Zebrafish gonadotropins and their receptors: I. Cloning and characterization of zebrafish follicle-stimulating hormone and luteinizing hormone receptors--evidence for their distinct functions in follicle development

In the present study, we cloned and characterized zebrafish FSH receptor (Fshr) and LH receptor (Lhr). Both fshr and lhr were abundantly expressed in the zebrafish gonads; however, they could also be detected in the kidney and liver, respectively. When overexpressed in mammalian cell lines together with a cAMP-responsive reporter gene, zebrafish Fshr responded to goldfish pituitary extract but not hCG, whereas Lhr could be activated by both. It was further demonstrated that Fshr was specific to bFSH, while Lhr could be stimulated by both bovine FSH and LH. Low level of fshr expression could be detected in the immature ovary, but the level steadily increased during vitellogenesis of the first cohort of developing follicles. In contrast, the expression of lhr could barely be detected in the immature ovary, but it became detectable at the beginning of vitellogenesis and steadily increased afterward with the peak level reached at the full-grown stage. At the follicle level, the expression of fshr was very weak in the follicles of primary growth stage but significantly increased with the follicles entering vitellogenesis. However, after reaching the maximal level in the midvitellogenic follicles, the level of fshr expression dropped slightly but significantly at the full-grown stage. In comparison, the expression of lhr obviously lagged behind that of fshr. Its expression became detectable only when the follicles started to accumulate yolk granules, but the level rose steadily afterward and reached the peak at the full-grown stage before oocyte maturation. These results suggest differential roles for Fshr and Lhr in zebrafish ovarian follicle development.     

Potential role of bone morphogenetic protein-15 in zebrafish follicle development and oocyte maturation

Bone morphogenetic protein (BMP)-15 is a member of the transforming growth factor beta (TGF-beta) superfamily and is closely related to growth and differentiation factor (GDF)-9, both structurally and functionally. In mammals, BMP-15 is predominantly produced by oocytes and exerts important regulatory functions within the ovary, such as promoting early folliculogenesis, preventing premature luteinization and enhancing cumulus cell expansion. The role of BMP-15 in mammalian ovary differs between monoovulatory and polyovulatory species. Recent studies in zebrafish have provided initial evidence that BMP-15 is also an important regulator of ovarian functions. BMP-15 is produced by the zebrafish ovary throughout follicle development and maturation. In vitro studies using zebrafish follicles have revealed that incubation with recombinant human BMP-15 or over-expression of BMP-15 in oocytes results in an inhibition of gonadotropin- and maturation inducing hormone (MIH)-induced oocyte maturation. Conversely, immnunoneutralization with BMP-15 antiserum or silencing of BMP-15 expression using morpholino antisense oligonueclotides enhances oocyte maturation. A key step in BMP-15 action is the sensitivity of follicles to MIH. In vivo injection of BMP-15 antiserum causes a significant decrease in maturation-incompetent (insensitive to MIH) small early growth phase follicles and a concomitant increase in mature follicles. These findings support a role in BMP-15 in preventing precocious oocyte maturation in zebrafish. We propose that the suppression of premature oocyte maturation by BMP-15 may be important to maintain oocyte quality and subsequent ovulation and fertilization.     

Expression Patterns of CREBs in Oocyte Growth and Maturation of Fish

In fish, oocyte meiotic maturation is regulated by 17α, 20β-dihydroxy-progesterone through cAMP. To study the role of cAMP response element binding protein (CREB) in meiotic maturation, we cloned and characterized the expression pattern of CREBs from two fish models, the Nile tilapia and catfish. In the Nile tilapia three different CREBs were identified where in CREB1 was found in many tissues including gonads with abundant expression in testis. CREB2, few amino acids shorter than CREB1, was expressed in several tissues with abundant expression in ovary. In addition, a 3’UTR variant form, CREB3 was exclusively found in ovary. During natural 14-day ovarian cycle of the Nile tilapia, CREB1 expression was stable throughout vitellogenesis with a sharp decrease on the day of spawning. In contrast, CREB2 remain unchanged throughout the ovarian cycle, however elevated in 11-day full-grown immature ovarian follicle and after hCG-induction. Interestingly, CREB3 expression was induced three folds on the day of spawning as well as during hCG-induced oocyte maturation. Based on the synergistic expression pattern, CREB1 is likely to control oocyte growth, whereas CREB 2 and 3 contribute to oocyte maturation in tilapia and the latter seems to be critical. In catfish, a single form of CREB showed a maximum expression during spawning phase and hCG-induced maturation both in vivo and in vitro augmented CREB expression. These results suggest that spatial and temporal expression of CREBs seems to be important for final oocyte maturation and may also regulate oocyte growth in fish.     

Pituitary adenylate cyclase-activating polypeptide (PACAP) and its receptors in the zebrafish ovary: evidence for potentially dual roles of PACAP in controlling final oocyte maturation

Pituitary adenylate cyclase-activating polypeptide (PACAP) is a neuropeptide originally purified from ovine hypothalamus for its potent activity to stimulate cAMP production. However, its presence and action have also been demonstrated in various peripheral tissues including the ovary. In the zebrafish, two forms of PACAP (PACAP(38)-1, adcyap1a; and PACAP(38)-2, adcyap1b) and three PACAP receptors (PAC(1)-R, adcyap1r1; VPAC(1)-R, vipr1; and VPAC(2)-R, vipr2) were all expressed in the ovary. Interestingly, although both follicle cells and oocytes express adcyap1b, the expression of adcyap1a was restricted to the oocytes only. Among the three receptors, adcyap1r1 and vipr2 were expressed in the oocytes, whereas the expression of vipr1 was exclusively located in the follicle cells. Temporal expression analysis of PACAP ligands and receptors during folliculogenesis suggested that PACAP might play differential roles in regulating follicle growth and maturation through different receptors. The two receptors that are expressed in the oocyte (adcyap1r1 and vipr2) showed a significant increase in expression at the transition from the primary growth (PG) stage to previtellogenic (PV) stage and their levels maintained high during follicle growth. However, when the follicle development approached full-grown (FG) stage, these two receptors both decreased significantly in expression. In contrast, vipr1, the receptor expressed in the follicle cells, showed little change in expression at the PG-PV transition and afterwards during follicle growth; however, its expression surged dramatically at the FG stage prior to oocyte maturation. Based on these results, we hypothesized that PACAP might play dual roles in regulating follicle growth and maturation through different receptors located in different compartments. PACAP may stimulate oocyte growth but block its maturation in early follicles by acting directly on the oocyte via PAC1-R and VPAC2-R, whose expression is dominant in growth phase; however, PACAP may promote oocyte maturation in the maturation phase via VPAC1-R on the follicle cells, whose expression surges in FG follicles prior to maturation and is consistently high in the follicles undergoing final maturation. This hypothesis was further supported by the observation that PACAP promoted maturation of follicle-enclosed oocytes but suppressed spontaneous maturation of denuded oocytes in vitro. This study provides strong evidence for a PACAP-mediated signaling network in the zebrafish ovarian follicle, which may play roles in orchestrating follicle growth and maturation via different types of receptors located in different compartments of the follicle.     

Developmental profiles of activin betaA, betaB, and follistatin expression in the zebrafish ovary: evidence for their differential roles during sexual maturation and ovulatory cycle

Our recent experiments showed that gonadotropin(s) stimulated activin betaA and follistatin expression through the cAMP-PKA pathway but suppressed betaB via a cAMP-dependent but PKA-independent pathway in cultured zebrafish follicle cells. Given that pituitary gonadotropins are the major hormones controlling the development and function of the ovary, the differential expression of activin betaA and betaB as well as follistatin in response to gonadotropin(s) raises an interesting question about the temporal expression patterns of these molecules in vivo during sexual maturation and ovulatory cycle. Three experiments were performed in the present study. In the first experiment using sexually immature zebrafish, we followed the expression of activin betaA, betaB, and follistatin at the whole ovary level during a 10-day period in which the ovary developed from the primary growth stage to the one with nearly full-grown follicles. Activin betaA expression was very low at the primary growth stage but significantly increased with the growth of the ovary, and its rise was accompanied by an increase in follistatin expression. In contrast, the expression of activin betaB could be easily detected in the ovary of all stages; however, it did not exhibit an obvious trend of variation during the development. The second experiment examined the stage-dependent expression of activin betaA, betaB, and follistatin at the follicle level in the adult mature zebrafish. The expression of activin betaA was again low in the follicles during the primary growth stage, but exhibited a phenomenal increase after the follicles entered vitellogenesis with the peak level reached at midvitellogenic stage; in contrast, activin betaB mRNA could be easily detected at all stages with a slight increase during follicle growth. The expression of follistatin, on the other hand, also increased significantly during vitellogenesis; however, its level dropped sharply after reaching the peak at the midvitellogenic stage. In the third experiment, we investigated the dynamic changes of the ovarian activin betaA, betaB, and follistatin expression during the daily ovulatory cycle. The expression of activin betaA and follistatin gradually increased from 1800 h onward and reached the peak level around 0400 h when the germinal vesicles had migrated to the periphery in the full-grown oocytes. In contrast, activin betaB expression steadily declined, although not statistically significant, during the same period, but increased sharply at 0700 h when mature oocytes started to appear in most of the ovaries collected. In conclusion, activin betaA and betaB exhibit distinct expression patterns during the development of the ovary and the daily ovarian cycle of the zebrafish. It seems that activin betaA is involved in promoting ovary and follicle growth, whereas activin betaB may have a tonic role throughout follicle development but becomes critical at the late stage of oocyte maturation and/or ovulation.     

In vitro inhibition of oocyte and follicular maturation and spawning in starfish (Asterias forbesi) by 2-4-dinitrophenol

The in vitro effects of 2-4-dinitrophenol (DNP) on spawning and follicular and oocyte maturation in starfish ovaries and its various cellular components were investigated. Spawning and oocyte and follicular maturation induced by starfish gonadotropin radial nerve factor (RNF) in isolated ovarian fragments were all inhibited by appropriate doses of DNP. DNP inhibits processes which occur shortly after addition of the gonadotropin; in ovarian fragments insensitivity to DNP inhibition occurred shortly after addition of RNF but prior to initiation of spawning. Spontaneous follicular and oocyte maturation which occurred following release of ovarian follicles into sea water was prevented by DNP. In non-spontaneously maturing follicles released from the ovary, DNP inhibited both follicle and oocyte maturation induced by the secondary stimulator of spawning and maturation, 1-methyladenine (1-MA). DNP also inhibited 1-MA induced meiotic maturation in isolated immature oocytes incubated in the absence of follicle cells. Inhibition of oocyte maturation was not associated with inhibition of 3H-1-MA incorporation by isolated oocytes. Immature oocytes incubated in the presence of DNP underwent maturation following washing and subsequent exposure to 1-MA. Immature oocytes initially exposed to both 1-MA and DNP, however, showed decreased maturation responsiveness following washing and re-exposure to 1-MA. The results suggest that the inhibitory effects of DNP on spawning and oocyte maturation are the result of direct effects on the oocytes and possibly other cells and tissues within the ovary.     

In vitro Inhibition of Oocyte and Follicular Maturation and Spawning in Starfish (Asterias forbesi) by 2-4-Dinitrophenol

The in vitro effects of 2-4-dinitrophenol (DNP) on spawning and follicular and oocyte maturation in starfish ovaries and its various cellular components were investigated. Spawning and oocyte and follicular maturation induced by starfish gonadotropin radial nerve factor (RNF) in isolated ovarian fragments were all inhibited by appropriate doses of DNP. DNP inhibits processes which occur shortly after addition of the gonadotropin; in ovarian fragments insensitivity to DNP inhibition occurred shortly after addition of RNF but prior to initiation of spawning. Spontaneous follicular and oocyte maturation which occurred following release of ovarian follicles into sea water was prevented by DNP. In non-spontaneously maturing follicles released from the ovary, DNP inhibited both follicle and oocyte maturation induced by the secondary stimulator of spawning and maturation, 1-methyladenine (1-MA). DNP also inhibited 1-MA induced meiotic maturation in isolated immature oocytes incubated in the absence of follicle cells. Inhibition of oocyte maturation was not associated with inhibition of ³H-1-MA incorporation by isolated oocytes. Immature oocytes incubated in the presence of DNP underwent maturation following washing and subsequent exposure to 1-MA. Immature oocytes initially exposed to both 1-MA and DNP, however, showed decreased maturation responsiveness following washing and re-exposure to 1-MA. The results suggest that the inhibitory effects of DNP on spawning and oocyte maturation are the result of direct effects on the oocytes and possibly other cells and tissues within the ovary.     

Temporal and spatial expression of the four Igf ligands and two Igf type 1 receptors in zebrafish during early embryonic development

The insulin-like growth factor (Igf) family is an evolutionarily conserved system essential for normal growth and development in vertebrates. Unlike mammals, four distinct Igf ligands (Igf1, Igf2a, Igf2b and Igf3) and two Igf type 1 receptors (Igf1ra and Igf1rb) are present in zebrafish. However, the localization of these multiple ligands and receptors especially the recently discovered igf3 during early development of zebrafish is poorly understood. In this study, detailed expression patterns of these components of the Igf system during embryogenesis of zebrafish were analyzed. It was found that igf1 is specifically expressed in the trigeminal ganglia region from 18 hpf to 72 hpf, while igf2a is restricted to the caudal regions of the notochord from 14 hpf to 18 hpf as well as in the midbrain, dorsal hind brain and otic vesicle at 24 hpf. On the other hand, igf2a is highly expressed in the midbrain and pharyngeal arch region at 48 hpf, followed by its appearance in the liver and brain at 72 hpf, while igf2b is restricted to the floor plate and hypochord from 12 hpf to 18 hpf, and strong expression is also detected in the midbrain and dorsal hind brain at 24 hpf. The teleost specific igf3 is highly expressed in the pharyngeal arch region before 24 hpf, but is then restricted to the sternohyoideus after 48 hpf. The receptor subtype igf1ra is ubiquitously expressed before 24 hpf but is confined to the brain at 72 hpf. However, igf1rb is widely expressed before 10 hpf, but is more confined to the brain region at 24 hpf and 72 hpf. This dynamic temporal-spatial expression during embryogenesis of zebrafish, together with the unique and overlapping expression patterns of the Igf ligands and receptors suggest the coordination of the divergent functions of the Igf system during early development in zebrafish.     

Spatial expression patterns of activin and its signaling system in the zebrafish ovarian follicle: evidence for paracrine action of activin on the oocytes

We have previously demonstrated that activin is likely an ovarian mediator of pituitary gonadotropin(s) and local epidermal growth factor in their stimulating oocyte maturation and maturational competence in the zebrafish. However, the downstream events controlled by activin remain unknown. One possible mechanism is that activin may directly work on the oocytes to promote the development of oocyte maturational competence. To substantiate this hypothesis, we performed the present study to demonstrate the expression of the activin system in different compartments of zebrafish follicles, namely, the follicle cells and oocytes. The proteins examined include activin subunits (betaA and betaB), activin-binding protein (follistatin), activin type II receptors (type IIA and IIB), the type I activin receptor-like kinases (ALK1-like, ALK2-like, and ALK4-like), and the intracellular activin signaling molecules (Smad2, Smad3, Smad4, and Smad7). The results showed that the entire activin signaling system is expressed by the full-grown immature zebrafish oocytes ( approximately 0.65 mm in diameter), including ALK4-like (ActRIB), ALK2-like (ActRIA), ActRIIA, ActRIIB, Smad2, Smad3, Smad4, and Smad7, therefore supporting our hypothesis that the oocytes are one of the direct targets of activin actions in the zebrafish ovary. In contrast, activin itself (betaA and betaB) and ALK1-like type I receptor are predominantly expressed in the follicle cells surrounding the oocytes. Interestingly, although follistatin is expressed in both the follicle cells and oocytes, its level of expression is significantly higher in the oocytes than the follicle cells, implying that follistatin may serve as a signal from the oocytes to modulate the activity of activin produced by the follicle cells. Taken together, the present study provides convincing evidence that although all members of the activin system are expressed in the whole follicle, they exhibit distinct spatial patterns of expression among different compartments of the follicle. It is likely that activin works directly on the oocytes in a paracrine manner to promote oocyte maturation and maturational competence. On the other hand, instead of being controlled passively by the follicle cells, the oocytes may actively participate in the regulation of follicle development by releasing various modulating molecules such as follistatin.     

Identification of calcitonin expression in the chicken ovary: influence of follicular maturation and ovarian steroids

Calcitonin (CALCA), a hormone primarily known for its role in calcium homeostasis, has recently been linked to reproduction, specifically as a marker for embryo implantation in the uterus. Although CALCA expression has been documented in several tissues, there has been no report of production of CALCA in the ovary of any vertebrate species. We hypothesized that the Calca gene is expressed in the chicken ovary, and its expression will be altered by follicular maturation or gonadal steroid administration. Using RT-PCR, we detected Calca mRNA and the calcitonin receptor (Calcr) mRNA in the granulosa and theca layers of preovulatory and prehierarchial follicles. Both CALCA and Calca mRNA were localized in granulosa and thecal cells by confocal microscopy. Using quantitative PCR analysis, F1 follicle granulosa layer was found to contain significantly greater Calca mRNA and Calcr mRNA levels compared with those of any other preovulatory or prehierarchial follicle. The granulosa layer contained relatively greater Calca and Calcr mRNA levels compared with the thecal layer in both prehierarchial and preovulatory follicles. Progesterone (P(4)) treatment of sexually immature chickens resulted in a significantly greater abundance of ovarian Calca mRNA, whereas estradiol (E(2)) or P(4) + E(2) treatment significantly reduced ovarian Calca mRNA quantity. Treatment of prehierarchial follicular granulosa cells in vitro with CALCA significantly decreased FSH-stimulated cellular viability. Collectively, our results indicate that follicular maturation and gonadal steroids influence Calca and Calcr gene expression in the chicken ovary. We conclude that ovarian CALCA is possibly involved in regulating follicular maturation in the chicken ovary.     

Induction and inhibition of meiotic maturation of amphibian (Rana dybowskii) follicular oocytes by forskolin and cAMP in vitro

Previous studies have demonstrated that direct or indirect elevation of cAMP levels in cultured amphibian ovarian follicles simultaneously stimulated production of oocyte maturation-inducing steroid (progesterone) by the follicles and inhibited oocyte maturation induced by endogenous or exogenous hormone. The duration of cAMP stimulation influenced arrest and reinitiation of oocyte meiotic maturation in ovarian follicles of Rana dybowskii. Addition of forskolin (adenylate cyclase stimulator) to cultured follicles inhibited both progesterone- and frog pituitary homogenate (FPH)-induced oocyte maturation. Similar inhibitory results were obtained when hormone-treated follicles were cultured in the continual presence of cAMP. Oocyte maturation increasingly occurred in follicular oocytes when cAMP or forskolin addition was delayed following treatment with FPH or progesterone. Transient exposure (6-8 hr) of ovarian follicles to forskolin or cAMP markedly stimulated oocyte maturation as well as accumulation of progesterone as measured by radioimmunoassay within the ovarian follicles. Forskolin was more effective than cAMP, at the dose tested, in stimulating progesterone production and accumulation by the follicles. The data demonstrate that transient manipulation (elevation) of cAMP levels in cultured follicles, without added FPH or steroid, was sufficient to initiate oocyte maturation. Results suggest that, with transient exposure to forskolin or exogenous cAMP, there is a sequential increase and decrease in endogenous cAMP levels in the somatic cells and germ cell components of the ovarian follicle. These changes appear to mediate production of maturation-inducing steroid and secondarily allow its effects on the oocyte to be expressed.     

Stimulatory and Inhibitory Effects of Human Follicular Fluid on Amphibian Oocyte Maturation and Ovulation hi vitro

Follicular fluid was collected from individual human ovarian follicles and its effects, alone or in combination with frog pituitary homogenate (FPH), on oocyte maturation and ovulation were assessed following incubation with amphibian ovarian follicles in vitro. Oocyte maturation, with little or no concomitant ovulation, was induced by variable amounts of follicular fluid. Some of the individual follicular fluid samples were very active in inducing oocyte maturation, whereas others were inactive. Frog pituitary homogenates exhibited biologic activity (induced oocyte maturation and ovulation) when incubated in the presence of most follicular fluid samples. However, follicular fluid samples from two individuals inhibited ovulation but not maturation in FPH-treated follicles. These results demonstrate that amphibian follicles remain viable and undergo a number of physiologic changes in the presence of unfractionated human follicular fluid. Under appropriate conditions both stimulatory and inhibitory effects of follicular fluid were observed. These data suggest that amphibian ovarian follicles may provide a simple and independent means for detecting and assaying a number of biologic activities present in follicular fluid obtained from single human and other mammalian ovarian follicles. Such results may provide the basis for dissociating endocrine and cellular interactions which occur during normal and abnormal follicular differentiation.     

Regulation of ovarian steroidogenesis in vitro by follicle-stimulating hormone and luteinizing hormone during sexual maturation in salmonid fish

The regulation of ovarian steroidogenesis in vitro by coho salmon FSH and LH was investigated in intact coho salmon follicles and isolated follicular layers at various stages of oocyte maturation, from late vitellogenesis until the completion of germinal vesicle breakdown (GVBD). In granulosa layers from all stages, LH, but not FSH, stimulated 17alpha,20beta-dihydroxy-4-pregnen-3-one (17, 20beta-P) production. In theca-interstitial layers from all stages, FSH and LH stimulated steroid production, LH being more potent than FSH. The basal steroid output of intact follicles was significantly lower than that of isolated follicular layers, and their response to FSH and LH also differed. First, the intact follicles produced 17alpha-hydroxyprogesterone in response to FSH during the central germinal vesicle stage while theca-interstitial layers did not. Second, estradiol-17beta production was not inhibited by LH during final oocyte maturation in intact follicles, as observed for granulosa layers. Our results indicate that LH is the determining factor regulating the production of the maturation-inducing steroid, 17,20beta-P, and the induction of GVBD in the salmonid ovary. In summary, we have provided evidence for maturation-associated changes in the effects of FSH and LH in the salmonid ovary, which further supports the hypothesis that FSH and LH have distinct functions in the teleost ovary.