on micronuclei of stage V or VI. Thus, that protein synthesis is not blocked by the drop of permeability of Chx into the conjugants during stage V or VI cannot be excluded. On the other hand, in the meiocytes of Lilium, the progress of meiosis immediately ceases by blockage of RNA synthesis at pachytene [26]. Small nuclear RNA which is synthesized at zygotene-pachytene regulates nuclease accessiblity in specific regions of meiotic cells [29]. RNA synthesis at the late period of conjugation Some differences were observed between the effects of AmD and Chx at the late period of conjugation (Fig. 2). If the pairs were treated with 98 J. YANO AND M. SUHAMA AmD at various stages during 9-14hr after the onset of pairing, AmD blocked either the nuclear events at ana- or telophase (segretaion of chromo- somes) of PM and PZII or the formation of a synkaryon. If AmD treatment was begun just before each stage, these nuclear events would not be affected. Thus, failure of segregation of chromosomes would not be caused directly by AmD, but by the inhibition of RNA synthesis of AmD. On the other hand Chx stopped nuclear events at nuclear differentiation or prophase of PM, PD and PZII [11]. Therefore, RNA for chromosomal segregation of PM and PZII would be synthesized later than RNA for nuclear dif- ferentiation and prophase of PM, PD and PZII. Except for nuclear division of PZII, nuclear events of PM, PD, SK and PZI may be controlled by RNA synthesized during meiosis I and II. If AmD treatment was begun at 12-14 hr, two nuclei produced by PZI were blocked at ana- or telophase of PZII. One of them was swollen like a macronuclear anlage. This result implies that one of 2 nucleus blocked at PZII can receive a signal for macronuclear development, though polytene chromosomes are not developed. On the other hand, the separation of the pair and the degenera- tion of old macronuclei safely proceeded in the presence of AmD. If pairs were treated with AmD or Chx at 15-19 hr, the development of macronuc- lear anlage was not affected by AmD, while it was stopped by Chx. These observations indicate that RNA necessary for cytological events in the late stage of conjugation is mainly synthesized until MII. Finally, nuclear events and separation of the pair into exconjugants during conjugation of S. pustulata require RNA synthesized sequentially and mainly by macronuclei from the onset of pairing to MII. RNA for meiosis is synthesized during meiotic prophase. The nuclear events in the late period of conjugation require RNA mainly during MI and MII. RNA may correspond to stable mRNA as suggested in S. mytilus by Sapra and Ammermann (3, 4]. Nuclear differentiation of PM, PD and MA, and prophase of PM, PD and PZII were blocked by Chx [11]. Thus, stable mRNA necessary for the progress of each stage may be transcribed just before or during the stage. 10 11 12 13 14 REFERENCES Raikov, I. B. (1972) Nuclear phenomena during conjugation and autogamy in ciliates. In “Research in protozoology, Vol.4”. Ed by T. T. Chen, Academic Press, New York, pp. 147-202. Raikov, I. B. (1982) The protzoan nucleus. Cell Biol. Monographs, Vol. 9. Springer-Verlag, Wien, New York. Sapra, G. R. and Ammermann, D. (1973) RNA synthesis and acquisition of actinomycin D insensi- tivity during conjugation in Stylonychia mytilus. Exp. Cell Res., 78: 168-174. Sapra, G. R. and Ammermann, D. (1974) An analysis of the developmental program in relation to RNA metabolism in ciliate Styloncychia mytilus. Dev. Biol., 36: 105-112. Martindale, D. W., Allis, C. D. and Burns, P. J. (1985) RNA and protein synthesis during meiotic prophase in Tetrahymena thermophila. J. Pro- tozool., 32: 644-649. McDonald, B. B. (1973) Nucleic acid in Tetrahyme- na during vegetative growth and conjugation. In “Biology of Tetrahymena”. Ed. by A. M. Elliot, Dowden, Hachinson and Ross, Inc., Strousdburg, Pa., pp. 287-306. : Ron, A. and Horovitz, O. (1977) Increased RNA synthesis during preconjugation and its effect on pair formation in Tetrahymena. Experimentia, 33: 1146-1149. Sugai, T. and Hiwatashi, K. (1974) Cytologic and autoradiographic study of the micronucleus at meio- tic phase in Tetrahymerna pyriformis. J. Protozool., 21: 542-548. Woodard, J., Woodard, M., Gelber, B. and Swift, H. (1966) Cytochemical studies of conjugation in Paramecium aurelia. Exp. Cell Res., 41: 55-63. Yano, J. (1986) Effects of actinomycin D and cycloheximide on pair formation and conjugant fu- sion in Stylonychia pustulata (Ciliophora). J. Sci. Hiroshima Univ., Ser. B, Div. 1, 32: 255-269. Yano, J. and Suhama, M. (1990) Effects of cyc- loheximide on meiosis and other nuclear changes during conjugation in the ciliate Stylonychia pustula- ta. J. Sci. Hiroshima Univ., Ser B, Div. 1, 34: 19-28. Yano, J. (1985) Characters of progeny clones from conjugant fusion of Stylonychia pustulata (ciliophora). J. Sci. Hiroshima Univ., Ser. B, Div. 1, 32: 209-211. Yano, J. (1985) Mating types and conjugant fusion with macronuclear union in Stylonychia pustulata (Ciliphora). J. Sci. Hiroshima Univ., Ser. B, Div. 1, 32: 157-175. Yano, J. (1985) Degeneration of the cortical organelles and nuclear changes in the split members from the early conjugating pairs in Stylonychia pus- 15 16 17 18 19 20 ay 22 RNA Synthesis during Conjugation 99 tulata (Ciliophora). J. Sci. Hiroshima Univ., Ser. B, Dived 32> 193-207. Prescott, D. M. Rao, M. V. N., Evenson, D. P., Stone, G. E. and Thrasher, J. D. (1966) Isolation of single nuclei and mass preparation of neuclei from several cell types. In “Method in cell physiology”, Vol. 2. Ed. by D. M. Prescott, Academic Press, New York, pp. 131-142. Dass, C. M., Sapra, G. R. and Kumar, R. (1982) Fine structure and nucleic acid synthesis during macronuclear development in Blepharisma muscu- lus var. seshachari. Arch. Protistenk., 126: 293-308. Rao, M. V. N. (1968) Macronucleus development in Euplotes woodruffi following conjugation. Exp. Cell Res., 49: 107-120. Kimball, R. F. and Perdue, S. W. (1964) Synthesis of RNA by fragments of the old macronucleus in Paramecium aurelia undergoing autogamy. J. Pro- tozool., 11 (suppl): 33. Wenkert, B. and Allis, C. D. (1984) Timing of the appearance of macronuclear-specific histone variant hvl and gene expression in developing new macro- nuclei of Tetrahymena thermophila. J. Cell Biol. 98: 2107-2117. Alonso, P. and Jareno, M. (1974) Incorporacion de uridina-H’ en el esbozo macronuclear de Stylony- chia mytilus. Microbiol. Esp., 27: 199-211. Gil, R. (1976) Possible nuclear-cytoplasmic trans- fers in the exconjugants of Stylonychia mytilus: ultrastructual observation. Microbiol. Esp., 29: 125-134. Rao, M. V. N. and Ammermann, D. (1970) 23 24 25 26 Di 28 29 Polytene chromosomes and nucleic acid metabolism during macronuclear development in Euplotes. Chromosoma, 29: 246-254. Spear, B. B. and Lauth, M. R. (1976). Polytene chromosomes of Oxytricha: biochemical and mor- phological changes during macronuclear develop- ment in a ciliated protozoan. Chromosoma, 54: 1- 3? Ammermann, D. (1968) Synthese und Abbau der Nucleinauren wahrend der Entwicklung des Macro- nucleus von Stylonychia mytilus (Protozoa, Ciliata). Chromosoma, 25: 107-20. Berger, J. D. (1973) Nuclear differentiation and nucleic acid synthesis in well-fed exconjugants of Paramecium. Chromosoma, 42: 247-268. Hotta, Y. (1971). Meiosis. In “Systematic cytology, Vol 5 (in Japanese)”. Ed. by K. Ogawa, Y. Oda, K. Kurozumi and S. Sugino, Asakura, Tokyo, pp. 74- NOS: Hotta, Y. and Stern, H. (1963) Syntheses of mes- senger-like ribonucleic acid and protein during meiosis in isolated cells of Trillium erectum. J. Cell Biol., 19: 45-58. Stern, H. (1981) Chromosome organization and DNA metabolism in meiotic cells. In “Chromosome today”, Vol. 7. Ed. by M. D. Bennett, M. Bobrow and G. Hewitt, George Allen and Unwin, London, pp. 94-104. Hotta, Y. and Stern, H. (1983) Small nuclear RNA molecules that regulate nuclease accessibility in spe- cific chromatin regions of meiotic cells. Cell, 27: 309-319. est , etait era pele? es i , : ~ \ - @ al ‘a ie on icaicemabinee aon oes wa) . j : 4 my « = 7 me aman ater Ame oe pelea es Pe bens.» Re Ra oe yh. aiaape + ‘- ee } i 3 pf xe at \ 15 1 i f if . ah, ng | . o . = ee yi : ‘ ee ana GE aioe bie inne Ce aan Yo ieee t- ‘ het oe sedan” { eS} gate see bray Seni: 4 ae eat, pee: in Sie) donne agate Wie oe Bes souls wanelies ahi ‘ ie aS ee e's er irs heap 7 teeny 2 soruge aan | | 2 oe fa ir oe it ‘$0 rcs at: ez oe Me, Ven? pi * ‘ x ae ‘ 0 Rient Hea ; a ee Pe oe Pa * pete ee ; - 4 =A 4 ; - a “sR hi at i ‘ ; “ane 7 has peo ioe ho a aa “We te “ % Sani 1S yh ie ' i¥ baa ae Reese Ms ees ¥ z ~ I, mee * x te . ad A es 3 ‘ i : jhe ‘ - ae ad Pos ; ; = Te, ; ites iy r ‘ Vane ‘ o i 3 ‘ i = f ¢ ane 4 3 eer ¥ . ait x = i r “ % 9 5 7 i i ig : 4 ‘ : . . 5 Fue a « ns. . ; %y 3 , i : = iar Pl 1 5 , 7 Re t K ‘ on i ” 1 4 a : Fs = 5 ) + 3 F ‘ 4 * ‘ F 2 : j “ t Nea i + nr ’ J “ one, rr By ‘ ; ZOOLOGICAL SCIENCE 9: 101-111 (1992) Autogamy and Autogamy Inheritance in Euplotes woodruffi Syngen 1 (Ciliophora) TOSHIKAZU KOSAKA Zoological Institute, Faculty of Science, Hiroshima University, Higashi-Hiroshima 724, Japan ABSTRACT— In the Euplotes woodruffi complex which consists of syngen 1, syngen 2, syngen 3 and the autogamy group, autogamy is known in all of the stocks belonging to syngen 2 and the autogamy group but not in syngens 1 and 3. However, a recent study has revealed that one exceptional stock of syngen 1 is endowed with autogamy ability. Stock KB-16 which was collected from brackish water could undergo both autogamy and conjugation. The viability of exautogamous cells was generally over 70%. The length of autogamy immaturity was between 43 and 65 cell divisions in the exautogamous clones. Through autogamy, the stock produced viable Fl, F2, and F3 exautogamous clones with autogamy ability. Since non-autogamous clones have never been obtained among them, the parental stock is considered to be homozygous for the autogamy trait. All exautogamous clones derived from the autogamy stock could also conjugate with complementary mating type stocks of syngen 1. When the stock was crossed with a doublet clone from non-autogamous stock, all of the singlet and doublet F1 exconjugants from the cross showed autogamy ability. Thus, autogamy ability is hereditary also through conjugation. Because the results showed that autogamy occurred in heterozygotes and the heterozy- gotes produced progenies with and without the autogamy trait at a ratio of 3:1 through the next autogamy, the trait is considered to be a dominant allele. To study the features of autogamy in this stock is very important in clarifying the evolution of sexual reproduction in the E. woodruffi complex as well as the evolution of sexual reproduction in ciliates. From the results, a possible evolutionary relationship © 1992 Zoological Society of Japan of the E. woodruffi complex is discussed. INTRODUCTION Sexual reproduction in ciliates occurs either as conjugation between two individuals or by auto- gamy of single individuals [1-3]. Although con- jugation is a widespread method of fertilization in ciliates, autogamy which is considered to be a specialized mode of conjugation has only been described in a relatively limited number species including the Paramecium aurelia complex [4-6], P. polycaryum [7], P. jenningsi [8], Tetrahymena rostrata [9], Frontonia leucas {10], Euplotes minuta [11-14], E. crassus [15-17], E. woodruffi [18-25], Paraurostyla weissei [26] and Aspidisca costata [27]. Of the ciliate species with autogamy ability, the Paramecium aurelia complex, P. polycaryum, Tet- rahymena rostrata, Frontonia leucas, Paraurostyla Accepted September 25, 1991 Received April 1, 1991 weissei and Aspidisca costata are freshwater dwel- lers, while Euplotes minuta and E. crassus live in sea water. In the E. woodruffi complex, however, three syngens and the autogamy group show differ- ence in habitats and form of sexual reproduction: syngen | undergoes conjugation and lives in brack- ish water to sea water [28]; syngen 2 has both conjugation and autogamy abilities and lives in freshwater, but only autogamy is effective in main- taining individuals of the syngen [20-23]; syngen 3 undergoes conjugation, especially intraclonal con- jugation or selfing, and is a freshwater dweller [29]; the autogamy group is distributed widely in freshwater rivers or lakes and has only autogamy ability [18, 19, 30]. These findings also indicate that there is a very close relationship between habitats of species (syngens and group) and their form of sexual reproduction. One stock with autogamy ability has recently been found in studies conducted during the last 15 years on sexual reproduction of many stocks of 102 T. KosSAKA syngen 1 of Euplotes woodruffi. Heredity of the autogamy trait in the E. woodruffi complex has never been analyzed because autogamy but not conjugation is the only effective way of sexual repoduction in the stocks of the autogamy group and syngen 2 in the E. woodruffi complex. In this paper, using this autogamous stock of syngen 1, several features of the stock and its descendants are reported, such as viability of exautogamous and exconjugant cells, length of autogamy im- maturity and inheritance of the autogamy trait. In the discussion, the possible evolutionary process of the E. woodruffi complex is considered based upon the form of sexual reproduction. MATERIALS AND METHODS Stocks and culture conditions Twenty-nine stocks of the hypotrichous ciliate Euplotes woodruffi GAW syngen 1, collected from Ariake Bay in Saga Prefecture in Japan, were used in the experiments. In addition to these stocks, five stocks of complementary mating types of the same syngen were used: stock MY-1 belonging to mating type I, MY-2 to II, ZN-3 to III, MJ-2 to V and AT-7 to VII. Two doublet strains which were experimentally produced by treatment with colchi- cine [31] were also employed. All stocks and descendant clones were cultured at 20-21°C in a medium containing 9 parts of a wheat grain infu- sion (40-50 wheat grains per liter of distilled water, boiled for 10 min) and 1 part of Van’t Hoff artificial sea water, and were fed with the colorless flagellate Chilomonas paramecium cultivated in the same medium. In testing the segregation of the autogamy trait, each of the F2 exautogamous clones was cultured in a plastic Petri dish 3.5 cm in diameter and was transferred into new culture medium every week for over 4 months. Tests for autogamy immaturity To measure autogamy immaturity of descendant clones in terms of the number of fissions from the beginning of a new clonal life cycle (i.e., autogamy or conjugation), all of the clones were cultured by the 2-day period isolation method [29]. To deter- mine the length of the immature period for auto- gamy, the left-over cultures of the clones were used. To induce autogamy, cells of the left-over cul- tures were transferred to 10% sea water without food organisms and were starved for one day. Then a small amount of the food organism, in- adequate to give rise to more cell divisions, was added to the culture. This method was more effective in inducing autogamy than the stimulus of starvation alone and was slightly more complicated than that used in earlier work [18-23, 25, 30], because the induction of autogamy in syngen 1 seemed to be more difficult, especially in young ages, than in the autogamy group and syngen 2 of this Euplotes species. Statistical analysis Statistical analysis was made by y°-test. RESULTS Isolation of the stock with autogamy ability Twenty-nine stocks which were collected from water samples of 5.5 to 10.5%o salinity in Ariake Bay, Japan, were cultured by the 2-day period isolation method and were examined in detail for their sexual reproduction and syngen to which they belonged. All the stocks belonged to Euplotes woodruffi syngen 1 and underwent conjugation when complementary mating type stock was added to any one culture of the stocks. In addition to conjugation, one exceptional stock, stock KB-16, showed autogamy ability. Of 600 or more stocks of E. woodruffi syngen 1 collected over a 15 year period, stock KB-16 was the first stock with auto- gamy ability found in this syngen. Autogamy trait and viability of the exautogamous cells Of 50 exautogamous cells that were isolated from the culture of stock KB-16 when cells of the stock underwent autogamy, 46 exautogamous cells survived and grew into clones. These 46 Fl clones were cultured by the isolation method for about 2 months to test whether non-autogamous clones appeared in the Fl clones. All Fl clones have Autogamy Inheritance in Euplotes 103 TABLE 1. Viability of exautogamous cells ee Exautogamous cells Eh Viability alive dead Ze) KB-16(Parent) 46 4 50 D2 Fl-1 40 10 50 80 F1-2 34 16 50 68 F1-3 42 8 50 84 F1-4 45 >) 50 90 F2-1 19 1 20 95 F2-2 4 16 20 20 F2-3 15 5 20 is F2-4 12 8 20 60 F2-5 20 0 20 100 F2-6 15 5 20 i F2-7 19 1 20 95 F2-8 19 1 20 95 F2-9 19 1 20 95 F2-10 17 3 20 85 F2-11 20 0 20 100 F2-12 12 8 20 60 F1 and F2 clones were derived from parental stock KB-16 and F1 clones, respectively, through auto- gamy. autogamy ability. F2 and F3 clones derived through autogamy from the F1 clones were also examined, as mentioned above. All 161 F2 ex- autogamous clones and all 191 F3 exautogamous clones also underwent autogamy. These results suggest that stock KB-16 and its descendant clones are homozygous for the autogamy trait. The viability of exautogamous cells is shown in Table 1. The viability was high in F1, F2 and F3 clones, with only one exceptional case. These results show that autogamy is fully effective as a mode of sexual reproduction in maintaining the autogamous strains. Autogamy immaturity and the relationship between frequency of autogamy and clonal age The length of autogamy immaturity, which im- plies immaturity for autogamy in this case, was examined. Twenty-two exautogamous cells were examined for the length of autogamy immaturity by the isolation method. The length of autogamy immaturity was between 43 and 65 cell divisions. The frequency of exautogamous cells was 1-8% when autogamy first occurred in the left-over cultures of the clones. These results show that there is a period of autogamy immaturity which varies from clone to clone, and that the frequency of exautogamous cells is low at a young age. The relationship between the frequency of ex- autogamous cells and clonal age was examined. Of the 22 clones, 12 were employed by using the isolation method for more 2 months. The reuslts (Figure 1) showed that the frequency does not increase with clonal aging. Even at the age of 110 cell divisions, no clones reached 100% of exauto- gamous cells in the left-over cultures. The results are very different from those of the autogamy group and syngen 2 of this Euplotes species. Conjugation between autogamous and_ non- autogamous stocks Cells mature for autogamy could usually conju- gate with cells of complementary mating type. Although the cause is still unknown, cells of non- autogamous (a ) stocks induced some cells of autogamous (a~) stock to undergo autogamy, and a* cells often brought about selfing in a_ stocks which usually did not undergo selfing. Heterotypic pairs between a’ and a cells were naturally formed at a frequency of 50% or less. When the viability of exconjugant cells was tested, conjugat- ing pairs were obtained from the union of one a* cell and one a cell. Pairs obtained from this method should be truly heterotypic. Three a’ clones (about the age of 100 cell divisions) and5a_ mature stocks were used in the crosses. The viability of the exconjugants were 75-100% (Table 2). The high viability of the exconjugants suggests that conjugation as well as autogamy is effective in maintaining a° strains. Cells mature for autogamy usually could conju- gate with a_ mature cells when a* and a cells were mixed together. However, it is unclear whether autogamy and conjugation abilities were expressed at the same time in the life cycle. To obtain heterotypic pairs more easily, a new method was used in which a” singlet cells and a~ doublet cells were mixed together. The doublets used here were bimicronucleates which consisted of two complete singlet cells. Any conjugating pair 104 T. KoSAKA (*/o) 0 40 50 60 70 80 90 100 110 (°lo) 40 ° e 2 20 e 20 e e ee iz e 10 oe e 0 ERIN The GEES OEY. DE SMIEVAD & BROLIN EE 0 40 50 60 70 80 90 100 110 (%e) 40 e 30 3 20 is e ee e tJ 10 ° z 3 5 0 e e e 5 e 0 40 50 60 70 80 90 100 110 (*e) 40 ‘ és e e 4 e 20 e 10 ee? C 0 Z ® eo ¥ ee ° ° 0 40 50 60 70 80 90 100 110 (*le) 40 e 30 Bes 5 56 5 ee ° 10 e e ee e 0) a 0 40 50 60 70 80 90 100 110 (*lo) 70 60 ° v 50 5 6 40 ae 30 v 20 e e d e? O 10 e e 0 e e s 10} 40 50 60 70 80 90 100 110 (°lo) 0 40 50 60 70 80 90 100 110 (Ie) 40 5 30 ° = e 20 e ote 10 e ° e é ° 0) FFF aaa SS (0) 40 50 60 70 80 90 100 110 (6) 40 50 60 70 80 90 100 110 (lo) 50 e 40 e e 1@ do : Ono e 20 ° e ee 10 .° 5 Pat 0) —————— SS 10} 40 50 60 70 80 90 100 110 (*/o) 40 3 me oe a al Oe 20 ° 10 ee ° * Y 0 Zz eielter? e 0 40 50 60 70 80 90 100 110 (“/o) 40 12 30 20 10 : e 0 5 eo S e O (p-—_—*—_*,___2._-s-2—_-s_2_, —__#_>—__#_#_— 0 40 50 60 70 80 90 100 110 Fic. 1. Change of the percentages of exautogamous cells with increase of clonal age. Ordinate: the percentages of exautogamous cells; abscissa: the total number of fissions from clonal initiation. Twelve clones were used. 1: Clones: 2223 5-5 4:50) De O 505 7 10 os 14 ONG O19 shh 203122 between a singlet and a doublet was thus a hetero- typic pair. The viability of the exconjugants from heterotypic pairs was 73% and 50% in two matings (Table 3). The viability of singlet participants was higher than that of doublet ones: 78% versus 69% in the cross between clone 7 and doublet stock AT-7 and 63% versus 38% in the cross between clone 11 and stock AT-7. Expression of autogamy ability in the exconjugant clones Expression of autogamy ability was studied in 30 Autogamy Inheritance in Euplotes 105 TABLE 2. Viability of the exconjugants (singlet with singlet) Exconjugants Viabilit Crosses : Total y ; One alive, q Both alive Ore 27:4 ‘Both dead (%) F2-1 x AT-7(VII) 14 1 1 16 91 F2-7 xMY-2(II) 10 3 1 14 82 F2-7 x AT-7(VID) DD, D 0 24 96 F2-11 x MY-1(1) 5 5 0 10 iS) F2-11 x MY-2(II) 9 0 0 9 100 F2-11 x ZN-3(III) 7 0 0 7! 100 F2-11 x MJ-2(V) 15 0 0 15 100 F2-11 x AT-7(VI) 26 2 0 28 96 TABLE 3. Viability of the exconjugants (singlet with doublet) Exconjugants Viabilit Crosses : : ; Total y Singlet alive, Singlet dead, (%) Both alive doublet dead doublet alive Both dead F2-7 x AT-7(VII) 30 15 10 3 58 73 F2-11 x AT-7(VII) 7 8 2 v) 24 50 Stock AT-7 was a doublet strain. synclones from the cross between clone 7 and stock AT-7 (doublet strain). To test both autogamy and conjugation abilities, two doublet strains originat- ing from stocks MY-2 and AT-7 and two singiet stocks (MY-2 and AT-7) were used as testers, although the number of testers was insufficient. Singlet exconjugant clones were mixed with doub- let tester stocks, while doublet exconjugant clones were mixed with singlet tester stocks. After 2 to 3 months in the new life cycle of the clones cultured by the isolation method, 27 of 30 singlet clones showed both autogamy and conjuga- tion. These clones were considered to be hetero- zygotes for the a‘ trait, because conjugation occurred between a’ homozygotes and a homozygotes. The other 3 clones died within a month after the beginning of clonal life. On the other hand, 10 of 30 doublet clones also underwent both autogamy and conjugation. The other 20 clones changed to singlet clones by cell division or died within a month. The results show that the a* trait is due to a dominant allele because autogamy occurred in the heterozygote clones. Further, since autogamy occurred in the doublet clones, it seems likely that reciprocal exchange of gamete nuclei occurred normally. Doublet clones very frequently lost the micro- nucleus in one member of the doublet during the course of the isolation method and eventually became unimicronucleate doublets. The doublet clones maintained autogamy ability after becoming unimicronucleate. Doublets often produced two singlets at cell division. When unimicronucleate doublet clones produced one amicronucleate and one micronucleate singlet, the amicronucleate singlet always died, while the micronucleate sur- vived (Figure 2). Usually, the micronucleate sing- lets from the unimicronucleate doublet clones could undergo autogamy. However, micronucle- ate singlets from the unimicronuclear doublet clone 41W completely lost autogamy ability. This shows that the macronucleus in the amicronucleate component of clone 41W doublets controlled the a’ trait; however, the viability of the amicronuc- lear component was dependent on the micronuc- lear component. 106 T. KOSAKA ee — cell A B division aN loss of one micronucleus Cc Fic. 2. Formation of singlets with or without micronuc- leus from a unimicronuclear doublet with heteroka- ryon. A: a micronucleate cell which can live and reproduce normally. B: an amicronucleate cell which always dies without cell division. C: a uni- micronucleate doublet. Segregation of the a~ trait Segregation of the a‘ trait in F2 clones was tested by using heterozygous exconjugant clones (a’/a_). Autogamy was induced in 5 exconjugant clones. Fl exconjugants showed high viability in crosses between a” cells anda cells. However, the viability of F2 exautogamous cells dropped drastically when the exconjugants underwent the next autogamy. In F2 exautogamous cells derived through autogamy from the exconjugant cells, the viability was 8-30%. For example, only 762 of 2555 and 178 of 2274 exautogamous cells isolated were viable in exconjugant clone 4W and clone 21W, respectively. The exautogamous cells obtained were cultured for 4 months. If the a* trait was due to a dominant allele and autogamy occurred in the heterozygous clones, it would be expected that approximately 3 times as many a* progeny would be expressed as a_ progeny in the phenotype of the exautogamonts. The results, shown in Table 4, indicate that the a* trait segre- gated at a ratio of about3:1tothea trait through autogamy. DISCUSSION The finding of the a* stock, KB-16, in Euplotes woodruffi syngen 1 made it possible to study the heredity of the a* trait which could not be analy- zed thus far in the E. woodruffi complex, because all stocks belonging to the autogamy group and syngen 2 of the E. woodruffi complex underwent autogamy and only autogamy was the effective sexual process for rejuvenation [18-23]. In the a* stock of E. woodruffi syngen 1, both autogamy and conjugation resulted in high viability of the prog- eny. These facts thus provided advantages for analyzing the trait. The present studies show that offspring inher- ited the a* trait from their parent through auto- gamy and conjugation. Autogamy ability was expressed in heterozygotes for the a* trait and even in heterokaryon doublets whose genotypes were a /a anda‘/a andthe gene dose of a~ in the doublets was assumed to be one fourth of the totala* anda genes. Heterozygotes from a cross between a’ and a_ stock produced a ratio of 3a*:la~ progeny following autogamy. These results suggest that a* is a dominant allele of a single locus with a pair of alleles; the mode of heredity thus is very similar to that of E. minuta [11, 12] and E. crassus [17]. In the a® stock and its descendants, there was autogamy immaturity fol- lowing the previous sexual reproduction. The TABLE 4. Segregation of autogamy ability in F2 clones Parental Autogamy ability in F2 clones Fl clones with without mae a ul 4S 74 20 94 0.695 0:3 P05 4W 89 WS 114 0.573 03<—PR0:5 21S 53 16 69 0.120 O7—P< Os 21W 135 34 169 2.148 01 P02 16W aS Dal 96 0.5 0.3<P<0.5 F1 clones were derived from the cross of singlets with a* trait and doublets, and were singlets. Clones 4W, 16W and 21W were produced by cell divisions of their doublet clones. and 21S and 21W were synclones. 4S and 4W, Autogamy Inheritance in Euplotes 107 results are the same as those characteristic for autogamy immaturity of exautogamous clones in the autogamy group and syngen 2 of the E. wood- ruffi complex. Autogamy immaturity also was reported in E. minuta [11, 13] and E. crassus [16], while there was no immature period following autogamy in the Paramecium aurelia complex [32]. Autogamy immaturity may be one of the charac- teristics of the life cycle of the genus Euplotes. The frequency of autogamy did not increase remarkably with increase of clonal age in the a* stock and its descendants. The results are different from those of the autogamy group and syngen 2 of the E. woodruffi complex [19, 20, 22, 25]. The frequency of autogamy in general increases with clonal age and reaches 100% after undergoing 40 to 50 cell divisions from the first appearance of autogamy in clones of the autogamy group and syngen 2. Similar results were also reported in E. minuta [11, 13]. When a™ cells anda cells were mixed, however, more autogamous cells appeared in the mixture than in an a* clone which was placed under the stimulus of food depletion for induction of autogamy, as reported in E. minuta [13] and E. crassus [16]. These facts may indicate that the presence of complementary mating type cells, in addition to the stimulus of food depletion, promotes the induction of autogamy in the a* strains of E. woodruffi syngen 1. According to Sonneborn’s theory [32], E. wood- ruffi syngen 1 is considered to be an outbreeder based upon several features such as the character- istics of life cycle, mating types, breeding system, and distribution [28, 33]. When inbreeding occurs in an outbreedng species, inbreeding depression is caused. In E. woodruffi syngen 1, successive selfings brought about inbreeding depression [34]. Inbreeding depression has been known in other outbreeding species such as Paramecium bursaria [35, 36] and Tetrahymena thermophila [37]. When autogamy, which is another form of inbreeding, occurs in outbreeding species, it is expected that inbreeding depression would result. However, contrary to expectation, the a~ progenies in E. woodruffi syngen 1 survived through successive autogamies without inbreeding depression in the laboratory. Here, it is worth noting that the parental a* stock collected from nature was homozygous for the a* trait. These facts indicate that the a* stock, KB-16, might carry out auto- gamy, probably successive autogamies, but not conjugation in nature, and that the stock might be one of the rare a* progenies which survived the inbreeding crisis. It is possible that com- plementary mating types make cells with a™ trait undergo autogamy instead of conjugation when a* and a_ strains coexist even in nature. Fl exconjugants showed high viability in crosses between a‘ cells anda cells, while the viability of F2 exautogamous cells dropped drastically when the exconjugants underwent the next autogamy. As E. woodruffi syngen 1 is an outbreeder, it is expected that the viability of exconjugants from outcrossing is high. However, once conjugation intervened in the process of successive autogamies, inbreeding depression might be expected to occur again in the following generations. Similar observations were made in E. minuta [11, 12] and E. crassus [16]. In E. minuta, the cause of the low percentage of viability among the exautogamous F2 clones was attributed to the existence of an isolating mechanism between the dominant mating type gene VII with a* and other six mating types [11, 12]. However, this phenomenon could be equally well explained by resumption of the in- breeding depression. On the other hand, conjuga- tion between a* cells or a* cells and a cells brought about lower viability than conjugation between a cells anda _ cells in E. crassus. This indicates that a degree of speciation might be more advanced between a* and a_ strains in E. crassus than that in E. minuta and E. woodruffi syngen 1. In view of the dominance of the a* trait, why does the trait not become more common in the population of E. woodruffi syngen 1? One possible answer to this question is that autogamy, which is used as an inbreeding strategy, might be rather harmful for outbreeders which have lived for a very long time under a conservative outbreeding strategy. Dini reported that autogamy was a conditional deleterious mutation in E. minuta and E. crassus [38]. The a~ trait, which is adverse to an outbreeding economy and probably originated from mutation, might be eliminated from the population by natural selection. As the a” trait of E. woodruffi syngen 1, however, is not lethal, it 108 TABLE 5. T. KOSAKA Sexual reproduction and several features of the Euplotes woodruffi complex in Japan Syngens and Sexual Features of sexual group reproduction reproduction De Syngen 1 Conjugation -13 mating types ‘Sea, rivers, lakes containing salinity of 2%o-30%e (Autogamy) — -Selfing very rare ‘Stocks not able to be cultured for a long time ‘Inbreeding depression without salinity ‘Distribution: Honshu, Shikoku, and Kyusht dis- tricts Syngen 2 Autogamy -5 mating types ‘Ditches and lakes containing salinity of less than Conjugation -Conjugation produces 1.3%c abortive progeny ‘Stocks able to be cultured for a long time without ‘Only autogamy effective salinity -Distribution: Hokkaido and Honsht districts, and Okinawa Is. Syngen 3 Conjugation -4 mating types ‘Spring water (freshwater) ‘Selfing occurs very often _-Distribution: Kumamoto Prefecture -No inbreeding depression (very limited distribution) Autogamy Autogamy ‘No conjugation ‘Freshwater rivers and lakes group ‘No inbreeding depression -Oligosaprobic to beta mesosaprobic -Autogamy immaturity of | -Distribution: Hokkaido, Honsht, Shikoku, and more than 30 cell divisions | Kyushu districts (very wide distribution) loss of conju. ability Syngen 2 ——————+ Autogamy group 4 ee ges au tog. + conju. eutog. with au tog. multiple mating types no conju. ee ee ineffective inbreeder extreme inbreeder A Syngen 1 eg Spee nO conju., no autog. with rare selfing selfing» Syngen 3 multiple mating types conju. no autog. inbreeding depression selfing out dreeder multiple mating types inbreeder Syngen | be ee eto h—| Syngen 2 kK Syngen 3 Autogamy group je -—- RB 35 oe 20 “loo 2 ‘lec 0 “foo salinity sea water brackish water fresh water Land Sea Fic. 3. A: Possible evolutionary relationship in the Euplotes woodruffi complex based upon the form of sexual reproduction, breeding systems and other features. conju: conjugation, autog.: autogamy. B: Habitats and distribution of the E. woodruffi complex. Autogamy Inheritance in Euplotes 109 may be possible that a* strains which undergo autogamy successively in nature make a bare living in nature, although the trait will not spread through conjugation throughout a population. If conjugation between a‘ and a_ cells occurs, the following autogamy causes inbreeding depression and results in elimination of cells with a~ trait from the population. Table 5 shows the results obtained thus far con- cerning several features of sexual reproduction, number of mating types, breeding system, and distribution in the E. woodruffi complex which live widely from sea water to freshwater [18-23, 28, 29, 33, 34, 39]. It can be seen from the table that inbreeding (autogamy, selfing) is common in fresh- water syngens 2 and 3 and the autogamy group, while outbreeding (conjugation between com- plementary mating types) is typical in marine syngen 1. The EF. woodruffi complex was un- doubtedly one species until the near-past, because morphological features connot distinguish the va- rious groups or syngens from one other. However, the characteristic habitats and sexual reproduction or breeding systems are very different among members of the E. woodruffi complex. Since a situation such as is found in the E. woodruffi complex has never before been uncovered in cili- ates, it seems worthwhile to discuss the evolution of the syngens and groups in the E. woodruffi complex. Figure 3. shows possible evolutionary rela- tionships within the Euplotes woodruffi complex. The relationship presented here is based upon several facts as well as speculations: (1) As auto- gamy is a specialized form of sexual reproduction, it is presumed that the course of evolution in sexual reproduction proceeded from conjugation to autogamy. (2) In syngen 2, conjugation does not have a role in rejuvenation and maintainance of strains [20-23]. Abortive conjugation in ciliates has been known only in syngen 2 of E. woodruffi. It is logical to suspect that there may have been a drastic change from functional conjugation to abortive conjugation. As conjugation is effective for rejuvenation in syngen 1, the ancestor of syngen 2 might be attributed to syngen 1. Sonne- born has previously discussed the evolution of the Paramecium aurelia complex, proposing that out- breeders were the ancestors of current inbreeders [32]. (3) The progression from syngen 2 to the autogamy group is probable because only auto- gamy is known in the autogamy group. (4) Syngen 1, which lives widely in brackish water to sea water of 2 to 30%o in salinity, is a euryhaline species. When the ancestors of syngen 2 and syngen 3 invaded freshwater, this feature might have played an important role in adapting easily to freshwater. (5) The a~ trait must not have been profitable for outbreeders which prefer to live in seas or lakes, rather than rivers or ponds. However when some individuals with a* trait adapt to freshwater, espe- cially rivers, the autogamy ability becomes very profitable. This is because individuals often cannot find proper mates due to disturbance of the river current. (6) Heckmann proposed an endosym- biont theory relating to the evolution of freshwater Euplotes species [40-42]. According to the theory, Euplotes species with a 9 type 1 or Euplotes patella cirrus pattern evolved from sea water to freshwater by association with Omikron or Omikron-like sym- bionts. This theory supports the idea that the ancestral a* strains of syngen 1 of E. woodruffi adapted to freshwater, because the autogamy group has the symbionts [41]. (7) The dominant a’ trait might be advantageous when the ancestral strains adapted to new freshwater environments, because autogamy ability can be expressed at once following their adaptation. The possibilities that the stock with autogamy ability seemed to undergo successive autogamies prior to adaptation to fresh- water and had already conquered inbreeding de- pression are also very important. (8) Syngen 3, in which autogamous stocks have never been col- lected, lives only in a very narrow range of habi- tats. As the syngen adapted to spring water where the water current was mild, the syngen could easily find mates. All stocks of syngen 3 undergo selfing which is another way of inbreeding. The syngen is likely to have originated from ancestral selfers of syngen 1. The finding of an a” stock in E. woodruffi syngen 1 is very significant, because the very close relationship between syngen 1 and syngen 2 or the autogamy group has been supported by the finding. Whether the a* stock was the direct ancestor of syngen 2 and the autogamy group is not 110 T. KOSAKA known. However, it is presumed that a~ strains have appeared repeatedly in syngen 1, and that some of the a* strains have successfully adapted to freshwater as inbreeders by overcoming inbreed- ing depression. ACKNOWLEDGMENTS The author thanks Dr. J. A. Kloetzel of the University of Maryland Baltimore County for reviewing the manu- script and making valuable comments. REFERENCES 1 Kudo, R. R. (1966) Protozoology, 5th ed. Charles C Thomas, Springfield, Illinois, 1174 pp. 2 Grell, K. G. (1973) Protozoology. Springer-Verlag, Berlin, Heidelberg, and New York, 554 pp. 3. Raikov, I. B. (1972) Nuclear phenomena during conjugation and autogamy in ciliates. In “Research in Protozoology”. Vol. 4, Ed. by T. T. Chen, Perga- mon Press, London and New York, pp. 147-289. 4 Diller, W. F. (1934) Autogamy in Paramecium aurelia. Science, 79: 57. 5 Diller, W. F. (1936) Nuclear reorganization proces- ses in Paramecium aurelia, with descriptions of autogamy and hemixis. J. Morph., 59: 11-67. 6 Sonneborn, T. M. (1947) Recent advances in the genetics of Paramecium and Euplotes. Adv. Genet., 1: 263-358. 7 Diller, W. F. (1954) Autogamy in Paramecium polycaryum. J. Protozool., 1: 60-70. 8 Mitchell, J. B. (1962) Nuclear reorganization in Paramecium jenningsi. J. Protozool., 9 (suppl.): 26. 9 Corliss, J. O. (1952) Le cycle autogamique de Tetrahymena rostrata. Compt. Rend. Acad. Sci., 235: 399-402. 10 Devi, R. V. (1961) Autogamy in Frontonia leucas (Ehrbg.). J. Protozool., 8: 277-283. 11 Siegel, R. W. and Heckmann K. (1966) Inheritance of autogamy and the killer trait in Euplotes minuta. J. Protozool., 13: 34-38. 12 Nobili, R. (1966) Mating types and mating type inheritance in Euplotes minuta Yocom (Ciliata, Hypotrichida). J. Protozool., 13: 38-41. 13. Luporini, P. (1970) Life cycle of autogamous strains of Euplotes minuta. J. Protozool., 17: 324-328. 14 Nobili, R. and Luporini P. (1967) Maintenance of heterozygosity at the mt locus after autogamy in Euplotes minuta (Ciliata, Hypotrichida). Genet. Res., 10: 35-43. 15 Luporini, P. (1974) Killer and autogamous strains of Euplotes crassus Dujardin from the Somalian coast. Monit. Zool. Ital. n. s., suppl. 5: 129-132. 16 17 18 19 20 Ze\h ip ji) 24 75) 26 zy) 28 29 30 Luporini, P. and Dini F. (1977) The breeding system and the genetic relationship between auto- gamous and non-autogamous sympatric populations of Euplotes crassus (Dujardin) (Ciliata, Hypotrichi- da). Monit. Zool. Ital. n. s., 11: 119-154. Dini, F. and Luporini P. (1980) Genic determina- tion of the autogamy trait in the hypotrich ciliate Euplotes crassus. Genet. Res., 35: 107-119. Kosaka, T. (1970) Autogamy in fresh-water Eu- plotes woodruffi (Ciliata). Zool. Mag., 79: 302-308. (In Japanese with English abstract) Kosaka, T. (1972) Effect of autogamy on clonal aging in Euplotes woodruffi (Ciliata). Zool. Mag., 81: 184-190. (In Japanese with English abstract) Kosaka, T. (1982) Predominance of autogamy over conjugation in a new syngen in fresh-water Euplotes woodruffi (Ciliophora). J. Sci. Hiroshima Univ., Ser. B, Div. 1, 30: 111-122. Kosaka, T. (1982) Cross-incompatibility in syngen 2 of Euplotes woodruffi (Ciliophora). J. Sci. Hiroshi- ma Univ., Ser. B, Div. 1, 30: 293-302. Kosaka, T. (1982) Several features of sexual repro- duction of new stocks related closely to syngen 2 in Euplotes woodruffi (Ciliophora) with reference to their distribution. J. Sci. Hiroshima Univ., Ser. B, Div. 1, 30: 303-313. Kosaka, T. (1982) Mortal conjugation between allopatric strains in syngen 2 of Euplotes woodruffi (Ciliophora). J. Sci. Hiroshima Univ., Ser. B, Div. 1, 30: 315-322. Murakami, T. (1986) The relation between feeding and autogamy in fasted cells of freshwater Euplotes woodruffi (Ciliophora). J. Sci. Hiroshima Univ., Ser. B, Div. 1, 32: 237-243. Murakami, T. (1988) Nuclear changes during auto- gamy in amicronucleates and unimicronuclear doub- lets of freshwater Euplotes woodruffi (Ciliophora). J. Sci. Hiroshima Univ., Ser. B, Div. 1, 33: 71-85. Jerka-Dziadosz, M. and Janus, I. (1975) Discon- tinuity of cortical pattern during total conjugation of a hypotrich ciliate Paraurostyla weissei. Acta Pro- tozool., 13: 309-333. Diller, W. F. (1975) Nuclear behavior and mor- phogenetic changes in fission and conjugation of Aspidisca costata (Dujardin). J. Protozool., 22: 221- 229} Kosaka, T. (1973) Mating types of marine stocks of Euplotes woodruffi (Ciliata) in Japan. J. Sci. Hiroshima Univ., Ser. B, Div. 1, 24: 135-144. Kosaka, T. (1990) Methods for inducing selfing, selfing and its role in the life cycle of Euplotes woodruffi syngen 3 (Ciliophora). J. Protozool., 37: 33-39. Murakami, T. (1986) Nuclear change and ciliature reorganization during autogamy in freshwater Eu- plotes woodruffi (Ciliophora). J. Sci. Hiroshima 31 372 33 34 3) Autogamy Inheritance in Euplotes Wnivesseh.b. Divs 132: 225-235) Kosaka T. (1986) Doublets obtained by colchicine treatment and their several features in Euplotes woodruffi. Jap. J. Protozool., 19: 39-40. (Abstract in Japanese) Sonneborn, T. M. (1957) Breeding systems, repro- ductive methods, and species problems in Protozoa. In “The Species Problem”. Ed. by E. Mayr, A. A. A. S., Washington D. C., pp. 155-324. Kosaka, T. (1974) Age-dependent monsters or macronuclear abnormalities, the length of life, and a change in the fission rate with clonal aging in marine Euplotes woodruffi. J. Sci. Hiroshima Univ., Ser. B, Div. 1, 25: 173-189. Kosaka, T. (1982) Inbreeding depression in marine Euplotes woodruffi (Ciliophora). J. Sci. Hiroshima Univ., Ser. B, Div. 1, 30: 101-110. Jennings, H. S. (1944) Paramecium bursaria: Life history. III. Repeated conjugations in the same stock at different ages, with and without inbreeding, in relation to mortality at conjugation. J. Exp. Zool., 96: 243-273. 36 3) 38 3) 40 4] 1 Jennings, H. S. (1944) Paramecium bursaria: Life history. IV. Relation of inbreeding to mortality of ex-conjugant clones. J. Exp. Zool., 97: 165-197. Nanney, D. L. (1957) Inbreeding degeneration in Tetrahymena. Genetics, 42: 137-146. Dini, F. (1984) On the evolutionary significance of autogamy in the marine Euplotes (Ciliophora, Hypotrichida). Am. Nat., 123: 151-162. Kosaka, T. (1982) Features of exconjugant clones derived from intraconjugation and crosses of selfer by non-selfer or selfer in marine Euplotes woodruffi (Ciliophora). J. Sci. Hiroshima Univ., Ser. B, Div. 1, 30: 83-99. Heckmann, K. (1975) Omikron, ein essentieller Endosymbiont von Euplotes aediculatus. J. Pro- tozool., 22: 97-104. Heckmann, K., ten Hagen, R., Gortz, H. -D. (1983) Freshwater Euplotes species with a 9 type 1 cirrus pattern depend upon endosymbionts. J. Protozool., 30: 284-289. Heckmann, K. (1983) Endosymbionts of Euplotes. Int. Rev. Cytol., suppl. 14: 111-144. spn = sti | aio ORO, (2089 a bering ‘te baal uf nk RERRREN GbE roteclinvantine : BEL? Be eaz i) | aeh) ag rs 86 Fauety DR: x A lah fe Tit, Neb ory ae ponue alge Vorewelaa yt ao oie A “bord dieting sis Saag pale ve Reid Piedeeyn: ich Ne 7 cima OE E angi te Aathcivetnt Heat | a Test eee a bscin ok “a se jie’ ANAS oP ine ai ‘ P rye, : af stew P : 4 * a CAs, a 7 f 7 "3 = joe as =f Sa e =s 7 Pree 34) < Y Fed met es OE oles ine Me rT =e ; - — ~ z 7™ . ‘e we - s+ 3 Ls : foc be Ds “i “1 za ihe a’ xe Mn ip mor a, 1 ' ¥ mad ¥ ' E 4 r v4 - 4 Fr ars 7 mes i pa} SE mis = ° iy s _— + 3 es arf in 4 2% + Ts wg > petenaie) Ang Res er : of, ate 7 a hE = > at - a i , ° c p : =) Pee 9 a oo es i -7 hb fy "i aA - = af =) 2 = t x ZOOLOGICAL SCIENCE 9: 113-118 (1992) © 1992 Zoological Society of Japan Granulocytes and Macrophages in Amphioxus HoncwE!I ZHANG)”, ZHE HUANG’, KazuHITO YAMAGUCHP and Susumu ToMoNAGA! 'School of Allied Health Sciences, Yamaguchi University, Ube 755, Japan, "Department of Biology, Shandong University, Jinan 250100, China, and *Institute of Laboratory Animals, Yamaguchi University Shool of Medicine, Ube 755, Japan ABSTRACT—Two types of mononuclear free cells were found in amphioxus. Cell type 1 with specific granules and cytoplasmic processes was usually observed in the lumen of blood vessels. The granules contained microtubule-like structures which are commonly present in insect granulocytes. Type 1 blood cell is referred to as “granulocyte”. The presence of an endothelial lining as an independent cell type was doubtful in blood vessels. Type 2 cell, the “macrophage” possessed numerous cell processes, lysosomal granules, vesicles and multivesicular bodies. Macrophages were detected not only in the coelom but also in other body spaces, the so-called lymph spaces. These two cell-types probably effect crucial immuno-defense functions in amphioxus. INTRODUCTION Amphioxus, a cephalochordate, often regarded as sharing a common ancestry with vertebrates, is an important experimental animal model for phy- logenetic analyses in the biological sciences includ- ing comparative immunology. However, little is known about the immune system. Although some workers believe that amphioxus possesses no blood cells in its vascular system [1-4], others have observed a few erythrocytes [5] and granulocytes [6, 7]. Moreover phagocytic coelomocytes occur in the body cavity of amphioxus after injection of bacteria [3]. In addition to the presence of these cells, there is a humoral factor i.e. a hemagglutinin which has been defined by Bretting and Renwrantz [8] and De Benedictis and Capalbo [9]. As one of the basic steps necessary for extending and clar- ifying information on the immuno-defense system of amphioxus, observations on the ultrastructure of blood cells, vascular system, coeiom and so- called lymph spaces [10, 11] were performed. Correspondence should be addressed to Dr. S. Tomo- naga. Accepted October 16, 1991 Received September 13, 1991 MATERIALS AND METHODS Amphioxus Adult Amphioxus (Branchiostoma _ belcheri tsingtauense), which were caught in the Yellow Sea, were supplied by The Chinese Institute of Oceanology, Oingtau, China. The animals were divided into three groups. Transmission electron microscopy (TEM) The first was used for transmission electron microscopy by fixing them in a mixture of 2% glutaraldehyde and 2% paraformaldehyde (GA/ FA) [12] for 8 hours at room temperature. The specimens were then postfixed in 1% osmium tetroxide for 1 hour and cut transversely into 1-2 mm blocks. Following dehydration with ethanol, tissue blocks were embeded in Epon 812. Ultrathin sections, stained with uranyl acetate and lead citrate, were observed under the JEM-200CX electron microscope (Japan Electron Optics Ltd.). Scanning electron microscopy (SEM) The second group was prepared for scanning electron microscopy by fixing them in GA/FA for 8 hr. Passing through tannic acid [13] the tissues were post-fixed in 1% osmium tetroxide, dehy- 114 H. ZHANG, Z. HUANG et al. drated in acetone and dried by the critical point dry method. Observations were made under the JSM T300 scanning electron microsocpe (Japan Electron Optics Ltd.). Light microscopy (LM) The third group was fixed for light microscopy in Bouin solution for 18 hours. After washing they were dehydrated in ethanol, embedded in paraffin and cut as longitudinal or transverse serial sections and stained with hematoxylin and eosin. RESULTS The coelom and lymph spaces Amphioxus has two kinds of spaces: coelom and the so-called lymph space. These spaces, in addi- tion to blood vessels, were confirmed by SEM (Figs. 1,2). Figure 1 was cut surface of pharyngeal region and Fig. 2 was between the atriopore and the anus. Granulocytes in blood vessels The dorsal aortae, the endostylar artery, the hepatic vein and the branchial vessels were ex- amined by both LM and EM. Under LM blood vessels appeared as simple structures entirely dif- ferent from those of vertebrates. One portion of the pharyngeal vascular wall consisted of only connective tissue and the other was composed essentially of epithelial cells and a basement mem- brane. The presence of endothelial cells was not confirmed in all blood vessels. A small number of mononuclear cells which stained consistently with eosin were often observed in the lumen of blood vessels. The cells were easily observed under the SEM (Fig. 3). By TEM, mononuclear cells in the vascular Space appeared to be variable in shapes, such a round, flattened or irregular (Figs. 4 and 5). Their most prominent feature was the presence of many specific cytoplasmic granules (Figs. 4 and 5). At higher magnification, the microtubule-like struc- tures, 15-20 nm in diameter were detected in the Fic. 1, 2. Scanning electron micrographs of cross sections of the amphioxus in the pharyngeal region (Fig. 1) and between the atriopore and anus (Fig. 2). intestine, ls: lymph space, n: notochord, p: pharynx, pv: postcardinal vein. a: atrium, c: coelom, cn: central nervous system, d: dorsal aorta, 1: x 100. Amphioxus Blood Cells 115 +0: NS < ae S ee lumen of blood vessel. 8,500. Fic. 4-6. Transmission electron micrographs of granulocytes (G) in the lumen of blood vessels. Fic. 4, 5. Lower power views of the granulocytes, which have pseudopods and specific granules. material in the luminal space of the blood vessel. 8,700 (Fig. 4), x 12,300 (Fig. 5). Fic. 6. A higher power view of specific granules having microtubule-like structure (arrows). 58,400. Note dense 116 H. ZHANG, Z. HUANG et al. granules (Fig.6). In addtion to granules, the | complex, free ribosomes and mitochondria, all cytoplasm usually contained an endoplasmic re- easily observed in ultrathin sections. These cells ticulum with rather dilated cisternae, the Golgi usually possessed cytoplasmic processes. We clas- = ee Se Ww OS Fic. 7. A scanning electron micrograph of macrophages (M) in the subdermal sp Cer one of the so-called lymph spaces, of amphioxus. 4,600. EE: epidermis, ct: connective tissue, ls: lymph space, M: macrophage Fic. 8. A transmission electron micrograph of macrophages (M) in the so-called lymph space. 14,400. Sis 5 aC Amphioxus Blood Cells 117 sified them as granular leukocytes because of their morphological characteristics. Although granulo- cytes were often observed near the internal wall of blood vessels, any intercellular junctions were never detected between two cells. In some sec- tions granulocytes were also found in the central area of the blood vessel lumen. Macrophages in so-called lymph spaces and the coelom Examination of longitudinal and transverse se- rial sections revealed the presence of free mono- nuclear cells in so-called lymph spaces, such as the metapleural spaces, spaces surrounding the dorsal aorta and those between muscle fasciae, the sheath of the notochord and subdermal spaces. The presence of cells in these spaces was confirmed by SEM (Fig. 7). Under SEM, macrophages appear- ed to be irregular in shape with many small cell surface pits. By TEM macrophages had numerous small vesicles which may correspond to the small pits observed by SEM. Numerous lysosomal gra- nules, phagosomes and multivesicular bodies were also found in the cytoplasm (Fig. 8). In addition to their presence in lymph spaces numerous macrophages were also found in many other specific regions of the coelom, such as the perigonadal coelom, the endostyle coelom and branchial coelom by both LM and SEM. Howev- er, macrophages were rarely observed in the supra- pharygeal or the subchordal coelom and in the perienteric coelom. Some macrophages lined the connective tissue of the coelomic wall. Such cells usually had more electron dense cytoplasmic bodies. DISCUSSION For a long time, many biologists generally be- lieved that there were no blood cells in amphioxus [1-4], although a few authors observed them [5- 7]. In our study we found mononuclear free cells having specific cytoplasmic granules. These gra- nules had specific microtubule-like structure simi- lar to those described in insect blood cells [14-17]. These findings strongly support our assigning them the name “granulocytes”. Moller and Philpott [2] investigated the ultra- structure of amphioxus blood vessels. They re- ported the absence of circulating blood cells but there were discontinuous endothelil lining cells within blood vessels. The features of endothelial cells reveals them clearly identical to granulocytes which we describe in this report. The contradic- tion between these two descriptions seems to be due only to classiflying them as endothelial cells. It should be emphasized that these cells, the granulo- cytes according to our terminology were observed not only in peripheral areas of blood vessels but also in central areas of the lumen of vessels. Granulocytes generally had cytoplasmic pro- cesses suggesting the capacity for ameboid move- ment. Moreover, cytochemical investigation by Moller and Philpott [18] demonstrated phagocytic activity and localization of acid phosphatase within cytoplasmic granules. Their findings together with our observations strongly suggest that the term “granulocyte” is entirely useful. There is one possibility that remains to be solved. Both the endothelial cells and the blood cells may originate from a common ancestor as cells which originally distribute on the vessel wall, since the presence of discontinuous endothelial cells is known in certain blood vessels of some invertebrate species [19]. The presence of coelomocytes in amphioxus was not clear until 1982 when Rhodes et al. [3] detected them after injecting bacteria. In the present study we found macrophages in non-stimulated, normal animals. The cells were observed not only in the coelom but also in the so-called lymph spaces described by Kampmeier [10] which were thought to be discontinuous to the coelom. Macrophages in various body spaces and granulocytes in blood vessels probably contribute to the immune defense mechanisms. Tunicates, (Urochordata) are another group of protochordates having many species and a univer- sal distribution. Clearly tunicates have diverse hemocyte types in their hemolymph [20]. In our recent studies using Halocynthia roretzi, about ten cell types including granulocytes and macrophages have been identified [21, 22]. We assume that these common cell-types, granulocytes and mac- rophages, present in both species, are essential cell members of the immuno-defense system of pro- tochordates. 118 ACKNOWLEDGMENTS We thank the staff of The Chinese Institute of Oceanology, Qingtau, China for kind support in sup- plying amphioxus. We express appreciation to Professor Edwin L. Cooper, Laboratory of Comparative Immunol- ogy, University of California, Los Angeles who critically read the manuscript. We also thank Mr. Akira Kuma- kura for technical assistances. 10 REFERENCES Franz, V. (1927) Morphologie der Akranier. Ergeb. Anat. Entwicklungsgesch., 27: 464-692. Moller, P. C. and Philpott, C. W. (1973) The circulatory system of amphioxus (Branchiostoma floridae) 1. Morphology of the major vessels of the pharyngeal area. J. Morphol., 139: 389-406. Rhodes, C. P., Ratcliffe, N. A. and Rowley, A. F. (1982) Presence of coelomocytes in the primitive chordate amphioxus (Branchiostoma lanceolatum). Science, 217: 263-265. Rowley, A. F., Rhodes, C. P. and Ratcliffe, N. A. (1984) Protochordate leucocytes: a review. Zool. J. Linn. Soc., 80: 283-295. Parker, T. J., Haswell, W. A. and Foster-Cooper, C. (1949) A Text-Book of Zoology. Vol II. Macmil- lan, London, 6th ed. p. 48. Hilton, W. A. (1943) The blood of Amphioxus (Branchiostoma). J. Entomol. Zool., 35: 31-32. Welsch, U. (1975) The fine structure of the pharynx, cyrtopodocytes and digestive caecum of amphioxus (Branchiostoma lanceolatum). Symp. Zool. Soc. Lon., 36: 17-41. Bretting, H. and Renwrantz, L. (1973) Unter- suchungen von Invertebraten des Mittelmeeres auf ihren Gehalt an hamagglutinierenden Substanzen. Z. Immun. -Forsch., 145: 242-249. De Benedictis, G. and Capalbo, P. (1981) An Amphioxus lanceolatus agglutinating factor for hu- man red cells. J. Immunogenetics, 8: 225-230. Kampmeier, O. F. (1969) Evolution and Compara- tive Morphology of the Lymphatic System. Charles C Thomas Publisher, Springfield, Illinois, pp. 160- At 12 113) 14 IS 16 ie, 18 19 20 21 aD H. ZHANG, Z. HUANG et al. WB: Lankester, E. R. (1889) Contributions to the know- ledge of Amphioxus lanceolatus, Yarrell. Quart. J. Micr. Sci., 29: 365-402. Karnovsky, M. J. (1967) The ultrastructural basis of capillary permeability studied with peroxidase as a tracer. J. Cell Biol., 35: 213-236. Murakami, T. (1974) A revised tannin-osmium method for non-coated scanning electron micro- scope specimens. Arch. Histol. Jap., 36: 189-193. Baerwald, R. J. and Boush, G. M. (1970) Fine structure of the hemocytes of Periplaneta americana (Orthoptera: Blattidae) with particular reference to marginal bundles. J. Ultrastruct. Res., 31: 151-161. Devauchelle, G. (1971) Etude ultrastructurale des hémocytes du Coléoptére Melolontha melolontha (L.). J. Ultrastruct. Res., 34: 492-516. Gupta, A. P. (1979) Hemocyte types: their struc- tures, synonymies, interrelationships, and taxono- mic significance. In “Insect Hemocytes. Develop- ment, Forms, Functions, and Techniques”. ed. by A. P. Gupta, Cambridge University Press, Cam- bridge, pp. 85-127. Rowley, A. F. and Ratcliffe, N. A. (1981) Insect. In “Invertebrate Blood Cells 2”. eds. by N. A. Ratcliffe and A. F. Rowley, Academic Press, London, pp. 421-488. Moller, P. C. and Philpott, C. W. (1973)” Whe circulatory system of amphioxus (Branchiostoma floridae) 11. Uptake of exogenous proteins by en- dothelial cells. Z. Zellforsch., 143: 135-141. Casley-Smith, J. R. (1980) Comparative fine struc- ture of the microvasculature and endothelium. Adv. Microcirc., 9: 1-44. Wright, R. T. (1981) Urochordate. In “Invertebrate Blood Cells 2”. eds. by N. A. Ratcliffe and A. F. Rowley, Academic Press, London, pp. 565-626. Sawada, T., Fujikura, Y., Tomonaga, S. and Fuku- moto, T. (1991) Classification and characterization of ten hemocytes types in the tunicate Halocynthia roretzi. Zool. Sci., 8: 939-950. Zhang, H., Sawada, T., Cooper, E. L. and Tomona- ga, S. (1991) Electron microscopic analysis of tuni- cate (Halocynthia roretzi) hemocytes. Zool. Sci., 9 (In press). ZOOLOGICAL SCIENCE 9: 119-125 (1992) The Purification and Characterization of Sepiaterin Reductase from Fat Body of the Silkworm Bombyx mori 'TERUHIKO IINO, 7KENJIRO DOHKE and *Motoo Tsusuf ‘Department of General Education, Nihon University, Sakurajosui, Setagayaku, Tokyo, 156. *Biological Laboratory, Kitasato University, Sagamihara, Kanagawa, 228. ABSTRACT—The enzyme sepiapterin reductase has been purified about 68,000 fold from the fat body of silkworm larvae by several column chromatographic procedures and its properties were compared with those of the same enzyme from other organisms. The molecular weight of the enzyme is 29,000 by SDS polyacrylamide gel electrophoresis and 59,000 by Ultrogel AcA 44 gelfiltration, suggesting that the native form of the enzyme consists of two identical subunits. The Km values for sepiapterin and NADPH are 10.2 uM and 1.3 uM respectively, and the V,,,, value for sepiapterin is 21.4 ~mol/min/ mg. Sepiapterin reductase from silkworm fat body and rat erythrocytes reduce several other nonpteridine carbonyl compounds in the presence of NADPH. Distribution of sepiapterin reductase activity in the larval tissues of the silkworm was examined and found in many tissues but most of the activity was detected in fat body. Viability similar to the normal strain was observed in the silkworm © 1992 Zoological Society of Japan lemon mutant which lacks sepiapterin reductase. INTRODUCTION The enzyme sepiapterin reductase (EC 1.1.1. 153) catalyzes the reversible formation of dihydro- biopterin (L-erythro-6-[1, 2-dihydroxypropyl]-7, 8- dihydropterin) and NADP from sepiapterin (6- lactyl-7, 8-dihydropterin) and NADPH [1-3]. This enzyme catalyzes the terminal biosynthesis of tet- rahydrobiopterin (L-erythro-5, 6, 7, & tetrahydrobiopterin) from 6-pyruvoyl-5, 6, 7, 8- tetrahydropterin. According to the recent reports, this enzyme catalyzes the isomerization of 6-1'- oxo-2’-hydroxypropyl tetrahydropterin to 6-1’- hydroxy-2 -oxopropyl _tetrahydropterin. This isomerization reaction may be part of the natural pathway for tetrahydrobiopterin biosynthesis [4, 5, 6]. Tetrahydrobiopterin is a natural cofactor for the pterin-dependent aromatic amino acid hyd- roxylases [7] and for the o-alkyl glycerolipid cleav- age enzyme [8], which is involved in neurotrans- mitter biosynthesis and lipid metabolism. Recent- ly, Mahmound et al. has reported that tetrahydro- biopterin is also a cofactor for N-hydroxylation of Accepted July 29, 1991 Received June 25, 1991 arginine [9]. Although sepiapterin reductase was first discovered in silkworm lavae [10], no further work was reported on the silkworm enzyme. It was however first purified from horse liver, subse- quently from rat erythrocytes and whole brain and its physicochemical and catalytic properties were studied in detail [11-13]. The medical significance of the enzyme has lead biochemists to study it in detail from mammalian sources. However pterin metabolism may have other significant roles for the insect since they recognized near ultraviolet light as a visible color. Thus, these fluorescent sub- stances effect their behavior. In this present paper, we describe a method for the purification of sepiapterin reductase from silk- worm fat body and compare its physicochemical and catalytic properties with those of the enzyme in other organisms. MATERIALS AND METHODS Experimental animals and tissue preparation Silkworm larvae were reared routinely on mul- berry leaves at 25°C. On the 5 th day of the 5 th instar, the silkworm larvae were dissected with 120 T. Inno, K. DOHKE AND M. Tsusu£ scissors and fat bodies were pooled and kept frozen at —80°C until use. : Chemicals Sepiapterin was purchased from Dr. Schircks, Switzerland and further purified and recrystallized by the method of Sugiura et al. [14]. Deoxysepiap- terin (2-amino-4-hydroxy-6-propionyl-7, 8-dihy- dropteridine) and _ 6-acetyl-dihydropterin (2- amino-4-hydroxy-6-acetyl-7, 8-dihydropteridine) were prepared from a Drosophila melanogaster mutant sepia [14]. Sepialumazin (2, 4-dihydroxy- 6-lactyl-7, 8-dihydropteridine) was prepared by enzymatic deamination of sepiapterin [15]. Erythropterin (2-amino-4, 6-dihydroxy-7-pyruvoy- Ipteridine) was a kind gift from Dr. W. L. Gyure of Seton Hall University. Butyl Toyopearl 650S and DEAE-Toyopearl 650S were from Toyo Soda Mfg. Co. (Tokyo). Cytochrome c, myoglobin, egg albumin and bovine serum albumin were from Sigma Chemical Co. (St. Louis, Mo.). All other chemicals were reagent grade from commercial sources. Enzyme assay The standard reaction mixture for sepiapterin reductase assay contained the following compo- nents (in ~ moles): potassium phosphate buffer, pH 6.5, 250; sepiapterin 0.1; NADPH 0.25; and 2 to 1200 vl of the enzyme, in a final volume of 2.5 ml. Enzyme activity was determined at 25°C by measuring the decrease in absorbance at 420 nm and/or 340 nm by a Hitachi U-3200 spectrophoto- meter. Carbonyl reductase activity was assayed by the decrease at 340 nm (16). The distribution of sepiapterin reductase in silk- worm larvae was determined as follows; The stan- dard reaction mixture of final volume 125 ul was incubated at 37°C for 10min. The reaction was stopped by adding 20 ul of 20% TCA and the precipitate was removed by centrifugation. To the supernatant 15 wl of a solution of 1 g I, dissolved in 100 ml of a solvent of 2 g KI in 100 ml water was added, the mixture was allowed to stand for 30 min at room temperature. From the mixture, 50 ul aliquots were injected into a Whatman SCX HPLC column. The mobile phase for the column system was 10mM potassium phosphate buffer pH 7.0, pumped at a flow rate of 1.0 ml/min. Biopterin was detected on elution by its fluorescence (excita- tion at 350 nm, emission at 450 nm). Substrates and inhibitors difficult to dissolve in water were first dissolved in aqueous ethyl alcohol and then added to the reaction mixture to give a final concentration of alcohol less than 0.5%, a concentration which had no detectable effect on the catalytic activity of the enzyme. One unit of the enzyme activity was defined as that amount catalyzing the conversion of 1 «mol sepiapterin per min, using an extinction coefficient for sepiapterin of 10.4x10°M~'xcm~' at 420nm [17] and NADPH of 6.210°>M~!xcm~! at 340 nm, re- spectively. Specific activity was expressed in m units per mg protein. Protein concentrations were measured with the Bio-Rad protein assay kit using bovine gamma globulin as a standard. Molecular weight determination The molecular weight of the enzyme was esti- mated by gel filtration based on the method of Andrews [18]. A column of Ultrogel AcA 44 (1.5 x 100 cm) was equilibrated with 20 mM potassium phosphate buffer pH 6.5. Polyacrylamide gel elec- © trophoresis in the presence of SDS was carried out as described by Laemmli [19]. RESULTS AND DISCUSSION Enzyme purification All procedures were carried out at 4°C, unless otherwise stated. Step 1. Preparation of crude extract. Frozen tis- sues (500 g) were homogenized with 3 vols. of 10 mM potassium phosphate buffer pH 6.5, contain- ing ammonium sulfate 20% saturation (Buffer A). The homogenate was centrifuged at 16,000 g for 40 min. Step 2. Amonium sulfate fractionation. Solid ammonium sulfate was added to the crude extract to 40% saturation and the mixture stirred for 30 min. The precipitate was removed by centrifuga- tion at 16,000g for 10min. Additional solid ammonium sulfate was added to the supernatant to 60% saturation. The mixture was stirred for 30 min and the precipitate was collected by centri- Purification and Characterization of Silkworm Sepiapterin Reductase 121 fugation at 16,000 g for 10 min, and dissolved in 0.3 vols (frozen tissue wet wt.) of Buffer A. Any insoluble materials was discarded by centrifugation at 33,000 g for 60 min. Step 3. Butyl-Toyopearl column chromatogra- phy. The enzyme solution was applied to a col- umn of Butyl-Toyopear!l 650S (2.5 x 43 cm) equili- brated with Buffer A. After washing the column with Buffer A, the adsorbed protein were eluted with 10 mM potassium phosphate buffer, pH 6.5, containing 10% saturation of ammonium sulfate. The flow rate was maintained at 40 ml/hr and fractions of 10 ml each were collected. The active fractions were pooled. Step 4. Phynyl-Sepharose column chromatogra- phy. The enzyme solution was applied to a col- umn of Phenyl-Sepharose (1.5<26cm) equili- brated with 10mM potassium phosphate buffer, pH 6.5, containing 0.1 M ammonium sulfate. Af- ter washing the column with 100 ml of the same buffer, the column was washed again with 10 mM potassium phosphate buffer, pH 6.5, containing 40 mM ammonium sulfate until the 280 nm absorb- ance reading of the eluates had decreased to the starting value. Elution was carried out with distil- led water. The flow rate was kept at 25 ml/hr and fractions of 7 ml each were collected. Step 5. DEAE-Toyopearl column chromatogra- phy. The enzyme solution from step 4 was ad- justed by adding 100mM potassium phosphate buffer, pH 7.2, containing 50% ethyleneglycol un- til a final concentration of 20 mM potassium phos- phate buffer, pH 7.2, containing 10% ethylenegly- col was reached. This solution was applied to a column of DEAE-Toyopearl 650S (1.5 x22 cm) equilibrated with 20 mM potassium phosphate buf- fer, pH 7.2, containing 10% ethyleneglycol. After washing the column with 100ml of the same buffer, elution was carried out with a linear gra- dient of 0-0.3 M KCl in the same buffer. The flow rate was maintained at 25 ml/hr and fractions of 2 ml each were collected. The activity was usually eluted in a narrow range of molarity (approximate- ly 150 mM KCl). Step 6. Hydroxylapatite column chromatogra- phy. The enzyme solution from step 5 was ad- justed to pH 6.8 by adding KH,PO,. This solution was applied to a column of hydroxylapatite (1 x 17 cm) equilibrated with 20 mM potassium phosphate buffer, pH6.8, containing 10% ethyleneglycol. After washing the column with 60 ml of the same buffer, elution was carried out with a linear gra- dient of 20-300 mM potassium phosphate buffer, pH 6.8, containing 10% ethyleneglycol. The flow rate was maintained at 7 ml/hr and fractions of 1 ml each were collected. The active fractions were combined and adjusted to 0.1 M ammonium sul- fate by solid ammonium sulfate. This solution was applied to a short column of Phenyl-Sepharose (0.55 cm) equilibrated with 10mM potassium phosphate buffer, pH6.5, containing 0.1M ammonium sulfate. Elution was carried out with distilled water. The flow rate was maintained at 3.5 ml/hr and fractions of 0.5 ml each were col- lected. A typical purification procedure is summa- rized in Table 1. Sepiapterin reductase from the silkworm fat body was purified about 68,000 fold with a 25% recovery. TABLE 1. Purification of sepiapterin reductase from silkworm fat body Total Total Specific : Step Activity Protein Activity a ecacn (unit) (mg) (mU/mg) (%) (fold) 1. Homogenate S)o45) 15800 0.33 100 1 2. A.S. Fractionation 5.10 3410 33) OF 4.5 3. Butyl-Toyopearl SoA 151 34.5 100 105 4. Phenyl-Sepharose 5.60 20.4 ZS) 107 832 5. DEAE-Toyopearl 4.23 2.6 1630 81 4930 6. Hydroxylapatite 1.30 0.058 22400 25 68,000 A.S: Ammonium sulfate 12? T. Inno, K. DOHKE AND M. TsusuE Determination of molecular weight The molecular weight of silkworm sepiapterin reductase was estimated to be 59kDa by gel filtration on an Ultrogel AcA 44 column (Fig. 1) and 29kDa by SDS polyacrylamide gel elec- trophoresis. Therefore, the enzyme consists of two identical subunits, each of which has a molecular weight of 29kDa. The molecular weight of horse liver sepiapterin reductase was reported to be 47 kDa [11], 68 kDa (monkey liver) [20] and 55 kDa (rat brain) [13]. Rat erythrocytes sepiapterin re- ductase is composed of two identical subunits of 27.5 kDa giving the enzyme an overall molecular weight of 55kDa [12]. This value is in good agreement with the silkworm sepiapterin reduc- tase. 100 90 E 2 80 EggAl SPR (59,000) a 70 60 1 Za 4 9G WC MOLECULAR WEIGHT (10°) Fic. 1. Molecular weight determination by Ultrogel AcA 44 gel filtration. Marker protein used were: Cyt-c, cytochrome c (M. W. 12,300); Myo, myoglo- bin (M. W. 18,800); Egg Al, egg albumin (M. W. 45,000); B.S.A., bovine serum albumin (M. W. 66,000); S.P.R., Sepiapterin reductase from fat body of silkworm pH optimum and enzyme kinetics The pH optimum of this enzyme was 5.2. The enzyme was assayed in 100 mM potassium phos- phate buffer between pH values of 4.8 and 7.2, while 100 mM Tris-HCl buffer was used for values from 6.5 to 8.0. The apparent Km values for sepiapterin and NADPH were determined according to the method of Lineweaver and Burk [21]. Activity tests were carried out at substrate concentrations of 5S—40 «eM sepiapterin (60 «~M NADPH) and 10- 60 «M NADPH (40 uM sepiapterin) at pH6.5. The apparent Km values for sepiapterin and NADPH were 10.2 «M and 1.3 uM, respectively, and the Vmax value for sepiapterin was 21.4 ymol/min/mg. The Km value for sepiapterin reductase from mammalian sources was reported as 14-23 uM [11, 12, 20, 22]. Biopterin synthase, an enzyme similar to sepiapterin reductase, which also catalyzes the NADPH-dependent conversion of sepiapterin to dihydrobiopterin, was found in Drosophila melanogaster |23|. The Km value for sepiapterin of this enzyme was reported to be 63 uM. A comparison of the results obtained for Km values from these several different organisms indi- cates that the silkworm sepiapterin reductase more closely resembles the mammalian enzymes than does the Drosophila melanogaster enzyme. Substrate specificity The pteridine derivatives listed in the Table 2 were tested for their susceptibility as substrates. Deoxysepiapterin and 6-acetyl-dihydropterin are 6-sidechain analogues of sepiapterin. Sepialuma- zine is a corresponding lumazine of sepiapterin. Erythropterin was completely inactive as a sub- strate, while the above described analogues were slightly affected by the enzyme (Table 2). Sepiap- terin reductase from silkworm fat body shows a strict specificity for sepiaptern in the same manner as mammalian enzyme. On the other hand, Katoh and Sueoka reported that sepiapterin reductase from rat erythrocyte reduced some vicinal dicar- bonyl nonpteridine derivatives in the presence of NADPH [16, 24]. The silkworm enzyme was examined for this property. The results show that some dicarbonyl and a quinone. (e.g. phenylpro- TABLE 2. Substrate specificity of sepiapterin reduc- tase from silkworm fat body eee Enzyme Activity (%) Substrate 420 nm 340 nm Sepiapterin 100 100 Deoxysepiaterin 4.0 6.0 6-acetyl-7,8-dihydropterin 3.6 6.0 Sepialumazin 335 520) Erythropterin — 0.0 a Purification and Characterization of Silkworm Sepiapterin Reductase 11723} TABLE 3. Km and Vmax values of silkworm sepiap- terin reductase for dicarbonyl compounds and a quinone Km Vmax Substrate (uM) (smol/min/mg) Sepiapterin 10.2 De? 1-phenyl-1,2-propanedion 3 Urrell Benzil 7.7 =~ Menadione 25.0 S)0) The aparent values of the Km and Vmax were estimated by double reciprocal plots with NADPH. panedion, benzil and menadione) are reduced in the presence of NADPH (Table 3). This property is quite similar to that of the rat erythrocytes enzyme. Inhibitors Inhibition of silkworm sepiapterin reductase us- ing sepiapterin and 1-phenyl-1, 2-propanedione as substrates was examined with various inhibitors of general aldo-keto reductase and mammalian sepiapterin reductase (Table 4). Phenobarbital and pyrazole slightly inhibited enzyme activity. Rutin and indomethacin, which are reported in- hibitors of rat erythrocytes sepiapterin reductase and human carbonyl reductase [24, 26], caused a 50% inhibition of enzyme activity at a concentra- tion of about 504M. N-acetylserotonine and 6-carboxypterin, which are reported to be potent TABLE 4. and sepiapterin inhibitors of mammalian sepiapterin reductase [24, 25], also inhibited silkworm sepiapterin reductase activity. Dicoumarol, which inhibit rat sepiapterin reductae and human carbonyl] reductase [24, 26], was the most effective inhibitor causing 50% in- hibition of the silkworm enzyme at a concentration of 0.04 uM. Sepiapterin reductase from silkworm fat body has properties similar to those reported for mammalian sources, especially for rat erythro- cytes [24]. Tissue distribution of sepiapterin reductase in silk- worm Several tissues of the silkworm larvae were removed and washed with Ringer’s solution. The tissues were homogenized 3 vol. of 10 mM potas- sium phosphate buffer, pH 7.0. The homogenate was centrifuged at 16,000 g for 30 min. and the supernatnat was brought to 70% saturation by the addition of solid ammonium sulfate. The precipi- tate was collected by centrifugation and dissolved in 0.3 vol. of 10 mM potassium phosphate buffer and dialyzed against 100 vol. of same buffer over- night. The dialysate was used as enzyme prepara- tion. Although the enzyme activities were distri- buted in many tissues of the silkworm, most of the ‘activity was detected in fat body as shown in Table 5), Sepiapterin reductase may be an indispensable enzyme for all animals, because it plays an impor- tant role in tetrahydrobiopterin biosynthesis. The Inhibition of silkworm sepiapterin reductase with phenylpropanedione Inhibition (Is) nN Sepiapterin 1-phenyl-1,2-propanedione (uM) (uM) Pyrazole 1,000 750 Phenobarbital 760 500 Indomethacin 43 Sy) Rutin 65 44 Dicoumarol 0.04 0.04 6-carboxypterin 47 40 N-acetyl-serotonin ies 1:3 The inhibiton was examined in the standard reaction mixture for 3 min in the presence of 30 uM sepiapterin and 60 ~M NADPH at pH6.5. 124 T. Inno, K. DOHKE AND M. TsusuE enzyme is widely distributed in mammalian tissues and the specific acitivity of this enzyme is uniform- ly equal in many tissues [27]. But specific acitivity of sepiapterin reductase in the larval fat body is 20-100 fold higher than found in other tissues (Table 5). TaBLE 5. Distribution of sepiapterin reductase in the larval tissues of the silkworm Enzyme activity Tissue (pmol BP/10 min/g protein) Fat body 19.4 Malpighian tube 0.258 Midgut 0.570 Posterior silk gland 0.754 Integument IIS) Body fluid 0.026 BP: Biopterin Viability study on lemon mutant Genetic background of the mutant /Jemon which lacks sepiapterin reductase was made uniform to a normal strain (C-108) by cross mating experiment. Then we compared the viability of the mutant with the normal strain. Throughout the developmental stage no significant difference was observed in body weight, mortality and number of eggs per egg-mass. Due to the lack of tetrahydrobiopterin caused by inborn or acquired deficiencies in its synthesis, many severe diseases result, such as a atypical phynylketonurea or hyperphenylalanaemia [28, 29, 30]. Lack of sepiapterin reductase, which catalyzes terminal biosynthesis of tetrahydrobiop- terin, has not yet been found in these patients, probably because a lack of this enzyme may be lethal. On the other hand, mutant /emon has the same viability as the normal strain described above. This fact may suggest that the cofactor of aromatic amino acids hydroxylase in the silkworm can be replaced by another reduced pterin (e.g., reduced sepiapterin or tetrahydroneopterin) instead of tet- rahydrobiopterin. We think that sepiapterin reductase in silkworm larvae may have functions besides tetrahydrobiop- terin biosynthesis. It may, for instance, have a role in pigment formation in insects. Further studies should clarify these points. Sepiapterin reductase from silkworm fat body has catalytic and physiological properties similar to those reported for mammalian sources, especially for rat erythrocytes. Recently, an important fea- ture in sepiapterin reductase has been discovered. It was reported that it has a domain which can bind with pterin. Anti-pterin binding site antibodies have demonstrated the structural relatedness be- tween the pterin site in sepiapterin reductase, dihydrofolate reductase and the aromatic amino acid hydroxylase, providing strong evidence for specific structures [31]. Citron et al. has reported that a pentapeptide sequence (Ala-Gly-Leu-Leu- Ser) exists in all five types of hydroxylase (human, rat, quail and rabbit), in sepiapterin reducatase and (partially) in aldose reductase [32]. They suggested that this sequence might be specific structure of pterin binding domain. Pterin binding enzymes in insects may have the same domain. Monoclonal antibodies against silkworm sepiapter- in reductase might be useful to clarify the property of the enzyme. Further studies on the enzyme are — now being undertaken and will be published else- where. ACKNOWLEDGMENTS We thank Dr. W. L. Gyure of Seton Hall University for his kind gift of erythropterin and for his correction of English. REFERENCES 1 Matsubara, M., Katoh, S., Akino, M. and Kauf- man, S., (1966) Sepiapterin reductase. Biochim. Biophys. Acta, 122: 202-212. 2 Kaufman, S., (1967) Metabolism of the phenylala- nine hydroxylation cofactor. J. Biol. Chem., 242: 3934-3943. 3. Nagai (Matsubara), M. (1968) Studies on sepiapter- in reductase; Further characterization of the reac- tion product. Arch. Biochem. Biophys., 126: 426- 435. 4 Katoh, S and Sueoka, T. (1987) Isomerization of 6-lactoyl tetrahydropterin by sepiapterin reductase. J. Biochem. (Tokyo), 101: 275-278. 5 Katoh, S and Sueoka, T. (1988) Coenzyme stimula- tion of isomerase activity of sepiapterin reductase in 10 iL 12 13 14 15 16 17 18 Purification and Characterization of Silkworm Sepiapterin Reductase the biosynthesis of tetrahydrobiopterin. J. Biochem., (Tokyo) 103: 286-289. Sueoka, T., Hikita, H and Katoh, S. (1990) Best-fit analysis of kinetic scheme for the stepwise reduction of the “diketo” group of 6-pyruvoyl tetrahydropterin by sepiapterin reductase. In “Enzymology and Molecular Biology of Carbonyl Metabolism III” Ed. by H. Weiner, et al., Plenum Press, New York. pp. 229-239. Kaufman, S. (1971) In Advances in Enzymology Ed. by A. Meister, Vol. 35, Interscience Publishers, New York, pp. 245-319. Tietz, A., Lindberg, M. and Kennedy, E. P. (1964) A new pteridine-requiring enzyme system for the oxidation of glyceryl ethers. J. Biol. Chem., 2339: 4081-4090. Tayeh M. A. and Marletta M. A. (1989) Mac- rophage oxidation of L-arginine to nitric oxide, nitrite, and nitrate. Tetrahydrobiopterin is required as a cofactor. J. Biol. Chem., 264: 19654-19658. Matsubara, M., Tsusué, M. and Akino, M. (1963) Occurrence of two different enzymes in the silk- worm, Bombyx mori, to reduce folate and sepiapter- in. Nature, 199: 908-909. Katoh, S. (1971) Sepiapterin reductase from horse liver; Purification and properties of the enzyme. Arch. Biochem. Biophys., 146: 202-214. Sueoka, T and Katoh, S. (1982) Purification and characterization of sepiapterin reductase from rat erythrocytes. Biochim. Biophys. Acta, 717: 265- Die. Katoh, S., Sueoka, T and Yamada, S. (1983) In “Chemistry and Biology of Pteridines”. Ed. by J. A. Blair, Walter de Gruyter, Berlin, pp. 789-793. Sugiura, K., Takikawa, S., Tsusué, M and Goto, M. (1973) Isolation and characterization of a yellow pteridine from Drosophila melanogaster mutant sepia. Bull. Chem. Soc. Jap., 46: 3312-3313. Tsusué, M. (1971) Studies on sepiapterin deaminase from silkworm, Bombyx mori. Purification and some properties of the enzyme. J. Biochem., (Tokyo), 69: 781-788. Katoh, S and Sueoka, T. (1984) Sepiapterin reduc- tase exhibits a NADPH-dependent dicarbonyl re- ductase activity. Biochem. Biophys. Res. Commun., 118: 859-866. Tsusué, M and Akino, M. (1965) Yellow pterin in mutant /emon of silkworm and mutant sepia of Drosophila melanogaster. Zool. Mag., 74: 91-94. Andrews, P. (1964) Estimation of the molecular weight of protein by sephadex gelfiltration. Biochem. J., 91: 222-223. 19 20 vA 22 7B) 24 25 26 27 28 2S) 30 Sil 32 125 Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacter- iophage T,. Nature, 227: 680-685. Katoh, S and Sueoka, T. (1982) Pteridine- metabolizing enzymes of Macaca _fascicularis. Comp. Biochem. Physiol., 71B: 33-39. Lineweaver, H and Burk, D. (1934) The determina- tion of enzyme dissociation constants. J. Am. Chem. Soc., 56: 658-666. Smith, G. Y. (1987) On the role of sepiapterin reductase in the biosynthesis of tetrahydropterin. Arch. Biochem. Biophys., 255: 254-266. Fan, C. L. and Brown, G. M. (1979) Partial purification and some properties of biopterin synth- ase and dihydropterin oxidase from Drosophila melanogaster. Biochem. Genet., 17: 351-369. Sueoka, T and Katoh, S. (1985) Carbonyl reductase activity of sepiapterin reductase from rat erythro- cytes. Biochim. Biophys. Acta, 843: 193-198. Katoh, S., Sueoka, T and Yamada, S. (1982) Direct inhibition of brain sepiapterin reductase by a catecholamine and an indoleamine. Biochem. Bio- phys. Res. Commun., 105: 75-81. Wermuth, B. (1981) Purification and properties of an NADPH-dependent carbonyl reductase from hu- man brain. Relationship to prostaglandin 9- ketoreductase and xenobiotic ketone reductase. J. Biol. Chem., 256: 1206-1213. Katoh, S., Nagai, M., Nagai, Y., Fukushima, T and Akino, M (1970) In “Chemistry and Biology of Pteridines” Ed. by K. Iwai, M. Akino, M. Goto and Y. Iwanami, Int. Acad. Print. Co., Tokyo, pp 225- 234. Kaufman, S. (1976) In “Advances in Neurochemis- try” VolIf Ed. by B. W. Agranoff and M. H. Aprisin, Plenum Press, New York, pp 1-132. Danks, D. M., Bartholomé, K., Clayton, B. E., Curtius, H-Ch., Grobe, H., Kaufman, S., Leeming, R., Pfleiderer, W., Rembold, H. and Rey, F. (1978) Malignant hyperphenylalaninaemia-current status (June 1977). J. Inher. Metab. Dis., 1: 49-53. Kaufman, S. (1985) | Hyperphenylalaninaemia caused by defects in biopterin metabolism. J. Inher. Metab. Dis., 8: 20-27. Jennings, I. and Cotton, R. (1990) Structural simi- larities among enzyme pterin binding sites as de- monstrated by a monoclonal anti-idiotypic antibody. J. Biol. Chem., 265: 1885-1889. Citron, B. A., Milstien, S., Gutierrez, J. C., Levine, R. A., Yanak, B. L and Kaufman, S. (1990) Isola- tion and expression of rat liver sepiapterin reductase cDNA. Proc. Natl. Acad. Sci. USA. 87: 6436-6440. Raed ae : n - Wea 5) ar: a ie ‘ 7 &) 4 oe es ere, ah as Pte ; . els vo, ey yi ee . hf batts A > ix See ae OSEL pega a ee “data bu, ye ees ae u Kf 24 venloe ee Aine ® gy Ssh ts Sig 4 one as 5m Oey ni spite Ve Pale l S| am ; : veto Dano 259 2 < ni 2 Bee Hiei a eh “4 Pos L aes my, eee AE nes bs J 5 mad ES t aie ps re er aint hit ‘nie ' - ‘ . J / ul j Yves, % “Y e208 1 eb tthe gy Pam cunt Pigg . 7 thts nehiee trees! settee rad iek, eps, c ‘ : } aie F re < te , (Rigas BY AS £ . i. : Cc be 7 ¥ > y Ly BAS te r * Z mt Ls A Seria sth 40 y + ? x } ree. Pyke f: , il y ( al : i a ttin ma rain 9 Fae ’ : } SS (75! = . “ys id “i 4 i i . Ff i : i hid + a 5 es ; v4 Lae § f H mS - € i ¢ . - 4 : I Re f ate @ PED OPO Dab ay Mee a2 hae ce = t a yi 7 ~ r a - ’ Bus pies * F = oye © ~ r 5 cs : = was Ms S72 R eats ~ af ¢ 3 fe ss ' , 4 * y i ma = =P a ; He x it = es 5 a ’ } ¢ "7 ‘ A , ep in VE ‘ ' j . a3 a dad Te t , a 3 3 q =F y ~ ae rt * i é 4 ¥ rie k ae a 4 3 5 aie pa! i" 1 j ii ag } 2 FF a y ‘ : 4 5 ahh ; £55 bal u x) ac 4 a | 7 Bos : 7 ae f3;) a, | ri¢ = ‘ * 4 a7 = c ¥ “ L ‘ = - aa ree - t * - Med Hy cle i Fe ' oy + Qe Ws i 7 Py pee i - ys . ¥ eh 3 d 4 A Si \ is i " ’ 2 es f os + a ca ; 7 « " ‘ cS , ah “ hy ( 7 = ata £ = 4 4s bs L re )