immunohistochemical localization of three different prepro-gnrhs in the brain and pituitary of the...

Immunohistochemical Localization ofThree Different prepro-GnRHs in the

Brain and Pituitary of the European SeaBass (Dicentrarchus labrax) Using

Antibodies to the CorrespondingGnRH-Associated Peptides

DAVID GONZALEZ-MARTINEZ,1 NILLI ZMORA,2 EVARISTO MANANOS,3

DANY SALIGAUT,4 SILVIA ZANUY,3 YONATHAN ZOHAR,5 ABIGAIL ELIZUR,2

OLIVIER KAH,4AND JOSE ANTONIO MUNOZ-CUETO1*

1Departamento de Biologıa, Facultad de Ciencias del Mar y Ambientales, Universidad deCadiz, 11510 Puerto Real, Spain

2National Center for Mariculture, IOLR. Eilat 88112, Israel3Instituto de Acuicultura Torre de la Sal, CSIC. 12595 Ribera de Cabanes, Castellon, Spain

4Endocrinologie Moleculaire de la Reproduction, UMR CNRS 6026, Campus de Beaulieu,35042 Rennes Cedex, France

5Center of Marine Biotechnology, University of Maryland Biotechnology Institute,Baltimore, Maryland 21202

ABSTRACTThe distribution of the cells expressing three prepro-gonadotrophin-releasing hormones

(GnRH), corresponding to salmon GnRH (sGnRH), seabream GnRH (sbGnRH), and chickenGnRH-II (cGnRH-II) forms, was studied in the brain and pituitary of the sea bass (Dicen-trarchus labrax) by using immunohistochemistry. To circumvent the cross-reactivity prob-lems of antibodies raised to GnRH decapeptides, we used specific antibodies generatedagainst the different sea bass GnRH-associated peptides (GAP): salmon GAP (sGAP), sea-bream GAP (sbGAP), and chicken-II GAP (cIIGAP). The salmon GAP immunostaining wasmostly detected in terminal nerve neurons but also in ventral telencephalic and preopticperikarya. Salmon GAP-immunoreactive (ir) fibers were observed mainly in the forebrain,although sGAP-ir projections were also evident in the optic tectum, mesencephalic tegmen-tum, and ventral rhombencephalon. The pituitary only receives a few sGAP-ir fibers. Theseabream GAP-ir cells were mainly detected in the preoptic area. Nevertheless, sbGAP-irneurons were also found in olfactory bulbs, ventral telencephalon, and ventrolateral hypo-thalamus. The sbGAP-ir fibers were only observed in the ventral forebrain, innervatingstrongly the pituitary gland. Finally, chicken-II GAP immunoreactivity was only detected inlarge synencephalic cells, which are the origin of a profuse innervation reaching the telen-cephalon, preoptic area, hypothalamus, thalamus, pretectum, posterior tuberculum, mesence-phalic tectum and tegmentum, cerebellum, and rhombencephalon. However, no cIIGAP-ir fiberswere detected in the hypophysis. These results corroborate the overlapping of sGAP- and sbGAP-expressing cells in the forebrain of the sea bass, and provide, for the first time, unambiguousinformation on the distribution of projections of the three different GnRH forms expressed in thebrain of a single species. J. Comp. Neurol. 446:95–113, 2002. © 2002 Wiley-Liss, Inc.

Indexing terms: GAP; central nervous system; perciform; immunohistochemistry; hypophysis

Grant sponsor: UE; Grant number: FAIR CT97-3785; Grant sponsor:CICYT; Grant numbers: CYTMAR MAR95-C03-02, MAR1988-1542-CE;Grant sponsor: CNRS.

*Correspondence to: Dr. Jose Antonio Munoz-Cueto, Departamento de Bio-logıa, Facultad de Ciencias del Mar y Ambientales, Universidad de Cadiz,Polıgono Rıo San Pedro, 11510 Cadiz, Spain. E-mail: [email protected]

Received 17 May 2001; Revised 17 December 2001; Accepted 2 January2002

DOI 10.1002/cne.10190Published online the week of March 18, 2002 in Wiley Interscience

(www.interscience.wiley.com).

THE JOURNAL OF COMPARATIVE NEUROLOGY 446:95–113 (2002)

© 2002 WILEY-LISS, INC.

Abbreviations

ALL anterior lateral line nerveAON anterior octaval nucleusAP accessory pretectal nucleus (part of the area pretectalis of

Schnitzlein, 1962)C caudal nucleus of the octavolateral areaCCe corpus of the cerebellumCM mammillary bodiesCP central posterior thalamic nucleus (part of the nucleus

dorsomedialis thalami of Schnitzlein, 1962)CZ central zone of the optic tectumDc1 subdivision 1 of the central part of the dorsal telencepha-

lonDc2 subdivision 2 of the central part of the dorsal telencepha-

lonDd dorsal part of the dorsal telencephalonDld dorsolateral part of the dorsal telencephalonDlp posterolateral part of the dorsal telencephalonDlv1 subdivision 1 of the ventrolateral part of the dorsal telen-

cephalonDlv2 subdivision 2 of the ventrolateral part of the dorsal telen-

cephalonDm1 subdivision 1 of the medial part of the dorsal telencepha-

lonDm2 subdivision 2 of the medial part of the dorsal telencepha-



lonDp posterior part of the dorsal telencephalonDP dorsal posterior thalamic nucleus (caudal part of the nu-

cleus dorsomedialis and nucleus dorsolateralis thalamiof Schnitzlein, 1962)

DWZ deep white zone of the optic tectumE entopeduncular nucleusECL external cellular layerEG eminentia granularisG granular layer of the cerebellumGL glomerular layerHCo horizontal commissureICL internal cellular layerIO inferior olivary nucleusIP interpeduncular nucleusIV fourth ventricleLFB lateral forebrain bundleLSO lateral septal organLT lateral thalamic nucleus (nucleus lateralis thalami of

Holmgren, 1920, and Bergquist, 1932, and nucleus post-glomerulosus of Schnitzlein, 1962)

M molecular layer of the cerebellumMAG magnocellular nucleus of the octavolateral areaMc Mauthner cellsMON medial octavolateral nucleusNAPv anterior periventricular nucleus (part of the suprachias-

matic nucleus of Crosby and Woodburne, 1940, nucleuspreopticus parvocellularis posterioris of Braford andNorthcutt, 1983)

NAT anterior tuberal nucleusNC nucleus corticalisNCLI central nucleus of the inferior lobeNDLI diffuse nucleus of the inferior lobeNDLIc caudal part of the diffuse nucleus of the inferior lobeNDLIl lateral part of the diffuse nucleus of the inferior lobeNDLIm medial part of the diffuse nucleus of the inferior lobeNGa anterior part of the nucleus glomerulosus (corpus glo-

merulosum pars anterior of Brickner, 1929)NGp posterior part of the nucleus glomerulosus (corpus glo-

merulosum pars rotunda of Brickner, 1929)NGS secondary gustatory nucleusNGT tertiary gustatory nucleusNI isthmic nucleusNLTd dorsal part of the lateral tuberal nucleusNLTi inferior part of the lateral tuberal nucleusNLTl lateral part of the lateral tuberal nucleusNLTm medial part of the lateral tuberal nucleusNLTv ventral part of the lateral tuberal nucleusNLVc central part of the nucleus of the lateral valvula

NLVp posterior part of the nucleus of the lateral valvulanMLF nucleus of the medial longitudinal fascicleNP nucleus pretectalis of Schnitzlein (1962)NPC central pretectal nucleus (part of the area pretectalis of

Schnitzlein, 1962)NPGc commissural preglomerular nucleus (nucleus supramam-

millaris of Bergquist, 1932)NPGl lateral preglomerular nucleus (part of the nucleus pre-

thalamicus of Meader, 1934)NPGm medial preglomerular nucleusNPOav anteroventral part of the parvocellular preoptic nucleus

(nucleus preopticus periventricularis of Crosby andWoodburne, 1940, part of the nucleus preopticus parvo-cellularis anterioris of Braford and Northcutt, 1983)

NPOpc parvocellular part of the parvocellular preoptic nucleus(nucleus preopticus periventricularis of Crosby andWoodburne, 1940, part of the nucleus preopticus parvo-cellularis anterioris of Braford and Northcutt, 1983)

NPPv posterior periventricular nucleus (nucleus preopticusperiventricularis of Crosby and Woodburne, 1940, partof the nucleus preopticus parvocellularis posterioris ofBraford and Northcutt, 1983)

NPT posterior tuberal nucleusNRLl lateral part of the nucleus of the lateral recessNRP nucleus of the posterior recessNSV nucleus of the saccus vasculosusNTe nucleus of the thalamic eminentia (eminentia thalami of

Bergquist, 1932, and Herrick, 1948)OC optic chiasmOLN olfactory nerve fibersOT optic tectumP pituitaryPCo posterior commissurePG periventricular granular cell mass of the caudal lobepgd dorsal periglomerular nucleusPGZ periventricular gray zone of the optic tectumPLI lateral part of the perilemniscular nucleusPLm medial part of the perilemniscular nucleusPMgc gigantocellular part of the magnocellular preoptic nucleusPPd dorsal periventricular pretectal nucleusPPv ventral periventricular pretectal nucleusPSp parvocellular superficial pretectal nucleus (nucleus lateral

geniculatus of Sheldon, 1912 and Schnitzlein, 1962)PVO paraventricular organ (nucleus posterioris periventricu-

laris and nucleus recessus lateralis pars medialis ofother authors)

RI inferior reticular nucleusRL lateral reticular nucleusRM medial reticular nucleusRS superior reticular nucleusSOF secondary olfactory fibersSR nucleus of the superior rapheSV saccus vasculosusSWGZ superficial white and gray zone of the optic tectumTLa nucleus of the torus lateralisTLo torus longitudinalisTN terminal nerve areaTNgc terminal nerve ganglion cellsTPp periventricular nucleus of the posterior tuberculumTSc central part of the semicircular torusTSl lateral part of the semicircular torusTSv ventral part of the semicircular torusV Descending trigeminal tractVII facial nerveVAO ventral accessory optic nucleusVc central part of the ventral telencephalonVCe valvula of the cerebellumVd dorsal part of the ventral telencephalonVi intermediate part of the ventral telencephalonVl lateral part of the ventral telencephalonVLo vagal lobeVM ventromedial thalamic nucleusVOT ventral optic tractVp postcommissural part of the ventral telencephalonVs supracommissural part of the ventral telencephalonVv ventral part of the ventral telencephalonXm motor nucleus of vagal nerve

96 D. GONZALEZ-MARTINEZ ET AL.

The decapeptide gonadotrophin-releasing hormone(GnRH) constitutes the main cerebral factor stimulatingthe production and secretion of the pituitary gonadotro-phins, follicle-stimulating hormone (FSH) and luteinizinghormone (LH) (Breton et al., 1972; Sherwood et al., 1993).In addition to its role in releasing gonadotrophins, GnRHseems to be implicated in other functions such as nestingand sexual behavior (Muske and Moore, 1994; Yamamotoet al., 1995, 1997) and their understanding requires soliddata and knowledge on the organization of these systems.However, GnRH does not represent a unique molecule. Infact, since the discovery of the mammalian GnRH form in1971 (mGnRH, Matsuo et al., 1971), the existence of fif-teen isoforms of GnRH in vertebrates and protochordateshas been reported, which have received the name of thespecies in which they have been discovered (Powell et al.,1996; Carolsfeld et al., 2000; Okubo et al., 2000a; Mon-taner et al., 2001). From these 15 GnRHs, eight forms arepresent in the brain of different teleosts: salmon GnRH(sGnRH: Sherwood et al., 1983), mGnRH (King et al.,1990), catfish GnRH (cfGnRH: Ngamvongchon et al.,1992), seabream GnRH (sbGnRH: Powell et al., 1994),chicken GnRH-II (cGnRH-II: King and Millar, 1982; Miy-amoto et al., 1984), herring GnRH (hGnRH, Carolsfeld etal., 2000), medaka GnRH (mdGnRH, Okubo et al., 2000a),and pejerrey GnRH (pGnRH, Montaner et al., 2001).

In teleosts, the basic pattern of distribution of GnRHcells suggested the existence of two major GnRH systems:one system along the ventral forebrain (terminal nerve,ventral telencephalon, preoptic area, and hypothalamus)expressing different GnRH forms (sGnRH, mGnRH, cf-GnRH, sbGnRH, etc.), and another system in the synen-cephalon expressing systematically cGnRH-II (Munz etal., 1981; Goos et al., 1985; Kah et al., 1986, 1991; Subhe-dar and Krishna, 1988; Batten et al., 1990; Amano et al.,1991; Grober and Bass, 1991; Yamamoto et al., 1995;Rodrıguez-Gomez et al., 1999). The organization of theseGnRH systems was mainly established via immunohisto-chemistry by using antibodies more or less specific for theendogenous decapeptides. However, in different teleostspecies, terminal nerve cells showed strong sGnRH and aweaker cGnRH-II immunoreactivity, whereas midbraintegmentum cells exhibited strong cGnRH-II and a weakersGnRH immunoreactivity (Yamamoto et al., 1995;Rodrıguez-Gomez et al., 1999). Although colocalization ofdifferent GnRH forms in the same cell cannot be com-pletely excluded, this observation seems to reflect thecross-reactivity of antisera with more than one GnRHform as a consequence of the high identity among thedifferent GnRH decapeptides.

In the European sea bass, a previous study based on theuse of antibodies to sGnRH showed the presence of nu-merous GnRH-immunoreactive neurons along a contin-uum of GnRH fibers extending from the olfactory bulbs tothe pituitary (Kah et al., 1991). On the other hand, thepresence of neurons expressing GnRH in the synencepha-lon of sea bass was not detected by using antibodies tosGnRH (Kah et al., 1991), but they were observed by usingan antiserum against cGnRH-II, which also revealed thepresence of cGnRH-II fibers in the pituitary (J.A. Munoz-Cueto, unpublished). In some, but not all species of te-leosts, cGnRH-II has also been detected in the pituitary(Yu et al., 1988; Schultz et al., 1993; Montero et al., 1995;Rodrıguez-Gomez et al., 1999) although its origin is stilluncertain.

Recently, it has been demonstrated that perciforms ex-press three different GnRH forms in their brains: salmonGnRH, seabream GnRH, and chicken GnRH-II (Powell etal., 1994; White et al., 1995). Furthermore, most studiesperformed up to now suggested the existence of a neuro-anatomical segregation in the expression of these threedifferent GnRHs (Gothilf et al., 1996; Okuzawa et al.,1997; White and Fernald, 1998). Thus, the expression ofsalmon GnRH appears restricted to the rostral olfactorybulbs; seabream GnRH is mainly expressed in the preopticarea, whereas chicken-II GnRH is synthesized in largecells at the diencephalic-mesencephalic transition. Basedon these results, it has been proposed that sGnRH neu-rons differentiate from the olfactory placode, whereassbGnRH and cGnRH-II neurons originate from preopticand mesencephalic primordia, respectively (Parhar, 1997;Parhar et al., 1998; Ookura et al., 1999). Although phys-iological data provided evidence that the three GnRHforms can stimulate the gonadotrophin secretion (Zohar etal., 1995), sbGnRH seems to be the main hypophysiotropichormone in perciforms (Powell et al., 1994; Holland et al.,1998), whereas the role of salmon GnRH and chicken-IIGnRH in reproduction is still unclear.

Recently, the cDNAs encoding the three GnRH isoformspresent in sea bass have been cloned (Zmora et al., 2002),providing information on the molecular structure of thethree preproGnRHs in this species. Sea bass GnRHcDNAs contain coding sequences for a signal peptide, thedecapeptide GnRH, a processing tripeptide, and a GnRH-associated peptide or GAP. These GAP sequences repre-sent valuable tools for in situ hybridization techniquesbecause they are longer when compared to GnRH se-quences and there is a lower sequence identity amongGAP sequences. By using these probes, we have recentlydemonstrated that, according to previous studies, each ofthe three GnRH prepro-mRNAs is preferentially ex-pressed in a specific brain region. However, we have de-scribed for the first time a clear overlapping of the sGAP-and sbGAP-expressing cells in the telencephalon and di-encephalon of the sea bass, suggesting that the anteriorGnRH systems are not clearly segregated, at least in thisspecies (Gonzalez-Martınez et al., 2001).

One of the aims of the present study was to corroboratewhich brain areas express the different GnRH prepro-forms in the sea bass by using immunohistochemical tech-niques. The second aim of this study was to obtain preciseinformation about the projections of each GnRH form intothe brain and pituitary of sea bass, determining whichGnRH form(s) is(are) hypophysiotrophic in this species. Toavoid the problem of cross-reactivity often encounteredwhen raising antibodies to highly related GnRH decapep-tides, we used antibodies generated against the differentsea bass GAPs, because it has been reported that thedistribution of immunoreactive GnRH was consistentlysimilar to that of the corresponding immunoreactive GAP(Ronchi et al., 1992; Polkowska and Przekop, 1993).

MATERIALS AND METHODS

Animals

Adult immature male (n � 5) and female (n � 5) spec-imens of sea bass ranging in weight from 200 to 250 g werepurchased from a local fishery (Cupimar, San Fernando,Spain) and kept in the laboratory in running seawater.Animals were treated in agreement with the European

97GnRHs IN THE BRAIN AND PITUITARY OF SEA BASS

Union regulation concerning the protection of experimen-tal animals.

Generation of specific antibodiesagainst GAPs

The procedures for the isolation of the sequences codingfor the three sea bass GAPs and the expression of recom-binant GAPs were reported previously (Zmora et al., 2002;Gonzalez-Martınez et al., 2001). The full-length sequencesof the three sea bass prepro-GnRH forms are available atthe GeneBank, accession numbers AAF62898 for sb-GnRH, AAF62899 for sGnRH, and AAF62900 for cGnRH-II. Briefly, for the production of recombinant GAPs, twoexpression vectors were used: the His-tagged expressionsystem for sGAP and sbGAP, and the Gluthathione-S-transferase (GST) expression system for cIIGAP. Analysison 17% sodium dodecylsulfate-polyacrylamide gel electro-phoresis (SDS-PAGE) to detect the existence of recombi-nant GAP peptides resulted in clear bands of the expectedsize (6.5 KD) for sGAP and sbGAP. For cIIGAP, SDS-PAGE analysis revealed two bands: the cIIGAP/GST fusedprotein, as a dominant band of 34 KD in size, and asmaller faint band of 27 KD which corresponds to the GSTprotein. The purified recombinants GAPs were used forthe immunization of 9 female guinea pigs (5/6 weeks old;345–365 g) obtained from CEGAV (France) and housed inlaboratory cages designed for guinea pigs. Animals, threefor each GAP, were acclimated for 1 week before the be-ginning of the immunization protocol. Each injection con-sisted of 25 �l of recombinant GAP (1 �g/�l), 75 �l dis-tilled water, and 100 �g of incomplete Freund’s adjuvant.Preimmune sera were collected on the day of the firstinjection and stored at �20°C. The first injection wasintradermal in the back (6–7 points) and the others, at1-month intervals, were given intramuscularly in the pos-terior legs. Eight days after the fourth injection, blood wascollected under anesthesia by cardiac punction, centri-fuged and the serum stored at �25°C.

Dot blot analysis

The specificity of the antisera used in the immunohis-tochemical study was assessed by dot blot analysis byusing the recombinant sGAP, sbGAP, and cIIGAP pro-teins as antigens. Briefly, 6–12 �g of recombinant sGAP,sbGAP, and cIIGAP proteins were added to individualnitrocellulose filter slices. The slices were saturated for anhour in Tris buffer saline pH 7.4 (TBS) containing 5% ofnonfat dry milk. Samples were then incubated for 12hours with the primary antisera (anti-sGAP, anti-sbGAP,and anti-cII-GAP) at a 1:2,000 dilution. After incubation,the slices were washed several times with TBS containing0.05% Tween 20 (Sigma, St. Louis, MO) and incubated for1 hour in peroxidase-conjugated anti-guinea pig IgG(Jackson, Peninsula Laboratories, Belmont, CA) diluted1:3,000. Slices were washed three times in 0.05% TBS-Tween 20, developed in 0.025% 4-chloro-1-naphtol (Sigma)containing 0.002% hydrogen peroxide in 0.05 M Trisbuffer, and washed thoroughly in distilled water.

Immunohistochemistry

Animals were anesthetized with 2-phenoxyethanol (0.3ml/liter of water; Sigma) and perfused via the aortic bulbwith Bouin’s fixative (4% paraformaldehyde and 0.2%picric acid in 0.1 M phosphate buffer, pH 7.4). Brains with

the pituitary attached, were then removed and postfixedin the same fixative for 5–7 hours in the dark at 4°C. Aftercryoprotection overnight in 0.1 M phosphate buffer con-taining 15% sucrose, brains were embedded in Tissue-Tek,frozen in cold isopentane and serial coronal and sagittalsections of 16- �m-thick were obtained with a cryostat.

Immunohistochemical staining was performed by usinga streptavidin-biotin-peroxidase complex method. Endog-enous peroxidase activity was blocked with 1% hydrogenperoxide in Coons buffer p H 7.4 (CBT, 0.01 M Veronal,0.15 M NaCl) containing 0.1–0.2% Triton X-100 for 30minutes. Before immunostaining, sections were trans-ferred for 5 minutes to CBT and then saturated in CBTcontaining 0.5% casein for 30 minutes. Sections were in-cubated overnight in a moist chamber at room tempera-ture with anti-sGAP, anti-sbGAP, and anti-cIIGAP anti-sera (1:500 to 1:1,000 dilution in CBT 0.5% casein).Sections were washed in CBT (2 � 15 minutes) and incu-bated for 1.5 hours at room temperature with Biotin-sp-Conjugated-AffiniPure Goat Anti-Guinea Pig–IgG (Jack-son Immuno Research Laboratories, West Grove, PA)diluted 1:1,000 in CBT. After washing in CBT (2 � 15minutes), sections were incubated 1.5 hour at room tem-perature with peroxidase-conjugated-streptavidin com-plex (Jackson Immuno Research Laboratories) diluted1:1,000 in CBT. Finally, sections were washed in CBTfollowed by Tris-HCl (0.05 M, pH 7.4) and peroxidaseactivity was visualized either in 0.05 M Tris-HCl, pH 7.6containing 0.025% 3,3 diaminobenzidine tetrahydrochlo-ride (Sigma) and 0.01% hydrogen peroxide or 0.04%4-chloro-1-naphthol (Sigma) and 0.01% hydrogen perox-ide. To confirm the specificity of the immunostaining, con-trols were performed by preabsorption of primary antiserawith their respective antigens, replacement of primaryantisera with the corresponding preimmune sera, andomission of primary or biotinylated antisera.

To check that GAP was a reliable marker of GnRH-producing cells, double immunofluorescence staining hasbeen carried out by incubating the sections concomitantlywith anti-sGAP and anti-sGnRH (donated by Dr. Breton)sera, anti-sbGAP and anti-mGnRH (donated by Dr.Tramu) sera, and anti-cIIGAP and anti-cGnRH-II (donat-ed by Dr. Peute) sera. The GAP and GnRH antisera usedwere obtained in guinea pigs and rabbits, respectively.The immunoreactive GAP cells and fibers were revealedby incubating the sections with an antibody againstguinea pig IgG coupled to fluorescein (Jackson Immu-noResearch), whereas the immunoreactive GnRH cellsand fibers were evidenced by using an antibody againstrabbit IgG coupled to 7-amino-4-methylcoumarin-3-aceticacid (AMCA) or Texas Red (Jackson ImmunoResearch).

Immunohistochemical sections were analyzed on aLeica photomicroscope and computer images were ob-tained with a Sony DKC-CM30 Digital Camera (Sony,Japan). The software used was Adobe Photoshop 5.0 andno subsequent alterations have been made with the ex-ception of the color of Figure 1B, which has been alteredfor purposes of illustration. Ten to fifty immunoreactivecells of each nucleus were measured in order to evaluatethe average cell size. For the precise localization of thedifferent GAP cells and projections, we have used a de-tailed sea bass brain atlas recently developed in our lab-oratory (Cerda-Reverter et al., 2001a, b).


RESULTS

Table 1 summarizes the results obtained by dot blotanalysis in order to test the specificity of the anti-sGAP,anti-sbGAP, and anti-cIIGAP sera. Immunostaining re-vealed a strong specific signal of the primary antiserumwith its own antigen, but no nonspecific reaction wasobserved with the two other recombinant proteins, con-firming the specificity of sera used in this immunohisto-chemical study. However, the specificity of the differentanti-GnRH sera used could not be confirmed at immuno-histochemical level because a cross-reactivity of antiseraappears evident on the same cell masses. At least in seabass (data not shown), the anti-cGnRH-II serum immuno-stained large cell bodies in the synencephalon but also inthe ventral telencephalon and olfactory bulbs (putativesGnRH and/or sbGnRH cells), whereas the anti-sGnRHantibody immunolabelled neurons in the ventral forebrainbut also detected immunoreactive perikarya in the dorsalsynencephalon (putative cGnRH-II cells).

As we used antisera against the GAP fragments of eachprepro-GnRH form in order to localize specifically thedifferent GnRH-expressing cells, this made it necessary todemonstrate previously that GAP-immunoreactive cellsalso expressed GnRH peptides in sea bass. Thus, by usingdouble immunofluorescence techniques we have corrobo-rated that salmon GAP-immunoreactive cells were alsoimmunoreactive for anti salmon GnRH (Fig. 1A and B),seabream GAP-immunoreactive cells were also immuno-stained with anti mammalian GnRH serum (Fig. 1C andD), and chicken-II GAP-immunoreactive cells also exhib-ited immunoreactivity with the chicken GnRH-II anti-serum (Fig. 1E and F).

All levels of the brain were examined with the threeanti-GAP sera on adjacent serial coronal and sagittal sec-tions and results obtained are summarized in Figure 1G.No immunostaining was observed when primary antiserawere preabsorbed with their respective antigens or re-placed with the corresponding preimmune sera, or when

primary and biotinylated antisera were omitted. The re-sults were highly consistent from one fish to the other andno evident differences could be noted between males andfemales, although the number of positive cells for eachantiserum was not quantified. The levels of transversesections illustrated in Figure 2 are shown in Figure 1G.

Salmon GAP-immunoreactive cells

Immunoreactivity for anti-sGAP serum was mostly de-tected in the olfactory bulbs, and particularly in ganglioncells of the terminal nerve associated to the glomerularlayer (Figs. 2A and 3A) and at the junction between thecaudal olfactory bulbs and the telencephalon (Figs. 2B and3B). These medium-sized to large cells (10–25 �m in di-ameter) were rounded, ovoid, or pyramidal in shape. Fur-thermore, sGAP-immunoreactive (sGAP-ir) cells were alsodetected further caudal, in the ventral telencephalon (Fig.2C) and in the ventral preoptic area, associated to theparvocellular and anteroventral subdivisions of the par-vocellular preoptic nucleus (Fig. 2D). Such cells were alsoround, bipolar, or pyramidal in shape but they were muchsmaller (4–8 �m in diameter) than in olfactory bulbs (Fig.3D and E). More caudally, no other sGAP-ir cells could bedetected in the brain of the sea bass.

The sGAP-ir fibers were observed mainly in the fore-brain. These sGAP-ir fibers were more evident in olfactorybulbs (Figs. 2A,B, 3C), ventral telencephalon (ventral, dor-sal, supracommissural, postcommissural, and intermedi-ate nuclei, Figs. 2C,D, 3D), preoptic area (parvocellularand magnocellular preoptic nuclei, anterior periventricu-lar nucleus, suprachiasmatic nucleus, Figs. 2D,E, 3F),hypothalamus (medial part of the lateral tuberal nucleus,anterior tuberal nucleus, nucleus of the lateral recess, Fig.2F,G), ventral thalamus (ventromedial and ventrolateralthalamic nuclei, Fig. 2E), pretectum (central prectectalnucleus, Fig. 2F) and dorsal tegmentum (semicirculartorus, secondary gustatory nucleus, perilemniscular nu-cleus, Fig. 2G,H). However, a conspicuous sGAP innerva-tion was also observed in the dorsal telencephalon (lat-eroventral and lateroposterior parts, posterior part,second subdivision of the medial zone, Fig. 2C,D), optictectum (central zone and periventricular gray zone, Figs.2E–H, 3G), and ventral rhombencephalon (reticular for-mation, anterior octaval nucleus, vagal lobe, inferior oli-vary nucleus, Figs. 2H–J, 3H). The pituitary only receivesa small number of sGAP-ir fibers (Fig. 2F). These fibersenter the neurohypophysis and reach the proximal parsdistalis of the adenohypophysis (Fig. 6A).

Seabream GAP-immunoreactive cells

The sbGAP immunoreactivity was more evident in cellsof the preoptic area associated with the parvocellular pre-optic nucleus (Figs. 2D, 4E) and the anterior periventricu-lar nucleus (Fig. 2E), but were also slightly more rostral,in the ventral nucleus of the ventral telencephalon (Figs.2C, 4D). These seabream GAP-immunoreactive (sbGAP-ir) cells were small or medium-sized (5–20 �m in diame-ter) and exhibited fusiform, bipolar, rounded, or ovoidshapes. However, sbGAP immunoreactivity was not re-stricted to these areas and some sbGAP-ir cells also ap-peared in the olfactory bulbs, associated with the glomer-ular layer (Figs. 2A, 4A) and the terminal nerve area(Figs. 2B, 4B), where the sGAP-ir cells were found. Thisresult did not represent a cross-reaction because labellingwith anti-sbGAP serum was not evident in large sGAP-ircells of the terminal nerve (Fig. 4B). Furthermore,

TABLE 1. Dot Blot Analysis of Anti-GAP Sera1

1Dot blot analysis using the recombinant sGAP, sbGAP, and cIIGAP proteins (6–12 �g)as antigens, and anti-sGAP, anti-sbGAP, and anti-cII-GAP (1:2,000 dilution) as pri-mary antisera. See Material and Methods for further description.2GAP, gonadotrophin-releasing hormone-associated peptides; sGAP, salmon GAP; sb-GAP, seabream GAP; cIIGAP, chicken-II GAP.


Figure 1.


sbGAP-ir cells of the olfactory bulbs were less numerousand clearly smaller than sGAP-ir cells of the same region(4–10 �m in diameter versus 10–25 �m in diameter).There were also some sbGAP-ir cells, further caudally, inthe ventrolateral hypothalamus (Figs. 2F, 4G), in whichonly one to two small and rounded immunolabelled cells(4–9 �m in diameter) could be observed per section.

The sbGAP-ir fibers were observed only in the ventralsurface of the forebrain, along the ventral telencephalon,the preoptic area, and the ventral hypothalamus, inner-vating strongly the pituitary gland of sea bass (Figs.2C–F, 4F–H). Nevertheless, some varicose immunoreac-tive fibers originating in sbGAP-ir cells of the olfactorybulbs did not run to the pituitary and remained in theolfactory area (Fig. 4C), in the vicinity of large ganglioncells of the terminal nerve, which appeared immuno-stained by using the anti-sGAP serum (Fig. 3B) but notappear immunostained with anti-sbGAP serum (Fig. 4C).Running rostrocaudally from the ventral telencephalon tothe pituitary stalk, there was clearly a major tract ofsbGAP-ir fibers along which sbGAP-ir cell bodies could bedetected. Furthermore, another tract of sbGAP-ir fibersoriginating in the hypothalamus arched in the ventrolat-eral surface of the brain, running caudorostrally to reachthe pituitary of sea bass. Both fiber tracts appeared moreevident in sagittal sections (Fig. 4F,H) than on transversesections (Fig. 4D,G). These sbGAP-ir fibers entered theneurohypophysis and invaded the proximal pars distalisof the adenohypophysis (Figs. 2F, 6B), and the border ofthe pars intermedia.

Chicken-II GAP-immunoreactive cells

Finally, chicken-II GAP-immunoreactive (cIIGAP-ir)cells appeared restricted to the medial zone of the dorsalsynencephalon, lying close to the fibers of the medial lon-gitudinal fascicle in a periventricular position (Figs. 2G,5A). These large cells (15–23 �m in diameter), roundedand ovoid in shape, appeared just caudal to the posteriorcommissure and rostral to the mesencephalic tegmentum.

In turn, cIIGAP-ir fibers were profusely distributed inthe brain of sea bass, being especially evident in the mid-brain and hindbrain. This innervation was conspicuous inthe dorsal telencephalon (medial, lateroventral, and pos-terior parts, Fig. 2C,D), ventral telencephalon (ventral,dorsal, supracommissural, central, and postcommissuralnuclei, Fig. 2C,D), periventricular preoptic area (parvocel-lular and magnocellular preoptic nuclei, anterior periven-tricular nucleus, Fig. 2D,E), and hypothalamus (lateraltuberal nucleus, anterior tuberal nucleus, nucleus of thelateral recess, diffuse nucleus of the inferior lobe, Fig.2F–H), dorsal and ventral thalamus (Fig. 2E,F), pretec-tum (superficial and central pretectal nuclei, Fig. 2F),posterior tuberculum (posterior tuberal nucleus, periven-tricular nucleus, preglomerular nuclear complex, Fig.2F,G), mesencephalic tectum (central zone, periventricu-lar gray zone, superficial white and gray zone, longitudi-nal torus, Figs. 2E–H, 5B) and tegmentum (semicirculartorus, oculomotor and trochlear nuclei, secondary gusta-tory nucleus, locus coeruleus, Figs. 2G,H, 5C), granularlayer of the valvula, corpus and eminentia granularis ofthe cerebellum (Figs. 2G–I, 5D,E), medulla oblongata (re-ticular formation, interpeduncular nucleus, octavolateralarea, visceromotor nuclei, vagal lobe, inferior olive, Figs.2H–J, 5F) and spinal cord. Although cIIGAP-ir fiberscould be observed in the ventromedial hypothalamus, inthe vicinity of the pituitary stalk, no cIIGAP-ir axons weredetected in sea bass hypophysis (Fig. 6C).

DISCUSSION

It is considered that the presence of three GnRH formsin the brain is not restricted to perciform fish (Powell etal., 1994; White et al., 1995; Gothilf et al., 1996; Gonzalez-Martınez et al., 2001) but is likely to be prevalent in most,if not all, vertebrates (Powell et al., 1985; Sherwood et al.,1986; Lescheid et al. 1997; Montaner et al., 1998,1999;Yahalom et al., 1999). These findings raise the question onthe precise roles of multiple GnRH forms in the control ofreproduction and other physiological processes. Emergingdata also point to a multiplicity of GnRH receptors, be-cause two GnRH receptor subtypes have been evidencedin the goldfish (Illing et al., 1999) and three in the bullfrog,a diploid species (Wang et al., 2001). One approach to-wards clarifying this puzzle is to clearly identify the brainnuclei expressing the different GnRH variants and theirrespective projections into the brain and pituitary.

In this study, we have presented the distribution ofimmunoreactive prepro-GnRH cells in the brain of theEuropean sea bass, Dicentrarchus labrax, by using specificantisera directed against the different sea bass GAP pep-tides. The results permit us to demonstrate that, in seabass, GAP and GnRH peptides are coexpressed in thesame cells, as described in mammals (Ronchi et al., 1992;Polkowska and Przekop, 1993), and confirm that GAPantisera represent valuable tools to localize specificallythe cells expressing the different GnRH forms, avoidingthe problem of cross-reactivity of antibodies directedagainst the GnRH decapeptides.

The immunohistochemical localization of sGAP, sbGAP,and cIIGAP cell bodies totally agrees with results ob-tained in the same species by using in situ hybridizationtechniques (Gonzalez-Martınez et al., 2001). Thus,sGAP-ir cells are predominant in the olfactory bulbs,whereas sbGAP-ir cells were more abundant in the caudaltelencephalon and preoptic area, and cIIGAP-ir cells ap-

Fig. 1. Evidence of the coexpresion of prepro-gonadotrophin-releasing hormone-associated peptides (GAPs; left column) andprepro-gonadotrophin-releasing hormones (GnRHs; right column) onthe same cells by using immunofluorescence methods. Note the im-munolabelling on the same cells and fibers when comparing A with B,C with D, and E with F. A: Section processed with an anti-salmonGAP (sGAP) serum obtained in guinea pig and anti-guinea pig IgGcoupled to Texas Red. B: Same section as A processed with an anti-sGnRH serum obtained in rabbit and anti-rabbit IgG coupled toAMCA. The color has been altered from blue to green for purposes ofillustration. C: Section processed with an anti-seabream GAP (sb-GAP) serum obtained in guinea pig and anti-guinea pig IgG coupled toTexas Red. D: Same section as C processed with an anti-mammalian(mGnRH) serum obtained in rabbit and anti-rabbit IgG coupled tofluorescein isothiocyanate (FITC). E: Section processed with an anti-chicken-II GAP (cIIGAP) serum obtained in guinea pig and anti-guinea pig IgG coupled to Texas Red. F: Same section as E processedwith an anti-chicken (cGnRH-II) serum obtained in rabbit and anti-rabbit IgG coupled to FITC. G: Sagittal drawing of sea bass brainsummarizing the distribution of sGAP (red triangles), sbGAP (greencircles) and cIIGAP (blue stars) cells and fibers (colored small dots).Lettered lines indicate the levels of the transverse sections shown inFigure 2 (? refers to the caution concerning the presence of sGAP cellsprojecting to the pituitary in the olfactory bulbs of the sea bass. Formore details, see the Discussion section). Hypot, hypothalamus; MO,medulla oblongata; OB, olfactory bulbs; Pit, pituitary; POA, preopticarea; SC, spinal cord; Syn, synencephalon;Tel, telencephalon. For allother abbreviations, see list. Scale bars � 100 �m in A–F.


Fig. 2. A–J: Series of schematic transverse sections through thebrain of sea bass showing the distribution of immunoreactive salmonprepro-gonadotrophin-releasing hormone-associated peptides (sGAP;triangles), seabream GAP (sbGAP; circles), and chicken-II (cIIGAP;

stars) perikarya and fibers (small dots). Thick lines represent fibersen passage. A constitutes the most rostral section and J the mostcaudal one. For abbreviations, see list. Scale bars � 1 mm.


Figure 2 (Continued)


Fig. 3. Distribution of immunoreactive salmon prepro-gonad-otrophin-releasing hormone-associated peptide (sGAP) cells and fi-bers in the sea bass brain. A: sGAP cells in the glomerular layer ofolfactory bulbs. Transverse section. B: sGAP immunoreactivity interminal nerve ganglion cells and fibers. Sagittal section. C: sGAP-immunoreactive fibers in the olfactory bulbs. Darkfield. Sagittal sec-tion. D: sGAP cells and fibers in the ventral nucleus of the ventraltelencephalon. Transverse section. E: sGAP cells in the parvocellularpart of the parvocellular preoptic nucleus. Transverse section.

F: sGAP-immunoreactive fibers in the preoptic area. Darkfield. Sag-ittal section. G: sGAP-immunoreactive fibers in the optic tectum.Darkfield. Transverse section. H: sGAP-immunoreactive fibers in cau-dal rhombencephalon. Darkfield. Sagittal section. Arrowheads in A,D,and E mark sGAP-immunoreactive cells. OB, olfactory bulbs; POA,preoptic area; RF, reticular formation; Tel, telencephalon. For allother abbreviations, see list. Scale bars � 50 �m in A,D,E, 200 �m inB,C,F,G,H.

Figure 4


peared restricted to the synencephalon. This study dem-onstrates that the overlapping of the sGAP and sbGAPfrom the olfactory bulbs to the preoptic region obtained atthe mRNA level (Gonzalez-Martınez et al., 2001) is alsoverified at the protein level. This overlapping did notrepresent a cross-reaction of antibodies because sbGAP-ircells in olfactory bulbs were clearly smaller than sGAP-ircells. Labelling with anti-sbGAP sera was never evident inlarge sGAP-ir cells, and conversely, anti-sGAP sera didnot immunostain smaller sbGAP-ir cells. Furthermore, asit was observed by using in situ hybridization (Gonzalez-Martınez et al., 2001), the ventrolateral hypothalamus ofsea bass only contained sbGAP-ir neurons and appeareddevoid of sGAP immunoreactivity on cell bodies. Finally,dot blot analysis and controls performed reinforce thespecificity of antisera used in this study.

The overlapping distribution of sGAP- and sbGAP-expressing cells from the olfactory bulbs to the preopticregion obtained by using both immunohistochemical (thisstudy) and in situ hybridization techniques (Gonzalez-Martınez et al., 2001) in sea bass brain is in contradictionto previous studies in perciforms showing a neuroana-tomical segregation in the expression of forebrain GnRHsystems, viz. sGnRH in terminal nerve and sbGnRH inpreoptic area (White et al., 1995; Gothilf et al., 1996;Okuzawa et al., 1997; White and Fernald, 1998; Parhar etal., 1998). Based on this expression pattern and ontogenicstudies, several authors proposed that sGnRH cells de-velop from the olfactory placode, whereas sbGnRH neu-rons originate from the preoptic area (Parhar, 1997, 1999;Parhar et al., 1998; Ookura et al., 1999). This is in con-trast with evidence obtained in other vertebrates, inwhich all forebrain GnRH neurons seem to have similardevelopmental origins. In amphibian, avian, and mam-mals, neurons destined to produce the preoptic/hypothalamic GnRH form develop from the olfactory pla-code and migrate centrally along the ventral surface of thebrain, adopting their final positions in a continuum fromthe olfactory bulbs to the hypothalamus (Wray et al., 1989;Muske, 1993; Muske and Moore, 1994; Norgren and Gao,1994; Schwanzel-Fukuda, 1999). Recently, in addition tothe mGnRH and cGnRH-II isoforms (White et al., 1998;Urbanski et al., 1999), the existence of a third form ofGnRH which seems to correspond with sGnRH, has beenreferred to in mammals, including humans (Yahalom et

al., 1999). Furthermore, there is growing evidence on theexistence of sGnRH in amphibian (Sherwood et al., 1986),reptilian (Powell et al., 1985), and eutherian mammals(Montaner et al., 1998, 1999). In calves and humans,sGnRH and mGnRH are localized in the same regions andthe total contents of sGnRH and mGnRH in the hypothal-amus are similar (Yahalom et al., 1999). In the rhesusmacaque, the existence of two different populations ofGnRH-immunoreactive neurons that arise during devel-opment has been demonstrated, exhibiting different mor-phological features (Quanbeck et al., 1997). A comparableresult was obtained in sea bass, in which two distinctGnRH cell populations, i.e., sGAP- and sbGAP-ir cells,were colocalized in the same brain areas. According tothese results, we have also observed morphological differ-ences between sGAP- and sbGAP-ir cells, especially in theterminal nerve area of sea bass, sGAP-ir cells being muchlarger than sbGAP-ir neurons. Taken together, ourpresent and previous (Gonzalez-Martınez et al., 2001) re-sults suggest that, as in other vertebrates, the forebrainGnRH systems of highly evolved teleosts arise from acommon olfactory primordium. Furthermore, it seems im-probable that sbGnRH cells in sea bass originate in thepreoptic area because this would imply that they migratein two different directions during development, rostrallytowards the olfactory bulbs, and caudally up to the ventralhypothalamus.

The distribution pattern of sGAP- and sbGAP-ir cells onthe same brain areas suggests that both forebrain GnRHisoforms have evolved from a common ancestor gene byduplications and mutations, probably coinciding with thegenomic duplication that occurred in early teleosts(Amores et al., 1998). Furthermore, this duplication ap-pears to have occurred more recently than the duplicationthat gave rise to the cGnRH-II form, which represents themost conserved GnRH form in phylogeny. The identifica-tion of three GnRH forms in a representative of an earlyevolving teleost group such as the herring (Carolsfeld etal., 2000) indicates that the presence of three GnRHs isnot restricted to the brain of highly evolved teleost. Con-sidering the monophyletic hypothesis of teleostean evolu-tion (De Pinna, 1996), all teleosts evolving after the her-ring should exhibit three forms of GnRH, but someeuteleosts such as salmonids or catfishes have only twoforms of GnRH. White and Fernald (1998) remarked onthe high similarities between the genomic sequence en-coding the sGnRH form in the perciform Haplochromisburtoni and the presumed releasing form of GnRH insalmonids, and suggested that the gene encoding the re-leasing form of GnRH in salmonids might not yet be de-scribed. In salmonids, specific in situ hybridization studiesshowed an intense sGnRH gene expression in the terminalnerve area and a weak expression in the preoptic area(Suzuki et al., 1992; Bailhache et al., 1994). A similarresult was obtained by us in sea bass by using in situhybridization with sGAP riboprobes, whereas sbGAPmRNA expression exhibited the inverse pattern, i.e., anintense labelling in the preoptic area and a lower expres-sion in the olfactory bulbs (Gonzalez-Martınez et al.,2001). Nevertheless, immunocytochemical studies usinganti-sGnRH sera in salmonids revealed an intense label-ling in preoptic GnRH cells (Amano et al., 1991; Navas etal., 1995). In the sea bass, a study performed before thecharacterization of sbGnRH and based on the use of anti-bodies to sGnRH also showed the presence of numerousGnRH-ir neurons in the anterior olfactory bulbs, terminal

Fig. 4. Distribution of immunoreactive seabream prepro-gonadotrophin-releasing hormone-associated peptide (sbGAP) cellsand fibers in the sea bass brain. A: sbGAP cell in the glomerular layerof olfactory bulbs. Transverse section. B: sbGAP-immunoreactive cellin the proximity of two unstained terminal nerve ganglion cells (openarrowheads). Transverse section. C: sbGAP-immunoreactive fibers inthe olfactory bulbs of sea bass surrounding unstained terminal nerveganglion cells (open arrowheads). Sagittal section. D: sbGAP cells andfibers in the ventral nucleus of the ventral telencephalon. Transversesection. E: sbGAP cells in the parvocellular part of the parvocellularpreoptic nucleus. Sagittal section. F: sbGAP-immunoreactive fibertracts in the preoptic area running to the pituitary. Darkfield. Sagittalsection. G: sbGAP-immunoreactive cells and fibers in the ventrolat-eral hypothalamus. Transverse section. H: sbGAP-immunoreactivefiber tract in the ventrolateral hypothalamus running rostrally toreach the pituitary. Darkfield. Sagittal section. Dark arrowheads inA,B,D, and G mark sbGAP-immunoreactive cells. Hypot, hypothala-mus; OB, olfactory bulbs; POA, preoptic area; Tel, telencephalon.Scale bars � 50 �m in A–E,G, 200 �m in F,H. For abbreviations, seelist.


nerve area, ventral telencephalon, ventrolateral preopticregion, and ventrolateral hypothalamus along a conspicu-ous tract of immunoreactive fibers running to the pitu-itary (Kah et al., 1991). The present study fully confirmsthis distribution but indicates that the sGnRH antibodiesused in this pioneer study probably recognizes bothsGnRH and sbGnRH of sea bass because labelling with

anti-sGnRH serum fits almost perfectly with the additionof sGAP and sbGAP immunoreactivity. In addition, an insitu hybridization and immunohistochemical study per-formed in catfish by using specific riboprobres and anti-bodies against the catfish GAP isoform revealed the dis-tribution of catfish GnRH-expressing cells in this species(Zandbergen et al., 1995). This distribution coincides al-

Fig. 5. Distribution of immunoreactive chicken-II prepro-gonad-otrophin-releasing hormone-associated peptide (cIIGAP) cells and fi-bers in the sea bass brain. A: cIIGAP cells in the dorsal synencepha-lon. Transverse section. B: cIIGAP-immunoreactive fibers in the optictectum. Darkfield. Transverse section. C: cIIGAP-immunoreactive fi-bers in the torus semicircularis of the tegmentum. Sagittal section.D: cIIGAP-immunoreactive fibers in the corpus of the cerebellum.Darkfield. Sagittal section. E: cIIGAP- immunoreactive fibers in thevalvula of the cerebellum. Darkfield. Sagittal section. F: cIIGAP-

immunoreactive fibers in the caudal rhombencephalon. Darkfield.Sagittal section. Arrowheads in A mark cIIGAP-immunoreactivecells. CCg, granular layer of the corpus of the cerebellum; CCm,molecular layer of the corpus of the cerebellum; RF, reticular forma-tion; Syn, synencephalon; Teg, tegmentum; TS, torus semicircularis;VCg, granular layer of the valvula of the cerebellum; VCm, molecularlayer of the valvula of the cerebellum. For abbreviations, see list.Scale bars � 100 �m in A,B, 200 �m in C–F.


most completely with the localization of sbGAP cells de-tected in sea bass by using both in situ hybridization(Gonzalez-Martınez et al., 2001) and immunohistochem-istry (this study). Furthermore, prepro-cfGnRH mRNA-expressing cells and cfGnRH-ir perykaryon present in theolfactory area were scarce and small (see Figs. 2 and 5 ofZandbergen et al., 1995, respectively), resembling the sb-GAP cells located in the olfactory bulbs of the sea bass anddiffering notably of the large and abundant sGAP cellspresent in this area. Taken together, this evidence sug-gests that a third GnRH form probably has yet to bediscovered in salmonids and catfishes.

Many researchers consider that cGnRH-II neurons inthe synencephalon (or rostral midbrain tegmentum) be-long to the nucleus of the medial longitudinal fascicle(nMLF), because of their similar location and spinal pro-jections of the two neuronal populations. However, theyhave connectional, cytological, and immunological differ-ences, and should not be regarded as the same neuronalpopulation. Neurons of the nMLF project only to the spi-nal cord, whereas synencephalic GnRH neurons project tothe telencephalon, diencephalon, mesencephalon, cerebel-lum, and medulla oblongata, in addition to the spinal cord(Yamamoto et al., 1995, 1998b). The nMLF neurons areinmunonegative to GnRH sera, located ventrolateral tothe GnRH neurons, and do not show ultrastructural fea-tures of peptidergic neurons (Miller and Kriebel, 1985,Miller and Kriebel, 1986b; Yamamoto et al., 1998b). Alsoimportantly, these previous studies show that the synen-cephalic GnRH neurons are located in the subventricularzone surrounded with ependymal matrix and that nMLFneurons are present beneath the ependyma surrounded bynMLF fibers. Thus, synencephalic GnRH neurons and the

nMLF neurons cannot be regarded as dorsal and ventralsubnuclei of a single cell mass. Miller and Kriebel (1985)used the term “dorsal tegmental magnocellular nucleus”to distinguish this synencephalic GnRH cell mass from thenMLF.

Moreover, we have obtained, for the first time, unam-biguous detailed information on the distribution of fibersimmunoreactive for the three different GnRH forms ex-pressed in the brain of a single species, which could pro-vide valuable information on the precise functions of eachGnRH form. The sGAP-ir fibers were observed mainly inthe forebrain (olfactory bulbs, ventral telencephalon, pre-optic area, periventricular hypothalamus, ventral thala-mus, pretectum) although a conspicuous sGAP innerva-tion was also observed in the dorsal tegmentum, optictectum, and ventral rhombencephalon. In turn, thesbGAP-ir fibers were observed only in the ventral surfaceof the forebrain, associated with the ventral telencepha-lon, preoptic area, and hypothalamus. Furthermore,cIIGAP-ir axons were profusely distributed in the brain ofsea bass, being especially evident in dorsal and ventraltelencephalon, periventricular preoptic area and hypo-thalamus, thalamus, pretectum, posterior tuberculum,mesencephalic tectum and tegmentum, cerebellum, andthe rest of the rhombencephalon. By using these specificanti-GAP sera, we have also demonstrated that the sb-GnRH neurons represent the main source of GnRH to thepituitary of sea bass. The sbGAP-ir fibers reach the prox-imal pars distalis and the border of the pars intermedia ofthe pituitary, where gonadotrophic cells and GnRH recep-tors are also found (Gonzalez-Martınez et al., unpublisheddata). This result corroborates physiological evidence stat-ing the main role of sbGnRH in the stimulation of the

Fig. 6. Serial transverse sections of the sea bass pituitary and hypo-thalamus immunostained with anti-salmon prepro-gonadotrophin-releasing hormone-associated peptides (sGAP; A), anti-seabream prepro-gonadotrophin-releasing hormone-associated peptides (sbGAP; B), andanti-chicken-II prepro-gonadotrophin-releasing hormone-associated pep-tide (cIIGAP; C) sera. Darkfield. Note the intense sbGAP innervation of

the pituitary (B), the presence of a few sGAP-immunoreactive fibers (A),and the absence of cIIGAP immunostaining (C) in the hypophysis of seabass (white arrowheads). NH, neurohypophysis; PPD, proximal parsdistalis of the pituitary. For abbreviations, see list. Scale bars � 400 �mfor A–C.


secretion of gonadotrophins in perciforms (Powell et al.,1994; Zohar et al., 1995; Gothilf et al., 1996,1997;Yamamoto et al., 1997; Holland et al., 1998; Senthilkuma-ran et al., 1999). However, sbGnRH does not represent theunique hypophysiotrophic GnRH in sea bass. Immunohis-tochemical results demonstrate that sGAP-ir axons alsoreach the pituitary of sea bass but this innervation isstrongly reduced as compared to sbGAP immunoreactiv-ity. In contrast, no cIIGAP-ir axons were detected in thepituitary of sea bass, suggesting that the putative role ofchicken-II GnRH in the control of reproduction does notinvolve a direct action of cerebral cGnRH-II on gonadotro-phic cells, at least in this species. According to our results,no immunoreactive fibers were detected in the pituitary ofcatfish by using a specific antibody against cIIGAP (Zand-bergen et al., 1995). Thus, the detection of cIIGnRH-irfibers in the pituitary of sea bass by using an anti-cGnRH-II serum seems to be the consequence of the cross-reactivity of this antibody with sbGnRH and/or sGnRHaxons entering the hypophysis.

Recently, Rodriguez et al. (2000) found that all threeGnRH forms were present in the pituitary of male seabass during sex differentiation and first spawning season,being sbGnRH levels 9-fold higher than cGnRH-II and17-fold higher than sGnRH levels. The contradiction ofthis finding to our immunohistochemical results could re-flect the existence of changes in the presence of cGnRH-IIin the sea bass pituitary in relation to sex, developmental,and/or physiological stages. Furthermore, the presence ofcGnRH-II in the pituitary of sea bass could be the conse-quence of an intrapituitary secretion because, at least inseabream, a local pituitary synthesis of GnRH has beendescribed (Zmora et al., 2000). However, the detection ofcGnRH-II in sea bass pituitary by using enzyme-linkedimmunoassays could also be the result of a cross-reactivityof antisera against the different GnRH peptides, as it wasobserved in immunohistochemistry.

In dwarf gourami and tilapia, it has been demonstratedby using immunohistochemistry combined with biocytintract tracing that only preoptic GnRH cells innervate thehypophysis, whereas terminal nerve and midbrain GnRHneurons do not project to the pituitary (Yamamoto et al.,1998a, b). These results are consistent with the lack ofchanges in the distribution of GnRH fibers in the pituitaryafter terminal nerve and olfactory tract lesions (Kim et al.,1993; Yamamoto et al., 1995) and the absence of labelledfibers in the pituitary after intracellular biocytin injec-tions into terminal nerve GnRH cells (Oka and Matsu-shima, 1993). In other fish species, cGnRH-II-ir fiberswere not detected in the pituitary despite the presence ofcGnRH-II-ir cell bodies in the dorsal synencephalon(Amano et al., 1991; Montero et al., 1994; Millar and King,1994). In agreement with these results, we have not foundcIIGAP-ir axons in the sea bass pituitary. However,sGAP-ir fibers reach the sea bass pituitary, the origin ofthese axons remaining unsolved. It seems improbable thatthe projection pattern of GnRH cells markedly differsbetween perciform species such as tilapia and sea bass. Ifthis is true, preoptic but not olfactory sGAP-ir cells mightcontribute to this pituitary innervation. However, in thegoldfish, small fusiform neurons, distinct from the termi-nal nerve cells, in the olfactory bulbs and tracts werelabelled following DiI injection in the pituitary, suggestingthat a minor contingent of GnRH fibers reaching the pi-tuitary may originate from the olfactory system (Angladeet al., 1993). In mammals, the presence of sGnRH in

neuronal fibers of the median eminence has been reported,suggesting that sGnRH is transported to the pituitarygland where it may interact with pituitary cells, butsGnRH content in pituitary stalk was reduced in compar-ison to mGnRH levels (Yahalom et al., 1999). A compara-ble result was obtained in sea bass, in which we notedhigher sbGAP fiber immunoreactivity compared to sGAPinnervation in the pituitary.

Should the presence of sGnRH be confirmed in highervertebrates, this could mean that sGnRH and cGnRH-IIhave been submitted to a strong evolutionary pressure, incontrast with the gonadotrophin-releasing GnRH form(lamprey GnRH-III, dogfish GnRH, catfish GnRH, sea-bream GnRH, herring GnRH, pejerrey GnRH, medakaGnRH, chicken GnRH-I, guinea pig GnRH, mammalianGnRH) which exhibits considerable variation among ver-tebrates. This selective pressure on sGnRH and cGnRH-IIstructure reinforces the assumption that these neurohor-mones could deserve important functions in brain otherthan reproduction (e.g., neurotransmitter and/or neuro-modulator). In this context, our recent ontogenic studiesin sea bass reveal that cGnRH-II and sGnRH are ex-pressed very early during development, much before thedifferentiation of gonadotrophic cells and gonads, whereassbGnRH is expressed much later (Gonzalez-Martınez etal., 2002). In turn, the diversity of hypophysiotrophic Gn-RHs might indicate a specialization of this GnRH towardsa single function, such as the stimulation of gonadotro-phin secretion. In fact, the sbGAP-ir fibers in sea bass canonly be detected in the ventral forebrain running to thepituitary, sGAP-ir and cIIGAP-ir fibers being profuselydistributed in other brain areas. It has been proposed thatsGnRH might coordinate the sensory and motivationalsystems, cGnRH-II might modify motor reproductive ac-tivity, whereas sbGnRH provoked the release of gonado-trophins (White et al., 1995). The results obtained in thisstudy might support these considerations becausesGAP-ir fibers were abundant in visual and gustatorysensory areas; cIIGAP-ir fibers represented the only irfibers detected in the cerebellum and innervated stronglythe medulla oblongata and spinal cord, whereas sbGAP-iraxons constituted the main hypophysiotrophic source ofGnRH. Nevertheless, the presence of sGAP-ir fibers in thepars distalis of the sea bass pituitary suggests a directeffect of sGnRH on adenohypophyseal functions. Theseeffects could be exerted directly on gonadotrophic cells toregulate the gonadotrophin secretion and reproductiveprocess. However, a direct action of sGnRH on other seabass pituitary cells can not be neglected because sGnRHwas reported to induce the release of growth hormone(Marchant et al., 1989) and prolactin (Weber et al., 1997)in teleost.

In sea bass, a previous immunohistochemical studyshowed the presence of varicose GnRH-ir fibers contactingand surrounding GnRH-ir cell bodies and dendrites, sug-gesting the existence of connections between differentGnRH cell groups (Kah et al., 1991). A similar observationis reported in this study, in which sGnRH and sbGnRHcells and fibers coexist in the same brain areas and seemto establish contacts. This is especially evident in theolfactory bulbs of sea bass, where varicose sbGAP-ir fibersappear surrounding the large perykaria of sGnRH cells. Inmammals, the presence of GnRH endings synapsing onGnRH perykaria has been well documented (Leranth etal., 1985; Witkin and Silverman, 1985; Witkin, 1999).Although ultrastructural studies are needed before con-


cluding as for the existence of synaptic connections be-tween different GnRH cells in sea bass brain, these obser-vations may support this assumption. Whether suchinteractions represent some way of influencing the activ-ity of other GnRH cells, or synchronizing the differentGnRH cell populations, is not known. Nevertheless, theoverlapping in the expression of sGnRH and sbGnRH cellsin the same regions, including the preoptic area, the co-existence of sGAP-ir and sbGAP-ir fibers in the proximalpars distalis of the sea bass pituitary, and the putativeexistence of direct interactions between both GnRH cellpopulations, suggest that sGnRH and sbGnRH could co-operate closely in the synchronization of developmentaland/or reproductive events.

Surprisingly, the hypophysiotrophic GnRH form (sb-GnRH) of perciforms is less potent than cGnRH-II andsGnRH in inducing gonadotrophin (Zohar et al., 1995) orprolactin (Weber et al., 1997) secretion. However, the highlevels of sbGnRH in the pituitary may compensate for itslower bioactivity. Furthermore, its reduced effectivenesson gonadotrophin secretion could also have a physiologicalrelevance by contributing to a significant response of pi-tuitary to GnRH when its levels increase in a significantmanner. Thus, the lower activity of sbGnRH could permitthe release of eggs in batches during an extended period oftime because only fully mature eggs ovulate in response tosbGnRH stimulation. The lower potency of sbGnRH maysuggest a faster enzymatic breakdown and/or more rapidclearance from the circulation. Also, this evidence could bethe consequence of a more recent emergence of sbGnRHand a lower affinity of this form for the pituitary GnRHreceptor, which might have evolved more slowly that itsligand. In fact, GnRH receptors are very similar in eelsand sea bass (80% of identity) despite the important phy-logenetic distance between these two species (Okubo et al.,2000b; T. Madigou, E. Mananos and O. Kah, unpublisheddata). The existence of at least three GnRHs in a partic-ular species leads to the emerging concept of the existenceof multiple GnRH receptors. Thus, two different GnRHreceptor subtypes are expressed in goldfish (Illing et al.,1999), and three distinct GnRH receptors seem to bepresent in several vertebrate classes (Troskie et al., 1998;Wang et al., 2001). The elucidation of the relative potencyof different GnRH forms to various GnRH receptors couldbe critical to understand their functions. Nevertheless,precise data on the localization of fibers containing GnRHvariants compared to the distribution of GnRH receptorsalso represent a crucial information, because the avail-ability of a given GnRH to a given receptor is determinantfor their real physiologically relevant interactions. In thiscontext, the GnRH receptor subtype 1 of bullfrog is ex-pressed predominantly in the pituitary, whereas GnRHreceptor subtypes 2 and 3 are expressed in the forebrainand hindbrain (Wang et al., 2001). Furthermore, it isrelevant to mention that a recently cloned GnRH receptorfrom trout is poorly expressed in the pituitary, whereas astrong expression was detected in particular brain nuclei(Madigou et al., 2000). Current studies in sea bass aim atdetermining how many GnRH receptors are expressedand at clarifying their respective distributions in the brainand pituitary of this species. As we now have informationon the pattern of projections of the three different GnRHforms expressed in sea bass, such information couldgreatly contribute to improving the knowledge on the pre-cise roles of multiple GnRH forms and receptors in thecontrol of reproduction and other physiological processes.

ACKNOWLEDGMENTS

We thank Dr. M.F. Wullimann, Dr. F.J. Rodrıguez-Gomez and Dr. C. Pendon for their helpful comments andassistance.

LITERATURE CITED

Amano M, Oka Y, Aida K, Okumoto N, Kawashima S, Hasegawa Y. 1991.Immunocytochemical demonstration of salmon GnRH and chickenGnRH-II in the brain of masu salmon, Oncorhynchus masou. J CompNeurol 314:587–597.

Amores A, Force A, Yan YL, Joly L, Amemiya C, Fritz A, Ho RK, LangelandJ, Prince V, Wang YL, Westerfield M, Ekker M, Postlethwait JH. 1998.Zebrafish hox clusters and vertebrate genome evolution. Science 282:1711–1714.

Anglade I, Zandbergen T, Kah O. 1993. Origin of the pituitary innervationin the golfish. Cell Tissue Res 273:345–355.

Bailhache T, Arazam A, Klungland H, Alestrom P, Breton B, Jego P. 1994.Localization of salmon gonadotropin-releasing hormone mRNA andpeptide in the brain of Atlantic salmon and rainbow trout. J CompNeurol 347:444–454.

Batten TF, Cambre ML, Moons L, Vandesande F. 1990. Comparativedistribution of neuropeptide-immunoreactive systems in the brain ofthe green molly, Poecilia latipinna. J Comp Neurol 302:893–919.

Bergquist H. 1932. Zur morphologie des zwischenhirns bei niederen wir-beltieren. Acta Zool 13:57–304.

Braford Jr MR, Northcutt RG. 1983. Organization of the diencephalon andpretectum of the ray-finned fishes. In: Northcutt RG, Davis RE, editors.Fish Neurobiology. Vol. 2 Ann Arbor: University of Michigan Press. p117–163.

Breton B, Weill C, Jalabert B, Billard R. 1972. Activite reciproque desfacteurs hypothalamiques de belier (Ovis aries) et de poissons teleos-teens sur la secretion in vitro des hormones gonadotropes c-HG et LHrespectivement par des hypophyses de carpe et de belier. CR Acad Sci(III) 264:2530–2533.

Brickner RM. 1929. A description and interpretation of certain parts of theteleostean midbrain and thalamus. J Comp Neurol 47:225–282.

Carolsfeld J, Powell JF, Park M, Fischer WH, Craig AG, Chang JP, RivierJE, Sherwood NM. 2000. Primary structure and function of threegonadotropin-releasing hormones, including a novel form, from an an-cient teleost, herring. Endocrinology 141:505–512.

Cerda-Reverter JM, Zanuy S, Munoz-Cueto JA. 2001a. Cytoarchitectonicstudy of the brain of a perciform species, the sea bass (Dicentrarchuslabrax). I. The telencephalon. J Morphol 247:217–228.

Cerda-Reverter JM, Zanuy S, Munoz-Cueto JA. 2001b. Cytoarchitectonicstudy of the brain of a perciform species, the sea bass (Dicentrarchuslabrax). II. The diencephalon. J Morphol 247:229–251.

Crosby EC, Woodburne RT. 1940. The comparative anatomy of the preopticarea and the hypothalamus. Res Publ Assoc Nerv Ment Dis 20:52–169.

De Pinna MCC. 1996. Teleostean Monophyly. In: Stiassny MLJ, ParentiLR, Johnson GD, editors. Interrelationships of fishes. New York: Aca-demic Press. p 147–162.

Gonzalez-Martınez D, Madigou T, Zmora N, Anglade I, Zanuy S, Zohar Y,Elizur A, Munoz-Cueto JA, Kah O. 2001. Differential expression ofthree different prepro-GnRH (gonadotrophin-releasing hormone) mes-sengers in the brain of the European sea bass (Dicentrarchus labrax).J Comp Neurol 429:144–155.

Gonzalez-Martınez D, Zmora N, Zanuy S, Sarasquete C, Elizur A, Kah O,Munoz-Cueto JA. 2002. Developmental expression of three differentprepro-GnRH (Gonadotrophin-releasing hormone) messengers in thebrain of the European sea bass (Dicentrarchus labrax). J Chem Neu-roanat (in Press.)

Goos HJ, de Leeuw R, de Zoeten-Kamp C, Peute J, Blahser S. 1985.Gonadotropin-releasing hormone-immunoreactive neuronal structuresin the brain and pituitary of the African catfish, Clarias gariepinus(Burchell). Cell Tissue Res 241:593–596.

Gothilf Y, Munoz-Cueto JA, Sagrillo CA, Selmanoff M, Chen TT, Elizur A,Kah O, Zohar Y. 1996. Three forms of gonadotrophin-releasing hor-mone in a teleost fish: cDNA characterization and brain localization.Biol Reprod 55:636–645.

Gothilf Y, Meiri I, Elizur A, Zohar Y. 1997. Preovulatory changes in thelevels of three gonadotropin-releasing hormone-encoding messengerribonucleic acids (mRNAs), gonadotropin beta-subunit mRNAs, plasmagonadotropin, and steroids in the female gilthead seabream, Sparusaurata. Biol Reprod 57:1145–1154.


Grober MS, Bass AH. 1991. Neuronal correlates of sex/role changes in labridfishes: LHRH-like immunoreactivity. Brain Behav Evol 38:302–312.

Herrick CJ. 1948. The brain of the tiger salamander. Chicago: Univiversityof Chicago Press.

Holland MCH, Gothilf Y, Meiri I, King JA, Okuzawa K, Elizur A, Zohar Y.1998. Levels of the native forms of GnRH in the pituitary of thegilthead seabream, Sparus aurata, at several characteristic stages ofthe gonadal cycle. Gen Comp Endocrinol 112:394–405.

Holmgren N. 1920. Zur anatomie und histology des vorder- und zwischen-hirns der knochenfische. Acta Zool 1:137–315.

Illing N, Troskie BE, Nahorniak CS, Hapgood JP, Peter RE, Millar RP.1999. Two gonadotropin-releasing hormone receptor subtypes with dis-tinct ligand selectivity and differential distribution in brain and pitu-itary in the goldfish (Carassius auratus). Proc Natl Acad Sci U S A96:2526–2531.

Kah O, Breton B, Dulka JG, Nunez-Rodrıguez J, Peter RE, Corrigan A,Rivier JE, Vale WW. 1986. A reinvestigation of the GnRH(gonadotrophin-releasing hormone) systems in the goldfish brain usingantibodies to salmon Gn-RH. Cell Tissue Res 244:327–337.

Kah O, Zanuy S, Mananos E, Anglade I, Carrillo M. 1991. Distribution ofsalmon gonadotrophin-releasing hormone in the brain and pituitary ofthe sea bass (Dicentrarchus labrax). Cell Tissue Res 266:129–136.

Kim MH, Oka Y, Amano M, Kobayashi M, Aida K, Kawashima S. 1993.Immunocytochemical localization of sGnRH and cGnRH-II in the gold-fish brain and effects of olfactory tract section on the distribution ofGnRH. Proc J Soc Comp Endocrinol 8:11.

King JA, Millar RP. 1982. Structure of chicken hypothalamic luteinizinghormone-releasing hormone II. Isolation and characterization. J BiolChem 257:10729–10732.

King JA, Dufour S, Fontaine YA, Millar RP. 1990. Chromatographic andimmunological evidence for mammalian GnRH and chicken GnRH II ineel (Anguilla anguilla) brain and pituitary. Peptides 11:507–514.

Leranth C, Segura LM, Palkovits M, MacLusky NJ, Shanabrough M, NaftolinF. 1985. The LH-RH-containing neuronal network in the preoptic area ofthe rat: demonstration of LH-RH-containing nerve terminals in synapticcontact with LH-RH neurons. Brain Res 345:332–336.

Lescheid DW, Terasawa E, Abler LA, Urbanski HF, Warby CM, Millar RP,Sherwood NM. 1997. A second form of gonadotropin-releasing hormone(GnRH) with characteristics of chicken GnRH-II is present in theprimate brain. Endocrinology 138:5618–5629.

Madigou T, Mananos-Sanchez E, Hulshof S, Anglade I, Zanuy S, Kah O.2000. Cloning, tissue distribution, and central expression of thegonadotropin-releasing hormone receptor in the rainbow trout. BiolReprod 63:1857–1866.

Marchant TA, Chang JP, Nahorniak CS, Peter RE. 1989. Evidence thatgonadotropin-releasing hormone also functions as a growth hormone-releasing factor in the goldfish. Endocrinology 124:2509–2518.

Matsuo H, Baba Y, Nair RMG, Arimura A, Schally AV. 1971. Structure ofthe porcin LH- and FSH-releasing hormone. I. The proposed amino acidsequences. Biochem Biophys Res Commun 43:1334–1339.

Meader RG. 1934. The optic system of the teleost, Holocentrus. J CompNeurol 60:361–407.

Millar RP, King JA. 1994. Plasticity and conservation in gonadotropins-releasing hormone structure and function. In Davey KG, Peter RE,Tobe SS, editors. Perspectives in comparative endocrinology. Ottawa:National Research Council of Canada. p 129–136.

Miller KE, Kriebel RM. 1985.Electron-microscopic identification of mesen-cephalic neurosecretory cells in teleosts. Cell Tissue Res 242:445–448.

Miller KE, Kriebel RM. 1986a. Peptidergic innervation of caudal neurose-cretory neurons. Gen Comp Endocrinol 64:396–400.

Miller KE, Kriebel RM. 1986b. Cytology of brain stem neurons projecting tothe caudal neurosecretory complex: an HRP-electron microscopicstudy. Brain Res Bull 16:183–188.

Miyamoto K, Hasegawa I, Nomura M, Igarashi M, Kangawa K, Matsuo H.1984. Identification of the second gonadotropin-releasing hormone inchicken hypothalamus: evidence that gonadotropin secretion is proba-bly controlled by two distinct gonadotropin-releasing hormones inavian species. Proc Natl Acad Sci U S A 81:3874–3878.

Montaner AD, Somoza GM, King JA, Bianchini JJ, Bolis CG, Affanni JM.1998. Chromatographic and immunological identification of GnRH(gonadotropin-releasing hormone) variants. Occurrence of mammalianand a salmon-like GnRH in the forebrain of an eutherian mammal: Hy-drochaeris hydrochaeris (Mammalia, Rodentia). Regul Pept 73:197–204.

Montaner AD, Affanni JM, King JA, Bianchini JJ, Tonarelli G, SomozaGM. 1999. Differential distribution of gonadotropin-releasing hormone

variants in the brain of Hydrochaeris hydrochaeris (Mammalia, Roden-tia). Cell Mol Neurobiol 19:635–651.

Montaner AD, Park MK, Fischer WH, Craig AG, Chang JP, Somoza GM,Rivier JE, Sherwood NM. 2001. Primary structure of a novelgonadotropin-releasing hormone in the brain of a teleost, pejerrey.Endocrinology 142:1453–1460.

Montero M, Vidal B, Vandesande F, King JA, Tramu G, Dufour S, Kah O.1994. Comparative distribution of mammalian GnRH (gonadotrophin-releasing hormone) and chicken GnRH-II in the brain of the Europeaneel. J Chem Neuroanat 7:227–241.

Montero M, Le Belle N, King JA, Millar RP, Dufour S. 1995. Differentialregulation of the two forms of gonadotropin-releasing hormone(mGnRH and cGnRH-II) by sex steroids in the European female silvereel (Anguilla anguilla). Neuroendocrinology 61:525–535.

Munz HP, Stumpf WE, Jennes L. 1981. LHRH systems in the brain ofplatyfish. Brain Res 221:1–13.

Muske LE. 1993. Evolution of gonadotropin-releasing hormone (GnRH)neuronal systems. Brain Behav Evol 42:215–230.

Muske LE, Moore FL. 1994. Antibodies against different forms of GnRHdistinguish different populations of cells and axonal pathways in aurodele amphibian, Taricha granulosa. J Comp Neurol 345:139–147.

Navas JM, Anglade I, Bailhache T, Pakdel F, Breton B, Jego P, Kah O.1995. Do gonadotrophin-releasing hormone neurons express estrogenreceptors in the rainbow trout? A double immunohistochemical study.J Comp Neurol 363:461–474.

Ngamvongchon S, Lovejoy DA, Fischer WH, Craig AG, Nahorniak CS,Sherwood NM. 1992. Primary structure of two forms of gonadotropin-releasing hormone, one distinct and one conserved, from catfish brain.Mol Cell Neurosci 3:17–22.

Norgren RB Jr, Gao C. 1994. LHRH neuronal subtypes have multipleorigins in chickens. Dev Biol 165:735–738.

Oka Y, Matsushima T. 1993. Gonadotrophin-releasing hormone (GnRH)-immunoreactive terminal nerve cells have intrinsic rhythmicity andproject widely in the brain. J Neurosci 13:2161–2176.

Okubo K, Amano M, Yoshiura Y, Suetake H, Aida K. 2000a. A novel formof gonadotropin-releasing hormone in the medaka, Oryzias latipes.Biochem Biophys Res Commun 276:298–303.

Okubo K, Suetake H, Usami T, Aida K. 2000b. Molecular cloning andtissue-specific expression of a gonadotropin-releasing hormone receptorin the Japanese eel. Gen Comp Endocrinol 119:181–192.

Okuzawa K, Granneman J, Bogerd J, Goos HJTh, Zohar Y, Kagawa H.1997. Distinct expression of GnRH genes in the red seabream brain.Fish Physiol Biochem 17:71–79.

Ookura T, Okuzawa K, Tanaka H, Gen K, Kagawa H. 1999. The ontogenyof gonadotrophin-releasing hormone neurons in the red seabream. In:Abstracts of the VI International Symposium on Reproductive Physi-ology of Fish. Bergen: University of Bergen. p 13.

Parhar IS. 1997. GnRH in tilapia: three genes, three origins and theirroles. In: Parhar IS, Sakuma Y, editors. GnRH neurons: gene to be-havior. Chapter 5. Tokyo: Brain Shuppan. p 99–122.

Parhar IS. 1999. Hormonal regulation of GnRH gene expression. In: Ab-stracts of the VI International Symposium on Reproductive Physiologyof Fish. Bergen: University of Bergen. p 12.

Parhar IS, Soga T, Ishikawa Y, Nagahama Y, Sakuma Y. 1998. Neuronssynthesizing gonadotropin-releasing hormone mRNA subtypes havemultiple developmental origins in the medaka. J Comp Neurol 401:217–226.

Polkowska J, Przekop F. 1993. Effect of protein deficiency on luteinizinghormone releasing hormone (LHRH), gonadotropin releasing hormoneassociated peptide (GAP) and luteinizing hormone (LH) immunocyto-chemistry in the hypothalamus and pituitary gland of prepubertalewes. Exp Clin Endocrinol 101:230–237.

Powell JF, Zohar Y, Elizur A, Park M, Fischer WH, Craig AG, Rivier JE,Lovejoy DA, Sherwood NM. 1994. Three forms of gonadotropin-releasing hormone characterized from brains of one species. Proc NatlAcad Sci U S A 91:12081–12085.

Powell JF, Reska-Skinner S, Prakash MO, Fisher WH, Park M, Rivier JE,Craig AG, Mackie GO, Sherwood NM. 1996. Two new forms ofgonadotropin-releasing hormone in a protochordate and the evolution-ary implications. Proc Natl Acad Sci U S A 93:10461–10464.

Powell RC, King JA, Millar RP. 1985. [Trp7,Leu8]LH-RH in reptilianbrain. Peptides 6:223–227.

Quanbeck C, Sherwood NM, Millar RP, Terasawa E. 1997. Two popula-tions of luteinizing hormone-releasing hormone neurons in the fore-brain of the rhesus macaque during embryonic development. J CompNeurol 380:293–309.


Rodriguez L, Carrillo M, Sorbera LA, Soubrier MA, Mananos E, Holland MC,Zohar Y, Zanuy S. 2000. Pituitary levels of three forms of GnRH in themale European sea bass (Dicentrarchus labrax, L.) during sex differenti-ation and first spawning season. Gen Comp Endocrinol 120:67–74.

Rodriguez-Gomez FJ, Rendon MC, Sarasquete C, Munoz-Cueto JA.1999.Distribution of gonadotropin-releasing hormone immunoreactive sys-tems in the brain of the Senegalese sole, Solea senegalensis. HistochemJ 31:695–703.

Ronchi E, Aoki C, Krey LC, Pfaff DW. 1992. Immunocytochemical study ofGnRH and GnRH-associated peptide in male Syrian hamsters as afunction of photoperiod and gonadal alterations. Neuroendocrinology55:134–145.

Schnitzlein HN. 1962. The habenula and the dorsal thalamus of someteleosts. J Comp Neurol 118:225–267

Schulz RW, Bosma PT, Zandbergen MA, Van der Sanden MC, Van Dijk W,Peute J, Bogerd J, Goos HJTh. 1993. Two gonadotropin-releasing hor-mones in the African catfish, Clarias gariepinus: localization, pituitaryreceptor binding, and gonadotropin release activity. Endocrinology 133:1569–1577.

Schwanzel-Fukuda M. 1999. Origin and migration of luteinizing hormone-releasing hormone neurons in mammals. Microsc Res Tech 44:2–10.

Senthilkumaran B, Okuzawa K, Gen K, Ookura T, Kagawa H. 1999. Distri-bution and seasonal variation in levels of three native GnRHs in the brainand pituitary of perciform fish. J Neuroendocrinol 11:181–186.

Sheldon RE. 1912. The olfactory tracts and centers in teleosts. J CompNeurol 22:177–339.

Sherwood N, Eiden L, Brownstein M, Spiess J, Rivier J, Vale W. 1983.Characterization of a teleost gonadotropin-releasing hormone. ProcNatl Acad Sci U S A 80:2794–2798.

Sherwood NM, Zoeller RT, Moore FL. 1986. Multiple forms ofgonadotropin-releasing hormone in amphibian brains. Gen Comp En-docrinol 61:313–322.

Sherwood NM, Lovejoy DA, Coe IR. 1993. Origin of mammaliangonadotropin-releasing hormones. Endocr Rev 14:241–254.

Subhedar N, Krishna NS. 1988. Immunocytochemical localization ofLH-RH in the brain and pituitary of the catfish, Clarias batrachus(Linn). Gen Comp Endocrinol 72:431–444.

Suzuki M, Hyodo S, Kobayashi M, Aida K, Urano A. 1992. Characteriza-tion and localization of mRNA encoding the salmon-type gonad-otrophin-releasing hormone precursor of the masu salmon. J Mol En-docrinol 9:73–82.

Troskie B, Illing N, Rumbak E, Sun YM, Hapgood J, Sealfon S, Conklin D,Millar R. 1998. Identification of three putative GnRH receptor subtypesin vertebrates. Gen Comp Endocrinol 112:296–302.

Urbanski HF, White RB, Fernald RD, Kohama SG, Garyfallou VT,Densmore VS. 1999. Regional expression of mRNA encoding a secondform of gonadotropin-releasing hormone in the macaque brain. Endo-crinology 140:1945–1948.

Wang L, Bogerd J, Choi HS, Seong JY, Soh JM, Chun SY, Blomenrohr M,Troskie BE, Millar RP, Yu WH, McCann SM, Kwon HB. 2001. Threedistinct types of GnRH receptor characterized in the bullfrog. Proc NatlAcad Sci U S A 98:361–366.

Weber GM, Powell JF, Park M, Fischer WH, Craig AG, Rivier JE, Nana-korn U, Parhar IS, Ngamvongchon S, Grau EG, Sherwood NM. 1997.Evidence that gonadotropin-releasing hormone (GnRH) functions as aprolactin-releasing factor in a teleost fish (Oreochromis mossambicus)and primary structures for three native GnRH molecules. J Endocrinol155:121–132.

White RB, Fernald RD. 1998. Genomic structure and expression sites ofthree gonadotropin-releasing hormone genes in one species. Gen CompEndocrinol 112:17–25.

White RB, Eisen JA, Kasten TL, Fernald RD. 1998. Second gene forgonadotropin-releasing hormone in humans. Proc Natl Acad Sci U S A95:305–309.

White SA, Kasten TL, Bond CT, Adelman JP, Fernald RD. 1995. Threegonadotropin-releasing hormone genes in one organism suggest novelrole for an ancient peptide. Proc Natl Acad Sci U S A 92:8363–8367.

Witkin JW. 1999. Synchronized neuronal networks: the GnRH system.Microsc Res Tech 44:11–18.

Witkin JW, Silverman AJ. 1985. Synaptology of luteinizing hormone-releasing hormone neurons in rat preoptic area. Peptides 6:263–271.

Wray S, Grant P, Gainer H. 1989. Evidence that cells expressing luteiniz-ing hormone-releasing hormone mRNA in the mouse are derived fromprogenitor cells in the olfactory placode. Proc Natl Acad Sci U S A86:8132–8136.

Yahalom D, Chen A, Ben-Aroya N, Rahimipour S, Kaganovsky E, Okon E,Fridkin M, Koch Y. 1999. The gonadotropin-releasing hormone familyof neuropeptides in the brain of human, bovine and rat: identification ofa third isoform. FEBS Lett 463:289–294.

Yamamoto N, Oka Y, Amano M, Aida K, Hasegawa Y, Kawashima S. 1995.Multiple gonadotropin-releasing hormone (GnRH) immunoreactivitysystem in the brain of the dwarf gourami, Colisa lalia: Immunohisto-chemistry and radioimmunoassay. J Comp Neurol 355:354–368.

Yamamoto N, Oka Y, Kawashima S. 1997. Lesions of gonadotropin-releasing hormone-immunoreactive terminal nerve cells: effects on thereproductive behavior of male dwarf gouramis. Neuroendocrinology65:403–412.

Yamamoto N, Parhar IS, Sawai N, Oka Y, Ito H. 1998a. Preopticgonadotropin-releasing hormone (GnRH) neurons innervate the pitu-itary in teleosts. Neurosci Res 31:31–38.

Yamamoto N, Oka Y, Yoshimoto M, Sawai N, Albert JS, Ito H.1998b.Gonadotropin-releasing hormone neurons in the gourami midbrain: adouble labeling study by immunocytochemistry and tracer injection.Neurosci Lett 240:50–52.

Yu KL, Sherwood NM, Peter RE. 1988. Differential distribution of twomolecular forms of gonadotropin-releasing hormone in discrete brainareas of goldfish (Carassius auratus). Peptides 9:625–630.

Zandbergen MA, Kah O, Bogerd J, Peute J, Goos HJTh. 1995. Expressionand distribution of two gonadotropin-releasing hormones in the catfishbrain. Neuroendocrinology 62:571–578.

Zmora N, Rosenfeld H, Meiri I, Gothilf Y, Elizur A, Zohar Y. 2000. Expres-sion of GnRH in the pituitary of gilthead seabream. In: Abstract of the4th International Symposium on Fish Endocrinology. Seattle: Univer-sity of Washington. p 75.

Zmora N, Gonzalez-Martınez D, Munoz-Cueto JA, Madigou T, Mananos-Sanchez E, Doste SZ, Zohar Y, Kah O, Elizur A. 2002. The GnRHsystem in the European sea bass, Dicentrarchus labrax: Cloning of thethree GnRH precursors and the production of antibodies against re-combinant GAP’s for the localization of preproGnRH expression in thebrain by means of immunohistochemistry. J Endocrinol 172:105–116.

Zohar Y, Elizur A, Sherwood NM, Powell JF, Rivier JE, Zmora N. 1995.Gonadotropin-releasing activities of the three native forms ofgonadotropin-releasing hormone present in the brain of gilthead sea-bream, Sparus aurata. Gen Comp Endocrinol 97:289–299.


immunohistochemical localization of three different prepro-gnrhs in the brain and pituitary of the...

Documents