General and Comparative Endocrinology 145 (2006) 263–269

www.elsevier.com/locate/ygcen

Seasonal changes in expression of LH-� and FSH-� in male and female three-spined stickleback, Gasterosteus aculeatus

Anna Hellqvist a,¤, Monika Schmitz b, Ian Mayer c, Bertil Borg a

a Department of Zoology, Stockholm University, S-106 91 Stockholm, Swedenb Department of Aquaculture, Swedish University of Agricultural Sciences, S-90183 Umea, Sweden

c Department of Biology, University of Bergen, N-5020 Bergen, Norway

Received 26 April 2004; revised 26 August 2005; accepted 6 September 2005Available online 11 November 2005

Abstract

In teleost Wshes, like in other vertebrates, the gonadal development is stimulated by two gonadotropic hormones; luteinizing hormone(LH) and follicle-stimulating hormone (FSH). To achieve a better understanding of the role of gonadotropins in teleost reproduction;expression of LH-� and FSH-� mRNA and the status of gonads and secondary sexual characters were analyzed over the annual cycle inmale and female three-spined sticklebacks, a species in which the development of male secondary sexual characters and spermatogenesisare separated in time. The kidney in the male stickleback hypertrophies during the breeding season and produces a glue used when build-ing nests. Kidney weights, as well as levels of 11-ketotestosterone (11KT), reached a peak in May. Both testosterone (T) levels and thegonadosomatic index (GSI, gonad weight/body weight £ 100) in females started to increase in April, and peaked in May as well. Later insummer, after the breeding season, these features declined. In females, LH-� expression followed the GSI and T levels closely, levels werelow during winter and early spring, increased to a peak in late May and declined to low levels again in July. FSH-� expression peaked ear-lier, in January and declined slowly over spring. In males, LH-� expression peaked in May. During June-September, when spermatogene-sis was active, LH-� levels were very low. FSH-� expression peaked in January, earlier than LH-� expression did, and reached the lowestlevels in July. Thus, when spermatogenesis started at the end of summer, the expression of both GTH-� mRNAs, and circulating 11KT,displayed their lowest levels. 2005 Elsevier Inc. All rights reserved.

Keywords: Stickleback; Seasonal; Expression; mRNA; Gonadotropin; GTH; Luteinizing; Follicle-stimulating; Hormone; LH; FSH

1. Introduction

In teleost Wshes, like in other vertebrates, the gonads arestimulated by two gonadotropic hormones secreted fromthe pituitary; luteinizing hormone (LH) and follicle-stimu-lating hormone (FSH) (Kawauchi et al., 1989). The sea-sonal pattern of LH and FSH in teleosts diVer betweendiVerent species. In salmonids for instance, FSH plasmalevels predominate during spermatogenesis and vitellogene-sis, whereas LH levels are high during spermiation in malesand during Wnal oocyte maturation and ovulation in

* Corresponding author. Fax: +46 08 167095.E-mail address: ahellqvist@nl.rogers.com (A. Hellqvist).

0016-6480/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ygcen.2005.09.012

females (Gomez et al., 1999; Prat et al., 1996). In addition,plasma levels of FSH in males are correlated with levels ofandrogens, and in females with estrogens, whereas circulat-ing levels of LH are correlated with those of 17�,20�-dihy-droxy-4-pregnen-3-one (17,20-P) in both sexes (Swanson,1991). Salmonids only develop one batch of eggs per sea-son, and vitellogenesis is only active in the prespawningperiod. In other species, such as the goldWsh (Carassiusauratus, Sohn et al., 1999; Yoshiura et al., 1997), blue gou-rami (Trichogaster trichopterus, Jackson et al., 1999) andJapanese Xounder (Paralichthys olivaceus, Kajimura et al.,2000) who are all repeat spawners, both FSH-� and LH-�mRNA levels increase during their spawning season.

The stickleback, a repeat spawner, has a marked sea-sonal reproductive cycle in which spermatogenesis and

264 A. Hellqvist et al. / General and Comparative Endocrinology 145 (2006) 263–269

the development of androgen-dependent secondary sex-ual characters in the male are temporally separated (e.g.,Borg, 1982). During the breeding season in late spring/early summer, the male stickleback displays distinct sec-ondary sexual characters such as red breeding colours anda hypertrophied kidney that secretes a glue, consisting ofthe 11-ketotestosterone (11KT) induced protein spiggin(Jakobsson et al., 1999; Jones et al., 2001), which is used tobuild the nest. At the end of the breeding period, when theandrogen levels drop (Mayer et al., 1990a) and secondarysexual characters regress, spermatogenesis starts. Andro-gen treatment inhibits the postbreeding commencement ofspermatogenesis in the stickleback (Andersson et al.,1988). Ovarian weights start to increase in late spring, andreach a peak during the breeding season (Borg and vanVeen, 1982). Since the female stickleback may spawn up to15 times during a breeding season (Baggerman, 1957),each time laying 10–20% her weight in roe, vitellogenesismust continue to be active in this period.

The aim of this study was to understand the roles of thetwo GTHs in stickleback reproduction. To that end, expres-sion of LH-� and FSH-� mRNA were analyzed in maleand female sticklebacks sampled every month for a year.To determine the reproductive state, testes were analyzedhistologically, and ovaries and male kidneys were weighed.In addition, plasma levels of 11KT and testosterone (T)were determined from March to October.

2. Materials and methods

2.1. Animals

Adult, non-breeding three-spined sticklebacks, werecaught in the Öresund on the 7–11 November 2001 andtransferred to a 1200 L aquarium. Since there were too fewfemales in the Wrst transfer, more were added on December19, the day after the second sampling. These females werecaught in the Öresund on the 15–16 December 2001. Thephotoperiod was changed four times a month and tempera-ture was changed as needed to follow the natural pattern inÖresund (Fig. 1). The aquarium contained periodicallychanged brackish water (0.5% salinity) which was con-

stantly Wltered and aerated, and the bottom was coveredwith sand. The Wsh were fed daily with frozen red midge lar-vae or mysids. Every month ten Wsh of each sex were anaes-thetized with 2-phenoxyethanol, weighed and dissected.The presence of running roe and male breeding colourswere noted. Plasma was sampled from March to October.The caudal peduncle was severed and blood collected inHematocrite tubes. Following centrifugation, plasma fromeach sex was pooled into Eppendorf tubes. The pituitarieswere collected in Eppendorf tubes and immediately frozenin liquid nitrogen. Both plasma and pituitaries were storedat ¡70 °C until analysis. The ovaries and male kidneys wereexcised and weighed, and kidneysomatic index (KSI, kid-ney weight/body weight £ 100) and gonadosomatic index(GSI, gonad weight/body weight £ 100) were calculated.Testes were Wxed for histology. Testes were not weighed,since testes weights are low and are no good indicators ofsexual maturation in the male stickleback (Borg, 1982).

2.2. Histology

Testes were Wxed in Bouin–Hollande’s solution, dehy-drated, embedded in Histosec (Merck), sectioned at 6 �mand stained with Ehrlich’s haematoxylin and eosin. Thepresence and amounts of spermatogonia, spermatocytes,spermatids, and spermatozoa were studied semi-quantita-tively and randomly on an entire mid-longitudinal section.Testes were classiWed into the following four categories: (1)quiescent testes containing no germ cells apart from sper-matozoa and small numbers of spermatogonia. (2) Com-mencement of spermatogenesis after breeding season,spermatogonia dominate. (3) Active spermatogenesis withall stages being present in large numbers. (4) Active sper-matogenesis, with large numbers of spermatocytes, sperm-atides, and spermatozoa, but only few spermatogonia.

2.3. Steroid measurement

Plasma levels of T and 11KT were measured by meansof radioimmunoassay (RIA) as described by Páll et al.(2002), in pooled plasma samples of 10 males and 10females, respectively. Prior to RIA, plasma samples were

Fig. 1. Simulated natural photoperiod and temperature during the experiment.

A. Hellqvist et al. / General and Comparative Endocrinology 145 (2006) 263–269 265

diluted with RIA buVer, and heat-treated for 1 h at 80°.After centrifugation, the supernatant was drawn oV andkept at 10 °C during the period of steroid determination.

2.4. Dot-blot analysis of LH-� and FSH-� expression

Total RNA was extracted from single pituitaries usingTrizol reagent (Invitrogen) following the manufacturer’sinstructions, denatured and blotted on Hybond-N+, Nylonmembrane (Amersham Pharmacia Biotech) through aHybri-blot manifold (Life Technologies). The membraneswere allowed to dry and then UV-crosslinked.

The membranes were hybridized using cRNA probesbased on stickleback LH-� and FSH-� (Accession No.AJ534969 and AJ534871, respectively. Hellqvist et al., 2004).The speciWcity of these probes have been earlier validated byNorthern blot analysis where total pituitary RNA hybrid-ized with labelled FSH-� or LH-� revealed single bands.

The antisense probes were synthesized by in vitro tran-scription using SP6 or T7 RNA polymerase, respectively, andlabeled with [�-32P]UTP according to the Strip-EZ RNA kit(Ambion). The blots were prehybridized for 1h at 65°C, withULTRAhyb hybridization solution (Ambion) and then incu-bated overnight with the radioactive probe (1£106 cpm/mlhybridization solution). The blots were washed twice in 1£SSC, 0.1% SDS at 65°C for 15min each and then once for15min at 65°C and once for 30min at 68°C in 0.1£ SSC,0.1% SDS. The hybridization signals of the blots werecounted and analysed using a biophosphorimager (Bio-Rad).The membranes were stripped according to the Strip-EZ pro-tocols and subsequently hybridized with a 32P-labeled stickle-back 18S rRNA probe to correct for unequal RNA loadings.

2.5. Statistical analysis

The data were analyzed using STATISTICA ’99 edition(StatSoft). Unless otherwise stated, the data were analyzedusing one-way ANOVA for multiple comparisons and

Tukey honest signiWcant diVerence (post hoc) test for com-parison between pair of groups. Prior to analysis, the datawere log transformed.

3. Results

Males started showing breeding colours in April, withhigher intensity in May-June. The kidneysomatic indexstarted to increase in March, from less than 0.63 in Novem-ber–February, up to 0.87 § 0.10 (means § SEM). In April itwas 1.83 § 0.37, and reached a peak at 2.43 § 0.28 in May, a4-fold increase compared to non-breeding animals. In June,KSI started to decline to 1.87 § 0.27, and in July–Octobermonthly means were below 0.75 (Fig. 2). During Novem-ber–May all males showed quiescent spermatogenesis(stage 1), with no germ cells apart from spermatozoa andsmall numbers of spermatogonia. In June some malesstarted to show commencing (2) or even active spermato-genesis (3), and in July–October the majority of the malesdisplayed active spermatogenesis (see Table 1). The 11KTlevels increased in April, from 5.4 ng/ml in March to12.3 ng/ml, reached a peaked in May, 23.8 ng/ml, anddeclined following breeding, down to 2–3 ng/ml (see Table2). T levels were always low.

Seasonal changes in both LH-� and FSH-� expressionsin males were signiWcant (p < 0.001 for both). LH � expres-sion increased signiWcantly (p < 0.001) from December(0.63 § 0.07 arbitrary units) to January (3.73 § 0.68). InFebruary–March expression remained around this level.LH � expression reached a peak in May (6.06 § 0.69),dropped signiWcantly in June (0.30 § 0.09, p < 0.001) andremained low at less than 0.4 arbitrary units during July–October. FSH-� expression in males increased signiWcantly(p < 0.001) between December (0.83 § 0.10) and January(4.75 § 1.01) where it reached a peak. From February–Maylevels were ca. 3 arbitrary units followed by a signiWcant(p < 0.001) decrease from May down to 0.45 § 0.12 in June,and further down to ca. 0.2 in July–August. A slight

Fig. 2. Seasonal cycle of LH-� and FSH-� mRNA expression, and of kidney somatic index (KSI) in stickleback males, means § SEM. N D 10. For statis-tics, see results.

266 A. Hellqvist et al. / General and Comparative Endocrinology 145 (2006) 263–269

increase occurred in September (0.64 § 0.10) and also inOctober (1.17 § 0.17) (Fig. 2).

In females, GSI was below 4 in November–March, andincreased in April to 9.34 § 2.24. GSI peaked in May, at11.45 § 1.42, a 3-fold increase compared to non-breedinganimals. During June GSI started to decline, 6.59 § 1.09,and from August to October onwards it was below 4 again(Fig. 3). Ripe females with swollen bellies were observed inthe aquarium April–June, but not during other times. A few

Table 1Annual cycle of spermatogenesis

(1) Spermatozoa and only a few spermatogonia.(2) Large number of spermatogonia.(3) All germ cell stages present in large numbers.(4) Large number of spermatocytes, spermatids, and spermatozoa, butonly few spermatogonia. Frequencies shown.

Sample date Stage of spermatogenesis

1 2 3 4

November 27 10December 18 10January 23 10February 28 10March 27 10April 29 10May 24 10June 19 5 4 1July 22 3 2 5August 19 2 7 1September 23 1 7 2October 28 4 6

Table 2Plasma levels of 11KT and T in males

Levels were measured in pooled plasma from 10 males at each sampling.

Sample date 11KT (ng/ml) T (ng/ml)

March 27 5.4 1.1April 29 12.3 2.6May 24 23.8 3.4June 19 3.5 1.0July 22 2.5 0.8August 19 2.1 ndSeptember 23 1.8 0.6October 28 1.9 nd

sampled females had running ripe roe in April–June (two inApril, one in May and one in June) but not during othertimes. The 11KT levels were always low whereas levels of Treached a peak during breeding season, from around 1 ng/ml in spring to 22.9 ng/ml in May (see Table 3).

Also in females, there were signiWcant seasonal changesin both LH-� and FSH-� expression (p < 0.001 in both).LH-� expression followed the GSI very closely, with verylow levels during November–February (less than 0.3 arbi-trary units) and an increase between March (0.54 § 0.15)and April (1.58 § 0.32). Expression reached a peak in May,(2.29 § 0.52) and decreased signiWcantly between May andJune (1.12 § 0.67, p < 0.005). LH-� expression remained lowthroughout July–October (less than 0.7). FSH-� expressionwere 1.05 § 0.10 in November, and increased from Decem-ber (0.60 § 0.07) to January when it reached a peak at2.02 § 0.39. First in June (0.61 § 0.25), levels were signiW-

cantly lower than in January (p < 0.05). Expression declinedfurther to less than 0.2 arbitrary units in July–August. InSeptember–October there was a slight increase again, up toca. 1 (Fig. 3). Levels signiWcantly higher (p < 0.03) than inJuly were reached Wrst in October.

Although the seasonal patterns in male and femaleGTH-expression displayed similarities (LH-� expressionpeaking in May and FHS-� peaking in January in bothsexes), there were also signiWcant diVerences (two-wayANOVA interaction between sex and month, LH: p < 0.001and FSH: p < 0.001). Males had generally higher expres-sions of both GTH-� (p < 0.001), compared to females. LH-

Table 3Plasma levels of 11KT and T in females

Levels were measured in pooled plasma from 10 females at each sampling.

Sample date 11KT (ng/ml) T (ng/ml)

Mar 27 1.2 0.8April 29 1.6 1.2May 24 3.0 22.9June 19 2.1 15.6July 22 1.4 1.6August 19 0.8 ndSeptember 23 1.0 0.5October 28 0.8 nd

Fig. 3. Seasonal cycle of LH-� and FSH-� mRNA expression, and of gonadosomatic index (GSI), in female sticklebacks, means § SEM. N D 10. Forstatistics, see results.

A. Hellqvist et al. / General and Comparative Endocrinology 145 (2006) 263–269 267

� expression started rising earlier in males than in females,in January and in March–April, respectively.

4. Discussion

Female GSI peaked in late May, and declined markedlyin July–August. In males KSI reached a peak in late Maythat diminished after the breeding season in August. Sper-matogenesis was quiescent in all males during winter andspring. In June males started to show commencing or evenactive spermatogenesis, and in July–October the majorityof the males displayed active spermatogenesis. The presentresults followed a similar pattern as found in Weld-caughtsticklebacks from the Öresund (Borg and van Veen, 1982and Borg, 1982), although the breeding season occurredsomewhat earlier (peak in May rather than in June) underthe present conditions.

In female sticklebacks, LH-� expression followed theGSI very closely, from low levels during winter and earlyspring, a peak in late May and down to low levels againfrom July. FSH-� expression, in contrast, peaked already inJanuary and declined slowly over spring. Also in maturingfemale ayu, Plecoglossus altivelis, expression of FSH-�peaks earlier than expression of LH-� (Yoshida et al.,2001). When FSH-� peaks in August, gonads have startedto grow but are still small; females are in the primary yolkstage. In September, when LH-� peaks, gonads haveincreased considerably in size and ovaries display second-ary yolk. In female striped bass (Morone saxatilis) FSH-�increased drastically during vitellogenesis, followed by apeak in LH-� expression a few months later (Hassin et al.,1999). In female goldWsh (Carassius auratus) both LH-�and FSH-� mRNA levels increased in parallell towards thespawning period in April–May and decreased duringgonadal regression in summer (Sohn et al., 1999). Femalered seabream had high levels of LH-� mRNA expressionfrom the beginning of early vitellogenesis to the spawningseason, after which levels declined. FSH-� expression, onthe other hand, was always low, but peaked at spawning(Gen et al., 2000). In the female Japanese eel (Anguillajaponica) expression of FSH-� mRNA was high in imma-ture Wsh, and declined during experimental inducedgonadal maturation, similar results have been observed inthe female European eel (Schmitz et al., 2005). LH-� levelsin the two eel species were very low initially and increasedmarkedly with ovarian maturation and ovulation (Schmitzet al., 2005; Suetake et al., 2002). In female rainbow trout(Oncorhynchus mykiss), plasma and pituitary levels of FSHand LH, as well as the expression of FSH-� and LH-�, alldisplayed high levels at full maturation. FSH, however,increased earlier than LH and there was also a peak in FSHplasma levels and expression in vitellogenic Wsh before Wnalmaturation (Gomez et al., 1999). To summarize, althoughexpression of FSH-� often increases earlier than that ofLH-�, the patterns in expression of GTHs vary consider-ably between diVerent species, even between species thathave similar patterns in ovarian development, as in gold-

Wsh, red seabream, and stickleback which are all repeatspawners.

Spermatogenesis was found to be most active in thepost-breeding period in late summer-autumn when theexpression of both GTHs as well as androgen-levels werelow and secondary sexual characters weakly developed.This is consistent with previous studies on sticklebacks.Lower levels of �-FSH and �-LH expression in postbreed-ing stickleback males with active spermatogenesis than inmales in breeding condition were also observed by Hellq-vist et al. (2003). In a study on natural seasonal changes inthe stickleback pituitary, Borg et al. (1988) found the lowestactivity in GTH-cells, as judged by their ultrastructuralappearance, in post-breeding males with active spermato-genesis. A seasonal separation between the appearance ofwell-developed secondary sexual characters, especially kid-ney hypertrophy, and spermatogenesis in sticklebacks wasWrst found by van Oordt (1924) in the nine-spined stickle-back, Pungitius pungitius, and has later been found in sev-eral studies on sticklebacks (e.g., Borg et al., 1988). Whensticklebacks are exposed to diVerent photoperiods and tem-perature in winter, spermatogenesis is active only under acombination of high temperature and short photoperiod(Borg et al., 1987; Hellqvist et al., 2004), where the kidneyepithelium height (KEH), 3-� hydroxysteroid dehydrogen-ease activity in the testes (Borg et al., 1987) and the expres-sion of both LH-� and FSH-� (Hellqvist et al., 2004) islower than in all other treatments. Borg (1981) alsoobserved that spermatogenesis and KEH showed a clearnegative correlation on the individual level. Mayer et al.(1990b) studied seasonal changes in circulating androgensin male sticklebacks and found low levels of all studiedandrogens in the postbreeding period, for 11KT the levelswere lower than at any other time of the year. Further, Borget al. (1989) observed that the androgen production in tes-tes incubated in vitro was low in postbreeding males, thesynthesis of 11-ketoandrogens was less than 2% of thatfound in breeding males.

Adminstration of androgens inhibits spermatogenesis inthe stickleback at the end of the breeding season (Anders-son et al., 1988; Borg, 1981) and under short photoperiodand high temperature in winter (Borg et al., 1986). Borg(1981) and Borg et al. (1986) used high doses of methyltes-tosterone (MT) administrated via the water. Treatmentswith high doses of T or MT (but not with 11-ketoandro-gens) have in several studies been shown to suppress sper-matogenesis and gonadal size in Wshes wherespermatogenesis is active when natural androgen levels arehigh (e.g., rainbow trout, Billard et al., 1981; Atlanticsalmon, Salmo salar, Berglund et al., 1995). This eVect isprobably due to the exogenous androgens inhibiting GTHsecretion, leading to decreased stimulation of androgenproduction and thus to lowered intratesticular androgenlevels. It is not likely that this is the reason for the suppres-sion of spermatogenesis in the stickleback where MT didnot suppress testes weights, but had the opposite eVect(Borg, 1981). Under short photoperiod and high

268 A. Hellqvist et al. / General and Comparative Endocrinology 145 (2006) 263–269

temperature in winter MT treatment suppressed spermato-genesis (Borg et al., 1986) even though it also activated theGTH cells (as judged by their ultrastructural appearance)and the Leydig cells (increased nuclear size). Anderssonet al. (1988) treated stickleback males with Silastic implantscontaining androstenedione or 11-ketoandrostenedione(11KA), at the end of the breeding season. Both androgenssuppressed spermatogenesis, 11KA, which can be convertedto 11KT by blood cells in the stickleback (Mayer et al.,1990a), being the most eVective. Though Andersson et al.(1988) did not measure the circulating levels of androgens,Mayer et al. (1990b) found similar 11KA capsules to resultin 11KT plasma levels of 22 ng/mL in castrated sticklebackmales. This level is similar to the peak level found in May inthe present study, and less than tenth of the levels found interritorial, nesting stickleback males in the sexual phase(Páll et al., 2002) and must be regarded as physiological.

The studies on the relationship between androgens andspermatogenesis in sticklebacks are generally consistentwith each other. They are, however, in marked contrastwith studies on other Wshes, which follow the general verte-brate pattern that GTHs stimulate spermatogenesis at leastpartly via a stimulation of androgen-production in the tes-tes (for review, see Schulz and Miura, 2002). Maturing malestriped bass showed an increase in FSH-� expression dur-ing early spermatogenesis, whereas LH-� peaked duringlate spermatogenesis (Hassin et al., 2000). In maturing maleayu, expression of FSH-� peaks earlier than expression ofLH-� (Yoshida et al., 2001). When FSH-� peaks in August,gonads have started to grow but are still small, and the ayumales have largely spermatocytes in the testes. In Septem-ber, when LH-� peaks, gonads have increased considerablyin size and testes contain largely spermatids. In male seabass (Dicentrarchus labrax) both FSH-� and LH-� subunitsincreased in parallel during spermatogenesis (Mateos et al.,2003). In male goldWsh (Sohn et al., 1999), LH-� expressiondisplayed a marked peak in the spawning season, whereasFSH-� expression changed much less seasonally, thoughthe highest levels were found in the late spawning period.FSH-� expression peaked at spawning in red seabreammales (Gen et al., 2000), LH-� was equally high prespaw-ning and spawning, but dropped in the post spawningperiod. In male rainbow trout (Gomez et al., 1999), bothLH-� and FSH-� expression, as well as LH and FSH pitui-tary content and LH plasma levels, reached their highestvalues at Wnal maturation (spermiation). FSH plasma lev-els, on the other hand, where highest at late spermatogene-sis, before spermiation. Furthermore, in early maturingparr male Atlantic salmon FSH-� subunit expressionincreased during spermatogenesis and decreased when sper-miation started, whereas the LH-� expression increased atthe beginning of spermiation and declined when the gonadsregressed (Schmitz, unpublished). Thus, seasonal patternsin GTHs are highly variable among Wshes. The diVerentpatterns in GTH expressions are also diYcult to correlatewith the diVerent seasonal patterns in spermatogenesis andsteroid production. The stickleback pattern in GTHs is very

diVerent from that in goldWsh, which appears to have arather similar gonadal cycle as the stickleback. On the otherhand, the stickleback GTH pattern resembles that found inthe Atlantic salmon, with the highest LH-� expressionin the spawning period and the highest FSH-� expression inthe months preceding spawning. However, in the salmonboth spermatogenesis and androgen levels are high in theprespawning period, whereas in the stickleback spermato-genesis is quiescent in the prespawning period when alsoandrogen levels are rather low (Mayer et al., 1990b), andmost active in late summer when 11KT levels as well as theexpression of both GTHs are at their lowest.

Among Wshes, the control of spermatogenesis is bestknown in the Japanese eel (for review, see Miura andMiura, 2001). In this Wsh, gonads remain immature in cap-tivity but injection of human chorionic gonotropin (hCG)stimulates the production of 11KT and induces all stages ofspermatogenesis (Miura et al., 1991a). Both hCG (Miuraet al., 1991b) and 11KT (Miura et al., 1991c) can alsoinduce all stages of spermatogenesis in vitro. Actions of11KT on spermatogenesis appear to be mediated via theinduction of activin in the Sertoli-cells (Miura and Miura,2001). Human activin can induce proliferation of eel sper-matogonia in vitro (Miura et al., 1995).

GTHs stimulate the production of androgens in the Japa-nese eel and in several other Wshes. The present and previousstudies support that this is also the case in the stickleback.11KT has been found to stimulate spermatogenesis in theJapanese eel and a number of other Wshes. The present andother studies on the stickleback, however, indicate that this isnot the case in this species. In the Japanese eel, at least someof the spermatogenetic eVects of 11KT are mediated via astimulation of activin production. Whether activin isinvolved in stickleback spermatogenesis is not known.

In conclusion, there is a great diversity in seasonal GTHpatterns in Wshes and also a great variety in seasonal pat-terns in gonadal steroids and gametogenesis. The diVerentseasonal GTH patterns can, however, not be consistentlyrelated to the divergent patterns in gonadal cycles. Thus, itappears likely that the physiological roles of LH and FSHdiVer considerably between diVerent Wshes.

References

Andersson, E., Mayer, I., Borg, B., 1988. Inhibitory eVect of 11-ketoandr-ostenedione and androstenedione on spermatogenesis in the three-spined stickleback, Gasterosteus aculeatus L. J. Fish Biol. 33, 835–840.

Baggerman, B., 1957. An experimental study on the timing of breeding andmigration in the three-spined stickleback (Gasterosteus aculeatus L.).Arch. Néerl Zool. 12, 105–317.

Berglund, I., Antonopoulou, E., Mayer, I., Borg, B., 1995. Stimulatory andinhibitory eVects of testosterone on testes in Atlantic salmon, Salmosalar, male parr. J. Fish Biol. 47, 586–598.

Billard, R., Breton, B., Richard, M., 1981. On the inhibitory eVect of somesteroids on spermatogenesis in adult rainbow trout (Salmo gairdneri).Can. J. Zool. 59, 1479–1487.

Borg, B., 1981. EVects of methyltestosterone on spermatogenesis and sec-ondary sexual characters in the three-spined stickleback, Gasterosteusaculeatus L. Gen. Comp. Endocrinol. 44, 177–180.

A. Hellqvist et al. / General and Comparative Endocrinology 145 (2006) 263–269 269

Borg, B., 1982. Seasonal eVects of photoperiod and temperature on sper-matogenesis and secondary sexual characters in the three-spined stick-leback, Gasterosteus aculeatusL. Can. J. Zool. 60, 3377–3386.

Borg, B., van Veen, T., 1982. Seasonal eVects of photoperiod and tempera-ture on the ovary of the three-spined stickleback, Gasterosteus aculea-tus L. Can. J. Zool. 60, 3387–3393.

Borg, B., Paulson, G., Peute, J., 1986. Stimulatory eVects of methyltestos-terone on pituitary gonadotropic cells and testes Leydig cells of thethree-spined stickleback, Gasterosteus aculeatus L., in winter. Gen.Comp. Endocrinol. 62, 54–61.

Borg, B., Peute, J., Reschke, M., van den Hurk, R., 1987. EVects of photo-period and temperature on testes, renal epithelium and pituitarygonadotropic cells of the threespine stickleback, Gasterosteus aculeatusL. Can. J. Zool. 65, 14–19.

Borg, B., Peute, J., Paulson, G., 1988. Seasonal changes in the gonadotropiccells of the male threespine stickleback, Gasterosteus aculeatus L. Can.J. Zool. 66, 1961–1967.

Borg, B., Schoonen, W.G.E.J., Lambert, J.G.D., 1989. Steroid metabolismin the testes of the breeding and non-breeding three-spined stickleback,Gasterosteus aculeatus L. Gen. Comp. Endocrinol. 73, 40–45.

Gen, K., Okuzawa, K., Senthilkumaran, B., Tanaka, H., Moriyama, S.,Kagawa, H., 2000. Unique expression of gonadotropin-I and -II sub-unit genes in male and female red seabream (Pagrus major) during sex-ual maturation. Biol. Reprod. 63, 301–319.

Gomez, J.M., Weil, C., Ollitrault, M., Le Bail, PY., Breton, B., Le Gac, F.,1999. Growth hormone (GH) and gonadotropin subunit gene expres-sion and pituitary and plasma changes during spermatogenesis andoogenesis in rainbow trout (Oncorhynchus mykiss). Gen. Comp. Endo-crinol. 100, 327–333.

Hassin, S., Claire, M., Holland, M.C.H., Zohar, Y., 1999. Ontogeny of folli-cle-stimulating hormone and luteinizing hormone gene expression dur-ing pubertal development in the female striped bass, Morone saxatilis.Biol. Reprod. 61, 1608–1615.

Hassin, S., Claire, M., Holland, M.C.H., Zohar, Y., 2000. Early maturity in themale striped bass, Morone saxatilis. Follicle stimulating hormone andluteinizing hormone gene expression and their regulation by gonadotropin-releasing hormone analogue and testosterone. Biol. Reprod. 63, 1691–1697.

Hellqvist, A., Schmitz, M., Lindberg, C., Olsson, P.-E., Borg, B., 2003. LH-�and FSH-� mRNA expression in nesting and post-breeding three-spined stickleback, Gasterosteus aculeatus, and eVects of castration onexpression of LH-�, FSH-� and spiggin mRNA. Fish Physiol. Bio-chem. 25, 311–317.

Hellqvist, A., Bornestaf, C., Borg, B., Schmitz, M., 2004. Cloning andsequencing of the FSH-� and LH �-subunit in the three-spined stickle-back, Gasterosteus aculeatus, and eVects of photoperiod and tempera-ture on LH-� and FSH-� mRNA expression. Gen. Comp. Endocrinol.135, 167–174.

Jackson, K., Goldberg, D., OWr, M., Abraham, M., Degani, G., 1999. Bluegourami (Trichogaster trichopterus) gonadotropic � subunits (I and II)cDNA sequence and expression during oogenesis. J. Mol. Endocrinol.23, 177–187.

Jakobsson, S., Borg, B., Haux, C., Hyllner, S.J., 1999. An 11-ketotestoster-one induced kidney-secreted protein: the nest building glue from malethree-spined stickleback, Gasterosteus aculeatus. Fish Physiol. Bio-chem. 20, 79–85.

Jones, I., Lindberg, C., Jakobsson, S., Hellqvist, A., Hellman, U., Borg, B.,Olsson, P.O., 2001. Molecular cloning and characterization of spiggin.An androgen-regulated extraorganismal adhesive with structural simi-larities to von Willebrand factor-related proteins. J. Biol. Chem. 21,17857–17863.

Kajimura, S., Yoshiura, Y., Suzuki, M., Aida, K., 2000. cDNA cloning oftwo gonadtropin � subunits (GTH-I� and-II�) and their expressionproWles during gametogenesis in the Japanese Xounder (Paralichthysolivaceus). Gen. Comp. Endocrinol. 122, 117–129.

Kawauchi, H., Suzuki, K., Itoh, H., Swanson, P., Naito, N., Nagahama, Y.,Nozaki, M., Nakai, Y., Itoh, S., 1989. The duality of teleost gonadotro-pins. Fish Physiol. Biochem. 7, 29–38.

Mateos, J., Mananos, E., Martinez-Rodriguez, G., Carrillo, M., Quérat, B.,Zanuy, S., 2003. Molecular characterization of sea bass gonadotropinsubunits (�, FSH-� and LH-�) and their expression during the repro-ductive cycle. Gen. Comp. Endocrinol. 133, 216–232.

Mayer, I., Borg, B., Schulz, R., 1990a. Conversion of 11-ketoandrostenedi-one to 11-ketotestosterone by blood cells of six Wsh species. Gen.Comp. Endocrinol. 77, 70–74.

Mayer, I., Borg, B., Schulz, R., 1990b. Seasonal changes in and eVects ofcastration/androgen replacement on the plasma levels of Wve andro-gens in the male three-spined stickleback, Gasterosteus aculeatus L.Gen. Comp. Endocrinol. 79, 23–30.

Miura, T., Yamauchi, K., Takahashi, H., Nagahama, Y., 1991a. Inductionof spermatogenesis in male Japanese eel, Anguilla japonica, by a singleinjection of human chorionic gonadotropin. Zool. Sci. 8, 63–73.

Miura, T., Yamauchi, K., Takahashi, H., Nagahama, Y., 1991b. Humanchorionic gonadotropin induces all stages of spermatogenesis in vitroin the male Japanese eel, (Anguilla japonica). Dev. Biol. 146, 258–262.

Miura, T., Yamauchi, K., Takahashi, H., Nagahama, Y., 1991c. Hormonalinduction of all stages of spermatogenesis in vitro in male Japanese eel(Anguilla japonica). Proc. Natl. Acad. Sci. USA 88, 5774–5778.

Miura, T., Miura, C., Yamauchi, K., Nagahama, Y., 1995. Human recom-binant activin induces proliferation of spermatogonia in vitro in theJapanese eel, Anguilla japonica. Fish. Sci. 61, 434–437.

Miura, T., Miura, C., 2001. Japanese eel: A model for analysis of spermato-genesis. Zool. Sci. 18, 1055–1063.

Páll, M., Mayer, I., Borg, B., 2002. Androgen and behavior in male three-spined stickleback, Gasterosteus aculeatus I. Developement of 11-ketotestosterone levels during the nesting cycle. Horm. Behav. 41, 377–383.

Prat, F., Sumpter, J.P., Tyler, C.R., 1996. Validation of radioimmunoassaysfor two salmon gonadotropins (GHT I and GTH II) and their plasmaconcentrations throughout the reproductive cycle in male and femalerainbow trout (Oncorhynchus mykiss). Biol. Reprod. 54, 1375–1382.

Schmitz, M., Aroua, S., Vidal, B., Le Belle, N., Elie, P., Dufour, S., 2005.DiVerential regulation of luteinizing hormone and follicle stimulatinghormone expression during ovarian development and under steroidfeedback in the European eel. Neuroendocrinology 81, 107–119.

Schulz, R.W., Miura, T., 2002. Spermatogenesis and its endocrine regula-tion. Fish Physiol. Biochem. 26, 43–56.

Sohn, Y.C., Yoshiura, Y., Kobayashi, M., Aida, K., 1999. Seasonal changesin mRNA levels of gonadotropin and thyrotropin subunits in the gold-Wsh, Carassius auratus. Gen. Comp. Endocrinol. 113, 436–444.

Suetake, H., Okubo, K., Sato, N., Yoshiura, Y., Suzuki, Y., Aida, K., 2002.DiVerential expression of two gonadotropin (GTH) �-subunit genesduring ovarian maturation induced by repeated injection of salmonGTH in the Japanese eel, Anguilla japonica. Fish. Sci. 68, 290–298.

Swanson, P., 1991. Salmon gonadotropins: reconciling old and new ideas.In: A.P Scott, J.P. Sumpter, D.E. Kime, M.S. Rolfe (Eds.), Proceedingsof the 4th International Symposium on Reproductive Physiology ofFish, University of East Anglia, Norwich, U.K. Fish Symp 91,SheYeld. pp. 2–7.

van Oordt, G.J., 1924. Die Veränderungen des Hodens während des Auf-tretens der sekundären Geschlechtsmerkmale bei Fischen. I. Gasteros-teus pungitius L. Arch. Mikr Anat. 102, 379–405.

Yoshida, D,, Nagae, M., Ito, F., Soyano, K., 2001. Molecular cloning ofcDNAs encoding pituitary glycoprotein hormone �, FSH � and LH �subunits in ayu, Plecoglossus altivelis. Zool. Sci. 18, 929–936.

Yoshiura, Y., Kobayashi, M., Kato, Y., Aida, K., 1997. Molecular cloningof the cDNAs encoding two gonadotropin � subunits (GTH-I� and -II�) from the gold Wsh, Carassius auratus. Gen. Comp. Endocrinol. 105,379–389.