vane what aa ana at eet yet He rh Daiwa snag Cerner aT Le ia) ” el ryan) fag a Wig i Piliporta ie AR pe pha saan wey f rnc vL) rd v nn. be his Aa NGAP Un tas Wt sil gee Md etree phi snare i ? atylanctr te Bal estates Cc “AC Lhe Carnes Mngt om itu csey cate alontachg ke ae uate ke Caudle Pye CEI ep ioe ORME EM Pan] Sphewal Tn tae U 2a try pitty Ty: ye Bi 1 Sy wae ane ae i , th f Bimal neat gan ey M Lae bauer Dy Dr H Ran ta ifs Sina SRE Aba ‘aliatiia Kane atae Doe a an Rare Cras Chon CPEMn Tres eos taiaes Faved) rf Te Me ibe onaute Pose Sigh itd Fatih ee Pea i pntge te 2 Rekyey CO hag cuter een cn total hon) try Bees henge rsbe EGE ies ig u Cae to et rete PRERME LSD ee eth Men eteh ree! nt ie sey Seria ae! Brealey y PN veut Re Tr aa ener) ant vind Mesut ¥ Lata am me “ZOOLOGICAL SCIENCE An International Journal ZOOLOGICAL SCIENCE The Official Journal of the Zoological Society of Japan Bers CMe The Zoological Society of Japan: POON LEVIS 8) Tosi) Toshin-building, Hongo 2—27-2, Bunkyo-ku, MidestnicobayastiaaTok yo) Tokyo 113, Japan. Phone (03) 3814-5675 Managing Editor: Ose USERS Narmayen (Oleie me) President: Hideo Mohri (Chiba) exssistanl Editors: : Secretary: Hideo Namiki (Tokyo) Vosm ase Lamsinima (Qlsayara) Treasurer: Makoto Okuno (Tokyo) Masaki Saket Csyane) Librarian: Masatsune Takeda (Tokyo) Akiyoshi Nida (Okayama) a Editorial Board: Howard A. Bern (Berkeley) Walter Bock (New York) Aubrey Gorbman (Seattle) Horst Grunz (Essen) Robert B. Hill (Kingston) Yukio Hiramoto (Chiba) Susumu Ishii (Tokyo) Yukiaki Kuroda (Tokyo) John M. Lawrence (Tampa) Koscak Maruyama (Chiba) Roger Milkman (Iowa) Kazuo Moriwaki (Mishima) Richard S. Nishioka (Berkeley) Chitaru Oguro (Toyama) Tokindo S. Okada (Okazaki) Andreas Oksche (Giessen) Hidemi Sato (Nagano) Mayumi Yamada (Sapporo) Ryuzo Yanagimachi (Honolulu) Hiroshi Watanabe (Tokyo) ZOOLOGICAL SCIENCE is devoted to publication of original articles, reviews and communications in the broad field of Zoology. The journal appears bimonthly. An annual volume consists of six numbers of more than 1200 pages including an issue containing abstracts of papers presented at the annual meeting of the Zoological Society of Japan. MANUSCRIPTS OFFERED FOR CONSIDERATION AND CORRESPONDENCE CONCERN- ING EDITORIAL MATTERS should be sent to: Dr. Tsuneo Yamaguchi, Managing Editor, Zoological Science, Department of Biology, Faculty of Science, Okayama University, Okayama 700, Japan, in accordance with the instructions to authors which appear in the first issue of each volume. Copies of instructions to authors will be sent upon request. SUBSCRIPTIONS. ZOOLOGICAL SCIENCE is distributed free of charge to the members, both domestic and foreign, of the Zoological Society of Japan. To non-member subscribers within Japan, it is distributed by Business Center for Academic Societies Japan, 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. © Copyright 1992 by the Zoological Society of Japan. In the U.S.A., authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by [copyright owner’s name], provided that designated fees are paid directory to Copyright Clearance Center. For those organizations that have been granted a photocopy license by CCC, a separate system of payment has been arranged. Copyright Clearance Center, Inc. 27 Congress St., Salem, MA, U.S.A. (Phone 508-744-3350; Fax 508-741-2318). Publication of Zoological Science has been supported in part by a Grant-in-Aid for Publication of Scientific Research Results from the Ministry of Education, Science and Culture, Japan. ZOOLOGICAL SCIENCE 9: 679-695 (1992) © 1992 Zoological Society of Japan REVIEW Hormonally Induced Synaptic Plasticity in the Adult Neuroendocrine Brain AKIRA MATSUMOTO Department of Anatomy, Juntendo University School of Medicine, Hongo, Tokyo 113, Japan CONTENTS MNGKOG WC HO Wee see eerie as nse nese eesaeisek tastes tetcctasiclee sires selaranistishaer sists jos aljceleisieininenintisiesinaynainslseubiepeainsleineten 679 ESHOsenicnesulatlonion synaptic plasticibyssseessesrsse-eseesare este sen dec ceceses-eeeeeercee-esecnerecesesesseceere 680 (i) Synaptic plasticity in the hypothalamus with brain lesion ...................0cceeeeceeeeecneceee eee ees 680 (ii) Synaptic plasticity in the hypothalamus without brain lesion ...................cecceeeeeeeeeeeeen ees 682 Gea) Synaptic plasticity mite VIMIN aii. fis. le osaetecGehdnceorl sane otsecnebensssentuautnssbtneeescce stencils 682 Gib) Synapticiplasticity mithe th OAG ea creeeec ee ceatein sec sensoh odes cetoecdadserdecurescdcceeceesecsecsss 683 (li-c) Synaptic plasticity during the estrous cycle ................. cee ceee cece eee eec eee eeeeeeeeeeeeseeeseeens 683 Androgenic regulation of synaptic plasticity ............. 0c. cece eee ee cece c eee ec eec eee eee eeu ereeneeeeeeeeereees 684 @ eS ynapticrplasticityaimithe:SNBccnas.cc-e heir ee pee scm once bariepct torhjans sesesepemauademeaeeceaesenciiee see 684 (ii) Plasticity of gap junctions in the SNB ............ceceeceececeeeeee eer eeneeeeteceee eee eeneeeeeeenereeeeees 685 Sex steroidal regulation of mRNA expression of structural proteins ..................cseceeeeeeeeeee eer eees 686 Gee xpressiomioige-actimvmRNA saicorecscsen crepe ceca sen eri oes sccenerinesiceessasnasherceccsnumececsranes 687 (i) aE xpressionlofocaphjunctionmmlRINAgeeescerre eee aterteeeeseesaecadeceeeseeteeeeeee cea seeee cae ceeeeae eae 688 (iii) Possible mechanisms of sex steroids for the expression of f-actin and gap junction IRIRINIANS sc nconesaboaaaedeb soba anSaeHeodndes seGaagtsepaP ental cScosadccnedr san sense Coreanadd Seas doce nama e er eee en 689 (COMGINGION '5.cooadadbaanoasbaceran Sap ROAUHBAB ENE CHGbn AGP BASAe nace Raa tc SPeUnACHE MICA SLEn tn ioe oan ana SR ener seers aren 690 IRNGRETEN CES Reeeee ereee reset Cierny are aamrae ayers rasa tail nian’ set clase aafo sili astiais aaiisieleeiisie Muir awe tinelnadnrediata sluineina 691 neuronal number [15-17], in dendritic morphology INTRODUCTION It is well established that sex steroids play a crucial role in reproductive neuroendocrine func- tions such as gonadotropin secretion and copula- tory behavior. The effects of gonadal steroid hormones on the sex steroid-accumulating neuro- nal structures have been classically subdivided into organizational and activational ones [1-3]. Estrogen or aromatizable androgen is responsible for modulating neuronal development and neuro- nal circuit formation during perinatal period [4- 10]. In certain brain regions these organizational actions of sex steroids can induce permanent sex- ual dimorphism in nuclear volume [11-14], in Received April 30, 1992 [18, 19], in neuronal membrane organization [20], in the distribution patterns of serotoninergic [21] and vasopressinergic fibers [22], in synaptic forma- tion [23-28] and in neuronal connectivity [29, 30]. In contrast, activation effects of sex steroids on adult brain tissues are considered to be imperma- nent and reversible. Although the effects have been thought to involve changes in neurophysiologi- cal and/or neurochemical events in the neuro- endocrine brain, evidence is now accumulating that even impermanent effects of sex steroids can in- volve major structural alterations in neuronal ele- ments and circuits, reflecting several kinds of re- productive neuroendocrine functions. Sex steroids have been found to regulate nuclear volumen [31, 32], neuronal number [33], neuronal size [34-36], 680 A. MAtTsumMoTo dendritic morphology [35, 37-43], vesopres- sinergic innervation patterns [44], and synaptic remodeling [45-51] in several areas of adult rodent neuroendocrine brain. It has also been reported that androgen regulates nuclear volume [52], dendritic length [53] and synaptic reformation [54, 55] in a portion of adult canary striatum. In the present paper, to clarify the activation effects of sex steroids on neuronal circuit forma- tion, the focus is placed on synaptogenic action of sex steroids on adult neuronal structures of the neuroendocrine brain such as hypothalamus and spinal cord which participate in reproductive func- tions [56-60]. ESTROGENIC REGULATION OF SYNAPTIC PLASTICITY The hypothalamic arcuate nucleus (ARCN), hypothalamic ventromedial nucleus (VMN) and preoptic area (POA), which are sex steroid- accumulating neuronal groups [61-64] are consid- ered to play a significant role in regulating gonado- tropin secretion and sexual behavior [56, 57]. Most of these functions are sex steroid-dependent and sexually dimorphic. Recently, the volume of the POA [11-14] and VMN [12], and the synaptic organization in the POA [23, 24], ARCN [25, 26] and VMN [27, 28] have been reported to be sexually dimorphic and are modified by neonatal sex steroid environment. Estrogen during neo- natal or prepubertal periods has a facilitatory effect on neuritic growth of mouse hypothalamic explants [65] or rat cultured hypothalamic cells [66, 67], dendritic spine formation in the VMN [68] and synaptogenesis in the ARCN [69-74], VMN [27, 28] and POA [75]. Since these hypothalamic neural tissues in neonatal animals are in an unde- veloped state [28, 69, 75-80], estrogen may act as growth-promoting factors on the developing hypothalamic neural components and play a criti- cal role in modulating postnatal neuronal matura- tion and neural circuit formation in the hypo- thalamus. (i) Synaptic plasticity in the hypothalamus with brain lesion Synaptogenic action of estrogen on several neu- ral tissues is not limited only at neonatal or pre- pubertal period. Estrogen has been found to have a facilitatory effect on synapse formation in the ARCN of adult [45-49] and aged female rate [47]. Matsumoto and Arai [45-47] have pointed out that estrogen has a facilitatory effect on synap- togenesis in the ARCN of adult hypothalamus. When the medial basal hypothalamus (MBH), including the ARCN, is isolated completely by means of Halasz knife (MBH-island), a remark- able reduction in the number of axodendritic shaft and spine synapses (Fig. la and b) in the ARCN of ovariectomized females with MBH-islands is caused by transection of the afferent fibers con- verging to the ARCN [81]. As shown in Table 1, the number of shaft and spine synapses decrease to 45% and 73% of the intact level, respectively. Administration of estrogen for three weeks follow- ing brain surgery can bring about a dramatic increase in the number of shaft and spine synapses in the ARCN of ovariectomized MBH-island females. Shaft synapses restore to almost 75% of the intact level, and the incidence of spine synapses is significantly greater than that in control animals. These suggest the possibility that estrogen can enhance the plasticity of intact arcu- ate neurons in adult brain, presumably stimulating not only axonal sprouting but also dendritic spine formation in intact arcuate neurons. There are a number of vacated synaptic sites in the ARCN after brain surgery. Estrogen fails to increase in the number of shaft and spine synapses in the ARCN of ovariectomized females without brain lesion (Table 1). This evidence indicates that synaptogenic potential of estrogen can only be expressed in the adult arcuate neurons when synaptic sites are vacated due to lesion-induced degeneration of axon terminals. In the ARCN, most axon terminals are associated with only one dendritic element, a shaft or a spine (Fig. la and b). In addition to these synapses, axon terminals contacting two postsynaptic elements in the plane (double synapse) are seen in the ARCN (Fig. Ic). Although the incidence of double synapses is usually very low, three types of double synapse, that is, spine-spine, spine-shaft and shaft-shaft double synapses, are recognized. Treatment of results Synaptic Plasticity in Response to Sex Steroids 681 nbd Poe One a Hf Fic. 1. (a) Axodendritic shaft synapse in the ARCN of a female rat at two months of age. « 33,000. (b) Axoden- dritic spine synapse in the ARCN of a female rat at two months of age. S= spine. 28,000. (c) Spine-spine dou- ble synapse in the ARCN of an ovar- iectomized MBH-island female rat treated with estrogen for three weeks. S=spine. 24,000. (Reproduced with permission from Ref. [46]) TABLE 1. Effect of estrogen on numbers of axodendritic shaft and spine synapses in the arcuate nucleus (ARCN) with or without complete deafferentation of the medial basal hypothalamus (Reproduced with permission from Ref. [46]) Axodendritic synapse Group ae ; Shaft synapse Spine synapse Intact controls 8 1544+91* 215+23 OVX 6 1539 + 87 207 +23 OVX+EB 6 1686 + 153 241 +37 MBH-island+ OVX 9 696+ 46 156+ 20 MBH-island+OVX-+ EB 8 1159+70 380+ 24 * Mean+S.E.M. MBH-island=complete deafferentation of the medial basal hypothalamus; OVX=ovariec- tomy; EB=treatment of estradiol benzoate. 682 A. Matsumoto TABLE 2. Effect of estrogen on numbers of double synapses in the arcuate nucleus (ARCN) with or without complete deafferentation of the medial basal hypothalamus (Reproduced with permission from Ref. [46]) Number Double synapse Guo of rats . Spine-spine Spine-shaft Shaft-shaft Intact controls 8 1.87+0.58* 3.75 +1.01 4.75+1.47 OVX 6 1.66+0.67 2.83+0.79 3.00 + 1.00 OVX+EB 6 2.83 £0.94 7-33 ste1e45 5.00+0.51 MBH-island+OVX 9 1.44+0.50 6.77 + 1.43 4.11+1.12 MBH-island+OVX+EB 8 15.00 +3.45 ES ae silk 6.00 + 1.03 * Mean+S.E.M. MBH-island=complete deafferentation of the medial basal hypothalamus, OVX=ovariectomy, EB= treatment with estradiol benzoate. estrogen causes a dramatic increase in the numbers of spine-spine and spine-shaft double synapses in the ARCN of ovariectomized MBH-island rats (Table 2). Although the total number of double synapses is very low as compared to regular types of synapses, an increase in the number of spine- spine and spine-shaft double synapses, but not of shaft-shaft double synapses, is remarkable in the deafferented ARCN of estrogen-treated females (Talbe 2). The findings provide further support for the ability of estrogen to stimulate synaptogenesis in the adlut ARCN. Collateral sproutings from the axon terminal, which have already synapsed on a shaft or a spine may not only reoccupy a vacated shaft or spine, but may also synapse with a spine newly formed by estrogen stimulation. Much remains to be clarified concerning the functional significance of this plastic response of adult neural tissues to estrogen. It seems unlikely that synaptic connections established at early developmental period persist throughout the life. In this context, it is worthy to note the coexistence of growth cones and degenerating axon terminals in the adult VMN [82], indicating that spontaneous synaptic degen- eration and synaptic remodeling can occur in adult brain. Estrogen might be responsible for such a process of synaptic reorganization. (ii) out brain lesion Synaptic plasticity in the hypothalamus with- Estrogen has been reported to exert a stimula- tory influence on synaptogenesis in the VMN [51, 83], in the lateral septum [84] and in the midbrain . central gray [85] of adult female rats even if the neural tissues are never lesioned. (ii-a) Synaptic plasticity in the VMN The VMN plays a crucial role in female sexual bahavior or lordosis, which is activated by estrogen and progesterone during the estrous cycle [57, 58]. Since the VMN contains a number of estrogen- accumulating neurons [61-64], and implants of estrogen into the VMN facilitate lordosis [58], the VMN is thought to be a direct site for estrogenic effect on lordosis. Administration of estrogen into ovariectomized female hamsters for two days in- duces a 48% increase in the dendritic length of VMN neurons compared with VMN neurons of control females [42]. The formation of dendritic spines on VMN neurons of ovariectomized female rats is also stimulated by estrogen treatment for two days [39, 40]. This estrogenic regulation of dendritic membrane area implies a concomitant change in synaptic inputs to dendrites. Ultra- structural study reveals estrogenic influence on synaptic inputs to VMN neurons [51, 83]. Accord- ing to Frankfurt and McEwen [51], axodendritic synapses cover 81% of the dendritic membrane in ovariectomized rats treated with estrogen for two days, and 53% in ovariectomized ones. The num- ber of axodendritic synapses per unit area of the VMN is 2.93+0.13 in ovariectomized rats treated with estrogen, and 1.79+0.11 in ovariectomized ones. Changes in the dendritic length, dendritic spine density and axodendritic synaptic population in the VMN have been detected as early as two Synaptic Plasticity in Response to Sex Steroids 683 days after estrogen treatment. These morpholo- gical findings suggest that the VNM might be responsive to even daily changes in estrogen level. (ii-b) Synaptic plasticity in the POA The luteinizing hormone-releasing hormone (LHRH) neurons in the POA form the final com- mon pathway regulating the gonadotropin secre- tion from the anterior pituitary [86]. The activity of LHRH neurons is considered to be regulated by intra- and extrahypothalamic neural elements as well as steroidal environment [56, 57]. Since LHRH neurons do not accumulate estrogen [87], it seems likely that other estrogen-sensitive inter- neurons in the POA transduce steroidal signals and in turn, alter the activity of LHRH neurons. In fact, LHRH-immunoreactive terminals have been reported to make synaptic contact with estrogen receptor-immunoreactive neurons in the POA [88]. Immunohistochemical ultrastructural investigations have demonstrated that LHRH neurons themselves in the POA are contacted with axon terminals containing amino acid, mono- amines and neuropeptides such as y-aminobutyric acid [89, 90], dopamine and norepinephrine [91, 92], serotonin [94], @-endorphin [92, 93], subst- ance P [95], pro-opiomelanocortin [96], adreno- corticotropic hormone [97] corticotropin-releasing factor [98], and LHRH [99-102]. According to Chen et al. [93], there is sexual dimorphism in the synaptic input to LHRH neurons in the rat POA. LHRH neurons in females receives approximately twice the number of synapses as do those of males. f-endorphin-immunoreactive terminals contribute to this dimorphism. Physiological differences in regulation of gonadotropin secretion may be reflected in a sexual dimorphic connectivity of the LHRH system. Witkin et al. [102] have demonstrated that synaptic input and glial apposition to LHRH neurons in the POA of adult rhesus monkeys is sensitive to circulating estrogen. Removal of estrogen by ovariectomy results in a significant increase in the apposition of glial processes to LHRH-immunoreacive neurons and a significant decrease in their innervation in the POA, and these changes are prevented by estrogen replace- ment. This anatomical study indicates the possibil- ity that LHRH system responds to alterations in the gonadal steroid environemtn with reversible changes in synaptic input and glial ensheathing. In relation to this finding, it is noteworthy that so- mata of LHRH-immunoreactive neurons in the POA of anestrous sheep are less innervated than non-immunoreactive cells around LHRH-im- munoreactive neurons, and that somata and de- ndrites of LHRH-immunoreactive neurons are almost entirely surrounded by glial processes [103]. Thus, natural seasonal changes in estrogen level probably influence the synaptic organization of LHRH system. (ii-c) Synaptic changes during the estrous cycle As mentioned above, estrogenic effects on changes in neuronal elements in the VMN occur two days after estrogen treatment. It seems quite possible that daily fluctuations in the level of estrogen during the estrous cycle cause significant alterations in the amount of membrane available for synaptic input and in the number of synapses. The dendritic spine density of VMN neurons of female rats at proestrus has been reported to be significant higher than that at estrus and diestrus [39]. It has not yet determined whether these changes in dendritic spines reflect changes in synaptic population in the VMN during the estrous cycle. On the other hand, Olmos et al. [104] found that the number of axosomatic synapses in the ARCN decreases at estrus when compared to other days of the estrous cycle. There is a reduc- tion of 31% in axosomatic synapses between proestrus and estrus. Because estrogen has a facilitatory effect on synaptic formation in the adult ARCN, it is reasonable to assume that synaptic remodeling of the ARCN occur during the estrous cycle. It seems likely, therefore, that the neuronal plasticity which has been demon- strated to occur during the estrous cycle may be a natural consequence of sex steroid feedback, and that the adult pattern of synaptic connections can be altered in the hypothalamus in relation to reproductive neuroendocrine and behavioral func- tions. Although the hippocampus is not considered to be directly involved in regulatory mechanisms of neuroendocrine functions, estrogen exerts a stim- 684 A. Matsumoto ulatory influence on dendritic spine formation of pyramidal neurons in the hippocampus (CAI re- gion) of ovariectomized adult rats [105]. More- over, spine density of the CA1 pyramidal neurons [106] and synaptic population in the neuropil of CA1 region alter during the estrous cycle (McEwen, B. S., personal communication). These evidence further supports the possibility that neuronal elements and synaptic connections reorganize during the estrous cycle. ANDROGENIC REGULATION OF SYNAPTIC PLASTICITY The spinal nucleus of the bulbocavernosus (SNB) of the rat lumbar spinal cord is a sexually dimorphic motoneuron pool [107-111]. The SNB motoneurons innervate the perineal striated mus- cles bulbocavernosus and levator ani that attach to the penis. They have an important role in copula- tory behavior, which is sensitive to alterations in circulating levels of androgen in adulthood [59, 60]. Both neuronal and muscular elements contain androgen receptors [107-114]. Castration of adult male rats rsults in a significant decrease in both somatic size [34, 36] and dendritic length [37, 38, 41] of these motoneurons, and androgen treatment of castrates reverses this effect. According to Kurz et al. [43], androgenic influences on somatic size and dendritic length are also detected in another androgen-sensitive motoneuron pool in the lumbar spinal cord, the dorsolateral nucleus (DLN), which innervates the ischiocavernosus and urethral sphincter muscles [113, 115]. These evidence suggests that the adult SNB and DLN motoneurons retain a great deal of plasticity to androgen, which is expressed when androgen levels fluctuate. This androgenic regulation of somatic and dendritic membrane area implies a Fic. 2. Percentage of SNB membrane covered by synaptic contacts (a), size of synaptic contacts (b) and frequency of synaptic contacts per 1 ~m of SNB neuronal membrane (c) of control (Controls), cas- trated (Castrates) and castrated male rats treated with testosterone for four weeks (Castrates + Testos- terone). Vertical lines indicate S.E.M. Numbers in the bottom of columns refer to the number of rats examined. (Reproduced with permission from Ref. [2]) concomitant change in synaptic inputs to these membranes. (i) Synaptic plasticity in the SNB Quantitative electron microscopic study reveals a | Soma | | Proximal dendrite 30+ 155 5 5 5 Controls Castrates Castrates Percent Covering of Synaptic Contacts + Testosterone b EI Soma HEE Proximal dendrite (um) 2.05 io iS) 2 r= cS) ra 9] a ° c bal a r) ao pe 7) 5 5 5 Controls Castrates Castrates + Testosterone Cc =] Soma HEE Proxima! denarite 9 ty a (o} r 0.100- Frequency of Synaptic Contacts (per jum of Neuronal Membrane) 5 5 5 Costrates Controls Castrates + Testosterone Synaptic Plasticity in Response to Sex Steroids 685 androgenic influences on synaptic inputs to SNB motoneurons [48, 50]. In intact adult males, synaptic structures cover about 40% and 47% of somatic and proximal dendritic membranes of SNB motoneurons, respectively (Fig. 2a) [48]. Four weeks after castration, these values are re- duced to about 14% and 19% of somatic and proximal dendritic membranes, respectively. However, treatment of testosterone for four weeks after castration prevents this decline. Castration and testosterone treatment also influence the synaptic size (Fig. 2b) and number of synaptic contacts per unit length of somatic and proximal dendritic membranes (Fig. 2c). These results indi- cate that androgen is critical for maintaining the reformation of synaptic inputs to SNB motoneurons in adult male rats. Together, the earlier light microscopic studies [34, 36-38, 41] and ultrastructural ones suggest that the SNB system can retain a great deal of plasticity in adulthood. Because androgen levels in rodents fluctuate with seasons [38], stress and aging [116, 117], the synaptic changes induced by castration may reflect an ongoing process of synaptic or- ganization and reorganization regulated by androgen throughout the adult life. Previous studies have suggested which kinds of neurotransmitters or neuromodulators may be contained in affectes to SNB motoneurons. Im- munohistochemical studies have suggested that the presence of norepinephrine, serotonin, substance P, leu- and met-enkephalin, thyrotropin-releasing hormone (TRH) and calcitonin gene-related pep- tide (CGRP) in fibers and terminals in close prox- imity to somata and dendrites of motoneurons in the rat SNB [118, 119], and in the cat or baboon Onuf’s nucleus which is homologous to the SNB [120-124]. These data indicate that the terminals of these fibers may make synaptic contact with SNB neurons. In fact, serotonin- and tyrosine hydroxylase-immunoreactive terminals make synaptic connection with SNB motoneurons [125]. Motoneurons in the Onuf’s nucleus are contacted with axon terminals containing norepinephrine, serotonin, substance P, leu- and met-enkephalin, TRH and CGRP [120, 124]. It has been clearly established that synaptic inputs to the SNB motoneurons can be regulated by androgenic in- fluence. Further studies are needed to determine the chemical identity of neuronal inputs to the SNB motoneurons which are affected by androgen. (ii) Plasticity of gap junctions in the SNB The SNB motoneurons exist in a cluster, and their somata come into contact with each other [107-111]. Numerous number of longitudinally oriented bundles of dendrites are packed in the neuropil of the SNB [126, 127]. Gap junctional plaques are found between these motoneurons, Fic. 3. A gap junction (arrow and inset) can be seen along the apposed membranes of a proximal den- drite (PD) of an SNB motoneuron and another SNB dendrite (D) in a testosterone-treated male rat four weeks after castration. Cholera toxin conjugated to HRP (CT-HRP) was injected into the bulbo- cavernosus muscles two day before autopsy, and the spinal cords containing SNB were dissected, proces- sed with a modified tetramethylbenzidine method for visualization of retrogradely transported CT- HRP. CT-HRP crystals (arrowheads) are found in both dendrites. x 16,400 (inset: x 175,500). (Re- produced with permission from Ref. [128]) 686 A. Matsumoto TABLE 3. Effect of androgen on the incidence and size of gap junctions between SNB motoneurons in control and experimental animals (Reproduced with permission from Ref. [128]) Number of neurons Number of gap Number - : : Length of gap containing ga t : ; Group of rats Finclone P ai junctions (nm) Controls 5 2.6+0.4* 3.4+0.4 168.8+25.2 Castrates 0.8+0.2' On On 64.74 8.1 Castrates + testosterone 2.8+0.5 3.6+0.2 154.4+19.1 * Mean+S.E.M. * P<0.001, * P<0.0005 (vs. Controls or Castrates+ testosterone). using thin-section and freeze-fracure technique [128]. The presence of gap junctions between SNB motoneurons is also ientified with electrophysiolo- gical and Lucifer yellow dye coupling techniques [129]. Translucent electron microscopic study re- veals that gap junctions show a characteristic fea- ture of pentalaminar structure (Fig. 3). Freeze- fracture replicas demonstrate plaques of in- tramembranous particles of the protoplasmic and complementary pits on the extracellular fracture faces. These particles are considered to be chan- nels of gap junctions, permitting electrical and/or metabolic coupling between motoneurons. Of the gap junctions observed in the SNB, 45% were somatodendritic, 35% dendro-dendritic and 20% somatosomatic. As shown in Table 3, the removal of testosterone by castration dramatically reduces the number and size of the junctional plaques, and these changes are prevented by testosterone re- placement. Because the size of the junctional plaques is related to the number of gap junctional channels between the cells [130], the findings that androgen regulates the total number of gap junc- tional channels would indicate that the degree of electrical and/or metabolic coupling is regulated by androgen. The enhanced electrical and/or metabolic coupling between SNB motoneurons may promote synchronous neuronal activity in this nucleus. Gap junctional plaques are also found between motoneurons in the DLN [131]. In addition to SNB motoneurons, gonadal ster- oids have been reported to regulate the size or frequency of gap junctional plaques in various kinds of cells. For example, estrogen increases gap junctions between granulosa cells of the rat ovary [132, 133], between smooth muscle cells of rodent uterus [134-139] and between neurons in the rat ARCN [140]. In contrast, removal of estrogen by ovariectomy induces an increase in the incidence of Lucifer yellow dye coupling between mag- nocellular neurons in the supraptic nucleus of the rat hypothalamus [141]. On the other hand, androgen stimulates the incidence of Lucifer yel- low dye coupling between magnocellular neurons in the paraventricular nucleus of the rat hypothala- mus [142]. Female Xenopus frogs possess dye- coupled laryngeal muscle fibers, whereas adult males do not [143]. Androgen treatment of females reduces the coupling in females. These evidence indicates that gonadal steroids regulates the number of gap junction channels in a wide variety of reproductive tissues, including the gonads, nervous system, and smooth and striated muscles. These tissues have in common the pre- sence of gonadal hormone receptors. Indeed, it become reasonable to ask whether steroid control of gap junctions and electrical coupling is a com- mon feature of all such steroid-sensitive cells. SEX STEROIDAL REGULATION OF mRNA EXPRESSION OF STRUCTURAL PROTEINS Although the molecular mechanisms underlying the neuronal plasticity in response to sex steroids are poorly understood, evidence is now accumulat- ing that sex-steroids play a significant role in regulating gene expression of structural proteins in sex steroid-sesitive neuronal tissues [144-149]. Synaptic Plasticity in Response to Sex Steroids 687 Riri ae of Bee bea) & eget teeta ra ; pate soir ea eH eRe ar ” + Ea a Aghines ones Eng re Boies se iat, ian z ea iets Fic. 4. Autoradiographic signals of £-actin mRNA on SNB motoneurons in a castrated male rat treated with testosterone for four weeks (a), and in a castrated male rat (b). X390. (Reproduced with permission from Ref. [149]) (i) Expression of B-actin mRNA Actin is one of the important cytoskeletal pro- teins that has been identified in neurons of the central nervous system [150-152]. Actin filaments are found to associate particularly with the neuro- nal membrane and postsynaptic density. As men- tioned before, androgen regulates the somatic and dendritic membrane area of SNB motoneurons and the number of synapses onto these motoneurons. It seems likely, therefore, that actin might be involved in the hormonally induced neuronal plasticity in the SNB motoneurons. In situ hybridization histochemical study using complementary DNA encoding chick /-actin re- veals androgenic influences on the expression of f-actin mRNA in SNB motoneurons of adult male rats [149]. Hybridizable G-actin mRNA are found to localized in the somata and proximal dendrites of SNB motoneurons (Fig. 4a and b). As shown in Figs. Sa, 5b and 5c, removal of androgen by castra- tion significantly reduces the neuronal area of SNB motoneurons and the expression level of -actin mRNA in SNB motoneurons, whereas these changes are prevented by testosterone treatment for four weeks. On the contrary, castration or testosterone treatment does not induce any change in the expression level of ~-actin mRNA in the motoneurons of the retrodorsolateral nucleus (RDLN) (Fig. 5a-c). The RDLN motoneurons innervating hindlimb musculature are examined, because the RDLN motoneurons are known to be androgen-insesitive [148, 149, 153]. These findings suggest that androgen exerts a regulatory influence on the f-actin mRNA expression in the SNB motoneurons. {-tubulin is known to be main component of microtubules which are another major cytoskeletal element in neuronal cells. The expression of #-tubulin mRNA in the SNB moto- nourons is also regulated by androgen (Matsumo- to, A. et al., unpublished data). Recent studies provide evidence for an increase in expression of cytoskeletal proteins such as actin, tubulin, and their mRNAs in neural plasticity processes like neuronal growth and synapse forma- tion [154-158]. In the SNB, androgen has been shown to regulate the somatic and dendritic size of motoneurons [34, 36-38, 41] and the number and size of synapses onto these motoneurons [48, 50], suggesting that f-actin participates in the under- lying mechanisms of plastic response of SNB motoneurons. Estrogen has been reported to exert a stimulatory influence on the expression of actin [159] and f£-actin mRNA in immature rat uterus [160]. These findings indicate the possibility that sex steroids generally play a significant role in the expression of G-actin genes in sex steroid-sensitive cells. 688 A. Matsumoto ‘E fs it 2 Ss sy} o = i) © Cc [e) i. > oO Zz SNB RDLN (b) IL ~ 400 oO o L oO ~ a Yc £9 S25 i= ‘op © Lez fe} i= o a iS =) z oO o SS Yn U0 is) De wo ee fo) = co) we} E 2) z RDLN Fic. 5. Neuronal area (a) and numbers of grains of B-actin mRNA per neuronal area (b) and per unit area (c) of motoneurons in the SNB and RDLN. Solid bars, control males; hatched bars, castrated males; open bars, castrated males treated with tes- tosterone. Six animals were examined in each group. Values given are the mean+S.E.M. (Repro- duced with permission from Ref. [149]) (11) Expression of gap junction mRNA Recent biochemical and molecular biological studies make possible to raise polyclonal and monoclonal antisera against several kinds of gap junction proteins (connexins) and to clone cDNAs coding for them [161-163]. It has been reported that neuronal tissues react immunohistochemically with antibody against gap junction protein of the rat liver (connexin 32) [164-169], and that the expression of gap junction mRNA is demonstrated in the neural tissues by a Northern blot analysis [170, 171] and by in situ hybridization histochemis- try using complementary DNA (cDNA) encoding connexin 32 [148, 149, 169, 172]. Androgenic influences on the expression of gap junction mRNA in the SNB motoneurons have been clarified by recent in sity hybridization his- tochemical studies using cDNA encoding connexin 32 [148, 149]. Hybridizable gap junction mRNA are localized on the somata and proximal dendrites of SNB motoneurons (Fig. 6a and b) [148]. Re- moval of androgen by castration significantly re- duces the expression level of gap junction mRNA in the SNB motoneurons, whereas this decline is prevented by testosterone treatment for two days (Fig. 7a-c). The recovered expression level is almost the same as that treated with testosterone for four weeks [149]. This evidence indicates that androgen can enhance and maintain the expression level of gap junction mRNA in the SNB motoneurons. On the contrary, castration or tes- tosterone replacement does not induce any change in the expression level of gap junction mRNA in the motoneurons of the RDLN (Fig. 7a-c). These results indicate that androgenic regulation of gap junction gene expression is rather specific to the SNB. The recovery of expression level of the gap junction mRNA is detected in the SNB motoneurons of castrates treated with testoster- one. This evevation of gap junction transcript is thought to be associated with the synthesis of gap junction protein, reflecting the formation of gap junction channels and its plaques. Indeed, we have recently pointed out that androgen increases the number and size of junctional plaques between SNB motoneurons [128]. In the uterine myomet- rial cells, an increase in the number and size of gap junctional plaques has been observed at parturi- tion [173]. Plasma levels of estrogen [174] and number of estrogen receptors in these cells [175] increase markedly, preceding the increment of gap junctions observed at parturition. Formation of gap junction plaques between these cells is en- hanced by treatment with estrogen [134-139]. In addition to these obervations, as increase in the expression of gap junction mRNA had been de- tected in these cells at parturition [176]. These Synaptic Plasticity in Response to Sex Steroids 689 ees Baas ce ees RAYS et at Ro Se ewhaces S — eh". Fic. 6. Autoradiographic signals of gap junction mRNA on SNB motoneurons in a castrated male rat treated with testosterone for two days (a), and in a castrated male rat (b). X390. (Reproduced with permission from Ref. [148]) Neuronal area( x10 °mm’) Number of grains/Neuronal area 500 250 ” Z 00) D S) [— Z SNB RDLN Number of grains/Unit area Lia Ya SNB RDLN Fic. 7. Neuronal area (a) and numbers of grains of gap junction mRNA per neuronal area (b) and per unit area (c) of motoneurons in the SNB and RDLN. Solid bars, control males; hatched bars, castrated males; open bars, castrated males treated with tes- tosterone for two dyas. Six animals were examined in each group. Values given are the mean+S.E.M. (Reproduced with permission from Ref. [148]) findings indicate that sex steroids generally exert influences on regulating the expression of gap junction genes in sex steroid-sensitive cells. (iii) Possible mechanisms of sex steroids for the expression of B-actin and gap junction mRNAs As mentioned above, removal of androgen by 690 A. MATSUMOTO castration reduces the expression of f-actin and gap junction mRNAs in SNB motoneurons and this decline is prevented by testosterone treat- ment. In contrast to the SNB motoneurons, androgen does not induce any change in the ex- pression of @-actin and gap junction rRNAs in the RDLN motoneurons. These evidence indicate there is regional difference in regulatory mecha- nisms of androgen on the ejxpression of /-actin and gap junction mRNAs in these lumber moto- neurons. According to Breedlove and Arnold [107, 113] and Jordan et al. [114], there is a regional differ- ence in the pattern of androgen accumulation by lumbar motoneurons, that is, the motoneurons in the SNB accumulate androgen more frequently and densely than those in the RDLN. This evi- dence may be correlated with the difference in androgenic influence on the expression of /-actin and gap junction mRNAs between these two motoneuron pools. It seems likely, therefore, that androgenic regulation of the expression of these genes is specific to the SNB, and that the express- ion of #-actin and gap junction genes in androgen- accumulating neurons is sensitive to androgen. The mechanism by which sex steroids influence on the expression of f-actin and gap junction genes is not yet known. The observed modulation of their mRNA expression may be attributed to al- teration in transcription rate of genes or in stability of mRNAs [177]. Steroid hormones have been shown to regulate the gene expression first by binding to their specific receptors and then by interaction of the hormone-receptor complex with specific steroid hormone responsive elements on the DNA which act as transrcriptional enhancers [178-181]. To clarify the regulatory mechanisms of the #-actin and gap junction gene expression, it will be important to identify the steroid hormone responsive elements in these genes. However, the steroid hormone responsive elements have not yet been detected in the rat Z-actin gene [182]. It is possible that there are no steroid hormone respon- sive elements in these genes. The conjecture may be associated with the idea that expression of these genes is mediated by some gene products which are under the direct influence of the androgen- receptor complex. There is another possible explanation for the regional difference in androgenic regulation of the expression of #-actin and gap junction mRNAs between SNB and RDLN motoneurons. It has been suggested that the androgenic regulation of the size of SNB motoneurons is mediated through their target muscles [183]. According to Araki et al. [184], androgen does not induce any change in neuronal size of the SNB motoneurons which have innervated androgen-insensitive soleus muscle. In contrast, the size of SNB motoneurons which have reinnervated to their androgen-sensitive target muscles is reduced by castration and enlarged by testosterone treatment. Recently, it is reported that androgen is involved in modulating the ex- pression of calcitonine gene-related peptide mRNA in SNB motoneurons through the effects on their target muscles [185]. On the other hand, in the RDLN system, Leslie et al. [153] pointed out that castration has no effect on the size of RDLN motoneurons and their target muscle, flexor digi- torum brevis. These findings indicate the possibil- ity that, in the SNB system, androgen plays a significant role in the production of muscle-derived trophic factors which can regulate the size of SNB motoneurons. It seems likely, therefore, that androgen regulates not only the neuronal size of SNB motoneurons but also the expression of /- actin and gap junction genes in them through its action on their target muscles. CONCLUSION The present paper provides ultrastructural evi- dence suggesting that sex steroids play a significant role in modulating neural circuit formation in sex steroid-sensitive areas in adult neuroendocrine brain. Electrical coupling and/or cell-to-cell com- munication via gap junctions may also be modified by sex steroid environment. Moreover, we now undertake to clarity the molecular mechanisms underlying the neurnal plasticity in response to sex steroids. Androgen exerts potent influences on modulating gene expression of structural proteins such as #-actin, 8-tubulin and gap junction in the SNB motoneurons. Expression of growth- associated protein 43 kDa (GAP-43) mRNA in the VMN is reported to be regulated by estrogen Synaptic Plasticity in Response to Sex Steroids 691 [147]. In relevant to these evidence, it is of interest to note that the expression of actin and tubulin mRNAs is facilitated by neonatal treatment with androgen in the preptic area-hypothalamus [145] and limbic system of female rats [144]. The expression pattern of @-tubulin mRNA in the developing rat hypothalamus is sexually dimorphic [146]. Exposure to androgen dring neonatal period has a permanent organizing effect on neuronal morphology and synaptic pattern in these regions [1-10] which contain abundant receptors for sex steroids [186, 187]. Thus, sex steroids may be involved in neuronal plasticity, that is, organi- zation and reorganizatin of neuronal elements and synaptic connections, through regulation of gene expression for neuronal structural proteins such as actin, tubulin, GAP-43 and gap junction protein. ACKNOWLEDGMENTS The present author’s studies were carried out in col- laboration with Prof. Y. Arai and Mrs. S. Murakami, Department of Anatomy, Juntendo University School of Medicine, Tokyo, Prof. A. Urano, Ocean Research Institute, University of Tokyo, Tokyo, Dr. S. Hyodo, Department of Biology, College of Arts and Sciences, University of Tokyo, Tokyo, and Profs. A. P. Arnold and P. E. Micevych, Laboratory of Neuroendocrinology, Brain Research Institute, University of California at Los Angeles, Los Angeles. These studies were supported by the grants from Ministry of Education, Culture and Science of Japan, Narishige Neuroscience Foundation, Yamada Science Foundation and NIH (NS23468 and HD15021). REFERENCES 1 Arnold, A. P. and Breedlove, S. M. (1985) Horm. Behav., 19: 469-498. 2 Matsumoto, A. (1991) Psychoneuroendocrinolo- gy, 16: 25-40. 3 Matsumoto, A. (1992) Rev. Neurosci., in press. 4 Goy, R. W. and McEwen, B. S. (1980) “Sexual Differentiation of the Brain”. MIT Press, Cam- bridge. 5 MacLusky, N. J. and Naftolin, F. (1981) Science, 211: 1294-1303. 6 Arnold, A. P. and Gorski, R. A. (1984) Annu. Rev. Neurosci., 7: 423-442. 7 Breedlove, S. M. (1984) Prog. Brain Res., 61: 147-170. 8 10 Toran-Allerand, C. D. (1984) Prog. Brain Res., 61: 63-98. Arai, Y., Matsumoto, A. and Nishizuka, N. (1986) In “Current Topics in Neuroendocrinology”. Ed. by D. Ganten and D. Pfaff, Springer-Verlag, Ber- lin, Vol. 7, pp. 291-307. Matsumoto, A. and Arai, Y. (1986) In “Pars Distalis of the Pituitary Gland-Structure, Function and Regulation”. Ed. by F. Yoshimura and A. Gorbman, Elsevier, Amsterdam, pp. 239-245. Gorski, R. A., Gordon, J. H., Shryne, J. E. and Southam, A. M. (1978) Brain Res., 148: 333-346. Matsumoto, A. and Arai, Y. (1983) Endocrinol. Jpn., 30: 277-280. Bleier, R., Byne, W. and Siggelkow, I. (1982) J. Comp. Neurol., 212: 118-130. Byne, W. and Bleier, R. (1987) J. Neurosci., 7: 2688-2696. Breedlove, S. M. and Arnold, A. P. (1983) J. Neurosci., 3: 424-432. Nordeen, E. J., Nordeen, K. W., Sengelanb, D. R. and Arnold, A. P. (1985) Science, 229: 671-673. Simerly, R. B., Swanson, L. W., Handa, R. J. and Gorski, R. A. (1985) Neuroendocrinology, 40: 501-510. Greenough, W. T., Carter, C. S., Steerman, C. and DeVoogd, T. J. (1977) Brain Res., 126: 63- WD. Ayoub, D. M., Greenough, W. T. and Juraska, J. M. (1983) Science, 219: 197-198. Garcia-Segura, L. M., Baetens, D. and Naftolin, F. (1985) Dev. Brain Res., 19: 146-149. Simerly, R. B., Swanson, L. W. and Gorski, R. A. (1982) J. Comp. Neurol., 225: 151-166. DeVries, G. J., Best, W. and Sluiter, A. A. (1983) Dev. Brain Res., 8: 377-380. Raisman, G. and Field, P. M. (1971) Science, 173: 731-733. Raisman, G. and Field, P. M. (1973) Brain Res., 54: 1-29. Matsumoto, A. and Arai, Y. (1980) Brain Res., 190: 238-242. Matsumoto, A. and Arai, Y. (1981) Neuroendocri- nology, 33: 166-169. Matsumoto, A. and Arai, Y. (1986) Neuroendocri- nology, 42: 232-236. Matsumoto, A. and Arai, Y. (1986) Lett., 68: 165-168. Dyer, R. G., MacLeod, N. K. and Ellendorff, F. (1976) Proc. R. Soc. Lond. B, 193: 421-440. Sakuma, Y. and Pfaff, D. W. (1981) Brain Res., 225: 184-188. Commins, D. and Yahr, P. (1984) Neurol., 224: 132-140. Bloch, G. J. and Gorski, R. A. (1988) J. Comp. Neurosci. J. Comp. 692 33 34 58 Neurol., 275: 613-622. Simerly, R. B. and Swanson, L. W. (1987) Proc. Natl. Acad. Sci. U.S.A., 84: 2087-2091. Breedlove, S. M. and Arnold, A. P. (1981) Brain Res., 225: 297-307. Gomez, D. M. and Newman, S. W. (1991) Anat. Rec., 231: 498-509. Sasaki, M. and Arnold, A. P. (1991) J. Comp. Neurol., 308: 11-27. Kurz, E. M., Sengelaub, D. R. and Arnold, A. P. (1986) Science, 232: 395-398. Forger, N. G. and Breedlove, S. M. (1987) J. Neurobiol., 18: 155-165. Frankfurt, M., Gould, E., Woolley, C. S. and McEwen, B. S. (1990) Neuroendocrinology, 51: 530-535. Frankfurt, M. and McEwen, B. S. (1991) Neuro- endocrinology, 54: 653-657. Goldstein, L. A., Kurz, E. M. and Sengelaub, D. R. (1990) J. Neurosci., 10: 935-946. Meisel, R. L. and Luttrell, V. R. (1990) Brain Res. Bull., 25: 165-168. Kurz, E. M., Brewer, R. G. and Sengelaub, D. R. (1991) J. Neurobiol., 22: 976-988. DeVries, G. J., Buijs, R. M. and Sluiter, A. A. (1984) Brain Res., 298: 141-145. Matsumoto, A. and Arai, Y. (1979) Cell Tissue Res., 198: 427-433. Matsumoto, A. and Arai, Y. (1981) J. Comp. Neurol., 197: 197-205. Matsumoto, A., Arai, Y. and Osanai, M. (1985) Neurosci. Res., 2: 412-418. Matsumoto, A., Micevych, P. E. and Arnold, A. P. (1988) J. Neurosci., 8: 4168-4176. Garcia-Segura, L. M., Baetens, D. and Naftolin, F. (1986) Brain Res., 366: 131-136. Leedy, M. G., Beattie, M. S. and Bresnahan, J. C. (1987) Brain Res., 424: 386-390. Frankfurt, M. and McEwen, B. S. (1991) Neuro- Report, 2: 380-382. Nottebohm, F. (1980) Brain Res., 189: 429-436. DeVoogd, T. and Nottebohm, F. (1981) Science, 214: 202-204. DeVogd, T., Nixdorf, B. and Nottebohm, F. (1985) Brain Res., 329: 304-308. Canady, R. A., Burd, G. D., DeVoogd, T. J. and Nottebohm, F. (1988) J. Neurosci., 8: 3770-3784. Sawyer, C. H. (1975) Neuroendocrinology, 17: 97-124. Fink, G. (1988) In “The Physiology of Reproduc- tion”. Ed. by E. Knobil, J. D. Neill, L. L. Ewing, G. S. Greenwald, C. L. Markert and D. W. Pfaff, Raven Press, New York, Vol. 1, pp. 1349-1377. Pfaff, D. W. (1980) “Estrogen and Brain Func- tion”. Springer, New York Heidelberg Berlin. 59 60 62 63 64 A. Matsumoto Hart, B. L. and Leedy, M. G. (1985) In “Hand- book of Behavioral Neurobiology”. Ed. by N. Adler, D. Pfaff and R. W. Goy, Plenum Press, New York, Vol. 7, pp. 373-422. Sachs, B. D. and Meisel, R. L. (1988) In “The Physiology of Reproduction”. Ed. by E. Knobil, J. D. Neill, L. L. Ewing, G. S. Greenwald, C. L. Markert, and D. W. Pfaff, Raven Press, New York, Vol. 2, pp. 1393-1485. Stumpf, W. E. (1970) Am. J. Anat., 129: 207-218. Pfaff, D. W. and Keiner, M. (1973) J. Comp. Neurol., 151: 121-158. Simerly, R. B., Chang, C., Muramatsu. M. and Swanson, L. W. (1990) J. Comp. Neurol. , 294: 76— 95. DonCarlos, L. L., Monroy, E. and Morrell, J. I. (1991) J. Comp. Neurol., 305: 591-612. Toran-Allerand, C. D. (1976) Brain Res., 106: 407-412. Uchibori, M. and Kawashima, S. (1985) Int. J. Dev. Neurosci., 3: 169-176. Ferreira, A. and Cacerres, A. (1991). J. Neurosci., 11: 392-400. Segarra, A. C. and McEwen, B. S. (1991) Neuro- endocrinology, 54: 365-372. Arai, Y. and Matsumoto, A. (1978) Psycho- neuroendocrinology, 3: 31—45. Matsumoto, A. and Arai, Y. Lett., 2: 79-82. Matsumoto, A. and Arai, Y. (1977) Brain Res., 129: 275-278. Matsumoto, A. (1984) J. Steroid Biochem., 20: 1420. Clough, R. W. and Rodriguez-Sierra, J. F. (1983) Am. J. Anat., 167: 205-214. Rodriguez-Sierra, J. F. and Clough, R. W. (1987) Synapse, 1: 258-264. Matsumoto, A., Murakami, S. and Arai, Y. (1988) Cell Tissue Res., 252: 33-37. Matsumoto, A. and Arai, Y. (1976) Cell Tissue Res., 169: 143-156. Reier, P. J., Cullen, M. J., Froelich J. S. and Rothchild, I. (1977) Brain Res., 122: 415-436. Koritsanszky, S. (1979) Cell Tissue Res., 200: 135- 146. Lawrence, J. M. and Raisman, G. (1980) Brain Res., 183: 466-471. Walsh, R. J. and Brawer, J. R. (1979) J. Anat., 128: 121-133. Matsumoto, A. and Arai, Y. (1978) Exp. Neurol., 59: 404-412. Van Houten, H. and Brawer, J. R. (1978) J. Comp. Neurol., 179: 719-738. Carrer, H. F. and Aoki, A. (1982) Brain Res., 240: 221-233. (1976) Neurosci. 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 Synaptic Plasticity in Response to Sex Steroids Miyakawa, M. and Arai, Y. (1987) Brain Res., 436: 184-188. Chung, S. K., Pfaff, D. W. and Cohen, R. S. (1988) Exp. Brain Res., 69: 522-530. Silverman, A.-J. (1988) In “The Physiology of Reproduction”. Ed. by E. Knobil, J. D. Neill, L. L. Ewing, G. S. Greenwald, C. L. Markert and D. W. Pfaff, Raver Press, New York, Vol. 1, pp. 1283-1304. Shivers, B. D., Harlan, R. E., Morrell, J. I. and Pfaff, D. W. (1983) Nature, 304: 345-347. Langub, Jr., M. C., Maley, B. E. and Watson, Jr., R. E. (1991) Endocrinology, 128: 27-36. Leranth, C., MacLusky, N. J., Sakamoto, H., Shanabrough, M. and Naftolin, F. (1985) Neuro- endocrinology, 40: 536-539. Leranth, C., Maclusky, N. J., Shanabrough, M. and Naftolin, F. (1988) Neuroendocrinology, 48: 591-602. Watanabe, T. and Nakai, Y. (1987) Arch. Histol. Jpn., 50: 103-112. Chen, W.-P., Witkin, J. W. and Silverman, A.-J. (1989) J. Comp. Neurol., 286: 85-95. Chen, W.-P., Witkin, J. W. and Silverman, A.-J. (1990) Endocrinology, 126: 695-702. Kiss, J. and Halasz, B. (1985) Neuroscience, 14: 69-78. Tsuruo, Y., Kawano, H., Hisano, S., Kagotani, Y., Daikoku, S., Zhang, T. and Yanaihara, N. (1991) Neuroendocrinology, 53: 236-245. Leranth, C., MacLusky, N. J., Shanabrough, M. and Naftolin, F. (1985) Brain Res., 449: 167-176. Third, K. K. and Goldsmith, P. C. (1988) Neuroendocrinology, 47: 203-216. MacLusky, N. J., Naftolin, F. and Leranth, C. (1988) Brain Res., 439: 391-395. Leranth, C., Segura, L. M. G., Palkovits, M., MacLusky, N. J., Shanabrough, M. and Naftolin, F. (1985) Brain Res., 345: 332-336. Pelletier, G. (1987) Neuroendocrinology, 46: 457- 459. Witkin, J. W. and Silverman, A.-J. (1985) Pep- tides, 6: 263-271. Witkin, J. W., Ferin, M., Popilskis, S. J. and Silverman, A.-J. (1991) Endocrinology, 129: 1083-1092. Lehman, M.N., Karsch, F. J., Robinson, J. E. and Silverman, A.-J. (1988) J. Comp. Neurol., 273: 447458. Olmos, G., Naftolin, F., Perez, J., Tranque, P. A. and Garcia-Segura, L. M. (1989) Neuroscience, 32: 663-667. Gould, E., Woolley, C S., Frankfurt, M. and McEwen, B. S. (1990) J. Neurosci., 10: 1286- 1291. 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 129 693 Woolley, C. S., Gould, E., Frankfurt, M. and McEwen, B. S. (1990) J. Neurosci., 10: 4035- 4039. Breedlove, S. M. and Arnold, A. P. (1980) Sci- ence, 210: 564-566. Schrgder, H. D. (1980) J. Comp. Neurol., 192: 567-587. McKenna, K. E. and Nadelhaft, I. Comp. Neurol., 248: 532-549. Ueyama, T., Arakawa, H. and Mizuno, N. (1987) Anat. Embryol., 177: 37-49. Collins, II, W. F., Erichsen, J. T. and Rose, R. D. (1991) J. Comp. Neurol., 308: 28-41. Dube, J. Y., Lesage, R. and Tremblay, R. R. (1976) Can. J. Biochem., 54: 50-55. Breedlove, S. M. and Arnold, A. P. (1983) J. Comp. Neurol., 215: 211-216. Jordan, C. A., Breedlove, S. M. and Arnold, A. P. (1991) J. Comp. Neurol., 313: 441-448. Sengelaub, D. R. and Arnold, A. P. (1989) J. Comp. Neurol., 280: 622-629. Ghanadian, R., Lewis, J. G. and Chisholm, G. D. (1975) Steroids, 25: 753-762. Peng, M. T., Hsu, H. K., Pi, W. P., Wu, K. M., Liao, C. F. and Wan, W. C.-M. Gerntology, 29: 32-40. Kojima, M., Matsuura, T., Tanaka, A., Amagi, T., Imanishi, J. and Sano, Y. (1985) Anat. Embryol., 171: 267-273. Micevych, P. E., Coquelin, A. and Arnold, A. P. (1986) J. Comp. Neurol., 248: 235-244. Konishi, A., Itoh, K., Sugimoto, T., Yasui, Y., Kaneko, T., Takada, M. and Mizuno, N. (1985) Neurosci. Lett., 61: 109-113. Tashiro, T. and Ruda M. A. (1988) Peptides, 9: 383-391. Tashiro, T., Satoda, T., Matsushima, R. and Mizu- no, N. (1989) Brain Res., 481: 392-398. Tashiro, T., Satoda, T., Matsushima, R. and Mizu- no, N. (1989) Brain Res., 497: 177-182. Rajaofetra, N., Passagia, J.-G., Marlier, L., Poulat, P., Pellas, F., Sandillon, F., Verschuere, B., Gouy, D., Geffard, M. and Privat, A. (1992) J. Comp. Neurol., 318: 1-17. Leedy, M. G., Bresnahan, J. C. and Beattie, M. S. (1991) Soc. Neurosci. Abstr., 17: 469. Kerns, J. M. and Peters, A. (1973) J. Neurocytol., 3: 533-555. Anderson, W. J., Stromberg, M. W. and Hinsman, E. J. (1976) Brain Res., 110: 215-227. Matsumoto, A., Arnold, A. P., Zampighi, G. A. and Micevych, P. E. (1988) J. Neurosci., 8: 4177- 4183. Collins, IJ, W. F. and Erichsen, J. T. (1988) Soc. Neurosci. Abstr., 14: 181. (1986) J. 694 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 154 155 Unwin, P. N. T. and Zampighi, G. (1980) Nature, 283: 545-549. Matsumoto, A., Arnold, A. P. and Micevych, P. E. (1989) Brain Res., 495: 362-366. Merk, F. B., Botticelli, C. R. and Albright, J. T. (1972) Endocrinology, 90: 992-1007. Burghardt, R. C. and Anderson, E. (1981) Cell Tissue Res., 214: 181-193. Bergman, R. A. (1968) J. Cell Biol., 36: 639-648. Dahl, G. and Berger, W. (1978) Cell Biol. Intern. Res., 2: 381-387. Garfield, R. E., Kannan, M. S. and Daniel, E. E (1980) Am. J. Physiol., 238: C81—C89. Merk, F. B., Kwan, W. L. and Leav, I. (1980) Cell Biol. Intern. Rep., 4: 287-294. Demianczuk, N., Towell, M. E. and Garfield, R. E. (1984) Am. J. Obstet. Gynecol., 169: 485-491. MacKenzie, L. W. and Garfield, R. E. (1985) Am. J. Physiol., 248: C296-C308. Perez, J., Tranque, P. A., Naftolin, F. and Garcia- Segura, L. M. (1990) Neurosic. Lett., 108: 17-21. Hatton, G. I., Yang, Q. Z. and Koran, L. E. (1992) Brain Res., 572: 291-295. Cobbett, P., Yang, Q. Z. and Hatton, G. I. (1987) Brain Res. Bull., 18: 365-370. Tobias, M. L. and Kelley, D. B. (1988) J. Neuro- sci., 8: 2422-2429. Adcock, I. M. and Greenstein, B. D. (1986) J. Endocrinol., 109: 23-28. Stanley, H. F. and Fink, G. (1986) Brain Res., 370: 223-231. Rogers, L. C., Junier, M.-P., Farmer, S. R. and Ojeda, S. R. (1991) Mol. Cell. Neurosci., 2: 130- 138. Lusting, R. F., Sudol, M., Pfaff, D. W. and Federoff, H. J. (1991) Mol. Brain Res., 11: 125- 132. Matsumoto, A., Arai, Y., Urano, A. and Hyodo, S. (1991) Neurosci. Lett., 131: 159-162. Matsumoto, A., Arai, Y., Urano, A. and Hyodo, S. (1992) Neurosci. Res., in press. Bamburg, J. R. and Bernstein, B. W. (1991) In “The Neuronal Cytoskeleton”. Ed. by R. D. Bur- goyne, Wiley-Liss, New York, pp. 121-160. Fifkova, E. (1985) Brain Res. Rev., 9: 187-215. Hirokawa, N. (1991) In “The Neuronal Cyto- skeleton”. Ed. by R. D. Burgoyne, Wiley-Liss, New York, pp. 5-74. Leslie, M., Forger, N. G. and Breedlove, S. M. (1991) Brain Res., 561: 269-273. Austin, L. (1985) In “Handbook of Neurochemis- try’. Ed. by A. Lajtha, Plenum, New York, 2nd ed, pp. 1-45. Cambray-Deakin, M. A. (1991) In “The Neuronal Cytoskeleton”. Ed by R. D. Burgoyne, Wiley-Liss, 156 157 158 159 160 161 162 163 164 165 166 167 178 179 180 181 A. MATSUMOTO New York, pp. 233-255. Grafstein, B. and McQuarrie, I. G. (1978) In “Neuronal Plasticity”. Ed. by C. W. Cotman, Raven Press, New York, pp. 155-195. Phillips, L. L. and Steward, O. (1990) Mol. Brain Res., 8: 249-257. Sutcliffe, J. G. (1988) Annu. Rev. Neurosci., 11: 157-198. Mairesse, N. and Galand, P. (1986) J. Steroid Biochem., 24: 287-288. Hsu, C.-Y. J. and Frankel, F. R. (1987) J. Biol. Chem., 262: 9594-9600. Beyer, E. C., Paul, D. L. and Goodenough, D. A. (1990) J. Membrane Biol., 116: 187-194. Dermietzel, R., Hwang, T. K. and Spray, D. S. (1990) Anat. Embryo., 182: 517-528. Bennett, M. V. L., Barrio, L. C., Bargiello, T. A., Spray, D. C., Hertzberg, E. and Saez, J. C. (1991) Neuron, 6: 305-320. Dermietzel, R., Traub, O., Hwang, T. K., Beyer, E., Bennett, M. V. L., Spray, D. C. and Willecke, K. (1989) Proc. Natl. Acad. Sci. U.S.A., 86: 10148-10152. Shiosaka, S., Yamamoto, T., Hertzberg, E. L. and Nagy, J. I. (1989) J. Comp. Neurol. , 281: 282-297. Yamamoto, T., Shiosaka, S., Whittaker, M. E., Hertzberg, E. L. and Nagy, J. I. (1989) J. Comp. Neurol., 281: 269-281. Yamamoto, T., Hertzberg, E. L. and Nagy, J. I. (1990) Brain Res., 527: 135-139. Yamamoto, T., Hertzberg, E. L. and Nagy, J. I. (1991) Synapse, 8: 119-136. Matsumoto, A., Arai, Y., Urano, A. and Hyodo, S. (1991) Neurosci. Lett., 124: 225-228. Paul, D. L. (1986) J. Cell Biol., 103: 123-134. Naus, C. C. G., Belliveau, D. J. and Bechberger, J. F. (1990) Neurosci. Lett., 111: 297-302. Micevych, P. E. and Aberson, L. (1991) J. Comp. Neurol., 305: 96-118. Garfield, R. E., Sims, S. and Daniel, E. E. (1977) Science, 198: 958-960. Puri, C. P. and Garfield, R. E. (1982) Reprod., 27: 967-975. Saito, Y., Sakamoto, H., MacLusky, N. J. and Naftolin, F. (1985) Am. J. Obstet. Gynecol., 151: 805-812. Risek, B., Guthrie, S., Kumar, N. and Gilula, N. B. (1990) J. Cell Biol., 110: 269-282. Darnell, Jr., J. E. (1982) Nature, 297: 365-371. Beato, M. (1989) Cell, 56: 335-344. Evans, R. M. (1988) Science, 240: 889-895. Yamamoto, K. R. (1985) Annu. Rev. Genet., 19: 209-252. Amero, S. A., Kertsinger, R. H., Moncrief, N. D., Yamamoto, K. R. and Pearson, W. R. (1992) Mol. Biol. Synaptic Plasticity in Response to Sex Steroids 695 Endocrinol., 6: 3-7. Neurosci., 11: 3025-3033. 182 Nudel, U., Zakut, R., Shani, M., Neuman, S., 185 Popper, P., Ulibarri, C. and Micevych, P. E. Levy, Z. and Yaffe, D. Nucleic Acids Res., 11: (1992) Mol. Brain Res., 13: 43-51. 1759-1771. 186 Sheridan, P. J., Sar, M. and Stumpf, W. E. (1974) 183. Breedlove, S. M. (1986) J. Neurobiol., 17: 157- Endocrinology, 94: 1386-1390. 176. 187 Sheridan, P. J., Sar, M. and Stumpf, W. E. (1974) 184 Araki, I., Harada, Y. and Kuno, M. (1991) J. Am. J. Anat., 140: 589-593. mae foe of oman —— 1 <r tf Ayeciy- a wine pry Rik As ral’ 3 A eae oieyeliay] nike Pah. te 2 es 3 BM pie’ @s cme ys ft ae + is Vaiiece, a, Hid Comeni, thetic, AAA” |* if iy Pa inh Feat, #4, ; Nag May 999s wh i: Ful, te 1 is on ‘Poiana A { tens LD or F : vie Eat ay rian bo vaso “Te i, vig’ fh SH i, 7 if rib | Pa vy APG2 Sis 5 ¢ : a 1 ' La i Gaia SNP aete ihe ae ~.' we e i: ay = Ww ZOOLOGICAL SCIENCE 9: 697-712 (1992) REVIEW The Species Concept in Theory and Practice WALTER J. Bock Department of Biological Sciences, Columbia University, New York NY 10027, U.S.A. ABSTRACT—The species concept has always been an integral part of basic biological theory and is not primarily a part of systematics. The biological species concept is based on evolutionary theory and applies only to sexually reproducing organisms. The important aspect of the species concept is lack of gene flow between different species, and hence the defining criterion of the species is genetic isolation. Species possess three different major properties, namely genetic isolation, reproductive isolation and ecological isolation; these properties evolve at different times and under the effect of different causes during the speciation process. Speciation requires an external isolating (geographic-ecological) barrier during the initial allopatric phase in which period genetic isolation evolves and must reach 100% efficiency. The subsequent sympatric phase of speciation occurs after the disappearance of the external isolating barrier when members of the two newly evolved species can interact with one another and exert mutual selective demands on one another. Much of the reproductive and ecological isolation evolves during this secondary sympatric phase. The practical application of the species concept in systematics and the recognition of species taxa requires many arbitrary decisions. No single Species notion exists which can be applied to all organisms, sexually and asexually reproducing. Recognition of species taxa in asexually reproducing organisms is based on amount of variation and gaps in the variation of phenotypic features associated with ecological attributes of these organisms as compared to similar attributes in sympatric species taxa of sexually reproducing organisms. Because recognition of species taxa, including those in sexually reproducing organisms, depends on many arbitrary decisions especially when dealing with broad geographical and temporal ranges, species taxa in general cannot be used as the foundation for developing and testing theoretical concepts in evolutionary theory. This can only be done with the non-dimensional species concept and with species taxa that match this concept closely. © 1992 Zoological Society of Japan INTRODUCTION Perhaps more problems exist in understanding the species concept than any other basic theoreti- cal idea in biology in spite of excellent analyses of this concept by Mayr [1-9] and others. Much of this confusion apparently arises from the erroneous belief accepted by most workers that the theoretical species concept lies within the purview of systematics and the resulting attempt to develop a single species concept applicable to all organisms [10]. This is simply not possible. Rather, the species concept is part of basic biological theory and as such is applied as closely as possible to Received May 29, 1992 objects in nature (=species taxa) by systematic biologisis. Moreover, the species concept has changed fundamentally over the decades and with these changes, our ideas on species taxa have modified accordingly. From the earliest days of biological thinking until sometime after Darwin published his “On the origin of species”, the species concept was based on typological essential- ism following the ideas of Plato and other early philosophers. The typological species concepts is usually, but erroneously, called the morphological species concept and was applied equally to all organisms, both those reproducing sexually and asexually. Species taxa were recognized morpho- logically. Geographic variants were treated as different species, as were many distinctive morphs 698 W. J. Bock within populations. As evolutionary ideas became better known, the basis of the species concept modified gradually to that of evolutionary theory. By the early years of the 1940’s many, but not all, workers based the species concept firmly on evolu- tionary theory with the biological species concept being the generally accepted one following Mayr [1, 2]. The biological species concept applies only to sexually reproducing organisms, but most syste- matists do not make this distinction when describ- ing species taxa which are still recognized morpho- logically. However, morphs within populations, when recognized, are no longer considered as species and geographic variants are grouped together into broadly based polytypic species taxa. Yet many workers continue to treat the species concept as if it was part of systematics and to believe that a single species concept exists and can be applied to all organisms, sexually and asexually reproducing. Although earlier comments may exist, the first clear statement that the species concept is not a part of systematics, but rather is an integral part of evolutionary theory, was by Szalay and Bock [11]. This review of the species concept will be found- ed on three fundamental points, namely: (a) The theoretical species concept (the non-dimensional species concept) is part of evolutionary theory and applies only to sexually reproducing organisms. (b) Recognition of species taxa in nature is part of systematics practice and for sexually reproducing organisms (the multi-dimensional species notion), species taxa should be as close to the theoretical concept as possible. (c) No single species idea exists which can be applied equally well to all organisms. SPECIES—THE THEORETICAL DEFINITION A) Species versus phyletic lineages If the species concept is based on evolutionary theory, the first decision is which unit within this theory is to be considered the species. This problem arises because two distinctly different but related units have been labelled as the species by different workers. These are: (1) the complex of interbreeding individual organisms co-existing at one point in time which are genetically isolated from other such complexes; and (2) the time-line of this complex of interbreeding organisms result- ing from it reproducting itself generation after generation. Only the first is properly termed the species; the second is the phyletic lineage [12, 13]. Several species concepts—the evolutionary, the phylogenetic and the cladistic—are actually seg- ments of phyletic lineages. One must be most careful in reading the biological and philosophical literature to determine whether the author actually means “the species” or really means “the phyletic lineage” when the term species is used; often it is not possible to determine which of these two units was means and the real meaning of the analysis cannot be ascertained. B) The nondimensional species concept In the decades following Darwin, biologists struggled to ascertain the core of the species concept and finally agreed that species are groups of interbreeding individuals which have vertical flow of genetic material during the production of offspring, and that the boundary between species is genetic isolation or the lack of vertical gene flow between members of different species. That is, members of a species can interbreed with one another and contribute genes to the production of offspring, but members of different species cannot do so. Genetic isolation became accepted during the evolutionary synthesis as the criterion for the species concept. Vertical gene flow bewteen spe- cies is achieved only by interbreeding (i.e., repro- duction) between them. But evolutionists con- fused genetic isolation with repreductive isolation. And hence the species concept was defined in the period of the evolutionary synthesis as “Species are groups of actually or potentially interbreeding natu- ral populations, which are reproductively isolated from other such groups” [2]. Unfortunately, evo- lutionists failed to recognize the asymmetry be- tween reproductive isolation and genetic isolation between speies. Species which are reproductively isolated are genetically isolated but species which are genetically isolated may not be reproductively isolated. The horse and donkey represent two species which are genetically isolated but not re- productively isolated, at least under captive condi- The Species Concept 699 tions if not in the wild. This confusion between genetic and reproductive isolation resulted in dis- agreements on aspects of speciation such as whether selective demands can increase genetic isolation from less than 100% to 100% after sym- patry has been established between newly evolved species (see Koopman [14] for an example of this problem) and much other confusion about the species concept and the process of speciation (see [39]). One must use care in reading the literature on species because the terms genetic isolation and reproductive isolation are frequently used inter- changeably. If reproductive isolation is used, it is usually in the meaning of genetic isolation. For these reasons, I proposed a minor, but significant, modification of the bilogical species concept of Bock [13], namely that: “A species is a group of actually or potentially interbreeding populations which is genetically isolated in nature from other such groups.” This definition is actually what was meant by the original proposers of the biological species concept as can be determined from their entire analysis. Genetic isolation be- tween species is achieved by the possession of intrinsic isolating mechanisms (I I M; see Mayr [4] pp. 91-109; and see below). Mayr [4] introduced the idea of the non-dimensional species concept (=the biological species concept) which is most fundamental, but has been little appreciated by evolutionary biologists and systematists. Unfortu- nately Mayr did not connect the non-dimensional species concept tightly with evolutionary theory and did not contrast it sufficiently with the multi- dimensional species notion which is applied in systematic practice to species taxa. It cannot be stressed too strongly that further development of evolutionary theory can be achieved only with the non-dimensional species concept, not with the multi-dimensional species notion or all recognized species taxa. The species concept applies only to sexually responducing organisms. It has no meaning for asexually reproducting organisms which are not organized into interbreeding uints. Species taxa appeared in the history of life at some (reasonably short) time after the evolution of sexual reproduc- tion, or possibly in conjuction with the evolution of sexual reproduction. The necessity for species results from the mechanisms of meiosis (the shuf- fling and recombination of existing genetic varia- tion during gamete formation) and of fusion of two haploid gametes into diploid zygotes of the next generation, all of which shuffles and recombines the genetic material existing in the genotypes of parental organisms into new offspring genotypes. If too much variation exists among the genotypes of the parental organisms, then the possibility of producing too many non-viable offspring is great. Species, and hence species boundaries, provide limits to the amount of genetic variation existing within each species. The arrangement of sexually reproducing individuals into discrete species taxa permits a more efficient subdivision of the environ- ment by sympatric species [9]. A species, therefore, is a group of individual organisms which are able to interbreed and which are genetically isolated from other species. As a species reproduces itself generation after genera- tion it forms a phyletic lineage which is the time line (the history) of the species. A cross-section through a phyletic lineage at any point in time represents a species. Cross-sections through the same phyletic lineage at different times represent species. But it is a non-question to ask whether these different time slices of a phyletic lineage (= species) represent the same species or different species. In theoretical considerations, limits can- not be placed on a species along a phyletic lineage because any such limits would necessitate drawing a boundary between one species and its successor; such boundaries would imply an evolutionary change “between species” other than the observed gradual change “within species” which has never been demonstrated. Therefore, if species can not be delimited along a phyletic lineage, it is not possible to speak of the origin or the brith of a species, nor is it possible to speak of the age of a species. All existing species are of equal age, or in other terms, all species are ageless. And it is not possible to speak of evolutionary change as being within the limits of a species or transgressing the limits of a species; evolution along a phyletic lineage never results in a new species and hence never passes a species limit regardless of the amount of phyletic evolutionary change that has taken place [12, 13]. 700 W. J. Bock KM 7 Species and speciation are not irreversible as assumed by some workers. Good species taxa may evolve and exist sympatrically for a period of time, but then may start to interbreed, exchange genetic mateiral and finally reunite into a single species [13]. An excellent example is the towhees (Pipilo erythrophthalmus and Pipilo ocaci) found in the central Mexican highlands [15] which started to hybridze some hundreds of years ago and have become a single genetic-reproductive-ecological unit over most of their range. Such subsequent interbreeding and merging of formally good spe- cies depends on the nature of the I I M previously separated them. C) Phylogenetic species concepts Herein I group several diverse species concepts proposed under the headings of the evolutionary, phylogenetic and cladistic species concepts [16- 18]. Regardless of other aspects of the definition for these species concepts, each includes some reference to a reasonable time dimension for the 7 ———lime T Time T, Fic. 1. Schematic diagram illustrating the concept of the phyletic lineage, which is a species taxon reproducing itself genera- tion after generation through time. Change in the species with respect to time is phyletic evolution. A cross-section through the phyletic lineage at any point in time is a species. Cross-sections at E different points in time are neither the fs same species of different species as no species boundary can separate them; they are simply different cross-sections of the same phyletic lineage. (From Bock, [12]). species. Hence these definitions confuse the spe- cies concept with that of the phyletic lineage and as such are not acceptable. Unless the concepts of the species and of the phyletic lineage are kept sharply separated, futher evolutionary theory de- velopment is not possible. For example, species can evolve, but phyletic lineages, which represent the past history of a species, cannot. D) Morphological species concepts Morphological species do not now or ever have existed as a proper theoretical species concept except as a misnomer for the older typological species definition. Some workers (e.g., Mishler and Brandon [10]) have proposed a morphological species definition in the attempt to advocate a single species concept for all organisms, including asexually reproducing forms. In essence, these definitions do little more than state that the species is what a taxonomist calls a species. Such recent efforts to propose morphological species concepts regardless of the wording of the definition, repre- The Species Concept 701 sent a confusion between the theoretical definition of the species concept and guidelines for recogniz- ing species taxa (multi-dimensional species no- tions). Unless authors proposing a morphological species concept also present a sweeping reanalysis of evolutionary causes and processes, including full considerations of causes and processes involved in sexual and asexual reproduction and their bearing on evolution which support their ideas on the species, morphological species concepts should be rejected out of hand. Recently some cladists (e.g., [17]) have argued that the species is the smallest recognizable tax- onomic unit in nature in addition to the usual cladistic concept that a species is the segment of a phyletic lineage between two successive branching points. This idea that the species in the smallest recognizable taxonomic unit in nature confuses the species concept with that of geographic variants within broadly ranging species. It confuses the theoretical species concept with the practical work of systematists, forgetting that the species concept is part of basic theoretical biology, not systematic practice. And it returns to a concept that is very similar to it not identical to the typological species concept which has been discarded along with the rejection of typology in biological thinking. E) Recognition species concept Paterson ([{19] and elsewhere) advocated a spe- cies concept based on the behavioral and other means by which members of one species recognize members of their species and distinguish them from members of other species. Paterson’s argu- ment on the invalidity of the biological species concept confuses completely genetic isolation and reproductive isolation. In addition, his argument confuse the basic criteria for a theoretical species concept with properties of species by which separa- tion between species is achieved. Several excellent critiques rejecting the recognition species concept have been published (e.g., [20, 21]) and need not be repeated here. None of these several species concepts, advo- cated as alternatives to the biological species con- cept, have a firm foundation in evolutionary theory and none provide a better understanding of evolu- tionary units designated by the term species [7-9, 21, 22]. Most fail because they confuse the theore- tical species definition which falls within the pro- vince of evolutionary theory and the practical application of this concept to species taxa in nature which falls within the province of systematics. F) Sphere of application The current species concept (based on evolu- tionary theory) applies only to sexually reproduc- ing organisms which is a major difference from the former typological species concept which applied to all organisms. This restriction developed with continued analysis of ideas associated with species concepts and the decision to focus on genetic isolation as the major criterion of the species concept. This was not the only possible develop- ment of the species concept under evolutionary theory, but is the one which occurred and hence the current restriction of the species concept to sexually reproducing organisms. Species are not an autonomic consequence of evolution. Actually there is nothing in basic evolu- tionary theory which necessitates the existence of species. Rather they exist in sexually reproducing organisms as a result of the processes of meiosis in the formation of gametes and the fusion of ga- metes in the formation of zygotes of the next generation, as discuseed above. G) Major properties of species Species possess three major sets of properties, or can be said to represent three different types of units which have been little discussed prior to Bock [13]; Mayr [3, 4] mentioned these three properties, but did not develop them. What is important is that the evolution of attributes possessed by spe- cies which are responsible for each of these prop- erties may occur at different times during the speciation process and under the action of differ- ent evolutionary causes. Fully evolved species possess all three sets of properties completely developed, but some species which are still in the process of speciation may not have all of these properties totally developed. Because most evolu- tionists have focused so strongly on genetic isola- tion in their analysis of the species concept and on the evolution of features associated with genetic isolation, there has been a serious lack of under- 702 W. J. Bock standing of the properties of reproductive and ecological isolation and of their evolution during speciation. Understanding of these properties clarifies considerably our comprehension of the entire process of speciation and of species taxa. (1) Genetic unit: Members of a species form a genetic community which is completely geneti- cally isolated (100%) from other species; that is, no vertical gene flow occurs between species. Hence the species is the largest unit in which genetic recombination can take place. This prop- erty exists for all species, with the exception that in a few cases a small amount of vertical gene flow may occur for a limited period of time between good species. Genetic isolation is maintained by the possession of intrinsic isolating mechanisms (II M) which exist in a variety of forms from premating to postmating mechanisms (see Mayr [4] pp. 91-109 for an excellent discussion and classification of IIM). Any of these diverse isolating mechanisms completely isolate species genetically, but they vary in other properties namely in their ability to isolate these species reproductively. Since the turn of the century, evolutionists have concentrated, and quite rightly sO, on genetic isolation and on IIM in their analysis of the species concept and have used genetic isolation as the fundamental criterion for the species definition. This decision was quite correct because without genetic isolation, the other properties of species could not evolve and be maintained. But evolutionists believed that with the clarification of genetic isolation between spe- cies, they have understood all that was essential about the species concept. Unfortunately this is simply not the true. Moreover, most workers have equated I I M and their evolution only with genetic isolation (generally termed reproductive isolation) and have not made the distinction between the evolution of IIM for genetic isolation and the evolution of II M for reproductive isolation. One must always keep in mind that a diversity of 11 M exist and that only some serve for reproductive isolation. Indeed, even is recent papers [21], the evolution of genetic isolation is discussed as the evolution of II M and almost never as the evolu- tion of genetic isolation which confuses completely the evolution of some IIM under the action of mutual selective demands (see below) following the sympatric phase of speciation. This confusion leads directly to further mis- understandings about the “historic adaptationistic baggage” of the biological speies concept and the process of speciation [21]. Adaptive aspects during the process of speciation well covered by Moore [23] must be reanalyzed with a clear understanding of the differences between genetic isolation and reproductive isolation. (2) Reproductive unit: Members of a species form a reproductive community which is reproduc- tively isolated (100%) from other species; that is, there is no hybridization between species. Species can be genetically isolated without being reproduc- tively isolated. Individual members of a species which attempt to or actually do interbreed with members of other species incur a reproductive cost relative to conspecific individuals which do not attempt to interbreed and hence these interbreed- ing individuals would be selected against because they have a lower fitness relative to non-inter- breeding individuals. Unfortunately most evolu- tionists have not distinguished between genetic and reproductive isolation [14], partly because both types of isolation result from II M. But full reproductive isolation results only from some I] M—namely those premating mechanisms pos- sessing low reproductive costs [categories 1(a) and 1(b) of Mayr’s [4] classification]. And these II M with low reproductive cost can evolve under selec- tive demands after species have achieved full gene- tic isolation. (3) Ecological unit: Members of a species form an ecological unit as individual members of one species have similar ecological requirements which differ from those of members of other species. Species can be genetically isolated with- out being ecologically isolated. Members of spe- cies which are not ecologically isolated will com- pete ecologically when sympatric, and those indi- viduals of the two species which compete ecologi- cally will have lower fitness relative to conspecific individuals which do not compete ecologically with members of other species. Ecological isolation between species results from the possession of phenotypic features associated with the ecological requirements of each species [24-26]. The Species Concept 703 H) Ontological nature of species taxa Although species taxa will be discussed below, one aspect, namely the ontological nature of spe- cies taxa, will be dealt with here. One point is whether species taxa are “real” or whether they are just abstractions with the only real units being the individual organisms. This paper is not the place to enter a lengthy discussion of the meaning of reality and the criteria by which various units in nature are recognized as real. It has been argued that species taxa are real because the same species taxa are recognized by groups of people with totally different traditions and training. Mayr [4] stated that the natives living in the Arfak Moun- tains of New Guinea recognize 136 of the 137 species taxa of birds found in that region as delimited by European systematists who have nev- er seen these birds alive. Others workers counter that those native people who may depend less on birds do not recognize the same species taxa as do tranined systematists—that the distinction be- tween species taxa depends on the importance of these organisms to the life of different human groups. Nevertheless, it has been demonstrated time and time again that in local regions, sexually reproducting organisms are organized, with few exceptions, into clearly demarcated species taxa (e.g., Mayr [27]) which are recognized by any person with the proper training. This agreement in the recognition of species taxa in local regions supports sufficiently the reality of species taxa [7]. The second point is whether these species taxa should be regarded as groups (sets, collections) or as individuals. This question was raised in the 1970’s by Ghiselin [28, 29]. Hull [30-32], and Mayr [8] among others who argued that species taxa do not represent groups, but individuals, with individual organisms being parts, not members, of the species taxon. These conclusions are not completely clear because it is not always definite that some workers advocating species as indi- viduals refer to species taxa or to phyletic lineages; in some cases, the contents of the papers suggest that the authors allude to phyletic lineages or to a combination of species taxa and phyletic lineages. It is argued that the characteristic integration of individuals is achieved in each species taxon by an integrated gene pool resulting from gene flow between all geographic parts of the species taxon [33]. Serious questions exist whether the gene pool of a species taxon is sufficiently tightly integrated as reguired for individuals [13]. The possession of some integration of the gene pool of a species taxon is not sufficient for individual status as certain groups may possess a degree of integration. What is importance is the tightness of this integra- tion [13]. Other authors, such as Hull, considered species taxa as “historical individuals” —that is, individuals in the sense of historical-narrative ex- planations (see [34]) which is quite a different concept from that of individuals in the sense of nomological-deductive explanations. Other work- ers (e.g., [35]) countered that species taxa are proper groups comprised of individual organisms as members, not parts, but that these groups are not characterized by typological essences. Analy- sis of the broad meaning of individuals versus groups in the philosophy of science suggests strongly that species taxa are groups, not indi- viduals, but are groups characterized by nontypo- logical attributes. Nontypological attributes are those which can vary in space and time, but are sufficiently similar to permit recongition of indi- viduals as members of the group—these are attri- butes which have been called “family-resemblance features” existing under the general approach of “population thinking.” THE PROCESS OF SPECIATION A) Evolutionary processes and causes Before it is possible to discuss the evolution of species diversity (speciation), it is necessary to outline the different processes of evolution and to correlate these with the species and the phyletic lineages. The process of evolutionary change, in the broadest sense, can be defined as change in living organisms with respect to time with the minmum time period being one generation; hence evolutionary change is difference observed be- tween organisms in one generation and those in descendent generations. Frequently cited defini- tions of evolution as “genetic changes over time” are far too restrictive. Evolutionary change con- 704 W. J. Bock sists of two basic, but quite different processes, namely phyletic evolution and speciation. Phyletic evolution can be defined as evolutionary change in a phyletic lineage. Phyletic evolution results from (a) the set of causes resulting in genetically-based phenotypical variation generation after genera- tion, and (b) selective demands arising from the external environment. Differences may exist in the causes and processes of phyletic evolution in sexually and asexually reproducing organisms, but these are of no immediate concern to this paper. Speciation occurs only in sexually reproducing organisms, and is the multiplication of a species or the splitting of a phyletic lineage into two or more lineages. No special causes of speciation exist other than those acting in phyletic evolution. The only additional factor occurring in speciation is a boundary condition in the form of an external reproductive barrier separating two portions of the original species into two separate populations and keeping them genetically isolated during the Origin of External Isolating , Barrier / Ancestral Species c Species B period in which members of the two populations could interbreed and exchange genetic material if they were in contact. During this period, intrinsic isolating mechanisms may evolve in the two populations by means of phyletic evolution. It is clear that phyletic evolution can occur without speciation, but that speciation can occur only with accompanying phyletic evolution in one or both populations isolated by the external barrier. In sexually reproducing organisms, the unit of evolutionary change is the interbreeding popula- tion which may be equivalent to the species if the geographic range of the species is sufficiently small. Different populations within species with a larger geographic range can evolve partly to com- pletely independently from one another; the changes in diverse populations are usually con- nected by gene flow between them, unless, of course, the different populations are separated by an external isolating barrier and are in the process of speciation (see below). The generalization that Fic. 2. Schematic diagram illustrating the relationship between speciation (split- ting of a phyletic lineage into two) and phyletic evolution. The process of speciation required the presence of an external isolating barrier and phyletic evolution in at least one of the two lineages. Species A and B are distinct and separate from one another, but each is not different from the ancestral species common to both phyletic lineages. (From Bock, [12]). The Species Concept 705 the species is the unit of evolutionary change is true only in a very rough sense. The phyletic lineage is never the unit of evolutionary processes because it is an abstraction composed of the time line of a species. The phyletic lineage is an historical representation and hence can no longer be changed. It is important to stress most strongly that although the interbreeding population or the spe- cies is the unit of evolutionary change, the process, it is not the unit on which evolutionary causes act. Causes of phyletic evolution act on individual organisms, and as a result of these causes including the reproduction of offspring organisms, evolu- tionary change occurs. In addition, it must be stressed that all observed or postulated phyletic evolutionary changes are gradual in that they are reducible to the causes and/or changes observable in populations genera- tion after generation. These evolutionary changes may have different rates, but they are all gradual in the sense just mentioned. Diverse workers have used other terms, such as vertical versus horizontal evolution or anagensis versus cladogensis, rather than phyletic evolution and speciation. I would recommend against use of these other terms because they are less clear and because they involve certain unnecessary connota- tions which are absent in the terms phyletic evolu- tion and speciation. Other terms such as phyletic speciation or vertical speciation have been used frequently, but in an ambiguous sense for phyletic evolution; they should be abandoned as they imply serious misconceptions and erroneous connota- tions with the process of speciation. B) Speciation Speciation, or splitting of a phyletic lineage, is the process by which two or more descendent species evolve from a single ancestral species. Although in most cases, speciation is a dichoto- mous splitting of phyletic lineages, speciation can- not be generalized to be always dichotomous as done by some workers, e.g., cladists. It is entirely possible to have a simultaneous subdivision of an ancestral species into three or more geographically defined portions (different lineages) by external barriers. The full process of speciation includes the evolution of the three fundamental properties —genetic isolation, reproductive isolation and ecological isolation—in the two or more descen- dent species, not just the evolution of genetic isolation as covered in many analyses of specia- tion. The evolution of each of the these properties will be discussed separately because the each can occur independently of the others and as a result of different sets of evolutionary causes. Moreover the timing of the evolution of these different properties is important. Speciation is a popula- tional phenomenon and is the result of the evolu- tionary causes underlying the process of phyletic evolution; hence speciation must be accompanied by phyletic evolution in at least one of the popula- tions. There are no distinct evolutionary causes of speciation, but the additional boundary condition of an external isolating barrier which divides the ancestral species into different units and remains in place sufficiently long to permit independent phyletic evolution of each of these units. The only external isolating barrier demonstrated to be suf- ficient for speciation is a geographical-ecological one [2, 4] giving rise to distinct allopatric (geo- graphically separated) populations and leading to geographic or allopatric speciation. Excluding speciation by hybridization and polyploidy in some groups of plants, no evidence exist for suitable external isolating barriers giving rise to distinct sympatric populations and leading to sympatric speciation [4, 26, 36]. Nor is there any evidence for parapatric speciation in which the two portions of the ancestral species inhabit different geo- graphic regions, but have always maintained a contact with one another as proposed by White [37]. It is erroneous to ask whether speciation occurs via geographic isolation or chromosomal rearrangements, as argued by some workers sup- porting the concept of parapatric speciation. Both aspects are involved in those cases where the intrinsic isolating mechanism results from chromo- somal differences (see [6]). The question exists of whether distinct types of speciation exists within allopatric or geographic speciation (Mayr, [6]). Diverse workers have discussed whether speciation occurs by “splitting” or by “budding”, or whether speciation occus when the ancestral species is divided into two large 706 W. J. Bock portions or only when there is a small geographic isolate—a founder population [4, 33]. Mayr, among others have argued for a “genetic revolu- tion” in the founder population (=the founder principle; Mayr [2] p. 237). The hypothesis of a genetic revolution postulates that the modification of the integrated gene pool of the ancestral species to the new integrated gene pool of the descendent species occurs at the origin of the founder popula- tion and is a rapid change. Genetic revolutions occur by genetic drift in small populations in which only a small portion of the genetic variation of the ancestral species is present in the founder popula- tion [5, 33, 38, 39]. No question exists that phyletic evolutionary change can be much faster in small populations, but serious doubt still exists on the role and importance of the founder principle in the process of speciation. No one has demonstrated the high degree of integration of the gene pool of a species as postulated by Mayr or the ways in which this “integrated gene pool” can modify over time. It has not been shown that the only way the “integrated gene pool” of a species can change is by genetic revolutions and no one has demon- strated the workings of a genetic revolution ex- perimentally in spite of numerous attempts. Nor has anyone demonstrated the existence of the “closed genetic system” of species which is mod- ified only during speciation [38]. Because speciation is more than the evolution of genetic isolation as will be discussed below, but also involves the evolution of reproductive isola- tion and ecological isolation both of which take place largely after the establishment of sympatry between the descendent species, additional doubt exists on the role of the founder principle and associated concepts such as “punctuated equilib- rium” as advocated by Eldredge and Gould [40]. C) The speciation process A full episode of the speciation process begins with the appearance of an external isolating bar- rier—a geographic-ecological one in the case of geographic speciation—and ends with the full achievement of reproductive and ecological isola- tion in addition to genetic isolation. This episode can be divided into two portions, namely (1) the initial allopatric phase and (2) the subsequent sympatric (or neosympatric) phase [12, 13]. It must be stressed that the speciation process is not completed with the disappearance of the external isolating barrier and the re-establishment of sym- patry of the two newly genetically isolated species as is ususally assumed in discussions of speciation. Rather this is just the end of the first phase and important changes take place during the subse- quent sympatric phase. The same sequence exists in speciation by hibridization and polyploidy in which the evolution of genetic isolation is achieved in a single sexual generation and hence the need for the alloptric phase is eliminated. (1) Evolution of genetic isolation: With the establishment of the external isolating barrier, the separated populations can undergo independent phyletic evolution. Intrinsic isolating mechanisms serving for genetic isolation may appear fortuitous- ly as pleiotropic consequences of these phyletic evolutionary changes during the initial allopatric phase of speciation. Which isolating mechanism appears and when it appears are strictly chance- based. Moreover, the evolution of these II M are never the result of their being favored as genetic isolating mechansims by selective demands arising from the external environment. For speciation to continue, genetic isolation must be 100% complete by the end of the initial allopatric phase when the external isolating barrier disappears and the two populations can expand their geographic ranges and become sympatric. If genetic isolation is less than 100% effective, then the members of the two populations will hybridize and exchange genetic material [23]; Moore’s excellent analysis must be read carefully as he is considering genetic isolation only, not reproductive isolation. The only out- come of sympatry of two populations with less than 100% genetic isolation is increased hybridization and subsequent merging of two populations. The speciation process will come to a halt and the two populations will continue to exist as a single spe- cies with the steepness of the secondary zone of integration between them depending on many factors. No theoretical argument or empirical evidence supports the oft-claimed idea [41] that selective demands from the external environment will disfavor hybrid individuals and lead to greater and greater genetic isolation until the level of The Species Concept 707 Sympatric Phase Time Allopatric Phase Onset of Speciation ~~ Termination of Speciation a)Reduction of Competition b)Reduction of cost of Geographical Barrier reproduction Initial Overlap Origin of Intrinsic Isolating Mechanism Subspeciation Geographical Spread Evolutionary Divergence Fic. 3. Schematic diagram to show the divergence of the two phyletic lineages during a speciation episode. Retes of evolutionary change and resulting divergence are low during the period of subspeciation and somewhat higher during the allopatric phase after the appearance of the geographic-ecological barrier. The rate of divergence increases sharply after the external barrier disappears and the two species have become sympatric. Mutual selective demands pushing the divergence during the sympatric phrase arise from exclusionary species interactions between members of the two species. terminates. (From Bock, [12]). 100% genetic isolation is reached [23]. The two populations having achieved sympatry and possessing 100% genetic isolation with respect to one another have reached the level of the species level based on the biological species defini- tion. Such populations will be recognized as different species taxa. But the process of specia- tion is not yet complete. (2) Evolution of reproductive isolation: As Speciation comes to an end when this period of rapid divergence noted earlier reporductive isolation differs from genetic isolation although these properties of spe- cies have been confused by most evolutionists. Both reproductive and genetic isolation are achieved by the possession of I I M, but not always the same ones. All II M achieve genetic isolation, but not all [1M achieve reproductive isolation. IIM (see Mayr [4] p. 92) differ considerably in their reproductive cost (see Bock [12] p. 35; [13] p. 708 W. J. Bock 43). If one examines the excellent classification of IIM presented by Mayr, the listed mechanisms vary from those possessing very low reproductive cost [mechanisms 1(1) and 1(b)] to those posses- sing very high reproductive cost [mechanism 2(d)]. Basically only seasonal, habitat and ethological isolation [mechanisms 1(a) and 1(b)] achieve com- plete reproductive isolation between species. The evolution of reproductive isolation may begin during the initial allopatric phase with the appearance of the original IIM and sometimes may even reach the 100% level by the time of the disappearance of the external isolating barrier. But this situation is rare. The more usual case is that given 100% genetic isolation between the two, newly sympatric populations, mutual selective de- mands by members of each species on the other will favor the evolution of additional 11 M with lower and lower reproductive costs until the two species become fully reproductively isolated. Genetic isolation does not increase, but remains at 100% as selective demands favors the evolution of I 1M of lower and lower reproductive costs. Re- productive isolation increases as do morphological differences between the two newly evolved spe- cies. This has been discussed as strengthening the IIM or strengthening the secondary isolating mechanisms ([4] pp. 552-3) which is valid, but is not sufficiently clear without making a sharp dis- tinction between genetic isolation and reproduc- tion isolation. This point is completely confused by many evolutionists (e.g., Coyne, et al. [21] p. 195) who state that they know of no earlier work- ers who claimed that ethological isolating mecha- nisns are selected for and perfected during specia- tion. This is in error because I have done so [12, 13], pointing out that such isolating mechanisms are selected for and perfected during the sympatric phase of speciation to increase reproductive isola- tion, not to increase genetic isolation. (3) Evolution of ecological isolation: At the time that newly evolved species re-estrablish sym- patry with the disappearance of the external isolat- ing barrier, they are generally rather similar ecolo- gically. Some ecological differences may have evolved during the allopatric phase of speciation, but if so, they are generally small [24, 25]. Lack discussed the several different possibilities of eco- logical interaction between newly evolved after the onset of sympatry and the evolutionary consequ- ences. Members of these species will generally compete ecologically with one another and exert mutual selective demands on one another once they overlap geographically and ecologically. The result in many cases will be increased divergence between the newly evolved species in those fea- tures associated with ecological requirements until the level of competition drops very low. Morphological divergence between descendant species results largely from the evolution of repro- ductive and ecological isolation as the result of mutual selective demands exerted by members of both species on one another following the estab- lishement of the secondary sympatric phase of speciation. The final end of the speciation episode occurs when morphological divergence between the two species ends or slows down to a very small rate with the disappearance of the mutual selective demands exerted by members of the two newly evolved species on one another. The fact that much to most of the differences between descen- dant species evolves during the secondary sympat- ric phase places doubt on such concepts as the genetic revolutions and punctuated equilibrium. SPECIES—THE TAXONOMIC UNIT A) Recognizing and cataloguing diversity in na- ture A primary task of the systematicst is to describe the observed diversity of living organisms in a system which will permit an efficient understand- ing of their atrributes. Many possible systems exist for arranging these organisms into primary units. The basis for recognizing species taxa has varied considerably during the history of biological re- search, but has generally reflected relatively close- ly the current theoretical species definition and changed with modifications in this definition. Cur- rently, the species definition chosen reflects most closely the evolutionary history of organisms. Nevertheless, application of the theoretical species definition to units (=species taxa) in nature has never been simple, and a considerable degree of judgment always had to be used. This was true The Species Concept 709 under the typological species definition as earlier systematists agonized over the question of how distinct a variety of a species had to be before it could be considered a distinct species. Under the biological nondimensional species definition, prob- lems developed immediately as species taxa were traced further and further in space and time from a single locality. Many arbitary decisions have to be made as no absolute standards exist for the recog- nition of species taxa. Hence, Mayr ([4] p. 19) coined the multidimensional species notion to deal with the delimitation of species taxa in space and time and to distinguish species taxa from the theoretical concept of the species. B) Species taxa Species taxa in sexually reproducting organisms recognized by taxonomists should correspond as closely as possible to the non-dimensional species definition, and hence consist of populations which actually or potentially interbreed (exchange genetic material) in nature. The actual or potential ability of interbreeding is judged on the dual bases of morphological similarities of organisms within a species taxon and of gaps between possible species taxa. Hence the degree of morphological similar- ity in a Series of actually interbreeding populations is used as the criterion for deciding on the species status of isolated populations in which actual inter- breeding cannot be observed directly. Such criteria must be used with great care and judge- ment as rigid standards cannot be established without running into the danger of a typological approach. Thus in various biochemical compari- sons of populations developed during the past two decades, one must avoid the temptation of estab- lishing fixed degrees of biochemical differences for subspecific status, specific status, etc., which can easily lead to typological thinking. The second problem is reaching decisions on the meaning and recognition of species taxa in asexual- ly reproducing organisms. Clearly such organisms are not arranged into species taxa similar to those found in sexually reproducing organisms. Yet systematists are faced with the problem of describ- ing and arranging the existing diversity observed in all organisms, be they sexual or asexual. Begin- ning at least two hundred years before the accept- ance of evolutionary theory, the primary units used in describing this diversity have been called species. And starting with Linnaeus in the mid- 18th century, the nomenclatural system used in describing this diversity has been based on species taxa. Novertheless, under evolutionary theory it is not possible to formulate a single theoretical spe- cies concept which is applicable to all organisms as has been advocated by a number of workers (e.g., [10]). Whatever species taxa are recognized for asexually reproducing organisms, these species taxa are fundamentally different from species taxa recognized in sexually reproducing organisms which are based on the biological species concept. It is only by tradition and convention that these primary units recognized by systematists in both sexually and asexually reproducing organisms are called species and are considered within systema- tics to be the same. Because no theoretical concept exists as the foundation for species taxa in asexually reproduc- ing organisms, no unambiguous foundation exists for criteria on which to recognize species taxa in these organisms. The best possibility is to reach some arbitrary decision on criteria for the arrange- ment of the diversity observed in individual asex- ual organisms into species taxa. Basically what systematists have done was to obtain a measure of the morphological differences observed in phe- notypic features of sympatric species taxa in sex- ually reproducing organisms which are associated with ecological isolation in these species. This degree of morphological difference was then used as the criterion for distinguishing species taxa in asexually reproducing organisms. Justification for this procedure comes from the argument that sympatric species taxa in sexual organisms serve to subdivide the local environment in such a way that the degree of variation in each species taxon is not too great to preclude the formation of ecologically viable offspring at each generation. One can argue, and completely correctly, that this degree of variation possessed in species taxa of sexual organ- isms is governed by the processes of recombination of genetic material during gamete formation and their fusion into zygotes, and that this degree of variation has nohting to do with the amount of variation contained in species taxa of asexual 710 W. J. Bock organisms. Therefore one could argue, and quite rightly, that applying a measure of ecological dis- tinctiveness observed in sexual species taxa to asexual organisms has no justification. But if this position is accepted, then no fundation exists for any standards on which to recognize species taxa in asexual organisms, and the recognition of such species taxa would become completely arbitrary. Therefore anything goes and the usefulness of species taxa in asexual organisms would disappear. The application of a standard based on ecological differences in sympatric sexual species taxa to asexual organisms results at least in species taxa which have ecological meaning and therefore con- siderable evolutionary meaning in terms of selec- tive demands exerted by members on one species taxon on others. C) Problems of geographic distribution Species taxa, both of sexual and of asexual organisms, often have a broad geographic distribu- tion, one that greatly exceeds the normal dispersal abilities of individuals during their lifetime. En- vironmental conditions generally vary over this geographical range resulting in diverse selective demands acting on local populations of the species and hence in geographic variation. Almost all species taxa with a broad geographic range show a certain (often considerable) degree of geographi- cal variation. Because the pattern of change in environmental factors is not gradual and because there may be gaps in the geographic distribution of species taxa, a particular species taxon may be subdivided recognizable, geographically arranged subunits, or subspecies (geographical races). Problems exist when these geographic subunits are not in direct contact and it is not possible to determine whether members of the different units can actually interbreed in nature. Or when the end forms of a long series of geo- graphic units meet in nature and do not interbreed, into as in the members of the Larus argentatus com- plex. The taxa argentatus—schistisagus—fuscus form a continuous ring of interbreeding popula- tions throughout the Holarctic Region with the two end fuscus— coexisting in northern Europe without interbreed- ing (see Mayr [2] pp. 180-185). Members of the forms—argentatus and end forms of the Rana pipiens complex in eastern United States do not produce viable zygotes in the laboratory [23, 42]. The central question is whether gene flow can actually or potentially occur in nature between the geographic populations of a species; hence the definition of the multi- dimensional species notion “as groups of popula- tions that actually or potentially interbreed with each other.” ([4] p. 19). Recently many cladists have advocated a species concept based on the idea that the species is the smallest taxonomically recognizable unit in nature (Cracraft, [17]). Under this concept, each of these geographical races, especially those in which geog- raphic gaps separate the ranges of neighboring races, whould be consideres as different species. This species concept should be rejected as it re- turns us to the typological species concept of the last century and earlier before the appreciation of geographic variation within species taxa and the development of the subspecies concept. It would result in species taxa of very different natures, such as sympatric species taxa and allopatric species taxa, which would greatly reduce the value of species taxa in any comparative analysis. D) Problems of chronological distribution Even more troublesome than geogrpahic dis- tribution is the time distribution of species taxa. The definition of the species concept indicates that the species has no time dimension and that there is no way to distinguish between different species along a phyletic lineage. Yet practical application of the multidimensional species notion to fossil material suggests that recognition of species taxa along a phyletic lineage is a necessary part of describing the diversity of organisms existing in the past and found in the fossil record. Because all the information a paleontologist can obtain from fossil specimens are morphological similarities and dif- ferences between them, decisions on the recogni- tion of species taxa must be based on analogy from the morphological variation in recent taxa. Hence the species taxa of fossil organisms are estimations at best, even for fossils of presumably sexually reproducing organisms. Little is gained from ex- tensive arguments about the limits of species taxa of fossil organisms. Generally a paleontologist will The Species Concept include fossil specimens over a time range in a single species taxon if they are morphologically similar, that is, do not shown any phyletic evolu- tionary change. Gaps in the fossil record are used as convenient points of delimitation between differ- ent species taxa arranged in time. Again, such recognition of species taxa in the fossil record is only a practical aspect of systematical study of fossils and must not be used as the basis for further development of theory. Application of the evolu- tionary or phylogenetic species concept to with similar fossil specimens over a reasonable time dimension is simply not valid. Moreover, it is difficult to impossible to use species taxa of fossil organisms to test theoretical concepts about evolu- tionary theory, such as support for ideas such as punctuated equilibrium. REFERENCES 1 Mayr, E. (1940) Amer. Nat., 24: 249-278. Mayr, E. (1942) Systematics and the origin of species, from the viewpoint of a zoologist., Col- umbia Univ. Press New York. 3 Mayr, E. (1957) Species concepts and definitions. pp. 1-22. Difficulties and importance of the biolo- gical species concept. pp. 371-388. In “The Species problem”. Ed. by E. Mayr, American Association for the Advancement of Science, Publ. 50, Washing- ton, DC. 4 Mayr, E. (1963) Animal species and evolution, Harvard Univ. Press, Cambridge, MA. 5 Mayr, E. (1982a) Evolution, 36: 1119-1132. 6 Mayr, E. (1982b) Process of speciation in animals. In “Mechanisms of speciation”. Ed. by C. Barigozzi, Alan R. Liss, New York, pp. 1-19. 7 Mayr, E. (1986) The species as category, taxon and population. In “Histoire du concept d’espéce dans les sciences de la vie”. Ed. by J. Roger and J. L. Fischer, Foundation Singer-Polignac, Paris, pp. 294-311. 8 Mayr, E. (1987) Biology and Philosophy, 2: 145- 166. 9 Mayr, E. (1988a) Biology and Philosophy, 3: 431- 441. 10 Mishler, B. D. and Brandon, R. N. (1987) Biology and Philosophy, 2: 397-414. 11 Szalay, F. S. and W. J. Bock (1991) Zeitschrift zool. Syst. Evolut-forsch., 29: 1-39. 12 Bock, W. J. (1979) Bulletin Carnegie Museum, 13: 20-69. 13 Bock, W. J. (1986) Species Concepts, Speciation 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 711 and Macroevolution. In “Modern Aspects of Spe- cies”. Ed. by K. Iwatsuki, P. H. Raven and W. J. Bock, Tokyo, University of Tokyo Press, pp. 31-57. Koopman, K. (1950) Evolution, 4: 135-148. Sibley, C. G. (1950) Univ. Calif. Publ. Zool., 50: 109-194. Simpson, G. G. (1951) Evolution, 5: 285-298. Cracraft, J. (1983) Current. Ornith., 1: 159-187. Hennig, W. (1966) Phylogenetic systematics. Uni- versity of Illinois Press, Urbana, IL. Paterson, H. E. H. (1985) The recognition concept of species. In “Species and speciation”. Ed. by E. S. Vrba, Transvaal Museum Pretoria, pp. 21-29. Raubenheimer, D. and Crowe, T. M. (1987) South African Jour. Sci., 83: 530-534. Coyne, J. A., Orr, H. A. and Futuyma, D. J. (1988) Syst. Zool., 37: 190-200. Mayr, E. (1988b) Toward a new philosophy of biology. Harvard Univ. Press, Cambridge, MA. Moore, J. A. (1957) An embryologist’s view of the species concept. In “The species problem”. Ed. by E. Mayr, Amer. Assoc. Adv. Sci., Publ., Washing- ton, DC. Lack, D. (1944) Ibis, 86: 260-268. Lack, D. (1949) The significance of ecological isolation. In “Genetics, paleontology and evolu- tion”. Ed. by G. L. Jepsen, E. Mayr, and G. S. Simpson, Princeton Univ. Press, Princeton, NJ, pp. 299-308. Mayr, E. (1947) Evolution, 1: 263-288. Mayr, E. (1992) Amer. Journ. Botany, 79: 222-238. Ghiselin, M. T. (1974) Syst. Zool., 23: 536-544. Ghiselin, M. T. (1987) Biology and Philosophy, 2: 127-143. Hull, D. (1975) History and Theory, 3: 253-274. Hull, D. (1976) Syst. Zool., 25: 174-191. Hull, D. (1978) Phil Sci., 45: 335-360. Mayr, E. (1954) Change of genetic environment and evolution. In “Evolution as a process”. Ed. by J. Huxley, A. C. Hardy and E. B. Ford, Allen and Unwin, London, pp. 157-180. Bock, W. J. (1991) Explanations in Konstruktions- morphologie and evolutionary morphology. In “Constructional morphology and evolution”. Ed. by N. Schmidt-Kitter & Vogel, Springer-Verlag, Heidelberg, pp. 9-29. Caplan, A. L. and W. J. Bock (1988) Biology and Philosophy, 4: 25-36. Mayr, E. (1949) Speciation and systematics. In “Genetics, Paleontology, and Evolution”. Ed. by G. L. Jepsen, E. Mayr and G. G. Simpson, Princeton University Press, Princteton, NJ, pp. 281-298. White, M. J. D. (1978) Modes of speciation. San Francisco, W. H. Freeman & Co. Carson, H. L. (1975) Amer. Nat. 109: 83-92. 39 40 712 Giddings, L. V., Kaneshiro, K. Y. and Anderson, W. W. (1989) Genetics, speciation, and the founder principle. Oxford Univ. Press, New York and Ox- ford. Eldredge, N. and S. J. Gould (1972) Punctuated equilibria: An alternative to phyletic gradualism. In “Models in Paleobiology”. Ed. by T. J. M. Schopf, 41 42 W. J. Bock Freeman, Cooper & Co, San Francisco, pp. 82-115. Dobzhansky, T. (1940) Amer. Nat., 74: 312-321. Moore, J. A. (1949) Patterns of evolution in the genus Rona. In “Genetics, Paleontology, and Evolution”. Ed. by G. L. Jepsen, E. Mayr and G. G. Simpson, Princeton University Press, pp. 315— 338. ZOOLOGICAL SCIENCE 9: 713-718 (1992) The Emetic Response of Urodele Amphibians Tomio Narrou! and RIcHARD J. WASSERSUG2 'Department of Biology, Shimane University, 1060 Nishikawatsu, Matsue, Shimane 690, Japan, and *Department of Anatomy and Neurobiology, Dalhousie University, Halifax, Nova Scotia, B3H 4H7, Canada ABSTRACT— Although it is well established that adult anurans can vomit, the one previous study with a urodele (Triturus) suggested that salamanders lack an emetic response. Using a salamandrid, Cynops pyrrhogaster, and a hynobiid, Hynobius nebulosus, we reexamined the possibility of emesis in urodeles. Apomorphine hydrochloride did not induce ejection of gastric contents in C. pyrrhogaster, while a small number of H. nebulosus showed gastric evacuation at high dosages of this emetic. Copper sulfate and antimonyl potassium tartrate, in contrast, caused vigorous vomiting in both species. Emetic behavior was similar in both species with both drugs. The urodeles ejected gastric contents only after substantive bending or twisting of their torso. This suggests that emesis in salamanders required elevation of intra-abdominal pressure by contractions of axial and abdominal wall muscles, as it does in anurans. However, unlike in frogs, the esophagus and stomach do not prolapse outside the mouth when © 1992 Zoological Society of Japan salamanders vomit. INTRODUCTION The ability to vomit is one of the more sophisti- cated protections that any animal can have to toxins in ingested food. Yet not all animals can vomit. Rats, for example, do not vomit, whereas cats do. Studies of vomiting in amphibians have established that adult anurans have a well de- veloped emetic response [1-5]. Compared to what is known about anurans, emesis in urodele amphi- bians has received little study. Mellinger [5] claimed that the newt Triturus (Salamandridae) did not vomit, when exposed to apomorphine, and that single observation has been cited [6] and miscited [7] several times over the last century. Given the broadly similar diets of anurans and urodeles, it would be surprising if Mellinger’s observation were true for all salamanders. Mellin- ger’s negative result for Triturus could be the result of the procedure that he used to induce and observe emesis. It would certainly be premature to generalize his conclusion to all urodeles. Given this background, the question remains: Do any Accepted April 13, 1992 Received January 22, 1992 " To whom correspondence should be addressed urodeles possess the ability to vomit and, if they do, how similar or different is their emetic be- havior to that of anurans? Our present study focuses on these questions. MATERIALS AND METHODS The salamanders Cynops pyrrhogaster (Sala- mandridae), weighing 1.9-6.3g, and Hynobius nebulosus (Hynobiidae), weighing 3.2—7.4 g, were used as experimental animals. They were field collected on the outskirts of Matsue, Japan, and maintained in the laboratory for up to three weeks on a diet of chopped earthworm, Eisenia foetida, fed to them twice weekly. Except for a small number of H. nebulosus, all animals were used only once. In order to determine if salamanders can vomit, three emetics were assessed: apomorphine hydrochloride (Sigma, St. Louis), copper sulfate (Wako Pure Chemical, Osaka) and antimonyl potassium tartrate (Wako Pure Chemical, Osaka). Apomorphine, which had been dissolved in 0.65% NaCl solution, was subcutaneously injected into the dorsum of the proximal hindlimb in dosages of 0.5, 2.0, 5.0, 20, 50, 100, 200, or 300 ug per g body weight (wet mass). Intraperitoneal injections of 714 T. NaAITOH AND R. J. WASSERSUG 50, 100 or 300 ug/g dosage were additionally attempted. Copper sulfate was given sub- cutaneously in the same way as apomorphine, or orally with catheter into stomach in a dosage of 0.2, 0.4, 1.0 or 2.0 mg per g body weight. Anti- monyl potassium tartrate was given orally in a dosage of 0.4 mg/g. The volume of apomorphine solution injected subcutaneously was 0.01 ml per g body weight with dosages equal to or below 50 ug/ g but 0.02, 0.04 and 0.06 ml at 100, 200 and 300 ug/g dosages, respectively. Copper sulfate and antimonyl potassium tartrate were given in a volume of 0.01 ml per g body weight. All salamanders were fed their normal labora- tory diet of chopped earthworm 30 to 60 min before testing with any emetic agent. Ejection of a piece of earthworm from the stomach was our criterion for vomiting. After the emetic was admin- istered, behaviors of the animals were directly observed or video-taped (with a JVC-GF-S1000H video camera) for up to three hours without inter- ruption. The container was checked again from ten to seventeen hours later, for the presence of vomitus. To correlate abdominal contractions with emetic behavior the proportions of the salamander torsos were measured with a mm ruler directly from their images on the video screen before, during, and after emesis. The length of torso was measured as the distance between the front and hind appendages. The width was the maximum width. Experiments were done in February and March for H. nebulosus and from March to April for C. pyrrhogaster. Room temperature ranged from 19.0 to 23.0°C. RESULTS In H. nebulosus, two of five animals vomited in 101.7 min and >180 min following subcutaneous injection of the 100 ug/g dose of apomorphine (Table 1). In C. pyrrhogaster, one of six animals vomited in 40.5 min at the 300 ug/g dosage, though it was uncertain if this was caused by apomorphine proper or by the high dose of saline. In this regard, one out of five control newts vomited after injection of 0.06 ml/g of 0.65% NaCl alone, but only after 180 min. At all other dosages of apomorphine, including the case of intraperitoneal injection, ejection of gastric con- tents did not occur. Nevertheless many specimens of both species, in particular C. pyrrhogaster, showed an atypically wide gaping of their mouth for a few seconds, plus strong lateral bending, ventral flexion, or complex twisting of the body following apomorphine injections at dosages higher than 5 ug/g. Furthermore, no subcutane- ous administration of 0.65% NaCl alone in a volume of 0.01 or 0.02 ml/g induced any trace of these same extreme postures in control specimens. Administration of copper sulfate, in contrast to apomorphine, dramatically induced vomiting in both species (see Table 1). An 0.65% NaCl admi- nistration was not effective. Cynops pyrrhogaster reacted slightly quicker than H. nebulosus to orally administered copper sulfate. At comparable dosages, C. pyrrhogaster tre- sponded much quicker to oral than subcutaneous administered copper sulfate. Hynobius nebulosus, however, did not show a similar difference in its response time to copper sulfate that could be associated with the route of administration. Although less extensively tested, antimony] potas- sium tartrate also proved to be an effective emetic (Table 1). Oral administration of this agent in- duced vomiting in all animals tested in both spe- cies. When emesis occurred, the behaviors associated with ejection of gastric contents were the same in both species and did not vary with the drug used or its pathway for administration. Ejection of gastric contents always occurred in a surge associated with one to several episodes of bending or twisting of the torso. In the extreme, the salamanders coiled into a circular, sigmoidal or reversed-sigmoidal posture (Figs. 1, 2 and 3). During the contortions associated with emesis the torso of C. pyrrhogaster shortened by an average of 12.8% (N=5, range 7.3-15.8%). In contrast, the maximum width of the body increased on average by only 3.6% (N= 5, range 0-8%). Thus the animals were becoming shorter without a proportionate expansion of cross-sectional area to maintain constant volume. Prolapse of the esophagus or stomach outside the mouth was never detected. The actual vomiting act took from about one to eighty sec depending Emesis in Urodeles 715 TABLE 1. The responses of salamanders to various emetic drug treatments Cynops pyrrhogaster Hynobius nebulosus No. vomited Reaction No. vomited Reaction No. tested times (min)* No. tested times (min)* Apomorphine HCI** 0.5 ug/g S.c. 0/6 2 0/6 5 0/7 0/5 20 0/8 50 0/6 0/5 100 0/6 2/5 101.7, >180 200 0/6 300 1/6 40.5 Copper Sulfate 0.2 mg/g s.c. 1/3 >180 3/3 38.9+11.8 1.0 mg/g s.c. 2/3 62.9, >180 2.0 mg/g s.c. 3/3 42.0 (n=2), >180 0.2 mg/g p.o. 3/3 26.5+3.2 4/4 31.4+10.6 0.4 mg/g p.o. 2/2 42.1, 6.6 2.0 mg/g p.o. 3/3 4.4+1.9 3/3 13.1+5.0 Antimony! potassium tartrate 0.4 mg/g p.o. 4/4 1.6+0.3 (n=3), TG 138.0+19.5 (n=6), >180 (n=1) >180 (n=1) * Mean+SE given for all animals that responded in less than 180 minutes. ** Intraperitoneal injections of 50, 100, or 200 ug/g did not induce vomiting in any of the 2-3 C. pyrrhogaster which received them. Note: s.c.=subcutaneous, p.o.=per oral. *** Includes 4 animals previously used in apomorphine tests. 3cm Fic. 1. Bending posture during vomiting in salamanders. Left: C. pyrrhogaster; right: H. nebulosus. Arrow indicates vomitus just appearing. Left and right correspond to Fig. 2, D and Fig. 3, C, respectively. (Images taken directly from videofilm using Sony Multiscan Video Printer UP-930). 716 T. NaAITOH AND R. J. WASSERSUG 444 O 3.5 2cm Fic. 2. Vomiting behavior in C. pyrrhogaster induced by orally administered copper sulfate at a dose of 2.0 mg/g body mass. A . Resting state, 50.0 sec before the start of the vomiting act. B through F=1.2, 3.5, 4.41, 4.44, 4.8 sec, respectively, after the start of vomiting. Vomitus (stippled) outside the mouth appears first in D to the left of the head. It is completely ejected by F. Note the characteristic bending of the body in D through E. Drawn from a videofilm. -—4.0 sec on the specimen and the concentration of the emetic administered. Often episodes of weak bending or twisting were repeated before culmi- nating in vomiting and gastric content ejection. DISCUSSION Apomorphine and copper sulfate are well estab- lished emetics in mammals [9]. The former is known to act centrally, while the latter, centrally and peripherally depending on how it is adminis- tered [10]. Antimony! potassium tartrate is also an effective emetic in mammals [9, 11]. In our previous study surveying the vomiting response in ten species of anurans [8], Xenopus laevis was typical for frogs in its sensitivity to apomorphine hydrochloride. Twenty ug/g and 50 ug/g were effective in Xenopus, whereas Rana Japonica and Rana ornativentris were less respon- sive to apomorphine even at dosages of 50 and 100 ug/g body mass. In contrast only two urodele studied here ejected its gastric contents after apo- morphine administration at these dosages, though gaping of the mouth and bending of the body commonly occurred in both species. Thus, these amphibians are an order of magnitude or more less sensitive to this agent than are carnivorous mam- mals on a zg per g body mass basis. However, the important point is that urodeles unequivocally vomit in response to relatively low doses of copper sulfate and antimonyl potassium tartrate. Mellin- ger [5], who seems to be the only author to previously explore emesis in a urodele, concluded that newts (in his case Triturus) could not vomit. Our present results with another salamandrid genus, as well as a hynobiid, do not support this conclusion. The response of urodeles to both copper sulfate Fic. 3. Vomiting behavior in H. nebulosus induced by copper sulfate administered orally at a dose of 0.4 mg/g body mass. A. Resting state, 4.0 sec before the start of the vomiting act. B through H=0.8, 31.3, 32.8, 35.0, 63.1, 69.4 and 72.3 sec, respective- ly, after the start of vomiting. Vomitus (stippled) first appears in C and is completely ejected from the mouth by H. Sigmoidal posture (C) and twisting of the body (C to E) are shown. In D, abdominal contraction is marked. Drawn from a videofilm. Emesis in Urodeles WAG and antimony! potassium tartrate leads us to tenta- tively suggest that most, if not all, amphibians possess an emetic capability. This would be consis- tent with the fact that salamanders and frogs both feed on relatively large and potentially toxic prey, with little or no oral processing. In mammals, carnivores that swallow large boluses of unmasti- cated food have a strong emetic response, whereas rodents, which extensively process and reduce food within their oral cavity exhibit a taste aver- sion response (see Fox, 1990 and other references cited therein [12]), but not the ability to vomit. It remains to be seen whether caecilians, which feed on quite different prey and in a quite different manner from frogs and salamanders [13], can vomit. Both subcutaneous injection and oral adminis- tration of copper sulfate were potently effective in inducing vomiting. The neurobiology of emesis in amphibians remains unstudied, but on pharmaco- logical grounds it seems similar to that of mammals described by Wang and Borison [10]. One of six newts vomited after subcutaneous injection of apomorphine at a very high (300 g/g) dose. However, this may not be the effect of apomorphine, since vomiting was also observed in one instance after a simple injection of the same volume of NaCl solution. It should be noted that relatively low doses of apomorphine, but not saline, induced marked behaviors such as twisting of the body and gaping of the mouth, without ejection of gastric contents. If such behaviors are indicative of retching, the neural mechanisms of apomorphine-induced vomiting may differ slightly from that of vomiting induced by systemically administered copper sulfate. The most conspicuous feature of vomiting in salamanders is the preparatory bending and twist- ing of the torso. These axial movements, in conjunction with tonic contraction of the muscles of the abdominal wall, elevate intra-abdominal pressure. Our observations suggest that salaman- ders raise intra-abdominal pressure by bending and twisting movements of the torso which com- press the abdominal cavity. Previous studies have shown that elevation of intra-abdominal pressure is necessary to eject stomach contents in frogs [2— 4]. Similar elevation of abdominal pressure seems necessary for emesis in urodeles as well. In this regard, emesis in amphibians is similar to that of mammals and is not a visceral response as simplis- tic as (reverse) antiperistalsis. Prolapse of stomach, however, which is common during emesis in anurans, was not observed in urodeles. This is likely due to the gross structural differences, such as the proportionately longer and narrower esophagus of salamanders and their longer bodies overall. This work was conducted following the “Guid- ing Principles for the Care and Use of Animals in the Field of Physiological Science” set by the Physiological Society of Japan. ACKNOWLEDGEMENTS We thank Mr. Masahiro Kirihara and Miss Motoko Matuura for maintaining the animals in the laboratory and their help in conducting the experiments. This work is part of a joint program on “Visceral Functions in Amphibians” supported by grants from the International Scientic Research Program of the Ministry of Education, Science and Culture (Japan) and the Natural Science and Engineering Research Council (Canada). REFERENCES 1 Tokita, K., Iwasaki, S. and Yui, F. (1953) On the Opening motion of mouth and vomiting of frogs after the digitalis preparations and on their mechanisms. Folia Pharmacol. Japon., 49: 138-142. 2 Hukuhara, T., Naitoh, T. and Kameyama, H. (1973) A peculiar phenomenon, the prolapsus of the esophagus-stomach, observed in the vomiting of the frog. Jpn. J. Smooth Muscle Res., 9: 1-8. 3 Naitoh, T., Nakai-Kameyama, H. and Hukuhara, T. (1981) The effect of denervation of abdominal musculature on the prolapse of esophagus and sto- mach in the vomiting of frogs. Mem. Fac. Sci. Shimane Univ., 15: 57-63. 4 Naitoh, T., Wassersug, R. J. and Leslie, R. A. (1989) The physiology, morphology, and ontogeny of emetic behavior in anuran amphibians. Physiol. Zool., 62: 819-843. 5 Mellinger, C. (1881) Beitrage zur Kenntniss des Erbrechens. Arch. Ges. Physiol. Pfltiger’s, 24: 232— 245. 6 Reeder, W. G. (1964) The Digestive System. In “Physiology of the Amphibia”. Ed. by J. A. Moore, Academic Press, New York, N.Y. pp. 99-149. 7 Nigam, H. C. (1977) Vomiting response in the Indian frogs Rana tigrina (Daud) and Rana cyanoph- 10 718 lyctis (Boulenger). Curr. Sci., 46: 791-792. Naitoh, T., Imamura, M. and Wassersug, R. J. (1991) Interspecific variation in the emetic response of anurans. Comp. Biochem. Physiol., 100C: 353- 359. Borison, H. L. and Wang, S. C. (1953) Physiology and pharmacology of vomiting. Pharmacol. Rev., 5: 193-230. Wang, S. C. and Borison, H. L. (1952) A new concept of organization of the central emetic mechanism: Recent studies on the sites of action of apomorphine, copper sulfate and cardiac glycosides. T. NAITOH AND R. J. 11 12 13 WASSERSUG Gastroenterology, 22: 1-12. Koppanyi, T. (1930) Studies on defecation, with special reference to a medullary defecation center. J. Lab. Clin. Med., 16: 225-238. Fox, R. A. (1990) Investigating motion sickness using the conditioned taste aversion paradigm. In “Motion and Space Sickness”. Ed. by G. H. Cramp- ton, CRC Press, Boca Raton, pp. 105-121. Bemis, W. E., Schwenk, K. and Wake, M. H. (1983) Morphology and function of the feeding apparatus in Dermophis mexicanus (Amphibia: Gymnophiona). Zool. J. Linn. Soc., 77: 75-96. ZOOLOGICAL SCIENCE 9: 719-724 (1992) Origin of Serotonergic Innervation of Olfactory Bulbs in the European Hamster, Cricetus cricetus: An Autoradiographic Study Using [°H]5-HT Retrograde Labelling JEAN Luc Miro!, SirviA ARANEDA~ and BERNARD CANGUILHEM Institute de Physiologie, Faculté de Médecine, Université Louis Pasteur, 67085 Strasbourg Cedex, France and URA CNRS 1332, and *Départment de Cytologie, Institut des Neurosciences, Université P. et M. Curie. 7, quai St-Bernard. 75252 Paris cedex 05, France ABSTRACT — Serotonergic afferents to the olfactory bulb of the European hamster, Cricetus cricetus were investigated by autoradiography following [*H]5-HT retrograde labelling techniques. Injections of [PH]5-HT into the main olfactory bulb resulted in radiolabelling of some nerve terminals in the granular and glomerular layers of the olfactory bulb, and in cell bodies of the median raphe and dorsal raphe. In the dorsal raphe, labelled cells were located in the dorso-ventral and lateral part of the nuclei. In the median raphe, labelled neurons were observed more caudally in the medial region. In this nucleus, the number of radiolabelled cells was a third of those in the raphe dorsalis. A few scattered radiolabelled cells were observed in the periaqueductal gray. The number of radiolabelled cells in both the raphe nuclei was not changed when animals were treated with a monoamine oxidase (MAO) inhibitor, Clorgyline. However, the intensity of labelling was increased in the treated animals. Treatment with 5,7-dihydroxytryptamine abolished the [°H]5-HT retrograde labelling in the olfactory bulb, as well as in the raphe cell bodies. It is concluded that in the Cricetus cricetus the granular and glomerular layers of © 1992 Zoological Society of Japan the ofactory bulbs receive a serotonergic innervation from the median and dorsal raphe. INTRODUCTION During the cold season hibernators display bouts of deep hibernation characterized by a low body temperature and a reduced basal metabolic rate. Hibernation is interrupted by spontaneous arousal and a return to an euthermic body temperature for a few hours or days. The time course of entrance into hibernation has only been studied in detail in few species by monitoring the body or hypotha- lamic temperature and recording electrical activ- ity. From these studies two main results have been obtained. Firstly, the decline in body or brain temperature is not a passive process but a control- led one [1]. Secondly, entrance into hibernation is thought to be an extension of normal slow wave Accepted April 21, 1992 Received November 12, 1991 1 Deceased on May 27th, 1992 sleep [2]. In certain species it seems that entrance into hibernation is achieved by way of.serotonergic system (see ref. in [3]). For example in the European hamster, Cricetus cricetus, serotonin (5- HT) concentrations in differents brain areas, espe- cially in the limbic system, were lower in hibernat- ing animals than in aroused animals in winter [4]. In addition, electrolytic or pharmacological lesion of the median but not of the dorsal raphe nucleus [3] as well as total ablation of the olfactory bulbs (OB) [5] has led to a suppression of hibernation. The evidence thus strongly suggests an anatomical connection between the raphe nuclei and the OB of the European hamster. Recently, retrograde axonal transport with triti- ated transmitters has been developed as a specific autoradiographic method for tracing pathways in the CNS [6]. Using this technique [*H]5-HT is specifically taken up at serotonergic nerve termi- nals and then transported by retrograde axonal 720 J. L. Miro, S. ARANEDA AND B. CANGUILHEM transport to their corresponding cell bodies [7-9]. The specificity of this methodology was dependent upon the concentration of injected [(PH]5-HT, and has been confirmed using combined studies of (SH]5-HT retrograde labelling and immuno- cytochemistry for serotonin [10]. Consequently, [(>H]5-HT retrograde radiolabelling is now re- garded as an important technique for studying the projections of serotonergic neurons of the mid- brain and medullary raphe in the rat [11]. In view of the involvement of the median raphe, olfactory bulb and serotonergic system in the process of entrance into hibernation, we have investigated the origin of the serotonergic innervation of the OB in the European hamster using [*H]5-HT retrograde labelling. It has been previously reported that the uptake [>H]amines by central catecholaminergic fibres is decreased in hibernating animals [12]. This result and those reported previously for the serotonergic system [3, 4, 14-16] suggest that the activity of the serotonergic neurons could be different between winter and summer hamsters. In order to test this hypothesis, we also designed experiments using retrograde axonal transport with [°H]S-HT in winter arising hamsters. MATERIALS AND METHODS Adult male European hamsters (Cricetus cri- cetus) were caught at the end of the hibernation period (April-May) in the field surrounding Stras- bourg (France). They were kept in individual cages with free access to food and water. For the summer experiments 19 hamsters were housed for one month (July) in a 12h light:12h dark cycle and constant temperature (20+1°C) before surgi- cal manipulation. The summer animals were di- vided into three groups. The first one was a control groups (n+5) which only received [?H]5- HT injections. The second group (n+5) was pretreated with clorgyline, a MAO inhibitor (10 mg/kg), 30 min before the [*H]5-HT injections in order to prevent the degradation of 5-HT into 5-hydroxyindole-3-acetic acid. The third group (n +9) was injected with 5,7-DHT (5,7-dihydro- xytryptamine creatinine sulfate, Regis Chemicals, 8 wg in 2 yl of saline with 0.1% ascorbic acid) which selectively destroys serotonergic nerve ter- minals. The 5,7-DHT solution was injected into the OB at the same stereotaxic cooredinates as the [>H]5-HT injections six days before. For the winter experiments, 4 animals were transferred into a cold room (7+1°C) with total darkness for one month (January) before the ex- periment. During this time hibernation was check- ed daily by visual inspection. Thirty min before the experiments, the hamsters were pinched in order to produce an arousal so that when the [H]S-HT injections began all the animals were normothermic. Hamsters from these experiments were not treated with any drugs before the [°H]5- HT injections. For the injections of [H]5-HT, animals were anaesthetised with Nembutal (50 mg/kg) and were placed in a David Kopf stereotaxic instrument. A mechanical obstruction of the olfactory diverticle was then carried out (AP=15, L=1, 2 and DV= 3, 4) in order to avoid difussion of the radioactive tracer as previously described in the rat by Ara- neday, ctqysalgauilail: [(H]-serotonin [°H]5- hydroxytryptamine creatinine sulfate, Amersham, 15 Ci/mol) was injected 20 wCi, 5x 10* M) in one or both OB (AP=17, L=1 and DV=3) over a Fic. 1. injections into the olfactory bulb. Autoradiograms of the olfactory bulb (A), dorsal raphe (B, C) and median raphe (D) following [*H]S-HT A) The PHI5-HT injection site in the olfactory bulb was located in the granular layer, where an intense autoradiographic reaction was observed which decreased in the plexiform and glomerular layers. Scale bar, 80 pm. B and C) Labelled cell bodies in the dorsal raphe, 24 hours after injections into the olfactory bulb. They were situated ipsilaterally when the [*H]5-HT was injected into one olfactory bulb. Some radiolabelled cells were located between both medial longitudinal bundles (mlb). Scale bar, 30 “m. D) In the median raphe, labelled cell bodies were observed in the mediolateral part of the nuclei and presented a dense autoradiographic labelling. Scale bar, 45 zm. Note that the silver grains were mainly distributed in the cytoplasm. The area indicated by arrows is enlarged in the inset. Scale bar, 20 um. Serotonin and Olfactory Bulbs in Hamster 722 J. L. Miro, S. ARANEDA AND B. CANGUILHEM period of 25 min. Twenty four hours later, animals were re-anaesthetised and the brains fixed by intra-aortic perfusion using glutaraldehyde (3.5%) in phosphate buffer (0.2 M, pH7.3). The whole brain was removed, dehydrated in ethanol and embedded in parafin-paraplast. Sections (7 ~m thick) were cut every 50 um. Then coated by dipping in Ilford KS emulsion (diluted 1:1), ex- posed for 10-30 days at 6°C, developed in phenisol (Ilford) and stained with cresyl violet. Light microscope observations were made using both bright and darkfield illumination. Radiolabelled cells were counted in the midbrain raphe and the Bartlett’s test for homogeneity of variance was used to determine statistical significance. RESULTS Injection site (H]5-HT injections into the OB of summer hamsters were characterised by a central zone of high density of silver grains confined to the granu- lar layer of the olfactory bulb. The intensity of this autoradiographic reaction gradually decreased towards the glomerular layer (Fig.1A), and caudally towards the anterior olfactory nucleus. A dense accumulation of silver grains was observed in the granular, plexiform and glomerular layers within 1 mm of the injection site. Accumulation of silver grains was restricted to the neuropile and not to the cellular components. The mitral tufted and granular cells were devoid of silver grains. Following 5,7-DHT treatment, a diffuse auto- radiographic reaction persisted but the accumula- tion of silver grains was abolished in all of the olfactory layers. Raphe cell bodies Dorsal and median raphe nuclei of the Euro- pean hamster are located in the mesencephalon ventral to the aqueduct. As has been described for Fic. 2. Schematic representation of dorsal (DR) and median raphe (MR) and the distribution of labelled cell bodies (dots) detected by autoradiography following retrograde axonal transport with [SH]5-HT. Brain sections were taken at 150 ~m intervals to be drawn using a camera lucida. Brain regions were identified with reference to the rat atlas of Paxinos and Watson (18). 3n: oculomotor nerve, mlb: medial longitudinal bundle, xscp: decussation of the superior cerebellar peduncle, CG: central gray, Aq: aqueduct, scp: superior cerebellar peduncle, TN: tegmental nucleus, ml: medial lemniscus, 5 Mo: motor trigeminal nucleus, LTN: lateral tegmental nucleus. Serotonin and Olfactory Bulbs in Hamster 723 the rat dorsal raphe [12], we also distinguished four principal regions in the dorsal raphe of the hamster: the ventromedian, the dorsomedian and the lateral groups, and the pars caudalis. The median raphe was situated ventrally to the supe- rior cerebellar peduncle in the mediolateral posi- tion. At the rostro-caudal levels, it extended from the interpenducular nucleus to the caudal levels of the tegmental nucleus (Fig. 2). Twenty four hours after the [>H]5-HT injections into the OB, a number (n=68) of dorsal raphe cell bodies showed specific autoradiographic labelling (Fig. 1B-D). Silver grains were distributed in the cytoplasm and sometimes extended to the de- ndrites but not to the nuclei. Radiolabelled cells were observed in the ventral position and scattered among many non-radiolabelled neurons. At the rostral level of the superior cerebellar peduncle, a moderate number of radiolabelled cells were observed in the medial region of the dorsal raphe. Caudally, there were more radiolabelled cells dis- tributed dorsal to and between the medial longitu- dinal bundle (Fig. 2). A few labelled cells were also observed in the adjacent central gray. In the median raphe, the largest number of retrogradely labelled cells (n=20) were observed in the ventral position, and situated more caudally to the labelled cells of the dorsal raphe (Fig. 2). In both raphe nuclei, radiolabelled cell bodies were situated mainly ipsilaterally when [*H]5-HT was injected into one OB (Fig. 1C). The locus coe- ruleus neurons were not radiolabelled. Results from the winter experiments indicate that the number of radiolabelled cells in the dorsal (n=45+20) and median raphe (n=17+9) were not statistically different from the summer hams- ters. The number of radiolabelled neurons in the dorsal raphe and median raphe of summer ham- sters was 50+18 and 17+9, respectively. DISCUSSION Injection of radiolabelled serotonin into the OB of European hamsters is followed by a selective accumulation of radioactivity in some cells of the dorsal and median raphe. This autoradiographic labelling depends upon uptake of the radioactive amine by the serotonergic nerve terminals since their selective destruction by 5,7-DHT did not result in a retrograde accumulation in the corres- ponding cell bodies. Moreover, the lack of label- ling of the noradrenergic neurons of the locus coeruleus indicated that this process was selective to 5-HT neurons. These results on the selectivity of labelling serotonergic projections are in good agreement with those obtained in the rat OB- raphe connections [10]. Using a concentration of 10* M [°H]5-HT, we have also previously demon- strated specific radiolabelling using a combination of autoradiography after [*H]5-HT retrograde transport and immunocytochemistry for serotonin in both the olfactory and spinal serotonergic pro- jections [11]. In the present study, inhibition of the enzyme monoamine oxidase by clorgyline did not affect the number of labelled neurons indicat- ing that the [*H]5-HT taken up by the terminals could be quickly bound or metabolized to another form in order to be transported retrogradely by axons as it has been described in the rat [9]. The present findings thus demonstrate a sero- tonergic innervation of the hamster OB by the dorsal and median raphe. Radiolabelling was located in the dorsomedian, ventromedian and lateral serotonergic cell groups in the dorsal raphe, and caudally in the median raphe. Moreover, radiolabelled neurons of both raphe innervate the ipsilateral OB. These results are in accordance with previous horseradish peroxidase anterograde- retrograde tracing [17] and [°H]5-HT retrograde autoradiographic studies in the rat [8, 11]. Although further studies will be required to deter- mine the precise anatomical site of serotonergic innervation of hamster OB structures, it appears likely from our results that many of the termina- tions are in the granular and glomerular layers. In the dorsal raphe, the distribution of sero- tonergic radiolabelled cells was mainly in the ven- tral and dorsal position which seems quite similar to that observed in the dorsal raphe of the rat [12]. In the median raphe, however, labelled neurons were located more caudally, and were not co- distributed in the same frontal section of the dorsal labelled cells as has been previously observed in the rat using the same methodology [11]. Our previous experiments have suggested that in the European hamster only specific groups of 724 serotonergic neurons in a small part of the median raphe are involved in the process of entrance into hibernation which may or may not be associated with the OB [3, 5]. Our present results lend some support to this assumption. If our hypothesis were true, however, we would expect to find a greater number of labelled cells in the summer than in the winter animals indicating increased accumulation of [PH|5-HT in serotonergic cell bodies before entrance into hibernation. This was not the case and further studies should thus be undertaken using true hibernating hamsters in order to address this question. In conclusion, we have demonstrated a sero- tonergic innervation of the hamster OB by a well defined population of neurons from the median and dorsal raphe. REFERENCES 1 Heller, H. C. (1979) Hibernation: neural aspect. Ann. Rev. Physiol., 41: 302-321. 2 Walker, J. M., Glotzbach, S. F., Berger, R. J. and Heller, H. C. (1977) Sleep and hibernation in ground squirrel (Citellus spp): electrophysiological observations. Am. J. Physiol., 233: R213—R221. 3 Canguilhem, B., Miro, J.-L., Kempf, E. and Schmitt, P. (1986) Does serotonin play a role in entrance into hibernation? Am. J. Physiol., 251: R755—R761. 4 Canguilhem, B., Kempf, E., Mack, G. and Schmitt, P. (1977) Regional studies of brain serotonin and norepinephine in the hibernating, awakening or active European Hamster, Cricetus cricetus, during winter. Comp. Biochem. Physiol., 57C: 175-179. 5 Miro, J.-L., Canguilhem, B. and Schmitt, P. (1980) Effects of bulbectomy on hibernation, food intake and body weight in the European hamster, Cricetus cricetus, Physiology and behaviour, 24: 859-862. 6 Cuenod, M., Bagnoli, P., Beaudet, A., Rustioni, A., Wiklund, L. and Streit, P. (1982) Transmitter specific retrograde labelling of neurons. In “Cytochemical Methods in Neuroanatomy”. Ed. by V. Chan-Palay and S. L. Palay, Alan R. Liss, New York, pp. 17-44. 7 Araneda, S., Bobillier, P., Buda, M. and Pujol, J. F. (1980) Retrograde axonal transport following injection of [*H]-serotonin in the olfactory bulb. I: 10 11 12 13 14 15 16 17 18 J. L. Miro, S. ARANEDA AND B. CANGUILHEM Biochemical study. Brain Res., 196: 405-415. Araneda, S., Font, C., Pujol, J. F. and Bobillier, P. (1983) Retrograde axonal transport after radioac- tive indole injections into the olfactory bulb. An autoradiographic study. Neurochem. Inter., 5: 741- 750. Araneda, S., Mermet, A., Buda, M., Bobillier, P., Pujol, J. F. (1984) Retrograde axonal transport after radioactive serotonin injections into the olfac- tory bulb. A biochemical analysis of radioactive material. Neurochem. Inter., 6: 27-39. Araneda, S., Magoul, R. and Calas, A. (1989) Tracing specific transmitter pathways in the rat CNS: combination of [*H]-serotonin retrograde labelling with immunocytochemical detection of en- dogenous transmitters. J. Neurosci. Methods, 30: 211-218. Magoul, R., Oblin, A., Calas, A. and Araneda, S. (1988) Serotonergic projections to the spinal cord but not those to the olfactory bulb also contain substance P. A combined immunocytochemical and autoradiographic study following retrograde axonal transport of [*H]-serotonin labeled products. Neurosci., 26: 959-969. Steinbucsh, H. W. M. and Niewenhuys, R. (1983) The raphe nuclei of the rat brainstem: a cytoar- chitectonic and immunocytochemical study. In “Chemical Neuroanatomy”. Ed. by P. C. Emson, Raven Press, New York, pp. 131-207. Faure, A. et Calas, A. (1976) Etude radioautog- raphique de l’incorporation in vitro de noradrena- line tritiée dans des fibres catécholaminergiques centrales chez I’hérisson actif et en hibernation C. R. Soc. Biol. 171: 136-140. Kempf, E., Mack, G., Canguilhen, B. and Mandel, P. (1978) Seasonal changes in the levels and the turnover of brain serotonin and noradrenaline in the European hamster kept under constant environ- ment, Experientia, 34: 1032. Miro, J.-L., Canguilhen, B. and Schmitt, P. (1979) Sérotonine cérébrale et sommeil hibernal, Brain Res., 170: 387-391. Miro, J.-L., Canguilhem, B. and Koch, A. (1986) Brain serotonin levels and metabolism in the non- hibernating bulbectomized European hamster. Life Sci., 39: 2551-2555. Mclean, H. J. and Shipley, M. (1987) Serotonergic afferents to the rat olfactory bulb: I. Origins and laminar specificity of serotonergic inputs in the adult rat. J. Neurosci., 7: 3016-3028. Paxinos, G. and Watson, C. (1982) The rat brain in stereotaxic coordinates. Academic Press, Sydney. ZOOLOGICAL SCIENCE 9: 725-731 (1992) Photoperiodic Control of the Determination of Two Different Seasonal Diphenisms of the Asian Comma Butterfly, Polygonia c-aureum L. KATSUHIKO ENDO, SATOSHI UENO, MiTSUNORI MATSUFUJI and YOSHIKI KAKUO Environmental Biology Laboratory, Biological Institute, Faculty of Science, Yamaguchi University, Yamaguchi 753, Japan ABSTRACT—The Asian comma butterfly, Polygonia c-aureum L. exhihits two different seasonal diphenisms in the morph and adult reproductive activity, both of which are governed by photoperiod and temperature existing during the post-embryonic development. In the complete and 8L-16D asymmetrical skeleton photoperiods at 20°C or 25°C, they formed similar response curves in the incidences of autumn morph and reproductive diapause in adult females. They required 4 and 5 16L-8D (long-day) days for achieving 50% occurrence of vitellogenic and summer morph adults, respectively. Critical stages of 16L-8D to 8L-16D (short-day) transfer obtained on 50% preventions of vitellogenic and summer morph female adults were located on day 0 and on day 2 of the 3rd instar. In contrast, the critical stages of adverse photoperiodic transfer greatly differed between these two kinds of seasonal diphenisms. The stages were day 3 of the Sth instar in the seasonal morphs and day 3 of pupae in the reproductive activity. Physiological mechanisms underlying photoperiodic control in P. c-aureum has been previously indicated to involve two neuroendrocrine systems secreting summer morph-producing hormone and/or a factor activating corpora allata in adults. Differential responsiveness of these two neuroendocrine systems to photoperiods may reflect on the difference in the incidences of the seasonal morphs and reproductive activity. In addition, the photoperiodic sensitive stage seems to be different with respect to © 1992 Zoological Society of Japan these two different seasonal diphenisms. INTRODUCTION In a number of insects, diapause and seasonal polyphenism are governed by environmental fac- tors, such as photoperiod and temperature they are subjected to after hatching [1-5]. Pupal diapause and spring morphs of the swal- lowtail, Papilio xuthus L., are induced by short days at 20°C. They become photoperiodically sensitive at around the first larval ecdysis [6]. The photoperiodic sensitiveness is lost at the middle of the 5th instar. In the photoperiod-sensitive stage, they require seven 8L-16D (SD: short-day) days for the 50% incidences (or require five 16L-8D (LD: long-day) days for the 50% preventions) of pupal diapause and spring morph development at 20°C [6]. The Asian comma butterfly, Polygonia c- aureum L., exhibits two different seasonal diphe- Accepted April 28, 1992 Received March 11, 1992 nisms, 1.€. seasonal morph and adult reproductive activity, both of which are determined by photo- period and temperature they are exposed to after hatching. Under LD at 20°C, P. c-awreum females developed into vitellognic/summer morph adults, whereas those held in SD at 20°C developed into reproductive diapause/autumn morph adults [7]. Physiological mechanism underlying photo- periodic control of these two different seasonal diphenisms involves cerebral factors secreted from corpus cardiacum and/or corpus allatum in the early pupal stages [7-9]. Critical stages at which summer-morph producers (long-day pupae) re- quire no more brain and/or nervi corporis cardiaci (I+1I) for summer morph and vitellogenic adult development lie at about 32 hr after larval-pupal ecdysis at 20°C in P. c-aureum [10, 11]. Our present study was initiated to determine the number of long-day required for the occurrence of summer morph and vitellogenic female adults of P. c-aureum. Then, the study was extended to deter- 726 K. Enpo, S. mine the beginning and the end of photoperiod- sensitive stages in the induction (or prevention) of these two different seasonal diphenism. MATERIALS AND METHODS Animals P. c-aureum collected in the Yama- guchi City was used. Eggs and larvae were held in transparent plastic containers of either 695 cm or 19x13xX5cm and were exposed to complete photoperiods, whose photophase was changed from 2hr (2L-22D) to 20hr (20L-4D) at 2-hr intervals. Rearing containers were placed in cabinets with constant temperature (20°C or 25°C) and were illuminated by two 20-W white fluores- cent tubes which were controlled by a 24-hour time-switch. In the light period, light intensity was at about 500 lux to the rearing containers. Larvae were fed on fresh leves of Humulus japonicus provided daily. Classification of seasonal morphs Summer- morph and autumn-morph females exhibit clear differences in background color of the wings. The wing background color of summer morph females is dark yellow on both sides, whereas that of autumn morph females is reddish-brown on the dorsal side and dark brown on the ventral side [12]. Seasonal morph classification was made only in females on the basis of the wing background color. Criterion for adult reproductive diapause When fed on 10% sucrose four days after emergence at 25°C, an vitellogenic (reproductive non-diapause) females accumulate green yolk in oocytes and/or produce eggs. In contrast, females entering repro- ductive diapause do not accumulate yolk at the same period. Female adults were caught on the day of emerg- ence, fed on 10% sucrose for 4 days at 25°C and dissected in 0.9% NaCl to see whether the oocytes in the ovaries accumulated yolk or not. Female adults accumulating green-colored yolk in oocytes were classified as vitellogenic, whereas others were classified as diapausing. UENO et al. RESULTS Fluctuations of the incidences of autumn morph and reproductive diapause female adults in the complete photoperiods at 20°C and 25°C To assess how the incidences of autumn morph and reproductive diapause adults fluctuate, groups of 150-200 larvae were subjected to complete photoperiods whose photophase was changed de- pending on larval groups from 2 hr (2L-22D) to 20 hr (20L-4D) at 2-hr intervals at 20°C or 25°C. Photoperiodic response curves were obtained on the basis of the percentages of autumn morph and reproductive diapause adult females at 20°C and 25°C (Fig. la and b). 100 Ape OE a ) rs) 7) a) £ ro) ~ 50 & v fo) o e a 0 @—@-—a—60 4 8 12 16 20 Photoperiod 100 b uw en U mS ee o os uv £ © 50 Soeike ~ fo) re $ wa o u a ot 8° @0—e0—00—00 4 8 12 16 20 Photoperiod Fic. 1. Photoperiodic response curves of P. c-aureum two different seasonal diphenisms in the complete photoperiods at 20°C (a) and 25°C (b). Open and solid circles show the incidences of autumn morph and reproductive diapause female adults, respec- tively. Seasonal Diphenisms of Polygonia 727 Under shorter than 12 hr photoperiodic regi- ments at 20°C, all females developed into autumn morph/reproductive diapause adults with an ex- ception of 2L-22D group in which 12 summer morph and 2 (out of 18) autumn morph adults were vitellogenic. In contrast, all females de- veloped into vitellogenic/summer morph adults under longer 14hr photoperiodic regimens at AIG: At 25°C, vetellogenic/summer morph adults appeared in all groups reared under complete photoperiods examined. Under short-day condi- tions at 25°C, percentages of autumn morphs varied in a range from 55% (2L-22D) to 95% (12L-12D) and 5% to 75% of the autumn morphs were judged as being vitellogenic. In contrast, all females developed into vitellogenic and summer morph adults under longer 14 hr photoperiodic regimens. Critical photoperiods for the determina- tion of these two different seasonal diphenisms were about 13 hr (13.1 hr and 12.7 hr) both at 20°C and at 25°C. The results indicate that P. c-aureum used the same critical daylength for determining both sea- sonal morphs and reproductive activity, in the complete photoperiods at 20°C and 25°C. How- ever, autumn morph development seems not al- ways to be accompanied by reproductive diapause at 25°C. Fluctuation of the incidences of autumn morph and reproductive diapause adult females in the asym- metrical skeleton photoperiods of 8L-16D To clarify how the incidences of autumn morph and reproductive diapause adults are fluctuated by a supplementary light pulse at 20°C, groups of 200-250 larvae were subjected to a photoperiodic regimen of 8L-16D, whose scottophase was inter- rupted by a supplementary light pulse of 0.5 hr. The giving time of light pulse was delayed from light-off of 8L-16D at 1-hr intervals. All females developed into autumn morph/re- productive diapause adults under 8L-16D at 20°C. All or a large part of females were prevented from exhibiting autumn morph and reproductive di- apause by a 0.5-hr supplementary light pulse on 3— 6hr after light-off of 8L-16D. However, a light pulse earlier or later in scotophase was ineffective where all or majority of females developed into autumn morph adults, 5% to 30% of which were vitellogenic (Fig. 2). WO eeeoes more -0-0~9-0-0-040-0 An 4 RO 2 © OS a awe @-0-0 vo 75) ro 2 / c= §50) H 5 é rod A ao (5) 2 a (oe P od osc Moen yaad 0) 4 8 12 16 20 2% Time of light pulse after dawn (hours) Fic. 2. Photoperiodic response curves of P. c-aureum two different seasonal diphenisms in the asymmet- rical skeleton photoperiods of 8L-16D at 20°C, where 0.5-hr light pulse interrupted the scotophase on the time given in the abscissa. Open and solid circels show the incidences of autumn morph and reproductive diapause female adults. Open trian- gles show the incidence of autumn morphs recorded under the corresponding complete photo-periods at 20°C, redrawn from Fig. 1. For the 50% prevention of autumn morph de- velopment, P. c-aureum has two clear critical photoperiods of 10.5-11.0 hr and 13.0—-13.5 hr in the asymmetrical skeleton photoperiods of 8L-16D at 20°C. The shorter critical daylength was almost the same value as, but the longer one was about 0.5-hr shorter than, the critical daylength of the 50% prevention of reproductive diapause (Fig. 2). Response curves to the number of 16L-8D (LD: long-day) days for inducing the development of summer morph and vitellogenic female adults To investigate how the effect of LD days is integrated and represented on the incidences of summer morph and vitelogenic adult females, 0- day-old 4th-instar larvae were selected from a stock culture raised from the egg stage under short days (SD-8L-16D) at 20°C. The LD-exposure was performed on 4th-instar or 4th- to 5th-instar larvae and the length of the exposure was changed from zero to 7 days depending on the groups. Any prevention in the incidence of autumn morph adults was not observed by a LD-exposure of shorter length than 3 days. But, a few percent of 728 K. Enpbo, S. UENo et al. the autumn morph adults became vitellogenic. Percentages of the summer morph adults rose gradually with the increase of the numbers of LD-exposure days. Fifty percent of the incidences of summer morph and vitellogenic adult females reached by a LD-exposure of five and four days, respectively. One hundred percents were achieved by the LD-exposure of seven days (Fig. 3). 100 @0 Percent of insects ol Ss) oO e on | ea oO OO Yo 1 3 5 7 Number of LD cycles Fic. 3. Response curves of P. c-aureum two different seasonal diphenisms to the number of long-day days, where 0-day-old 4th-instar larvae from SD condi- tions were exposed to LD cycles as given in the abscissa and then transferred to SD regime. Open and solid circles show the occurrence of summer morph and vitellogenic female adults at 20°C. Stage-dependent changes of the effects of LD-to-SD transfer on the incidences of summer morph and vitellogenic female adults To examine how the effects of LD-to-SD trans- fer on the incidences of summer morph and vitel- logenic adult females change with the developmen- tal stages, 0-day-old larvae of the 2nd, 3rd, 4th and Sth instar were selected from a stock culture raised from the egg stage under SD at 20°C. About two-hundred and fifty larvae of the same stages were grouped and subjected to a LD-to-SD trans- fer at 20°C. The LD-to-SD transfer was carried out at different/difinite developmental stages which were changed depending on groups at one- day intervals from day 2 of the first instar to day 2 of the Sth instar. When larvae were subjected to a LD-to-SD transfer before day 2 of the 2nd instar, all females responded to SD and developed into autumn morph adults, 13%-27% of which became vitel- logenic. Percentages of females showing a re- sponse to SD decreased gradually when the trans- ferred stages were delayed from day 2 of the 2nd instar. Fifty percent of the incidences of summer morph and vitellogenic adults reached by a group subjected to a LD-to-SD transfer on day 2 and day 0 of the 3rd instar, respectively. Critical stages of LD-to-SD transfer after which P. c-aureum did not show any response to SD were day 2 (for seasonal morphs) and day 0 of the 4th instar (for reproduc- tive activity) (Fig. 4). 100 76 Oe as vy) 3) cD) w) iS iz — ee — = 504 G =— 5days — © oa 0 J e“d-0-0 Lye Slimesl peleeant 0 10 i VO. a. keen Stage of Pee transfer (LD to SD) Fic. 4. Stage-dependent changes of the effects of LD- to-SD transfer on the determination of P. c-aureum two different seasonal diphenisms. Open and solid circles show the occurrence of summer morph and vitellogenic female adults at 20°C. “I0”, “II0”, “TII0”, “IVO” and “VO” show the first days of the first-, 2nd-, 3rd-, 4th- and Sth-instar larval stages, respectively. The results indicate that day 0 and day 2 of the 3rd instar seem to be the critical stages before which P. c-aureum can experience sufficient num- bers of LD days for achieving the 50% incidences of summer morph and vetellogenic adults, respec- tively. Stage-dependent changes of the effects of SD-to-LD transfer on the incidence of summer morph and vitellogenic adult females To assess how the effects of SD-to-LD transfer on the incidences of summer morph and vitel- Seasonal Diphenisms of Polygonia 729 logenic adults change with the developmental stages, 0-day-old 4th- and Sth-instar larvae and 0-day-old pupae were selected from a stock culture raised from the egg stage under SD at 20°C. Two-hundred larvae and two-hundred pupae of the same stages were grouped and were subjected to a SD-to-LD transfer at 20°C. In this experi- ment, the stage of the SD-to-LD transfer was changed depending on the groups at 1-day inter- vals from day 2 of the 4th instar to day 3 of pupa. All females introduced to LD on days preceding day 0 of the 4th instar developed into summer morph adults. Percentage of the summer morphs lowered quickly when the stage of the transfer was delayed from day 0 of the 4th instar. Fifty percent and one hundred percent of the incidences of the summer morphs were achieved by groups intro- duced to LD on day 3 and on day 5 of the Sth instar, which corresponded to 3 days and one day before larval-pupal ecdysis, respectively (Fig. 5). With respect to the vitellogenic activity deter- mination, P. c-aureum showed responses to a SD-to-LD transfer different from those in morphs. Seventy-four percent of females introduced to LD on day 2 of pupae became vitellogenic. Critical stage of the SD-to-LD transfer based on the 50%- incidence of vitellogenic adults was found to lie on day 2 of pupae, which is 5 days latter than that obtained on the 50%-incidence of the summer morphs (Fig. 5). The results indicate that the developmental stages at which P. c-aureum looses the photo- 100 5 un U o wo sg —4days Oo 504 = wo 1S) o - e ° si \oret sie, he TAP Tioger Vara 0 eae eee T ua T T T TT T T Teal: 1Y{0) vO PO Adult Stage of photoperiodic transfer (SD to LD) Fic. 5. Stage-dependent changes of the effects of SD- to-LD transfer on the determination of P. c-aureum two different seasonal diphenisms. Open and solid circles show the occurrence of summer morph and vitellogenic female adults at 20°C. “IV0”, “VO”, “PO” and “adult” show the first days of 4th-instar larval, Sth-instar larval, pupal and imagimal stages. periodic responsiveness seems to vary depending on the kinds of seasonal diphenisms of the insects. DISCUSSION Almost all adult females of autumn morphs in P. c-aureum developed under complete short-day photoperiods at 20°C entered reproductive di- apause. Critical daylengths for occurrence of autumn morph and reproductive diapause adult females were about 13 hr, as has been reported for morph determination in the same species [7]. The coincidental responses in the determination of these two kinds of seasonal diphenisms dis- appeared in the complete photoperiods at 25°C (Fig. 1b). Differential fluctuations of the inci- dences of these two kinds of seasonal diphenisms were also observed in the larval groups subjected to the asymmetrical skeleton photoperiods of 8L- 16D at 20°C (Fig. 2). Physiological mechanism underlyint the photo- periodic control of these two kinds of seasonal diphenisms may involve clock and neuroendocrine systems as had been demonstrated in many other insects [13-18]. For the regulation of two kinds of seasonal diphenisms, P. c-aureum may share clock components functioning time-measurement of photoperiod and accumulation of photoperiodic information. This suppositon is based on almost the same values of critical daylengths of these two kinds of seasonal diphenisms which P. c-aureum showed in the complete and asymmetrical skeleton photoperiods, respectively (Figs. 1 and 2). The differential responses of these two kinds of seasonal diphenisms mentioned as above may de- pend on differential responsiveness of neuroen- docrine systems secreting summer-morph-produ- cing hormone [12] and a factor activating corpora allata in adults, which seems to do similar function as allatotropin in Manduca sexta [19]. P. c-aureum required a LD-exposure of diffe- rent length (five and four days) for achieving 50% summer morph and vitellogenic adults (Fig. 3). Photoperiod-sensitive stage in which P. c-aureum shows the responsiveness to photoperiods seems to begin at around the first larval ecdysis (Fig. 6) as has been demonstrated in the determination of P. xuthus pupal diapause [6]. This supposition is 730 Peaks of hemolymph ecdysteroid titer LD-to-SD Critical stages transfer for seasonal morph determination Photoperiod-sensitive stage Critical stages for reproductive diapause determination LD-to-SD transfer Photoperiod-sensitive stage K. Enpbo, S. UENO et al. SD-to-LD Brain transfer NCCI+I for seasonal morph determination Brain SD-to-LD ag transfer for reproductive diapause determination 3rd instar First instar 4th instar 2nd instar Egg Fia. 6. 5th Pupa Adult instar Schematic representation of P. c-aureum photoperiod-sensitive stage with the critical stages of summer morph and vitellogenic adult development. “NCC(I+II)” shows the nervi corporis cardiaci. based on a hypothesis that groups of insects receiv- ing a LD-exposure of 5 and 4 days before LD-to- SD transfer should represent 50%-incidences of summer morph and vitellogenic adult females. On that day, P. c-aureum larvae undergo the first larval ecdysis at 20°C (Fig. 4). A similar explanation may be available for the end of the photoperiod-sensitive stage; the stage may come on day 1 of pupae, which is 5 days later than the critical day of SD-to-LD transfer of the incidences of summer morphs (Fig. 5). The stage is one-day earlier than the critical days at which summer-morph-producers no more require the brain and the nervi corporis cardiaci which are essential for summer morph and vitellogenic adult development (Fig. 6) [10]. However, P. c-aureum pupae are thought to have photoperiodic respon- siveness on days following larval-pupal ecdysis. This is based on the evidence that large numbers of pupae introduced to LD on 6 days following larval-pupal ecdysis showed a clear response and developed into vitellogenic (autumn morph) adults (Fig. 5). It may be that the a factor activating corpora allata in P. c-aureum adults may be se- creted in the late pupal stage according to the photoperiodic imformation accumulated from day 0 of the 2nd instar to two or three days before eclosion by the clock component (Fig. 6). Here, we could not provide any evidence about neuroendocrine system secreting a factor activat- ing corpora allata (allatotropin?) in P. c-aureum adults. But, it may become clear by further studies on physiological mechanism underlying the photo- periodic control of seasonal diphenisms of P. c- aureum. ACKNOWLEDGMENTS This work was supported in part by Grant-in-Aid from the Ministry of Education, Science and Culture, Japan No. 03640606 to K.E. REFERENCES 1 Tanaka, Y. (1950) Studies on hibernation with special reference to photoperiodicity and breeding of Chinese tussar silkworm. I. J. Seric. Jpn., 19: 358-371. 2 Miller, H. J. (1955) Die Saisonformenbildung von Araschnia levana, ein photoperiodisch gesteuerter Diapause-effekt. Naturwiss., 42: 134-135. 3 Lees, A. D. (1959) The role of photoperiod and temperature in the determination of parthenogene- tic and sexual forms in the aphid Megoura viciae Buckton—I. The influence of these factors on apter- ous virginoparae and their progeny. J. Insect Phys- iol., 3: 92-117. 4 Danilevskii, A. S. (1961) Fotoperiodism i sezonne razvitie nasekomykh. Izd. Leningradskogo Univer- sititeta, Leningrad, p. 238. 5 Beck, S. D., Clutier, E. J. and McLeod, D. G. R. 10 11 12 Seasonal Diphenisms of Polygonia (1962) Photoperiod and insect development. In “Proc. 23rd Biol. Colloq. Orgon State Univ.” pp. 43-64. Endo, K., Ito, T. and Chiba, Y. (1986) Photo- periodic control of pupal diapause in the swallow- tail, Papilio xuthus L. (Lepidoptera, Papilionidae) —Sensitive stage and number of required cycles. Zool. Sci., 3: 351-356. Hidaka, T. and Takahashi, H. (1967) Temperature condition and maternal effect as modifying factor in the photoperiodic control of seasonal forms in Poly- gonia c-aureum (Lepidoptera, Nymphalidae). Annot. Zool. Japon., 40: 200-204. Fukuda, S. and Endo, K. (1966) Hormonal control of the development of seasonal forms in the but- terfly, Polygonia c-aureum L. Proc. Japan Acad., 42: 1082-1087. Endo, K. (1970) Relation between ovarian matura- tion and activity of corpora allata in seasonal forms of the butterfly, Polygonia c-aureum L. Devel. Growth and Differ., 11: 279-304. Endo, K. (1972) Activation of the corpora allata in relation to ovarian maturation in the seasonal forms of the butterfly, Polygonia c-aureum L. Devel. Growth and Differ., 14: 263-274. Endo, K. (1984) Neuroendocrine regulation of the development of seasonal forms of the Asian comma butterfly, Polygonia c-aureum L. Devel. Growth and Differ., 26: 217-222. Endo, K., Masaki, T. and Kumagai, K. (1988) 13 14 15 16 17 18 19 731 Neuroendocrine regulation of the development of seasonal morphs in the Asian comma butterfly, Polygonia c-aureum L.: Difference in activity of summer-morph-producing hormone from _brain- extracts of the long-day and short-day pupae. Zool. Sci., 5: 145-152. Bunning, E. (1960) Circadian rhythms and time measurement in photoperiodism. Cold Spring Har- bor Symp. Quant. Biol., 25: 249-256. Tyshchenko, V. P., Gorshin, N. I. and Azaryan, A. G. (1972) The role of circadian processes in insect photoperiodism. Zhur. Obshch. Biol., 33: 21-31. Saunders, D. S. (1976) Insect Clock. Pergamon Press, New York, p. 279. Beck, S. D. (1980) Insect Photoperiodism. Academic Press, New York, 2nd ed., p. 387. Saunders, D. S. (1971) The temperature-compen- sated photoperiodic clock “programming” develop- ment and pupal diapause in the fleshfly, Sarcophaga argyrostoma. J. Insect Physiol., 17: 801-812. Gorshin, N. I. and Tyshchenko, V. P. (1970) Ther- mostability of the process of perception of photo- periodic information in the moth Acronucta rumicis (Lepidoptera, Noctuidae). Dokl Acad. Nauk. USSR, 193: 458-461. (in Russian) Kataoka, H., Toschi, A., Li, J. P., Carney, R. L., Schooly, D. A. and Kramer, S. J. (1989) Identifica- tion of an allatotropin from adult Manduca sexta. Science, 243: 1481-1483. hiiyt aditilieanlion 0 t ay We Adleprcty (st oe ‘ A orruatiatane & Buivess yu ‘ ile mepeleyhine va day thy telhins gic asl) Wh ah ‘a oc SRT Ree re b VGOMS) Vd rind (eet Title a by Maoist? AREA anit il Mia 3 lone wir ott faery ") hive fy See te natn PL) 4 Al ead we beh den ® a: sie | | ‘ had ae ha oe i Arete fl tr fl. t aie? FEA yepearte| LAE J PGE aT AYA bes Wi i t] eer it 1h aes inodirvobil peernraven staneena fi py bo ae 4 ed Punk ; State tn Ninbpn mererwsrsut eeatinl poumpy ite Elomi Ayu Agiopri a Pa {Ae a ot Op gabled wisn Stk ir Priuth b nema brshaaay von f tne 1. oulemt es "1 ea Mee ohhelas rH mialie Garris Sf] ke Gualt Step i ih bry PET UA ns ah UO Oi 1 so upobiers teat} 4 Wind ty ey weet 4 p i eel ot ek { ( erie: ¢ eiyay We ea es hy ue Ali la ae f iB r Apes dere Bol, bide (aeTy pine | j . a Tee ee be 1 “- > (Ne « r* Ly : a oe \ re ae Per ty Vind Big DETY « : 7 tut : red thal, ZOOLOGICAL SCIENCE 9: 733-739 (1992) Scanning Electron Microscopy of Differentiating Chick Ovaries during Embryonic Period ATSUMI UKESHIMA Kumamoto University, College of Medical Science, Kumamoto 862, Japan ABSTRACT— Asymmetrically developing right and left ovaries of the chick embryos from 2 to 19 days of incubation were observed by scanning electron microscope (SEM). Surface epithelium of undifferentiated gonad showed rough appearance with many vesicular protrusions and fillopodia, as compared with the other regions of coelomic surface. After the ovarian differentiation started, surface feature of the ovary was different between the right and left. In the left ovary, which made a normal development, the surface was still rugged, and possessed many vesicular protrusions and fillopodia like in undifferentiated period. In contrast, the right ovarian surface showed degenerating profile, possessing few numbers of cytoplasmic protrusions. The most significant feature of the right ovary was the fact that from 11 days onward, many holes were found on the surface. These holes exhibited a variety in size and continued to inner medullary lacunae. Sometimes, a cell or cell aggregate which, © 1992 Zoological Society of Japan supposedly, were abandoned germ cells, were found at the opening of the holes. INTRODUCTION In the female chick embryo, the ovary develops only on the left side, and the right degenerates to atrophic remnant during an embryonic period. This is due to the failure of the cortex formation in the right gonad. Although primary sex cords are equally formed on both sides during an early differentiating period of the ovary, the secondary cords, which is to form the cortex from 7 days of incubation onward, do not appear on the right. Eventurally the growth of the right ovary stops. Although some ultrastructural studies on the developing avian ovaries were present [1-4], they described on the inner structure by transmission electron microscopy (TEM). In the present study, changes of the surface morphology of the right and left ovaries which take the different pattern of formation with the advance of development, were compared by scanning electron microscopy (SEM). MATERIALS AND METHODS Chick ovaries from stage 16 (staged by Hambur- Accepted May 8, 1992 Received March 12, 1992 ger and Hamilton [5], about 2 days of incubation) to 19 days were employed for this study. As the growth of the right ovary stopped at the time of the occurrence of secondary cords, female gonads (Ovaries) were apparently different in size on the right and left sides. Therefore, identification of the gonadal sex was available. Ovaries removed from the embryos were fixed in 2.5% glutaraldehyde (phosphate buffer, pH 7.4) for 1.5 hr, and postfixed in 1% osmic acid (phosphate buffer pH 7.4) for 1 hr. Some ovaries were sectioned into thin pieces by free hand tech- nique to observe the cut face. After dehydration, materials were dried by critical point drying method, coated with platinum-paradium, and finally observed by JEOL scanning electron micro- scope (Model JSM-6400F). RESULTS In the present study, a period of the ovarian development is divided into two sections. The first is an undifferentiated period. At the initiation of this period (stage 16, about 2 days), gonadal anlages are composed of columnar (thickened) coelomic epithelium, or what is called the “germin- al ridges”. Germ cells (primordial germ cells) 734 A. UKESHIMA Fic. 1. Low magnification figure of the epithelial surface of undifferentiated gonad of a stage-16 embryo. Free surface of the cells is swollen out and possesses many fillopodia, vesicles and short microvilli. 2,400. Fic. 2. Mesonephric surface of a stage-16 embryo. As compared with the gonadal surface, few cytoplasmic processes are seen, although cells are swollen out like in the gonadal surface. 2,400. Fic. 3. Surface of the left ovary of a 8-day embryo. On the cell surface, numerous vesicular protrusions and some fillopodia are present. 3,600. Fic. 4. Right ovarian surface of a 8-day embryo. In contrast to the left ovary, the cell surface are flattened, and many short microvilli are lined at the border of each cell. 2,400. SEM of Embryonic Chick Ovaries 735 migrate into the germinal ridges via blood vessels _ epithelium is still thickned and pseudostratified, from the extraembryonic sites [6-8]. and contains many immigrated germ cells. In the The second is a differentiated period. At the right ovary, however, secondary cords are lacking. beginning (7 days) of this period, secondary cords _‘The epithelium exhibits a thin, squamous structure appear in the left ovary to form the cortex. The with no germ cells [4]. Fic. 5. Lower magnification figure of the left ovary of a 11-day embryo. The surface is very rugged, showing depression and jutting out. 540. Fic. 6. Left ovarian surface of a 16-day embryo. Cell surfaces are also covered with many vesicles, fillopodia as in younger embryo. The cells in center retain the trace of cell division (arrow). The mitotic features according to the ovarian growth were frequently observed, especially on the left ovary. 2,700. @ E mg Fic. 7. A cut surface of a 9-day left ovary. The thick surface epithelium (Ep) is composed of columnar cells. Note a germ cell (asterisk) enclosed by the epithelial cells. 2,400. Fic. 8. A sectioned view of a 9-day right ovary. Unlike the left ovary, the epithelium (Ep) is thin, and is composed of squamous cells. Under the epithelium, lacunae of various sizes (asterisk) are frequently seen. 2,400. 736 A. UKESHIMA Undifferentiated period Surface of the germinal ridges (gonadal anlages) as a whole showed a rough profile at low mag- nification (Fig. 1). This was based on the presence of vesicles of various sizes and fillopodia on the cell surface. Free surface of each epithelial cell, measuring 3.5 to 8 u~m in diameter, was swollen out into the coelomic cavity. These surface fea- tures of the undifferentiated gonads were apparently different from those on the other re- gions of the coelomic epithelium. For example, epithelial surface of mesonephros was smooth ow- ing to poor protrusions, although each cell was swollen out like in the gonadal surface (Fig. 2). Fic. 9. Whole view of the right ovary from a 16-day embryo. Many holes are found on the ovarian surface. Note a posterior end (arrows) of the ovary (Ov) incorporated to the mesonephros (Ms). X90. Fic. 10. oval shape. 250. Surface of a 19-day right ovary. Holes of various sizes are seen. The opening of the holes is usually round or SEM of Embryonic Chick Ovaries 737 Differentiated period From 7 days onward, which ocrresponded to the beginning of the secondary cord formation, surface of the developing left ovary still showed similar profile as in the undifferentiated gonads; each cell surface was usually swollen out, possessing vesicu- lar protrusions of various sizes (Fig. 3). Among those, long fiber-like structures were seen extend- ing from the edge of the cells. The free surface of the cell was 3 to 6 ~m in diameter, usually about 4 ym. On the other hand, the mght ovary, which ceased to make further development because of the failure of the secondary cord formation, exhi- bited a smooth and flattened surface (Fig. 4). The free surface of each cell was somewhat larger than that in the left ovary, measuring 4.6 to 14 wm in diameter. At the margin of each cell surface, short microvilli or protrusions were seen. The surface of the left ovary became more prominent with developmental stages, showing irregularly rugged features at lower magnification (Fig. 5). Some parts were protruded, the others hollowed. Small hollows were often observed. Higher magnification revealed the rough surface of the epithelial cells covered by many vesicular protrusions. Some cells were elongated and con- Fic. 11. nected with other cells by the intercellular bridge, showing the vestige of cell division (Fig. 6). The cut surface revealed difference in the epithelial composition of both ovaries. The epithe- lium of the left ovary was thickned like in earlier stages. Germ cells were frequently seen to be involved among these columniated epithelial cells (Fig. 7). In the right ovary, however, the epithe- lium was thin and composed of the squamous cells (Fig. 8), possessing no germ cells. Just under the epithelium, medullary lacunae were often obser- ved. The most prominent feature of the right ovary was the holes on the surface, which appeared from 11 days onward (Figs. 9 and 10). Although 11-day ovary contained several holes on the surface, they increased in number with the advance of develop- ment. In 17 days, almost whole surface of the right ovary were occupied with the holes. The holes varied in size, measuring 12 to 47 um in diameter. The right ovary in advanced stage of embryo was seen to be incorporated partly to the mesonephros, the border partly being obscured (Fig. 9). The epithelial surface was flat and scanty of microvilli. In the surroundings of the holes, however, the cytoplasmic protrusions such as fillopodia and mic- rovilli were seen increasing in number (Fig. 11). The cut surface of right ovary revealed the holes ie : j eee (12) A high magnification figure of a hole of the right ovary from a 13-day embryo. Around the hole many microvilli or long fillopodia are seen, although superficial epithelial cells have few microvilli. 1,800. Fic. 12. A sectioned view of the right ovary from a 19-day embryo. This figure shows that the hole on the ovarian surface continues with the medullary lacunae (La). Arrow: pathway of the hole to the lacunae. x700. 738 A. UKESHIMA Fic. 13. This figure shows a cell aggregate coming out of the opening of a hole on the right ovary from 19-day embryo. The aggregate is composed of several adhered cells. The arrows show the adhered portions of the cells. x3.600. continuing to the medullary lacunae (Fig. 12), while the hollows found on the left ovarian surface were not continuous to the inner structure. At the opening of the holes, free cells were occasionally found after 16 days of incubation. These cells were observed as a single cell, or a cell aggregate in which several cells adhered to each other (Fig. 13). The number of the cells were not alterable with increasing embryo age. DISCUSSION In the chick embryo, the ovaries were different in size between the right and left sides. This is due to the failure of development of the right ovary. The surface morphology of the asymmetrically developing ovaries were first revealed by SEM in the present study. On the surface of the right atrophic ovaries, many holes were present at later stage of develop- ment and continued to inner medullary lacunae. Although it is not clear what mechanism exists about this continuity in the medullary lacunae of the atrophic right ovary, it might be concerned with the degeneration of germ cells from under- standing in our previous study [4], which described on germ cells released in the medullary lacunae of degenerating right ovaries. In some cases, a cell or cell aggregate were found at the opening of the holes. Also judging from our previous study [4], the cells found at the opening may be exhibiting a terminal phase of those germ cells. More accurate identification of these cells, however, should be done by the histo- logical technique. As to birds, microscopic studies on the lacunar system of the normal left ovary have been per- formed by Callebaut er al. [9, 10]. They indicated that the lacunar system continued to the coelomic (peritoneal) cavity at the dorsal (proximal) side of the left ovary. In the present study, SEM observa- tion have been made only on the surface of both ovaries, and not on the dorsal dide, demonstrating that no holes were present on the left ovarian surface. The cell aggregates found at the opening of the holes in the right ovary were often adhered tightly to each other. This aggregate is, presumably, a mass made of germ cells and lacunar epithelial cells. Similar structures were found in the right medullary lacunae in our previous study [4]. At the beginning of ovarian differentiation, SEM of Embryonic Chick Ovaries germ cells were observed in the epithelium of the left ovary, but never in that of the right. Initially, migrating germ cells were equally existed in the columnar epithelium of both gonads. Thereafter, almost all germ cells in the right ovary have moved to the medulla with a part of the epithelium, while epithelial cells have changed to a squamous shape. Therefore, germ cells in the right ovary were localized in the medulla, but none in the epithe- hum. At the cut surface of the ovary, the germ cells within a epithelium were easily identified, because the germ cells in the epithelium were round in shape and larger than the mitotic epithelial cells [7, Sh, Lil], As secondary cords are formed into a cortex only on the left side, the left ovary developed larger in size than the right. At 7 days of incuba- tion, however, difference in ovarian size between the right and left sides was imperceptible, because it was just after initiation of the cortex formation. Therefore, materials from 7-day embryo were not employed. REFERENCES 1 Narbaitz, R. and Adler, A. (1966) Submicroscopic- al observations on the differentiation of chick gonads. J. Embryol. Exp. Morphol., 16: 41-47. 2 Yamada, K. and Amanuma, A. (1980) Fine struc- ture of Balbiani body in germ cells of the chick embryo. J. Predental Fac. Gifu Coll. Dentistry, No. 10 739 6: 87-103. (in japanese). Yamada, K. and Amanuma, A. (1985) Ultrastruc- ture of interstitial cells in the gonad of quail embryo. J. Liberal Arts Asahi Univ., No. 11: 113-125. Ukeshima, A. and Fujimoto, T. (1991) A fine morphological study of germ cells in asymmetrically developing right and left ovaries of the chick. Anat. Rec., 230: 378-386. Hamburger, V. and Hamilton, H. L. (1951) A series of normal stages in the development of the chick embryo. J. Morphol., 88: 49-92. Fujimoto, T., Ukeshima, A. and Kiyofuyi, R. (1976) The origin, migration and morphology of the pri- mordial germ cells in the chick embryo. Anat. Rec., 185: 139-154. Ukeshima, A. and Fyimoto, T. (1984) Ultrastruc- ture of primordial germ cells in the early chick embryo. In: Ultrastructure of Reproduction. (J. Van Blerkom and P. M. Motta eds.) Martinus Nijhoff Pub., Boston, 12-18 pp. Ukeshima, A., Yoshinaga, K. and Fujimoto, T. (1991) Scanning and transmission electron micros- copic observations of chick primordial germ cells with special references to the extravasation in their migration course. J. Electron Microsc., 40: 124-128. Callebaut, M. (1979) The avian ovary is an open organ.; A study of the lacunar system. Anat. Embryol., 158: 103-119. Callebaut, M., Meeussen, C. and Nassauw, L. V. (1988) The early development of the lacunar system in the avian ovary. Med. Sci. Res., 16: 1131-1133. Ukeshima, A. and Fujimoto, T. (1978) Scanning electron microscopy of the primordial germ cells in early chick embryos. J. Electron Microsc., 27: 19- 24. ahnet) ee om bia } 4 Teh his NGG OF Son x aI ih rh ab f RR TE A a sae Lie rere | han oo Si t a x p Meier Ferd airy nf oF 4 SW Heese ete mort 3 e be ok F vo 5 i i ; 1 j fad fF te al { { i hel of \ j 4 ah {rtaret & b a | Aeraes| . a \ 4 fs fi ~<. t- ? © = ¢ Yi ZOOLOGICAL SCIENCE 9: 741-747 (1992) © 1992 Zoological Society of Japan The Acid-Insoluble Organic Matrix of Spicules in the Sea Urchin Hemicentrotus pulcherrimus SHOGO NAKAMURA, RYOKO KAGOTANI, HIROE FUJISAKI and MANABU K. KoJIMA Department of Biology, Faculty of Science, Toyama University, Toyama 930, Japan ABSTRACT— The structure of the acid-insoluble organic matrix (ATOM) of the spicules of sea urchin larvae was examined by light microscopy and transmission electron microscopy. The spicules were isolated from pluteus larvae of Hemicentrotus pulcherrimus, and AJOM was obtained by demineraliza- tion of the spicules with 0.1 N acetic acid. We found that AIOM has many fine fibers (diameter: 16-27 nm) and two types of granules. INTRODUCTION A mineralized skeleton of spicules is one of the most conspicuous structures in sea urchin embryos. Previous investigations of the develop- ment of the skeleton in sea urchin larvae showed that the primary mesenchyme plays a major role in spicule formation [1-4]. However, the control of spicule formation has not yet been fully resolved. Calcite, CaCO3, and small amount of MgCO; are the inorgaic components of the spicules [5]. Oka- zaki [1] showed that spicules contain an organic matrix which was insoluble after acid-treatment. Benson ef al. [6] revealed that EDTA- or acid- insoluble organic matrix contains fibrous lamella. Although acid- or EDTA-soluble organic matrix had several glycoproteins and proteins [7, 8], the components of acid- or EDTA-insoluble matrix are not known. To obtain further information about biomineralization in the sea urchin embryo, we have observed the structure of the acid- insoluble organic matrix (AIOM) by light and electron microscopy. We found that AIOM con- tains many fibers and two types of granules. Accepted May 28, 1992 Received March 9, 1992 MATERIALS AND METHODS Gametes Hemicentrotus pulcherrimus were collected from the coast of Toyama Bay and were maintained at 14°C in aquaria with circulating sea water. Gametes were obtained by injecting 0.5 M KCl into the body cavity. Sperm were collected directly from the coelomic cavity and stored “dry” in the refrigerator. Eggs were collected and washed 2 times in filtered natural sea water (FNSW). Isolation of organic matrix Fertilized eggs were washed in FNSW and cultured to plutei at 18°C with constant stirring (~1lround/sec). Larvae were collected by centrifugation (at 3,000 g for 10 min) and washed once in cold (4°C) Ca*t- and Mg**-free sea water (CMFSW: NaCl 26.5 g, KCI 0.7 g, pH was adjusted to 8.2 by the addition of 0.5 M NaHCoO;) [9]. They were resuspended in CMFSW, shaken vigorously, then centrifuged at 1,000 g for 5 min. The pellet was resuspended in 10 vol of 10 mM Tris buffer (pH 7.5) which con- tained 0.5 mM phenylmethylsulfonyl fluoride, and pipetted vigorously. The suspension was centri- fuged at 1,000 g for 5 min. The pellet was resus- pended in 10 vol of 2% TritonX-100, 4% sodium deoxycholate, 20mM Tris buffer (pH7.4) and pipetted strongly, then centrifuged at 1,000 g for 5 min. This procedure was repeated 2-3 times to 742 S. Nakamura, R. KAGOTANI et al. obtain crude fraction of spicules as the pellet. The crude fraction was resuspended in 5vol of the same Tris buffer containing detergents and 5 vol of 5% sodium hypochlorite in 20 mM Tris buffer (pH 7.5). The suspension was pipetted and centrifuged at 25 g for 3 min. The pellet was treated with 5% sodium hypochlorite for 10min and washed 3 times with artificial sea water (ASW: NaCl 26.5 g, MgSO,:7H20 11.9g, KCl 0.7 g, CaCl 1.2 g, pH was adjusted to 8.2 by the addition of 0.5M NaHCOs) [9]. Micrographs of isolated spicules were taken by drak-field microscopy (Olypus BH2 microscope with a BH2-DCW dark-field condenser). Electron microscopy Isolated spicules were de- mineralized by resuspending them in equal vol of 0.1 N acetic acid and incubating for 30 min at 0- 4°C. The suspension was spun at 16,000 g for 10 min, and the acid-insoluble organic matrix (AIOM) was obtained as the pellet. The AIOM was fixed in 2% glutaraldehyde in 80% ASW for 1 hr at room temperature, and then fixed again in 1% glutaraldehyde in 0.1M sodium cacodylate (pH 7.0) for 1 hr. The samples were post-fixed in 1% OsOy, in 0.1 M sodium cacodylate (pH 7.2) for 1 hr. After rinsing 0.1 M sodium cacodylate buffer (pH 7.2), the samples were dehydrated through a ethanol series. They were infiltrated and embed- ded in epoxy resin (Epok 812, Oken Shoji Co., Ltd., Tokyo). Sections were cut with glass knives on a Sorvall MT1 Porter-Blum ultramicrotome, and stained with uranyl acetate and Reynold’s lead citrate. To survey the purity of isolaed spicule, whole mounts were prepared. Ultrathin sections and whole mounts were examined with a JEOL 100SX electron microscope operated at 80 kV. RESULTS We isolated many spicules in almost intact form from pluteus larvae of Hemicentrotus pulcherrimus (Fig. 1b). There was very little contamination in the fraction of isolated spicules (Fig. 1b) and the surface of the spicules was clean (Fig. 2). When the isolated spicules were treated with 0.1 N acetic acid for 30 min, demineralization occurred and thread-like structures remained (Fig. lc), as re- Fic. 1. sea urchin, Hemicentrotus pulcherrimus. Preparation of AIOM from pluteus larvae of the Pluteus larvae (a: 72 hr after the insemination) were used for isolation of spicules which appeared free of con- tamination by dark-field microscopy (b). After isolated spicules were demineralized with 0.1 N ace- tic acid, thread-like structures with granular varico- sities remained (c). Bar, 100 «~m. ported by Okazaki [1]. granules on the thread-like structures. There were many fine Those granules were never seen in the fraction of isolated spicules. We call this thread-like structure the acid-insoluble organic matrix (AIOM). AIOM was examined with transmission electron micros- copy. Figure 3 shows that AIOM has many fibers Organic Matrix of Spicules 743 which are twined loosely (diameter: ~27 nm) with a few filaments (Fig. 4a) or tightly (diameter: ~ 16 nm) (Fig. 4b). We could not see a periodic pattern on the fibers. The fibers ran parallel with each other in some regions and they made an eddy pattern in other regions. Moreover, AIOM contained two types of gran- ules (Fig. 3). Granules whose appearances re- minded us of a cell nucleus (N-granules) lacked limiting membranes (Figs. 3, 5a). Many fine fibrous structures were present in the electron Fic. 2. Whole mount EM of ioslated spicules. No translucent compartment of N-granules (Fig. 5a). contamination is evident around the spicule surface. | We often observed disintegrating N-granules with Bar, 5 wm. discharged fibrous structures (Fig. 5a) which looked like filaments in the loosely twined fibers (Fig. 4a). The other type of granules (V-granules) Fic. 3. A typical ultrathin section of AIOM. Many fine fibers and two types of granules were found in this specimen. The fibers run parallel in some regions, and in eddy patterns in other regions. N: N-granule. V: V-granule. Bar, 2 pm. Fic. 4. High magnification TEM views of the fibers in AIOM. The fibers are twined loosely with a few filaments (a) or more tightly (b). Bar, 0.2 ~m. Fic. 5. High magnification views of two types of granules in AIOM. The nucleus-like granule (N-granule) (a) has no clear membrane and many fine fibers in the electron translucent parts. This N-granule was disintegrating with discharged filaments (arrows). The other type of granules (V-granules) have vacuoles surrounded with electron dense deposite (b). Bar, 0.5 um. 744 S. Nakamura, R. KaGorani et al. Organic Matrix of Spicules 745 746 S. NAKAMURA, R. KAGOTANI et al. showed homogenous and low electron density, and had large vacuoles with peripheral deposits of electron dense materials in it (Figs.3, 5b). V- granules had no outer membrane and took various shapes. They were seen more frequently than N-granules. Preliminary observations show that N-granules and V-granules have various diameters (N-granules: 2.5 ~m-—3.2 um; V-granules: 1.4 ~m— 2.4 wm). DISCUSSION We have found that spicules of sea urchin embryos contain many fine fibers and two types of granules. Our preliminary observations show that the same types of fibers and granules that we observed in Hemicentrotus pulcherrimus embryos exist in spicules of Pseudocentrotus depressus embryos (data not shown). Benson ef al. [6] reported that the organic matrix of spicules had fibrous lamellae. However, they did not show the fine structure of the fibrous lamella and did not report the existence of granules. In the blastocoel of the sea urchin embryo, there is an abundunce of fibrillar matrix which may consist of collagen [10]. However, we believe that fibers and granules in our preparation are not contaminants of blastocoelic fibrillar matrices or substances, since (1) no contaminants were evident around the surface of isolated spicules by light and electron microsopy (Figs. 1b, 2), (2) we observed only one type of fiber and two types of granules in AIOM. Although the diameter of the fibers varied from 16 nm to 27 nm, we think they are made from the same material(s) and that the difference in diameter reflects their formation or assembly. It is well known that vertebrate bone predomi- nantly contains tvpe I collagen fiber. Pucci- Minafra et al. [11] have reported that the spicule of a sea urchin embryo contains collagen and which may play a role in the formation of the organic matrix of the spicules. Several investigators re- ported that collagen metabolism is required for spicule formation [12-14]. On the other hand, Benson et al. failed to detect the presence of hydroxyproline-rich collagen by structural [6] and biochemical studies [7], and they claimed that spicules isolated by the methods of Pucci-Minafra et al. were impure. We think the fibers seen in AIOM are probably not collagen since they did not show periodic banding, and the proteins did not have typical mobilities of collagen in SDS-PAGE (preliminary data not shown). We suppose the fibers give strength and flexibil- ity to the spicule and the granules are related to the fiber formation. Whole AIOM may have an important role in the formation of the spicule. Biochemical studies are under way to define their natures and functions. ACKNOWLEDGMENTS We thank Drs. S. L. Tamm and S. Pimple (BUMP, MBL, Woods Hole) for their critical readings of this manuscript. REFERENCES 1 Okazaki, K. (1960) Skeleton formation of sea urchin larvae. II. Organic matrix of the spicule. Embryologia, 5: 283-320. 2 Wolpert, L. and Gustafson, T. (1961) Studies on the cellular basis of morphogenesis of the sea urchin embryo. Development of the skeletal pattern. Exp. Cell Res., 25: 311-325. 3 Okazaki, K. (1965) Skeleton formation of sea urchin larvae. V. Continuous observation of the process of matrix formation. Exp. Cell Res., 40: 585-596. 4 Okazaki, K. (1975) Spicule formation by isolated micromeres of the sea urchin embryo. Amer. Zool., 15: 567-581. 5 Okazaki, K. and Inoué, S. (1976) Crystal property of the larval sea urchin specicule. Develop., Growth and Differ., 18: 413-434. 6 Benson, S., Jones, E. M. E., Crise-Benson, N. and Wilt, F. (1983). Morphology of the organic matrix of the spicule of the sea urchin larva. Exp. Cell Res., 148: 249-253. 7 Benson, S. C., Benson, N. C. and Wilt, F. (1986) The organic matrix of the skeletal spicule of sea urchin embryos. J. Cell Biol., 102: 1878-1886. 8 Venkatesan, M. and Simpson, R. T. (1986) Isola- tion and characterization of spicule proteins from Strongylocentrotus purpuratus. Exp. Cell Res., 166: 259-264. 9 Okazaki, K. (1956) Skeleton formation of sea urchin larvae. I. Effect of Ca concentration of the medium. Biol. Bull., 110: 320-333. 10 Crise-Benson, N. and Benson, S. C. (1979) Ultra- structure of collagen in sea urchin embryos. Wilhelm 11 12 Organic Matrix of Spicules Roux’s Arch Dev. Biol., 186: 65-70. Pucci-Minafra, I., Casano, C. and La Rosa, C. (1972) Collagen synthesis and spicule formation in sea urchin embryos. Cell Differ., 1: 157-16513. Golob, R., Chetsanga, C. J. and Doty, P. (1974) The onset of collagen synthesis in sea urchin embryos. Biochim. Biophys. Acta, 349: 135-141. 13 14 747 Gould, D. and Benson, S. (1978) Selective inhibi- tion of collagen synthesis in sea urchin embryos by a low concentration of actinomycin D. Exp. Cell Res., 112: 73-78. Blankenship, J. and Benson, S. (1984) Collagen metabolism and spicule formation in sea urchin micromeres. Exp. Cell Res., 152: 98-104. ; | chee ; { eee wee . nia) bes 4 haa o i ‘Li vl a OG ROR pha so yeiy Hips aeirnaiedie Naveen WI oo “hy “ae Pal) wee aa, ett Vs ite i perenne pti & yeh destabal odes ub A wah nn, wadigdetian ‘ ih ert ea Lt 4 & ; : Ye ‘Bima tite: Shy): iit ine a die ‘inden G oe ne anil Wnfea Cah. (141 CONROY | Ea = at vm : ok, Gre So Tie ' : ; a : ee ee a — a fs 7, i Ss J . Pt - i T ay ) 4 i ZOOLOGICAL SCIENCE 9: 749-755 (1992) © 1992 Zoological Society of Japan Scleral Fibroblasts of the Chick Embryo Can Proliferate without Transferrin in Protein-Free Culture YosHie Ouya!, KAzUuo WATANABE!”*, NoBuo SHIMAMOTO2 and Minoru Amano! 'Cell and Developmental Biology Laboratory, Faculty of Integrated Arts and Sciences, Hiroshima University, Hiroshima 730, and 2DNA Research Center, National Institute of Genetics, Mishima 411, Japan ABSTRACT—Scleral fibroblasts of the chick embryo can proliferate in vitro without transferrin. We prepared a monoclonal antibody, TFY004, which was reactive against avian-transferrin, and checked the possibility of an autocrine secretion of transferrin in the cells. Dot-blot immunoassay of the conditioned medium by the use of TFY004 indicated that the conditioned medium did not contain detectable amount of transferrin. We tested muscle trophic activities of the conditioned medium to myogenic cells to detect transferrin activities in the conditioned medium. The results also indicated no presence of transferrin in the conditioned medium. Furthermore, the supplementation of transferrin had no effect for survival and proliferation of the scleral fibroblasts. All these results suggested that, in the scleral fibroblast culture, mechanisms different from the external transferrin-mediated one must be involved to uptake iron from medium into cells. INTRODUCTION Living cells inevitably requie iron. It is generally accepted that the cells utilize environmental iron by an efficient transport system with the use of transferrin, a soluble protein with high iron- binding capacity [1, 2]. One of the reason for adding serum in the cell cultures is to supply a sufficient amount of serum-transferrin. The trans- ferrin binds efficiently irons in the culture medium, and the iron-bound transferrin is internalized by endocytosis [3] after specific binding to the cellular receptor [4]. In consequence, the supplementation of high concentration of transferrin, 5-10 “g/ml, is a general format for serum-free cultures [5]. Interestingly, the scleral fibroblasts of chick embryo do not require supplementation of trans- ferrin to the protein-free culture medium for their proliferation [6]. Then, a question arises, how the Accepted May 28, 1992 Received May 6, 1992 3 To whom all correspondence should be addressed. scleral fibroblasts uptake environmental iron, which is necessary for cell survival. One possible mechanism in an autocrine secre- tion of transferrin by the cells themselves. In fact, we reported that the scleral fibroblasts proliferated by producing multiple autocrine growth factors into the conditioned medium [6, 7]. Dittmann et al. reported that HL-60 cells produced a transfer- rin-like protein as an autocrine factor, which reacted with anti-transferrin antibody [8], and also Kohgo et al., reported K562 cells could proliferate without transferrin by secreting immuno-reactive transferrin as an autocrine factor [9]. In this study, we tried to detect transferrin in the conditioned medium of the scleral fibroblast cul- ture. MATERIALS AND METHODS Chick embryos were obtained by incubating eggs of randomly bred Japanese fowl purchased from Yamagishism Association, Japan. 750 Y. Onya, K. WATANABE et al. 1. Cell culture Scleral fibroblasts Procedures for culturing scleral fibroblasts were described previously [6, 7]. In brief, the scleral fibroblast layers were isolated from the eyeballs of chick embryos (stage 38 or stage 39) [10], and incubated with 2 ml of 0.2% collagenase at 37°C for 40 min, and flushed gently in 2 ml of EF- medium (1:1 mixture of Dulbecco’s modified Eagle medium and Ham’s F-12 medium), using a Pasteur pipette to dissociate tissues into single cells. The dissociated cells were inoculated at an initial density of 5 x 10° cells in 100-mm plastic dish with 8ml of EF-medium and cultured in 5% CO,-95% moist air at 38°C. Myogenic cells Muscle tissues were isolated from the femoral region of stage 38 embryos. Procedures for dis- sociation were the same as in the case of scleral fibroblasts. The cells were inoculated at an initial density of 5X10? cells in 35-mm plastic dish (Fal- con 3001; Becton Dickinson Co., U.S.A.) with 2 ml of EF-medium supplemented with 10% horse serum (GIBCO, U.S.A.) and were incubated in humidified 5% CO,-incubator at 38°C. 2. Preparation of conditioned medium and SAF-I from the scleral fibroblast culture After the first medium change on the third day of culture, the conditioned media were harvested every second day, i.e., the fifth, seventh and nineth days of culture. Cell densities were 120 cells/mm? at the third day and 650 cells/mm” at the nineth day. The procedures for preparing the SAF-I (scleral autocrine factor-I) and its characterization were described previously [6]. In brief, pooled con- ditioned medium was concentrated by the use of ultrafiltration, removing molecules less than 10 kDa. The concentrated conditioned medium was applied to DEAE-Sepharose column chroma- tography at pH 8.0, and proteins were eluted with 0-0.35 M NaCl gradient. We called a fraction, eluted at about 0.2 M and showed a heat-stable growth-promoting activity, as SAF-I. 3. Preparation of ovotransferrin as an avian trans- ferrin Ovotransferrin was extracted from chick egg white after Kimura’s methods [11]. In brief, acetone powder of egg white was prepared and dissolved in FeCl; solution. The solution was subjected to preparative DEAE-Sepharose, and next to CM-Sepharose column chromatography. A fraction which showed a single peak (absorb- ance 280 nm), manifested also a clear single band with molecular weight of ca. 80kDa by SDS polyacrylamide-gel electrophoresis (SDS-PAGE) [12]. We call the fraction as purified ovotrans- ferrin. 4. Preparation of monoclonal antibody against transferrin As an antigen, 0.8 mg of the purified ovotrans- ferrin was dissolved in 0.25 ml of distilled water, mixed with the same volume of complete Freund’s adjuvant and injected intraperitoneally to a BALB/c mouse. Twelve days after the primary injection, booster injection was made with an adjuvant-free antigen. Three days later, spleen cells were harvested and fused with myeloma cells (Sp2/0-Ag-14) [13] by the use of polyethylenegly- co16000 (Koch-Light Co., Germany) as described by Galfre et al., [14]. HAT-selected hybridoma cells were cloned by limiting dilution and the supernatant media were screened by ELISA method using horse-radish peroxidase (HRP)- conjugated goat anti-mouse immunoglobulins (ZYMED LAB. Inc., U.S.A.). We obtained 15 hybridoma clones producing antibodies against transferrin. But unfortunately, most positive clones attenuated rapidly, i.e., they ceased to produce antibodies, or stopped to pro- liferate and died. Finally a hybridoma clone, TFY004 clone, was remained. We recloned the TFY004 cells, which left sufficient amount of hy- bridoma supernatant (TFY004), which specifically reacted with purified ovotransferrin on nitrocellu- lose filter (Fig. 1). 5. Dot-blot immunoassay Every one ml of the solutions containing various concentration of the test samples was blotted to Cell Proliferation without Transferrin 751 ovalbumin BSA va — Ing 300ng eS 100ng 30ng 10ng 3ng Fic. 1. Dot-blot immunoassay of purified ovotransfer- rin, ovalbumin and BSA (bovine serum albumin) with hybridoma supernatant, TFY004. Each 1 ml of serially-diluted ovalbumin (No. 300-00711; Wako, Japan), BSA (No. 017-12961; Wako) or purified ovotransferrin (Tf) were blotted to a nitrocellulose filter and reacted with TFY004. The concentrations were indicated at the right. nitrocellulose filter (No. 162-0115; BIO-RAD, U.S.A.) by the use of Milliblot (Millipore Ltd., U.S.A.). The filter was washed with Ca**- and Mg**-free phosphate buffered saline (PBS) and reacted with non-diluted hybridoma supernatant, TFY004, for 1hr. Then, 60-times dilution of HRP-labelled goat anti-mouse immunoglobulins (ZYMED LAB., Inc.) was added and visualized by diaminobenzidine. All procedures were done at room temperature and filters were washed with PBS containing 0.05% Tween-20. Non-specific binding of antibodies was blocked with PBS con- taining 3% of commercial skim milk. 6. Measurement of growth-promoting activity Growth-promoting activity was assayed as an enhancement in *H-thymidine incorporation into acid-insoluble fraction of the cells. Detailed condi- tions for culturing and labelling were given pre- viously [6, 7] and also in figure legends for Figure 5 and 6. After labelling, the cell monolayer was lysed with 0.1 N NaOH and lysate was poured onto a glass microfiber filter after Kawahara’s methods [15]. The filter was washed three times with 5% TCA and once with absolute ethanol. Radioactivity was counted with an Aloka LSC- 7000 liquid scintillation counter [15]. RESULTS Conditioned medium did not contain transferrin according to dot-blot immunoassay SDS-PAGE of the conditioned medium indi- cated the presence of a protein with molecular weight near 80 kDa, which was similar to transfer- Its concentration in the medium was esti- We purified rins. mated to be 150 ng/ml (Fig. 2). conditioned medium with DEAE-sepharose chro- CM SAF-| «— 80kDa—, aay hae sae 320° > Fic. 2. SDS-PAGE of the conditioned medium (CM) and the partially purified growth promoting factor (SAF-I) separated from the conditioned medium. Ten micro litter of the acid-insoluble materials of the conditioned medium was applied, corresponding to 0.1 ml of conditioned medium. In the case of SAF-I, 10 wl of the solution containing 4 ug of protein was applied. 752 Y. Onya, K. WATANABE et al. matography. The partially purified fraction named SAF-I [6], also showed a band of ca. 80kDa protein (Fig. 2). However, there found no substance to react with anti-transferrin antibody in the conditioned medium by dot-blot immunoassay. If transferrin existed, it should be less than 1 ng/ml in the conditioned medium (Figs. 1, 3). SDS-Tf tml 300, 30 pI 10 pl Sp! 1ng tpl 0.1ng Fic. 3. Dot-blot immunoassay of the conditioned medium of scleral fibroblasts with hybridoma super- natant, TFY004. Each 1 ml of serially-diluted con- ditioned medium were blotted to nitrocellulose filter. The volume of conditioned medium used was indicated at the left. As the controls, each | ml of ovotransferrin and SDS-treated one (SDS-Tf) were blotted to the same filter and reacted with TFY004. Their concentrations were indicated at the right. Conditioned medium had no effect on myogenic cells Addition of 3 “g/ml of purified ovotransferrin to myogenic cells brought about prominent muscle trophic effects (Fig. 4), as reported by Ozawa’s group [2, 16]. On the other hand, 50-fold concen- trated conditioned medium of the scleral fibro- blasts did not promote the proliferation and dif- ferentiation of myogenic cells (Fig. 4). Consistently, the conditioned medium and SAF- I, which showed DNA-synthesis promoting activ- ity to the scleral fibroblasts, did not promote any DNA synthesis of the myogenic cells, under the condition that the supplementation of ovotransfer- rin specifically promoted it (Fig. 5). The addition of ovotransferrin had no effect on the growth of scleral fibroblasts The addition of 6 4g/ml of ovotransferrin, a routine concentration in serum-free cultures, had no DNA-synthesis promoting effect to the scleral fibroblasts, although insulin had prominent effect to promote DNA synthesis (Fig. 6). Increasing the concentration of ovotransferrin from 6 ng/ml to 50 ug/ml, we obtained the same results (data not shown). DISCUSSION The scleral fibroblasts can rapidly proliferate in vitro under protein-free medium [6]. We tested a postulation that the scleral fibroblasts secreted transferrin into conditioned medium as an auto- crine factor. It is known that the biological activities of transferrins manifest class-specificity among verte- brate animals [2, 17], i.e., mammalian-type trans- ferrin does not manifest its biological activity to avian cells, and vice versa. So, we purified ovo- transferrin from chick egg white as an avian-type transferrin [11, 17] and used it for positive control experiments in the biological assays (Figs. 4, 5, 6). At the same time, we prepared an specific mono- clonal antibody, TFY004, against the ovotransfer- rin. Since the TFY004 could not recognize SDS- treated ovotransferrin (Fig. 3), the epitope of TFY004 seemed to be protein conformation, not to be carbohydrate moiety. The obtained results indicated; (1) the con- ditioned medium did not contain immunochemi- cally-detectable amount of transferrin. The dot- blot analysis indicated that it contained, if existed, less than 1 ng/ml (Figs. 1, 3), much less than the effective concentration, 800 ng/ml [18], of trans- ferrin on growth of myogenic cells. (2) The Cell Proliferation without Transferrin 753 conditioned medium did not mainfest any muscle trophic effects (Fig. 4) or growth-promoting effect (Fig. 5) upon myogenic cells even with the use of concentrated one, which estimated to contain ca. 7-10 ug/ml of the 80 kDa protein. These results indicated that there are no significant amount of transferrin in the conditioned medium, and lead us to discard the possibility of autocrine secretion of transferrin in the scleral fibroblasts. The protein with molecular weight of ca. 80 kDa in the con- ditioned medium was considered not to be trans- ferrin. 100 hm Fic. 4. Myogenic cell cultures treated with ovotransferrin (3 ~g/ml) or conditioned medium (final concentration was 50-fold concentrated from original conditioned medium), showing prominent differences in growth and differentiation on the 4th day after inoculation. For culture conditions, see Materials and Methods. Then, how do the scleral fibroblasts satisfy iron requirement for their survival and proliferation? Interestingly, the supplementation of 6 ng—S0 ug/ ml of ovotransferrin showed no effects on their proliferation. This fact seems to suggest the scleral fibroblasts lack transferrin receptor, since it is generally accepted that the transferrin receptors increase in the rapidly proliferating cells [2]. In another word, the fact implicates the uptake of iron without participation of soluble transferrin in the scleral fibroblasts. One possible mechanism is the scleral fibroblasts =) Nn ms 3H-Thymidine incorporation (cpm x 10-3) 3H-Thymidine incorporation (cpm x 10-2) = (o) , ] Y. Onya, K. WATANABE et al. BSA = oO Tf Fic. 5. SAF-| Control Insulin Fic. 6. BSA Tf Control 24 hr 48 hr 72 hr — Lt Fic. 5. Effects of ovotransferrin, BSA, conditioned medium and SAF-I, on *H-thymidine incorporation of myogenic cells. 5X10*cells/6-mm microwell (Falcon 3072) were inoculated with 2001 of DMEM (Dulbecco’s modified Eagle medium) con- taining 10% horse serum. After 12 hr cultivation, ovotransferrin (final concentration; 5 ug/ml), BSA (final concentration; 5 ug/ml), conditioned medium (30 wl/well) or SAF-I (5 pl/well), was added to become final volume of 200 wl. After further 12 hr cultivation, cells were labelled for 12 hr with 0.3 nCi of ?H-thymidine (Amersham Co., U.K., 40-60 Ci/ mmol) in 60 “«l of DMEM, and radioactivities were measured. Control means no addition of test sam- ple. The data were averages of three different cultures with SD. Fic. 6. Effects of insulin, BSA and ovotransferrin on 3H-thymidine incorporation of the scleral fibro- blasts. 510° cells/6-mm microwell (Falcon 3072) were inoculated with 200 «l of DMEM containing 6 yg/ml of insulin (from bovine pancreas; No. I-1882; Sigma Co., U.S.A.), BSA or ovotransferrin. After cultivation for 24, 48 and 72 hr, cells were labelled for 12 hr with 0.3 Ci of *H-thymidine and radioacti- vities were measured. Control means no addition of test sample. The data were averages of three different cultures with SD. uptake iron through membrane-bound transferrin. In fact, melano-transferrin (p 97), a membrane- bound transferrin like molecule, was found in human melanoma [19]. Unfortunately, we could not test the mechanism now, because hybridoma (TFY004) had already attenuated. On the other hand, Saito et al. [18] reported that chick embryonic cells could grow without transfer- rin in 10-100 “M Fe?*- or Fe**-ion, suggesting an existence of unknown iron-uptake mechanism in the chick embryonic cells. We confirmed that the scleral fibroblasts grow even in the condition of less than 0.2 ~M of Fe**-ion (data not shown). So, if such iron-uptake mechanism without trans- ferrin actually operated in the scleral fibroblasts, it should be more enhanced. Anatomically, the scleral fibroblasts are derived from perichondrial tissue, and we recently found the scleral fibroblasts possessed a capacity to dif- ferentiate into cartilage in soft agar [20]. It is of Cell Proliferation without Transferrin 755 interest if such nutritional characteristics as iron- uptake without soluble transferrin are specific for precartilagenous perichondrial cells, because they manifest an adaptive proliferation and differentia- tion in either skeletal remodeling during develop- ment or fracture healing [21] under restricted vascular environment. ACKNOWLEDGMENTS The authors are indebted to Dr. Makoto J. Tabata in our laboratory for technical advices in hybridoma pro- duction, and also to Dr. Akira Kawahara for his gener- ous advices. The work was supported by a Grant-in-Aid for Scientific Research to K.W. (No. 03833023; No. 03304009) from the Ministry of Education, Science and Culture, Japan. REFERENCES 1 Neuman, R. E. and Tytell, A. A. (1961) Iron replacement of lactalysate and embryo extract in growth of cell culture. Proc. Soc. Exp. Biol. Med., 107: 876-880. 2 Ozawa, E. (1989) Transferrin as a muscle trophic factor. Rev. Physiol. Biochem. Pharmacol., 113: 89-141. 3 Karin, M. and Mintz, B. (1981) Receptor-mediated endocytosis of transferrin in developmentally totipo- tent mouse terato-carcinoma stem cells. J. Biol. Chem., 256: 3245-3252. 4 Jandle, J. H., Inman, J. K., Simmons, R. L. and Allen, D. W. (1959) Transfer of iron from serum iron-binding protein to human reticulocytes. J. Clin. Invest., 38: 161-185. 5 Barnes, D. and Sato, G. (1980) Serum-free cell culture: A unifying approach. Cell, 22: 649-655. 6 Watanabe, K., Fujioka, M., Takeshita, T., Tsuda, T., Kawahara, A. and Amano, M. (1989) Scleral fibroblasts of the chick embryo proliferate by an autocrine mechanism in protein-free primary cul- ture: differential secretion of growth factors depend- ing on the growth state. Exp. Cell Res., 182: 321- 329. 7 Fujioka, M., Shimamoto, N., Kawahara, A., Ama- no, M. and Watanabe, K. (1989) Purification of an autocrine growth factor in conditioned medium obtained from primary culture of scleral fibroblasts of the chick embryo. Exp. Cell Res., 181: 400-408. 8 Dittmann, K. H. and Petrides, P. E. (1991) A 41 kDa transferrin related molecule acts as an auto- crine growth factor for HL-60 cells. Biochem. Bio- 12 13 14 15 16 17 18 20 21 phys. Res. Commun., 176: 473-478. Kohgo, Y., Kato, J. and Kondo, H. (1982) Iron binding proteins, transferrin and its receptor: Reg- uratory mechanism of gene expression and clinical application (in Japanese). Protein-Nucleic Acid- Enzyme (Kyoritsu-Shuppan, Tokyo, Japan), 37: 26— 36. Hamburger, V. and Hamilton, H. L. (1951) A series of normal stages in the development of chick embryo. J. Morphol., 88: 49-92. Kimura, I., Hasegawa, T. and Ozawa, E. (1982) Indispensability of iron-bound chick transferrin for chick myogenesis in vitro. Develop. Growth and Differ., 24: 369-380. Laemli, U. K. (1970) Cleavage of structural pro- teins during the assembly of the head of bacte- riophage T4. Nature, 227: 680-685. Shulman, M., Wilde, C. D. and Kohler, G. (1978) A better cell line for making hybridomas secreting specific antibodies. Nature, 276: 269-270. Galfre, G., Home, S. C., Milstein, C., Butcher, B. W. and Howard, J. C. (1977) Antibodies to major histocompatibility antigen produced by hybrid cell lines. Nature, 266: 550-552. Kawahara, A., Ishikawa, S. and Amano, M. (1985) In vitro growth of adult amphibian (Xenopus laevis) hepatocytes and characterization of hepatocyte- proliferating activity in homologous serum. Exp. Cell Res., 159: 344-352. li, I., Kimura, I. and Ozawa, E. (1982) A myo- trophic protein from chick embryo extract: its purification, identity to transferrin, and indispensa- bility for avian myogenesis. Dev. Biol., 94: 366-377. Shimo-Oka, T., Hagiwara, Y. and Ozawa, E. (1986) Class specificity of transferrin as a muscle trophic factor. J. Cell. Physiol., 126: 341-351. Saito, K., Hagiwara, Y., Hasegawa, T. and Ozawa, E. (1982) Indispensability of iron for the growth of cultured chick cells. Develop. Growth and Differ., 24: 571-580. Rose, T. M., Plowman, G. D., Teplow, D. B., Dreyer, W. J., Hellstrom, K. E. and Brown, J. P. (1986) Primary structure of human melanoma- associated antigen p 97 (melanotransferrin) deduced from the mRNA sequence. Proc. Natl. Acad. Sci. USA., 83: 1261-1265. Watanabe, K., Yagi, K., Ohya, Y. and Kimata, K. (1992) Scleral fibroblasts of the chick embryo diffe- rentiate into chondrocytes in soft-agar culture. In Vitro Cell. Dev. Biol. 28A: (in press) Rosen, V. and Thies, R. S. (1992) The BMP proteins in bone formation and repair. Trends Genet., 8: 97-102. aye + BT) nui AH ar —— i - nei Ht cote pele one? ry, ae H lve aorresmamiyy aevcadind), acriboriel Senne alihe he aad twig” nf © dbs t “Ail toTetcorlep yn ay | 1) ANE oAy y ot RP H 7 he -, ‘Ae § bh gaa - 4 Ero rove anah) Te eget til ou nye as 2% ifiaeprie Ye bat May ‘ " reoahy b . ‘ p Avy Siete WN waa MY, chee ieg eben ; a Neral ate I = ' a) ‘endo at ; ae oe: Ele Hy ‘a Obi i ( » ee eT ree alts eorhyrh aie 0 rere eT Yoriety Lt oy) dane hae ro _ i bas hi a #). oe vith re ‘ mn writ aoe ees Ae) pe Mi : moins? i iw Ani a ae rid i ey f Testi ne “ ice i aity il ij ~ hon 4 : , ‘ 1 (1h) tp Tentyay fr n i Tine? et ‘ A een A : | ertey etna Fis pitied Menderes eke Vie hb ted en le i) Oe, we al i) Ce Ar Yanai arr" f buadat ly eee ; shu derah eee ea « 123 ey <i D Pe chide Sed iwnb ri ti sry med jee ace if (torre) yh word ewae t ror reve itery dae oI waty apes Facial i Gp} @ ibs vi, naiey bo] P cay xiiiolk (eh a td : 1, -aarpout Vihelainn sro ev) yrae ks : Ait cui hal hak hae Ronee Amba a my ’ mae ty oh iw os i. 1 } A. Atta cr Gas WaT rye, ; #4 i . py Ad S i cane ane, VV 7 OUT) Wier Rae oh, * Tieng, Ler ee es Ge mt . oe lt ey ity, Gah ater. om foal ra ee Bein by awitioy j tals he Foal) ute , , i rae ig, q nghe te iis. } é iP oye . t mr Or ar ee HM, a! elas. oy wwreKlioy ‘ ee Le A 8 : I { ; HOM. Hig ; te Cet OM tee 7 py Tap he j ‘ gi , ase ’ . aly vy A VM NE anh ty iy Aik iy » ‘ } ni" t Ales QA ak shone i oat) Ge ZOOLOGICAL SCIENCE 9: 757-764 (1992) Symbiosis between Cytaeis sp. (Hydrozoa) and Niotha livescens (Gastropoda) Starts during Their Larval Stage Cute! Inoue! and YosHIKO KAKINUMA~ ‘National Institute for Basic Biology, Laboratory of Reproduction, 38, Nishigonaka, Myodaijicho, Okazaki, 444, and * Department of Biology, Faculty of Science, Kagoshima University, 1-21-35, Kourimoto, Kagoshima 890, Japan ABSTRACT—The mechanism of symbiosis between Cytaeis sp. (Hydrozoa) and Niotha livescens (Gastropoda) has been investigated. The annual cycle of hydrozoan colonies on shells of gastropods was observed in the field while the initiation of the symbiosis was investigated in the laboratory. In the field, growth of colonies of polyps took place during spring, and formation of medusa buds in early summer. From summer to autumn the buds were liberated and the polyps degenerated into stolons. Copulation of N. livescens was observed from spring to early summer. In the laboratory, sumbiosis started at the planula and veliger stage. The rate of adhesion and metamorphosis of planulae on newly metamor- phosed gastropods reached a peak at 3 to 4 days after fertilization of the hydroids. The rate or reception of one-week planulae by N. livescens was highest at the gastropodan stage just after metamorphosis. The planulae settled on the surface of N. livescens at the base of the siphon, and then differentiated into polyps. The metamorphosed polyps elongated their stolons over the growing juvenile gastropod, and formed subsequent generations by asexual reproduction. © 1992 Zoological Society of Japan INTRODUCTION The hydrozoan Cytaeis sp. colonizes the surface of shells of living Niotha livescens. It has been reported that the differentiation from polyp to medusa of Cytaeis sp. depends on the physiological activity of N. livescens in the breeding season [1]. However, it is still unknown when and how the symbiosis occurs between the two species. To investigate the mechanism of symbiosis between the two species, we first observed their interrela- tion in the field, as defined seasonally by the appearance of gastropods, the formation of polyp colonies on the shells, and the formation of medu- sa buds from polyps. Next we investigated the initiation of symbiosis in the laboratory. Accepted June 2, 1992 Received April 8, 1992 " To whom reprint requests should be addressed. MATERIALS AND METHODS Field observations Gastropods, Niotha livescens, covered with a colony of the hydroids, Cytaeis sp. (Fig. 1B), were collected from the sand shore of Yojiro-ga-hama in Kagoshima city (Kagoshima Pref., Japan). Gas- tropods were captured with fresh fish pieces at several points on the shore every month (Fig. 1A). Within 30 minutes about 20 to 50 gastropods gathered around the bait. Of these, the number of gastropods with a colony of hydrozoa averaged 5 to 7. The pattern of colony distribution and the number of polyps and/or medusa buds were re- corded on schematic drawings of the shell and are summarized in Fig. 2. The animals were released at the same points of capture during following day. Laboratory observations Gastropods with a colony of hydrozoa were cultured in laboraotry aquaria with circulating seawater of 25°C. N. livescens were fed with fish and Cytaeis sp. with brine shrimp. Freshly Fic. Number of polyps and medusa buds Fic. 1. The hydrozoan colony on shells of N. livescens. A. The gastropods were trapped with fish meat. Scale, 1 cm, B. Hydrozoan colony on a shell. Scale, 1 mm. nN oO Temperature (°C) 15 400 tl 10 300 200 100 0 JFMAMJJASOND J FM 1985 1986 2. Annual reproductive cycle of Cylaeis sp. and N. livescens. The largest number of polyps or medusa buds on captured gastropoods of each collection was plotted. Ordinate left; Number of polyps (—@—), medusa buds (—O—), and right, temperature (—A—). Abscissa: Months in which gastropods were collected. N. /. (@), breeding season of N. livescens. C. INOUE AND Y. KAKINUMA spawned eggs from N. livescens (3 individuals) were collected from the walls of the auqgaria (about 20 liters) and transferred to petri dishes where they were allowed to develop into veligers and were then hatched. As soon as the gastropods spawned eggs, another group of adult gastropods carrying col- onies of polyps were transferred to petri dishes (about 8cm in height) and kept at 25°C. They were fed brine shrimps to obtain hydrozoan medu- sae. Newly released medusae were also fed on brine shrimp. Spawning was induced by exposing mature medusa, previously kept in the dark, to light. Fertilization took place without further treatment. The resulting fertilized eggs were allowed to develop, and subsequent planula larvae were maintained and used in the experiments. The behavior of planula larvae, young polyps, veligers and juvenile gastropods after the estab- lishment of symbiosis was continuously observed. In order to examine the ability of attachment of planulae, one gastropod was mixed with 6 planulae of the same age, just after metamorphosis and the number of attached planulae was counted within 10 days (Fig. 4). Six trials were carried out, each with planulae of a different age. Seven-eight hr old planulae were designated as 0-day planula. In order to examine the rate of reception of 1-week planula by N. livescens of various ages, six 1-week planulae were mixed with one gastropod of vari- able age (Fig. 5, arrows: time point of mixing). The number planulae found to be attached within 10 days was counted. Six trials, each with gastro- pods of a different age were carried out. Cytological examinations of planulae Planulae were fixed in glutaraldehyde (2% in sea water) for 1 hour at room temperature. They were dehydrated in graded ethanol series, embedded in JB-4 plastic medium (Polyscience Inc.), and sec- tioned at 1-2 um with glass knives. The sections were stained routinely with Delafield’s hematoxy- lin and eosin and observed under a light micro- scope. Symbiosis between Hydrozoa and Gastropoda 169) RESULTS Coexistence of the two species in the field The number of polyps varied between gastro- pods, and fluctuated throughout the year. The maximum number of polyps and medusa buds for each observation through the year are shown in Fig. 2. In January of 1985 no polyp was observed. In February and April, a total of 47 and 90 polyps were recorded respectively. In May, the polyps (336) covered most of the surface of the shells, and 122 medusa buds were counted. In June and early July, the number of polyps was 255 and 234, respectively. During the same period, 160 and 132 medusa buds were counted. In late July, the number of polyps had decreased to 4. Medusa buds were not observed. These results show that the colony of polyps degenerates from autumn to winter, and appear again in early spring. From spring tc summer, they grow and cover the surface of shells. Medusa buds are formed during early summer and are probably released in the same period. From summer to winter, the number of polyps fluctuate, but medusa buds are not observed in this period. WN. livescens displayed copulation from spring to early summer, which coincided with the appearance of hydrozoan medusae. Interrelation of the two species in the laboratory Egg-laying and the development of N. livescens The total number of spawned eggs collected was 722 from May to July, 1985, and about 300 from July to August, 1986. The diameter of the egg capsule was about 350 ~m. Embryos inside the egg capsules developed normally (larval shells, vela, eyes, feet, and siphons). Embryos hatched and developed into veliger larvae about 15 days after egg-laying (25°C). The earliest veliger, with a small velum and no siphon, was observed on the 9th day after egg-laying. Veligers in this stage displayed only planktonic (swimming) behavior and were designated as early veligers. Veligers 15 to 16 day old were observed to have a broad velum and a siphon, and were designated as late veliger. From 17 to 30 days after egg-laying, they metamorphosed to juvenile gas- tropods, revealed by withered vela, and entered into completely benthonic life. Development of Cytaeis sp. Colonies of Cytaeis sp. formed medusa buds during the breeding season of N. livescens. The medusa buds grew for about 1 week up to 1 mm in diameter and then separated from the colony. About 4 days after their release (25°C) they spawned gametes in early morning. The diameter of the eggs was about 100 um. The living planulae (Fig. 3A) and the sections (Fig. 3B) gave the following observations. At 7 to 8 hr after spawning (20°C), the planulae were just after hatching. Blastocoels were still present and differentiation of endoderm by polarization at the posterior ectoderm had begun to occur. The total number of the cells was small. In 1-day planulae, blastocoels had disappeared, and differentiation of endoderm from ectoderm was completed. How- ever, nematocysts were not observed. The planu- lae began to swim at this stage. In 4-day planulae, the nematocysts were recognized at the vegetal pole. The planulae swam most actively in this stage, whilst the number of cells was highest. The cells contained many secretions which stained red with eosin. Twenty days after spawning, the size of planulae and the number of cells decreased and vacuoles were observed in the endoderm. Sixty days after spawning, the planulae gradually be- came motionless and were observed to be lying on the surface of the sediment. The shape of ectodermal cells became flat and the endoderm degraded with a large central vacuole. Rate of attachment of planulae of various ages on newly metamorphosed N. livescens Planulae did not attach to shells within the first 1 or 2 days after mixing (Fig. 4). The rate of attachment of 0-day planula was 40%. The highest rate of attachment was obtained with 4-day planu- lae. At this age almost all planulae were attached to the shells within a few days after mixing. Ten-day planulae still showed a high rate of attach- ment, but the rate decreased rapidly with increas- ing age of planula. Sixty-day planulae did not show any capability of attaching to shells. 760 C. INOUE AND Y. KAKINUMA Fic. 3. Early development of Cyraeis sp. A. Living planulae (bright-field micrographs). B. l-m sections of planulae (hematoxylin-eosin stained). 1, 2, 3, 4, and 5. 7-8 hours, 1 day, 4 days, 20 days, and 60 days after fertilization, respectively. Scale, 50 «m. Symbiosis between Hydrozoa and Gastropoda —_ 00 50 Attachment of planulae (%) 0 10 20 30 40 50 60 Age of planulae (days) Fic. 4. Rate of attachment of planulae of different ages on newly metamorphosed gastropods. Ordinate: Attachment of planulae (%). Abscissa: The age of planulae (days). One gastropod just after meta- morphosis was mixed with six planulae at different ages. Rate of reception of I-week planula by N. livescens shell of various ages The veliger stage of N. livescens was observed to begin on the ninth day after egg-laying (Fig. 5, A) while metamorphosis took place about 1 week after the beginning of the veliger stage (Fig. 5, B). The oldest veliger was observed 27 days after egg-laying (Fig.5, C). Therefore, the period of metamorphosis could be regarded as the double- shadowed area shown in Fig. 5 (between B and C). When the planulae were mixed with 10-day gastro- A B Cc ~ 109 “Lt hy Reception of planulae (%) ao (=) i Yj Age of gastropods (days) Fic.5. Rate of reception of one-week planulae by gastropods of various ages. Six l-week planulae were mixed with one gastropod of various ages. Ordinate: Reception of planulae (%). Abscissa: The age of the gastropods (days). Arrows indicate time point of mixing. 761 pods, the rate of reception was 40%. In this case, no planulae attached to the shell within the first 5 days after mixing. Consequently, the data-plotting should not have begun until the 15th day after egg laying, and the data were therefore re-plotted with open cricles. The highest rate or reception was observed during the period of metamorphosis. The rate of reception remained at 50% throughout the early juvenile stages. Settlement, metamorphosis and colony formation of Cytaeis sp. As soon as the juvenile gastropods and planulae encountered each other, the gastropods showed an active movement and then immediately became motionless. On the other hand, the planulae displayed position-searching behavior, initially creeping around the siphon and finally selectively settling at the base of the siphon (Fig. 6). Most planulae settled on this place of the shell. 1 Fic. 6. The location of settlement of planulae on young gastropods. Note the display of position-seaching behavior by planula on a motionless gastropod. Scale, 100 un. The planulae metamorphosed into a polyp with- in 4 days after settlement on the shell (Fig. 7). The process of the metamorphosis was divided into 5 C. INOUE AND Y. KakINUMA Rate of metamorphosed hydrozoa (%) (3)O0 @o 00 oe 9 , ’ ’ . ff ate 1 '¢ a Ge TR 4 we tae Ar ‘ tf uw 4 , ’ ve s . ‘ ‘ S 5 2 3 1 Days after attachment Rate of metamorphosed individuals was expressed as %. Fic. 7. Metamorphosis of Cytaeis sp. on the shell. Ordinate: Stages of metamorphosis of hydrozoa. Abscissa: Days after attachment. Scale, 100 pm. I-V. Stages of metamorphosis (see text). 26 28 (days) 18 Formation of the polyp colony on the shell. Numbers indicate days after attachment of planula on the shell. 1 8 Fic. 8. Scale, 1 mm. Symbiosis between Hydrozoa and Gastropoda 763 stages: 1; up to just after settlement, 2; beginning of formation of the gastric cavity, 3; completion of the gastric cavity, 4; formation of tentacle rudi- ments, and 5; completion of metamorphosis to a polyp with tentacles. Typically (43% of observa- tions), a planulae metamorphosed to a polyp with- in 2 days along the course shown by the thickened line in Fig. 7. When planulae were cultured in a petri dish, almost all of the planulae were on the bottom of the dish and died. Only a few planulae differentiated into a polyp (about 0.01%). As mentioned above a polyp initially located at the base of the siphon. This polyp elongated a stolon (initial one) in the direction of the shell growth. The second polyp was then formed, again at the base of the siphon, around 15 days after the first planula settlement. After the second polyp formation, the initial polyp degenerated and be- came engulfed by the growth of the shell (Fig. 8). Thus, the metamorphosed polyp produced subse- qunet generations of the polyp colony asexually. Food was considered to be an important factor for the asexual reproduction of Cytaeis sp., because lack of food (brine shrimp, in the present study) hindered polyp formation. Rate of survival and growth of N. livescens The symbiotic relationship and the rate of sur- vival of the gastropods were examined ex- perimentally. First, veligers were mixed with 1-week planulae. Out of 86 veligers, the number of veligers on which planula settled was 18, of which 14 veligers metamorphosed and grew while 4 died before metamorphosis. Of the remaining 68 veligers, on which planulae did not settle, 18 veligers died before metamorphosis while 50 veli- gers metamorphosed and grew. After metamorphosis, a second trial was carried out, in which gastropods which were still without hydrozoa were mixed with 1-week planulae. Among 50 gastropods, the number of gastropods which were settled with planula was 37 and these gastropods all survived. Among the rest, i.e. 13 gastropods on which planulae did not settle, 9 gastropods died within 10 days and gastropods grew (Fig. 9). The survival rate of the gastropods with VP* (Veligers with planulae) and GP* (young gastropods with planulae) or without VP~ 37) GP +,100%) | 4(GP ,30%) Number of veligers or gastropods o o Lo | <+——Veliger stage ————> | ~~— Young gastropod stage —> Metamorphosis Fic. 9. Symbiotic relationship and the rate of survival of the gastropods. VP*: Veligers with planulae, VP: Veligers without planulae, GP*: young gas- tropods with planulae, GP: young gastropods with- out planulae. Ordinate: length of each line (i.e. numeral on each line) indicates numbers of veligers or gastropods. Abscissa: stage of the gastropod. a (Veligers without planulae) and GP (young gas- tropods without planulae) was estimated as 89% [(77% + 100%) /2] and 52% [(73% +30%)/2], re- spectively. If the calculation is simply done at the last observation (i.e. the right-most line of Fig. 9), the rate of survival of gastropods without hydroids is about 7% [4/(14+37+4) x 100]. DISCUSSION In the laboratory symbiosis took place between planula and late veliger stage respectively. The ability of planulae to go through metamorphosis appears to be maintained for about 20 days and then gradually disappears. This in turn implies that suppression of metamorphosis also lasts for this period provided a stimulation for triggering metamorphosis is not given. As for stimulation, it has been reported that in Hydractinia echinata the stimulus causing initiation of metamorphosis emanates from certain marine bacteria [2]. The stimulus for Cytaeis sp. may also be a chemical factor, because the planulae of Cytaeis sp. always settle at the base of the siphon of N. livescens. It is considered that a receptive system for a metamorpnosis-triggering factor must exist in planulae. It has been reported that neural dif- 764 C. INOUE AND Y. KAKINUMA ferentiation begins in the planular ectoderm 24 hours after fertilization and continues throughout larval development [3, 4]. One reason explaining the inability of 60-day planulae to attach is consi- dered to be a lack of swimming ability which reduces the chance for planulae to encounter the gastropods. It is also possible that in 60 day planulae a degeneration of the neural system re- duces their ability to perceive a stimulation factor. As mentioned before, a few planulae differenti- ated into a polyp when cultured in a petri dish. It is considered that metamorphosis of Cytaeis. sp. can be triggered by other factors, and once triggered, metamorphosis seems to progress without any further exogenous stimulation. However, in the field we have never seen a colony of Cytaeis sp. except on the surface of the N. livescens. The larvae of N. livescens clearly offer a substra- tum on which the planulae adhere and meta- morphose. Since the planulae and the juvenile gastropod display specific behavior as soon as a planulae and a juvenile gastropod encounter each other, and since the position of planula settlement is always at the base of the siphon, we can specu- late that the gastropods secrete some planula- attacting substance from the base of the siphon, to which the palnula reacts. On the other hand, the hydrozoan planula may secrete some factor to make the gastropod motionless. Although one trial was not enough to determine the survival rate, the rate of survival of gastropods that are attached by planula tends to be higher than that of gastropods without planulae. There- fore the adhesion of planula may provide some adequate condition for the gastropodan physiology as Hirai and Kakinuma previously reported for the adult sage [1, 5]. Among the collected gastropods, the percentage of gastropods without hydrozoa is fairly high (about 70%). And much higher than calculated from the laboratory trials (finally about 7%). Therefore, in the laboratory, symbiosis may be preferable for the gastropod particularly in the juvenile stage. Since in the field the breeding seasons of the two species coincide with each other, symbiosis be- tween the two species must take place at the larval stage, although the gastropods can survive without hydrozoa. Whether the loose symbiosis between these species inclines to a more strict relationship or declines towards dissolution requires long term observation. For the hydroids, however, the cur- rent relation seems to be necessary to survive and to diverse in the field. ACKNOWLEDGMENTS We thanks Dr. H. Shirai (Okayama University) for critical reading of the manuscript, Messrs. M. Tabata and Y. Ikimori (Kagoshima Marine Park) for help in collect- ing the gastropods, and Dr. J. Komen and Dr. Cocket (National Institute for Basic Biology) for reading the manuscript. REFERENCES 1 Hirai, E. and Kakinuma, Y. (1971) On symbiotic relations between differentiation of a hydrozoan Cytaeis uchidae with a gastropod Niotha livescens. Bull. Mar. Biol. Stat. Asamushi, 14: 65-77. 2 Spindler, K. D. and Muller, W. A. (1972) Induction of metamorphosis by bacteria and by a lithium-pulse in the larvae of Hydractinia echinata (Hydrozoa). Wilhelm Roux’ Archiv 169: 271-280. 3 Martin, V. J. (1988) Development of nerve cells in hydrozoan planulae: I. Differentiation of ganglionic cells. Biol. Bull. 174: 319-329. 4 Martin, V. J. (1988) Development of nerve cells in hydrozoan planulae: II. Examination of sensory cell differentiation using electron microscopy and im- munocytochemistry. Biol. Bull. 175: 65-78. 5 Hirai, E. and Kakinuma, Y. (1973) Differentiation and symbiosis in two hydrozoans. Publications of the Seto Marine Biological Laboratory, 20: 257-273. ZOOLOGICAL SCIENCE 9: 765-773 (1992) Effects of Olfactory Tract Section on Brain GnRH Distribution, Plasma Gonadotropin Levels, and Gonadal Stage in Goldfish Maxito Kopayasui!, MAsSAFUMI AMANO!, YoOSHIHISA HASEGAWA2 Koicut Okuzawa? and Katsumi Arpa! ‘Department of Fisheries, Faculty of Agriculture, University of Tokyo, Bunkyo, Tokyo 113, *Department of Obstetrics and Gynecology, Gunma University, School of Medicine, Maebashi, Gunma 321-16, and *Inland Station, National Research Institute of Aquaculture, Tamaki, Watarai, Mie 519-04, Japan ABSTRACT— Goldfish in early sexual recrudescence were olfactory tract-sectioned (OTX) or sham- operated in order to examine the distribution of brain gonadotropin releasing-hormone (GnRH) originating from the terminal nerve (IN), and the involvement of GnRH in reproduction. Fish were kept at 20°C, and one month following surgery, brain contents of two types of GnRHs, salmon type (sGnRH) and chicken-II type (CGnRH-II), were measured by specific radioimmunoassays. Plasma gonadotropin (GTH) levels and the stages of gonadal development were also determined. After OTX, sGnRH contents showed a marked decrease in all brain areas (telencephalon, hypothalamus, optic tectum-thalamus, cerebellum, medulla oblongata) except in the olfactory bulbs, whereas cGnRH-II contents in the brain and the pituitary gland showed no clear changes. Pituitary sGnRH contents and plasma GTH levels showed slight decreases in OTX males, but not in females. In spite of large decreases in brain sGnRH contents, gonadal development in OTX fish was as active as that in sham-operated fish with the progression of spermatogenesis in males and vitellogenesis in females. These results suggest that most of the sGnRH in the brain originates from the TN, and that TN-GnRH is © 1992 Zoological Society of Japan not essential to gonadal development in goldfish. INTRODUCTION Gonadotropin releasing-hormone (GnRH) is a decapeptide which regulates reproductive activi- ties, such as the stimulation of gonadotropin (GTH) release from the pituitary gland [1] and the potentiation of sexual behavior [2]. Recent studies have shown that more than one type of GnRH exists in the brain of teleosts, and most of the species examined possess salmon type (sGnRH) and chicken-II type (CGnRH-II) [3]. Immunocyto- chemical studies revealed that major populations of GnRH-producing cells in teleosts are localized in the preoptic area, the terminal nerve (TN, homologous to the nucleus olfactoretinalis), the midbrain tegmentum, and the nucleus lateralis tubelis [4-12], although particular GnRH type was Accepted April 19, 1992 Received February 14, 1992 not necessarily examined in all the cases, in con- text of the antibodies used in each study. Among these GnRH populations, the GnRH system in the preoptic area is considered to be morphologically equivalent to that of the pre- optico-infundibular pathway of other vertebrates [13]. The GnRH in this system is presumed to function as a hypophysiotropic hormone in the stimulation of GTH release [7]. However, direct evidence of the release of GnRH from this system has not been obtained in teleosts. Since teleosts lack the hypothalamo-pituitary portal system, the measurement of GnRH near its site of release is quite difficult. Many studies on the TN system in teleosts have been conducted regarding morphological aspects [14], but the physiological role of the system is not fully understood. Demski and Northcutt [15] suggested that the TN system mediates the be- havioral response to sex pheromones in goldfish. 766 M. KosayAsut, M. AMANo et al. However, this possibility now seems less probable, since it has been demonstrated that chemosensory responses to sex pheromones in the goldfish are mediated by the medial olfactory system instead of the TN system [16]. Although the TN system is known to produce GnRH, it is not clear whether GnRH originating from the TN is involved in gonadal maturation through GTH secretion. There are some studies on changes in brain GnRH contents in relation to gonadal maturation. However, the results of these studies are not consistent. In caribe colorado, Pygocentrus notatus [17] and masu_ salmon, Oncorhynchus masou [18], brain GnRH contents (sGnRH in masu salmon) were high when gonads were mature, but no clear correlation was observed between brain GnRH contents and gonadal maturity in eel, Anguilla anguilla [19], goldfish, Carassius auratus [20] and rainbow trout, Oncorhynchus mykiss [21]. In brown trout, Salmo trutta [22], a correlation was observed between pituitary GnRH contents and plasma GTH levels. These studies measured brain and pituitary GnRH contents without considering the origin of GnRH fibers, and moreover, most of these studies were conducted using radioimmunoassay (RIA) whose antibodies were not specific to different types of teleost GnRH. In the goldfish, two types of GnRH, sGnRH and cGnRH-II, were chromatographically and im- munologically identified in the brain [23]. Major populations of GnRH-producing cells were found in the IN, preoptic area, midbrain tegmentum, and nucleus lateralis tubelis by immunocyto- chemical studies [4, 7]. Since the olfactory bulbs of goldfish are of the pedunculated type with elon- gated olfactory tracts, we can easily block the axonal transport of GnRH from the TN to other brain regions by sectioning the olfactory tracts. In the present study, the effects of olfactory tract section were studied in goldfish in order to ex- amine the brain distribution of GnRH of the TN. Brain GnRH was measured by RIAs using highly specific antibodies to sGnRH and cGnRH-II [21]. Plasma gonadotropin (GTH) and gonadal de- velopmental stages were also determined in order to examine whether TN-GnRH was involved in gonadal development. MATERIALS AND METHODS Goldfish Goldfish were obtained from a commercial source and kept in stock tanks under natural condition. Fish weighing 30-40 g were used for the experiment. Experimental Protocol Fish in early sexual recrudcence were olfactory tract-sectioned (male N=13; female, N=7) or sham-operated (male, N=13; female, N=5) on October 19. Initial body weights of each group were as follows (mean+SEM g): male; sham 34.9 +0.8, OTX 33.7+0.8: female; sham 32.1+1.8, OTX 32.4+1.9. The experiment was originally designed to use males since the involvement of olfactory system in reproduction has been reported mostly in male goldfish [15, 24, 25]. Female groups were supplemtned as an additional part of the experiment, although sample size was small. For olfactory tract section, four-sided flap was cut in the frontal bone using a disc saw. The first cut was made aligned with the center of the iris. If this cut were made more posteriorly, damage to the pineal gland would be likely to occur. The second cut, parallel to the first, was made just behind the nares. The third and fourth cuts, parallel to the longitudinal axis of the fish, con- nected to the first and the second cuts. After removal of a square flap of the frontal bone, fat and cranial fluid were gently wiped out with paper. The olfactory tracts were bilaterally sectioned at two places with Wecker’s scissors (Fig. 1), and the resultant sections were removed to prevent regen- eration of the olfactory tract fibers (OTX fish). The cavity resulting from the operation was filled with gelatin sponge (Spongel, Yamanouchi Phar- maceutical Co.). A sham operation was per- formed in the same manner without cutting the olfactory tracts (sham-operated fish). After the operation, fish were kept in 25 ppm oxyteracyclin (Sigma) overnight, and then trans- ferred to a 60 liter experimental tank (males and females in the same tank) which was kept at 20°C under ISL (lights on at 0500 hr). Fish were fed with commercial trout pellets once a day. Fish Brain GnRH in OTX Goldfish 767 were sampled on November 20. After being anesthetized with 0.02% tricaine methansulfonate, their body weights were measured. Blood samples were taken from caudal vasculature with a hepari- nized needle and syringe. Fish were sacrificed by decapitation, and the brain and pituitary glands were collected. The brains were dissected into six parts (olfactory bulbs, telencephalon, hypothala- mus, optic tectum-thalamus, cerebellum, medulla oblongata) as shown in Fig. 1. The tissues were frozen on dry ice after being weighed. The gonads were dissected out and weighed to determine the gonadosomatic index (GSI). For histological observation, the gonads were fixed with Bouin’s solution and embedded in paraffin. Sections were stained with hematoxylin and eosin. Ovarian developmental stages were identified according to the classification by Yamamoto and Yamazaki [26]. PIT Fic. 1. Schematic diagram of a sagittal section of goldfish brain showing sites of severance of the olfactory tracts (arrows) and disection of the brain for the determination of GnRH. OLT, olfactory tracts; OB, olfactory bulbs; TEL, telencephalon, including preoptic area; OT-THAL, optic tectum- thalamus, including midbrain tegmentum; CER, cerebellum; MED, medulla oblongata; PIT, pituitary. RIA Salmon GnRH and cGnRH-II were measured by specific RIAs as described by Okuzawa et al. [21]. The cross-reactivity of cGnRH-II in the sGnRH RIA (antibody to sGnRH, Lot No. 2) was 1.58%, and the cross-reactivity of sGnRH in the cGnRH-II RIA (antibody to cGnRH-II, Lot R-E) was 0.051%. Measurement of GnRH in the goldfish brain by the RIA was validated by examing parallelism of the displacement curves of goldfish brain extract. High performance liquid chromatography (HPLC) analysis followed by RIA was also employed for the validation of the RIA. To test parallelism, goldfish brain tissue was homogenized by sonica- tion in 0.1 N HCl. The homogenate was centri- fuged at 10,000 xg for 30 min at 4°C. The super- natant was frozen, lyophilized, and reconstituted in assay buffer. The supernatant obtained after centrifuging again at 10,000 g for 30 min at 4°C was used for the GnRH RIAs. The brain extract of goldfish produced displacement curves which Salmon GnRH RIA 100 Dilution of brain extract _———_—— el Brain extract = 50 Salmon GnRH {e) gl a 0 9.8 39 156 625 (pg/ml) Chicken GnRH-Il RIA 100 Dilution of brain extract Brain extract SS = 501 Chicken GnRH-II a oO 9.8 39 156 625 (pg/ml) Fic. 2. Top. Competition curves for salmon GnRH and brain extract of goldfish in the salmon GnRH RIA. Bottom. Competition curves for chicken GnRH-II and brain extract of goldfish in the chicken GnRH-II RIA. The scale for dilution of brain extract indi- cates a twofold serial dilution. Each point repre- sents the mean of duplicate determinations. 768 M. Kopayasui, M. AMANO et al. were parallel to sGnRH and cGnRH-II standards in respective RIAs (Fig. 2). For the measurement of brain GnRH contents of the experimental fish, the brains were extracted in the same method as for the paralellism test. For HPLC analysis, the brain was extracted as described by Okuzawa et al. [21]. The extract was injected through a 2 ml-injection loop onto a TSK- ODS 80TM column (0.4625 cm, Tosoh Co. Ltd., Tokyo, Japan). A Hitachi high performance liquid chromatography (Model 655 A) was pro- grammed for a flow rate of 1 ml/min. The extract was applied at the beginning of a 10-min isocratic period of 17% acetonitrile (CH3CN) in 0.1% trifluoroacetic acid; CH3CN was then increased to 27% over a 10-min period. GnRHs were eluted under isocratic conditions at 27% CN3CN. Frac- tions (1 ml) were collected and aliquots were assayed for GnRH. HPLC analysis revealed that goldfish brain contained a peptide chromato- graphically and immunologically identical to sGnRH and a peptide identical to cGnRH-II (Fig. 3). Although there was some immunoreactive materials which were chromatographycally diffe- rent from sGnRH and cGNRH-II in the goldfish brain, their amount was considered negligible, and therefore it was not necessary to separate sGnRH and cGnRH-II by HPLC prior to the RIAs. Plasma GTH levels were measured using silver carp GTH RIA as described by Kobayashi et al. [27]. Purified silver carp GTH (Lot SS) was used as a standard and for iodination [28], and an antibody (Lot No. 1) raised against the silver carp GTH (Lot SS) was used for the RIA [29]. Valida- tion of the RIA system for use in goldfish was achieved by obtaining parallelism for a serial dilu- tion of plasma samples collected from ovulated females. Recoveries at three different doses, high, middle, and low, were 95.2, 102.2, and 115.4%, respectively. The GIH measured in this study is considered to be equivalent to GTH II according to Kawauchi’s designation [30] by physicochemical characters of the silver carp GTH [28], an im- munocytochemical study of goldfish pituitary GTH cells using the silver carp GTH antibody (Lot No. 3) [29], and plasma profiles of GTH during ovula- tion in goldfish measured by the silver carp GTH RIA [31]. Salmon GnRH RIA Salmon Chicken-ll 4 > 8 5 3 3 3 = S z= Chicken GnRH-II RIA H 5 I I Elution time (min) Fic. 3. Reverse-phase HPLC of goldfish brain extract followed by salmon GnRH RIA (top) and chicken GnRH-II RIA (bottom). Arrows indicate the elu- tion time of synthetic chicken GnRH-II and salmon GnRH. The mobile phase was CH3CN containing 0.1% TFA. Statistics Statistical analysis of the results were done by Student’s t-test or Cochran-Cox method. RESULTS One month following surgery, the experimental fish exhibited gains in body weight; and there were no significant differences in final body weights between sham-operated and OTX fish (mean+ SEM g): male; sham 42.6+0.7, OTX 43.7+1.4; female; sham 41.8+2.8, OTX 44.4+2.2 (Fig. 4). After OTX, sGnRH contents in OTX fish of both sexes showed a marked decrease (P<0.01) in all brain areas (telencephalon, hypothalamus, optictectum-thalamus, cerebellum, medulla oblon- gata) except in the olfactory bulbs (Fig. 5 and 6). Chicken GnRH-II contents in OTX fish showed a tendency toward decrease, but only the decrease in hypothalamus of males was significant (P<0.05). Brain GnRH in OTX Goldfish 769 Male Female Body weight 60 y g 40 i) 20 S A S N 13 13 7 GSI 4 3 & 2 if S A S N 12 13 5 7 Plasma GTH 20 = * ro.) Cc — 10 S S 0 N 12 12 BY Fic. 4. Effects of olfactory tract section on body weight, gonadosomatic index (GSI), and plasma levels of gonadotropin (GTH) in goldfish. S, sham- operated fish; O, olfactory tract-sectioned fish. Each column represents the mean and SEM. Level of significance; *, P<0.05. Salmon GnRH contents in the olfactory bulbs showed no significant difference between sham- operated and OTX fish in both sexes. Chicken GnRH-II contents of the olfactory bulbs were mostly low and no clear change was detected after the operation. Pituitary sGnRH contents showed decreased levels in OTX males (P<0.05), but not in females. Pituitary cGnRH-II contents showed no significant changes after the operation in both males and females. Plasma GTH levels in OTX males were slightly lower (P<0.05) than those in sham-operated males, but no significant difference was observed in females (Fig. 4). There was no significant differences in GSI values between OTX and sham-operated fish in both sexes (Fig.4). Histological observation showed that under conditions which are favorable for gonadal maturation in goldfish (20°C, 15L) both OTX and sham-operated fish had developing gonads regardless of the difference in brain GnRH contents. All males had spermatogenic testes and lobules filled with spermatozoa. The oocytes of the ovaries in all females had advanced to the primary yolk stage except one OTX fish which had the ovaries with secondary yolk stage oocytes. DISCUSSION The brain GnRH distribution and contents in sham-operated goldfish showed a pattern similar to those of goldfish reported in other studies [20, 23] and those of salmonid fishes [18, 21]: Salmon GnRH was distributed in a larger amount in the olfactory bulbs, telencephalon, hypothalamus, and the pituitary gland than in other regions, whereas cGnRH-II was distributed widely throughout the brain with the highest concentration in the medulla oblongata. The difference between goldfish and salmonid fishes is that the goldfish pituitary con- tains cGnRH-II, which is absent in salmonid pituitaries [12, 18, 21]. After OTX, brain sGnRH contents showed a marked derease in all brain areas except in the olfactory bulbs. These results suggest that sGnRH in the goldfish brain mostly originates from the TN. That is, sGnRH contents in various brain regions reflect sGnRH in fibers which are trans- ported from the TN cell bodies in the olfactory bulbs. Immunocytochemical studies in other spe- cies support this hypothesis. In the catfish, Clarias batrachus [9] and the dwarf gourami, Colisa lalia [11] GnRH fibers from the TN send projections to various regions of brain parts. Electrical lesion of the TN cells in gourami caused disappearance of most of immunoreactive fibers in the brain leaving GnRH cells and fibers in preoptic area intact [32]. Furthermore, it has been also demonstrated by 770 GnRH (ng/mg tissue) M. KopayasHi, M. AMANO et al. OB TEL HYP OT CER MED PIT THAL Salmon GnRH 70 60 150 50 40 100 30 as 20 50 10 0 s O Chicken GnRH-II Fic.5. Effects of olfactory tract section on salmon GnRH and chicken GnRH-II concentrations in discrete brain areas in male goldfish. OB, olfactory bulbs; TEL, telencephalon, including preoptic area; OT-THAL, optic tectum-thalamus, including midbrain tegmentum; CER, cerebellum; MED, medulla oblongata; PIT, pituitary. S, sham-operated fish (N=12); O, olfactory tract-sectioned fish (N=13). ND, Nondetectable. Each column represents the mean and SEM. Level of significance: *, P<0.05; **, P<0.01; ***, P<0.001. GnRH (ng/mg tissue) OB TEL HYP OT CER MED PIT THAL Salmon GnRH 40 100 30 15 20 10 50 10 5 * * * * * * * * 0 0 0 Ss oO Ss Oo Ss Oo s O Ss Oo s Oo Chicken GnRH-II Fic. 6. Effects of olfactory tract section on salmon GnRH and chicken GnRH-II concentrations in discrete brain areas in female goldfish. Abbreviations and symbols as in Fig. 5. S, N=5; O, N=7. Brain GnRH in OTX Goldfish 771 intracellular staining in gourami that a single TN- GnRH cell projects to all brain regions where GnRH fibers has been demonstrated but not to the pituitary [33]. Other immunocytochemical studies are also suggestive of the existence of projections of GnRH fibers from the TN to other brain regions [5, 7]. It seems likely that a common anatomical feature in teleosts is projection by the TN system of some type of GnRH to various regions of the brain, although the physiological function of such GnRH remains to be elucidated. It is not clear from the present study whether TN cells of goldfish produce cGnRH-II. Chicken GnRH-II contents were mostly undetectable in the olfactory bulbs and showed no clear changes after OTX in the other brain areas except in the hypothalamus of males. In masu salmon, cell bodies of cGnRH-II neurons were localized in the midbrain tegmentum, and only a few fibers were found in the olfactory bulbs [12]. Small amounts of cGnRH-II in the goldfish olfactorty bulbs mea- sured by RIA may be provided by fibers from other regions of the brain, such as the midbrain tegmentum. To confirm the anatomical distribu- tion of cell bodies and fibers of sGnRH and cGnRH-II immunocytochemical approach is required. Although OTX caused a marked decrease in brain sGnRH contents, no gonadal regession was observed both in males and females in contrast to the example of hypophysectomized female goldfish [34]. There was no difference in the stages of gonadal development between OTX and sham- operated fish. Under conditions which enhance gonadal maturation of goldfish, fish of both groups showed active gonadal development which would not normally occur during non-spawning period. These results suggest that TN-sGnRH is not essen- tial to gonadal development in goldfish. Although abnormal changes were seen in the gonads after lesion of nucleus olfactoretinalis in platyfish, Xiphophorus maculatus [8], reasons for different responses to a similar treatment between the two species is unclear. OTX males showed decreased levels of pituitary sGnRH which may be responsible for the decrease in plasma GTH levels. There are some possible interpretations for this decrease, such as blockade neurons, an of a direct innervation of sGnRH fibers from the TN to the pituitary gland, or a removal of stimula- tion to sGnRH neurons which send fibers to the pituitary, etc. 17a,20-Dihydroxy-4-pregnen-3- one, one of the goldfish sex pheromones, is known to stimulate GTH secretion in males during spawn- ing [35]. If female goldfish release sex pheromones which stimulate basal GTH release in males, the decrease in the pituitary sGnRH contents and plasma GTH levels in OTX males may be partly explained by inability in receiving sex pheromones since OTX fish were anosmic in the present study. However, a decrease in pituitary sGnRH contents and plasma GTH levels to the extent observed in this study did not seem to affect basal testicular development. As well as the gonadal stages, no differences were observed in body weight between OTX and sham-operated fish. Reduction of the number of feeding acts (bites directed toward gravel and artificial vegetation) after OTX is reported in goldfish [24], but the actual feeding of the OTX fish in the present study did not seem to be impaired by OTX, since both sham-operated and OTX fish gained weight similarly during the one- month experimental period. It is known in the goldfish that GnRH stimulates growth hormone release from the pituitary gland [36]. The results of present study suggest that GnRH in the TN is not likely to be involved in the regulation of growth hormone secretion. Neuromodulatory effects are suggested as one of physiological func- tions of TN-GnRH in teleosts [11, 33], but what aspects of brain activities are modulated by the TN-GnRH should be elucidated. Although there are some studies on changes in brain GnRH contents in relation to gonadal maturation, a clear correlation between brain GnRH content and gonadal maturity has not al- ways been observed [17-22]. Since the present study suggests that most of sGnRH contents in the brain reflect those in fibers originating from TN cell bodies, changes in GnRH contents measured in those studies may reflect the activities of GnRH producing-cells in the TN system. In fact in masu salmon, sGnRH contents of most of discrete brain areas showed seasonal changes parallel to those of the olfactory bulbs [18], where cell bodies of 772 sGnRH neurons shown by immunocytochemistry to be abundant [12]. More recently, Suzuki et al. [37] revealed by in situ hybridization techniques that the intensity of hybridization signals were greater in the olfactory neurons and the olfactory bulbs than in the hypothalamic area in masu salmon. These results combined with the present study suggest that changes in the contents of GnRH attributed to the TN system seem to mask changes in the hypothalamic GnRH which are considred to regulate GTH release. Since the axonal transport of GnRH from the TN system to other brain areas can be easily blocked by OTX in goldfish, OTX fish appear to be good models for studying relationship between gonadal maturation and GnRH of hypothalamic origin. In summary, the present study suggests that most of the sGnRH in the brain originates from the TN system, and that this GnRH is not essential to gonadal maturation in goldfish. The physio- logical functions of the TN system and sGnRH of TN origin remain unknown. The origin of brain cGnRH-II and its functions should also be eluci- dated. Presently, we are further examining the involvement of YTN-sGnRH in vitellogenesis, ovulation, and sex behavior, and using specific antisera, the distribution of sGnRH and cGnRH-II neurons are being immunocytochemically studied. ACKNOWLEDGMENTS We thank Dr. Yoshitaka Oka, Zoological Institute, University of Tokyo, for his critical reading of the manuscript. This study was supported in part by Grants- in-Aid for Scientific Research from the Ministry of Education, Science, and Culture and by a Grant-in-Aid (Bio Media Program 91-II-2-5 from the Ministry of Agriculture, Forestry, and Fisheries of Japan. REFERENCES 1 Fink, G. (1988) Gonadotropin secretion and its control. In “The Physiology of Reproduction” Ed. by E. Knobil and J. Neil, Raven Press, Ltd., New York, pp. 1349-1377. 2 Pfaff, D. W. and Jorgensen, K. and Kow, L.-M. (1987) Luteinizing-hormone releasing hormone in rat brain: Gene expression, role as neuromodulator, and functional effects. Ann. N. Y. Acad. Sci., 519: 323-410. 3 11 12 M. Kosayasut, M. AMANO et al. Sherwood, N. M. and Lovejoy, D. A. (1989) The origin of the mammalian form of GnRH in primitive fishes. Fish Physiol. Biochem., 7: 85-93. Kah, O., Chambolle, P., Dubourg, P. and Dubois, M. P. (1984) Immunocytochemical localization of luteinizing hormonereleasing hormone in the brain of the goldfish Carassius auratus. Gen. Comp. En- docrinol., 53: 107-115. Nozaki, M., Fujita, I., Saito, N., Tsukahara, T., Kobayashi, H., Ueda, K. and Oshima, K. (1985) Distribution of LHRH-like immunoreactivity in the brain of the Japanese eel (Anguilla japonica) with special reference to the nervus terminalis. Zool. Sci., 2: 537-547. Goos, H. J. Th., de Leeuw, R., de Zoeten-Kamp, C., Peute, J. and Blahser, S. (1985) Gonadotropin- releasing hormone-immunoreactive neuronal struc- tures in the brain and pituitary of the African catfish, Clarias gariepinus (Burchell). Cell Tissue Res., 241: 593-596. Kah, O., Breton, B., Dulka, J. G., Nunez- Rodriguez, J., Peter, R. E., Corrigan, A., Rivier, J. E. Vale, W. W. (1986) A reinvestigation of the Gn-RH (gonadotropin-releasing hormone) systems in the goldfish brain using antibodies to salmon Gn-RH. Cell Tissue Res., 244: 327-337. Schreibman, M. P. and Margolis-Nunno, H. (1987) Reproductive biology of the terminal nerve (nucleus olfactoretinalis) and the other LHRH pathways in teleost fishes. Ann. N.Y. Acad. Sci., 519: 60-68. Subhedar, N. and Rama Krishna, N. S. (1988) Immunocytochemical localization of LH-RH in the brain and pituitary of the catfish, Clarias batrachus (Linn.). Gen. Comp. Endocrinol., 72: 431-442. Schafer, H., Schulz, R. and Bliim, V. (1989) Im- munoreactivity to gonadotropin-releasing hormone and gonadotropic hormone in the brain and pituitary of the rainbow trout Salmo gairdneri. Cell Tissue Res. 257: 227-235. Oka, Y. and Ichikawa, M. (1990) Gonadotropin- releasing hormone (GnRH) immunoreactive system in the brain of the dwarf gourami (Colisa lalia) as revealed by light microscopic immunocytochemistry using a monoclonal antibody to common ammino acid sequence of GnRH. J. Comp. Neurol., 300: $11-522. Amano, M., Oka, Y., Aida, K., Okumoto, N., Kawashima, S. and Hasegawa, Y. (1991) Im- munocytochemical demonstration of salmon GnRH and chicken GnRH-II in the brain of masu salmon, Oncorhynchus masou. J. Comp. Neurol., 314: 587- 597. Silverman, A.-J. (1988) The gonadotropin-releasing hormone (GnRH) neuronal systems: Immunocy- tochemistry. In “The Physiology of Reproduction” Ed. by E. Knobil and J. Neil, aven Press, Ltd., New 15 16 17 18 19 20 21 22 23 24 Brain GnRH in OTX Goldfish York, pp. 1283-1304. Stell, W. K., Walker, S. E. and Ball, A. K. (1987) Functional-anatomical studies on the terminal nerve projection to the retina of bony fishes. Ann. N.Y. Acad. Sci., 519: 80-96. Demski, L. S. and Northcutt, R. G. (1983) The terminal nerve: a new chemosensory system in the vertebrates. Science, 202: 435-437. Fujita, I., Sorensen, P. W., Stacey, N. E. and Hara, T. J. (1991) The olfactory system, not the terminal nerve, functions as the primary chemosensory path- ways mediating responses to sex pheromones in male goldfish. Brain Behav. Evol., 38: 313-321. Gentile, F., Lira, O. and Marcano-de Cotte, D. (1986) Releationship between brain gonadotropin- releasing hormone (GnRH) and seasonal reproduc- tive cycle of the “caribe colorado” Pygocentrus notatus. Gen. Gomp. Endocrinol., 64: 239-245. Amano, M., Aida, K., Okumoto, N. and Hasega- wa, Y. (1992) Changes in salmon GnRH and chicken GnRH-II contents in the brain and pituit- ary, and GTH content in the pituitary in female masu salmon, Oncorhynchus masou, from hatching through ovulation. Zool. Sci., 9: 375-386. Dufour, S., Pasqualini, C., Kerdelhue, B. and Fon- taine, Y. A. (1982) Presence and distribution of radioimmunoassayable LHRH in the Europian eel, Anguilla anguilla. Neuropeptides, 3: 159-171. Yu, K. L., Nahorniak, C. S., Peter, R. E., Corri- gan, A., Rivier, J. E. and Vale, W. W. (1987) Brain distribution of radioimmunoassayable gonadotro- pin-releasing hormone in female goldfish: Seasonal variation and periovulatory changes. Gen. Comp. Endocrinol., 67: 234-246. Okuzawa, K., Amano, M., Kobayashi, M., Aida, K., Hanyu, I., Hasegawa, Y. and Miyamoto, K. (1990) Differences in salmon GnRH and chicken GnRH-II contents in discrete brain areas of male and female rainbow trout according to age and stage of maturity. Gen. Comp. Endocrinol., 80: 116-126. Breton, B., Motin, A., Billard, R., Kah, O., Geof- fre, S. and Precigoux, G. (1986) Immunoreactive gonadtropin-releasing hormone-like material in the brain and the pituitary gland during the periovula- tory period in the brown trout (Salmo trutta L.): Relationships with the plasma and pituitary gona- dotropin. Gen. Comp. Endocrinol., 61: 109-119. Yu, K. L., Sherwood, N. M. and Peter, R. E. (1988) Differencial distribution of two molecular forms of gonadotropin releasing hormone in discrete brain areas of goldfish (Carassis auratus). Peptides, 9: 625-630. Stacey, N. E. and Kyle, A. N. (1983) Effects of olfactory tract lesions on sexual and feeding be- havior in the goldfish. Physiol. Behavior, 30: 621- 628. 25 26 27 28 29 30 31 32 33 34 35 36 37 773 Kobayashi, M., Aida, K. and Hanyu, I. (1986) Pheromone from ovulatory female goldfish induces gonadotropin surge in males. Gen. Comp. Endocri- nol., 63: 451-455. Yamamoto, K. and Yamazaki, F. (1961) Rhythm of development in the oocyte of the goldfish, Carassius auratus. Bull. Fac. Fish. Hokkaido Univ., 12: 93- 110. Kobayashi, M., Aida, K. and Hanyu, I. (1985) Radioimmunoassay for silver carp gonadotropin. Bull. Japan. Soc. Sci. Fish., 51: 1085-1091. Kobayashi, M., Aida, K., Hanyu, I. and Ishi, S. (1985) Application of redioreceptor assay to the purification of silver carp gonadotropin. Bull. Japan. Soc. Sci. Fish., 51: 405-411. Kaneko, T., Kobayashi, M., Aida, K. and Hanyu, I. (1985) Ultrastructural immunocytochemistry of gonadotrophs in the goldfish pituitary gland. Cell Tissue Res., 239: 337-342. Kawauchi, H., Suzuki, K., Itoh, H., Swanson, P., Naito, N., Nagahama, Y., Nozaki, M., Nakai, Y. and Itoh, S. (1989) The duality of teleost gonadotro- pins. Fish Physiol. Biochem., 7: 29-38. Kobayashi, M., Aida, K. and Hanyu, I. (1987) Hormone changes during ovulation and effects of steroid hormones on plasma gonadotropin levels and ovulation in goldfish. Gen. Comp. Endocrinol., 67: 24-32. Yamamoto, N., Oka, Y., Amano, M., Aida, K., Hasegawa, Y. and Kawashima, S. (1991) Investiga- tion of brain GnRH systems in the dwarf gourami: immunohistochemistry and radioimmunoassay. Proc. Japan Soc. Comp. Endocrinol., 6: 46. Oka, Y. (1991) Intracellular recording and staining of terminal nerve-GnRH cells in a fish brain in vitro, Soc. Neurosci. Abstr., 17: 1335. Yamazaki, F. (1965) Endocrinological studies on the reproduction of the female goldfish, Carassius auratus 1L., with special reference to the function of the pituitary gland. Mem. Fac. Fish. Hokkaido Univ., 13: 1-64. Dulka, J. D., Stacey, N. E., Sorensen, P. W. and Van Der Kraak, G. J. (1987) A steroid sex pher- omone synchronizes male-female spawning readi- ness in goldfish. Nature, 325: 251-253. Peter, R. E., Habibi, H. R., Chang, J. P., Nahor- niak, C. S., Yu, K. L., Huang, Y. P. and Marchant, T. A. (1989) Actions of gonadotropin-releasing hormone (GnRH) in the goldfish. Progress in clinic- al and biological research, 342: 393-398. Suzuki, M., Hyodo, S., Kobayashi, M., Aida, K. and Urano, A. (1992) Chaaracterization and local- ization of mRNA encoding the salmon-type go- nadotrophin-releasing hormone precursor of the masu salmon. J. Mol. Endocrinol., in press. erred oli it inary Of cthnerpartes “4 ! S——<e Ps ke uc Wola oun, ‘es sbi yep jd. ® now? * (FWD cial Aidt Menem ts 1 ate i. asta dry eke tenretet Stee «tna inant | PR HONE mg lini wi ti ce Di tei. A, COP apiA ie ees. Twi 1 Ve I ere sath PCy Bee ml wei" (oat Peet’ carnyatity itis cant nee i ne : A as ae at Barna Sat « M. idcauded-OFe sachs Se paergnene ae _ incites auiedog RRO rT ie) 1 ieee. Cl WGarereeLs Vin y +) 7 a — Ot aueteol oa page cas Aik. a a7 ee 2 Thar ht: alae Ae oy) j bt yal” pe i" oe srhig! rita tot beh - ee ais Te vw reer te, | Kebondndared i Z Tide) sin ar tortie ‘ ngu teal Reyne Vf its otumliae 1 ee = ne’, ois Que wvivileedace ae yee : we lontsh Vinee, ae by ofl _aasie HMO 270) Mie TS y Me ndioae » Gn RS * Ht Bes potter, naar ay! mt fy eon ity siae Both Sane, 7 7 psy dale Heu macy escort r PLAN : 1 > Ti hah heteieat ENT epee Ay te SHELA aye hs Boe ; OD VA Athen oie aii ral phage sheen) ara tired ‘eteranle ted) 4 ’ A"s } Opry belt «1%, as vita. Be Y. ave = , iis 4ese }Aeeaere line a ; ‘awnin ® i o=lby re yf gern Hen) wr e i dy eae | qa , ik th qian . cowry! «A! >a). ei + ' i ' ) he Oe Gap ett rien? ly. haceulge, ary io swted? Heath jonny Si “tem ft read ft Am 7 ] Im Jeareetee aa rh 07a baeols Moline roller i> ieibuds + 4 il 4 a) tts tee pea ae , “ be (ber em eunteel o overeat Herd do lhinae ZOOLOGICAL SCIENCE 9: 775-784 (1992) Antagonism of Prolactin and Growth Hormone: Impact on Seawater Adaptation in Two Salmonids, Salmo trutta and Oncorhynchus mykiss STEFFEN S. MADSEN!” and Howarp A. BERN* ‘Institute of Biology, Odense University, Campusvej 55, DK-5230 Odense M, Denmark; *Department of Integrative Biology, Cancer Research Laboratory and Bodega Marine Laboratory, University of California, Berkeley, California 94720, U.S.A. ABSTRACT—The effect of simultaneously-injected prolactin and growth hormone on short-term adaptation to seawater was investigated in two salmonid teleosts, Salmo trutta and Oncorhynchus mykiss. Freshwater-adapted fish were given 5—6 injections of prolactin, growth hormone or combina- tions of the two hormones on alternate days and subsequently challenged by transfer to seawater. Major plasma ions, osmolarity, muscle water content and gill Nat, K*-ATPase activity were measured. In both species, increased plasma ions and decreased muscle water were observed 2-3 days after transfer to seawater. Growth hormone significantly reduced post-transfer changes in plasma ions and muscle water, whereas prolactin injections alone had no effect on either freshwater-adapted or seawater- challenged fish. However, when prolactin was injected simultanously with growth hormone, prolactin significantly abolished the seawater-adaptive effect of growth hormone in a dose-related manner. Prolactin also significantly reduced the growth hormone-induced stimulation of gill Na*, K ‘-ATPase activity. We conclude that prolactin is antagonistic to growth hormone during seawater adaptation and © 1992 Zoological Society of Japan development of hypoosmoregulatory mechanisms in salmonids. INTRODUCTION Despite a close similarity in the primary struc- tures of prolactin (PRL) and growth hormone (GH), there is much evidence to suggest an antagonistic relationship during development of hypoosmoregulatory mechanisms in salmonids. Plasma PRL and GH levels are normally inversely related during both smoltification [1, 2] and sea- water (SW) adaptation [1, 3-7], with PRL levels decreasing and GH levels increasing in both events. PRL is essential for freshwater (FW) survival and osmoregulation in several euryhaline teleosts [8-10]. The importance of PRL in FW adaptation of salmonids is, however, less clear, as hypophysectomized fish survive well in FW with little or no ionic disequilibrium [11-13]. PRL causes hypernatremia when injected into FW- and Accepted April 19, 1992 Received March 7, 1992 SW-adapted teleosts including salmonids (see [10], [14], [15]). Furthermore, PRL decreases Cl~ excretion by opercular membrane chloride cells [16] and reduces chloride cell size in seawater- adapted tilapia, Oreochromis mossambicus [17]. The importance of GH during the early phase of SW adaptation and in smoltification is well estab- lished for several salmonids including sea trout, Salmo trutta, and rainbow trout, Oncorhynchus mykiss [18-22]. GH has a hypoosmoregulatory effect, stimulating gill chloride cell proliferation and Na‘, K*-ATPase activity in vivo [20-22], but not in vitro [23]. Osmoregulatory effects of GH in nonsalmonid fishes have not yet been reported. As the above evidence suggests that GH and PRL may act antagonistically during SW adapta- tion, the present study aimed to examine their interaction following simultaneous injection into intact fish. The impact of hormone manipulation was studied for a period before and during the early phase of SW adaptation. Major plasma ion levels, muscle water content and gill Na‘, K*- 7716 S. S. MADSEN AND H. A. BERN ATPase activity were measured. MATERIALS AND METHODS Food was withheld from all fish used in the following experiments beginning two days before and lasting throughout each experiment. Fish were briefly anesthetized in 0.01% phenoxyetha- nol prior to tagging, injection and sampling. Experiment 1 Fish and protocol Immature sea trout (Salmo trutta, age 1+, mixed sexes, mean weight 66.1+ 1.9g, N=160) were obtained in October 1989 from a commercial trout farm in Egtved, Jutland, Denmark, where they had been reared under natural photoperiod and water temperatures. The fish were brought to the Boegebjerggaard Marine Laboratory (Odense University), separated into four groups by fin clipping, and acclimated to the experimental conditions in fresh water (FW; 10°C, 12:12 L:D artificial photoperiod). After two weeks of acclimation (=day 0), fish in the four groups were injected intraperitoneally according to body weight with saline, ovine growth hormone (NIADDK-oGH-15: 2 g/g), ovine pro- lactin (NIADDK-oPRL-19: 2 ug/g) or combined oGH-+o0PRL (2 ug/g each). Hormones were dis- solved in saline (0.9% NaCl, 0.5% bovine serum albumin, pH 9.0) prior to injection, and the injec- tion volume was 2 yl/g. Hormone doses were based on those used in previous studies [17, 21, 22]. During the following 14 days, all groups were given five additional injections with the above dosages (on days 2, 4, 7, 9 and 11). Three days after the last injection (day 14), subgroups from each group were directly transferred to 25 ppt natural seawater (SW; 10°C), where they were held for 3 days or until sampling. Sampling and analyses Subgroups of eight fish were sampled from each group 3 days after the last injection in FW, and 2 and 3 days after transfer to SW. Eight additional fish were sampled from the control group on day 0 before the first injection. Blood was drawn from the caudal vessels using heparinized syringes and centrifuged at 5000Xg; the plasma was frozen (—20°C) until analysed. After bleeding, the fish were decapitated, and a piece of paraxial muscle was dissected and weighed for determination of muscle water content. Plasma sodium and potassium were determined by flame photometry (Instrumentation Laboratory 243), and plasma magnesium and total calcium analysed by atomic absorption spectrophotometry (Perkin Elmer 2380, Mountain View, CA). Plasma chloride was determined by coulometric titration (Radiometer CMT 10, Copenhagen). Muscle water content was determined as weight loss after drying at 105°C until constant weight was attained (48 hours). Experiment 2 Fish and protocol Steelhead trout (Oncorhyn- chus mykiss, age 0+, mixed sexes, mean weight 71.5+1.4 g, N=130) used in this experiment were obtained from the California Department of Fish and Game Warm Springs Hatchery in November 1991. They were transported to the Bodega Marine Laboratory and acclimated in outdoor raceways for 2 weeks in freshwater prior to experi- mentation (12-13°C). During that period the fish were randomly separated into 6 groups and tagged using visible implants in the clear postocular tissue. On days 0, 2, 4, 6 and 8, fish in the six groups were injected intraperitoneally with one of the following hormone doses/combinations: Con: saline; GH: 2 ug/g oGH; PRL: 2 ug/g oPRL; GH +PRLI1: 2 ug/g oGH+0.2 u/g oPRL; GH+ PRL2: 2 ug/g oGH+1 ug/g oPRL; GH+PRL3: 2 ug/g OGH+2 ug/g oPRL. Saline, hormones and injection volumes were the same as described in experiment 1. One day after the last injection (day 9) sub- groups were transferred directly to 32 ppt natural SW (12°C) and held for two days. Sampling and analyses Eight to twelve fish were sampled from the treatment groups on day 9 in FW and again on day 2 after transfer to SW. The sampling procedure was the same as in experi- ment 1, except that gill filaments were also sam- pled. Filaments were placed in sucrose-EDTA- imidazole buffer (SEI; 300mM sucrose, 20 mM EDTA, 50 mM imidazole, pH 7.3), frozen on dry PRL-GH Antagonism in Salmonid Osmoregulation 7717 ice and stored at —80°C until analyzed for Na‘, K*-ATPase activity. Plasma osmolarity was analyzed on a micro vapor pressure osmometer (Wescor, 5100C, Logan, UT). Muscle samples were used to deter- mine muscle water content. Gill filaments were analyzed for Nat, K*-ATPase activity using a method slightly modified from McCormick and Bern [24]. Gill tissue was thawed and homog- enized in SEI buffer with 0.1% sodium deoxycho- late and 10 mM mercaptoethanol using a handheld glass homogenizer. The homogenate was centri- fuged at 5000 g for 30 sec, and the supernatant was kept on ice until analysed for Na‘, K*- ATPase (within 30-60 minutes). Ten microliters of gill homogenate were added to 990 yl assay mixture (50 mM imidazole, 1 U/ml L-lactic dehydrogenase, 2.5 U/ml pyruvate kina- se, 2mM _ phosphoenolpyruvate, 0.05mM {- NADH, 0.5mM ATP, 0.4mM KCN, 45 mM NaCl, 2.5 mM MgCh, 10 mM KCl, pH 7.5). After mixing, the decrease in absorbance at 340 nm was recorded over 6 min at 25°C in a Beckman spec- trophotometer (DU 70; Fullerton, CA). A second cuvette containing 0.5 mM ouabain was run simul- taneously, and the difference in rate of NADH oxidation between assays with or without ouabain was used to calculate specific Na*, K*-ATPase activity. Protein content of the gill homogenate was analyzed according to Lowry et al. [25], and enzyme activity was expressed as micromoles of ADP hydrolyzed per mg of protein per hour. Statistics Statistical differences among groups were analyzed using CRISP software (CRUNCH interactive statistical package, San Fransisco, CA). One-way ANOVA was performed and followed by Student-Newman-Keuls multicomparison _ test. Significant differences are reported if the confi- dence level was >95%. RESULTS Experiment 1 The results obtained in Experiment 1 are summarized in Figure 1. Within each treatment group, changes in each of the major plasma ions followed a similar pattern after transfer to SW. Control, PRL-treated and GH+PRL-treated fish all showed a major increase in plasma ions on day 2 and a stabilization on day 3. The only significant difference among these groups was in plasma Mg** on day 3 in SW (Fig. 1D), when the GH+ PRL-treated group showed a lower value than the control group. In contrast, the GH-treated fish showed less deviation from FW values than did all other treatment groups transferred to SW. In this group, the posttransfer increase was significantly less than that seen in controls for plasma Na‘, Cl, Mg’? and total Ca and less than that in the GH+ PRL-treated for plasma Na* and Cl-. Further- more, plasma ion levels had returned to FW values on day 3 after transfer. Changes in muscle water (Fig. 1F) were inverse- ly related to changes in plasma ions, with dehydra- tion taking place on day 2 in SW. GH- and GH+ PRL-treated fish were less dehydrated than con- trols, with GH-+PRL-treated fish losing more muscle water than GH-injected fish. Experiment 2 Plasma osmolarity increased in all groups after SW transfer (Fig. 2A). However, the magnitude of the increase varied among the groups and showed the following order: Con=PRL>GH+ PRL3 >GH+PRL2>GH+PRL1>GH. Changes in muscle water (Fig. 2B) were inversely related to changes in plasma osmolarity after SW transfer and the order of the degree of dehydration was PRL >Con >GH+ PRL3 >GH+ PRL2>GH +PRL1I=GH. There was a significant negative correlation between plasma osmolarity and muscle water in SW-challenged fish (r7=0.9361, P< 0.01). The effect of hormone treatment on gill Na‘, K*-ATPase activity is shown in Figure 3. PRL significantly decreased and GH significantly in- creased gill ATPase activity. None of the com- bined treatment groups showed enzyme activities significantly different from the control group; nevertheless, the higher doses of PRL inhibited gill Na‘, K*-ATPase activity more effectively than the lower dose. Regression analyses showed that there was a significant correlation between pretransfer gill 778 Plasma Nat (mmol/|) Plasma Cl” (mmol/|) Plasma Kt (mmol/|) 200 180 160 140 120 100 200 180 160 140 120 100 6.0 5.0 4.0 3.0 2.0 SW transfer i | b SW transfer C oD » AC SW transfer Plasma Mg** (mmol/!) Plasma Ca (mmol/l) Muscle water (%) 5.0 4.0 3.0 2.0 1.0 3.0 2.0 1.0 S. S. MADSEN AND H. A. BERN SW transfer \ 44 SW transfer a — SW transfer PRL-GH Antagonism in Salmonid Osmoregulation 779 400 [_] control IN) 2 wg GH + 02 wg PRL/g 2ug GH/g MM 2 ug GH + 1 ug PRL/g fX] 2 ug PRL/g FA 2 ug GH + 2 ug PRL/g 370 340 310 Osmolarity (mOsm/|) 280 250 Ix<x<h 80 ) (-) w ~ 79 Cc ~ O° o 78 ® oO 77 = ® 2 76 =>} = 75 noe fa OS Xi] FW groups after 5 injections 48 hr in SW Fic. 2. Effect of various hormone treatments on plasma osmolarity (A) and muscle water content (B) in O. mykiss in FW and 48 hr after transfer to SW. Fish were given 5 injections of each dose and combination of PRL and GH in FW and transferred to 32 ppt SW one day after the last injection. Values are mean+SEM of 8-12 fish. Values with shared symbols are not significantly different (P<0.05). Fic. 1. Changes in plasma sodium (A), chloride (B), potassium (C), magnesium (D), and total calcium (E) and in muscle water content (F) in S. trutta given 6 injections of saline (O), 2 ug/g oRPL (©), 2 ug/g oGH (A) or 2 pg/¢g oPRL +2 ug/g oGH (@) and subsequently transferred to 25 ppt SW two days after the last injection. Values are mean+SEM of 8 fish. a, Significantly different (P<0.05) from the control group; b, significantly different from the GH+PRL-treated group. 780 = L] control oO 2 ug GH/g E 8 E] 2 ug PRL/g O = at 6 ® a ao 4 - < 7 Mew D < = J = 0 Fic. Osmolarity (mOsm/l) Muscle water content (%) S. S. MADSEN AND H. A. BERN a ist 2 ug GH + 0.2 ug PRL/g 2 ug GH + 1 wg PRL/g 2 ug GH + 2 ug PRL/g FW groups dea 5 injections 3. Effect of various doses and combinations of PRL and GH on gill Na* , K*-ATPase activity in O. mykiss. Fish were given 5 injections in FW and sampled 1 day after the last injection. Values are mean +SEM of 8-10 fish. Values with shared symbols are not significantly different (P>0.05). 400 375 350 Zo 325 79.0 78.0 77.0 ag 76.0 Geri Ctr tae ia TURE Oe Le eel ep anoee aeal (0) 2.0 4.0 6.0 8.0 Na*,K*-ATPase (umol/mg/hr) Na‘, K*-ATPase activity and both posttransfer plasma osmolarity (Fig. 4A: r>=0.7637, P<0.05) and muscle water content (Fig. 4B: r>=0.6652, P <0.05). DISCUSSION GH treatment was used to induce osmoregula- tory changes comparable to those expected during smolting. Both experiments confirmed the SW- adaptive effect of GH reported earlier for S. trutta and O. mykiss [19, 21, 22] as well as for other salmonids [18, 26]. The mechanism responsible has been shown to involve stimulation of branchial chloride cell development and Na‘, K*-ATPase activity in vivo [20-22], but not in vitro [23]. However, more studies are needed to clarify the exact mechanisms. The ion-retaining effect of PRL, well established Fic. 4. Regression analyses of pretransfer gill Na“, K*-ATPase activity and plasma osmolarity (A) and muscle water content (B) in O. mykiss 48 hr after SW transfer. Analyses are based on mean values (+ SEM, N=8-12) from hormone-manipulated fish in Experiment 2 (see Fig. 2). A: r°-=—0.7637, P< 0.05; B: r= —0.6652, P<0.05. PRL-GH Antagonism in Salmonid Osmoregulation 781 and developed into a bioassay for PRL bioactivity in several euryhaline teleosts [27-31], remains controversial in salmonids [31]. Generally, PRL seems to play a minor role in FW osmoregulation in salmonids compared with that seen in other teleosts. For instance, PRL treatment in the present study did not have any effect on ionic balance in FW-acclimated fish, and hypophysecto- mized (Hx) O. mykiss and O. kisutch survive well in FW with only minor ion losses [11-13]. Also, salmonid pituitary PRL secretion in vitro is not altered by changes in extracellular osmolarity in the physiological range as occurs in other species [32, 33]. One study by Oduleye [34], however, indicated that brown trout depends on pituitary factors for FW survival. In contrast to the clear effect of GH, pretreat- ment with PRL had no effect on iono-osmotic regulation during the subsequent early phase of SW adaptation in either species used. By compari- son, only a minimal hypernatremic effect after 3 injections of 2.5 ug/g chum salmon PRL (sPRL) into O. mykiss was seen by Bolton et al. ({[19]—see also Hirano [14]). Any difference in the results reported could arise from differences in sampling time, the use of mammalian vs. teleost PRL [29- 31] and size of fish used. In most salmonids, a hyperosmoregulatory (i.e., anti-SW-adaptive) effect becomes evident when PRL is injected into either SW-adapted fish or into fish simultaneously injected with GH. Hasegawa et al. [31] reported that both sPRL (0.5 ug/g) and oPRL (10 ug/g) induced hypernatremia in O. mykiss adapted to 50% SW. Similarly, S. Vester- vang and S. S. Madsen (unpublished) found that oPRL (2 ug/g) induced hypernatremia and re- duced muscle water content when injected into O. mykiss adapted to 28 ppt SW but not when fish were adapted to 20 ppt SW. Hasegawa et al. [31], however, did not find any effect of sPRL or oPRL on ion levels in SW-adapted O. keta fry, suggesting that either interspecific differences in sensitivity may exist or that the effect was compensated for by high levels of SW-adaptive hormones (such as GH). In our studies, PRL-GH antagonism was evident when the two hormones were coinjected. PRL then inhibited the SW-adaptive effect of GH in a dose-related manner, with the effect of GH almost abolished when the two hormones were injected at equal doses. Increasing doses of GH may antagonize PRL in a similar manner. How- ever, in a converse experiment conducted at a different time of year (March instead of Novem- ber), evidence for the ability of PRL to antagonize the effect of GH in O. mykiss, although occa- sionally indicated, could only be considered equivocal over all (S. S. Madsen, unpublished). The endocrine effects on hypoosmoregulatory performance are well correlated with the effects on gill Nat, K*-ATPase activity (Fig. 5A, B), sug- gesting a causal relationship. GH stimulated gill Nat, K*-ATPase in Exp. 2, whereas PRL has a small but significant inhibitory effect when injected alone, but in particular when coinjected with GH. This may be viewed as a “FW-adaptive” effect of PRL, since lower enzyme levels are normally associated with FW adaptation in salmonids and most other teleosts. The effect is in accordance with a similar inhibitory effect of PRL on gill Na‘, K*-ATPase in SW-adapted Chelon labrosus [35], Hx FW-transferred Anguilla japonica [36], Hx FW-adapted Fundulus heteroclitus [37] and SW- adapted Morone saxatilis (S. S. Madsen, in prep). Interestingly, in SW-adapted O. mossambicus, PRL had no effect on gill Na*, K*-ATPase [17, 38], even though it reduced average chloride cell size [17] and decreased opercular membrane Cl~ excretion rate [16]. Additional osmoregulatory effects of PRL that may antagonize SW adaptation include reduced osmotic permeability of the gill epithelium [39], reduced intestinal water and ion transport rates [40], and increased glomerular filtration rate in vitro in O. mykiss (B. Dunne and C. Rankin, personal comm.). There are as yet no reports on the effects of GH on these targets. During “normal” development of hypo- osmoregulatory mechanisms (e.g., in smolting and/or SW adaptation), PRL levels are decreasing while GH levels are increasing [1—7, 41], suggest- ing that endogenous release and turnover of PRL and GH may be inversely controlled during these processes. Therefore, our results imply that any development-associated increase in PRL levels during a hypoosmoregulatory phase of the salmo- nid life cycle may be critical and incompatible with hypoosmoregulation due to PRL’s antagonism of 782 S. S. MADSEN AND H. A. BERN GH. Increased PRL levels reported during sexual maturation in SW [14, 42-44] could ultimately cause osmoregulatory maladaptation to SW. Stress-associated elevations in PRL levels during smolting [45] may be deleterious to smolt quality and subsequent performance in SW. Finally, the effect of experimentally administered hormones may potentially be antagonized by endogenous hormones. The mechanism for the observed antagonism remains unclear from our in vivo experiments. Competition between PRL and GH at the receptor level is unlikely to be a major factor, since dis- placement studies in various salmonid tissues show that the GH receptor has only 1-10% of the affinity for PRL that it has for GH [46-48]. Furthermore, each hormone has its own specific receptor in several osmoregulatory organs in sal- monids [36-49]. Although hypothetical at this moment, it is possible that other endocrine path- ways may be involved in the observed antagonism. Insulin-like growth factor-I and “synlactin” are possible hepatic mediators or synergists for GH and PRL, according to the somatomedin and “syn- lactin” hypotheses, respectively [50, 51]. Recent research in our laboratory has indicated that IGF-I may stimulate gill Na*, K*-ATPase in O. kisutch in vitro, thereby supporting the suggestion of McCormick ef al. [52] that IGF-I is the mediator for the osmoregulatory action of GH in salmonids. In conclusion, our study shows that the SW- adaptive effect of exogenous GH may be abolished by increasing PRL levels, thus adding another facet to the multihormonal control of osmoregula- tion in salmonids. ACKNOWLEDGMENTS S.S.M. was a postdoctoral fellow of the Carlsberg Foundation (Denmark) for part of this study. We are grateful to Drs. S. D. McCormick and R. S. Nishioka for their reviews of the manuscript and to H. Ahm (Odense), A. Corrales and J. S. Endersen (Berkeley) for their help with the analytical work. NIH and the National Pituitary Program (Baltimore, U.S.A.) pro- vided the ovine prolactin and growth hormone used in this study, and Zenyaku Kogyo Co. of Tokyo provided research support. This work is a result of research sponsored in part by NOOA, National Sea Grant Col- lege Program, Department of Commerce, under grant number NA89AA-D-SG138, project number R/F-117, through the California Sea Grant College, and in part by the California State Resources Agency. The U.S. Gov- ernment is authorized to reproduce and distribute for governmental purposes. REFERENCES 1 Prunet, P., Boeuf, G., Bolton, J. P. and Young, G. (1989) Smoltification and seawater adaptation in atlantic salmon (Salmo salar): Plasma prolactin, growth hormone, and thyroid hormone. Gen. Comp. Endocrinol., 74: 355-364. 2 Young, G. Bjérnsson, B. T., Prunet, P., Lin, R. J. and Bern, H. A. (1989) Smoltification and seawater adaptation in coho salmon (Oncorhynchus kisutch): Plasma prolactin, growth hormone, thyroid hor- mones, and cortisol. Gen. Comp. Endocrinol., 74: 346-354. 3 Prunet, P., Boeuf, G. and Houdebine, L. M. (1985) Plasma and pituitary prolactin levels in rainbow trout during adaptation to different salinities. J. Exp. Zool., 235: 187-196. 4 Sakamoto, T., Ogasawara, T. and Hirano, T. (1990) Growth hormone kinetics during adaptation to a hyperosmotic environment in rainbow trout. J. Comp. Physiol., 160B: 1-6. 5 Yada, T., Takahashi, K. and Hirano, T. (1991) Seasonal changes in seawater adaptability and plas- ma levels of prolactin and growth hormone in land- locked sockeye salmon (Oncorhynchus nerka) and amago salmon (O. rhodurus). Gen. Comp. Endocri- nol., 82: 33-44. 6 Yada, T. and Hirano, T. (1992) Influence of seawa- ter adaptation on prolactin and growth hormone release from organ-cultured pituitary of rainbow trout. Zool. Sci., 9: 143-148. 7 Sakamoto, T., Iwata, M. and Hirano, T. (1991) Kinetic studies of growth hormone and prolactin during adaptation of coho salmon, Oncorhynchus kisutch, to different salinities. Gen. Comp. Endocri- nol., 82: 184-191. 8 Pickford, G. E. and Phillips, J. G. (1959) Prolactin, a factor in promoting survival of hypophysectomized killifish in fresh water. Science 130: 454-455. 9 Ball, J. N. and Ensor, D. M. (1965) Specific action of prolactin on plasma sodium levels _ in hypophysectomized Poecilia latipinna (Teleostei). Gen. Comp. Endocrinol., 8: 432-440. 10 Clarke, W. C. and Bern, H. A. (1980) Comparative endocrinology of prolactin. In “Hormonal Proteins and Peptides”. Ed. by C. H. Li, Academic Press, New York, Vol. 8, pp. 105-197. 11 Donaldson, E. M. and McBride, J. R. (1967) The effects of hypophysectomy in the rainbow trout, Salmo gairdnerii (Rich.) with special reference to 12 13 14 15 16 17 18 19 20 21 22 23 PRL-GH Antagonism in Salmonid Osmoregulation the pituitary-interrenal axis. Gen. Comp. Endocri- nol., 9: 93-101. Nishioka, R. S., Richman III, N. H., Young, G., Prunet, P. and Bern, H. A. (1987) Hypophysec- tomy of coho salmon (Oncorhynchus kisutch) and survival in fresh water and seawater. Aquaculture, 65: 343-352. Yamauchi, K., Nishioka, R. S., Young, G., Ogasa- wara, T., Hirano, T. and Bern, H. A. (1991) Osmoregulation and circulating growth hormone and prolactin levels in hypophysectomized coho salmon (Oncorhynchus kisutch) after transfer to fresh water and seawater. Aquaculture, 92: 33-42. Hirano, T. (1986) The spectrum of prolactin action in teleosts. In “Comparative Endocrinology: De- velopment and Directions”. Ed. by C. L. Ralph, Alan R. Liss, New York, pp. 53-74. Hirano, T., Ogasawara, T., Bolton, J. P., Collie, N. L., Hasegawa, S. and Iwata, M. (1987) Osmoregulatory role of prolactin in lower verte- brates. In “Comparative Physiology of Environmen- tal Adaptations”. Ed. by R. Kirsch and B. Lahlou, Karger, Basel, Vol. 1, pp. 112-124. Foskett, J. K., Machen, T. E. and Bern, H. A. (1982) Chloride secretion and conductance of tele- ost opercular membrane: Effects of prolactin. Am. J. Physiol., 242: E380-R389. Herndon, T. M., McCormick, S. D. and Bern, H. A. (1991) Effects of prolactin on chloride cells in opercular membrane of seawater-adapted tilapia. Gen. Comp. Endocrinol., 83: 283-289. Komourdyjian, M. P., Saunders, R. L. and Fenwick, J. C. (1976) The effect of porcine somatotropin on growth, and survival in seawater of Atlantic salmon (Salmo salar) parr. Can. J. Zool., 54: 531-535. Bolton, J. P., Collie, N. L., Kawauchi, H. and Hirano, T. (1987) Osmoregulatory actions of growth hormone in rainbow trout (Salmo gairdneri). J. Endocrinol., 112: 63-68. Richman III, N. H. and Zaugg, W. S. (1987) Effects of cortisol and growth hormone on osmoregulation in pre- and desmoltified coho salmon (Oncorhyn- chus kisutch). Gen. Comp. Endorcinol., 65: 189- 198. Madsen, S. S. (1990) The role of cortisol and growth hormone in seawater adaptation and development of hypoosmoregulatory mechanisms in sea trout parr (Salmo trutta trutta). Gen. Comp. Endrocrinol., 79: 1-11. Madsen, S. S. (1990) Enhanced hypoosmoregula- tory response to growth hormone after cortisol treatment in immature rainbow trout, Salmo gaird- neri. Fish Physiol. Biochem., 8: 271-279. McCormick, S. D., Dickhoff, W. W., Dustin, J., Nishioka, R. S. and Bern, H. A. (1991) Develop- mental differences in the responsiveness of gill Na“ , 24 25 26 27 28 29 30 31 32 33 34 35 36 783 Kt-ATPase to cortisol in salmonids. Gen. Comp. Endocrinol., 84: 308-318. McCormick, S. D. and Bern, H. A. (1989) Jn vitro stimulation of Na*, K*-ATPase activity and oua- bain binding by cortisol in coho salmon gill. Amer. J. Physiol., 256: R707—R715. Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193: 265-275. Boeuf, G., Prunet, P. and Le Bail, P.-Y. (1990) Un traitement a l’hormone de croissance peut-il stimul- er la smoltification du saumon atlantique? C.R. Acad. Sci. Paris, 310: 75-80. Ensor, D. M. and Ball, J. N. (1986) A bioassay for fish prolactin (paralactin). Gen. Comp. Endocrinol., 11: 104-110. Clarke, W. C. (1973) Sodium-retaining bioassay of prolactin in the intact teleost Tilapia mossambica acclimated to seawater. Gen. Comp. Endocrinol., 21: 498-512. Grau, E. G., Prunet, P., Gross, T., Nishioka, P. S. and Bern, H. A. (1984) Bioassay for salmon prolac- tin using hypophysectomized Fundulus heteroclitus. Gen. Comp. Endocrinol., 53: 78-85. Specker, J. L., King, D. S., Nishioka, R. S., Shir- ahata, K., Yamaguchi, K. and Bern, H. A. (1985) Isolation and partial characterization of a pair of prolactins released in vitro by the pituitary of a cichlid fish. Proc. Nat. Acad. Sci. USA, 82: 7490- 7494. Hasegawa, S., Hirano, T. and Kawauchi, H. (1986) Sodium-retaining activity of chum salmon prolactin in some euryhaline teleosts. Gen. Comp. Endocri- nol., 63: 309-317. Gonnet, F., Prunet, P., Tonon, M. C., Dubourg, P., Kah, O. and Vaudry, H. (1988) Effect of osmotic pressure on prolactin release in rainbow trout: in vitro studies. Gen. Comp. Endocrinol., 69: 252-261. Kelley, K. M., Nishioka, R. S. and Bern, H. A. (1990) In vitro effect of osmotic pressure and cortisol on prolactin cell physiology in the coho salmon (Oncorhynchus kisutch) during the parr- smolt transformation. J. Exp. Zool., 254: 72-82. Oduleye, S. O. (1976) The effects of hypophysec- tomy, prolactin therapy and environmental calcium on freshwater survival and salinity tolerance in the brown trout Salmo trutta L. J. Fish Biol., 9: 463- 470. Gallis, J.-L., Lasserre, P. and Belloc, F. (1979) Freshwater adaptation in the euryhaline teleost, Chelon labrosus \. Effects of adaptation, prolactin, cortisol, and actinomycin D on plasma osmotic balance and (Na*, K*)-ATPase in gill and kidney. Gen. Comp. Endocrinol., 38: 1-10. Kamiya, M. (1972) Hormonal effect on Na-K- ATPase activity in the gill of Japanese eel, Anguilla By 38 39 40 41 42 43 44 784 japonica, with special reference to seawater adapta- tion. Endocrinol. Japan., 19: 489-493. Pickford, G. E., Griffith, R. W., Torretti, J., Hend- lez, E. and Epstein, F. H. (1970) Branchial reduc- tion and renal stimulation of (Na*, K*)-ATPase by prolactin in hypophysectomized killifish in fresh water. Nature, 228: 378-379. Young, P. S., McCormick, S. D., Demarest, J. R., Lin, R. J., Nishioka, R. S. and Bern, H. A. (1988) Effects of salinity, hypophysectomy, and prolactin on whole-animal transepithelial potential in the tila- pia Oreochromis mossambicus. Gen. Comp. Endoc- rinol., 71: 389-397. Ogawa, M. (1977) The effect of hypophysectomy and prolactin treatment on the osmotic water influx into the isolated gills of the Japanese eel (Anguilla japonica). Can. J. Zool., 55: 872-876. Utida, S., Hirano, T., Oide, H., Ando, M. Johnson, D. W. and Bern, H. A. (1972) Hormonal control of the intestine and urinary bladder in teleost osmoregulation. Gen. Comp. Endocrinol. Suppl., 3: 317-321. Avella, M., Young, G., Prunet, P. and Schreck, C. B. (1990) Plasma prolactin and cortisol concentra- tions during salinity challenges of coho salmon (Oncorhynchus kisutch) at smolt and post-smolt stages. Aquaculture, 91: 359-372. McKeown, B. A. and Van Overbeeke, A. P. (1972) Prolactin and growth hormone concentrations in the serum and pituitary of adult migratory sockeye salmon (Oncorhynchus nerka). J. Fish. Res. Board Can., 29: 303-309. Nagahama, Y. (1973) Histo-physiological studies on the pituitary glands of some teleost fishes, with special reference to the classification of hormone- producing cells in the adenohypophysis. Mem. Fac. Fish. Hokkaido Univ., 21: 1-63. Andersen, @., Skibeli, V., Haug, E. and Gautvik, K. M. (1991) Serum prolactin and sex steroids in 45 46 47 48 49 50 51 52 S. S. MADSEN AND H. A. BERN Atlantic salmon (Salmo salar) during sexual matura- tion. Aquaculture, 95: 169-179. Avella, M., Schreck, C. B. and Prunet, P. (1991) Plasma prolactin and cortisol concentrations of stressed coho salmon, Oncorhynchus kisutch, in fresh water or salt water. Gen. Comp. Endocrinol., 81: 21-27. Gray, E. S., Young, G. and Bern, H. A. (1990) Radioreceptor assay for growth hormone in coho salmon (Oncorhynchus kisutch) and its application to the study of stunting. J. Exp. Zool., 256: 290- 296. Yao, K., Niu, P. D., Le Gac, F. and LeBail, P.-Y. (1991) Presence of specific growth hormone binding sites in rainbow troup (Oncorhynchus mykiss) tis- sues: Characterization of the hepatic receptor. Gen. Comp. Endocrinol., 81: 72-82. Sakamoto, T. and Hirano, T. (1991) Growth hor- mone receptors in the liver and osmoregulatory organs of rainbow trout: characterization and dyna- mics during adaptation to seawater. J. Endocrinol., 130: 425-433. McKeown, B. A., Fargher, R. C. and van Over- beeke, A. P. (1987) The localization of labelled fish prolactin in various tissues of rainbow trout (Salmo gairdneri). Cell Tiss. Res., 250: 57-62. Daughaday, W. H., Hall, K., Raben, M. S., Sal- mon, W. D. Jr., Van Den Brande, J. L. and Van Wyk, J. J. (1972) Somatomedin: proposed designa- tion for sulphation factor. Nature, 235: 107. Anderson, T. R., Pitts, D. S. and Nicoll, C. S. (1984) Prolactin’s mitogenic action on the pigeon crop-sac mucosal epithelium involves direct and indirect mechanisms. Gen. Comp. Endocrinol., 54: 236-246. McCormick, S. D., Sakamoto, T., Hasegawa, S. and Hirano, T. (1991) Osmoregulatory actions of insulin-like growth factor I in rainbow trout (Oncorhynchus mykiss). J. Endocrinol. , 130: 87-93. ZOOLOGICAL SCIENCE 9: 785-789 (1992) © 1992 Zoological Society of Japan Effects of Ca Concentrations in Culture Medium on the Release of Calcitonin from Incubated Ultimobranchial Glands of the Bullfrog, Rana catesbeiana Hiroyuki Kal-yA, JUNKO OKUYAMA, TAKASHI ISHIJIMA, YUICHI SASAYAMA, Hipext YosHizAwa! and Cuitaru OGcuro Department of Biology, Faculty of Science, Toyama University, Toyama 930, and ‘Department of Oral Anatomy II, Matsumoto Dental College, Shiojiri, Nagano 399-07, Japan ABSTRACT—Calcitonin released from ultimobranchial glands incubated in culture media having different Ca concentrations was determined by rat bioassay. In the first incubation of 30 min, a large quantity of calcitonin was released into various culture media such as normal medium, and high Ca media which have 2 times, 3 times and 6 times Ca concentrations compared to the normal medium. In this term, there was no significant difference in the quantity of calcitonin released into the media among those groups. However, during 48 hr following the first incubation, in the Ca 2 times medium, calcitonin quantity released into the medium was maximum. In the Ca 3 or 6 times medium, the amount of calcitonin released was less than that of Ca 2 times medium. These results imply that in the bullfrog, secretion of calcitonin from ultimobranchial glands in vivo may be induced by a suitable rise in the serum Ca concentration. INTRODUCTION Calcitonin is secreted from thyroid glands in mammals or from ultimobranchial glands in non- mammals. It has been well known that in mam- mals, calcitonin is secreted against the rise of blood Ca levels. In pigs and rabbits, blood calcitonin level is increased when the thyroid gland is per- fused with high Ca solution [1-3]. Furthermore, under in vitro condition, calcitonin secretion from the thyroid gland of pigs is accelerated by moder- ate elevation of Ca concentrations in culture medium [4]. In birds, when the ultimobranchial gland is perfused with high Ca solution, blood calcitonin level is increased [5, 6]. Incubated avian ultimobranchial glands also show the same reac- tions as in mammals [7, 8]. These facts imply that in higher vertebrates, the rise of blood Ca level is one of the factors which accelerate calcitonin secretion. On the other hand, it has been reported that in Accepted May 20, 1992 Received March 16, 1992 bony fishes such as trouts and eels, rise of blood Ca level does not cause calcitonin secretion from ultimobranchial glands [9, 10]. Therefore, it is suggested that mechanisms which trigger calcitonin secretion may be different between higher verte- brates and lower ones. In the present study, effects of Ca concentra- tions in culture medium on calcitonin release from incubated ultimobranchial glands of the bullfrog, Rana catesbeiana, were examined by rat bioassay. MATERIALS AND METHODS Male bullfrogs (body weight, 200-300 g) were purchased from commercial source. One pair of ultimobranchial glands found near the glottical sphincter was dissected out carefully under the binocular microscope, and was immediately put into an incubation chamber (Lab-Tek 4804, Nunc Inc.) which contained 1 ml of amphibian Ringer’s solutions with different Ca concentrations as de- scribed below. The ultimobranchial glands were kept at room temperature for 30 min (the first incubation). Then, the ultimobranchial glands 786 were transferred into a separate chamber which contained the same kind of incubation medium, and incubated for 24 hr at 25°C (the second in- cubation). After that, those were again displaced into a distinct chamber with the same kind of the solution, and incubated furthermore for 24 hr at 25°C (the third incubation). Each of the culture media obtained at each incubation time, which included released calcitonin, was lyophilized and frozenly stocked —50°C until use. Three sets of incubation chambers were pre- pared according to the incubation period, such as: 30 min (the first incubation), 24 hr (the second incubation), and next 24 hr (the third incubation). Each set included four culture media having dif- ferent Ca concentrations as follows: the normal amphibian Ringer’s solution, which served as the standard incubation medium, was composed of NaCl 6.50 g/l, KCl 0.14g/l, CaCl, 0.12 g/], NaHCO; 0.20 g/1l, glucose 1 g/l; the other three were composed of Ringer solutions that contained 2, 3, and 6 times amount of Ca of the standard medium (Ca 2X, Ca 3X, and Ca 6X incubation medium, respectively). Bullfrog calcitonin released from ultimobran- chial glands into incubation medium was detected by rat bioassay [11]. Each of the pooled lyophi- lized samples which were obtained after incubating 3 pairs of ultimobranchial glands separately, was adjusted to 0.4 ml by saline solution (0.9% NaCl) and administered to a rat. As a control, the saline solution was administered. Furthermore, salmon calcitonin (Novabiochem Inc.) (10, 25 and 100 mU) was administered as a calcitonin standard for comparison with samples. Blood was samples just before medium administration and at 0.5, 1, 2, and 3 hr after. Serum Ca concentrations in rats were determined by atomic absorption spectropho- tometry (Hitachi-Zeeman 180-70 type). In the present study, changes in serum Ca levels were exhibited as decline rates from the initial Ca level. Furthermore, the areas which were lower than the initial level were determined to examine the dura- tion of hypocalcemic effect of the culture medium, and were exhibited as serum Ca graphs. Student’s f-test was applied to evaluate the data. Fic. H. Kari-ya, J. Okuyama et al. Oro a fe) N iS) serum Ca 4%, we We \ -30 QO Of 1 2 3 time(hr) O vs 0 = oe Te =| (3) serum Ca 4%, tn (e) il * {2} 9 -20 ue a tC SO TOS A 3 time(hr) Oro —_ Wea = a pe) N Ss) serum Ca 4 ¥, as \ ron S os TES a 2 3 time(hr) 1. Time courses (0-3 hr) of serum Ca concentra- tion fall (delta % from the initial levels) in rats after administrations of the first incubation (30 min) medium (A) and the second incubation (24 hr) medium (B) and the third incubation (24 hr) medium (C). Vertical bar shows mean+SE. Each symbol means normal Ca medium (4), Ca 2 times medium (@), Ca 3 times midium (™), Ca 6 times medium (*) and saline solution (©). The numbers in parentheses mean number of rats used. Signif- icantly different from the value of normal Ca medium: * P<0.05, ** P<0.005, *** P<0.001. Ca Effect on CT Release in vitro in Frog 787 RESULTS When incubation media were administered to rats, serum Ca levels were decreased. Decline patterns by the first, the second and the third incubation media having different Ca concentra- tions are exhibited in Figure 1A, B, C, respective- ly. In Figure 2, decline patterns are shown when salmon calcitonin (10, 25 and 100 mU) was admi- nistered as a calcitonin standard. In Figure 3A, areas (cm’) declined by the admi- nistration of normal Ca Ringer (1 x), Ca 2 times medium (2x), Ca 3 times medium (3X), Ca 6 times medium (6X), salmon calcitonin (sCT: 10, 25 and 100 mU) and saline solution as a control, which were lower than the initial Ca level during 0-3 hr in each of the incubation time (the first culture, the second culture and the third culture), are shown in histograms. Among the areas made by administration of 4 kinds of the first incubation media, there was no significant difference, when 153 150 A 100 7 yh ac (cm?) serum Ca 4% ND oe) =e= Oo | JA us (ft SC OOISNE D) 2 time(hr) Fic. 2. Time courses (0-3 hr) of serum Ca concentra- KEE ! 1 1 1 1 1 i} i] 1 ' 1 200 I I I ! ! ! * 1 1 100 | 1 1 i] if ! 1 I ; a ! 0 tion fall (delta % from the initial levels) in rats after administrations of salmon calcitonin 10 mU (@), 25 mU (4) and 100 mU (™). The numbers in parenthe- ses mean number of rats used. * P<0.05, ** P< 0.01. B 300 (cm?) Saline 10 25 100 1X 2X 3X 6X 1X 2X 3X 6X 1X 2X 3X 6X 1X 2X 3X 6X sCT (ml) Ist culture Qnd culture 3rd culture Total Fic. 3A. Histograms showing areas (cm7), which are lower than the initial Ca level in Figure 1 and 2, declined by administrations of saline solution, salmon calcitonin (sCT: 10, 25 and 100 mU), normal Ca Ringer (1x), Ca 2 times medium (2x), Ca 3 times medium (3X) and Ca 6 times medium (6X) in each incubation term of the first incubation (1st culture), the second incubation (2nd culture) and the third incubation (3rd culture). Significantly different from 1 x area in each incubation time, but the value was calculated after subtraction of saline area: * P <0.05, ** P<0.005, *** P<0.001. Fic. 3B. Histograms showing total areas (cm)), which include all areas during 48.5 hr declined by adiminstrations of normal Ca Ringer (1x), Ca 2 times medium (2X), Ca 3 times medium (3x) and Ca 6 times medium (6x). 788 H. Kat-ya, J. Okuyama et al. they were compared after the subtraction of the area by the saline administration. The calcitonin released in this incubation time was in large quan- tities for its short term. Those amounts were approximately compared to 30-72% of salmon calcitonin 10 mU judging from the area. However, among the second incubation media, the area made by Ca 2X medium was the largest (P< 0.001 to the normal Ringer), which corresponded to 118% of salmon calcitonin 25 mU. The area by Ca 3X medium was also significantly larger than that by saline control (P<0.05). When the third in- cubation media were administered, the area made by Ca 2x medium was also the largest (P<0.005 to the normal Ringer), which was comparable with 87% of salmon calcitonin 25mU. Furthermore, the areas by Ca 3X and Ca 6X media were also larger than that by the saline control (P<0.05). In Figure 3B, the total areas, which include all areas declined during 48.5 hr, are exhibited in histo- grams. The area made by Ca 2 medium was 1.9 times larger than the normal Ringer. The area made by Ca 3X medium was 1.4 times larger than the normal Ringer. The area by Ca 6X was only 1.1 times larger. DISCUSSION It has been reported that in the incubated thy- roid gland of the pig, secretion of calcitonin into the medium was concentrated in the first 15 min period [4]. In the present study, it was known that also in the bullfrog, release of large quantity of calcitonin from the incubated ultimobranchial gland occurred in the first 30 min incubation time. It has been reported that in the leopard frog, Rana pipiens, ultimobranchial glands are innervated, and that the secretion of calcitonin is suppressed by the nervous system [12]. Therefore, it is possible that also in the bullfrog, under in vitro condition, suppressive control of the nervous sys- tem for ultimobranchial secretion was eliminated. We reported previously that in the ultimobran- chialectomized bullfrog tadpoles kept in high Ca water (20mg/100 ml) for 48 hr, a conspicuous hypercalcemia was observed [13]. In that experi- ment, Ca concentration of the environmental wa- ter was about 3 time that of the serum Ca level (the average value of the serum Ca before the treat- ment was 7.4mg/100 ml). Final serum Ca value was 14.3 mg/100 ml which was 1.9 times that of the initial value. However, the serum Ca levels of the sham-operated group did not show any increases. This fact suggests that in bullfrog tadpoles, the ultimobranchial glands can secrete calcitonin effec- tively against at least 2 times rise of the serum Ca concentrations. Furthermore, when 0.5 ml of ex- tremely high Ca water (200 mg/100 ml) was in- fused in the intestine of the ultimobranchialecto- mized bullfrog tadpoles, the serum Ca concentra- tions was raised 1.6 times that of the sham-control group at 24hr later [14]. These observations suggest that in the ultimobranchialectomized bull- frog tadpoles, serum Ca level does not elevate more than twice of the normal level at least for a short term, even if any treatments are done. Therefore, also in the adult bullfrog, serum Ca level may not elevate more than twice of the normal level. In mammals, it has been known that secretion rate of calcitonin is directly and linearly related to the actual increment in plasma Ca level [1-3]. On the other hand, ultimobranchial glands of teleosts do not respond to rises of serum Ca level [9, 10]. In the persent study using adult bullfrog, it was the Ca 2X medium that was most effective on releasing calcitonin from incubated ultimobranchial glands. Therefore, calcitonin secretion from ultimobranchial glands may be re- lated to the rise of serum calcium levels, to some extent in bullfrogs. The Ca 3x medium and Ca 6 x medium were also effective, but less potent than Ca 2x medium. In anuran amphibians, delicate control system of calcitonin secretion as in mam- mals may not have been established. REFERENCES 1 Care, A. D., Cooper, C. W., Duncan, T. and Orimo, H. (1968) A study of thyrocalcitonin secre- tion by direct measurement of in vivo secretion rates in pigs. Endocrinology, 83: 161-169. 2 Lee, M. R., Deftos, L. J. and Potts, J. T. Jr. (1969) Control of secretion of thyrocalcitonin in the rabbit as evaluated by radioimmunoassay. Endocrinology, 84: 36-40. 3 Cooper, C. W., Deftos, L. J. and Potts, J. T. Jr. (1971) Direct measurement of in vivo secretion on pig thyrocalcitonin by radioimmunoassay. Endocri- Ca Effect on CT Release in vitro in Frog 789 nology, 86: 747-754. Bell, N. H. (1975) Further studies on the regulation of calcitonin release in vitro. Horm. Metab. Res., 7: 77-83. Ziegler, R., Telib, M. and Preiffer, E. F. (1968) The secretion of calcitonin by perfused ultimobran- chial gland of the hen. Horm. Metab. Res., 1: 39- 40. Boelkins, J. N. and Kenny A. D. (1973) Plasma calcitonin levels in Japanese quail. Endocrinology, 92: 1754-1760. Feinblatt, J. D. and Raisz, L. G. (1971) Secretion of thyrocalcitonin in organ culture. Endocrinology, 88: 797-804. Feinblatt, J. D., Raisz, L. G. and Kenny, A. D. (1973) Secretion of avian ultimobranchial calcitonin in organ culture. Endocrinology, 93: 277-284. Ross, B. A. and Deftos, L. J. (1976) Calcitonin secretion in vitro II. Regulation effects enteric mammalian polypeptide hormone on the trout C- cell cultures. Endocrinology, 98: 1284-1288. 10 11 12 13 14 Shiraki, M., Hasegawa, S., Hirano, T. and Orimo, H. (1982) Calcitonin secretagogues in the eel. In “Comparative Endocrinology of Calcium Regula- tion” Ed. by C. Oguro and P. K. T. Pang. Japan Sci. Soc. Press, Tokyo, pp. 125-133. Oguro, C. Nagai, K., Tarui, H. and Sasayama, Y. (1980) Hypocalcemic factor in the ultimobranchial gland of the frog, Rana rugosa. Comp. Biochem. Physiol., 68: 95-97. Robertson, D. R. (1968) The ultimobranchial body in Rana pipiens. Z. Zellforsh., 90: 273-288. Sasayama, Y. and Oguro, C. (1976) Effects of ultimobranchialectomy on calcium and sodium con- centrations of serum and coelomic fluid in bullfrog tadpoles under high calcium and high sodium en- vironment. Comp. Biochem. Physiol., 55A: 35-37. Sasayama, Y. and Oguro, C. (1985) The role of the ultimobranchial glands on Ca balance in bullfrog tadpoles. In “Current Trends in Comparative En- docrinology” Ed. B. Loft and W. H. Holmes. Hong Kong Univ. Press, Hong Kong, pp. 837-838. ii iW “ihe oe Ae EE at Ta i ¥ berth pene ret heave ek) PLE: e A NEA: RPM BRD AAD, ey Rea wi 1 Fi by ue via rf oh Sh : ; ‘yw ‘ / th iay ; rose] Pensa F si) a7 ac ate tet ‘ bul tira ny eh 5 ef ‘4 i 7 : rita : P = n : L a/ y i eRe er on ua ac : Ae wi \2 vidi vad 7 1 i - wits tt Atte wT AD are a a UL RD A vovlat* hs AEVO totter Jolog pelle crawl | i wash Bebe a wb uf hed , - Crone cere TES, el Patt ae! 4 ; tet bs wu peer t Corny een Le i i eras Bi | " ’ ' iv 4 : t | la © h “ x a i j 4 - ’ § i ‘ ( : ; ‘aan “ ; \ i ‘ ay ry ral HW" igus, ayer nat ‘sa ZOOLOGICAL SCIENCE 9: 791-802 (1992) © 1992 Zoological Society of Japan Effects of Chronic Treatment with Chlorpromazine on the Aging of Hypothalamo-Pituitary-Ovarian Axis in the Rat Yuta Kopayasut'”, Sencuiro KAwAsHima!, SuMIO TAKAHASHI? and KatsuMI WAKABAYASHI* ‘Zoological Institute, Faculty of Science, University of Tokyo, Tokyo 113, Department of Pharmacology, Shimane Medical University, Izumo 693, 3Department of Biology, Faculty of Science, Okayama University, Okayama 700, and *Institute of Endocrinology, Gunma University, Maebashi 371, Japan ABSTRACT— Effects of chlorpromazine (CPZ) on the age-related changes of the estrous cycle and the glial histology in the hypothalamus were studied in female Wistar rats. CPZ (3.0 or 5.5 mg/day/rat) was orally administered for 6 months beginning at 8 months of age. About 28% of rats given no CPZ treatment showed prolonged estrus at 13-14 months of age, while about 41% of rats was cycling. In contrast, none of high-dose CPZ-treated rats of the same age was in a prolonged estrous state and the difference was statistically significant. The ovarian weight was heavier and the number of Graafian follicles in the ovary was less in CPZ-treated rats. In the hypothalamic arcuate nucleus, CPZ inhibited the increase in the number of astrocytic granules. These results indicate that CPZ treatment effectively delayed the occurrence of some age-related changes of the ovarian function and the hypothalamus in the rat. INTRODUCTION Chlorpromazine (CPZ) is a common antipsycho- tic drug showing complex biological effects [1]. Although many patients are given CPZ for years, the reports on the effect of long-term administra- tion of CPZ on aging processes are a few [2]. Reduction of the accumulation of lipofuscin pig- ments during aging in neurons of the nucleus reticularis gigantocellularis by long-term oral administration of CPZ in C57BL/10 female mice was observed [3]. In addition, the effect of CPZ on neurons of rat cerebral cortex in primary culture was reported by Ohtani and Kawashima [4]. They observed that CPZ was effective in reducing the accumulation of lipofuscin pigments in cultured neurons. The lipofuscin pigments accumulate in various tissues during aging and they are regarded Accepted May 28, 1992 Received April 8, 1992 Reprint requests should be addressed to: Dr. S. Kawa- shima, Zoological Institute, Faculty ef Science, Universi- ty of Tokyo, Tokyo 113, Japan. as one of the parameters of aging [5, 6]. These results led us to propose a working hypothesis that CPZ affects some aging processes in the brain. The age-related increase in the astrocytic and microglial activities in the hypothalamic arcuate nucleus in female rats was reported [7, 8] and it is suggested that the age-related increase in the hypothalamic gliosis may be closely related with the deterioration of hypothalamic function. Simi- larly, an increase in the number of reactive micro- glia (microglia containing more than five granules) and the number of astrocytic granules in the arcu- ate nucleus with aging in the rat and mouse was reported [9]. The decline in reproductive function in female rats with aging has been observed as the loss of regular estrous cycles and the manifestation of irregular cycles, usually characterized by pro- longed vaginal estrus at middle ages [10-14]. The age-related deviation of the estrous cycle was suggested to be dependent on the hypothalamus (for review, see 13). The present paper will deal with the effects of 792 Y. KosBAyAsHI, S. KAWASHIMA et al. chronic administration of CPZ on the occurrence of prolonged vaginal estrus and the microglial and astrocytic histology in the arcuate nucleus in female rats during aging. MATERIALS AND METHODS Experimental design One hundred fifty female rats of the Wistar strain at 4 weeks of age were purchased from Nihon Rat Co. Ltd. (Urawa, Japan). Upon arri- val, they were housed 5 each in an aluminum cage lined with wood shavings at 22°C under 12 hr-light (6:00-18:00) and 12hr-dark cycle, and were given rat chow (CA-1, Clea Japan Inc., Tokyo) and tap water ad libitum. These animals generally live to a maximum of 30 months of age. From 31 weeks of age, the commercial chow was replaced by hand-made chow containing no CPZ. CPZ administration in hand-made chow began in some rats at 32 weeks of age. From 56 weeks of age, commercial chow was supplied again. The procedures to prepare hand-made chow were as follows: The chow powder (CA-1) was kneaded with tap water equivalent in weight to 90% of the chow powder with or without CPZ (Contomin, Yoshitomi Pharmaceutical Ltd., Osa- ka) and chopped into cubes of about 15 mmx 15 mm 30mm. The cubes were dried in an oven about 10 hr at 80°C to 60% of the original weight. Rats were divided into the following five groups; (1) High dose CPZ-treated group. Forty rats were given chow containing 0.35 mg of CPZ per g CA-1 powder. (2) Low dose CPZ-treated group. Forty rats were given chow containing 0.175 mg CPZ/g. (3-1) Food restricted group to the level of the food intake of high-dose CPZ-treated group. Twenty rats were given the corresponding restricted amount of hand-made chow containing no CPZ. (3-2) Food restricted group to the level of the food intake of low dose CPZ-treated group. Twenty rats were given the corresponding restricted amount of hand-made chow containing no CPZ. (3-3) Control ad libitum feeding group. Hand- made chow containing no CPZ was given ad libitum to thirty rats. Food restricted groups were prepared because the food intake of CPZ-treated groups was less than that of control group of ad libitum feeding. The consumption of food per day in each cage of CPZ-treated groups was recorded once a week and the same mean daily amount of chow without CPZ was given to the corresponding food restricted group in the next week. In high dose and low dose CPZ-treated groups, the aver- age uptake of CPZ was 5.5 mg/day/rat and 3.0 mg/day/rat, respectively. During 41st week of age, water consumption in each cage was mea- sured daily. Vaginal smears were recorded during 25-27, 40-42, 53-55 and 60-62 weeks of age daily at 15 :00-17:00. Animals for plasma hormone determination and ovarian histology were killed by decapitation at 56, 63 and 68 weeks of age between 11:00 and 14:00. The plasma was separated from the trunk blood and kept frozen at —20°C until radioimmunoassay for luteinizing hormone (LH) and _ prolactin (PRL). The ovaries and uterus were weighed, and the ovaries were fixed in Bouin’s fluid. Histological observation of ovaries Paraffin sections of the right and left ovaries were cut at 10 ~m thickness, and one each section at the first, second and third quarters of the serial sections, i.e. six sections per animal, was stained with hematoxylin and eosin. The numbers of Graafian follicles and corpora lutea larger than 500 ym in diameter were counted. Histological observation of microglial and astrocy- tic granules At 59 weeks of age, five control rats (four persistent estrous and one cycling rats), five low dose CPZ-treated rats (one persistent estrous and four cycling rats) and five high dose CPZ-treated cycling rats were used for this study. For cycling rats those at estrus were used. They were anesthe- tized by i.p. injection of nembutal and perfused with Lock’s solution from the heart ventricle, followed by perfusion with 0.12 M phosphate buf- fer containing 1% glutaraldehyde and 1% para- formaldehyde. The hypothalami were dissected out and fixed in 0.1 M cacodylate buffer containing 1% glutaraldehyde and 1% paraformaldehyde for 4hr at 4°C. Then the hypothalami were postfixed Effects of Chlorpromazine on Rat Aging in 1% OsO, in the same buffer for 2 hr and embedded in epon. The ovaries were fixed in Bouin’s fluid for histological observations. The microglial and astrocytic changes were quantified according to the method described by Brawer et al. [7]. Briefly, coronal semithin sections 1 ym thick were cut at the mid-level of the arcuate nucleus (4-6 sections). These semithin sections were stained with toluidin blue and were observed under a light microscope. The microglia were identified as relatively small ellipsoidal cells con- taining toluidin blue-positive granules. The astro- cytes are small cells generally stained pale with less frequent granules. Four sections from the ipsi- lateral arcuate nucleus in each rat were used for measurements, and the number of reactive micro- glia (the cell containing more than five granules) and the number of granules in the astrocyte were counted. Radioimmunoassay Plasma concentrations of LH and PRL in rats at 56 and 63 weeks of age were determined with radioimmunoassay kits, NIADDK rat LH-I-5, anti-rat LH S-4, rat LH RP-1 (0.03 NIH unit/mg), rat PRL-I-3, anti-rat PRL S-6 and rat PRL PR-1 (11 IU/mg) provided by Rat Pituitary Hormone Distribution Program, NIADDK, National Insti- tutes of Health, Bethesda, Md. The assay results for LH and PRL were expressed as mg of NIH-rat- LH-S, (=NIH unit), and mIU, respectively. Statistics The differences in body weight, food intake and water intake were tested by analysis of variance (ANOVA). Tissue or organ weights, results of vaginal smear and blood LH and PRL concentra- tions were compared by Mann-Whitney’s U-test, as the distribution of the data did not show appa- rent “normal distribution”. Concerning the results of histological observation of the ovaries, the numbers of follicles and large corpora lutea were compared by Mann-Whitney’s U-test, and intersti- tial tissue hypertrophy was compared using Fisher’s exact probability test as the results were non-parametric. The numbers of reactive micro- glia and astrocytic granules were tested by ANO- VA and Duncan’s multiple range test. 793 RESULTS Body weight, food intake and water intake The number of death by 56 weeks of age and/or apparently unhealthy rats at 56 weeks was as follows; (1) high dose CPZ-treated group, 4; (2) low dose CPZ-treated group, 1; (3-1) food res- tricted group for high dose CPZ group, 0; (3-2) food restricted group for low dose CPZ group, 3; (3-3) ad libitum feeding group, 2. There was no significant difference among groups (Fisher’s exact probability test). Changes in body weight during experiments are shown in Figure 1. The body weight temporarily decreased in all groups after changing commercial chow to hand-made chow, and then, it began to increase. The inhibition of body weight gain in CPZ-treated groups was dose-dependent (Fig. 1, Table 1). After the withdrawal of CPZ from the chow at 56 weeks of age, the rate of body weight gain recovered to the level of ad libitum feeding group (Fig. 1). The body weights of food res- tricted groups were smaller than those of the corresponding CPZ-treated groups and ad libitum feeding group, although consumed food amount 450) ++ Ad lib feeding (19) °-—- CPZlowdose (19) *-—s €PZ high dose (19) a Body weight (g) 300° —_ 25 30 35 40 45 50 55 60 65 Age (weeks) Fic. 1. Changes in body weight by long-term chlorpro- mazine (CPZ) administration. Hand-made chow containing 0.175 mg of CPZ per g chow powder was given to low dose group and that containing 0.35 mg/g, to high dose group of rats from 32 to 55 weeks of age. The period of CPZ administration is shown as solid column in abscissa. The numbers in pa- rentheses indicate the numbers of rats. * P<0.05 compared with the control ad libitum feeding group (ANOVA). 794 Y. KOBAYASHI, S. KAWASHIMA et al. TaBLe 1. Effects of chlorpromazine (CPZ) on body weight, food intake and water intake Caer Body weight Body weight Food intake Water intake P (g) gain (g) (g/day/rat) (ml/day/rat) Ad libitum feeding group 398+9 (28) 47 +3 (28) 18.6+0.3 (6) 28.8+1.6 (6) low dose CPZ group Bw aeY) (39) suiae se (GY) 17.4+0.3* (8) 24.7+0.6 (8) high dose CPZ group 362+7** (36) 10+3** (36) IS). 7/sE0).3'* (() 19.0+0.6** (8) Food restricted groups ie to low dose group 368+8* (17) 8+6** (17) 17.5+0.3* (4) 31.54+2.2* (4) to high dose group 352+7** (20) 6+3** (20) 16.0+0.3** (4) 28.44+1.17* (4) Mean+S.E. Numbers in parentheses indicate the number of animals. Body weight at 55 weeks of age. Body weight gain during 32 and 55 weeks of age. Food intake shows the mean during 32 and 55 weeks of age. Water intake at 41 weeks of age. * P<0.05, ** P<0.01 compared with ad libitum feeding group (ANOVA). # P<0.05, ** P<0.01 compared with corresponding CPZ-treated group. was adjusted to be equal to that of CPZ-treated groups (Table 1). The difference in body weight gain between low dose CPZ group and matched food restricted group was statistically significant (Table 1). The food intake in CPZ-treated groups was significantly and dose-dependently decreased. The water intake in high dose CPZ-treated group was significantly less than that in ad libitum feeding group, while the water intake in low dose CPZ- treated group and food restricted groups was not significantly different from that in ad libitum feed- ing group. Water intake in food restricted groups was significantly greater than that in matched CPZ-treated groups. Estrous cycle, ovarian histology and plasma LH and PRL concentrations In the gonadal functions of control rats, there were no significant differences between ad libitum feeding group and food restricted groups. There- Fic. 2. Effects of chlorpromazine (CPZ) adiministra- tion on frequency distribution of the total number of proestrous and estrous days (left) and the number of estrous cycles (right) in young rats at 25-27 weeks of age (top panel) and rats at 40 to 42 weeks of age (lower three panels). Arrow indicates the median of the distribution. n=number of rats. The frequency distribution of high dose CPZ-treated group was significantly different from that of the control (P< 0.05) (U-test). fore, data of ad libitum feeding group and food restricted groups were combined and served as the control in the following analyses. At 25-27 weeks of age, regular 4- or 5-day estrous cycles were observed (Fig. 2). At 40-42 weeks of age, a few control rats showed prolonged % Young n=30 Control n=65 CPZ low dose n=39 CPZ high dose n=36 6 2 4 6 8101214161820 135 7 9111315171921 Number of days in proestrus and estrus during 3-week period 01234567 Number of estrous cycles during 3-week period Effects of Chlorpromazine on Rat Aging estrus (Fig. 2). Some CPZ-treated rats showed estrous cycle with relatively long diestrous phase and a few rats showed prolonged diestrus. The median of the number of proestrous plus estrous days of the control group at 42—42 weeks was not significantly different from those of rats at 25-27 weeks and of high dose CPZ-treated group. The frequency distribution of prostrous and estrous days of high dose CPZ-treated group was signi- ficantly different from that of the control. The number of estrous cycles during a 3-week period was about 5 in the young and high dose CPZ- treated groups and it was about 4 in the control at 40—42 weeks of age (P<0.05, vs. young group). Fic. 3. Effects of long-term chlorpromazine (CPZ) administration on frequency distribution of the total number of proestrous and estrous days (left) and the number of estrous cycle (right) in rats at 53 to 55 weeks of age. Arrow indicates the median of the distribution. n=number of rats. The frequency distribution of the control group was significantly different from that of high dose CPZ-treated group (P<0.01) and that of low dose CPZ-treated group (P<0.02). The frequency distribution of low dose CPZ-treated group was significantly different from that of high dose CPZ-treated group (P<0.01). The frequency distribution of the control group was significantly different from that at 40 to 42 weeks of age shown in Fig. 2 (P<0.01) (U-test). 795 The distribution of vaginal smear pattern at 53- 55 weeks of age is shown in Figure3. In the control group about 28% of rats showed prolonged estrus, resulting in a mass of distribution locating toward the right. Another mass was present in the % Control n=65 30 u \/ 20 10 0 CPZlowdose n=39 v 30 y 1| 20 10 CPZ high dose n=36 Vv 0 0 2 4 6 810121416 18 20 01234567 135 7 9111315171921 Number of days in proestrus and estrus during 3-week period Number of estrous cycles during 3-week period TaBLE2. Effects of chlorpromazine (CPZ) administration (from 32 weeks of age to 55 weeks of age) on some organ weights and plasma LH and prolactin (PRL) concentrations at 56 weeks of age (upper part), and changes in these parameters at 63 weeks of age, 7 weeks after the cessation of CPZ treatment (lower part) Control low dose CPZ group high dose CPZ group At 56 weeks of age Ovarian weight (mg/100 g BW) 2.8) 38 IY) (18) A oY SEAS) (10) 28.5+1.0* (9) Uterine weight (mg/100 g BW) 191+19 (18) 152+ 16 (10) 146+ 13 (9) LH concentration (mg S,/ml) 0.88 + 0.33 (18) 0.39+0.06 (10) 0.51+0.11 (9) PRL concentration (mIU/ml) 3.22 +0.99 (18) 1.96 +0.66 (10) 2.49 + 0.74 (9) At 63 weeks of age Ovarian weight (mg/100 g BW) WD) ae MF (19) 23.2+2.4% (10) ipo rial (10) Uterine weight (mg/100 g BW) 192+ 14 (19) 185+ 12 (10) 173+10 (10) LH concentration (mg S,/ml) 0.52+0.11 (19) 0.44 +0.09 (10) 0.57+1.10 (10) PRL concentration (mIU/ml) 3.32 £0.66 (19) 2.15+0.46 (10) 1.80+0.28 (10) Mean+S.E. BW; body weight. Numbers in parentheses indicate the number of rats. * P<0.05; compared with control group, * P<0.05, compared with the values at 56 weeks of age (U-test). 796 Y. KoBAYASHI, S. KAWASHIMA et al. middle showing cyclic estrous pattern (about 41%). The number of estrous cycles of the control group showed a large variation indicating irregular cycle of some animals. In contrast, many rats showed cyclic estrous pattern with a fragmentary mass at the right in low dose CPZ-treated group. In high dose CPZ-treated group, the median of distribution of proestrous and estrous days was located in the middle with a small mass at the left showing a prolonged diestrous phase. The fre- quency distribution was significantly different among the three groups. The frequency distribu- tion of the control group at 53-55 weeks of age was significantly different from that at 40-42 week of age (Fig.3 vs. Fig.2). The difference in the frequency distribution of CPZ-treated groups be- tween 40-42 and 53-55 weeks was statistically not significant. Some organ weights and plasma levels of LH and PRL in three groups at 56 weeks of age are shown in Table 2. The ovarian weight was signi- ficantly heavier in high dose CPZ-treated group than the control group. The low dose CPZ-treated group showed the same tendency. The uterine weight was lighter, and LH and PRL levels were lower in CPZ-treated groups than the control group, although these differences were statistically not significant. TaBLeE 3. Summary of histological observations of the ovaries of rats given chlorpromazine (CPZ) Follicles ale Number of follicles (>500 ~m) Total nue 0-2.0 2.1-4.0 4.1-7.0 CLES Control ay) 16 6 26 CPZ low 10 7 3 20 CPZ high** 10 4 3 17 Corpora lutea Number of corpora lutea (>500 um) Total number Group ¥ n sre of rats Control 5 3 18 26 CPZ low y, 0 18 20 CPZ high 0 16 17 Interstitial tissue Interstitial tissue hypertrophy Total number Group 2 ak a of rats Control 15 3 8 26 CPZ low 13 5 2 20 CPZ high 14 2 1 17 *) Number of rats. observation of hypothalamus) Follicles Ovaries were obtained at 56 weeks (Table 2) and at 59 weeks (for the Number of follicles per section is shown. ** P<0.01 compared with the control by U-test. Corpora (—) indicates no C.L. in 6 sections. lutea (C.L.) (+) indicates less than one C.L. per section. (+) indicates more than one C.L. per section. Interstitial (—) indicates no hypertrophy. tissues (+) indicates slight hypertrophy. (+) indicates marked hypertrophy. Effects of Chlorpromazine on Rat Aging 797 When control rats were divided into subgroups, one showing prolonged estrus (more than 18 days were in estrus or proestrus out of 21 days) and the other showing regular estrous cycles (8-13 days of estrus or proestrus out of 21 days), the ovarian weight of the former was significantly lighter than the latter (prolonged estrus, 15.2+2.2 mg/100 g body weight, n=7; regular cycle, 28.2+2.1 mg/ 100 g body weight, n=7; P<0.01). The uterine weight was significantly heavier (prolonged estrus, 226+23 mg/100 g body weight; regular cycle, 126 +17 mg/100 g body weight; P<0.01) in prolonged estrus group than regular cycle group. The LH level was greater in the former group (prolonged estrus, 1.61+0.80 mg S,/ml; regular cycle, less than detectable level of 0.25 mg S;/ml; P<0.01 by Fisher’s exact probability test). The PRL level was greater (prolonged estrus, 5.6+2.3 mIU/ml; regu- lar cycle, 1.5+0.5 mIU/ml) in prolonged estrus group than regular cycle group, however, the difference was not significant. In CPZ-treated rat subgroups, the results were similar to those in the control subgroups. The ovaries of rats showing regular estrous cycles in three groups contained some follicles and corpora lutea, and the hypertrophy of interstitial tissue was not observed in these ovaries. On the contrary, the ovaries of rats showing prolonged estrus in the three groups contained numerous follicles but no or few large corpora lutea, and showed the hypertrophy of interstitial tissue. Sum- mary of ovarian histology at 56 and 59 weeks of age is shown in Table 3. In the ovaries of 85% of control rats, more than two large follicles per section were observed. In the ovaries of 59% of high dose CPZ-treated group rats and 50% of low dose CPZ-treated rats, less than two follicles per section were present. The difference in the num- ber of follicles between the control and high dose CPZ-treated group was statistically significant. On the other hand, less than one large corpora lutea per section was observed in the ovaries of about 30% of control rats, and in those ovaries the hypertrophy of interstitial tissue was apparent. In 94% of the ovaries of high dose CPZ-treated rats and 90% of them of low dose CPZ-treated rats, more than one large corpora lutea per section was present. Difference in the number of large corpora lutea was not statistically significant. Seven weeks after the cessation of CPZ treat- ment the same parameters were studied. The occurrence of estrous and proestrous days during a 3-week vaginal smear test at 60-62 weeks of age is shown in Figure 4. In CPZ-treated groups, the incidence of rats showing prolonged estrus became greater as compared with that at 53-55 weeks of age, and there was no significant difference from the value of the control group. The shift of the frequency distribution was significant in the high and low dose CPZ-treated groups at 60-62 weeks of age compared with that of the same animals at % Control n=39 fo y } 10 0 — CPZlowdose n=19 | 30 | i? 20} 1 CPZ highdose n=19 30 | ' ' | Vi 20 | 10 j 0 0 2 4 6 8101214161820 135 7 9111315171921 Number of days in proestrus and estrus during 3-week period 01234567 Number of estrous cycles during 3-week period Fic. 4. Effects of the cessation of chlorpromazine (CPZ) administration on frequency distribution of the total number of proestrous and estrous days (left) and the number of estrous cycle (right) in rats at 60 to 62 weeks of age. Solid arrow indicates the median of the distribution. n=number of rats. The difference in the frequency distribution of the con- trol group and CPZ-treated group was not statisti- cally significant. The median of the frequency distribution of high and low dose CPZ-treated groups became significantly larger than those (broken arrows) of the identical groups at 53 to 55 weeks of age (P<0.01) (U-test). 798 53-55 weeks of age. The ovarian weight of CPZ-treated rats at 63 weeks of age was not significantly different from that of the control (Table 2). The ovaries reduced in weight seven weeks after the cessation of CPZ treatment compared with that during CPZ treat- ment in low dose CPZ-treated group (Table 2). Plasma levels of LH and PRL at 63 weeks of age were similar to those at 56 weeks of age (Table 2). Observations of ovarian histology revealed no significant differences among the three groups at 63 weeks of age, although the number of follicles in high dose CPZ-treated group tended to be smaller than the other two groups (Table 4). The increase in the number of follicles in low dose CPZ-treated group and the increase in the inci- dence of interstitial tissue hypertrophy in high dose CPZ-treated group were significant after the cessa- Y. KoBAyYAsHI, S. KAWASHIMA et al. tion of CPZ treatment compared with that during treatment (Tables 3 and 4). Vaginal smear pattern, organ weights and ova- rian histology in rats killed at 68 weeks of age (9 high dose CPZ-treated, 9 low dose CPZ-treated and 20 control rats; data not shown) were similar among the three groups and were almost the same compared with those at 63 weeks. The number of reactive microglia and astrocytic granules Figure 5 shows an example of reactive microglia and astrocytic granules in the hypothalamic arcu- ate nucleus of a control rat. The effects of CPZ on the arcuate glial histology are presented in Table 5. In the number of reactive microglia, no signi- ficant differences were detected among the three groups, although the number of reactive microglia TABLE 4. Summary of histological observations of the ovaries of rats 7 weeks after the cessation of chlorpromazine (CPZ) treatment Follicles Gat Number of follicles (>500 «m) Total Aono 0-2.0 2.1-4.0 4.1-7.0 of rats Control 6» 7 6 19 CPZ low* 2 4 ti CPZ high A i rm Corpora lutea Number of corpora lutea (>500 «m) Total number GrONp a u ie of rats Control 1 3 15 19 CPZ low 1 8 10 CPZ high 0 2 8 10 Interstitial tissue Interstitial tissue hypertrophy Totalbwumbee Group i of rats = ae + Control 12 2 5) 19 CPZ low 4 4 2 10 CPZ high* 4 3 3 10 @) Number of rats. # P<0.05, compared with the score during CPZ administration (Table 3) by U-test (follicles) or Fisher’s exact probability test (interstitial tissue). Follicles; Number of follicles in a section is shown. interstitial tissue, refer to Table 3. For indices of corpora lutea and Effects of Chlorpromazine on Rat Aging 799 Fic. 5. Mid-level of hypothalamic arcuate nucleus of a control rat. Toluidin blue preparation. Microglia (arrows) and astrocytic granules (arrow heads) are visible. Lipofuscin-like granules are present in the neurons (N). Scale: 5 um. TaBLE5. Glial histology in the hypothalamic arcu- ate nucleus in chlorpromazine (CPZ)-treated rats killed at 59 weeks of age Number Number of Number of Group of reactive astrocytic rats microglia® granules Control 5 7.6+1.1 SERS Ie CPZ low 5) 7.0+1.9 ADS seas CPZ high 5 3.2+0.6 Aide Mean+S.E. a; Number of microglia containing more than five granules in four semithin sections. b; Total number of granules in astrocytes in four sections. ** P<(0.01, among groups (ANOVA). ## P<(.01, difference from the control (Duncan’s multiple range test). tended to decrease in high dose CPZ-treated group. On the other hand, the ANOVA clearly showed a singificant effect of CPZ treatment on the number of astrocytic granules. The numbers of astrocytic granules in low and high dose CPZ- treated groups were significantly less than that in the controls (P<0.01 in both comparisons). DISCUSSION Reduction in body weight gain by chronic CPZ treatment in rats in the present study agrees with the previous report using mice [3]. A significant decreases in food intake may be the main reason for the reduction in the rate of body weight increase. Although food intake was manipulated to be the same by food restriction, body weight gain was significantly less in food restricted control rats than CPZ-treated rats. We have already reported that 5 months treatment with high dose of CPZ reduced the ambulatory activity in open field test in 13-month-old rats [15]. Such a change in behavior in CPZ-treated rats may correlate to the energy metabolism and may result in larger body weight gain in CPZ-treated groups than food in- take-matched control rats. Food restriction increases longevity [16-18]. This is the reason why we prepared a food res- tricted group for each dose group. However, there was no difference between ad libitum feeding group and food restricted groups as concerns the parameters of gonadal function in the present study. Therefore, it is considered that the effects of CPZ on gonadal function observed in the pre- sent study were not the consequences of the reduc- tion in food intake and body weight gain. Previous studies restricted food to 40 to 60% of ad libitum intake [17-18]. In the present study, the reduction of food intake by low and high doses CPZ-treated groups were to 94 and 84% of ad libitum intake, respectively. Such a small reduction of food intake may have no influence on aging. Samorajski and Rolsten [3] could not detect any change in water intake by CPZ treatment. The present results disagree with theirs. On the other hand, water intake of food restricted groups was almost the same as that of ad libitum feeding group, indicating that the reduction in water intake by CPZ treatment is mediated not through the reduction of food intake. Vaginal smear records showed that the adminis- tration of CPZ was effective for the maintenance of regular estrous cycle, when prolonged estrous State was predominant in age-matched control rats. Ovarian weight and the number of follicles in high dose CPZ-treated rats were significantly diffe- rent from those in the control, due to the differ- ence in the proportion of subtypes of gonadal function between the two groups. Any differences on some other organ weights but ovaries and on ovarian histology were not observed between CPZ-treated rats showing estrous cycle and control 800 Y. Kopayasui, S. KAWASHIMA et al. cyclic rats, indicating normal gonadal function in the cyclic CPZ-treated rats. The blood hormone levels, uterine weight and ovarian histology did not conflict with this interpretation. At 63 weeks of age, 7 weeks after the cessation of CPZ adminis- tration, regular estrous cycles disappeared in most CPZ-treated rats. In our separate study, high dose of CPZ chow was administered for 10 days to 12-month-old rats showing persistent estrus, but any recovery to regular estrous cycles was not observed (unpub- lished observation). The maintenance of estrous cycle by CPZ seems to be related with long-term administration of this agent from younger ages. Aschheim [12, 19] demonstrated that the trans- plantation of ovaries from young rats into ovar- iectomized aged animals which had been in persis- tent estrous state did not result in reinitiation of estrous cycles, and he suggested that the origin for the loss of estrous cycle in aged rats was not at the ovarian level. Electrical stimulation of the preop- tic area of aged rats restored resproductive func- tion [20]. Maintenance of cyclic cornification of the vaginal epithelium was observed in young adult female rats given heterochronic transplantation of the pituitary and the ovary from aged female rats. These results indicate that the age-related devia- tion of the estrous cycle was not dependent on the ovary but on the hypothalamus. The hypothalamus or higher central nervous system is important as the action site of CPZ in the present study. CPZ treatment decreased the accu- mulation of astrocytic granules compared with controls. The increase in the glial activities in microglia and astrocytes with aging was reported [9]. In a preliminary study, we also found the age-related increase in glial activities. In Wistar/ Tw strain maintained in our laboratories, the num- ber of reactive microglia in the arcuate nucleus in 19-21-month-old females showing persistent estrus (33.8+3.5 cells/4 sections, n=6) was signi- ficantly greater (P<0.001) than that in 5-month- old females at estrus (5.6+0.3 cells/4 sections, n= 5). Similarly, the number of astrocytic qranules in the arcuate nucleus in aged females (871+43 granules/4 sections, n=7) is significantly higher (P <0.001) than that in young females (215+25 granules/4 sections, n=5). Similar changes were reported in other brain areas: for example, the increase in the numerical density of hypertrophied astrocytes in the hippocampus in rats with aging [22]. The reduction in astrocytic granules by CPZ treatment in the present study may have some relevance to the delay in gonadal aging. CPZ treatment reduced the accumulation of lipofuscin pigments in the nucleus reticularis giagantocellularis of mice [3]. They interpreted the effect of CPZ on its action on the biomem- brane properties. It is suggested that the forma- tion of lipofuscin was related with peroxidation of lipids in biomembranes [6]. In support of this view, some anti-oxidants or free-radical scavengers have been reported to reduce the accumulation of lipofuscin pigments and show other anti-aging acti- vities [23-26]. The anti-oxidative activity or effects of CPZ on biomembrane properties may have some relation with the present anti-aging effects of CPZ. Acute administration of CPZ (25 mg/kg body weight) stimulates PRL release from the pituitary due to its action as a dopamine antagonist [1, 27— 29]. Inhibition of gonadal function by enhanced PRL release and the occurrence of some pro- longed diestrous rats in CPZ-treated group may be correlated with the PRL release activity of CPZ. However, in the present study, blood PRL level was somewhat lower in CPZ-treated rats at 56 weeks of age. Measurement of blood PRL level in CPZ-treated rats at earlier period, especially dur- ing prolonged diestrous state, was necessary to clarify the correlation. The possibility that other effects of CPZ, such as a non-competitive acetyl- choline blocking action [30] on the central nervous system, caused the delay of aging in the hypothala- mo-hypophysial-gonadal system cannot be ex- cluded. To conclude, delay of the occurrence of some age-related changes of the ovarian function and the hypothalamic glial histology by chronic CPZ treatment in the rat was evident. ACKNOWLEDGMENTS The authors wish to express their gratitudes to Dr. S. Raiti and the Pituitary Hormone Distribution Program, the National Institute of Arthritis, Diabetes and Diges- Effects of Chlorpromazine on Rat Aging tive and Kidney Diseases (NIADDK) and Dr. A. F. Parlow, Pituitary Hormones/ Antisera Center, Harbor- UCLA Medical Center for kind supply of RIA kits. This work was supported in part by Grants-in-Aid from the Ministry of Education, Science and Culture, Japan (Nos. 63304008, 02404007). 10 11 1, REFERENCES Baldessarini, R. J. (1990) Drugs and the treatment of psychiatric disorders. In “Goodman and Gilman’s The Pharmacological Basis of The Therapeutics”. Eds. by A. Goodman Gilman, T. W. Rall, A. S. Nies and P. Tayler, Macmillan Publ. Co., New York, pp. 387-445. Stavroula, G. Y., Manoussakis, M. N., Mavridis, A. K., Tzioufas, A. G., Loukas, S. B., Plataris, G. K., Liakos, A. D. and Moutsopoulos, H. M. (1990) Factors related to the presence of autoantibodies in patients with chronic mental disorders. Biol. Psychiatry, 27: 747-756. Samorajski, T. and Rolsten, C. (1976) Chlorproma- zine and aging in the brain. Exp. Geront., 11: 141- 147. Ohtani, R. and Kawashima, S. (1983) Reduction of lipofuscin by centrophenoxine and chlorpromazine in the neurons of rat cerebral hemisphere in primary culture. Exp. Geront., 18: 105-112. Strehler, B. L., Mark, D. D., Midran, A. S. and Gee, M. V. (1959) Rate and magnitude of age pigment accumulation in the human myocardium. J. Gerontol., 14: 430-439. Constantinides, P. (1984) Ageing. In “Ultrastructu- ral Pathobiology”. Ed. by P. Constantinides, Else- vier, Amsterdam, pp. 321-335. Brawer, J. R., Naftolin, F., Martin, J. and Sonnens- chein, C. (1978) Effects of a single injection of estradiol valerate on the hypothalamic arcuate nuc- leus and on reproductive function in the female rat. Endocrinology, 103: 501-512. Brawer, J. R., Schipper, H. and Naftolin, F. (1980) Ovary-dependent degeneration in the hypothalamic arcuate nucleus. Endocrinology, 107: 274-279. Schipper, H., Brawer, J. R., Nelson, J. F., Felicio, L. S. and Finch, C. E. (1981) Role of the gonads in the histologic aging of the hypothalamic arcuate nucleus. Biol. Reproduct., 25: 413-419. Ingram, D. L. (1959) The vaginal smear of senile laboratory rats. J. Endocrinol., 19: 182-188. Takahashi, S. (1980) Age-related changes in the vaginal smear pattern in rats of the Wistar/Tw strain. J. Fac. Sci. Univ. Tokyo Sec. IV, 14: 345- 349. Aschheim, P. (1983) Relation of neuroendocrine system to reproductive decline in female rats. In: “Neuroendocrinology of Aging”. Ed. by J. Meites, 13 15 20 21 22 23 24 25 26 801 Plenum Press, New York, pp. 73-101. Kawashima, S. (1983) Aging of hypothalamic- pituitary function in the rat. In: “Recent Trends in Life Sciences”. Eds. by A. Gopalkrishna, S. B. Singh and A. K. Saxena, Manu Pub., Kanpur, pp. 59-92. Scarbrough, K. and Wise, P. M. (1990) Age-related changes in pulsatile luteinizing hormone release precede the transition to estrous acyclicity and de- pend upon estrous history. Endocrinology, 126: 884-890. Kawashima, S. and Kobayashi, Y. (1985) Age- related changes and effects of long-term administra- tion of chlorpromazine in open field behavior of female rats. J. Sci. Hiroshima Univ., Ser. B, Div. 1., 32: 1-14. McCay, C. M., Crowell, M. F. and Maynard, L. A. (1935) The effect of retarded growth upon the length of life span and upon the ultimate body size. J. Nutrition, 10: 63-79. Yu, B. P., Masoro, E. J. and McMahan, C. A. (1985) Nutritional influences on aging of Fischer 344 rats: I. Physical, metabolic, and longevity character- istics. J. Gerontol., 40: 657-670. Stokkan, K.-A., Reiter, R. J., Vaughan, M. K. Nonaka, K. O. and Lerchl, A. (1991) Endocrine and metabolic effects of life-long food restriction in rats. Acta Endocrinol., 125: 93-100. Aschheim, P. (1964) Resultats fournis par la greffe heterochrone des ovaires dans l’etude de la regula- tion hypothalamo-hypophyso-ovarienne de la ratte senile. Gerontologia, 10: 65-75. Clemens, J. A., Amenomori, Y., Jenkins, T. and Meites, J. (1969) Effects of hypothalamic stimula- tion, hormones and drugs on ovarian function in old rats. Proc. Soc. Exp. Biol. Med., 132: 561-563. Peng, M.-T. and Hung, H.-H. (1972) Aging of hypothalamic-pituitary-ovarian function in the rat. Fertil. Steril., 23: 535-542. Lindsey, J. D., Landfield, P. W. and Lynch, G. (1979) Early onset and topographical distribution of hypertrophied astrocytes in hippocampus of aging rats: a quantitative study. J. Gerontol., 34: 661-671. Harman, D. (1968) Free radical theory of aging: Effect of free radical reaction inhibitors on the mortality rate of male LAF, mice. J. Gerontol., 23: 476-482. Constantinides, P., Harkey, M. and McLaury, D. (1986) Prevention of lipofuscin development in neurons by anti-oxidants. Virchow’s Arch., 409: 583-593. Heicklen, J. and Brown, E. (1987) Increase in life expectancy for mice fed diethylhydroxylamine (DEHA). J. Gerontol., 42: 674-680. Carney, J. M., Starke-Reed, P. E., Oliver, C. N., Landum, R. W., Cheng, M. S., Wu, J. F. and Floyd, 27 28 802 R. A. (1991) Reversal of age-related increase in brain protein oxidation, decrease in enzyme activity, and loss in temporal and spatial memory by chronic administration of the spin-trapping compound N- tert-butyl-a-phenylnitrone. Proc. Natl. Acad. Sci. USA., 88: 3633-3636. Lu, K.-H., Amenomori, Y., Chen, C.-L. and Meites, J. (1970) Effects of central acting drugs on serum and pituitary prolactin levels in rats. Endocri- nology., 87: 667-672. Takahashi, S., Kawashima, S. and Wakabayashi, K. (1979) Effects of chlorpromazine and estradiol benzoate on prolactin secretion in gonadectomized male and female rats. Endocrinol. Japon., 26: 419- 29 30 Y. KospayAsHI, S. KAWASHIMA et al. 422. Baron, J. C., Martinot, J. L., Cambon, H., Boulen- ger, J. P., Poirier, M. F. Caillard, V., Blin, J., Huret, J. D., Loch, C. and Maziere, B. (1989) Striatal dopamine receptor occupancy during and following withdrawal from neuroleptic treatment: correlative evaluation by positron emission tomography and plasma prolactin levels. Psychopha- rmacology., 99: 463-472. Changeux, J.-P., Pinset, C. and Ribera, A. B. (1986) Effects of chlorpromazine and phencyclidine on mouse C2 acetylcholine receptor kinetics. J. Physiol. , 378: 497-513. ZOOLOGICAL SCIENCE 9: 803-809 (1992) Differential Effects of Testosterone and 5a-Dihydrotestosterone on Growth in Mouse Submandibular Gland KAZUHIKO SAWADA and TETSuO NOUMURA Department of Regulation Biology, Faculty of Science, Saitama University, Urawa, Japan ABSTRACT—Testosterone and 5a-dihydrotestosterone (DHT) were compared their morphometrical effects on the mouse submandibular gland in relation to the developmental profiles of their serum levels. In male mice, serum levels of both testosterone and DHT were dramatically increased between days 20 and 30, when sexual difference of the submandibular gland occurred morphometrically. In 30-day-old males, serum testosterone and DHT levels were 2.9+0.59 ng/ml and 1.0+0.21 ng/ml, respectively (3:1 in the ratio). Daily injection of testosterone and DHT for 1-20 days starting on day 20 induced morphometrical changes in the glands of neonatally-castrated mice. The glandular weight gain was caused only in males by treatment with testosterone for 10-20 days, but in both sexes by DHT for 7-20 days. The relative occupied area (ROA) of the granular convoluted tubules (GCT) in castrated mice of both sexes were increased by treatment with testosterone for 10-20 days, but by DHT for 7-20 days. The mitotic rates of GCT were increased only in males by treatment with testosterone for 10-13 days, but in both sexes by DHT for 2-7 days. The present study suggests that androgenic effects of these two androgens on growth in the mouse submandibular gland are different, being more affected by DHT than by testosterone, and also that © 1992 Zoological Society of Japan androgen responsiveness of the gland is sexually different, being higher in male than in female. INTRODUCTION In rodents, the male submandibular gland is larger and has more complex morphology than the female one. The glandular contents of biologically active polypeptides including nerve growth factor, epidermal growth factor, renin and proteases, are higher in the male than in the female, being responsive to androgens [1-7]. By histological, ultrastructural and morphometrical studies, both sexes in the mice experience a similar mor- phogenesis of the mouse gland during perinatal development, and then the sexual difference arises at 3-4 weeks of age, when the granular convoluted tubules (GCT) grows more rapidly in the male than in the female [8-11]. The masculine develop- ment of GCT in the gland is caused by androgens, testosterone and 5a-dihydrotestosterone (DHT), but not by estradiol-17 [11]. In a completely androgen-independent state (neonatally-castration Accepted June 2, 1992 Received April 8, 1992 and androgen-insensitive Tfm mutation), the gland displays the feminine development [11, 12]. The present study was aimed to compare the effects of testosterone and DHT on morphometric- al changes in the mouse submandibular gland in relation to the developmental profiles of their circulating levels. MATERIALS AND METHODS Animals CD-1 mice were obtained from Charles River Japan Co. and maintained by randomly mating in our laboratory. The animals were given a com- mercial diet (CRF-1: Charles River Japan Co.) and tap water ad libitum and were kept at 23+1°C under 12 hr artificial illumination (from 8:00 to 20:00). Hormone assay In mice of both sexes on days 20, 30 and 90, blood sample was collected from the tail vein 804 K. SAWADA AND T. Noumura under the ether anesthesia. The serum was sepa- rated and stored at —20°C until assayed. Serum levels of testosterone and 5a-dihydrotestosterone were separately determined by a radioimmunoas- say. The inter- and intraassay variations of testos- terone were 11.5% and 5.2%, and those of DHT were 12.4% and 5%, respectively. Individual values were the mean of the duplicate determina- tions and were expressed as nanogram per millili- ter of serum. Treatments of animals and histological procudure Both male and female mice were castrated on day 0 and were subcutaneously given daily treat- ments with testosterone (10 or 100 ug), DHT (10 or 100 ug) or sesame oil vehicle alone (0.1 ml) for 1, 2, 4, 7, 10, 13, 16 and 20 days starting on day 20. Mice were killed on days 21, 22, 24, 27, 30, 33 36 and 40, respectively. All animals were sub- cutaneously given a single injection of colchicine (5 pg/g body weight) dissolved in 0.9% NaCl, 5 hr before they were killed to arrest cell division at metaphase. Submandibular glands were weight and fixed in Bouin’s solution, embedded in par- affin and sectioned at 8 wm. Sections were stained with Delafield’s hematoxylin and eosin. Morphometrical Methods The sectional figures of submandibular glands in the microscopic enlargements (x 400) were traced by the camera lucida, and the areas cf GCT were measured by the picture analyzer (Logitec K-510, Kantou Denshi Co., Japan) connected with a microcomputer (NEC PC-9801DA). The areas were expressed as percentages of total area (the relative occupied area, ROA). Cell height of GCT was measured in randomly chosen five sections. Mitotic Activities Cells in division per 5S00-1,000 cells were counted in the GCT of the glands. Mitotic rate was estimated by counting the cell at metaphase per 5 hr in 100 cells. Statistical Analysis Data were statistically analyzed by Student’s t-test or by Cochran-Cox test. RESULTS Serum levels of testosterone and DHT in mice On day 20, serum testosterone levels showed no sexual difference statistically, but were slightly higher in males than in females (0.35 +0.08 ng/ml and 0.23+0.01 ng/ml). In male mice, serum tes- tosterone levels dramatically raised on day 30 and attained adult levels (Fig. 1A). Similarly, serum DHT levels were slightly higher in males than in females on day 20 (0.45+0.10 ng/ml and 0.11+ 0.01 ng/ml) and then the DHT levels in males dramatically increased to adult levels by day 30 (Fig. 1B). At 30 and 90 days of ages, serum levels of testosterone were 3-fold higher in males and 2-fold higher in females than those of DHT, re- spectively (Fig. 1A and 1B). testosterone A serum levels (ng/ml) ~~ IN) (et) ag > 20 30 90 age in days Fic. 1. Serum levels of testosterone and DHT in mice. a: P<0.05, b: P<0.01 vs age-matched females (Cochran-Cox test). Androgens and Mouse Submandibular Gland 805 male A female 150 testosterone e oil ©1049 ia} 10049 100 50 weight of the glands (mg) 0) 10 20 0 10 20 treatment periods (days) Fic. 2. Change in the weight of the submandibular glands of neonatally-castrated mice, treated with testosterone (A, C) and with DHT (B, D) for 1-20 days from day 20. a: P<0.05, b: P<0.02, c: P<0.01, d: P<0.002, e: P<0.001 (Student’s ¢-test). male i female 60 testosterone c @ oil ° 10u9 40; © 100yu9 ~ ° 20 ~— i— © © — S D <x ie) x ro) 10 20 0 10 20 treatment periods (days) Fic. 3. Change in the relative occupied areas (ROA) of GCT of the submandibular glands of neonatally-castrated mice, treated with testosterone (A, C) and with DHT (B, D) for 1-20 days from day 20. 806 K. SAWADA AND T. NOUMURA Changes in the mouse submandibular gland after treatments with testosterone and DHT As circulating levels of androgens, both testos- terone and DHT, showed a dramatically increase between days 20 and 30, we compared the effects of these androgens on morphometrical changes in the submandibular glands of neonatally-castrated mice for 20 days from day 20. The gland weight in castrated males gradually increased and became significantly higher after 7 days of treatment with testosterone (Fig. 2A). In castrated females, the gland weight gain was ultimately induced after 16 days by treatment with 100 ug testosterone, but was failed by 10 ug testosterone during the period of 20-day treatment (Fig. 2C). Seven days and longer treatment with DHT caused to increase significantly the gland weight in both sexes. DHT was more effective on the gain in gland weight than testosterone (Fig. 2B and 2D). In castrated males, the ROA of GCT were male 25 fF testosterone height of GCT cells (um) gradually increased and reached plateau after 13— 16 days by treatment with 10-100 ug testosterone or 10 wg DHT, and after 10 days by 100 ng DHT (Fig. 3A-D). Similarly, the height of GCT cells in both sexes significantly increased after 7 days by treatment with testosterone and after 4 days by DHT (Fig. 4A-D). These androgenic effects were dose-dependent, and DHT was more effective on the GCT stimulation than testosterone (Fig. 3 and 4). Testosterone and DHT had no effect on the ROA of the intercalated ducts (ID) and excretory striated ducts (SD), and both androgens did not caused any changes in the size of acini in the glands (data not shown). The mitotic rates of GCT in castrated males were significantly increased by treatment with 10- 100 4g testosterone and with 10 “7g DHT only after 10-13 days (Fig.SA and 5B). In castrated females, 10-100 ug testosterone and 10 ~.g DHT failed to increase the mitotic rates during treat- ment periods of 20 days (Fig. 5C and 5D). By female 20 0 10 20 treatment periods (days) Fic. 4. Change in the height of GCT cells of the submandibular glands of neonatally castrated mice, treated with testosterone (A, C) and with DHT (C, D) for 1-20 days from day 20. a: P<0.05, b: P<0.005, c: P<0.001 (Student’s t-test), d: P<0.001 (Cochran-Cox test). Androgens and Mouse Submandibular Gland 807 male one testosterone e ol/ fe) 109 mitotic rate of GCT cells (%) female C D e d c (6) 10 20 treatment periods (days) Fic. 5. Change in the mitotic rates of GCT of the submandibular glands of neonatally-castrated mice, treated with testosterone (A, C) and with DHT (B, D) for 1-20 days from day 20. a: P<0.05, b: P<0.01, c: P<0.01, d: P< 0.005 (Student’s t-test), e: P<0.05, f: P<0.01 (Cochran-Cox test). treatment with 100 ~g DHT, however, the mitotic rates were significantly increased after 2 days, reached a peak after 4 days, and then gradually decreased to the control levels after 10 days (Fig. 5B and 5D). Neither testosterone nor DHT had influence on the mitotic rates of the other three regions in the glands (data not shown). DISCUSSION In adult mice, the submandibular gland consists of acini, ID, GCT and SD. Among these struc- tures, the GCT develops more in the male than in the female. The mouse gland has been studied histologically, ultrastructurally and morphometri- cally during the pre- and postnatal development. Both sexes experience a similar morphogenesis of the gland during early postnatal development, and then the sexual differences occur at 3-4 weeks of age, when the GCT grows more rapidly in the male than in the female [8-11]. The mouse subman- dibular gland contains both the cytosolic and the nuclear androgen receptor [13-17]. The cytosolic androgen receptor in the glands of male mice increases during postnatal development and attains adult levels by day 20 [13, 16], while circulating testosterone levels begin to rise on day 20 [16]. In the present study, serum levels of testosterone and DHT dramatically increased be- tween days 20 and 30 (Fig. 1A and 1B). Adminis- tration of both androgens to neonatally-castrated mice for 20 days from day 20 caused to increase the gland weight and the ROA of GCT, the cell height of GCT and the mitotic rates of GCT, DHT being more effective than testosterone (Fig. 2-5). In our previous study, estradiol-178 did not induce any morphometrical changes in the mouse gland [11]. These results suggest that the masculine develop- 808 ment of the mouse gland is attributed to an in- crease in the circulating androgen levels during development. In the mouse gland, testosterone acts in its own right, but not as the precursor of DHT or other biologically active metabolites, be- cause it is not converted in vitro to its metabolic derivatives by the glands from 20- and 30-day-old mice (Sawada and Noumura, unpublished observations). In our male mice at 30 days of ages, circulating DHT levels were 1.0+0.21 ng/ml and were about 35% of the testosterone levels (Fig. 1B). In comparison with male rats at puberty reported [18], the DHT levels were approximately 4-fold over and the serum DHT/ testosterone ratio was approximately 2-fold higher in our male mice. Therefore, both testosterone and DHT may act independently on the masculine development of the mouse gland, but it is not yet solved which is a predominant androgen. Research is required to examine further the properties of the androgen receptor of the gland by using both androgens as ligands. In our previous study, administration of 100 ng DHT for 10 days from day 20 failed to increase the mitotic rates of GCT in neonatally castrated mice of both sexes [11]. In the present study, however, daily 100 ng DHT increased the mitotic rates of GCT between 2-7 days only, indicating a peak at 4 days, and then gradually declined to the control levels at 10 days (Fig. 5B and 5D). Therefore, it turned out that the observation after 10 days was not suitable for examining the effects of DHT on the mitotic activity in our previous study. In the present study, both testosterone and DHT were more effective in the males than in the females to induce increases in the gland weight and the mitotic activity of GCT (Fig. 2 and 5), corres- ponding to our previous study [11]. In the male gland, the amounts of androgen receptors (un- occupied cytosolic receptors plus occupied nuclear receptors) are approximately 3-fold over those in the female gland [17]. This may support our results that the androgen responsiveness of the mouse gland is sexually different. REFERENCES | Byyny, R. L., Orth, D. N. and Cohen, S. (1972) K. SAWADA AND T. NOuMURA Radioimmunoassay of epidermal growth factor. En- docrinology 90: 1261-1266. Roberts, M. L. (1974) Testosterone-induced accu- mulation of epidermal growth factor in the subman- dibular salivary glands of mice, assessed by radioim- munoassay. Biochem. Pharmacol. 23: 3305-3308. Michelakis, A. M., Yoshida, H., Menzie, J., Mura- kami, K. and Inagami, T. (1974) A radioimmunoas- say for the direct measurement of renin in mice and its application to submaxillary gland and kidney studies. Endocrinology 94: 1101-1105. Barka, T. (1980) Biologically active polypeptides in submandibular glands. J. Histochem. Cytochem. 28: 836-859. Walker, P., Weichsel, M. E. Jr., Hoath, S. B., Poland, R. E. and Fisher, D. A. (1981) Effect of thyroxine, testosterone, and corticosterone on nerve growth factor (NGF) and epidermal growth factor (EGF) concentrations in adult female mouse sub- maxillary gland: dissociation of NGF and EGF responses. Endocrinology 109: 582-587. Gresik, E. W., Schenkein, I., van der Noen, H. and Barka, T. (1981) Hormonal regulation of epidermal growth factor and protease in the submandibular gland of the adult mouse. Endocrinology 109: 924— 929. Kasayama, S., Yoshimura, M. and Oka, T. (1989) The regulation by thyroid hormones and androgen of epidermal growth factor synthesis in the subman- dibular gland and its plasma concentrations in mice. J. Endocr. 121: 269-275. Gresik, E. W. and MacRae, E. K. (1975) The postnatal development of the sexually dimorphic duct system and of amylase activity in the subman- dibular glands of mice. Cell Tissue Res. 157: 411- 422. Gresik, E. W. (1980) Postnatal developmental changes in submandibular glands of rats and mice. J. Histochem. Cytochem. 28: 860-870. Jayasinghe, N. R., Cope, G. H. and Jacob, S. (1990) Morphometric studies on the development and sex- ual dimorphism of the submandibular gland of the mouse. J. Anat. 172: 115-127. Sawada, K. and Noumura, T. (1991) Effects of castration and sex steroids on sexually dimorphic development of the mouse submandibular gland. Acta Anat. 140: 97-103. Sawada, K. and Noumura, T. (1992) Sexually dimorphic duct system of the submandibular gland in mouse with testicular feminization mutation (Tfm/Y). Acta Anat. 143: 241-245. Takuma, T., Nakamura, T., Hosoi, K. and Kume- gawa, M. (1977) Binding protein for Sa- dihydrotestosterone in mouse submandibular gland. Biochim. Biophys. Acta. 496: 175-181. Verhoeven, G. (1979) Androgen binding proteins in 15 16 Androgens and Mouse Submandibular Gland mouse submandibular gland. J. Steroid Biochem. 10: 129-138. Sato, N., Nemoto, T., Baba, R. and Ota, M. (1985) Dialysis-induced transformation of mouse subman- dibular androgen receptor. Biochem. Int. 10: 771- 776. Minetti, C. A. S. A., Valle, L. B. S., Fava-De- Moraes, F., Romaldini, J. H. and Oliveira-Filho, R. M. (1986) Ontogenesis of androgen receptors in the mouse submandibular gland: correlation with the developmental profiles of circulating thyroid and 17 18 809 testicular hormones. Acta Endocr. 112: 290-295. Kyakumoto, S., Kurokawa, R., Ohara-Nemoto, Y. and Ota, M. (1986) Sex difference in the cytosolic and nuclear distribution of androgen receptor in mouse submandibular gland. J. Endocr. 108: 267- iB. George, F. W., Johnson, L. and Wilson, J. D. (1989) The effect of a Sa-reductase inhibitor on androgen physiology in the immature male rat. Endocrinology 125: 2434-2438. i } i. a iat Ae a Hegde, AC coed lil - F ae wen oe Lae | 7 Sneath cri verre es IRD ANE AR RR ce |” | ti heey mache. de) celia vevtoety eas ; \ a Ao ME ohecl 1 telels Wile ence cao) ee et te WL Mae Tee Pitcairn i, ' poly edie eae a aie “habe ye een yer = ie if 11 VE) Ole Wiad eo nine ea Li@ he ae Pieces Ls ik ‘ q 7) Ay " laa r f] \ aul erty sl } i+ ie (ple t cane bite 4 ee ee i ie ete: sihaereert ' } 4 i abs Nees kh Fhe if Lihcwetsamy wal hel te» al Lie : ‘fs > AS abe = . \ 1% Loi ’ ral ri ' re » +) Oe ss cts ante a - oa \ 4 . — a, o 1 ss : 18) ZOOLOGICAL SCIENCE 9: 811-821 (1992) © 1992 Zoological Society of Japan Functional Morphology of Feeding Apparatus of the Cookie-Cutter Shark, /sistius brasiliensis (Elasmobranchii, Dalatiinae) SHIGERU SHIRAI’ and KAzuHIRO NAKAYA2 ‘Department of Zoology, University Museum, University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo 113, and ? Laboratory of Marine Zoology, Faculty of Fisheries, Hokkaido University, 3-1-1, Minato, Hakodate 041, Japan ABSTRACT— External morphology, skeletons, and muscles related to the peculiar feeding mechanism of Isistius brasiliensis are described and compared with other squaloid sharks. On the feeding function of this shark, a few characteristic conditions have been emphasized (e.g., fleshy oral margins, scoop-like lower jaw with a sharp tooth row, and developed basihyal and throat muscle). Serious morphological differences are, however, found in the whole of neurocranium, viscerocranium, shoulder girdle, and related musculature. These characteristic conditions are not suitable for the clutching-cutting feeding mechanism of typical squaloids but should effectively work on scooping the flesh of preys, making a strong oral vacuum, and rotating the body itself. These functions explain the cause of a crater-like wound on fish and cetaceans, and /sistius must apply them not only on large preys but on small fish and squids. INTRODUCTION The cookie-cutter shark, Isistius brasiliensis (Quoy et Gaimard), is a pelagic species of world- wide distribution throughout tropical and temper- ate waters. Among squaloid members, it is rather small reaching a size of 50cm TL. Recently this shark attracted many oceanological and fishery scientists because of its curious feeding habit re- vealed by Jones [1]. Isistius brasiliensis often feeds by biting chunks out of teleostean swimmers (e.¢., dolphinfish, marlin, tuna, albacore, and wahoo), cetaceans, seals, and large pelagic sharks including the megacasmids [1-5]. This small shark gives these preys a crater-like wound, which has been thought to be caused by the bacteria, parasites, or cyclostomes [6]. The purpose of this unusual feeding habit is ambiguous, because I[sistius’s sto- mach contents are found largely occupied with pelagic squids, as large as the body of this shark in most specimens, small fishes including Vinciguerria (Photichthyidae), and crustaceans [1, 7, 8]. In its Accepted April 13, 1992 Received February 14, 1992 feeding mechanism, we do not still understand enough; Jones [1] referred to only a few character- istic structures, which will be of help to scoop the flesh of prey and to induce the strong oral vacuum. The purpose of the present paper is to describe anatomical features of the feeding apparatus of Isistius brasiliensis in detail and to compare with those of closely related groups. Also, we discuss these functions on the morphological basis. MATERIALS AND METHODS Examination of skeletons and muscles of Jsistius brasiliensis was based on one male and four females. Several dalatiine species and other squaloids were dissected for comparison. Collec- tion information of these specimens is listed below. Study materials are deposited in HUMZ (Labora- tory of Marine Zoology, Faculty of Fisheries, Hokkaido University) and FSFL (Far Seas Fisher- ies Research Laboratory, Shimizu, Japan). Defini- tion of the Dalatiinae follows Compagno [5]. Dalatiinae TIsistius brasiliensis: one male—HUMZ 119287, 425mm TL; four females—HUMZ 69173, 364 812 S. SHIRAI AND K. NAKAYA mm TL, HUMZ 86585, 480mm TL, HUMZ 87120, 438 mm TL, HUMZ 89900, 516 mm TL. Squaliolus aliae: HUMZ 119285, male, 185 mm TL, S. laticaudus : two females—HUMZ 74972, 250 mm TL, HUMZ 74974, 245 mm TL. Euprotomicrus bispinatus: HUMZ 90413, male, 197 mm TL. Dalatias licha: two males—HUMZ 74603, 390 mm TL, HUMZ 95229, 621 mm TL. Other squaloids Deania quadrispinosum: FSFL-EE422, female, 325 mm TL. Etmopterus lucifer: HUMZ 90230, male, 337 mm TL. Oxynotus bruniensis: HUMZ 91383, male, 545 mm TL. Somniosus microcephalus : female, 1,425 mm TL. Squalus acanthias : HUMZ 87733, male, 495 mm TL. Zameus squamulosus : HUMZ 75872, male, 526 mm TL. HUMZ 112816, Figure drawings were made in the camera lucida image with dissected specimens stained by alcian blue. Anatomical terminology mainly follows Daniel [9] and Edgeworth [10]. Abbreviations used in figures are as follows: alc +—anterior upper labial cartilage am —adductor mandibulae ams —adductor m. superficialis bb —basibranchial bh —basihyal btp —basitrabecular process cb —ceratobranchial ch —ceratohyal chd —constrictor hyoideus dorsalis chy —constrictor hyoideus ventralis co —coracoid coa —coraco-arcuales cob —coraco-branchialis cod —constrictor dorsalis coh _—coraco-hyoideus es —eye stalk fbc —articular fossa of basihyal with ceratohyal fopp —foramen for ophthalmicus profundus gco —genio-coracoideus gcob —f-slip of genio-coracoideus gm —gum of upper jaw gop