Group-synchronous ovarian development , ovulation and spermiation in the European sea bass ( Dicentrarchus labrax L . ) could be regulated by shifts in gonadal steroidogenesis *

European sea bass (Dicentrarchus labrax L.) is a species with group-synchronous ovarian development. A mechanism is required which enables maturation to occur in the first clutches of oocytes without inducing maturation in subsequent clutches. The present study examined the individual plasma variations of testosterone (T), 17,20β-dihydroxy-4pregnen-3-one (17,20βP) and 17,20β,21-trihydroxy-4-pregnen-3-one (20βS) in both sexes, estradiol (E2) and vitellogenin (VTG) in females (n=15) and 11-ketotestosterone (11KT) in males (n=21), in an effort to elucidate the hormonal control of the reproductive cycle in this species. A sample of oocytes was obtained at every sampling from each female and the stage of development of the most advanced clutch of oocytes was determined and related to the individual hormone plasma levels. Total expressible milt was collected from males at each sampling during the spawning period and variations in the sperm production were related to hormone plasma levels. Successive elevations of plasma T and E2 levels were observed prior to peaks of progestagens, which resulted from the shift in gonadal steroidogenesis and coincided with the maturation-ovulation of the different clutches of oocytes or with increases in the sperm production. Following each progestagens wave, a new shift in gonadal steroidogenesis, resulting in a new elevation in plasma T and E2, was observed. This hormonal pattern was repeated several times depending on the number of ovulations per female. Results from the present study suggest a mechanism, based on shifts in gonadal steroidogenesis, which may be responsible for regulation of group-synchronous ovarian development, ovulation and spermiation in this species. In addition, evidence is presented which supports a role for both 17,20βP and 20βS as the maturation-inducing steroids (MIS) in male and female European sea bass.


INTRODUCTION
European sea bass (Dicentrarchus labrax L.) has a group-synchronous ovarian development (Carrillo et al., 1989;Mayer et al., 1990;Alvariño et al., 1992). Thus, during gonadal recrudescence, each ovary contains two or more clutches of oocytes in different stages of development that are successively ovulated. Recent studies have shown that one female of this species can have up to four ovulations during the natural reproductive period (Asturiano et al., 2000).
Several studies have described the seasonal variations of plasma levels of testosterone (T), 17βestradiol (E 2 ) and vitellogenin (VTG) in the female European sea bass (Prat et al., 1990(Prat et al., , 1999Mañanós et al., 1994Mañanós et al., , 1997aZanuy et al., 1995;Navas et al., 1998) as well as the seasonal plasma variations of T and 11-ketotestosterone (11KT) levels in the males (Prat et al., 1990(Prat et al., , 1999Cerdá et al., 1997). However, these previous studies were based on monthly samplings and results were expressed as the mean hormonal plasma level from several fish. Due to the group synchronous nature of this species and the unpredictability of the multiple ovulations that occur during the reproductive season, large variations in cycling are evident both in males and females. Therefore, using monthly sampling method, rapid changes in hormonal levels may be missed. New studies, based on more intense strategies of sampling and an individualized study of the gonadal morphology are required to determine the precise relationship between plasma levels of T, 11KT, E 2 and VTG and the progression of the reproductive cycle. This information would provide a better understanding of the hormonal regulation of reproductive processes in this species.
In salmonids, FSH (previously named GTH1, Quérat, 1994) is released during the period of gonadal growth and stimulates the follicular production of E 2 and T (Prat et al., 1996;Nagahama, 1997).
When follicles reach the postvitellogenic stage, LH (previously named GTH2, Quérat, 1994) regulates final maturation and ovulation/spermiation and stimulates a shift in steroidogenesis from the synthesis of T and E 2 to that of maturation-inducing steroids (MIS) thus promoting the maturation and ovulation of oocytes (reviewed by Nagahama, 1994Nagahama, , 1997Nagahama et al., 1995;Prat et al., 1996). This shift in gonadal steroidogenesis has been observed in several species of the genus Morone, closely related to the European sea bass (King et al., 1994(King et al., , 1995Mylonas et al., 1997a).
In the female European sea bass, several studies have described the potential role of progestagens during final oocyte maturation (Prat et al., 1990;Scott et al., 1990). Both 17,20βP) and 17,20β, were shown to induce maturation in vitro (Sorbera et al., 1999) and Asturiano et al. (2000) showed plasma peaks of 17,20βP and 20βS in vivo coinciding with the beginning of maturation and with final maturation, respectively, suggesting that both progestagens may be the MIS in this species. However, until now no studies have been done concerning the role of progestagens in the regulation of the male reproductive process in this species.
The aim of the present study was to elucidate the hormonal mechanism which enables maturation to occur in the first clutches of oocytes without inducing maturation in subsequent clutches. Moreover, this study attempted to confirm the involvement of 17,20βP and 20βS in the reproductive cycle of both male and female European sea bass.

Fish
Three-year-old female (882.4 ± 57 g B.W.; n=15) and five-years-old male (852.16 ± 23 g; n=21) Euro-pean sea bass were maintained under natural photoperiod and temperature conditions in separate fibre-glass tanks supplied with aerated flow-through sea water (37.8 g l -1 salinity). Fish were hand fed three days a week with a natural diet consisting of Boops boops at the rate of 2.2% tank biomass per day from June to October and 1.2% the rest of the year. The amount of food given to broodstock was adjusted monthly after sampling.

Sampling protocol
All samplings were performed between 9:30 and 10:30 am. At each sampling, males and females were anesthetized with ethylenglycol-monophenylether (1 ml l -1 ) and blood was taken from the caudal vein and dispensed into heparinized tubes kept on ice. Plasma was obtained by centrifugation (1600 g, 20 min at 4ºC), aliquoted and stored at -20ºC until analysis.
Males were bleed and checked for the presence of sperm by gently stripping, between 9:30 and 10:30 am, at 7-day intervals from October to April. After cleaning the genital area with fresh water and throughly drying to avoid contamination of samples with feces, urine and sea water, total expressible milt was collected (from January to April) by applying gentle abdominal pressure to anesthetized males and by placing a vacuum glass tube around the genital area to collect sperm in a graduate cilinder. Total expressible milt was noted in ml.
Females were first sampled at 1-month intervals from the beginning of September to mid-January and later at 7-day intervals until April. A sample of oocytes was obtained from each ovary of every female by inserting a polyethylene tubing into the oviduct to the mid-portion of the ovary and sucking orally as the cannula was withdrawn. Oocytes were placed in Ringer's solution modified for the sea bass (SBR; Sorbera et al., 1999) and maintained on ice for measurement of oocyte diameter using an ocular micrometer under light microscopy immediately after obtaining the sample. In each sample, 60-100 of the largest oocytes were measured. A sample of each group of measured oocytes was fixed separately in 4% formaldehyde, 1% glutaraldehyde fixative (McDowell and Trump, 1976). Follicles were dehydrated in a 75-90% ethanol series and embedded in plastic Technovit (Kulzer). The blocks were cut in 2µm sections and were stained with Cleveland-Wolfe (Herlant, 1960). These preparations were used to determine the developmental stage of largest oocytes in each female at every sampling. The stag-ing criteria used for oocyte development was as described by Asturiano et al. (2000).

Plasma vitellogenin and steroid hormone analyses
All plasma VTG and steroid hormone samples were assayed in duplicate. Plasma levels of T were measured by an EIA as described by Rodríguez et al. (2001). T was triple extracted with methanol using 100 µl of plasma diluted 1:32 in potassium phosphate buffer (0.1 M, pH 7.4) containing 0.01% sodium azide, 0.4 M NaCl, 0.001 M EDTA and 0.1% BSA. The interassay coefficient of variation was 7.54% (n=10).
Plasma levels of 11KT were measured by EIA as described by Cuisset et al. (1994) and the inter-assay coefficient of variation was 10.04% (n=3).
Plasma levels of VTG were quantified using a homologous ELISA following the procedure described by Mañanós et al. (1994). Prior to analysis, plasma samples were diluted 1:10,000 or 1:100,000-fold dilution depending on the reproductive stage of the females in order to ensure that all measurements were within the confidence range of the standard curve. The inter-assay coefficient of variation was 10.39% (n=5). E 2 was extracted from 100 µl of plasma using cyclohexane:ethylacetate (1:1, v/v) and E 2 concentrations were determined by RIA according to Prat et al. (1990). The intra-and inter-assay coefficients of variation were 8.34% (n=4) and 4.27% (n=12), respectively.
Plasma concentrations of 17,20βP and 20βS were measured by specific ELISAs based on the method described by Nash et al. (2000). Cross-reactivities of the antisera to 17,20βP and 20βS are as in Canario et al. (1989). We have not carried out a full test of cross-reactivity using ELISA for these steroids, but previous studies (Cuisset et al., 1994) and our own more limited trials have shown no significant difference in cross-reactivity between ELISA and RIA.

Females
Individual samples collected from weekly samplings from September to April revealed that in all females the first ovulation took place after the maximum levels of plasma VTG (Figs. 1 and 2).
Both T and E 2 plasma levels showed important variations during the spawning season, with peaks observed prior to ovulation of the successive clutches of oocytes in each female ( Figs. 1 and 2). In addition, all the females showed a wave-like profile of plasma 17,20βP and 20βS. Peaks of these progesta-gens occurred slightly before or coincided with maturation-ovulation of the different clutches of oocytes detected in each female. Figures 1 and 2   analysis of ovarian samples collected at each sampling and measurement of the diameters of the largest oocyte led to the conclusion that these two females had three and four ovulations, respectively.
For instance, Figure 1 shows the first increases of T and E 2 during the 2nd week of January. Next, a shift in gonadal steroidogenesis was observed resulting in diminishing plasma T and E 2 and increases in 17,20βP (from the 2nd week) and 20βS (between the 3rd-4th weeks of January); the first ovulation was detected in the 1st week of February. ovulation of the most advanced clutch of oocytes, a new shift in gonadal steroidogenesis was observed. This shift was associated with a new elevation of plasma T and E 2 and a reduction in MIS. This pattern was repeated several times depending on the number of ovulations per female. Figure 1 illustrates how new increases of 17,20βP (3rd week of February) and 20βS (4th week of February) resulted in a second ovulation during the 1st week of March and how increases of 17,20βP (1st-2nd weeks of March) and 20βS (3rd week of March) resulted in a third ovulation during the 3rd week of March. This mechanism was observed in all ovulating females (80%); unovulated oocytes from the remaining females appeared to regress and undergo atresia in spite of reaching a postvitellogenic oocyte diameter.

Males
A similar mechanism was also observed in the males. Figures 3 and 4 show the total sperm volume collected from two individual males together with 278 J.F. ASTURIANO et al. Numbers indicate the peaks in sperm production during this period and the associated peaks of androgens and progestagens, respectively.
Vertical long-dotted lines correlate different peaks with every increase in sperm production. plasma variations of T, 11KT, 17,20βP and 20βS. In the 21 considered males, total sperm volume was not constant, instead exhibiting successive peaks in production. T and 11KT peaks preceded the MIS elevations responsible for increases in sperm production. For instance, results from a male shown in Fig. 4 displayed the first peaks of T and 11KT during the 2nd week of January that were followed by 17,20βP and 20βS peaks during the 3rd week (coinciding with a reduction in androgens) and by the first increase in sperm production during the 3rd-4th weeks of the same month. A second peak in androgens was observed in the 1st week of February and was followed by progestagen peaks during the 2nd-3rd weeks of this month which coincided with a second increase in sperm production. New increases in androgens and progestagens were detected as well as a third peak of sperm production during the 1st week of March of March. Although experimental males were maintained separated from females throughout the experiment, the length of the spermiation period was not affected. Thus other males maintained in the same conditions and fed the same diets, but housed with females, exhibited a similar duration of the spermiation period (data not shown).

DISCUSSION
In a species with group-synchronous ovarian development such as the European sea bass, a mechanism is required which enables maturation to occur in the first clutches of oocytes without inducing maturation in subsequent clutches. Results from the present study suggest, for the first time in this species, the existence of this mechanism which is based on shifts in gonadal steroidogenesis. Moreover, results from this study corroborate that both 17,20βP and 20βS have a role as MIS in male and female European sea bass.
All the females that presented ovulations displayed successive short-term elevations of plasma T and E 2 prior to peaks of 17,20βP and 20βS. Each MIS wave resulting from the shift in gonadal steroidogenesis coincided with a reduction in T and E 2 plasma levels and occurred slightly prior to ovulation of the most advanced clutch of oocytes. Following each MIS wave which resulted in the maturation and ovulation of one clutch of oocytes, a new shift in gonadal steroidogenesis was observed. This shift signified a new elevation in T and E 2 plasma levels and a reduction in MIS plasma levels preventing maturation of the rest of clutches.
A similar pattern of hormonal regulation was observed in males. Each wave of plasma progestagens was observed to coincide with increases in the sperm production and was preceded by increases in the synthesis of androgens.
Studies in vivo and in vitro in salmonids have demonstrated that when follicles reach the postvitellogenic stage, an increase in LH level induces the synthesis of androgens and a decrease in their aromatization to E 2 while the enzymes involved in the synthesis of MIS are activated (Zohar et al., 1986;Nagahama, 1994Nagahama, , 1997Nagahama et al., 1995). A similar mechanism was observed by Vizziano et al. (1996) in male rainbow trout (Oncorhynchus mykiss) in which the synthesis of 17,20βP was inhibited by E 2 until the spermiation period, when the release of E 2 became discontinuous. The same mechanism was previously proposed to explain the changes in the release of 17,20βP during the period of final maturation of oocytes in the ovary of the rainbow trout (Jalabert and Fostier, 1984). Murayama et al. (1994) found that the changes in plasma levels of E 2 correlated with the successive spawnings in the Japanese sardine (Sardinops melanostictus). This shift in gonadal steroidogenesis from the production of T and E 2 to that of MIS has also been observed in the striped bass (Morone saxatilis) and other species of the genus Morone, closely related to the European sea bass. King et al. (1994) observed in vitro that human chorionic gonadotropin (hCG) induced the production of MIS and the reduction of E 2 and T synthesis in ovarian fragments of striped bass (M. saxatilis). Similar results were reported both in vitro and in vivo in M. americana and M. chrysops (King et al., 1995). Recently, Rahman et al. (2001) reported a similar steroidogenic shift in hCG-treated ovarian fragments of Japanese yellowtail (Seriola quinqueradiata) during final oocyte maturation.
In the white bass (M. chrysops) in vivo treatment with gonadotropin releasing hormone agonist (GnRHa) induced final oocyte maturation associated with a significant elevation in plasma LH levels and important shifts in gonadal steroidogenesis (Mylonas et al., 1997b). In female European sea bass, Navas et al. (1998) observed a peak of LH at the middle of the ovulation. Moreover, males exhibited the highest LH plasma levels at the middle of the spermiation period (Navas et al., 2001;Rodríguez et al., 2001). Although these studies did not examine plasma progestagen variations, female treatment with GnRHa resulted in an important elevation of plasma LH levels that was followed by a peak of 17,20βP and the appearance of ovulated oocytes (Fornies et al., 2000). Both 17,20βP and 20βS have been detected in the plasma of female European sea bass (Prat et al., 1990;Scott et al., 1990;Asturiano et al., 2000) and both progestagens have been shown to induce maturation in vitro (Sorbera et al., 1999). Results from the present study support a role for both 17,20βP and 20βS as MIS both in male and female European sea bass.
The correlation between the beginning of spermiation and increases in plasma levels of progestagens has been observed in several species. Ueda et al. (1985) reported that injections of 17,20βP causes an increase in sperm volume in the goldfish (Carassius auratus) and suggested that this steroid could play a role as a testicular mediator in the pro-duction of sperm induced by LH. A similar suggestion has been made for salmonids (reviewed by Nagahama, 1994). In the striped bass (M. saxatilis) GnRHa treatment resulted in a significant increase of LH, 11KT, and 17,20βP plasma levels as well as a significant elevation in total expressible milt (Mylonas et al., 1997c). In accordance with our results, Baynes and Scott (1985) and Barry et al. (1990) proposed that LH induces the production of progestagens which causes an inhibition in androgen synthesis thus regulating final maturation of sperm in fish.
In the plaice (Pleuronectes platessa), which also exhibits group-synchronous ovarian development, Scott et al. (1998) found that 17,20βP plasma levels during oocyte maturation were not higher than basal levels, although its metabolites were considerably higher in females that ovulated. The present results support the idea suggested by Scott and Canario (1987) that 17,20βP has a paracrine function, influencing the follicles where it is synthesized and that enzymes could inactivate it before it enters the circulatory system. If it entered the circulatory system, 17,20βP would have an endocrine role and induce the maturation of all follicles. This mechanism is necessary in species like the European sea bass, that have group-synchronous ovarian development resulting in several spawnings during a reproductive season, and could explain the slight variations in the plasma levels of 17,20βP and 20βS detected in this study in comparison with the levels reached in other species, especially salmonids, that display synchronous oocyte development.
Furthermore, studies have also revealed rapid changes in MIS levels in some teleosts and a weekly sampling strategy, as used in this study, or monthly sampling would give a limited view of the real amplitude of the progestagens' variations. For instance, in the Japanese sardine (S. melanostictus), final oocyte maturation occurs daily between 18:00 and 23:00 hours (Matsuyama et al., 1994;Murayama et al., 1994), and only during this period plasma concentrations of 17,20βP are detected. Moreover, it is known that the gonad transforms free steroids to glucuronides and sulfates (Kime, 1993) which are undetectable by the assays used in this study. In addition to these changes there may also be short term diurnal changes or a pulsatile release of steroids throughout the day as described by Nash et al. (1999Nash et al. ( , 2000 for rainbow trout. Thus, all these factors could be affecting our measurements of progestagen plasma level variations. This study has identified a possible mechanism based on shifts in gonadal steroidogenesis which may be responsible for regulation of group-synchronous ovarian development, ovulation and spermiation in the European sea bass. However, in vitro studies are required to confirm this mechanism and to explain the involvement of GTHs, gonadal steroids, enzymes, and follicular factors in the regulatory system of reproduction in this species.