Inhibins, activins and follistatins

The Molecular Mechanism of Sex Hormones on Sertoli Cell Development and Proliferation

Sustaining and maintaining the intricate process of spermatogenesis is liable upon hormones and growth factors acting through endocrine and paracrine pathways. The Sertoli cells (SCs) are the major somatic cells present in the seminiferous tubules and are considered to be the main regulators of spermatogenesis. As each Sertoli cell supports a specific number of germ cells, thus, the final number of Sertoli cells determines the sperm production capacity. Similarly, sex hormones are also major regulators of spermatogenesis and they can determine the proliferation of Sertoli cells. In the present review, we have critically and comprehensively discussed the role of sex hormones and some other factors that are involved in Sertoli cell proliferation, differentiation and maturation. Furthermore, we have also presented a model of Sertoli cell development based upon the recent advancement in the field of reproduction. Hence, our review article provides a general overview regarding the sex hormonal pathways governing Sertoli cell proliferation and development.

Identification of a second follistatin gene in grass carp (Ctenopharyngodon idellus) and its regulatory function in myogenesis during embryogenesis

Follistatin can antagonize the function of myostatin as a competitive binding protein and promote muscle growth in vivo. Here, we report the isolation and characterization of a second follistatin gene fst2 in grass carp (Ctenopharyngodon idellus). The grass carp fst2 cDNA was 1,376 bp in length, with an open reading frame (ORF) encoding 350 amino acid residues. A relatively low sequence identity of 78% was found between grass carp Fst2 and its paralog Fst1. Sequence and phylogenetic analyses suggest that the grass carp fst2 originated from fish-specific gene duplication. In adult fish, fst2 mRNA expression was observed in most tissues but was strongly expressed in the eyes, muscles, skin and ovary. Grass carp fst2 mRNA could be detected as early as 16 h post-fertilization (hpf), while fst1 mRNA was detected throughout embryogenesis. Using in situ hybridization, fst2 transcripts were detected in the anterior somites at 24 hpf and in the brain and posterior somites at 36 hpf. Meanwhile, fst1 mRNA was transcribed mainly in the optic vesicle and at the cephalic mesoderm at 12 hpf, in the eyes, cephalic mesoderm and at the lateral edge of most somites at 24 hpf, and mainly in the brain at 36 hpf. Furthermore, overexpression of fst2 mRNA markedly affected the formation of the embryonic midline and somite structures. Based on comparisons with fst1, our findings suggest that fst2 retained the ancestral functions of regulating muscle development and growth during embryogenesis in grass carp.

Intra-ovarian roles of activins and inhibins

Granulosa cells are the main ovarian source of inhibins, activins and activin-binding protein (follistatin) while germ (oogonia, oocytes) and somatic (theca, granulosa, luteal) cells express activin receptors, signaling components and inhibin co-receptor (betaglycan). Activins are implicated in various intra-ovarian roles including germ cell survival and primordial follicle assembly; follicle growth from preantral to mid-antral stages; suppression of thecal androgen production; promotion of granulosa cell proliferation, FSHR and CYP19A1 expression; enhancement of oocyte developmental competence; retardation of follicle luteinization and/or atresia and involvement in luteolysis. Inhibins (primarily inhibin A) are produced in greatest amounts by preovulatory follicles (and corpus luteum in primates) and suppress FSH secretion through endocrine negative feedback. Together with follistatin, inhibins act locally to oppose auto-/paracrine activin (and BMP) signaling thus modulating many of the above processes. The balance between activin-inhibin shifts during follicle development with activin signalling prevailing at earlier stages but declining as inhibin and betaglycan expression rise.

Modulation of Androgen Receptor Transactivation by FoxH1

Androgen signaling plays key roles in the development and progression of prostate cancer, and numerous ongoing studies focus on the regulation of androgen receptor (AR) transactivity to develop novel therapies for the treatment of androgen-independent prostate cancer. FoxH1, a member of the Forkhead-box (FOX) gene family of transcription factors, takes part in mediating transforming growth factor-beta/activin signaling through its interaction with the Smad2.Smad4 complex. Using a series of experiments, we found that FoxH1 repressed both ligand-dependent and -independent transactivation of the AR on androgen-induced promoters. This action of FoxH1 was independent of its transactivation capacity and activin A but relieved by Smad2.Smad4. In addition, the repression of the AR by FoxH1 did not require deacetylase activity. A protein-protein interaction was identified between the AR and FoxH1 independently of dihydrotestosterone. Furthermore, a confocal microscopic analysis of LNCaP cells revealed that the interaction between the AR and FoxH1 occurred in the nucleus and that FoxH1 specifically blocked the foci formation of dihydrotestosterone-activated AR, which has been shown to be correlated with the AR transactivation potential. Taken together, our results indicate that FoxH1 functions as a new corepressor of the AR. Our observations not only strengthen the role of FoxH1 in AR-mediated transactivation but also suggest that therapeutic interventions based on AR-coregulator interactions could be designed to block both androgen-dependent and -independent growth of prostate cancer.

Novel action of activin and bone morphogenetic protein in regulating aldosterone production by human adrenocortical cells

We have uncovered a functional bone morphogenetic protein (BMP) and activin system complete with ligands (BMP-6 and activin betaA/betaB), receptors (activin receptor-like kinase receptors 2, 3, and 4; activin type-II receptor; and BMP type-II receptor), and the binding protein follistatin in the human adrenocortical cell line H295R. Administration of activin and BMP-6 to cultures of H295R cells caused concentration-responsive increases in aldosterone production. The mRNA levels of steroidogenic acute regulatory protein or P450 steroid side-chain cleavage enzyme, the rate-limiting steps of adrenocortical steroidogenesis, were enhanced by activin and BMP-6. Activin and BMP-6 also activated the transcription of steroidogenic acute regulatory protein as well as the late-step steriodogenic enzyme CYP11B2. Activin enhanced ACTH-, forskolin-, or dibutyryl-cAMP- but not angiotensin II (Ang II)-induced aldosterone production, whereas BMP-6 specifically augmented Ang II-induced aldosterone production. Activin and ACTH but not BMP-6 increased cAMP production. Follistatin, which inhibits activin actions by binding, suppressed basal and ACTH-induced aldosterone secretion but failed to affect the Ang II-induced aldosterone level. Furthermore, MAPK signaling appeared to be involved in aldosterone production induced by Ang II and BMP-6 because an inhibitor of MAPK activation, U0126, reduced the level of aldosterone synthesis stimulated by Ang II and BMP-6 but not activin. In addition, Ang II reduced the expression levels of BMP-6 but increased that of activin betaB, whereas ACTH had no effect on these levels. Collectively, the present data suggest that activin acts to regulate adrenal aldosterone synthesis predominantly by modulating the ACTH-cAMP-protein kinase A signaling cascade, whereas BMP-6 works primarily by modulating the Ang II-MAPK cascade in human adrenal cortex in an autocrine/paracrine fashion.

Testicular endocrine function, seasonality and semen quality of the stallion

To gain further information on gonadal function of the stallion, concentrations of testicular steroids in blood plasma (bpl) and seminal plasma (spl) and their distribution in the ejaculate were determined. Blood and semen samples from a total of 11 stallions were collected from November to July. Estrone (E1), estrone sulfate (E1S), estradiol-17beta (E2beta) and testosterone (T) were determined in bpl and spl, and in addition androstenedione (A), dehydroepiandrosterone (DHEA) and 5alpha-dihydrotestosterone (5alpha-DHT) were measured in spl. At certain points of time, aliquots of an ejaculate were centrifuged, washed and the distribution of E1, E1S, E2beta and T into seminal plasma and the sperm fraction was assessed. Hormone assay was by RIA, partly after prior separation by HPLC. Mean concentrations (X(g) x DF) were as follows: E2beta (bpl) 31.1 (1.16), (spl) 24.2 (1.42) pg ml(-1); E1 (bpl) 143.3 (1.21), (spl) 117.7 (1.53) pg ml(-1); E1S (bpl) 157.3 (1.44), (spl) 2.92 (1.42) ng ml(-1); T (bpl) 570.6 (1.43), (spl) 23.1 (1.68) pg ml(-1); A (spl) 17.9 (1.39) pg ml(-1); DHEH (spl) 12.4 (1.51) pg ml(-1); 5alpha-DHT (spl) 9.7 (1.29) pg ml(-1). Except for E2beta and A in seminal plasma, a seasonal pattern was established for all other steroids with lowest mean values occurring from November to April. From the semen parameters determined, only motility was correlated to season. There was a higher correlation among oestrogen in blp than in spl and the only correlation identified between oestrogenic and androgenic steroids was between T and E2beta in blp. In spl, T was correlated with A and 5alpha-DHT. T was the dominant free steroid in bpl while it was E1 in spl; T and E1S concentrations were about 23- and 54-fold lower in spl compared to bpl with E1S, however, showing the highest absolute values in both fluids. In the fractionated ejaculate an association of free oestrogens, particularly E2beta, with spermatozoa was observed.

Expression of bone morphogenetic protein-7 mRNA in normal and ischemic adult rat kidney

BMP-7, a member of the bone morphogenic protein subfamily (BMPs) of the transforming growth factor-beta superfamily of secreted growth factors, is abundantly expressed in the fetal kidney. The precise role of this protein in renal physiology or pathology is unknown. A cDNA that encodes rat BMP-7 was cloned and used as a probe to localize BMP-7 mRNA expression by in situ hybridization in the adult rat kidney. The highest expression of BMP-7 mRNA could be seen in tubules of the outer medulla. In glomeruli, a few cells, mainly located at the periphery of the glomerular tuft, showed specific and strong signals. Also, high BMP-7 mRNA expression could be localized to the adventitia of renal arteries, as well as to the epithelial cell layer of the renal pelvis and the ureter. Preliminary evidence suggests that BMP-7 enhances recovery when infused into rats with ischemia-induced acute renal failure. We examined BMP-7 mRNA expression in kidneys with acute renal failure induced by unilateral renal artery clamping. BMP-7 mRNA abundance as analyzed by solution hybridization was reduced in ischemic kidneys after 6 and 16 h of reperfusion compared with the contralateral kidney. In situ hybridization in ischemic kidneys showed a marked decrease of BMP-7 mRNA in the outer medulla and in glomeruli. Utilizing rat metanephric mesenchymal cells in culture, we also demonstrate that BMP-7 induces epithelial cell differentiation. Taken together, these data suggest that BMP-7 is important in both stimulating and maintaining a healthy differentiated epithelial cell phenotype.

Decreased immunoexpression of prostate inhibin peptide in prostatic carcinoma: a study with monoclonal antibody

Prostate inhibin peptide (PIP) is a follicle-stimulating hormone (FSH) regulating peptide produced by the prostate. The mechanism of its endocrine role in regulating prostate growth is believed to be androgen-independent but FSH-dependent. Previous studies using polyclonal antibody proposed PIP as a prostatic-specific marker in cancer diagnosis. However, the recently available monoclonal antibody has not yet been evaluated. Paraffin sections of 72 prostatectomy specimens for prostate cancer with or without hormonal blockage therapy and 10 nonneoplastic prostate tissues from autopsy were stained by using PIP monoclonal antibody (clone: 4A6A6) with the avidin-biotin complex method. PIP reactivity was semiquantitatively estimated in prostatic carcinoma (PCA), prostatic intraepithelial neoplasia (PIN), benign prostatic hyperplasia (BPH) and normal tissue in each case when ever present. Statistical analyses were performed accordingly. PIP expression is predominantly cytoplasmic. Urothelium, seminal vesicles, inflamed prostatic glands, basal cells, and squamous metaplasia were negative for PIP. Average percentage of cells expressing PIP was significantly decreased in PIN (40%) and PCA (14%) when compared with BPH (81%) and normal tissue (68%). There was no correlation of tumor PIP level with patient's age, tumor size, Gleason score, tumor stage, or the usage of preoperative hormonal blockage therapy. PIP monoclonal antibody should be used with caution as a prostate-specific marker in surgical pathology. The mechanism for this alteration and the effect of PIP on prostatic tumor growth, particularly in patients under a variety of hormonal therapies, needs further study.

Collapsin-1 covalently dimerizes, and dimerization is necessary for collapsing activity

Chick collapsin-1, the first identified vertebrate member of the semaphorin family of axon guidance proteins, repels specific growth cones. Like all family members, collapsin-1 contains within its sequence a semaphorin domain that is necessary for specifying activity. Two additional structural domains of collapsin-1, the immunoglobulin (Ig) domain and the basic tail, each potentiate collapsin-1 activity. We identify in this study another structural feature of collapsin-1 that is necessary for its function. Collapsin-1 covalently dimerizes, and dimerization is necessary for collapse activity. This dimerization is mediated through a cysteine at residue 723, between the Ig domain and basic tail. The semaphorin domain alone is not active since it cannot dimerize. The collapsing activity of the semaphorin domain can be reconstituted when made as a chimeric construct with an immunoglobin Fc domain, which promotes dimerization.

Identification of a novel member of the TGF-beta superfamily highly expressed in human placenta

While conducting a gene discovery effort targeted to transcripts of the prevalent and intermediate frequency classes in placenta throughout gestation, we identified a novel member of the TGF-beta superfamily that is expressed at high levels in human placenta. Hence, we named this factor 'Placental Transforming Growth Factor Beta' (PTGFB). The full-length sequence of the 1.2-kb PTGFB mRNA has the potential of encoding a putative pre-pro-PTGFB protein of 295 amino acids and a putative mature PTGFB protein of 112 amino acids. Multiple sequence alignments of PTGFB and representative members of all TGF-beta subfamilies evidenced a number of conserved residues, including the seven cysteines that are almost invariant in all members of the TGF-beta superfamily. The single-copy PTGFB gene was shown to be composed of only two exons of 309 bp and 891 bp, separated by a 2.9-kb intron. The gene was localized to chromosome 19p12-13.1 by fluorescence in-situ hybridization. Northern analyses revealed a complex tissue-specific pattern of expression and a second transcript of 1.9 kb that is predominant in adult skeletal muscle. Most importantly, the 1.2-kb PTGFB transcript was shown to be expressed in placenta at much higher levels than in any other human fetal or adult tissue surveyed.

Opposing effects of activin A and follistatin on developing skeletal muscle cells

Activin and the activin-binding protein follistatin modulate a variety of biological processes and are abundant at sites of muscle development. Activin and follistatin were expressed in developing chick pectoral muscle in vivo and in primary cell culture. Addition of recombinant activin inhibited muscle development in a dose-dependent manner as measured by the number of nuclei in myosin heavy chain positive cells and creatine phosphokinase activity. Conversely, follistatin potentiated muscle development. The effects of activin were found to be distinct from those of the related protein transforming growth factor (TGF) beta1. Muscle development was repressed by activin at all time points investigated and did not recover with the removal of activin following a limited exposure. In contrast, while myogenic differentiation in TGFbeta1 was initially repressed, muscle marker expression recovered to control levels--even in the continued presence of TGFbeta1. Fibroblast growth factor (FGF) had little effect on inhibiton of muscle development caused by activin A. However, inhibition of development produced by TGFbeta increased with increasing concentrations of FGF. Finally, early expression of myoD and myf5 mRNA by muscle cultures in the presence of activin and follistatin was analyzed. Activin-treated cultures expressed reduced myoD and myf5 levels at 1.5 days after plating. Myf5 levels in follistatin-treated cultures were elevated, but, surprisingly, these cultures showed a reduction in myoD levels. These data suggest that endogenously expressed activin and follistatin are important modulators of muscle development.

Activin A and follistatin expression in developing targets of ciliary ganglion neurons suggests a role in regulating neurotransmitter phenotype

The avian ciliary ganglion contains choroid neurons that innervate choroid vasculature and express somatostatin as well as ciliary neurons that innervate iris/ciliary body but do not express somatostatin. We have previously shown in culture that activin A induces somatostatin immunoreactivity in both neuron populations. We now show in vivo that both targets contain activin A; however, choroid expressed higher levels of activin A mRNA. In contrast, follistatin, an activin A inhibitor, was higher in iris/ciliary body. Iris cell-conditioned medium also contained an activity that inhibited activin A and could be depleted with anti-follistatin antibodies. These results suggest that development of somatostatin is limited to choroid neurons by differential expression of activin A and follistatin in ciliary ganglion targets.

Comparative regulation of inhibin, activin and human chorionic gonadotropin production by placental trophoblast cells in culture

In the present study, we investigated the roles of cyclic adenosine monophosphate (cAMP), intracellular calcium, glucocorticoids, protein kinase-C and gonadotrophin-releasing hormone (GnRH) in regulating human chorionic gonadotrophin (hCG), inhibin and activin production in cultured human term placental trophoblast cells. Inhibin and hCG were measured in conditioned media by radioimmunoassay, while putative forms of inhibin and activin were characterized by western blotting using affinity-purified antisera directed against the inhibin alpha- and beta A-subunits. Inhibin and hCG secretion were stimulated by dexamethasone (0.2 microM), GnRH (5-25 microM), calcium ionophore A23187 (0.2-1 microM), phorbol-12-myristate-13-acetate (22 nM) and epinephrine (1 microM), with increasing response over successive 24-h treatment periods. Two molecules Mr approximately 30 and 32 kDa appeared to be the predominant dimeric forms of inhibin secreted by the cells, while 26 kDa activin was present in excess over inhibin. Large amounts of 40-44 kDa protein were detected by the alpha-directed antisera only, which may be a form of the inhibin alpha-subunit precursor protein. Secretion of activin was responsive to phorbol ester-mediated stimulation but not to the presence of GnRH or elevated cAMP concentrations. The divergence in maternal serum inhibin and hCG concentrations during late pregnancy remains unexplained by these findings.

Development and validation of a radioimmunoassay for follistatin in human serum

OBJECTIVE

Follistatin (FS/FSH-suppressing protein/activin-binding protein) is a single-chain glycoprotein, structurally distinct from inhibin, that has been shown to have inhibin-like activity in suppressing FSH secretion both in vivo and in vitro. The aim of these studies was to develop and validate a radioimmunoassay (RIA) for FS in human serum, and to describe the physiological variations of serum FS in humans.

PATIENTS

Serum was collected from normal men, and normal women in the follicular and luteal phases of the menstrual cycle. Clinical samples were also collected from male patients with hypogonadism, post-menopausal women and pregnant women in the first, second and third trimesters.

MEASUREMENTS

A RIA for FS in human serum was developed using antisera raised against purified bovine FS and using bovine FS as tracer and standard. Serial dilutions of serum were non-parallel to purified bovine FS standard in the RIA. The addition of sodium dodecyl sulphate (SDS 0.05%) to the assay buffer resolved the non-parallelism suggesting that the interference of serum in the RIA was an assay matrix effect. To characterize the serum FS immunoactivity further, serum was fractionated by gel filtration on Sephadex G-100. At neutral pH, FS immunoactivity eluted as a major peak in the molecular weight range > 200 kDa. In 0.1 M HCl, a second peak of FS immunoactivity eluted in the molecular weight range 30-60 kDa consistent with the known sizes of FS. This suggested that FS was dissociating from a larger complex. Fractions taken from the low molecular weight region diluted in parallel with bovine FS in contrast with fractions from the high molecular weight region which were non-parallel. It is concluded that the non-parallelism of human serum in the FS RIA is due to the binding of serum FS to an unknown high molecular weight factor. This interference is eliminated by the inclusion of 0.05% SDS in the assay buffer. The RIA in 0.05% SDS has been applied to the description of the normal and pathophysiological variations of FS in human serum.

RESULTS

There were no significant differences between FS levels in serum from hypogonadal men, normal women in the follicular phase of the menstrual cycle, post-menopausal women and pregnant women from the first, second and third trimesters. FS levels in serum of women in the luteal phase of the menstrual cycle were significantly lower than all other groups. FS levels in normal men were significantly higher than those in both luteal phase women and women in the first trimester of pregnancy (P < 0.05).

CONCLUSIONS

These findings argue against a role for circulating follistatin in the control of gonadotrophin secretion and suggest that the gonads and/or conceptus are not the primary source of follistatin immunoactivity in serum. The lower follistatin levels in the luteal phase of the menstrual cycle remain unexplained.

Inhibin as a marker for hydatidiform mole: a comparative study with the determinations of intact human chorionic gonadotrophin and its free beta-subunit

OBJECTIVE

The purpose of the study was to evaluate plasma inhibin as a marker of hydatidiform mole and to compare the results with intact human chorionic gonadotrophin (hCG) and its free beta-subunit.

DESIGN

Serial determinations of the plasma concentrations of inhibin, intact human chorionic gonadotrophin and its free beta-subunit in cases of hydatidiform mole over an average period of 140 days.

PATIENTS

Five cases of hydatidiform mole, including patients with spontaneous remission after evacuation or persistent trophoblastic disease.

MEASUREMENTS

Immunoreactive inhibin, hCG and free hCG beta-subunit were measured using standard enzyme immunoassays.

RESULTS

Inhibin and free hCG beta-subunit levels were greater than in normal pregnant women at the same gestational age. Only intact hCG could detect the persistence of trophoblastic tissue.

CONCLUSIONS

Our data suggest that inhibin, intact human chorionic gonadotrophin and free beta-subunit might be useful as diagnostic markers of molar pregnancies. However, the original method of intact hCG determination is still superior for follow-up.

The TGF-beta superfamily: new members, new receptors, and new genetic tests of function in different organisms

Collapse

The induction of divergent gonadotropin secretion in vitro by variations in luteinizing hormone releasing hormone pulse regimen

The effects of luteinizing hormone releasing hormone (LHRH) pulse amplitude, duration, and frequency on divergent gonadotropin secretion were examined using superfused anterior pituitary cells from selected stages of the rat estrous cycle. Cells were stimulated with one of five LHRH regimens. With low-amplitude LHRH pulses (regimen 1) in the presence of potentially estrogenic phenol red, LH response in pituitary cells from proestrus 1900, estrus 0800, and diestrus 1,0800 were all significantly larger (P less than 0.05) than the other stages tested. In the absence of phenol red, responsiveness at proestrus 1900 was significantly larger than proestrus 0800, proestrus 1500, and estrus 0800 (P less than 0.01, 0.05, and 0.05, respectively); other cycle stages tested were smaller. No significant differences were observed between cycle stages for follicle-stimulating hormone (FSH) secretion in the presence or absence of phenol red. Because pituitary cells at proestrus 1900 were the most responsive to low-amplitude 4 ng LHRH pulses, they were also used to study the effects of LHRH pulses of increased amplitude or duration and decreased frequency. Increasing the amplitude (regimen 2) or the duration (regimens 3 to 5) increased FSH secretion; this effect was greatest with regimens 3 and 5. When regimens 3 and 5 were studied in pituitary cells obtained at proestrus 1500, FSH was significantly increased by both regimes, but most by regimen 5; furthermore, LH release was significantly reduced. When regimens 3 and 5 were studied in pituitary cells obtained at estrus 0800, FSH release was elevated most significantly by regimen 5. Thus, variations in LHRH pulse regimen were found to be capable of inducing significant divergence in FSH release from superfused anterior pituitary cells derived from specific stages of the estrous cycle.

Effect of luteinizing hormone releasing hormone pulse characteristics on comparative luteinizing hormone and follicle stimulating hormone secretion from superfused rat anterior pituitary cell cultures

We have shown that 4 ng luteinizing hormone releasing hormone (LHRH) pulses induced significantly greater luteinizing hormone (LH) release from proestrous rat superfused anterior pituitary cells with no cycle related differences in follicle stimulating hormone (FSH). Current studies gave 8 ng LHRH in various pulse regimens to study amplitude, duration and frequency effects on LH and FSH secretion from estrous 0800, proestrous 1500 and proestrous 1900 cells. Regimen 1 gave 8 ng LHRH as a single bolus once/h; regimen 2 divided the 8 ng into 3 equal 'minipulses' given at 4 min intervals to extend duration; regimen 3 gave the 3 'minipulses' at 10 min intervals, thereby further extending duration: regimen 4 was the same as regimen 2, except that the 3 'minipulses' were given at a pulse frequency of 2 h rather than 1 h. In experiment 1, all four regimens were employed at proestrus 1900. FSH was significantly elevated by all 8 ng regimens as compared to 4 ng pulses; further, 8 ng divided into 3 equal 'minipulses' separated by 4 min at 1 and 3 h frequencies (regimens 2 and 4) resulted in FSH secretion that was significantly greater than with either a single 8 ng bolus (regimen 1) or when the 'minipulses' were separated by 10 min (regimen 3). In experiment 2, at proestrus 1500, FSH response to the second pulse of regimen 4 was significantly greater than in regimen 2; LH release was significantly suppressed at pulse 2 compared to regimen 2 accentuating divergent FSH secretion. At estrus 0800, FSH response to the second pulse of regimen 4 was significantly stimulated FSH at proestrus 1900, 1500 and estrus 0800, FSH divergence was most marked at proestrus 1500. These data indicate a potential role for hypothalamic LHRH secretory pattern in inducing divergent gonadotropin secretion in the rat.