| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Neuroendocrinology; |
Environmental and Molecular Fish Biology Group, School of Biosciences, Hatherly Laboratories, University of Exeter, Exeter, Devon, EX4 4PS, United Kingdom
ABSTRACT
The mechanisms underlying the initiation of puberty in fish are poorly understood, and whether the Kiss1 receptor (Kiss1r; previously designated G protein-coupled receptor 54; GPR54) and its ligands, kisspeptins, play a significant role, as has been established in mammals, is not yet known. We determined (via real-time PCR) temporal patterns of expression in the brain of kiss1r, gnrh2, and gnrh3 and a suite of related genes in the hypothalamo-pituitary-gonadal (HPG) axis and analyzed them against the timing of gonadal germ cell development in male and female fathead minnow (Pimephales promelas). Full- or partial-length cDNAs for kiss1r (736 bp), gnrh2 (698 bp), and gnrh3 (804 bp) cloned from fathead minnow were found to be expressed only in the brain, testis, and ovary of adult fish. Localization of kiss1r, gnrh2, and gnrh3 within the brain provided evidence for their physiological roles and a likely hypophysiotropic role for GnRH3 in this species (which, like other cyprinids, does not appear to express gnrh1). In both sexes, kiss1r expression in the brain increased at the onset of puberty and reached maximal expression in males when spermatagonia type B appeared in the testis and in females when cortical alveolus-stage oocytes first appeared in the ovary, the timings of which differed for the two sexes. However, kiss1r expression was considerably lower during more advanced stages of spermatogenesis and oogenesis. The expression of kiss1r closely aligned with that of the gnrh genes (gnrh3 in particular), suggesting the Kiss1r/kisspeptin system in fish has a similar role in puberty to that occurring in mammals, and this hypothesis was supported by the induction of gnrh3 (2.25-fold) and kiss1r (1.5-fold) in early-mid pubertal fish injected with mammalian kisspeptin-10 (2 nmol/g wet weight). An intriguing finding, and contrasting that in mammals, was an elevated expression of esr1, ar, and cyp19a2 (genes involved in sex steroid signaling) in the brain at the onset of puberty, and in females slightly in advance of the elevation in the expression of kiss1r.
brain, fathead minnow, fish, gonadotropin-releasing hormone, GPR54, kiss1 receptor, kisspeptin, puberty, sex steroids
In vertebrates, puberty is defined as the process by which a sexually immature animal acquires, for the first time, the capacity to reproduce [1]. In fish, puberty occurs following gonadal sex differentiation and is characterized by the onset of spermatogenesis in males [2] and oogenesis in females [3]. The hypothalamo-pituitary-gonadal (HPG) axis plays a key role in regulating puberty in vertebrates, and is believed to acquire full physiological capacity at this time. GnRH from the hypothalamus stimulates the synthesis and release of the gonadotropins FSH and LH from the anterior pituitary; these hormones act on the gonads to induce oogenesis and spermatogenesis, most crucially via stimulating production of sex steroids [4]. Although this cascade has been well characterized in many fish species, the triggers and mechanisms involved in its activation at the start of puberty are not known for any fish species [4, 5]. An understanding of this process in fish is both of fundamental and comparative interest, but could also have significant importance for the aquaculture industry, where an ability to modify the timing of puberty would be of great benefit in the farming of many fish species [1].
In mammals, the pulsatile secretion of GnRH occurs during late fetal and early neonatal development, but subsequently enters a dormant state until puberty, when there is a reactivation of GnRH release, which is the key event triggering the onset of puberty [6]. The release of GnRH is itself governed by the interplay of excitatory and inhibitory signals (neurohormones and neurotransmitters) acting at the level of the hypothalamus. A "maturational switch" resulting in the enhancement of excitatory signals, together with the lowering of inhibitory signals, is believed to occur at puberty, reactivating GnRH neuronal function. For example, one hypothesis proposed for the initiation of puberty in mammals suggests that it occurs when GnRH neurons become less sensitive to sex steroids that normally inhibit GnRH release, resulting in an increase in GnRH secretion (the "gonadostat" hypothesis) [6].
Recently, exciting developments have been made in the study of puberty in mammals, where it has been demonstrated that a signaling mechanism mediated by the KISS1 receptor (KISS1R), a G protein-coupled receptor and member of the galanin receptor family (previously designated G protein-coupled receptor 54; GPR54), is crucially involved [7]. In 2003, inactivating mutations in the KISS1R gene were identified as a cause for hypogonadotropic hypogonadism, a syndrome characterized by failure of pubertal development due to impaired secretion of FSH and LH [8, 9]. Subsequent studies have indicated essential roles for KISS1R and its natural ligands, the kisspeptins (encoded by the KISS1 gene [10]) in reproduction in mammals, particularly in stimulating GnRH release [11] and thus acting as a gatekeeper for puberty. Additionally, KISS1R and kisspeptins are believed, for example, to mediate the feedback effects of sex steroids on gonadotropin secretion [12, 13] and seasonal aspects of reproduction [14].
In contrast to the situation in mammals, information on the Kiss1r/kisspeptin signaling pathway in fish is extremely limited. To date, cDNAs for kiss1r have been cloned in only three fish species (tilapia, Oreochromis niloticus [15]; cobia, Rachycentron canadum [16]; and grey mullet, Mugil cephalus [17]) and cDNA/protein sequences for kiss1/kisspeptins have not been reported in any non-mammalian species. It has been established, however, that within the fish brain, kiss1r expression is localized to GnRH neurons [15] and is greater in fish at the start of puberty compared with pre- or postpuberty, at a time when gnrh expression is also increased [16, 17]. These findings provide initial evidence that GnRH cells are direct targets for kisspeptins in fish and that kisspeptins likely induce GnRH release at puberty via interactions with Kiss1r, as occurs in mammals. A much more detailed understanding of the expression of kiss1r and gnrh in relation to gonadal development and in both males and females is, however, necessary if we are to determine the precise physiological role(s) of the Kiss1r/kisspeptin system in the regulation of puberty in fish. In addition, information on how the expression patterns for kiss1r and gnrh relate to the expression patterns of other genes in the HPG axis may provide valuable insights into the overall coordination and timing of neuroendocrine signaling during puberty in fish. For example, in immature fish, sex steroids stimulate (rather than inhibit, as in mammals) the development of the HPG axis [18], suggesting that the production of sex steroids and/or the expression of their cognate receptors may be (or may be a part of) a missing link for the initiation of puberty [5].
In this work, full- or partial-length cDNAs for kiss1r, gnrh2 (formerly known as chicken GnRH-II), and gnrh3 (formerly known as salmon GnRH) were first cloned from fathead minnow (Pimephales promelas), an established fish model for studies on reproduction and in ecotoxicology. Furthermore the fathead minnow, being a cyprinid fish, is believed to lack the GnRH1 subtype and so represents an intriguing model for studying alternative relationships between the Kiss1r/kisspeptin and GnRH systems. The expression of the genes cloned was initially determined in different body tissues and different regions of the brain in adults. Detailed investigations were then undertaken into their temporal expression profiles throughout puberty in the brains of males and females in relation to gonadal development and analyzed against those for a suite of related genes in the HPG axis. These genes comprised dopamine d2 receptor (drd2), the receptor that modulates the inhibitory effects of the neurotransmitter dopamine on the actions of GnRH in fish [19, 20]); the steroid hormone receptors estrogen receptor 1 (esr1; previously known as estrogen receptor 
) and androgen receptor (ar); and cytochrome P450 19a2 (cyp19a2), the enzyme responsible for estrogen biosynthesis from testosterone [21]. As part of this work, we also tested for a possible functional role of kisspeptin in stimulating the reproductive axis in fathead minnow via an injection study using mammalian kisspeptin-10. The data presented provide novel insights into the molecular signaling of puberty in fish, highlighting for the first time intriguing differences in kiss1r localization in the brain in fish, kiss1r developmental expression between males and females, and sex steroid signaling and the sensitivity of the brain to sex steroids at the onset of puberty in fish compared with mammals.
The fathead minnow used in these studies were obtained as embryos on the day of fertilization or, for the kisspeptin administration study, as juveniles, from stocks at AstraZeneca's Brixham Environmental Laboratory, Brixham, U.K. Hatching took place between 4 and 6 days postfertilization (dpf), and fish were cultured as described previously [22]. All animal-use procedures were carried out ethically according to U.K. Home Office guidelines.
Cloning of cDNAs for kiss1r, gnrh2, and gnrh3 in Fathead Minnow
To enable the design of primers for real-time PCR amplification of kiss1r, gnrh2, and gnrh3 mRNAs in fathead minnow, their partial- (kiss1r) or full- (gnrh2 and gnrh3) length cDNA sequences were cloned from fathead minnow brain using RT-PCR, 5'-/3'-Rapid Amplification of cDNA Ends (RACE), cloning, and automated fluorescence sequencing as described previously [22]. To obtain a fragment for each gene, RT-PCRs were carried out using pairs of oligonucleotide primers designed from regions of kiss1r, gnrh2, and gnrh3 cDNA sequences conserved between fish species available in the GenBank database (kiss1r: 5'-CAGACAGATGAGGACGGCTA-3'/5'-GGAAGGGAAAGGTCTTCCTG-3'; gnrh2: 5'-GGGGATGTTGCTGTGTCTAA-3'/5'-GTAGGAACTGCTGCAAATGG-3'; gnrh3: 5'-ATGGAGTGGAAGGAAGGTT-3'/5'-TTACACTCTTCCCCGTCTGT-3'; sense and antisense, respectively; MWG-Biotech). For gnrh2 and gnrh3, the partial-length sequences were subsequently extended to full-length via 5'-/3'-RACE using primers specific to the respective fathead minnow sequence. We also attempted to isolate cDNA for gnrh1 (previously known as sea bream GnRH) from fathead minnow by RT-PCR using two sets of oligonucleotide primers (sense: 5'-GCACTGGTCGTATGGACTGA-3' or 5'-GTCAGCACTGGTCGTATGGA-3'; antisense: 5'-TCCTCGGCACAGCCCAGG-3' or 5'-TCCTTTCATTCTGTACATTTTGG-3'). Sequence similarity was analyzed using BLASTn [23] and multiple sequence alignment.
Tissue Expression of kiss1r, gnrh2, and gnrh3
For determining the tissue expression of kiss1r, gnrh2, and gnrh3 in fathead minnow, eight male and eight female adult fish were killed by terminal anesthesia with benzocaine (0.5 g/L; ethyl-p-aminobenzoate; Sigma, Poole, U.K.). Brain, gill, liver, gonad, intestine, and muscle tissues were collected from each fish, snap-frozen in liquid nitrogen, and stored at –80°C until use for RNA extraction.
Expression of kiss1r, gnrh2, and gnrh3 in Different Brain Regions
For determining the expression of kiss1r, gnrh2, and gnrh3 in different regions of the brain, seven male and seven female adult fish were killed as described above. The brain of each fish was divided into eight discrete regions: (a) olfactory bulbs and tracts, (b) telencephalon and preoptic area, (c) optic nerves, (d) optic tectum, (e) pituitary, (f) hypothalamus/thalamus including midbrain tegmentum, (g) cerebellum, and (h) medulla oblongata. For each sex, the tissue for each brain region from the seven individuals was pooled, snap-frozen in liquid nitrogen, and stored at –80°C until use for RNA extraction.
Developmental Expression of kiss1r, gnrh2, gnrh3, drd2, esr1, ar, and cyp19a2
For determining the developmental expression of kiss1r, gnrh2, gnrh3, drd2 (GenBank: DT162482), esr1 (GenBank: AY775183), ar (GenBank: AY727529), and cyp19a2 (GenBank: AJ277866) in fathead minnow, fish were killed as described above at 16 incremental time points during development (25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 90, 100, 120, and 140 dpf). At each time point, 15 fish were sampled. Each fish was measured for fork length (mm) and wet weight (g; for lengths and weights of the fish sampled, see Supplemental File 1 available at www.biolreprod.org) and whole brains were dissected from individual fish, immediately snap-frozen in liquid nitrogen, and stored at –80°C until use for RNA extraction. In order to determine the sex and stage of gonadal development of each fish, the body of each fish was fixed in Bouin fixative (Raymond A. Lamb, Ltd., Eastbourne, U.K.) and transferred into 70% ethanol (Fisher Scientific, Loughborough, U.K.) until use for gonadal histology.
Expression of kiss1r, gnrh2, and gnrh3 Following Administration of Mammalian Kisspeptin
To test for a functional role of kisspeptin in stimulating the reproductive axis in fathead minnow, early-mid pubertal fish were injected with mammalian kisspeptin-10 (the minimal kisspeptin sequence necessary for Kiss1r activation in mammals), and expression levels of kiss1r, gnrh2, and gnrh3 in the brain were compared with those in sham-injected controls. For each treatment, 15 fish (mean ± SEM fork length: 36.11 ± 1.189 mm; mean ± SEM wet weight: 0.599 ± 0.062 g) were injected into the i.p. cavity with either mouse kisspeptin-10 (2 nmol/g wet weight, at a volume of 20 µl/g wet weight; Phoenix Europe GmbH, Germany) or water (sham-controls). At 10 h postinjection, all fish were killed as described above, and individual brains were collected, snap-frozen in liquid nitrogen, and stored at –80°C until use for RNA extraction.
Whole body transverse sections (5–10 mm thick) were cut from the mid-body region of each fish between the operculum and the anal fin to encompass the gonad, dehydrated in industrial methylated spirits, and embedded in paraffin wax using an automated tissue processor (Shandon Hypercenter XP; Shandon Life Sciences Ltd., Runcorn, U.K.). Sections were cut at 5-µm thickness at different points along the length of the gonads, mounted, stained with hematoxylin and eosin, and analyzed under light microscopy. Germ cells present in the gonad were staged as described [24] and, for the earliest stages, gonads were examined for other structural elements that helped to identify sex (e.g., the manner in which somatic cells were distributed relative to primordial germ cells, which is different in males compared with in females [24], and gonadal ducts).
RNA Extraction and Reverse Transcription
Total RNA was extracted from the tissue samples, DNase-treated, and reverse transcribed to cDNA as previously described [22]. For the brain samples in the developmental expression study, the cDNA was RNase-treated, purified, and the concentration quantified using the Quant-iT OliGreen ssDNA reagent kit (Invitrogen) as previously described [25]. The samples were then normalized (by dilution with molecular-grade water) such that all samples contained equal amounts of cDNA template for real-time PCR. This normalization was conducted such that we could carry out real-time PCR analysis on these samples without the need for subsequent normalization to a "housekeeping" gene (to account for differences in the amounts of input cDNA) because we could not find such a gene whose expression did not change across all 16 developmental stages (see below).
Development of real-time PCR assays for kiss1r, gnrh2, and gnrh3. Primers specific for fathead minnow kiss1r, gnrh2, gnrh3, and drd2 were designed for use in real-time PCR with Beacon Designer 3.0 software (Premier Biosoft International, Palo Alto, CA) according to manufacturer's guidelines and purchased from MWG Biotech (kiss1r: 5'-GGCTTGAGGACGGCTATTG-3'/5'-AGATGGTAATGACAGGCAGTAG-3'; gnrh2: 5'-AAAGCGAGCAGATAGACATTTACG-3'/5'-TGAGGGCATCCAGCAGTATTG-3'; gnrh3: 5'-AGCACTGGTCATACGGTTGG-3'/5'-ACCTTCAGCATCCTCCTCATTC-3'; drd2: 5'-ACCACCACCAACTACCTGATAG-3'/5'-TGCTGAACCGCCACTCTC-3'; sense and anti-sense, respectively). Assays were optimized and validated for real-time PCR using SYBR Green chemistry as described previously [22]. Assays had detection ranges of at least three orders of magnitude. Specificity of primer sets throughout this range of detection was confirmed by the observation of single amplification products of the expected size and melting temperature. All assays were quantitative with standard curve (mean threshold cycle [Ct] vs. log cDNA dilution) slopes of –3.079 (kiss1r), –4.595 (gnrh2), –3.646 (gnrh3), and –3.194 (drd2), translating to high PCR efficiencies (E) of 2.11 (kiss1r), 1.65 (gnrh2), 1.88 (gnrh3), and 2.06 (drd2). Over the detection range, the linear correlation (R2) between the mean Ct and the logarithm of the cDNA dilution was >0.99 in each case.
Real-time PCR analyses.
Real-time PCR using SYBR Green chemistry was performed for fathead minnow kiss1r, gnrh2, gnrh3, drd2, esr1, ar, and cyp19a2 with the iCycler iQ Real-time Detection System (Bio-Rad Laboratories, Inc., Hercules, CA) using the protocols described previously [22]. For the kiss1r, gnrh2, gnrh3, and drd2 assays, annealing was at 62.0°C, 61.8°C, 63.0°C, and 61.0°C, respectively. The assays for fathead minnow esr1, ar, and cyp19a2 were as previously described [22, 26]. For each gene, each sample was analyzed in triplicate. Relative expression levels were calculated using an efficiency-corrected version of the arithmetic comparative 2-
Ct method, as described previously [22]. The housekeeping gene ribosomal protein l8 (rpl8) also measured in each sample (as in [22]) was used for relative quantitation because its expression did not change between the tissues analyzed or following any of the treatments. For the developmental expression study, expression levels were calculated using the arithmetic 2–Ct method [27] but using the actual measured E value for each gene rather than assuming E = 2 (efficiency correction). The 2–Ct method is a modification of the 2–Ct method that was developed to enable normalization to a measurement external to the PCR experiment (in this case, input cDNA) because we could not find a housekeeping gene that did not change in its expression between all 16 developmental stages for use in relative quantitation.
Statistical differences (P < 0.05) between experimental groups were assessed by Student t-test or one-way ANOVA, followed by Dunn post hoc test (SigmaStat 3.10; Jandel Scientific Software) or nonparametric alternatives when appropriate. All data are shown as means ± SEM.
Cloning of kiss1r, gnrh2, and gnrh3 in Fathead Minnow
A partial-length cDNA for fathead minnow kiss1r (736 bp) and full-length cDNAs for fathead minnow gnrh2 (698 bp) and gnrh3 (804 bp) were amplified by RT-PCR and RACE techniques, which coded for putative partial-length Kiss1r and full-length GnRH2 and GnRH3 proteins of 245, 86, and 94 amino acids, respectively (submitted to GenBank, accession nos. EF672266, EF672264, and EF672265). These cDNAs and putative proteins displayed high sequence identities with the respective sequences in other fish and mammalian species, confirming their identities. Attempts to isolate a cDNA for gnrh1 in fathead minnow were unsuccessful.
Tissue Expression of kiss1r, gnrh2, and gnrh3
In adult fish, expression of kiss1r, gnrh2, and gnrh3 occurred only in brain and gonad and was not detected in gill, liver, intestine, or muscle (Fig. 1). All three genes were more highly expressed in brain compared with in gonad (on average, 14-, 40-, and 3-fold greater for kiss1r, gnrh2, and gnrh3, respectively). There was no evidence for sexual dimorphism in the expression of kiss1r in brain. Although there appeared to be greater expression of gnrh2 and gnrh3 in the brains of females compared to that of males, these differences were not statistically significant (P = 0.095 and P = 0.145, respectively). In contrast, however, in gonad, kiss1r and gnrh2 were more highly (5- and 2-fold, respectively) expressed in males compared with in females, while gnrh3 showed the opposite pattern, with 3-fold higher expression in females compared with in males.
|
Expression of kiss1r, gnrh2, and gnrh3 in Different Regions of the Brain
Expression of kiss1r, gnrh2, and gnrh3 in different regions of the brain in adult males and females is shown in Figure 2. Expression of kiss1r was detected in all brain regions, but predominantly in the telencephalon (which included the preoptic area), where it had approximately 140-fold higher expression than in the optic nerves, pituitary, and cerebellum. Moderate levels of kiss1r expression occurred in the olfactory bulbs and tracts, optic tectum, and hypothalamus/midbrain tegmentum but, elsewhere in the brain, kiss1r expression was low. Expression of gnrh2 was predominantly in the region that included the hypothalamus and midbrain tegmentum (approximately 250-fold higher expression than in the forebrain regions, optic nerves, optic tectum, and hindbrain regions, where only very low levels of expression were observed). No expression of gnrh2 was detected in the pituitary. Expression of gnrh3 was predominantly in the forebrain regions (both the olfactory bulbs and tracts and telencephalon, which included the preoptic area), where it was approximately 200- to 1000-fold more highly expressed than in the optic tectum and hindbrain regions (cerebellum and medulla oblongata). Moderate levels of expression were observed for gnrh3 in the pituitary of females and in the hypothalamus/midbrain tegmentum region, but no expression was detected in the optic nerves. There were apparent sex differences for kiss1r, gnrh2, and gnrh3 in some regions of the brain, including for kiss1r in the optic tectum and hypothalamus/midbrain regions (higher expression in males), for gnrh2 in the optic tectum (higher expression in males) and telencephalon (higher in females), and for gnrh3 in the pituitary, olfactory bulbs and tracts, and telencephalon (higher expression in females).
|
Expression of kiss1r in the Brain During Puberty in Relation to Gonadal Development
Males. Expression of kiss1r in the brains of males during puberty and how this related to the progression of testicular development is shown in Figure 3. Expression of kiss1r was low in males aged between 25 and 30 dpf. At this time, the gonads of males contained only a few primordial germ cells (i.e., spermatogenesis had not been initiated) and they were categorized as males based on a single point of attachment of the gonad to the peritoneal wall and an even distribution of somatic and germ cells across the gonad. At 35 dpf, expression of kiss1r had increased to a significantly higher level (2.5-fold higher than that at 25 dpf) and this coincided with the appearance of spermatogonia type A in the testis. Subsequently, however, between 40 and 50 dpf, expression dropped back to the level seen at 25 dpf. At 55 dpf, there was a second and sharp significant increase in kiss1r expression, associated with a large increase in the number of spermatogonia type A (arising through mitosis) and a rapid increase in testis size. Maximal expression of kiss1r in males (to a level 8-fold greater than at 25 dpf) occurred at 60 dpf and corresponded with the onset of meiosis (appearance of spermatogonia type B) and the formation of lobules within the testis. Expression of kiss1r first declined from its maximal level at 70 dpf and, subsequently, declined further until it was maintained at a stable level between 100 and 140 dpf. During this time, spermatocytes (at 100 dpf) and spermatids/spermatazoa (at 120 dpf) appeared in the testis, signifying the final stages of spermatogenesis. The level of kiss1r expression between 100 and 140 dpf was not significantly different from that in prepubertal (25 dpf) males.
|
Females. Expression of kiss1r in the brain of females during puberty and how this related to germ cell development in the ovary is shown in Figure 4. At 25 dpf, ovarian differentiation had already occurred (the ovary was characterized by two points of attachment to the peritoneal wall and it was surrounded by an ovarian cavity) and, between 25 and 35 dpf, development progressed from ovaries containing predominantly oogonia and early-meiotic oocytes (up to pachytene stage; at 25 dpf) to ovaries containing predominantly transforming pachytene stage oocytes and primary oocytes at the early diplotene (perinucleolar) stage (35 dpf). At this time (25–35 dpf), expression of kiss1r in the brain was low, but significantly increased from 40 dpf, coincident with rapid growth of the ovary and an increase in the number and size of perinucleolar stage oocytes, to a maximal level at 70 dpf (which was 11-fold higher than at 25 dpf). Expression of kiss1r in the brain was reduced again at 75 dpf (to 30% of the level at 70 dpf and similar to the level at 25 dpf), coincident with the appearance of the first oocytes at the cortical alveolus stage (in most females). Moderate and relatively stable levels of kiss1r expression occurred in brains of females at the more advanced stages of ovarian development (vitellogenic-stage oocytes at 120–140 dpf). Expression of kiss1r was higher (4-fold) in sexually mature females (140 dpf) compared with prepubertal females (25 dpf; P < 0.001).
|
Expression of gnrh2, gnrh3, drd2, esr1, ar, and cyp19a2 Compared with kiss1r in the Brain During Puberty
Males. Expression of gnrh2, gnrh3, drd2, esr1, ar, and cyp19a2 in the brain of males during puberty and how this related to expression of kiss1r is shown in Figure 5. The developmental expression of gnrh3 and drd2 most closely matched that for kiss1r, in that low and stable expression was observed between 25 and 30 dpf, followed by an initial increase in expression at 35 dpf and a subsequent reduction to the level at 25 dpf. There was then a second increase in expression at 55 dpf to maximal expression at 60 dpf (when expression was 5-fold greater than at 25 dpf), and a drop in expression at 65 dpf (for drd2) or 70 dpf (for gnrh3). However, unlike that observed for kiss1r and drd2, gnrh3 expression increased again at 120 dpf (P = 0.032) and was greater (3-fold) in sexually mature males compared with in prepubertal males (P = 0.005). In contrast, for gnrh2, esr1, and ar, there was only a single peak in expression during testicular development. For esr1 and ar, the first significant increase in expression above that at 25 dpf occurred at 50 dpf, which was delayed compared with the first small increase in expression of kiss1r (that occurred at 35 dpf) but in advance of the second (and main) increase in kiss1r expression (that occurred at 55 dpf). Maximal expression of esr1 and ar occurred at 55 dpf, which was also slightly in advance of the time of the maximal expression of kiss1r (at 60 dpf). For cyp19a2, the expression pattern during early development (up to 100 dpf) was similar to those for esr1 and ar (with expression first increasing at 50 dpf—again after the first peak in expression of kiss1r but slightly in advance of the second and main peak of kiss1r expression), but further increases in expression occurred later during development, between 100 and 140 dpf, with maximal expression (10-fold greater than in prepubertal males) in sexually mature males at 140 dpf.
|
Females. Expression of gnrh2, gnrh3, drd2, esr1, ar, and cyp19a2 in the brain of females during puberty and how this related to expression of kiss1r is shown in Figure 6. During ovarian development, the neural expression profiles for gnrh2, gnrh3, drd2, esr1, ar, and cyp19a2 mapped closely with those for kiss1r in that they were characterized by a single peak in expression during the middle of the study period (generally between 40 and 80 dpf). For gnrh2 and gnrh3, consistently low-level expression occurred between 25 and 35 dpf, followed by an increase in expression at 40 dpf (5- and 2-fold, respectively) to a maximal level at 70 dpf (when expression was 12- and 18-fold higher than at 25 dpf, respectively). Expression of gnrh3, however, then first dropped from this maximal level slightly later than that for kiss1r and gnrh2 (80 dpf compared to 75 dpf). From this reduced expression, both gnrh2 and gnrh3 progressively increased (4- and 3-fold, respectively) to 140 dpf (an increase that was not seen for kiss1r). For esr1, ar, and cyp19a2, the initial increase in expression occurred slightly in advance of the elevation in kiss1r expression (at 30 dpf, compared with at 40 dpf) and then was reduced from maximal levels from 70 dpf onwards (again, in advance of the reduction in the expression of kiss1r). For cyp19a2, as for gnrh2/gnrh3, after the initial peak in expression around the time of the onset of puberty, there was a further progressive increase with sexual development to maturity (of 3.5-fold) while expression of esr1 and ar remained at a static level during this time. Expression of drd2 was also elevated slightly in advance of the elevation of kiss1r (at 35 dpf compared to 40 dpf) but did not first decrease from its maximal level until 80 dpf, and remained at a low and static level thereafter. As for kiss1r, expression of gnrh2, gnrh3, and cyp19a2 was higher in sexually mature females than in prepubertal females (8- to 10-fold), but this was not observed for esr1, ar, and drd2.
|
Expression of kiss1r, gnrh2, and gnrh3 Following Administration of Mammalian Kisspeptin
There was significantly higher expression of both kiss1r (approximately 1.5-fold) and gnrh3 (approximately 2.25-fold) in the brains of the early- to mid-pubertal fish 10 h following i.p. injection with 2 nmol/g wet weight mammalian kisspeptin-10 compared with in the sham-injected controls (Fig. 7). There was, however, no significant difference in the expression of gnrh2 in the brains of the kisspeptin- and sham-injected fish (Fig. 7).
|
To understand the coordination and timing of neuroendocrine signaling regulating the initiation of puberty in fish, we cloned cDNAs for kiss1r, gnrh2, and gnrh3 and studied their expression in relation to a wider suite of transcripts in the HPG axis (drd2, esr1, ar, cyp19a2) in the brain throughout sexual development in both male and female fathead minnow. Most families of fish (including all of those in which kiss1r has been studied to date) have three distinct GnRH systems that express gnrh1, gnrh2, and gnrh3 and are located in the preoptic area, midbrain tegmentum, and terminal nerve areas of the brain, respectively. In these fish species, the GnRH1 subtype originating from the preoptic area appears to be the only form that induces gonadotropin release from the pituitary. Cyprinid fish, however, appear to have GnRH2 and GnRH3 subtypes only (e.g., goldfish, Carassius auratus [28]; roach, Rutilus rutilus [29]; zebrafish, Danio rerio [30]; and fathead minnow [our work in this study]), providing intriguing models for studying alternative relationships between the Kiss1r/kisspeptin and GnRH systems.
Our initial studies on kiss1r, gnrh2, and gnrh3 in adult fathead minnow established that their expression was localized to the brain and gonad only, with significantly higher levels of expression in the brain. For kiss1r in females, this finding is in contrast to that observed in mature female grey mullet, where kiss1r expression was 2-fold higher in ovary compared with in brain, although in immature females the situation was reversed [17]. Indeed, in mammals, ovarian expression of kiss1 is believed to play a local role in ovulation and correlates with ovarian maturation and plasma LH levels [31], effects which are presumably mediated by Kiss1r. Our finding on the expression of kiss1r in testis of fathead minnow is the first such report in fish and, given its relatively high level of expression in this tissue, further investigation is warranted to determine its function here. In mammals, kiss1r is also expressed in other tissues, including placenta, pituitary, pancreas, spinal cord, and, in contrast with our findings, in liver and intestine [32, 33]. Its function(s) in these tissues, however, is (are) not yet known. Expression of gnrh has also been demonstrated in tissues other than the brain in both fish and mammals, including in the pituitary, skeletal muscle, heart, liver, kidney, spleen, placenta, mammary gland, ovary, and testis [34, 35], again suggesting diverse physiological roles for GnRHs (and additional to effecting gonadotropin secretion), as for kiss1r.
To evaluate the potential roles of neural Kiss1r, GnRH2, and GnRH3 in fathead minnow, we compared the expression of their mRNAs between eight discrete regions of the brain. As expected, the two gnrh subtypes showed different patterns of expression and these were generally consistent with those observed in the other cyprinid fish species [28–30]. Expression of gnrh2 predominated in the region that included the midbrain tegmentum, consistent with its localization in other jawed vertebrates studied. However, gnrh2 expression also occurred (albeit at far lower levels) in many other regions of the brain, as has been reported for the goldfish (but contrasting with that in roach and zebrafish). In common with other cyprinids, gnrh3 was expressed both in the olfactory bulbs and tracts (where the terminal nerve GnRH system is located) and in the region corresponding to the telencephalon and preoptic area. In species that possess three GnRH subtypes, the preoptic area is typically where gnrh1 expression occurs, and several lines of evidence (including axon projection to the pituitary, relative abundance of peptide in the pituitary, and seasonal fluctuation of mRNA/peptide levels in the brain and pituitary) have lead to a consensus that the GnRH subtype produced in this region is the form that regulates pituitary function. Therefore, these data provide evidence that GnRH3 is likely to be the hypophysiotropic form of GnRH in the fathead minnow, as has been demonstrated in other fish species that possess only two GnRH subtypes [30] (an exception to this is in the goldfish, where both GnRH2 and GnRH3 regulate pituitary function [36–38]). Interestingly, kiss1r showed a much more widely dispersed expression within the brain compared with the gnrh genes. The highest level of kiss1r expression, however, occurred in the regions where expression of gnrh2 and, most notably, gnrh3 predominated. There are no other studies available on regional expression of kiss1r in the brain for fish, but in mammals, kiss1r is also expressed throughout the brain, including in regions where GnRH neurons are not present [39, 40], suggesting some neuronal Kiss1r signaling may be unrelated to GnRH secretion.
Consistent with the lack of any sexual dimorphism in kiss1r expression in whole brain in fathead minnow, there did not appear to be any differences between the sexes in kiss1r expression in the majority of the dissected regions of the brain analyzed. A possible exception was in regions of the midbrain where kiss1r expression was slightly higher in males. To date, there have not been any other studies in fish examining sex-related differences in kiss1r expression in the brain. Contrasting with that seen for kiss1r, for both gnrh subtypes in whole brains there was a trend for higher expression in females, but the high level of variation between fish meant these differences were not statistically significant. Sex differences in GnRH content or cell number in whole brain have been observed in studies on other animal classes [41, 42], but with no apparent consistency, suggesting there may be species or developmental-stage specificity. In contrast, within certain regions of the brain we did observe some sex-related differences in the expression of both gnrh2 and gnrh3. Two examples of this were for gnrh3 in the pituitary, where much higher expression was observed in females, and in the hypothalamus, where expression was higher in males. Interestingly, the latter observation was also seen in studies on sexually mature goldfish and was attributed to the decrease in hypothalamic gnrh3 expression that is known to occur in females during ovarian maturation and vitellogenesis [28, 43].
In the brain, the significant increase in expression of kiss1r at the onset of puberty in both males and females is in agreement with recent findings in fish [16, 17] and mammals [13, 44]. Interestingly, the main increase in neural kiss1r expression in fathead minnow occurred immediately following (rather than prior to) the appearance of spermatagonia type A (in testis) and perinucleolar-stage oocytes (in ovary) at a time when the numbers of these germs cells increased dramatically. In males, however, there was also a smaller additional increase in kiss1r expression coincident with the appearance of spermatagonia type A in the testis (and prior to the main increase). It is difficult to compare our findings with those for the two published studies on other fish species (grey mullet and cobia) due to the limited scale of the temporal analysis and gonadal histology in those studies. In grey mullet, expression of kiss1r was 4-fold higher in females at an early stage of puberty, the ovaries of which contained perinucleolar oocytes, compared with in females at intermediate or advanced stages of puberty, when the ovaries contained oocytes at the primary/secondary or tertiary yolk globule stages, respectively [17]. In cobia, kiss1r expression was 3-fold higher in males at early puberty (when the testis contained many spermatocytes and a few spermatids) compared with in fish of unknown sex and gonadal development from an earlier sampling point, and was greater than in females of the same age in which puberty had not been initiated [16]. Together, these findings are strongly supportive of a role of the Kiss1r/kisspeptin system in puberty in fish. In which cells/regions kiss1r expression changes during puberty, however, and the possibility of differential regulation of kiss1r expression in the different cells/regions (some of which may not relate to the regulation of the pubertal process) has yet to be delineated in the regulation of the pubertal process in fish.
The differences noted in the dynamics of kiss1r expression during puberty between male and female fathead minnow has not been shown previously for any fish species and likely result from the differences in the timing of sexual development in this species (e.g., the main pubertal increase in kiss1r expression occurred earlier in females [at 40 dpf] than in males [at 55 dpf], consistent with the advanced timing of sexual development in females compared with males). A further intriguing difference between the sexes in the expression of kiss1r in the brain was the higher level in females at sexual maturity compared to prepuberty, which did not occur in males. Interestingly, this difference between the sexes has also been observed in rats, where adult females had significantly higher expression of Kiss1r than prepubertal females, while prepubertal males had similar levels of Kiss1r expression compared to adult males [13]. In fish, a higher neural expression of kiss1r at sexual maturity was also observed in female mullet [17], and, in contrast with our findings, kiss1r was expressed in a significantly higher percentage of GnRH neurons in mature male tilapia compared with immature males [15].
The close alignment of the expression profiles for kiss1r and both gnrh genes in the brain during puberty is in keeping with findings for multiple gnrh genes in other fish [16, 17]. If kisspeptin stimulates GnRH in fish, as occurs in mammals, an increased expression of kiss1r in GnRH cells at puberty would likely function to increase their sensitivity to kisspeptin at this time (assuming enhanced expression of kiss1r translates into an increase in the number of Kiss1r proteins), and thus contribute to an increase in GnRH release at the activation of puberty. Such an increase in sensitivity to kisspeptin at puberty has been reported in the rat and mouse [45, 46], although the magnitude of this sensitization suggests that further factors, such as an increase in the efficiency of Kiss1r coupling to its intracellular effectors, additionally contribute to this response [45].
To establish whether kisspeptin (acting via the Kiss1r) stimulates the GnRH system in fish, we measured gnrh expression in pubertal fish injected i.p. with mammalian kisspeptin-10 (2 nmol kisspeptin-10 per gram wet weight) and found an enhanced expression of gnrh3 in the brain. This finding in the fathead minnow, together with some other very recent data showing stimulatory effects (in a time- and dose-dependent manner) of mammalian kisspeptin on the HPG axis (including an increase in plasma sex steroids) in tilapia [47], provides support for a similar role of kisspeptin in fish as in mammals. Moreover, our result suggests that an increase in endogenous kisspeptin at the onset of puberty may have been responsible for the pubertal activation in gnrh transcription seen. Although the majority of evidence for reproductive effects of kisspeptin in mammals has come from the demonstration of stimulatory effects on release of the GnRH protein [13], a very recent study with immortalized mammalian GnRH neuron cells has also shown stimulatory effects of kisspeptin on gnrh expression in the brain (5-fold up-regulation by 10 nM kisspeptin-10) [48]. Of further interest, our results demonstrate, for the first time, direct effects of kisspeptin on GnRH3. This is again in contrast to the situation in mammals, where there is no evidence that kisspeptin regulates any form of GnRH other than GnRH1 (mammalian GnRH). Some differences between the GnRH system in mammals and fish are already well known, in particular the lack of a hypothalamo-hypophysial-portal system in fish (GnRH is instead delivered to the pituitary by direct neuronal innervation). Additional quantification of kiss1r expression following kisspeptin-10 administration in our study demonstrated an autoregulatory effect of kisspeptin on its own receptor. As for gnrh, this provides evidence that the increases in kiss1r expression observed at puberty in fathead minnow brain may have been (at least partially) due to increases in the levels of kisspeptin.
Interestingly, a closer association with kiss1r expression was observed for gnrh3 compared with gnrh2 in fathead minnow. Differences between the expression of the different gnrh subtypes and the expression of kiss1r have also been observed in other fish species [15–17], and this is consistent with the differences in function of the different GnRH subtypes in fish (discussed above). The tighter alignment of kiss1r expression with that of gnrh3 compared with gnrh2 is consistent with the likely role of GnRH3 as the hypophysiotropic form of GnRH in this species. Further support for this hypothesis comes from the induction of gnrh3, but not gnrh2, by mammalian kisspeptin-10 in our injection study. By contrast, the midbrain GnRH2 form, the most highly conserved GnRH isoform, is not thought to play a major role in the release of gonadotropins from the pituitary and is postulated to have neuromodulatory or paracrine functions that may influence reproduction in a less direct manner [49]. There are data in mammals, birds, amphibians, and fish, for example, supporting a role for GnRH2 in reproductive behavior [50, 51].
In parallel with the increases in expression of kiss1r/gnrhs at puberty, we observed increases in the expression of drd2, the receptor that modulates the inhibitory effects of the neurotransmitter dopamine on the actions of GnRH, which operates via both effects on GnRH neurons blocking the synthesis of the peptide or inhibiting its release [19], and direct effects at the level of the pituitary gonadotrophs [20]. The high pubertal expression of drd2 is consistent with the high dopaminergic inhibition reported for other cyprinid species [52] and the known stimulatory effect of estrogens (which dramatically increase in the plasma in fish at puberty) on dopamine production and drd2 expression [53, 54]. Our results suggest that an inhibition of the GnRH signal is required for a tight control in HPG axis output in fathead minnow. However, though the role of dopamine in puberty and early gametogenesis in fish has yet to be fully established (in contrast to its role in ovulation/spermiation), a drop in dopaminergic inhibition at puberty does not appear to be responsible for the initiation of puberty in this species, as was proposed for eel (Anguilla anguilla [55]) and spadefish (Chaetodipterus faber [56]).
The increases in the expression of genes involved in estrogen biosynthesis (cyp19a2) and sex steroid signaling (esr and ar) at puberty are in agreement with the central importance of sex steroids as positive feedback regulators of the HPG axis in fish, for example via enhancing brain and pituitary GnRH levels, pituitary responsiveness to GnRH, and/or pituitary gonadotropin levels [57, 58]. Because estrogens of gonadal origin are not believed to pass the blood-brain barrier, cytochrome P450 aromatase (the enzyme encoded by cyp19a2 that is responsible for the conversion of androgens into estrogens) is believed to be of particular importance in estrogen signaling in the brain [21]. An increased responsiveness of the brain to sex steroids at puberty in fish is in contrast to the situation in mammals, where decreased responsiveness to sex steroids (which have inhibitory rather than stimulatory effects on the HPG axis) is believed to occur, enabling puberty to take place. Indeed, in mammals at puberty, decreases (rather than increases) occur in estrogen receptor immunoreactivity in the ventral medial hypothalamus, an area of the brain that mediates negative feedback regulation of the HPG axis [59], and in hypothalamic aromatase activity [60]. In males, it is possible that the first small increase in kiss1r expression at 35 dpf served as the stimulus for the subsequent increases in expression of esr1, ar, and cyp19a2 expression, because up-regulations in these genes are normally associated with increased plasma levels of androgens and estrogens [21], which occur downstream of GnRH production. However, interestingly, in females the pubertal increases in esr1, ar, and cyp19a2 expression occurred slightly in advance of those for kiss1r and the gnrhs.
Whether an increase in sex steroids or sex steroid signaling contributed to the high kiss1r expression (or vice versa) cannot yet be established. Although few studies in mammals (and none in fish) have investigated possible regulation of kiss1r expression by sex steroids, a very recent mammalian study demonstrated a 6-fold up-regulation in immortalized GnRH neurons (GT1–7 cells) after 24-h treatment with 10 nM estradiol [48]. Further evidence, albeit indirect, for kiss1r regulation by sex steroids comes from the finding that the increase in KISS1R in monkeys during development occurred only in intact (and not in agonadal) animals [44]. Expression of the KISS1 gene (which codes for kisspeptins, the ligands for Kiss1r) is also modulated by sex steroids in the mammalian brain, and thus (in mammals) the Kiss1r/kisspeptin system is a very likely candidate for relaying the feedback effects of sex steroids on GnRH secretion. Interestingly, this regulation occurs in a brain region-specific manner that correlates with the different functions of the different brain regions, with kiss1 inhibited by sex steroids in the arcuate nucleus (arc; which mediates sex steroid negative feedback), but increased in the anteroventral periventricular nucleus (AVPV) (which mediates sex steroid positive feedback effects on the preovulatory LH surge in females) [61, 62]. Furthermore, while the effects of estrogens in both regions are mediated via the estrogen receptor (interestingly ESR1, not ESR2 [13, 61]), the effects of androgens are mediated by different steroid receptors depending on the brain region, with both androgen and estrogen receptors playing a role in the arc, but only estrogen receptors (after aromatization of testosterone to estradiol) in the AVPV [62].
In conclusion, we have provided further evidence supporting a role for the neural Kiss1r/kisspeptin system in puberty in fish and have provided novel insights into the coordination and timing of molecular signaling of puberty, including the role of sex steroid signaling, in fish, which clearly differs from that observed in mammals. Our data have further provided evidence of a likely role for the Kiss1r/kisspeptin system in regulating ovarian and testicular function locally in fish, and this warrants further investigation. The cloning of kiss1 in fish, production of homologous fish kisspeptin proteins for functional studies, and further studies on the neural circuits controlling kiss1r/kiss1 expression and their regulation by gonadal steroids in fish are now required.
ACKNOWLEDGMENTS
The authors would like to acknowledge Gregory Paull at the University of Exeter for carrying out the kisspeptin injections.
Authors' contributions. A.L.F. carried out the experiments, fish culture, real-time PCR, histology and data analyses, and wrote the manuscript. JW.D. carried out the cloning of the gnrh genes. R.vA. carried out the cloning of kiss1r and participated as a supervisor to JW.D. C.R.T. and R.vA. conceived the idea for the study, oversaw the experiment designs and edited the manuscript.
FOOTNOTES
1A.L.F. was funded by a Ph.D. studentship from the Biotechnology and Biosciences Research Council (BBSRC) and various research grants (C.R.T.). R.v.A. was funded by the Natural Environment Research Council (NERC) and the U.K. Environment Agency (Grant NE/C002369/1). J.W.D. was supported by the European Community (Socrates/Erasmus Programme). 
Correspondence: 2FAX: 44 1392 263700; e-mail: a.l.filby{at}exeter.ac.uk
3These authors contributed equally to this work.
Received: 14 June 2007.
First decision: 10 July 2007.
Accepted: 23 October 2007.
REFERENCES
CT method Methods 2001 25402–408[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  F. Gaytan, M. Gaytan, J. M. Castellano, M. Romero, J. Roa, B. Aparicio, N. Garrido, J. E. Sanchez-Criado, R. P. Millar, A. Pellicer, et al. KiSS-1 in the mammalian ovary: distribution of kisspeptin in human and marmoset and alterations in KiSS-1 mRNA levels in a rat model of ovulatory dysfunction Am J Physiol Endocrinol Metab, March 1, 2009; 296(3): E520 - E531. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kitahashi, S. Ogawa, and I. S. Parhar Cloning and Expression of kiss2 in the Zebrafish and Medaka Endocrinology, February 1, 2009; 150(2): 821 - 831. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. S. Mechaly, J. Vinas, and F. Piferrer Identification of Two Isoforms of the Kisspeptin-1 Receptor (kiss1r) Generated by Alternative Splicing in a Modern Teleost, the Senegalese Sole (Solea senegalensis) Biol Reprod, January 1, 2009; 80(1): 60 - 69. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kanda, Y. Akazome, T. Matsunaga, N. Yamamoto, S. Yamada, H. Tsukamura, K.-i. Maeda, and Y. Oka Identification of KiSS-1 Product Kisspeptin and Steroid-Sensitive Sexually Dimorphic Kisspeptin Neurons in Medaka (Oryzias latipes) Endocrinology, May 1, 2008; 149(5): 2467 - 2476. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
