Biological and Immunological Characterization of Multiple GnRH in an Opisthobranch Mollusk, Aplysia...

BMA

La*†a

A

GmitbGpbscsttaGGnaliTvl

dm8

General and Comparative Endocrinology 118, 77–89 (2000)doi:10.1006/gcen.2000.7457, available online at http://www.idealibrary.com on

0CA

iological and Immunological Characterization ofultiple GnRH in an Opisthobranch Mollusk,plysia californica

ihong Zhang,* Nancy L. Wayne,† Nancy M. Sherwood,‡ Hector R. Postigo,*nd Pei-San Tsai*,1

Department of Physiology and Neurobiology, University of Connecticut, Storrs, Connecticut 06269-4156;Department of Physiology, UCLA School of Medicine, Los Angeles, California 90095;nd ‡Biology Department, University of Victoria, Victoria, British Columbia V8W 3N5, Canada

ccepted December 22, 1999

wGGiwctwteapAdtcAGhGmA

H

onadotropin-releasing hormone (GnRH) is a neurohor-one central to the regulation of reproductive functions

n vertebrates. Recently, several studies have reportedhe presence of GnRH immunoreactivity (IR) in a num-er of mollusks, suggesting that the distribution ofnRH may not be restricted to Phylum Chordata. In theresent study, we extend our investigations to an opistho-ranch mollusk, Aplysia californica, to characterize theource, chemical nature, and biological activity of mollus-an GnRH-related molecules. Specific radioimmunoas-ays (RIAs) of various tissue extracts of Aplysia revealedhat only ovotestis, hemocytes, and hemolymph con-ained significant amounts of GnRH that crossreacts withntisera raised against tunicate-I (tI) and mammalian (m)nRH. Further RIAs and extractions revealed that thenRH-IR in the hemolymph is biochemically and immu-ologically distinct from the GnRH-IR in the hemocytesnd ovotestis. Using reverse-phase high-performanceiquid chromatography coupled with RIAs, the GnRH-IRn the hemolymph was resolved into two major peaks.he first peak eluted earlier than most known forms ofertebrate GnRH, and the later peak coeluted with m,amprey I, chicken II, and tI-GnRH. However, both peaks

1 To whom correspondence and reprint requests should be ad-ressed at Department of Environmental, Population and Organis-ic Biology, Campus Box 334, University of Colorado, Boulder, CO

t0309-0334.

77016-6480/00 $35.00opyright r 2000 by Academic Pressll rights of reproduction in any form reserved.

ere broad and may contain a heterogeneous mixture ofnRH-IR. Immunocytochemical study showed that tI-nRH-IR was present in the connective sheath surround-

ng the central nervous system, with a strong presence inhat appeared to be vascular space, again suggesting the

lose association between Aplysia GnRH-IR and circula-ion. Finally, treatment of the neuroendocrine bag cellsith chicken II GnRH significantly decreased the dura-

ion of the afterdischarge (AD, a characteristic pattern oflectrical firing in bag cell neurons) and the number ofction potentials fired during an AD, indicating theresence of a corresponding GnRH receptor in theplysia central nervous system. Overall, the resultsemonstrated the presence of multiple forms of GnRH-IRhat crossreact with tI- and mGnRH antisera in A.alifornica and the ability of a vertebrate GnRH to alterplysia neural activity. Together, these data suggest thatnRH may be a factor released by the ovotestis andemocytes into the circulation to alter neural functions.nRH-IR produced by the latter may serve as a novelediator of the neural and immune functions inplysia. r 2000 Academic Press

Key Words: Aplysia; GnRH; mollusk; invertebrate;PLC; evolution.

Gonadotropin-releasing hormone (GnRH) is central
o the initiation and maintenance of reproduction in

vfggfrPfPcnanHGepfm

GsmbaDcICdcoHrbavoantftambs

ite

78 Zhang et al.

CA

ertebrates. The mature GnRH decapeptide is releasedrom the axon terminals to stimulate the secretion ofonadotropins from the anterior pituitary. In turn, theonadotropins promote the development of gonadalunctions. To date, the GnRH family is composed of 13elated members (reviewed by King and Millar, 1995;owell et al., 1996; Jimenez-Linan et al., 1997; Carols-

eld et al., 2000). GnRH is found in all major groups ofhylum Chordata, including an invertebrate, the tuni-ate (Powell et al., 1996). In all cases, GnRH serves as aeuroendocrine regulator of reproductive functionsnd as a neurotransmitter/neuromodulator in theervous system (reviewed by King and Millar, 1995).owever, despite the extraordinary conservation ofnRH in chordates, until recently, there was very little

vidence to support the presence of GnRH-relatedeptides in other phyla. Consequently, the GnRH

amily of peptides were thought to be unique amongembers of Phylum Chordata.Recently, at least four studies have reported strongnRH immunoreactivity (IR) in the central nervous

ystem (CNS) of several species of mollusks. Theseollusks include two pulmonate gastropods (Gold-

erg et al., 1993; Young et al., 1999), two bivalves (Pazosnd Mathieu, 1999), and a cephalopod (Di Cosmo andi Cristo, 1998). The most comprehensive reports

ame from studies on a pulmonate, Helisoma trivolvis.n this freshwater snail, GnRH-IR was observed in theNS neurons and fibers that innervate the male repro-uctive tract. This GnRH-IR was immunologically andhromatographically similar to the mammalian formf GnRH (mGnRH), and the CNS extract isolated fromelisoma was capable of stimulating gonadotropin

elease from dispersed goldfish pituitary cells (Gold-erg et al., 1993; Young et al., 1999). In addition, thepplication of mGnRH to Helisoma CNS elicited di-erse electrophysiological and morphological effectsn individual neurons, including alterations of postsyn-ptic potentials and input resistance, and inhibition ofeurite outgrowth (Goldberg et al., 1993). The associa-

ion of Helisoma GnRH with CNS and reproductiveunctions suggests that the roles of GnRH as a reproduc-ive and neural regulator may have been evolution-rily conserved. The reproductive and neural involve-ent of the molluscan GnRH was further corroborated

y studies on the cephalopod and bivalves. These
tudies demonstrated a strong mitogenic action of c
opyright r 2000 by Academic Pressll rights of reproduction in any form reserved.

GnRH on gonadal cells (Pazos and Mathieu, 1999) anda spatial distribution of GnRH-IR in the CNS that ishighly suggestive of reproductive regulation (Di Cosmoand Di Cristo, 1998).

The results on GnRH bioactivity and immunoreactiv-ity from mollusks and an invertebrate chordate (thetunicates) suggest striking functional and structuralconservation of the GnRH family through evolution.Unfortunately, the precise structure of the nonchordateGnRH remains elusive. Nevertheless, the presence ofGnRH-IR in various invertebrates suggests that thepresence of an ancestral GnRH may have predated theemergence of vertebrates. If this is true, GnRH may bea very ancient molecule that is present throughoutvertebrate and invertebrate phyla.

In this study, we extend our investigations beyondpulmonate gastropods to an opisthobranch gastropod,Aplysia californica, to characterize the presence and thebiological activities of GnRH-related molecules. Tworeasons led us to the choice of this animal. First, wewish to examine whether the GnRH system observedin the pulmonates (Young et al., 1999) is conserved in adifferent subclass, the opisthobranchs. Second, the bagcell neurons of Aplysia are perhaps one of the mostwell-characterized systems of neuroendocrine neuronsknown to regulate molluscan reproductive functions.These neurons are functionally analogous to vertebrategonadotropes and release a peptide, egg-laying hor-mone (ELH), following a highly characteristic patternof electrical firing called the afterdischarge (AD; Kup-fermann and Kandel, 1970). ELH then acts on Aplysiaovotestis and CNS to induce ovulation and egg-layingbehavior (reviewed by Conn and Kaczmarek, 1989).The well-defined Aplysia reproductive axis shouldgreatly facilitate our effort in elucidating the relation-ship between GnRH and reproduction, allowing us togo beyond structure to biological functions.

In this study, we utilized a combination of radioim-munoassays (RIAs) and reverse-phase high-perfor-mance liquid chromatography (rp-HPLC) to immuno-logically and biochemically characterize the GnRH-IRin Aplysia. In addition, we utilized immunohistochem-stry (IHC) to localize GnRH-IR in the CNS. Finally, weested whether chordate forms of GnRH can alter thelectrical and secretory activities of the neuroendo-
rine bag cells. Unexpectedly, we found an unconven-

to

M

A

rSrafiwm

G

pImfAb

G

ficoctoWwvbrafinµa

cpcaifcf2

WecMS4nMcss

G

frrs22

pufa

H

0l(eUwt

GnRH in Aplysia californica 79

ional GnRH system with little similarity to the previ-usly described molluscan system.

ATERIALS AND METHODS

nimals

Wild-caught A. californica were purchased from Alac-ity Marine Supply (Redondo Beach, CA) and Marinepecimen Unlimited (San Francisco, CA). Upon ar-ival, animals were kept in 50-gallon tanks containingrtificial sea water (Instant Ocean) recirculated andltered through biological and chemical filters. Theater temperature was maintained at 15–18°C. Ani-als were fed Romaine lettuce daily.

nRH and General Reagents

Synthetic mGnRH, salmon GnRH (sGnRH), lam-rey I GnRH (lGnRH-I), and chicken II GnRH (cGnRH-

I) were purchased from Peninsula Laboratory (Bel-ont, CA). Tunicate I GnRH (tI-GnRH) was obtained

rom Dr. Jean Rivier (The Salk Institute, La Jolla, CA).ll HPLC reagents were purchased from Fisher (Pitts-urgh, PA).

nRH Extraction with Acid/Acetone

Tissues. GnRH extraction procedure was adoptedrom Kelsall et al. (1990). Briefly, tissues from Aplysia,ncluding the CNS, ovotestis, atrial gland, hepatopan-reas, and dorsal skin, were dissected and snap-frozenn dry ice. The frozen tissues were powdered with aold mortar and pestle held on dry ice. The powderedissues were extracted in acetone:1 N HCl (100:3, v/v)vernight on dry ice and cleared by filtration throughhatman no. 4 filters. The solids were reextractedith acetone:0.01 N HCl (4:1, v/v) in 40% of the

olume of the original extract and refiltered. Com-ined filtrates were extracted with petroleum ether at aatio of 4:1 (filtrate:petroleum ether, v/v) to removecetone, lipids, and other hydrophobic substances. Thenal aqueous phase was evaporated to complete dry-ess in a vacuum concentrator and reconstituted in 500l water. The tissue extracts were stored at 220°C until
ssayed by RIAs. m
Hemolymph. To collect hemolymph, animals wereooled in a sea water bath prechilled with ice packs torevent animals from inking during the hemolymphollection. A small cut was made at the base of the footnd the hemolymph was drained into a beaker contain-ng 100 µg/ml bacitracin. The hemolymph was centri-uged at 600g for 10 min to pellet the hemocytes. Theell-free hemolymph was precipitated overnight withour volumes of prechilled acetone:0.03 N HCl at

20°C and cleared by sequential filtrations throughhatman no. 4 and no. 1 filters. The filtrate was

xtracted with petroleum ether as described above andoncentrated using Sep-Pak C18 cartridges (Waters,

ilford, MA). Briefly, the extract was loaded onto aep-Pak cartridge with 0.01 M ammonium acetate, pH.6, washed with 20% acetonitrile in 0.01 M ammo-ium acetate, and eluted with 40% acetonitrile in 0.01

ammonium acetate. The eluate was evaporated toomplete dryness in a vacuum concentrator and recon-tituted in 500 µl water. The hemolymph extracts weretored at 220°C until assayed by RIAs.

nRH Extraction with Water

Hemocytes. Hemolymph was centrifuged at 600gor 10 min to pellet the hemocytes. The cell pellet wasinsed once with artificial sea water, centrifuged, andesuspended in water. Cell pellets were disrupted byonication and cleared by centrifugation at 14,000g for

min. The homogenate of hemocytes was stored at70°C until assayed by RIAs.Other tissues. CNS, ovotestis, atrial gland, hepato-

ancreas, and dorsal skin were sonicated in 10 vol-mes of water and cleared by centrifugation at 14,000g

or 2 min. The homogenates were stored at 270°C untilssayed by RIAs.

PLC Separation of Hemolymph Extract

The hemolymph extract was filtered through a.2-µm filter and injected through a 200-µl injectionoop into a Waters Symmetry C18 HPLC column3.9 3 150 mm, particle size 5 µm). The solvent gradi-nt was delivered by Waters 515 HPLC pumps and theV absorbance monitored by Waters 2487 dual-avelength detector. The mobile phase was 0.05%

rifluoroacetic acid in acetonitrile with a flow rate of 1
l/min. The separation program consisted of an
Copyright r 2000 by Academic PressAll rights of reproduction in any form reserved.

ifn1aitossise

G

f(au0carw2

maemmrDTdtsltt

I

E(

gpwcdentwepra1sp

I

hamcspfiTa8g1K

T

Ci

LSCCM

80 Zhang et al.

CA

nitial isocratic elution of 15% acetonitrile for 15 min,ollowed by a linear gradient to 23% acetonitrile in theext 5 min, an isocratic elution at 23% acetonitrile for0 min, followed by another linear gradient to 40%cetonitrile in 25 min, and an increase to 80% acetonitrilen 10 min. One-milliliter fractions were collected, concen-rated, reconstituted in water, and assayed for the presencef GnRH-IR by RIAs. With this separation program,ynthetic mGnRH, lGnRH-I, cGnRH-II, tI-GnRH, andGnRH could be resolved in this order. Each samplenjection was preceded by a blank run that wasubsequently assayed for the presence of GnRH tonsure that column contamination did not occur.

nRH Iodination and RIAs

Radioiodination of GnRHs and RIAs were per-ormed according to the protocol previously describedTsai and Licht, 1993) with some modifications. Iodin-ted GnRHs were purified by Sep-Pak C18 cartridgessing a step gradient of 0, 20, and 30% acetonitrile in.01 M ammonium acetate, pH 4.6 (for tI-GnRH andGnRH-II), or a step gradient of 0, 20, and 25%cetonitrile (for mGnRH). Under these conditions,adiolabeled GnRHs eluted at 25 or 30% acetonitrile,hereas free 125I2 and lactoperoxidase eluted at 0 and

0% acetonitrile, respectively.Four GnRH RIAs were used to detect tI-GnRH,GnRH, and cGnRH-II. The tI-GnRH RIA utilized an

ntiserum generated against tI-GnRH (81CR-3; Powellt al., 1996) and iodinated tI-GnRH as the tracer. TheGnRH RIAs utilized two antisera generated againstGnRH (R1245, obtained from Dr. Terry Nett, Colo-

ado State University; and Antibody DJC, obtained fromr. David D. Chase) and iodinated mGnRH as the tracer.he cGnRH-II RIAutilized an antiserum generated againstogfish GnRH (7CR-10) and iodinated cGnRH-II as the

racer. Antiserum 7CR-10 was found to crossreact moretrongly to cGnRH-II than to dogfish GnRH (unpub-ished data). RIA crossreactivities of these four antiserao seven known forms of GnRH have been tested inhis study and are listed in Table 1.

n Vitro Bag Cell Recording and Perfusion

In vitro monitoring of bag cell electrical activities andLH secretion was performed as previously described

Wayne and Wong, 1994). Briefly, excised abdominal 7


anglia including bag cell clusters were placed in aerfusion chamber. Synthetic tI-GnRH and cGnRH-IIere administered by a peristaltic pump through the

annulated artery to facilitate hormone delivery. After-ischarges were stimulated and recorded with suctionlectrodes placed on the pleurovisceral connectiveerve and bag cell clusters, respectively. Action poten-

ials were monitored by a Gould bioelectric amplifierith a chart recorder. Since bag cell neurons are

lectrotonically coupled, extracellular recording of com-ound action potentials from the cluster gives accurateepresentation of individual bag cell activity (Dudeknd Blankenship, 1977; Haskins and Blankenship,979). Perifusates were collected every 5 min andtored at 220°C. ELH RIA was carried out using therocedure described by Wayne and Wong (1994).

HC

Tissues were immersion-fixed in 4% paraformalde-yde in 0.1 M phosphate-buffered saline (PBS) for 6 hnd cryoprotected in 25% sucrose for 24 h. Sixteen-icrometer frozen coronal sections were cut using a

ryostat and thawed-mounted onto gelatin-coatedlides. Thus, each section represents an anterior–osterior cut. Sections were incubated with 0.5% H2O2

or 5 min to quench the endogenous peroxidase activ-ty, washed five times with PBS containing 0.4%riton X (PBST), and incubated with various GnRHntisera for 48 h at 4°C. These antisera included1CR-3 (anti-tI-GnRH; 1:1000), LR-1 (anti-mGnRH; aift from Dr. R. Benoit, McGill University, Canada;:10,000), 675 (anti-cGnRH-II; a gift from Dr. Judying, University of Cape Town, South Africa; 1:1000),

ABLE 1

rossreactivities of GnRH Antisera with Various Forms of GnRHn Radioimmunoassays

81CR-3 DJC R1245 7CR-10

amprey I 0.75% 0.1% 0.1% 6.00%almon ,0.03% ,0.03% 0.2% ,0.03%hicken I ,0.03% 7.8% 2.7% ,0.03%hicken II ,0.03% ,0.03% 1.4% 100%ammalian ,0.03% 100% 100% ,0.03%

Tunicate I 100% ,0.03% ,0.03% ,0.03%Tunicate II 0.1% ,0.03% ,0.03% ,0.03%

CR-10 (anti-cGnRH-II; 1:1000), 3952 (anti-lamprey-III-

GHSgtpaiiibhwLtecmcb

R

CE

pgcsshatcGo(tWoa(tf

hmacbRsccoDobfttlof

r

GwrlRGGG2f

T

BA

OHH

so


nRH; a gift from Dr. Stacia Sower, University of Newampshire; 1:1000), and Sigma Ab (anti-mGnRH;

igma, St. Louis, MO; 1:50). All these antisera wereenerated with GnRH conjugated to either bovinehyroglobulin or bovine serum albumin; keyhole lim-et hemocyanin was not involved in the generation ofny of these antisera. All antisera were diluted in PBSTn the presence of 4% normal donkey serum. Afterncubation, sections were given three 5-min rinses andncubated with a biotinylated donkey anti-rabbit anti-ody (Jackson ImmunoResearch, West Grove, PA) for 1

at room temperature. Sections were subsequentlyashed and incubated with the ABC reagents (Vectoraboratories, Burlingame, CA) for 1 h at room tempera-

ure, washed three times, and visualized for the pres-nce of GnRH-IR using diaminobenzidine (DAB) as ahromagen. Some sections were counterstained for 10in with hematoxylin to identify the nuclei. Negative

ontrols included sections incubated with normal rab-it serum in place of the primary antiserum.

ESULTS

haracterization of GnRH-IR in Aplysia Tissuextracts by RIAs

The tI-GnRH RIA was initially utilized to detect theresence of GnRH-IR in the CNS, ovotestis, atrialland, hepatopancreas, skin, hemolymph, and hemo-ytes of Aplysia. Among acid/acetone extracts of tis-ues, only hemolymph contained tI-GnRH-IR. A sub-tantial amount of tI-GnRH-IR was present in theemolymph extract, averaging 0.51 6 0.13 ng pernimal (n 5 10; Table 2). After performing water extrac-ions of the CNS, ovotestis, atrial gland, hepatopan-reas, skin, and hemocytes, we found additional tI-nRH-IR in the ovotestis and hemocytes. The amountsf GnRH-IR in the ovotestis (n 5 7) and hemocytesn 5 11) were approximately 130- and 5-fold higherhan that in the hemolymph, respectively (Table 2).

hen ovotestis and hemocyte extracts were loadednto C18 Sep-Pak cartridges, the GnRH-IR failed todsorb to the column and eluted in the buffer washdata not shown). Thus, the tI-GnRH-IRs in the ovotes-is and hemocytes are biochemically distinct from that
ound in the hemolymph. (
The GnRH-IRs in the ovotestis, hemocytes, andemolymph were further characterized by displace-ent bindings in four RIAs described under Materials

nd Methods (Fig. 1). Neither hemolymph nor hemo-yte extracts competed with the cGnRH-II tracer ininding to 7CR-10 (Fig. 1B; ovotestis not tested in thisIA). However, considerable parallelism was ob-erved in the binding of all three extracts to 81CR-3ompared to synthetic tI-GnRH (Fig. 1A) and to R1245ompared to synthetic mGnRH (Fig. 1C). When an-ther antiserum generated against mGnRH (AntibodyJC) was used in the RIA, different results werebtained from these three extracts (Fig. 1D). Whereasoth hemocyte and ovotestis extracts competed success-ully in the RIA binding in a fashion seemingly parallelo that of synthetic mGnRH, hemolymph extract failedo compete (Fig. 1D). These tests revealed that immuno-ogical characteristics of GnRH-IR in hemocytes andvotestis are identical to each other but are distinctrom the hemolymph.

p-HPLC Analysis of GnRH-IR in the Hemolymph

Since the hemolymph extract contained the onlynRH-IR that could be retained on the C18 column,e characterized this GnRH-IR using a combination of

p-HPLC and RIAs. As the crude extracts of hemo-ymph contain GnRH-IR that could be detected inIAs using R1245 (anti-mGnRH) and 81CR-3 (anti-tI-nRH), these two RIAs were again used to detectnRH-IR in HPLC eluates. Two prominent peaks ofnRH-IR were observed using the tI-GnRH RIA (Fig.

A). The first peak eluted earlier than most knownorms of vertebrate GnRH, between 12 and 16 min

ABLE 2

iochemical Characteristics and Distribution of GnRH-IR inplysia Tissues

TissuesAcid/acetone

extractableRetainedon C18

tI-GnRH-IR(ng/animal)

votestis No No 64.83 6 18.16emocytes No No 2.61 6 0.52a

emolymph Yes Yes 0.51 6 0.13a

a Hemocyte and hemolymph GnRH contents may represent alight underestimation of the true amount since about 10% (or less)f the fluid still remained in the animal after bleeding.

Fig. 2A). The second peak eluted between 18 and 22


F

tacamsrc

82 Zhang et al.

CA

IG. 1. Representative displacement curves of ovotestis, hemocyte, and hemolymph extracts in various RIAs. (A) RIA displacement curves ofhree tissue extracts and synthetic tI-GnRH in the tI-GnRH RIA (antiserum, 81CR-3; tracer, iodinated tI-GnRH). Note that displacement curves ofll three extracts are parallel to that of synthetic tI-GnRH. (B) RIA displacement curves of two tissue extracts and synthetic cGnRH-II in theGnRH-II RIA (antiserum, 7CR10; tracer, iodinated cGnRH-II). Both hemocyte and hemolymph extracts failed to compete in the binding to thentibody. Ovotestis extract was not tested in this RIA. (C, D) RIA displacement curves of three tissue extracts and synthetic mGnRH in twoGnRH RIAs (antisera, R1245 and DJC; tracer, iodinated mGnRH). Note that displacement curves of all three extracts are parallel to that of

ynthetic mGnRH in the RIA using R1245 (C), but the hemolymph extract failed to compete in the binding to antibody DJC (D). Each pointepresents the mean of duplicate determinations. The same experiment has been performed three times, and, each time, identical displacement
urves were observed.

mIG(a

eRew

F

8mg


in, at elution positions of mGnRH, lGnRH-I, cGnRH-I, and tI-GnRH. The elution profile of this laternRH-IR was broad and cannot be resolved precisely

Fig. 2A). Regardless, both peaks were relatively broad

IG. 2. HPLC elution profile of GnRH-IR in Aplysia hemolymph e1CR-3 and (B) the mGnRH RIA using antiserum R1245. AcetonitriGnRH (m), lGnRH-I (l-I), cGnRH-II (c-II), tI-GnRH (tI), and sGnRH

enerated from 1 L of hemolymph pooled from 10 animals.

nd may contain a heterogeneous mixture of mol- e

cules. Using a mGnRH RIA system with antiserum1245, three peaks were detected (Fig. 2B). The earlyluting peak previously detected by the tI-GnRH RIAas now resolved into two narrower peaks, one

GnRH-IR was measured by (A) the tI-GnRH-RIA using antiserumient is indicated by the dashed lines. Elution positions of synthetice indicated by solid triangles at top of each graph. The extract was

xtract.le grad

(s) ar

luting at 11 min and one at 14 min. The second peak


rpq

L

twsLnanitngsogtctoasper

F

fain(tsGndiSacgp

84 Zhang et al.

CA

emained broad and had the same elution profile asreviously detected by the tI-GnRH RIA. The relativeuantity of GnRH-IR varied with the RIAs used.

ocalization of GnRH-IR in the CNS by IHC

The inability of our RIAs to detect any GnRH-IR inhe CNS was unexpected. To ensure that Aplysia CNS

as truly devoid of GnRH-IR, we performed IHCtudies on the CNS with six polyclonal antisera (675,R-1, 7CR-10, 3952, 81CR-3, and Sigma Ab). No immu-ostaining was observed in the Aplysia CNS with fiventisera (675, LR-1, 7CR-10, 3952, and Sigma Ab; dataot shown). However, using 81CR-3 (anti-tI-GnRH),

ntense immunostaining was observed in the connec-ive sheath (fibrous sheath lining neuron clusters anderve bundles), but not neuron cell bodies, of allanglia. Since the patterns of staining are highlyimilar in all ganglia, we present photomicrographs ofnly two representative ganglia (pedal and abdominalanglia) in Fig. 3. As mentioned before, the neuronshemselves were never stained (Figs. 3A and 3B). Aloser examination of the immunostaining revealedhat the majority of GnRH-IR was associated withpen space in the connective sheath (Fig. 3) and someppeared acellular (Figs. 3A and 3C). Frequently, thetaining was immediately adjacent to and outlinedarts of the neuron clusters (Figs. 3A and 3B). How-ver, some stainings were observed distal to the neu-on clusters in the open space of connective sheath

IG. 3. GnRH IHC of two representative Aplysia ganglia. Sectionsrom (A) pedal and (B, C) abdominal ganglia were stained withntiserum 81CR-3 (anti-tI-GnRH). Dark brown color indicates DABmmunoperoxidase reaction; deep blue color indicates hematoxylinuclear staining. Neurons are large cells with prominent nuclei150–500 µm, some indicated by arrowheads in A). Fibrous connec-ive tissue surrounds neuron cell bodies and contains numerousmall nonneuronal cells (10–20 µm). (A) There was abundantnRH-IR in the open space of the connective tissue surrounding theeuron cluster, but neuron cell bodies (arrowheads) were completelyevoid of GnRH-IR. Arrows indicate acellular staining. (B) Strong

mmunostaining could be found to outline the neuron clusters. (C)trong immunostaining was also found in large open space, presum-bly the vascular space, in the connective sheath distal to the neuronlusters. Arrows indicate acellular staining. Black bar, 250 µm. Otheranglia (cerebral, buccal, and pleural) displayed similar staining
atterns (data not shown).

(tn

Ac

iwaccotaasR5sbnEa(

D

GlApbawnaphttdmc

(Pacol

F

snatt


Fig. 3C). This pattern of staining is highly representa-ive of all five ganglia (pleural, pedal, buccal, abdomi-al, and cerebral) that we have processed for IHC.

lteration of Bag Cell Neuron Electrical Activity byGnRH-II

Since bag cell neurons of Aplysia produce ELH, anmportant peptide regulator of reproduction, we tested

hether vertebrate GnRH could alter bag cell electricalctivities and ELH release. Application of 100 nMGnRH-II to bag cell preparations significantly de-reased the duration of the AD (a characteristic patternf compound action potentials fired by bag cells torigger ELH secretion; Fig. 4A) and the number ofction potentials fired during an AD (Fig. 4B). The totalmount of ELH secreted during an AD was notignificantly altered by the GnRH treatment (Fig. 4C).epresentative tracings of AD fired in the absence (Fig.A) and presence (Fig. 5B) of 100 nM cGnRH-II arehown. We also tested the ability of tI-GnRH to alterag cell electrical activity. The duration of AD, theumber of action potentials fired, and the amount ofLH secreted were not significantly altered by thedministration of 100 (n 5 3), 300 (n 5 3), and 600n 5 4) nM of tI-GnRH (data not shown).

ISCUSSION

In the present study, we describe the presence ofnRH-IRs in the ovotestis, hemocytes, and hemo-

ymph of an opisthobranch mollusk, A. californica. Theplysia GnRH system is strikingly different from thereviously described molluscan GnRH systems (Gold-erg et al., 1993; Di Cosmo and Di Cristo, 1998; Pazosnd Mathieu, 1999; Young et al., 1999) in three uniqueays. First, GnRH-IR was not detected in Aplysia

eurons. Second, there is a close tie between GnRH-IRnd the circulatory component of Aplysia. Third, theresumed GnRH source tissues, the ovotestis andemocytes, contained a GnRH-IR that cannot be ex-

racted and purified by conventional techniques usedo isolate chordate forms of GnRH. Together, theseata suggest the presence of a novel GnRH system in aollusk that may be biochemically and physiologi-

ally distinct from the chordate GnRH system. r

In striking contrast with other molluscan reportsGoldberg et al., 1993; Di Cosmo and Di Cristo, 1998;azos and Mathieu, 1999; Young et al., 1999), our RIAsnd IHC failed to detect any GnRH-IR in the neuronalomponent of Aplysia CNS. Given the strong presencef GnRH in the CNS of diverse chordates and mol-

usks, this lack of GnRH-IR in Aplysia CNS probably

IG. 4. Perfusion of bag cell clusters with 100 nM cGnRH-IIignificantly decreased (A) the duration of AD (P , 0.01) and (B) theumber of action potentials fired during an AD (P 5 0.02) withoutltering (C) the total amount of ELH secreted. Each bar representshe mean 6 SEM, n 5 6. Statistical significance was calculated by thewo-tailed Mann–Whitney u test.

eflects our inability to detect GnRH-IR with the existing


aAstcC

vostsgspraoG

b

li

F

ecc

86 Zhang et al.

CA

ntisera rather than the complete absence of GnRH inplysia neurons. Regardless, our IHC results contrasted

harply with previously published studies reportinghe presence of mGnRH and cGnRH-II-IR in mollus-an neurons (Goldberg et al., 1993; Di Cosmo and Diristo, 1998; Pazos and Mathieu, 1999; Young et al., 1999).A closer examination of the CNS immunostaining re-

ealed that the majority of GnRH was associated withpen space in the connective sheath, presumably vascularpace draining the neurons (Fig. 3). Some immunoreac-ivity appeared acellular. At present, the identity of thetructure containing GnRH-IR is not known. However,iven its location and pattern, we believe that thesetructures may be associated with blood vessels thatenetrate into the connective sheath. If this is true, thisesult would indicate that GnRH-IR in Aplysia may ben intimate component of the circulation, consistent withur RIA data demonstrating the presence of tI-nRH-IR in the hemolymph and hemocytes.Several conclusions may be drawn from the com-

IG. 5. Tracings of AD fired by bag cell neurons in the (A) abslectrotonically coupled, each vertical spike represents a compoundGnRH-II shortened the duration of AD. Data were chosen from exorner corresponds to a 5-s period.

ined HPLC–RIA results obtained from the hemo-


ymph extract (Fig. 2). First, multiple forms of circulat-ng GnRH-IR are present in Aplysia. Second, most

known forms of GnRH can be excluded from the earlyeluting peak, with the possible exception of lamprey IIIGnRH (lGnRH-III), also an early eluting GnRH (Soweret al., 1993). Third, the elution position of the secondGnRH-IR peak coincided with four forms of chordateGnRH and was consistently detected by mGnRH andtI-GnRH RIAs. However, we do not believe that thispeak contained either mGnRH or tI-GnRH, since thesetwo forms crossreacted minimally in each other’sRIAs. In other words, for the tI-GnRH RIA to detectauthentic mGnRH, the level of mGnRH must begreater than 30 ng/fraction, but this was clearly not thecase (Fig. 2B). The reverse is also true for tI-GnRH. Inaddition, the second peak should not contain cGnRH-IIbecause no GnRH-IR in the hemolymph extract couldbe detected by the cGnRH-II RIA (Fig. 1B). lGnRH-Iremains as a possibility, but again given the specifici-ties of mGnRH and tI-GnRH RIAs, the likelihood is

nd (B) presence of 100 nM cGnRH-II. Since bag cell neurons arepotential fired by all bag cell neurons in the cluster. The addition ofntal bag cell clusters used in Fig. 4. The black bar at the lower left

ence aactionperime

small. Overall, the lack of a perfect match between the

egnWssr

ocGbGsgkli(imcedaTaGao1gfla

cdtrhfssnbht

vpcF(vtip1

castaifmperG(wf1oAdfinhsirc


lution time and the immunological properties sug-ests that the GnRH-IR in the hemolymph is probablyot structurally identical to the known forms of GnRH.hether these GnRH-IRs resolved by rp-HPLC repre-

ent multiple cleavage or degradation products of aingle molecule or true products of different genesemains to be investigated.

It is interesting to note that the GnRH-IR in thevotestis and hemocytes is immunologically and bio-hemically distinct from that in the hemolymph.nRH-IR in ovotestis and hemocytes was extractabley water but not by acid/acetone, suggesting that thisnRH-IR may be a large molecule, possibly a precur-

or, that has denatured and precipitated during or-anic extractions. A large GnRH precursor of 26–28Da was previously described in the rodent hypotha-

amic extract, demonstrating the possibility of recover-ng such a large GnRH precursor from tissue extractsMillar et al., 1977; Gautron et al., 1981). Also, GnRH-IRn ovotestis and hemocytes failed to adsorb to the C18

atrix. This observation parallels findings in a hemi-hordate in which a large amount of GnRH-IR in tissuextracts of Saccoglossus bromophenolosus was eluteduring the washout period in rp-HPLC without inter-cting with the C18 column (Cameron et al., 1999).hus, invertebrates may possess variants of GnRH thatre far more hydrophilic than the vertebrate forms ofnRH. We hypothesize that extensive proteolytic cleav-

ge of the larger, more hydrophilic GnRH-IR may haveccurred extracellularly by proteases (Squire et al.,991; Bawab et al., 1992, 1993; Zappulla et al., 1999) toenerate a shorter, acid/acetone-extractable GnRH-IRound in the hemolymph. However, whether thisarger, more hydrophilic GnRH-IR is a true precursornd, if so, how it is processed for release is still unclear.Two tissues in Aplysia, hemocytes and ovotestis,

ontained significant amounts of GnRH (Table 2). Weid not conduct IHC on these two tissues to investigate

he cellular details of the peptide source; however, theelatively high levels of GnRH-IR in the ovotestis andemocytes suggest that these two are good candidatesor source tissues, unless an unconventional transportystem was involved to deliver GnRH from otherources to these two tissues. Further IHC studies areeeded to localize GnRH-immunopositive cell types inoth the ovotestis and the hemocytes. Our finding thatemocytes and ovotestis contained GnRH-IR, al-

hough unusual, is not entirely unconventional. In

ertebrates, numerous studies have demonstrated theresence of GnRH in gonads, placenta, and immuneells (see King and Millar, 1995; Marchetti et al., 1995).or example, GnRH messenger ribonucleic acidmRNA) and GnRH-like molecules were found in theertebrate immune system, including rat splenic andhymic lymphocytes, peripheral T-lymphocytes, themmune mast cells in the dove brain, and humaneripheral blood mononuclear cells (Emanuele et al.,990; Azad et al., 1991, 1993; Zhuang et al., 1993; Wilson

et al., 1995; Chen et al., 1999). Molluscan hemocytes, aomponent of the immune system (Stefano et al., 1989),re a rich source of neuropeptides, including neuroten-in, oxytocin, somatostatin, substance P, opioid pep-ides, and many others (Stefano et al., 1989; Ottavianind Cossarizza, 1990). Hence, it is possible that GnRHs produced by Aplysia hemocytes to regulate CNSunctions, thus serving as a novel link between im-

une and neural functions. On the other hand, theresence of gonadal GnRH in Aplysia suggests anxtraordinary functional conservation of GnRH ineproduction. In bivalve mollusks, vertebrate forms ofnRH potently stimulated gonadal DNA synthesis

Pazos and Mathieu, 1999). In addition, in two freshater snails, GnRH immunoreactive nerve fibers were

ound to innervate the penial complex (Young et al.,999). All these findings suggest a direct involvementf GnRH in molluscan reproduction. We did not testplysia penial complex in our RIAs because of theifficulty that we encountered in extracting such abrous tissue. However, we have conducted prelimi-ary IHC using the tunicate GnRH antiserum andave found no GnRH-IR in the penis proper (data nothown). A more detailed characterization of GnRH-IRn the peripheral reproductive organs by IHC is cur-ently underway. Although the physiological signifi-ance of gonadal GnRH-IR in Aplysia is presently

unclear, it is tantalizing to speculate that GnRH is asignaling molecule that coordinates bag cell activitiesand ovulation or a paracrine regulator of gonadalfunctions.

The application of 100 nM cGnRH-II to Aplysia bagcell preparations significantly decreased the durationof the AD and the number of action potentials firedduring an AD (Fig. 4). We tested cGnRH-II because it isthought to be the most ancient form of GnRH, isubiquitously found in all jawed vertebrates, and is
biologically active in diverse chordates (see King and

MdbbbTberpattrieG

IceaiaaFnphGlaii

cs(msihnfbimf

alr

AmCYbavptoeMis

A

aI

R

A

A

B

B

C

C

88 Zhang et al.

CA

illar, 1995). Since there was no immunological evi-ence for the presence of cGnRH-II in Aplysia, weelieve that it was acting as a hormonal agonist,inding to a receptor whose endogenous ligand maye structurally similar, but not identical, to cGnRH-II.he ability of bag cell neurons to respond to a verte-rate form of GnRH suggests the presence of anndogenous GnRH-like molecule and a correspondingeceptor. This possibility was further supported by arevious study demonstrating the ability of mGnRHnd its agonist to increase the firing rates of neurosecre-ory white cells in Aplysia (Seaman et al., 1980). Al-hough cGnRH-II did not significantly decrease ELHelease in the current study (Fig. 4C), the large variabil-ty and small sample size suggest that additionalxperiments are needed to conclusively rule out anRH effect on ELH secretion.At present, the cellular events underlying the cGnRH-

I-induced shortening of AD and the functional signifi-ance of this phenomenon are unclear. The possibilityxists that cGnRH-II may act as a hyperpolarizinggent to decrease bag cell membrane excitability, lead-ng to the early termination of AD. A similar mode ofction has been described in an Aplysia neuropeptide,lpha bag cell peptide (Redman and Berry, 1991).urther investigations using intracellular recording iseeded to verify whether cGnRH-II acts through thisarticular mechanism. It is interesting to note that thisypothesis is consistent with the reported mode ofnRH action in vertebrates. Elevated IP3 and intracel-

ular Ca21, both intracellular changes induced by thectivation of vertebrate GnRH receptor, are capable ofnducing transient hyperpolarization (Fink et al., 1988)n bag cell neurons.

The functional significance of decreased AD is diffi-ult to interpret, but several possibilities exist. First,ince ELH synthesis was shown to be triggered by ADLee and Wayne, 1998), an inhibition of bag cell

embrane excitability and AD may attenuate ELHynthesis. Therefore, GnRH in Aplysia may act as annhibitory regulator of long-term ELH synthesis and,ence, its availability for release. Second, the bag celleurons enter a refractory period of up to 20 hollowing an AD, during which no additional AD cane triggered (Kupfermann and Kandel, 1970). By inhib-

ting membrane excitability of these neurons, GnRHay prolong the refractory period, resulting in less

requent egg laying. Regardless of these possibilities,


dditional functional data at cellular and organismalevels are required to better ascertain the functionalole of GnRH in Aplysia.

In conclusion, we have identified a GnRH system inplysia that is distinct from the previously describedolluscan GnRH systems (Goldberg et al., 1993; Diosmo and Di Cristo, 1998; Pazos and Mathieu, 1999;oung et al., 1999). The Aplysia GnRH-IR is detectabley antisera against tI- and mGnRH, and is closelyssociated with the ovotestis and the blood. Further, aertebrate GnRH is biologically active in altering thehysiology of reproductive neuroendocrine cells. Taken

ogether, these data are in accord with a growing bodyf evidence suggesting that GnRH is an ancient mol-cule that arose before the emergence of chordates.oreover, the functions of GnRH in reproductive,

mmune, and neural regulation may have been exten-ively conserved.

CKNOWLEDGMENTS

We thank Steven Hsu and Lu Pan for technical assistance in IHCnd biochemical purification. This work was supported by NSFBN-9733237.

EFERENCES

zad, N., Emanuele, N. V., Halloran, M. M., Tentler, J., and Kelley,M. R. (1991). Presence of luteinizing hormone-releasing hormone(LHRH) mRNA in rat spleen lymphocytes. Endocrinology 128,1679–1681.zad, N., La Paglia, N., Jurgens, K. A., Kirsteins, L., Emanuele, N. V.,Kelley, M. R., Lawrence, A. M., and Mohagheghpour, N. (1993).Immunoactivation enhances the concentration of luteinizing hor-mone-releasing hormone peptide and its gene expression inhuman peripheral T-lymphocytes. Endocrinology 133, 215–223.

awab, W., Querido, E., Crine, P., and DesGroseillers, L. (1992).Identification and characterization of aminopeptidases from Aply-sia californica. Biochem. J. 286, 967–975.

awab, W., Aloyz, R. S., Crine, P., Roques, B. P., and DesGroseillers,L. (1993). Identification and characterization of a neutral endopep-tidase activity in Aplysia californica. Biochem. J. 296, 459–465.

ameron, C. B., Mackie, G. O., Powell, J. F., Lescheid, D. W., andSherwood, N. M. (1999). Gonadotropin-releasing hormone inMulberry cells of Saccoglossus and Ptychodera (Hemichordata:Enteropneusta). Gen. Comp. Endocrinol. 114, 2–10.

arolsfeld, J., Powell, J. F. F., Park, M., Fischer, W. H., Craig, A. G.,
Chang, J. P., Rivier, J. E., and Sherwood, N. M. (2000). Primary

C

C

D

D

E

F

G

G

H

J

K

K

K

L

M

M

O

P

P

R

S

S

S

S

T

W

W

Y

Z

Z


structure and function of three gonadotropin-releasing hormones,including a novel form, from an ancient teleost, herring. Endocrinol-ogy 141, 505–512.

hen, H. F., Jeung, E. B., Stephenson, M., and Leung, P. C. (1999).Human peripheral blood mononuclear cells express gonadotropin-releasing hormone (GnRH), GnRH receptor, and interleukin-2receptor gamma-chain messenger ribonucleic acids that are regu-lated by GnRH in vitro. J. Clin. Endocrinol. Metab. 84, 743–750.

onn, P. J., and Kaczmarek, L. K. (1989). The bag cell neurons ofAplysia. Mol. Neurobol. 3, 237–273.i Cosmo, A. D., and Di Cristo, C. D. (1998). Neuropeptidergiccontrol of the optic gland of octopus vulgaris: FMRF-amide andGnRH immunoreactivity. J. Comp. Neurol. 398, 1–12.udek, F. E., and Blankenship, J. E. (1977). Neuroendocrine cells ofAplysia brasiliana. I. Bag cell action potentials and afterdischarge. J.Neurophysiol. 40, 1301–1311.

manuele, N. V., Emanuele, M. A., Tentler, J., Kirsteins, L., Azad, N.,and Lawrence, A. M. (1990). Rat spleen lymphocytes contain animmunoactive and bioactive luteinizing hormone-releasing hor-mone. Endocrinology 126, 2482–2486.

ink, L. A., Connor, J. A., and Kaczmarek, L. K. (1988). Inositoltrisphosphate releases intracellularly stored calcium and modu-lates ion channels in molluscan neurons. J. Neurosci. 8, 2544–2555.autron, J. P., Pattou, E., and Kordon, C. (1981). Occurrence of highermolecular forms of LHRH in fractionated extracts from rathypothalamus, cortex and placenta. Mol. Cell. Endocrinol. 24, 1–15.oldberg, J. I., Garofalo, R., Price, C. J., and Chang, J. P. (1993).Presence and biological activity of a GnRH-like factor in thenervous system of Helisoma trivolvis. J. Comp. Neurol. 336, 571–582.askins, J. T., and Blankenship, J. E. (1979). Interactions betweenbilateral clusters of neuroendocrine cells in Aplysia. J. Neurophysiol.42, 356–367.

imenez-Linan, M., Rubin, B. S., and King, J. C. (1997). Examinationof guinea pig luteinizing hormone-releasing hormone gene re-veals a unique decapeptide and existence of two transcripts in thebrain. Endocrinology 138, 4123–4130.

elsall, R., Coe, I. R., and Sherwood, N. M. (1990). Phylogeny andontogeny of gonadotropin-releasing hormone: Comparison ofguinea pig, rat, and a protochordate. Gen. Comp. Endocrinol. 78,479–494.

ing, J. A., and Millar, R. P. (1995). Evolutionary aspects of gonadotropin-releasing hormone and its receptor. Cell Mol. Neurobiol. 15, 5–23.

upfermann, I., and Kandel, E. R. (1970). Electrophysiologicalproperties and functional interconnections of two symmetricalneurosecretory clusters (bag cells) in abdominal ganglion ofAplysia. J. Neurophysiol. 33, 865–876.

ee, W., and Wayne, N. L. (1998). The roles of transcription andtranslation in mediating the effect of electrical afterdischarge onneurohormone synthesis in Aplysia bag cell neurons. Endocrinology139, 5109–5115.archetti, B., Morale, M. C., Gallo, F., Batticane, N., Farinella, Z., andCioni, M. (1995). Neuroendocrineimmunology (NEI) at the turn ofthe century: Towards a molecular understanding of basic mecha-nisms and implications for reproductive physiolopathology. Endo-
crine 3, 845–861.
illar, R. P., Aehnelt, C., and Rossier, G. (1977). Higher molecularweight immunoreactive species of luteinizing hormone releasinghormone: Possible precursors of the hormone. Biochem. Biophys.Res. Commun. 74, 720–731.ttaviani, E., and Cossarizza, A. (1990). Immunocytochemical evi-dence of vertebrate bioactive peptide-like molecules in the im-muno cell types of the freshwater snail Planorbarius corneus (L.)(Gastropoda, Pulmonata). FEBS Lett. 267, 250–252.

azos, A. J., and Mathieu, M. (1999). Effects of five natural gonadotro-pin-releasing hormones on cell suspensions of marine bivalvegonad: Stimulation of gonadal DNA synthesis. Gen. Comp. Endocri-nol. 113, 112–120.

owell, J. F. F., Reska-Skinner, S. M., Prakash, M. O., Fischer, W. H.,Park, M., Rivier, J. E., Craig, A. G., Mackie, G. O., and Sherwood,N. M. (1996). Two new forms of gonadotropin-releasing hormonein a protochordate and the evolutionary implications. Proc. Natl.Acad. Sci. USA 93, 10461–10464.

edman, R. S., and Berry, R. W. (1991). Temperature-dependentpeptidergic feedback: Potential role in seasonal egg laying inAplysia. J. Neurosci. 11, 1780–1785.

eaman, R. L., Lynch, M. J., and Moss, R. L. (1980). Effects ofhypothalamic peptide hormones on the electrical activity ofAplysia neurons. Brain Res. Bull. 5, 233–237.

ower, S. A., Chiang, Y. C., Lovas, S., and Conlon, J. M. (1993).Primary structure and biological activity of a third gonadotropin-releasing hormone from lamprey brain. Endocrinology 132, 1125–1131.

quire, C. R., Talebian, M., Menon, J. G., Dekruyff, S., Lee, T. D.,Shively, J. E., and Rothman, B. S. (1991). Leucine aminopeptidase-like activity in Aplysia hemolymph rapidly degrades biologically activealpha-bag cell peptide fragments. J. Biol. Chem. 266, 22355–22363.

tefano, G. B., Leung, M. K., Zhao, X., and Scharrer, B. (1989).Evidence for the involvement of opioid neuropeptides in theadherence and migration of immunocompetent invertebrate hemo-cytes. Proc. Natl. Acad. Sci. USA 86, 626–630.

sai, P. S., and Licht, P. (1993). Differential distribution of chicken-Iand chicken-II GnRH in the turtle brain. Peptides 14, 221–226.ayne, N. L., and Wong, H. (1994). Persistence of hormone secretionfrom neuroendocrine cells of Aplysia after termination of electricalafterdischarge. Endocrinology 134, 1046–1054.ilson, T. M., Yu-Lee, L. Y., and Kelley, M. R. (1995). Coordinate geneexpression of luteinizing hormone-releasing hormone (LHRH)and the LHRH-receptor after prolactin stimulation in the rat Nb2T-cell line: Implications for a role in immunomodulation and cellcycle gene expression. Mol. Endocrinol. 9, 44–53.

oung, K. G., Chang, J. P., and Goldberg, J. I. (1999). Gonadotropin-releasing hormone neuronal system of the freshwater snailsHelisoma trivolvis and Lymnaea stagnalis: Possible involvement inreproduction. J. Comp. Neurol. 404, 427–437.

appulla, J. P., Wickham, L., Bawab, W., Yang, X. F., Storozhuk, M. V.,Castellucci, V. F., and DesGroseillers, L. (1999). Cloning andcharacterization of Aplysia neutral endopeptidase, a metallo-endopeptidase involved in the extracellular metabolism of neuro-peptides in Aplysia californica. J. Neurosci. 19, 4280–92.

huang, X., Silverman, A. J., and Silver, R. (1993). Reproductivebehavior, endocrine state, and the distribution of GnRH-like
immunoreactive mast cells in dove brain. Horm. Behav. 27, 283–295.

Biological and Immunological Characterization of Multiple GnRH in an Opisthobranch Mollusk, Aplysia...

Documents

Transcript of Biological and Immunological Characterization of Multiple GnRH in an Opisthobranch Mollusk, Aplysia...