Gonadotropin signaling plays a pivotal role in the spermatogenesis of vertebrates, but exactly how gonadotropins regulate the process in non-mammalian species remains elusive. Using a gene knockout approach in zebrafish, we have previously demonstrated the non-canonical action of gonadotropin signaling on spermatogenesis by analyzing four single mutant lines (lhb, lhr, fshb and fshr) and three double mutant lines (lhb;fshb, lhr;fshr and fshb;lhr). In this study, we further investigated the actions of gonadotropins on the testis by establishing three other double-mutant zebrafish lines (lhb;lhr, fshb;fshr and lhb;fshr). All lhb;lhr and fshb;fshr mutant males were fertile. Analysis on the gonadosomatic index and testicular histology in these lhb;lhr and fshb;fshr mutants demonstrated that Lh signaling and Fsh signaling could functionally compensate each other in the testis. Intriguingly, it was found that the lhb;fshr mutant male fish were also morphologically and histologically normal and functionally fertile, a phenomenon which could be explained by the cross-activation of Lhr by Fsh. We have demonstrated this cross-reactivity for the first time in zebrafish. Fsh was shown to activate Lhr using three different assay systems, in which Lh-Fshr activation was also confirmed. Taken together, we conclude that the action of Lh signaling and Fsh signaling is redundant in that either alone can support zebrafish spermatogenesis based on two observations. First, that either Lh signaling or Fsh signaling alone is sufficient to support male fertility. Second, that the two gonadotropin ligands could promiscuously activate both receptors. Apart from revealing the complexity of gonadotropin signaling in controlling male reproduction in zebrafish, this study also shed light toward a better understanding on the evolution of gonadotropin signaling in vertebrates from fish to mammals.

To ascertain the neuroendocrine function of the kisspeptin/GPR54 system in non-mammalian species, full-length cDNAs encoding for Kiss1 and Kiss2 as well as their putative cognate receptors GPR54a and GPR54b, were isolated from goldfish (Carassius auratus). The deduced protein sequences between Kiss1 and Kiss2 in goldfish share very low similarity, but their putative mature peptides (kisspeptin-10) are relatively conserved. RT-PCR analysis demonstrated that the goldfish kiss1 gene (gfkiss1) is highly expressed in the optic tectum-thalamus, intestine, kidney, and testis, while the goldfish kiss2 gene (gfkiss2) is mainly detected in the hypothalamus, telencephalon, optic tectum thalamus, adipose tissue, kidney, heart, and gonads. The two receptor genes (gfgpr54a and gfgpr54b) are highly expressed in the brain regions including telencephalon, optic tectum thalamus, and hypothalamus. Both mature goldfish kisspeptin-10 peptides (gfKiss1–10 and gfKiss2–10) are biologically active as they could functionally interact with the two goldfish receptors expressed in cultured eukaryotic cells to trigger the downstream signaling pathways with different potencies. The actions of gfKiss1–10 and gfKiss2–10 on LH secretion were further investigated in vitro and in vivo. Intraperitoneal administration of gfKiss1–10 to sexually mature female goldfish could increase the serum LH levels. However, this peptide does not significantly influence LH release from goldfish pituitary cells in primary culture, indicating that the peptide does not exert its actions at the pituitary level. On the other hand, gfKiss2–10 appears to be a much less potent peptide as it exhibits no significant in vivo bioactivity and is also inactive on the primary pituitary cells.

Two GPR39 transcripts, designated as sbGPR39-1a and sbGPR39-1b, were identified in black seabream (Acanthopagrus schlegeli). The deduced amino acid (aa) sequence of sbGPR39-1a contains 423 residues with seven putative transmembrane (TM) domains. On the other hand, sbGPR39-1b contains 284 aa residues with only five putative TM domains. Northern blot analysis confirmed the presence of two GPR39 transcripts in the seabream intestine, stomach, and liver. Apart from seabream, the presence of two GPR39 transcripts was also found to exist in a number of teleosts (zebrafish and pufferfish) and mammals (human and mouse). Analysis of the GPR39 gene structure in different species suggests that the two GPR39 transcripts are generated by alternative splicing. When the seabream receptors were expressed in cultured HEK293 cells, Zn^{2 +} could trigger sbGPR39-1a signaling through the serum response element pathway, but no such functionality could be detected for the sbGPR39-1b receptor. The two receptors were found to be differentially expressed in seabream tissues. sbGPR39-1a is predominantly expressed in the gastrointestinal tract. On the other hand, sbGPR39-1b is widely expressed in most central and peripheral tissues except muscle and ovary. The expression of sbGPR39-1a in the intestine and the expression of sbGPR39-1b in the hypothalamus were decreased significantly during food deprivation in seabream. On the contrary, the expression of the GH secretagogue receptors (sbGHSR-1a and sbGHSR-1b) was significantly increased in the hypothalamus of the food-deprived seabream. The reciprocal regulatory patterns of expression of these two genes suggest that both of them are involved in controlling the physiological response of the organism during starvation.

Recent studies have indicated that ghrelin stimulates growth hormone release from the pituitary via the growth hormone secretagogue receptor (GHSR). We have previously isolated two GHSR subtypes from the pituitary of the black seabream Acanthopagrus schlegeli. In the present study, we have cloned and characterized ghrelin from the same fish species at both the cDNA and gene levels. The full-length seabream ghrelin cDNA, isolated from sea-bream stomach using a novel approach by exploiting a single conserved region in the coding region, was found to encode a prepropeptide of 107 amino acids, with the predicted mature ghrelin peptide consisting of 20 amino acids (GSSFLSPSQKPQNRGKSSRV). Embedded in this full-length cDNA is a putative fish orthologue of the recently reported mammalian obestatin peptide. The ghrelin gene in black seabream, obtained by genomic PCR, was found to encompass four exons and three introns, possessing the same structural organization as in tilapia and goldfish, but different from that in rainbow trout. In addition, a 2230-bp 5′-flanking region of the seabream ghrelin gene was obtained by genome walking. Sequence analysis revealed that, as in the case of the human ghrelin gene, there is neither a GC box nor a CAAT box present in the isolated 5′-flanking region. However, a number of putative transcription factor-binding sites different from the human counterpart were found in the 5′-flanking region of the seabream ghrelin gene, suggesting that different cis- and trans-acting elements are involved in controlling their gene expression. Functional activity of this 5′-flanking region was examined by cloning it into the pGL3-Basic vector upstream of the luciferase reporter gene and transfected into various cell lines. Positive promoter activity could only be recorded in the colon-derived Caco-2 cells, suggesting that the cloned 5′-flanking region represents the functional promoter of the seabream ghrelin gene, which exhibits tissue-specific promoter activity. Using reverse transcriptase PCR analysis, expression of ghrelin was detected only in the seabream stomach, but not in the other tissues examined, including the brain, gill, intestine, kidney, liver and spleen. This stomach-specific expression of ghrelin in seabream is subject to regulation, as administration of growth hormone or ipamorelin to the fish in vivo was demonstrated to enhance its expression. Reminiscent of the homologous upregulation found in the transcriptional control of the seabream GHSR gene, a similar homologous regulatory mechanism might also exist in controlling the expression of seabream ghrelin. The identification of both GHSR and ghrelin from a single fish species would facilitate our subsequent studies on the elucidation of the physiological functions of the ghrelin/GHSR system in teleost. The possible existence of obestatin in teleost opens up new research avenues on the somatotropic axis in fish.

Two prolactin receptors (PRLRs) encoded by two different genes were identified in the fugu and zebrafish genomes but not in the genomes of other vertebrates. Subsequently, two cDNA sequences corresponding to two PRLRs were identified in black seabream and Nile tilapia. Phylogenetic analysis of PRLR sequences in various vertebrates indicated that the coexistence of two PRLRs in a single species is a unique phenomenon in teleosts. Both PRLRs in teleosts (the classical one named as PRLR1, the newly identified one as PRLR2) resemble the long-form mammalian PRLRs. However, despite their overall structural similarities, the two PRLR subtypes in fish share very low amino acid similarities (about 30%), mainly due to differences in the intracellular domain. In particular, the Box 2 region and some intracellular tyrosine residues are missing in PRLR2. Tissue distribution study by real-time PCR in black seabream (sb) revealed that both receptors (sbPRLR1 and sbPRLR2) are widely expressed in different tissues. In gill, the expression level of sbPRLR2 is much higher than that of sbPRLR1. In the intestine, the expression of sbPRLR1 is higher than that of sbPRLR2. The expression levels of both receptors are relatively low in most other tissues, with sbPRLR1 generally higher than sbPRLR2. The sbPRLR1 and sbPRLR2 were functionally expressed in cultured human embryonic kidney 293 cells. Both receptors can activate the β-casein and c-fos promoters; however, only sbPRLR1 but not sbPRLR2 can activate the Spi promoter upon receptor stimulation in a ligand-specific manner. These results indicate that both receptors share some common functions but are distinctly different from each other in mobilizing post-receptor events. When challenged with different steroid hormones, the two PRLRs exhibited very different gene expression patterns in the seabream kidney. The sbPRLR1 expression was up-regulated by estradiol and cortisol, whereas testosterone had no significant effect. For sbPRLR2, its expression was down-regulated by estradiol and testosterone, while cortisol exerted no significant effect. The 5′-flanking regions of the sbPRLR1 and sbPRLR2 genes were cloned and the promoter activities were studied in transfected GAKS cells in the absence or presence of different steroid hormones. The results of the promoter studies were in general agreement with the in vivo hormonal regulation of gene expression results. The sbPRLR1 gene promoter activity was activated by estradiol and cortisol, but not by testosterone. In contrast, the sbPRLR2 gene promoter activity was inhibited by estradiol, cortisol, and testosterone.

To address the roles of doublesex and mab-3-related transcription factor 1 (Dmrt1), forkhead transcription factor gene 2 (Foxl2), and aromatase in sex differentiation of Southern catfish, the cDNA sequences of these genes were isolated from the gonads. Dmrt1a and Dmrt1b were found to be expressed in the gonads, being higher in the testis. A low expression level of Dmrt1b was also detected in the intestine and kidney of the male. Foxl2 was found to be expressed extensively in the brain (B), pituitary (P), gill and gonads (G), with the highest level in the ovary, indicating the possible involvement of Foxl2 in the B–P–G axis. Cytochrome P450 (Cyp)19b was found to be expressed in the brain, spleen, and gonads, while Cyp19a was only expressed in the gonads and spleen. All-female Southern catfish fry were treated with fadrozole (F), tamoxifen (TAM), and 17β-estradiol (E₂) respectively, from 5 to 25 days after hatching (dah). The expression levels of these genes were measured at 65 dah. In the F-, TAM-, and FTAM-treated groups, Dmrt1a and Dmrt1b were up-regulated in the gonad, whereas Foxl2 and Cyp19a were down-regulated, while the expression of Cyp19b in the gonad remained unchanged. Furthermore, down-regulation of Foxl2 and Cyp19b was also detected in the brain. In the E₂-treated group, Dmrt1a and Dmrt1b were down-regulated to an undetectable level in the gonad, whereas Foxl2 and Cyp19b were up-regulated in the brain. Consistent with the observed changes in the expressions of these genes, 56, 70, and 80% sex-reversed male individuals were obtained in the F-, TAM-, and F + TAM-treated groups respectively. These results indicate the significant roles of Dmrt1, Foxl2, and Cyp19 in the sex differentiation of Southern catfish.